You are on page 1of 146

EQUITY RESEARCH

INITIATING COVERAGE
Biotechnology October 3, 2013
Simos Simeonidis, Ph.D. simos.simeonidis@cowen.com 646.562.1386 Yatin Suneja yatin.suneja@cowen.com 646.562.1388 Raymond Chang, M.D. raymond.chang@cowen.com 646.562.1337 Recommendation
Rating: Price Target (in $): Expected Return: Dividend: Enterprise Value (MM): Market Perform $4.50 (10.0)% NA $994.0 2013E $(0.09)A $0.18A $(0.13) $(0.13) $(0.16) 2014E $(0.08) $(0.08) $(0.08) $(0.04) $(0.28)

ARENA PHARMACEUTICALS (NASDAQ:ARNA)

Initiation: BELVIQ's modest efficacy keeps us on the sidelines


We are initiating coverage of Arena Pharmaceuticals with a Market Perform rating and $4.50 price target. Despite wellstructured partnerships and the potential for an efficacious combination follow-on product, BELVIQ's modest efficacy keeps us on the sidelines on ARNA.
BELVIQ has a role in the new obesity landscape, but its efficacy is modest, and its safety profile is not without its issues.
We expect BELVIQ (lorcaserin) to find a niche in the obesity space, possibly among female and diabetic obese patients, both large patient groups. However, we view its efficacy as marginal, and we expect this to be a major roadblock to this drug's commercial success. We expect BELVIQ will have a difficult time competing head-to-head with Qsymia and Contrave, should it get approved. In addition, BELVIQ's safety is not pristine either, since it comes with warnings about serotonin syndrome and valvular heart disease.

Earnings Per Share


Q1 Q2 Q3 Q4 FY 2012A $(0.18) $(0.12) $(0.07) $(0.10) $(0.45)

Stock Statistics as of 10/02/2013 (in $)


Price: 52W Range: Shares Out (MM): Market Cap (MM): Net Debt (MM): Net Cash Per Share: $5.00 $11.00-$4.78 234.8 $1,089.7 0.0 0.76

Fundamentals
Revenue (MM) ('12A) Revenue (MM) ('13E) Revenue (MM) ('14E) 27.6 81.1 37.5

BELVIQ/Phen combo may be the future for Arena, but if it is, that will likely not be for a number of years.
Arena and Eisai recently announced plans to test the BELVIQ-plus-phentermine combination. We believe this combination has the potential for significant weight loss, similar to that of FenPhen, its predecessor, and theoretically without the disastrous side effects, given lorcaserin's selectivity for the serotonin 2C (vs. the 2B) receptor. However, 1) it could be a number of years until "Bel-Phen" gets to market & 2) the regulatory path, including CVOT requirements, is unclear. Despite that, 3) we have assumed that "Bel-Phen" is approved, have included it in our revenue estimates, and had it account for a significant portion of NPV in ARNA.

Close to 52-week low, ARNA still trades on par or at premium to obesity peers.
ARNA shares, close to a 52-week low, still trade at EV ~$1B, either on par or even at a premium to obesity peers VVUS ($10.19, Outperform) ($1B), OREX ($6.06 Outperform) ($750M). Our sum-of-the-parts NPV analysis points to a fair value of $4.59/share. Therefore, and despite recent decline in ARNA shares, we consider ARNA fairly valued and as even having some potential room for further downside. Please see addendum of this report for important disclosures. MEMBER: FINRA/SIPC www.cowen.com

Cowen and Company, LLC

EQUITY RESEARCH

Company Description
Arenas lead product, BELVIQ (lorcaserin), a selective serotonin (5-HT) 2C receptor agonist, was approved by the FDA for the treatment of obesity (chronic weight management) in June 2012 and was launched in the U.S. in June 2013. BELVIQ has demonstrated modest weight loss with a relatively benign safety profile in three Phase III trials. BELVIQ has been classified by the U.S. Drug Enforcement Administration (DEA) as a Schedule IV drug. Arena has partnered BELVIQ with Eisai in North and South America. Under this agreement, Arena sells BELVIQ to Eisai at transfer prices (i.e. royalty rates) ranging from 31.5%-36.5% of Eisais annual net sales in the U.S. Additionally, Arena is entitled to receive up to $1.2B in milestones and purchase price adjustments (i.e. sales milestones) from Eisai. Through its G-proteincoupled receptor (GPCR)-focused drug discovery platform, Arena has internally developed early stage pipeline candidates, which include: 1) temanogrel, an inverse agonist of the serotonin 2A receptor, which has completed two Phase I trials and is being developed in partnership with South Korean biopharma Ildong for treatment of thrombotic diseases; 2) APD811, an orally available agonist of the prostacyclin (IP) receptor for the treatment of pulmonary arterial hypertension (PAH), which has recently completed Phase I testing in healthy volunteers, and will be evaluated in a Phase II trial in 1Q14; 3) APD334, an S1P1 receptor agonist, which is being developed as a potential treatment for autoimmune diseases, including multiple sclerosis and rheumatoid arthritis, and recently completed Phase I dosing in healthy volunteers; and 4) APD371, a CB2 receptor agonist, currently in preclinical development as a potential treatment for pain. Arena was founded in 1997, is based in San Diego, CA, and currently has approximately 290 employees. Arena: R&D Pipeline
Candidate name BELVIQ BELVIQ + Phentermine Temanogrel APD811 APD334 APD371 Total Drugs in Development San Diego, CA Source: Cowen and Company Indication Obesity Obesity Thrombotic disease Pulmonary arterial hypertension (PAH) Autoimmune diseases Pain P-C I 0 1 II III FILING MKT Comments Launched in the US on June 7, 2013 by Eisai Eisai to initiate a 12-week pilot study YE13/1Q14 Partnered with Ildong Phase II in PAH to be initiated 1Q14 Completed dosing in a Phase I trial in healthy subjects

1 4 0 0 Investor Relations Contact: Cindy McGee - 858.453.7200 x 1479

MEMBER: FINRA/SIPC

www.cowen.com

Cowen and Company, LLC

EQUITY RESEARCH

Arena: Expected Milestones


Milestones BELVIQ Regulatory submission in South Korea Regulatory submission in Brazil Initiate a 12-week pilot study of BELVIQ in combination with Phentermine ( conducted by Eisai ) Phase I PK data for BELVIQ in combination with Phentermine ( conducted by Arena ) Decision on Mexican MAA filing Decision on Canadian MAA filing Other pipeline Ildong to initiate a Phase I trial of temanogrel Initiation of a Phase II trial of APD811 in PAH
Source: Cowen and Company

Timing YE13/1Q14 YE13/1Q14 YE13/1Q14 2013/14 2014 2014 YE13 1Q14

MEMBER: FINRA/SIPC

www.cowen.com

Cowen and Company, LLC

EQUITY RESEARCH

Investment Thesis
We are initiating coverage of Arena Pharmaceuticals (ARNA) with a Market Perform rating and a 12-month price target of $4.50/share. Our thesis on ARNA is that the companys lead asset, BELVIQ (lorcaserin), a novel, twice-a-day agent approved in the U.S. for the treatment of obesity (chronic weight management), which was launched in the U.S. market in June 2013, will have a very difficult time garnering significant market share, mainly due to its modest efficacy. In addition, the compounds safety is not without its issues, since in addition to the carcinogenicity and valvulopathy signals that emerged in its preclinical studies, the BELVIQ label includes warnings about serotonin syndrome (a potentially significant issue, since many overweight and obese patients present with depression and take SSRIs) and valvular heart disease (with the Fen-Phen history still very fresh in physicians minds). Our view of BELVIQs overall clinical profile is based on A) our own analysis of the BELVIQ clinical dataset, in conjunction with the datasets from the Qsymia and Contrave clinical trials, and B) a 100-physician survey that we conducted in order to gain an understanding of prescribers views of these new anti-obesity agents, which we have included at the end of our initiation. Qsymia could make things very difficult for BELVIQ in the next 12 months Our detailed analysis of the clinical datasets for the three obesity compounds, which is included in the back of our report, leads us to believe that Qsymia will end up being the leader in the market, given its strong weight loss efficacy, which comes with once-a-day convenience. Furthermore, and despite the fact that Vivus recently ousted management team and board have faced significant difficulties in the Qsymia launch thus far, most of which have been self-inflicted by, for example, not partnering the drug, as Arena management did appropriately, in our view, we expect that the recent changes at Vivus may serve to relieve some of these issues, for example, if the new team manages to secure a commercial partner for Qsymia. Were such a partnership to materialize, especially one in which Qsymia would be partnered with a big pharma or specialty pharma company with considerable experience in marketing primary care products in the metabolic space, and one in which a substantial level of commercial resources would be committed, this could pose an even more daunting commercial threat to BELVIQ, which until recently was facing a theoretically much easier opponent: Qsymia was approved with a fairly ominous REMS program, was only available via mail-order, and was (still is) being detailed by a 180-person salesforce. Right now, the REMS has been modified, and the drug is available in retail pharmacies; again, should the new regime at Vivus manage to convert its formidable Rolodex into a strong partnership, BELVIQ, with its modest-to-underwhelming weight loss, and the shadows (justified or not) of carcinogenicity and valvulopathy association in some physicians minds (again, fair or unfair), could be in very serious commercial trouble. and after that, things could get even tougher, should Contrave make it to the market. We expect the Qsymia-BELVIQ duopoly to become a three-way fight approximately 12 months from now, with the U.S. approval of Contrave. Orexigens Contrave is a combination of bupropion, an antidepressant with which U.S. PCPs are very familiar and very comfortable prescribing (26M scripts annually), and naltrexone, an agent that is approved for the treatment

MEMBER: FINRA/SIPC

www.cowen.com

Cowen and Company, LLC

EQUITY RESEARCH

of addiction and the main function of which is to reduce cravings. This combination product is currently being tested in the LIGHT cardiovascular outcomes trial (CVOT). Interim data from this trial are expected by early December 2013, and if positive, Orexigen is expecting to resubmit the Contrave NDA by YE13; this drug could thus be on the market in 2H14. We view Contrave as a formidable commercial opponent for both BELVIQ and Qsymia in the obesity/weight loss space, and as one that could quickly garner significant market share and pose a formidable competitive threat to both Qsymia and BELVIQ for the following reasons: 1) We view Contraves mechanism of action, given the combination of the two components that are present in Contrave, as uniquely positioned to be very effective in the treatment of obesity, since A) depression is a comorbidity in a large proportion of the obese population (25-35%, according to a number of sources) and B) helping to deal with cravings and binge-eating are viewed as especially helpful by the obesity experts with whom we have consulted, 2) Contrave would be the only anti-obesity agent on the U.S. market that would have been tested in and received the FDAs OK to get to the market following (interim) data from a CVOT study, 3) Contrave has a North American (plus Mexico) partnership with Takeda, which is a company with a significant diabetes presence in the U.S., given the Actos/alogliptin franchise, and appears to have committed substantial commercial resources behind this product in terms of a primary care salesforce first position calling effort, 4) Contrave would be the only one of the three anti-obesity agents that could be sampled, since both Qsymia and BELVIQ are DEA-scheduled drugs. We dont see the upside in ARNA shareseven though we have included BELVIQ/Phentermine, E.U. BELVIQ revenues, and $1.2B in WW sales. Our Market Perform rating on ARNA is based on our sum-of-the-parts NPV analysis, which includes revenues from sales of both BELVIQ and the BELVIQ/Phentermine combination product in the U.S., and from BELVIQs E.U. sales, where we have assumed Arena will be able to secure a partnership with similar terms as in North America and other territories. We have modeled total peak U.S. sales of BELVIQ and its combination product with Phentermine of $915M in 2029. We have also modeled peak E.U. BELVIQ sales of $246M in 2026, and total WW sales of $1.2B in 2026. In addition, we have assumed revenues from sales of BELVIQ and the BELVIQ/Phentermine combination product in the other two territories for which Arena has secured partnerships, namely South Korea and Taiwan. While one could certainly make the argument that our revenue projections may have left room for upside in BELVIQ sales, especially when compared to our projections for Qsymia and Contrave, both of which we consider more efficacious products and thus more commercially robust for a therapeutic area in which efficacy, in addition to safety, is of paramount importance, it is also important to acknowledge that in our modeling assumption, we have included projections for BELVIQ/Phentermine, a potential product for which we dont know if

MEMBER: FINRA/SIPC

www.cowen.com

Cowen and Company, LLC

EQUITY RESEARCH

and when it will get to the market, and, if and when it does, how it will do commercially, given its inevitable association (fair or unfair, and thats definitely one argument its competitor s will attempt to make) in physicians minds with Fen-Phen. In addition, we have also included in our NPV calculations revenues from E.U. sales of BELVIQ, while again, we dont know if and when BELVIQ will get to the E.U. market. Based on all these modeling assumptions, we have arrived at our 12-month price target of $4.50/share and our Market Perform rating on ARNA.

MEMBER: FINRA/SIPC

www.cowen.com

Cowen and Company, LLC

EQUITY RESEARCH

Arena Partnered With Eisai to Market BELVIQ in North & South America
On July 1, 2010, two months before the first FDA AdCom in September 2010, which voted 9-5 against BELVIQs approval, Arena and Eisai announced a partnership agreement to market BELVIQ in the U.S. The agreement was amended on May 10, 2012, the day of the second FDA AdCom, which voted 18-4 to recommend BELVIQs approval, to include most of North and South America, including Canada, Mexico, and Brazil. Under the agreement, Arena received $50M as an upfront payment, $5M for the amendment, a $20M milestone payment for inclusion in the BELVIQ label of data from the Phase III BLOOM-DM trial in T2D patients, a $65M milestone upon DEA scheduling and launch, and a $1M milestone for regulatory submissions in Mexico and Canada, for a total of $141M in milestone payments to date. Under the terms of the agreement, Arena manufactures BELVIQ at its facility in Switzerland and sells finished product to Eisai. The transfer price (NOTE: this is the term used in the agreement and one that Arena management has been using in its communications with investors, and we believe it is equivalent to what most companies and investors commonly refer as royalty rate) starts at 31.5% of Eisais annual net product sales and will increase on a tiered basis, reaching a maximum rate of 36.5% on the portion of annual net product sales exceeding $750M. Under the agreement, Arena is also entitled to receive up to $53.5M for regulatory filings and approvals. In total, Arena is eligible to receive up to $1.19B in one-time purchase price adjustments (NOTE: again, this is the term used in the agreement and one that Arena management has been using in its communications with investors, and we believe it is equivalent to what most companies and investors commonly refer as sales milestones). These one-time purchase price adjustments would be paid in seven payments, and begin to be triggered at annual net BELVIQ sales of $250M. The caveat here is that, in order for Arena to receive all the milestones for which it is eligible, BELVIQ must achieve annual sales of $2.5B in all the territories covered by the agreement.

MEMBER: FINRA/SIPC

www.cowen.com

Cowen and Company, LLC

EQUITY RESEARCH

Arena and Eisai Partnership Summary


Partner Geographies Partnership date Upfront payment Transfer price Eisai Eisai owns rights to North and South America (including Canada, Mexico and Brazil) Partnered with Eisai in July 2010; amended May 2012 $50M US: 31.5%-36.5% Other territories: 30.75%-35.75% $141M ($50M in upfront; $5M for amending the agreement, and $20M in milestones for including Phase III BLOOM-DM in the label; $65M received on DEA scheduling and delivery of launch supply to Eisai; $1M for regulatory filings in Mexico and Canada ) $53.5M for regulatory filings and approvals $330M in one-time purchase price adjustments ($300M) and milestones ($30M) with annual net sales from $250M-$1B in the US.

Amount received thus far Regulatory milestones remaining

Purchase price adjustments/milestones

$185M in one-time purchase price adjustments for annual net sales in ex-US territories Total potential purchase price adjustment payments Patent life Post-marketing costs up to $1.2B (would have to achieve at least $2.5B in annual sales in all the territorries covered by the agreement) US patents expire mid-2023; (Arena has guided it believes it can receive up to an additional three years of patent extension under Hatch-Waxman, extending US patents until mid2026) Eisai will pay for 90% of CVOT expenses and Arena will pay 10%; Arena is responsible for 50% of certain pediatric development costs

Source: Cowen and Company, SEC Filings

Breakdown of Seven Purchase Price Adjustment Payments

Company disclosure Total Eisai Annual Adjustment Payments to Arena in Net Sales Addition to Transfer Price $250M $X $X $X $X $X $2.5B
Source: Cowen and Company, SEC filings

$25M $X $X $X $X $X $X Total: $1.16B

Arena Partnered with Ildong for South Korean Rights


In November 2012, Arena and Ildong Pharmaceuticals entered into a collaboration agreement to market BELVIQ in South Korea. Under the terms of the agreement, Arena received $5M as an upfront payment and is entitled to receive $3M upon approval. Ildong is responsible for development, regulatory approval, and ultimately, marketing and distribution of BELVIQ in South Korea, including all related costs and expenses. Similarly to the agreement between Arena and Eisai, Arena will sell finished product to Ildong. The purchase price will start at 35% of Ildongs annual net product sales, and will increase on a tiered basis up to 45% on the

MEMBER: FINRA/SIPC

www.cowen.com

Cowen and Company, LLC

EQUITY RESEARCH

portion of annual net sales exceeding $15M. In its 2Q13 earnings call, Arena announced that Ildong would file for regulatory approval in South Korea around YE13.

Arena Partnered with CY Biotech for Taiwanese Rights


In July 2013, Arena announced a marketing and supply agreement with CY Biotech Company (CYB) for BELVIQ in Taiwan. Under the terms of the agreement, Arena received $2M as an upfront payment and is eligible for a milestone payment upon approval of the first additional BELVIQ indication in Taiwan. CYB is responsible for development, regulatory approval, marketing, and distribution of BELVIQ in Taiwan, including all related costs and expenses. Similarly to the agreement between Arena and Eisai, Arena will sell finished product to CYB. The purchase price will be 45% of CYBs annual net product sales, and Arena is eligible for purchase price adjustment payments (which we understand to be equivalent to sales-based milestone payments) based on annual net sales by CYB.

MEMBER: FINRA/SIPC

www.cowen.com

Cowen and Company, LLC

EQUITY RESEARCH

Valuation
To value ARNA shares, we use a sum-of-the-parts methodology, and estimate the probabilityadjusted NPV of: 1) the BELVIQ royalty stream, 2) the four pipeline compounds, and 3) the companys current net cash position.

1) BELVIQ royalties and milestones ($3.62/share)


i) Eisai collaboration: In exchange for North and South American (including Canada, Mexico, and Brazil) rights for BELVIQ, Arena is entitled to receive up to $1.2B in sales milestones and purchase price adjustments (i.e sales milestones) from Eisai. Under the terms of the agreement, Arena will sell finished product to Eisai. The transfer price (i.e. royalty rate) will start at 31.5% of Eisais annual net product sales and will increase on a tiered basis up to 36.5% on the portion of annual net product sales exceeding $750M. In June 2013, Arena received a $65M cash milestone from Eisai for DEA scheduling and the U.S. launch of BELVIQ. In addition, according to the agreement with Eisai, on the year when annual sales in the North and South American regions reach $250M, Arena will receive $55M as a cash milestone from Eisai: a milestone payment of $30M, and the first purchase price adjustment of $25M. In our model, we have assumed this happens in 2018. We have also assumed that similar purchase price adjustments of $55M, $100M, and $150M will materialize in 2020, 2022, and 2026, when BELVIQ sales in these regions reach $500M, $750M, and $1B respectively. Cowen Revenue Model: Eisai Selling Price Assumptions
T ie rs 1 2 3 < $ 2 5 0 M : 3 1 .5 % $250-$750M : 34% > $ 7 5 0 : 3 6 .5 %

Source: Cowen and Company

ii) E.U. collaboration: We have assumed that Arena will enter into a collaboration agreement with a pharmaceutical company for E.U. rights to BELVIQ. We have also assumed that Arena will be able to secure similar economics for E.U. rights as for North/South American rights with Eisai. We have assumed that Arena will receive an upfront payment of $100M in 2015 and $50M in 2016 for commercial launch milestones. Arena will also receive $50M as a first commercial milestone in 2018, $50M as a second commercial milestone in 2020, $50M as a third commercial milestone in 2022, and a final commercial milestone of $75M in 2024. Arena will sell finished product to its E.U. partner at a purchase price of 36% of annual net product sales. iii) Ildong collaboration: In exchange for South Korean rights to BELVIQ, Arena received $5M as an upfront payment and is entitled to receive $3M on approval. Arena will sell finished product to Eisai at a purchase price that will start at 35% of Ildongs annual net product sales

MEMBER: FINRA/SIPC

10

www.cowen.com

Cowen and Company, LLC

EQUITY RESEARCH

and will increase on a tiered basis up to 45% of the portion of annual net sales exceeding $15M. The table below lists the selling price assumptions which we have used in our model for South Korean sales. Cowen Revenue Model: Ildong Selling Price Assumptions
T ie rs 1 2 3 <$5M : 35% $5-$15M : 40% >$15M : 45%

Source: Cowen and Company, SEC filings

iv) CY Biotech collaboration: In exchange for Taiwanese rights to BELVIQ, Arena received $2M as an upfront payment and is eligible for a milestone payment upon approval of an additional BELVIQ indication in Taiwan. Similar to the agreement between Arena and Eisai/Ildong, Arena will sell finished product to CYB. The purchase price will be 45% of CYBs annual net product sales, and Arena is eligible for purchase price adjustment payments (which we understand to be equivalent to sales-based milestone payments) based on annual net sales by CYB. Patent life assumptions: BELVIQ is covered by issued U.S. and E.U. patents that expire in mid-2023. Arena has already filed for extension under Hatch-Waxman. In our NPV calculations, we have assumed that Arena will receive a 3-year patent extension under HatchWaxman, resulting in U.S. patent expiry in mid-2026. Further, we have assumed that the BELVIQ/Phentermine combination will be introduced to the U.S. in 2018, which will extend the patent life beyond 2026, because of the longer patent life of BELVIQ/Phentermine. In the E.U., BELVIQ will be eligible to receive 10-year exclusivity, which will end in mid-2026, assuming a 2016 launch. U.S./E.U./South Korea/Taiwan BELVIQ sales: We have modeled that BELVIQ could reach peak sales of $915M, $246M, $50M, and $24M in the U.S., E.U., South Korea, and Taiwan, respectively, for total peak sales of $1.2B in 2026. Discount Rate and Probability of Success (POS): In calculating the net present value of BELVIQs free cash flows, we use a 10% discount rate. We have also probability -adjusted the E.U., South Korean (SK), and Taiwan royalties to Arena by assigning a 50%, 80%, and 80% probability of success (POS) that the compound is approved and reaches the market in the E.U., South Korea, and Taiwan, respectively. Using these assumptions, as shown in the table below, we arrive at a probability-adjusted NPV for BELVIQ of $3.62/share.

MEMBER: FINRA/SIPC

11

www.cowen.com

Cowen and Company, LLC

EQUITY RESEARCH

BELVIQ NPV analysis US


($MM) Total US Sales Total revenue on US sales to ARNA Total US revenues to ARNA COGS SG&A R&D expenses Milestone payments - Eisai Tax adjusted EBIT
Tax rate

2013E
18 6 6 2 33 17 65 18
0%

2014E
54 17 17 5 35 20 10 (33)
0%

2015E
88 28 28 5 36 20 10 (23)
0%

2016E
122 39 39 7 36 20 10 (15)
0%

2017E
158 50 50 10 37 20 0 (17)
0%

2018E
290 92 92 14 38 20 55 71
5%

2019E
372 120 120 19 19 0 0 76
8%

2020E
462 151 151 23 20 0 55 139
15%

2021E
634 209 209 32 20 0 0 134
15%

2022E
806 269 269 40 21 0 100 247
20%

2023E
753 250 250 38 21 0 0 153
20%

2024E
797 266 266 40 21 0 0 153
25%

2025E
844 283 283 42 22 0 0 164
25%

2026E
894 301 301 45 22 0 150 288
25%

2027E
901 303 303 45 23 0 0 177
25%

2028E
907 305 305 45 23 0 0 178
25%

2029E
915 308 308 46 24 0 0 179
25%

BELVIQ free cash flow


% y/y growth

18

(33)
-286%

(23)
-29%

(15)
-35%

(17)
10%

71
-525%

76
6%

139
83%

134
-3%

247
84%

153
-38%

153
0%

164
7%

288
75%

177
-39%

178
1%

179
1%

Discount Period Discount Factor

0.24 0.98

1.24 0.89

2.24 0.81

3.24 0.73

4.24 0.67

5.24 0.61

6.24 0.55

7.24 0.50

8.24 0.46

9.24 0.41

10.24 0.38

11.24 0.34

12.24 0.31

13.24 0.28

14.24 0.26

15.24 0.23

16.24 0.21

PV of BELVIQ Free Cash Flow

$17

($29)

($19)

($11)

($11)

$43

$42

$69

$61

$102

$58

$53

$51

$81

$45

$42

$38

Discount Rate Perpetual Growth Rate Final year FCF Terminal Value
Discount Factor

10% 0% $0 $0 $0 $615 $615 235 $2.62 100% $2.62

Present Value of Terminal Value Present Value of Cash Flows Present Value of Total Cash Flows Fully Diluted Shares Outstanding Present Value of Cash Flows Per Share Probability of success BELVIQ NPV

Source: Cowen and Company

MEMBER: FINRA/SIPC

12

www.cowen.com

Cowen and Company, LLC

EQUITY RESEARCH

BELVIQ NPV analysis E.U.


($MM) Total EU Sales Royalties on EU sales Total EU revenues to ARNA COGS Milestone payments - EU Tax adjusted EBIT
Tax rate

2013E
-

2014E
0

2015E
0 100 100
0%

2016E
16 6 6 1 50 55
0%

2017E
35 13 13 2 0 10
0%

2018E
55 20 20 3 50 64
5%

2019E
78 28 28 4 0 22
8%

2020E
101 36 36 5 50 69
15%

2021E
128 46 46 6 0 34
15%

2022E
157 56 56 8 50 79
20%

2023E
176 63 63 9 0 44
20%

2024E
198 71 71 10 75 102
25%

2025E
221 79 79 11 0 51
25%

2026E
246 88 88 12 0 57
25%

2027E
25 9 9 1 0 6
25%

2028E
12 4 4 1 0 3
25%

2029E
12 4 4 1 0 3
25%

0%

0%

BELVIQ free cash flow


% y/y growth

100

55
-45%

10
-81%

64
509%

22
-65%

69
211%

34
-51%

79
134%

44
-45%

102
134%

51
-50%

57
11%

6
-90%

3
-50%

3
0%

Discount Period Discount Factor

0.24 0.98

1.24 0.89

2.24 0.81

3.24 0.73

4.24 0.67

5.24 0.61

6.24 0.55

7.24 0.50

8.24 0.46

9.24 0.41

10.24 0.38

11.24 0.34

12.24 0.31

13.24 0.28

14.24 0.26

15.24 0.23

16.24 0.21

PV of BELVIQ Free Cash Flow

$0

$0

$81

$40

$7

$39

$12

$35

$15

$33

$16

$35

$16

$16

$1

$1

$1

Discount Rate Perpetual Growth Rate Final year FCF Terminal Value
Discount Factor

10% 0% $0 $0 $0 $348 $348 235 $1.48 50% $0.74

Present Value of Terminal Value Present Value of Cash Flows Present Value of Total Cash Flows Fully Diluted Shares Outstanding Present Value of Cash Flows Per Share Probability of success BELVIQ NPV - EU

Source: Cowen and Company

MEMBER: FINRA/SIPC

13

www.cowen.com

Cowen and Company, LLC

EQUITY RESEARCH

BELVIQ NPV analysis South Korea/Taiwan


($MM) Total South Korean Sales Total Taiwan Sales Total South Korea/Taiwan Sales Total revenue on South Korean sales to ARNA Total revenue on Taiwan sales to ARNA Total SK/Taiwan revenues to ARNA COGS Milestone payments received Tax adjusted EBIT
Tax rate

2013E
-

2014E
-

2015E
2.1 1.0 3.1 0.7 0.5 1 0 3 4
0%

2016E
4.6 2.2 6.7 1.6 1.0 3 0 0 2
0%

2017E
6.6 3.2 9.8 2.4 1.4 4 1 0 3
0%

2018E
8.8 4.2 13.1 3.3 1.9 5 1 0 4
5%

2019E
19.9 9.6 29.5 8.0 4.3 12 1 0 10
8%

2020E
24.5 11.7 36.2 10.0 5.3 15 2 0 11
15%

2021E
33.1 15.9 49.0 13.9 7.2 21 2 0 16
15%

2022E
34.8 16.7 51.5 14.7 7.5 22 3 0 16
20%

2023E
40.7 19.6 60.3 17.3 8.8 26 3 0 18
20%

2024E
42.8 20.6 63.4 18.3 9.3 28 3 0 18
25%

2025E
45.0 21.7 66.8 19.3 9.8 29 3 0 19
25%

2026E
47.4 22.9 70.3 20.3 10.3 31 4 0 20
25%

2027E
49.9 24.1 73.9 21.4 10.8 32 4 0 21
25%

2028E
50.0 24.1 74.1 21.5 10.9 32 4 0 21
25%

2029E
50.1 24.2 74.3 21.5 10.9 32 4 0 22
25%

BELVIQ free cash flow


% y/y growth

2
-46%

3
48%

4
34%

10
130%

11
15%

16
38%

16
-1%

18
18%

18
-1%

19
5%

20
5%

21
5%

21
0%

22
0%

Discount Period Discount Factor

0.24 0.98

1.24 0.89

2.24 0.81

3.24 0.73

4.24 0.67

5.24 0.61

6.24 0.55

7.24 0.50

8.24 0.46

9.24 0.41

10.24 0.38

11.24 0.34

12.24 0.31

13.24 0.28

14.24 0.26

15.24 0.23

16.24 0.21

PV of BELVIQ Free Cash Flow

$0

$0

$3

$2

$2

$3

$5

$6

$7

$7

$7

$6

$6

$6

$6

$5

$5

Discount Rate Perpetual Growth Rate Final year FCF Terminal Value
Discount Factor

10% 0% $0 $0 $0 $75 $75 235 $0.32 80% $0.25

Present Value of Terminal Value Present Value of Cash Flows Present Value of Total Cash Flows Fully Diluted Shares Outstanding Present Value of Cash Flows Per Share Probability of success BELVIQ NPV

Source: Cowen and Company

2) Arenas pipeline ($0.21/share)


We believe that the market assigns minimal value to the companys four pipeline compounds, given their early stage of development and lack of clinical data: 1) temanogrel, an inverse agonist of the serotonin 2A receptor, which has completed two Phase I trials and is being developed in partnership with South Korean biopharma Ildong for treatment of thrombotic diseases, 2) APD811, being developed for PAH, which has recently completed Phase I testing in healthy volunteers and will be evaluated in a Phase II trial in 1Q14, 3) APD334, being developed for autoimmune disease and currently in a Phase I trial in healthy volunteers, and 4) APD371, which is currently in preclinical development. We believe that the value of this pipeline: A) is very difficult to accurately quantify, and B) could increase or decrease significantly, as more data become available. For the purposes of valuing these early stage assets at this point, we have decided to assign them a total value of $50M.

3) Arenas current cash position ($0.76/share)


Arena ended 2Q13 with $178.9M or $0.76/fully diluted share in cash.

MEMBER: FINRA/SIPC

14

www.cowen.com

Cowen and Company, LLC

EQUITY RESEARCH

ARNA Sum-of-the-Parts Analysis

BELVIQ NPV - US BELVIQ NPV - EU BELVIQ NPV - SK/Taiwan Pipeline Cash Sum-of-the-parts value for ARNA
Source: Cowen and Company

$2.62 $0.74 $0.25 $0.21 $0.76 $4.59

BELVIQ Revenue Model


Obesity is a chronic condition that affects approximately one-third of the United States population, and its prevalence has doubled over the past two decades. The CDC (Health, United States 2012, National Center for Health Statistics 2013) estimates that 35.7% of US adults are obese (35.5% of men and 35.8% of women). This represents 71.3M adults between the ages of 18 and 64 (35.4M men and 35.9M women) who can technically be classified as obese based on their BMI. We have used this group as the starting point in our estimation of the addressable market for BELVIQ in the U.S. The prevalence of obesity is slightly lower in the E.U., and according to a number of literature sources, male and female obesity rates vary from 4% to 28% and 6% to 36%, respectively. In our revenue model, we have assumed that 20% of adults are obese in the E.U. This represents 61.9M adults between 18 and 64 years of age (30.7M men and 31.3M women). Using similar assumptions, we estimate that there are approximately 6.2M obese adults in South Korea (3.1M men and 3.1M women) and 2.9M obese adults in Taiwan (1.5M men and 1.4M women). We estimate that this is the patient population that would be eligible for treatment with BELVIQ in the E.U., South Korea, and Taiwan. Pricing, penetration rates and sales: Eisai has priced BELVIQ at an average monthly cost of $199.50, and we have estimated that BELVIQ will be launched at a 25% discount to U.S. pricing in the E.U., and at a 50% discount to U.S. pricing in South Korea and Taiwan. We estimate that BELVIQ will be launched in 2015 in South Korea and Taiwan, and in mid-2016 in the E.U. We have also assumed that the BELVIQ/Phentermine combination will be introduced to the U.S. in 2018, and to South Korea and Taiwan in 2019. Peak penetration in men between 18 and 64 years of age: We estimate that WW BELVIQ + BELVIQ/Phentermine sales in men between 18 and 64 years of age could be ~$70M in 2018, and that at 0.51%, 0.18%, 0.59%, and 0.59% penetrations in the U.S., E.U., South Korea, and Taiwan, respectively, BELVIQ could reach peak WW sales of $240M in 2026. Peak penetration in women between 18 and 64 years of age: We have assumed that penetration in women will be four times that of men, and we estimate that WW

MEMBER: FINRA/SIPC

15

www.cowen.com

Cowen and Company, LLC

EQUITY RESEARCH

BELVIQ + BELVIQ/Phentermine sales in women between 18 and 64 years of age could be ~$290M in 2018, and that at 2.04%, 0.72%, 2.36%, and 2.36% penetrations in the U.S., E.U., South Korea, and Taiwan, respectively, BELVIQ could reach peak WW sales of $971M in 2026. BELVIQ + BELVIQ/Phentermine revenue model ($MM) - US
BELVIQ+BELVIQ/Phentermine Revenue Model (US)
US population
Population growth

2012E 318,769,185
0.88%

2013E 321,574,353
0.88%

2014E 324,404,208
0.88%

2015E 327,258,965
0.88%

2016E 330,138,844
0.88%

2017E 333,044,066
0.88%

2018E 335,974,853
0.88%

2019E 338,931,432
0.88%

2020E 341,914,029
0.88%

2021E 344,922,872
0.88%

2022E 347,958,193
0.88%

2023E 351,020,225
0.88%

2024E 354,109,203
0.88%

2025E 357,225,364
0.88%

2026E 360,368,948
0.88%

2027E 363,540,194
0.88%

2028E 366,739,348
0.88%

2029E 369,966,654
0.88%

Total men # of men between 18 and 64 years old


% of men between 18 and 64 years old % of men between 18 and 64 years old that are obese

155,453,282 98,739,136
63.5% 35.5%

156,821,271 99,608,040
63.5% 35.5%

158,201,298 100,484,591
63.5% 35.5%

159,593,469 101,368,855
63.5% 35.5%

160,997,892 102,260,901
63.5% 35.5%

162,414,673 103,160,797
63.5% 35.5%

163,843,923 104,068,612
63.5% 35.5%

165,285,749 104,984,416
63.5% 35.5%

166,740,264 105,908,279
63.5% 35.5%

168,207,578 106,840,272
63.5% 35.5%

169,687,805 107,780,466
63.5% 35.5%

171,181,057 108,728,934
63.5% 35.5%

172,687,451 109,685,749
63.5% 35.5%

174,207,100 110,650,984
63.5% 35.5%

175,740,123 111,624,712
63.5% 35.5%

177,286,636 112,607,010
63.5% 35.5%

178,846,758 113,597,951
63.5% 35.5%

180,420,610 114,597,613
63.5% 35.5%

# of men between 18 and 64 years old that are obese


BELVIQ penetration in men

35,052,393

35,360,854
0.025%

35,672,030
0.05%

35,985,944
0.07%

36,302,620
0.09%

36,622,083
0.11%

36,944,357
0.04%

37,269,468
0.03%

37,597,439
0.02%

37,928,297
0.02%

38,262,066
0.02%

38,598,772
0.02%

38,938,441
0.02%

39,281,099
0.02%

39,626,773
0.02%

39,975,488
0.00%

40,327,273
0.00%

40,682,153
0.00%

# of obese men treated with BELVIQ


BELVIQ/Phentermine penetration in men

8,840

17,836

25,190

32,672

40,284

14,778
0.15%

11,181
0.20%

7,519
0.25%

7,586
0.33%

7,652
0.40%

7,720
0.35%

7,788
0.35%

7,856
0.35%

7,925
0.35%

800
0.35%

403
0.35%

407
0.35%

# of obese men treated with BELVIQ/Phentermine Average # of prescriptions (Rx)/patient % compliance Average cost/prescription (Rx)
% price increase

55,417 4.5 70% $200


2%

74,539 6 70% $255


5%

93,994 6 70% $268


5%

125,163 6 70% $281


5%

153,048 6 70% $295


5%

135,096 6 70% $310


5%

136,285 6 70% $325


5%

137,484 6 70% $341


5%

138,694 6 70% $358


5%

139,914 6 70% $376


0%

141,145 6 70% $376


0%

142,388 6 70% $376


0%

6 70% $203
3%

6 70% $210
5%

6 70% $220
5%

6 70% $231
5%

6 70% $243
5%

Discount offered BELVIQ - total US Sales in men ($MM) BELVIQ/Phentermine - total US Sales in men ($MM) Total women # of women between 18 and 64 years old
% of women between 18 and 64 years old % of women between 18 and 64 years old that are obese

35% $0 $0 160,262,826 99,491,200


62.1% 35.8%

30% $11 $0 163,095,862 101,249,949


62.1% 35.8%

22% $17 $0 164,531,106 102,140,949


62.1% 35.8%

20% $24 $0 165,978,979 103,039,789


62.1% 35.8%

20% $31 $0 167,439,594 103,946,539


62.1% 35.8%

20% $12 $45 168,913,063 104,861,269


62.1% 35.8%

20% $10 $64 170,399,498 105,784,048


62.1% 35.8%

20% $7 $84 171,899,013 106,714,948


62.1% 35.8%

20% $7 $118 173,411,725 107,654,039


62.1% 35.8%

20% $8 $152 174,937,748 108,601,395


62.1% 35.8%

20% $8 $141 176,477,200 109,557,087


62.1% 35.8%

20% $9 $149 178,030,199 110,521,189


62.1% 35.8%

20% $9 $158 179,596,865 111,493,776


62.1% 35.8%

20% $10 $167 181,177,318 112,474,921


62.1% 35.8%

20% $1 $177 182,771,678 113,464,700


62.1% 35.8%

20% $1 $179 184,380,069 114,463,190


62.1% 35.8%

20% $1 $180 186,002,613 115,470,466


62.1% 35.8%

$4 $0 161,673,138 100,366,722
62.1% 35.8%

# of women between 18 and 64 years old that are obese


BELVIQ penetration in women

35,617,849

35,931,287
0.1%

36,247,482
0.20%

36,566,460
0.28%

36,888,245
0.36%

37,212,861
0.44%

37,540,334
0.16%

37,870,689
0.12%

38,203,951
0.08%

38,540,146
0.08%

38,879,299
0.08%

39,221,437
0.08%

39,566,586
0.08%

39,914,772
0.08%

40,266,022
0.08%

40,620,363
0.01%

40,977,822
0.00%

41,338,427
0.00%

# of obese women treated with BELVIQ


BELVIQ/Phentermine penetration in women

35,931

72,495

102,386

132,798

163,737

60,065
0.60%

45,445
0.80%

30,563
1.00%

30,832
1.32%

31,103
1.60%

31,377
1.40%

31,653
1.40%

31,932
1.40%

32,213
1.40%

3,250
1.40%

1,639
1.40%

1,654
1.40%

# of obese women treated with BELVIQ/Phentermine Average # of prescriptions (Rx)/patient % compliance Average cost/prescription (Rx)
% price increase

225,242 4.5 70% $200


2%

302,966 6 70% $255


5%

382,040 6 70% $268


5%

508,730 6 70% $281


5%

622,069 6 70% $295


5%

549,100 6 70% $310


5%

553,932 6 70% $325


5%

558,807 6 70% $341


5%

563,724 6 70% $358


5%

568,685 6 70% $376


0%

573,690 6 70% $376


0%

578,738 6 70% $376


0%

6 70% $203
3%

6 70% $210
5%

6 70% $220
5%

6 70% $231
5%

6 70% $243
5%

Discount offered BELVIQ - total US Sales in women ($MM) BELVIQ/Phentermine - total US Sales in women ($MM) BELVIQ - total US Sales ($MM) BELVIQ/Phentermine - total US Sales ($MM) BELVIQ + BELVIQ/Phentermine- total US Sales ($MM) Total revenue on US sales to ARNA $0 $0 $0 $0 $0 $0

35% $15 $0 $18 $0 $18 $6

30% $43 $0 $54 $0 $54 $17

22% $70 $0 $88 $0 $88 $28

20% $98 $0 $122 $0 $122 $39

20% $127 $0 $158 $0 $158 $50

20% $49 $184 $61 $229 $290 $92

20% $39 $259 $48 $323 $372 $120

20% $27 $343 $34 $428 $462 $151

20% $29 $480 $36 $598 $634 $209

20% $31 $616 $38 $768 $806 $269

20% $33 $571 $41 $712 $753 $250

20% $35 $605 $43 $754 $797 $266

20% $37 $641 $46 $799 $844 $283

20% $39 $679 $48 $846 $894 $301

20% $4 $719 $5 $896 $901 $303

20% $2 $726 $3 $904 $907 $305

20% $2 $732 $3 $912 $915 $308

Source: Cowen and Company

MEMBER: FINRA/SIPC

16

www.cowen.com

Cowen and Company, LLC

EQUITY RESEARCH

BELVIQ revenue model ($MM) EU


BELVIQ Revenue Model (EU)
EU population
Population growth

2012E 504,832,526
0.10%

2013E 505,337,358
0.10%

2014E 505,842,695
0.10%

2015E 506,348,538
0.10%

2016E 506,854,887
0.10%

2017E 507,361,742
0.10%

2018E 507,869,103
0.10%

2019E 508,376,972
0.10%

2020E 508,885,349
0.10%

2021E 509,394,235
0.10%

2022E 509,903,629
0.10%

2023E 510,413,533
0.10%

2024E 510,923,946
0.10%

2025E 511,434,870
0.10%

2026E 511,946,305
0.10%

2027E 512,458,251
0.10%

2028E 512,970,710
0.10%

2029E 513,483,680
0.10%

Total men
% of men between 18 and 64 years old

241,263,217
63.5%

241,504,480
63.5%

241,745,984
63.5%

241,987,730
63.5%

242,229,718
63.5%

242,471,948
63.5%

242,714,420
63.5%

242,957,134
63.5%

243,200,091
63.5%

243,443,291
63.5%

243,686,735
63.5%

243,930,421
63.5%

244,174,352
63.5%

244,418,526
63.5%

244,662,945
63.5%

244,907,608
63.5%

245,152,515
63.5%

245,397,668
63.5%

# of men between 18 and 64 years old


% of men between 18 and 64 years old that are obese

153,242,963
20.0%

153,396,206
20.0%

153,549,602
20.0%

153,703,152
20.0%

153,856,855
20.0%

154,010,712
20.0%

154,164,722
20.0%

154,318,887
20.0%

154,473,206
20.0%

154,627,679
20.0%

154,782,307
20.0%

154,937,089
20.0%

155,092,026
20.0%

155,247,118
20.0%

155,402,366
20.0%

155,557,768
20.0%

155,713,326
20.0%

155,869,039
20.0%

# of men between 18 and 64 years old that are obese


BELVIQ penetration in men

30,648,593

30,679,241 -

30,709,920 -

30,740,630 -

30,771,371
0.02%

30,802,142
0.04%

30,832,944
0.06%

30,863,777
0.08%

30,894,641
0.10%

30,925,536
0.12%

30,956,461
0.14%

30,987,418
0.15%

31,018,405
0.16%

31,049,424
0.17%

31,080,473
0.18%

31,111,554
0.02%

31,142,665
0.01%

31,173,808
0.01%

# of obese men treated with BELVIQ Average # of prescriptions (Rx)/patient % compliance Average cost/prescription (Rx)
% price increase

6,154 6 70% $158


5%

12,321 6 70% $166


5%

18,500 6 70% $174


5%

25,000 6 70% $183


5%

30,895 6 70% $192


5%

37,111 6 70% $202


5%

43,339 6 70% $212


5%

46,481 6 70% $222


5%

49,629 6 70% $233


5%

52,784 6 70% $245


5%

55,945 6 70% $257


0%

5,600 6 70% $257


0%

2,803 6 70% $257


0%

2,806 6 70% $257


0%

Discount offered BELVIQ - total EU Sales in men ($MM) Total women


% of women between 18 and 64 years old

25% $0 251,591,877
62.1%

20% $7 252,852,355
62.1%

20% $11 253,105,207


62.1%

20% $15 253,358,312


62.1%

20% $20 253,611,671


62.1%

20% $25 253,865,282


62.1%

20% $31 254,119,148


62.1%

20% $35 254,373,267


62.1%

20% $39 254,627,640


62.1%

20% $43 254,882,268


62.1%

20% $48 255,137,150


62.1%

20% $5 255,392,287
62.1%

20% $2 255,647,679
62.1%

20% $2 255,903,327
62.1%

$0 251,843,469
62.1%

$0 252,095,312
62.1%

$0 252,347,408
62.1%

$3 252,599,755
62.1%

# of women between 18 and 64 years old


% of women between 18 and 64 years old that are obese

156,188,296
20.0%

156,344,484
20.0%

156,500,829
20.0%

156,657,330
20.0%

156,813,987
20.0%

156,970,801
20.0%

157,127,772
20.0%

157,284,900
20.0%

157,442,184
20.0%

157,599,627
20.0%

157,757,226
20.0%

157,914,983
20.0%

158,072,898
20.0%

158,230,971
20.0%

158,389,202
20.0%

158,547,592
20.0%

158,706,139
20.0%

158,864,845
20.0%

# of women between 18 and 64 years old that are obese


BELVIQ penetration in women

31,237,659

31,268,897

31,300,166

31,331,466

31,362,797
0.08%

31,394,160
0.16%

31,425,554
0.24%

31,456,980
0.32%

31,488,437
0.40%

31,519,925
0.48%

31,551,445
0.56%

31,582,997
0.60%

31,614,580
0.64%

31,646,194
0.68%

31,677,840
0.72%

31,709,518
0.07%

31,741,228
0.04%

31,772,969
0.04%

# of obese women treated with BELVIQ Average # of prescriptions (Rx)/patient % compliance Average cost/prescription (Rx)
% price increase

25,090 6 70% $158


5%

50,231 6 70% $166


5%

75,421 6 70% $174


5%

101,921 6 70% $183


5%

125,954 6 70% $192


5%

151,296 6 70% $202


5%

176,688 6 70% $212


5%

189,498 6 70% $222


5%

202,333 6 70% $233


5%

215,194 6 70% $245


5%

228,080 6 70% $257


0%

22,831 6 70% $257


0%

11,427 6 70% $257


0%

11,438 6 70% $257


0%

Discount offered BELVIQ - total EU Sales in women ($MM) BELVIQ - total EU Sales ($MM) Total revenue on EU sales to ARNA $0 $0 $0 $0 $0 $0 $0 $0 $0 $0 $0 $0

25% $12 $16 $6

20% $28 $35 $13

20% $44 $55 $20

20% $63 $78 $28

20% $81 $101 $36

20% $103 $128 $46

20% $126 $157 $56

20% $142 $176 $63

20% $159 $198 $71

20% $177 $221 $79

20% $197 $246 $88

20% $20 $25 $9

20% $10 $12 $4

20% $10 $12 $4

Source: Cowen and Company

MEMBER: FINRA/SIPC

17

www.cowen.com

Cowen and Company, LLC

EQUITY RESEARCH

BELVIQ + BELVIQ/Phentermine revenue model ($MM) South Korea


BELVIQ +BELVIQ/Phentermine Revenue Model (SK)
South Korea population
Population growth

2012E 49,060,054
0.20%

2013E 49,160,137
0.20%

2014E 49,260,423
0.20%

2015E 49,360,915
0.20%

2016E 49,461,611
0.20%

2017E 49,562,513
0.20%

2018E 49,663,620
0.20%

2019E 49,764,934
0.20%

2020E 49,866,454
0.20%

2021E 49,968,182
0.20%

2022E 50,070,117
0.20%

2023E 50,172,260
0.20%

2024E 50,274,611
0.20%

2025E 50,377,172
0.20%

2026E 50,479,941
0.20%

2027E 50,582,920
0.20%

2028E 50,686,109
0.20%

2029E 50,789,509
0.20%

Total men
% of men between 18 and 64 years old

24,538,235
63.5%

24,588,293
63.5%

24,638,454
63.5%

24,688,716
63.5%

24,739,081
63.5%

24,789,549
63.5%

24,840,119
63.5%

24,890,793
63.5%

24,941,570
63.5%

24,992,451
63.5%

25,043,436
63.5%

25,094,525
63.5%

25,145,717
63.5%

25,197,015
63.5%

25,248,417
63.5%

25,299,923
63.5%

25,351,535
63.5%

25,403,252
63.5%

# of men between 18 and 64 years old


% of men between 18 and 64 years old that are obese

15,585,931
20.0%

15,617,727
20.0%

15,649,587
20.0%

15,681,512
20.0%

15,713,502
20.0%

15,745,558
20.0%

15,777,679
20.0%

15,809,865
20.0%

15,842,117
20.0%

15,874,435
20.0%

15,906,819
20.0%

15,939,269
20.0%

15,971,785
20.0%

16,004,367
20.0%

16,037,016
20.0%

16,069,732
20.0%

16,102,514
20.0%

16,135,363
20.0%

# of men between 18 and 64 years old that are obese


BELVIQ penetration in men

3,117,186

3,123,545

3,129,917

3,136,302
0.04%

3,142,700
0.08%

3,149,112
0.11%

3,155,536
0.14%

3,161,973
0.10%

3,168,423
0.05%

3,174,887
0.05%

3,181,364
0.05%

3,187,854
0.05%

3,194,357
0.05%

3,200,873
0.05%

3,207,403
0.05%

3,213,946
0.05%

3,220,503
0.05%

3,227,073
0.05%

# of obese men treated with BELVIQ


BELVIQ/Phentermine penetration in men

1,255 6 70% $105


5%

2,514 6 70% $110


5%

3,464 6 70% $116


5%

4,418 6 70% $121


5%

3,162
0.20%

1,584
0.30%

1,587
0.40%

1,591
0.40%

1,594
0.45%

1,597
0.45%

1,600
0.45%

1,604
0.45%

1,607
0.45%

1,610
0.45%

1,614
0.45%

# of obese men treated with BELVIQ/Phentermine Average # of prescriptions (Rx)/patient % compliance Average cost/prescription (Rx)
% price increase

6,324 6 70% $127


5%

9,505 6 70% $134


5%

12,700 6 70% $140


5%

12,725 6 70% $147


5%

14,345 6 70% $155


5%

14,375 6 70% $163


5%

14,404 6 70% $171


5%

14,433 6 70% $179


5%

14,463 6 70% $188


0%

14,492 6 70% $188


0%

14,522 6 70% $188


0%

Discount offered BELVIQ - total South Korean Sales in men ($MM) BELVIQ/Phentermine - total South Korean Sales in men ($MM) Total women
% of women between 18 and 64 years old

22% $0 $0 24,521,819
62.1%

20% $1 $0 24,722,530
62.1%

20% $1 $0 24,772,964
62.1%

20% $2 $0 24,823,501
62.1%

20% $1 $3 24,874,141
62.1%

20% $1 $4 24,924,884
62.1%

20% $1 $6 24,975,731
62.1%

20% $1 $6 25,026,681
62.1%

20% $1 $7 25,077,736
62.1%

20% $1 $8 25,128,894
62.1%

20% $1 $8 25,180,157
62.1%

20% $1 $9 25,231,525
62.1%

20% $1 $9 25,282,997
62.1%

20% $1 $9 25,334,574
62.1%

20% $1 $9 25,386,257
62.1%

$0 $0 24,571,843
62.1%

$0 $0 24,621,970
62.1%

$0 $0 24,672,199
62.1%

# of women between 18 and 64 years old


% of women between 18 and 64 years old that are obese

15,223,151
20.0%

15,254,206
20.0%

15,285,325
20.0%

15,316,507
20.0%

15,347,752
20.0%

15,379,062
20.0%

15,410,435
20.0%

15,441,872
20.0%

15,473,374
20.0%

15,504,939
20.0%

15,536,569
20.0%

15,568,264
20.0%

15,600,023
20.0%

15,631,847
20.0%

15,663,736
20.0%

15,695,690
20.0%

15,727,710
20.0%

15,759,794
20.0%

# of women between 18 and 64 years old that are obese


BELVIQ penetration in women

3,044,630

3,050,841

3,057,065

3,063,301
0.16%

3,069,550
0.32%

3,075,812
0.44%

3,082,087
0.56%

3,088,374
0.40%

3,094,675
0.20%

3,100,988
0.20%

3,107,314
0.20%

3,113,653
0.20%

3,120,005
0.20%

3,126,369
0.20%

3,132,747
0.20%

3,139,138
0.20%

3,145,542
0.20%

3,151,959
0.20%

# of obese women treated with BELVIQ


BELVIQ/Phentermine penetration in women

4,901 6 70% $105


5%

9,823 6 70% $110


5%

13,534 6 70% $116


5%

17,260 6 70% $121


5%

12,353
0.80%

6,189
1.20%

6,202
1.60%

6,215
1.60%

6,227
1.80%

6,240
1.80%

6,253
1.80%

6,265
1.80%

6,278
1.80%

6,291
1.80%

6,304
1.80%

# of obese women treated with BELVIQ/Phentermine Average # of prescriptions (Rx)/patient % compliance Average cost/prescription (Rx)
% price increase

24,707 6 70% $127


5%

37,136 6 70% $134


5%

49,616 6 70% $140


5%

49,717 6 70% $147


5%

56,046 6 70% $155


5%

56,160 6 70% $163


5%

56,275 6 70% $171


5%

56,389 6 70% $179


5%

56,504 6 70% $188


0%

56,620 6 70% $188


0%

56,735 6 70% $188


0%

Discount offered BELVIQ - total South Korean Sales in women ($MM) BELVIQ/Phentermine - total South Korean Sales in women ($MM) BELVIQ - total South Korean Sales ($MM) BELVIQ/Phentermine - total South Korean Sales ($MM) BELVIQ + BELVIQ/Phentermine- total South Korea Sales ($MM) Total revenue on South Korean sales to ARNA $0 $0 $0 $0 $0 $0 $0 $0 $0 $0 $0 $0 $0 $0 $0 $0 $0 $0

22% $2 $0 $2 $0 $2 $1

20% $4 $0 $5 $0 $5 $2

20% $5 $0 $7 $0 $7 $2

20% $7 $0 $9 $0 $9 $3

20% $5 $11 $7 $13 $20 $8

20% $3 $17 $3 $21 $24 $10

20% $3 $23 $4 $29 $33 $14

20% $3 $25 $4 $31 $35 $15

20% $3 $29 $4 $37 $41 $17

20% $3 $31 $4 $39 $43 $18

20% $4 $32 $5 $41 $45 $19

20% $4 $34 $5 $43 $47 $20

20% $4 $36 $5 $45 $50 $21

20% $4 $36 $5 $45 $50 $21

20% $4 $36 $5 $45 $50 $22

Source: Cowen and Company

MEMBER: FINRA/SIPC

18

www.cowen.com

Cowen and Company, LLC

EQUITY RESEARCH

BELVIQ + BELVIQ/Phentermine revenue model ($MM) Taiwan


BELVIQ +BELVIQ/Phentermine Revenue Model (Taiwan)
Taiwan population
Population growth

2012E 23,299,716
0.27%

2013E 23,362,625
0.27%

2014E 23,425,704
0.27%

2015E 23,488,954
0.27%

2016E 23,552,374
0.27%

2017E 23,615,965
0.27%

2018E 23,679,728
0.27%

2019E 23,743,664
0.27%

2020E 23,807,772
0.27%

2021E 23,872,053
0.27%

2022E 23,936,507
0.27%

2023E 24,001,136
0.27%

2024E 24,065,939
0.27%

2025E 24,130,917
0.27%

2026E 24,196,070
0.27%

2027E 24,261,400
0.27%

2028E 24,326,905
0.27%

2029E 24,392,588
0.27%

Total men
% of men between 18 and 64 years old

11,699,718
63.5%

11,731,307
63.5%

11,762,981
63.5%

11,794,741
63.5%

11,826,587
63.5%

11,858,519
63.5%

11,890,537
63.5%

11,922,641
63.5%

11,954,833
63.5%

11,987,111
63.5%

12,019,476
63.5%

12,051,928
63.5%

12,084,469
63.5%

12,117,097
63.5%

12,149,813
63.5%

12,182,617
63.5%

12,215,510
63.5%

12,248,492
63.5%

# of men between 18 and 64 years old


% of men between 18 and 64 years old that are obese

7,431,300
20.0%

7,451,365
20.0%

7,471,483
20.0%

7,491,656
20.0%

7,511,884
20.0%

7,532,166
20.0%

7,552,503
20.0%

7,572,895
20.0%

7,593,341
20.0%

7,613,843
20.0%

7,634,401
20.0%

7,655,014
20.0%

7,675,682
20.0%

7,696,406
20.0%

7,717,187
20.0%

7,738,023
20.0%

7,758,916
20.0%

7,779,865
20.0%

# of men between 18 and 64 years old that are obese


BELVIQ penetration in men

1,486,260

1,490,273 -

1,494,297 -

1,498,331
0.04%

1,502,377
0.08%

1,506,433
0.11%

1,510,501
0.14%

1,514,579
0.10%

1,518,668
0.05%

1,522,769
0.05%

1,526,880
0.05%

1,531,003
0.05%

1,535,136
0.05%

1,539,281
0.05%

1,543,437
0.05%

1,547,605
0.05%

1,551,783
0.05%

1,555,973
0.05%

# of obese men treated with BELVIQ


BELVIQ/Phentermine penetration in men

599 6 70% $105


5%

1,202 6 70% $110


5%

1,657 6 70% $116


5%

2,115 6 70% $121


5%

1,515
0.20%

759
0.30%

761
0.40%

763
0.40%

766
0.45%

768
0.45%

770
0.45%

772
0.45%

774
0.45%

776
0.45%

778
0.45%

# of obese men treated with BELVIQ/Phentermine Average # of prescriptions (Rx)/patient % compliance Average cost/prescription (Rx)
% price increase

3,029 6 70% $127


5%

4,556 6 70% $134


5%

6,091 6 70% $140


5%

6,108 6 70% $147


5%

6,890 6 70% $155


5%

6,908 6 70% $163


5%

6,927 6 70% $171


5%

6,945 6 70% $179


5%

6,964 6 70% $188


0%

6,983 6 70% $188


0%

7,002 6 70% $188


0%

Discount offered BELVIQ - total Taiwan Sales in men ($MM) BELVIQ/Phentermine - total Taiwan Sales in men ($MM) Total women
% of women between 18 and 64 years old

22% $0 $0 11,725,987
62.1%

20% $0 $0 11,853,141
62.1%

20% $1 $0 11,885,145
62.1%

20% $1 $0 11,917,235
62.1%

20% $1 $1 11,949,411
62.1%

20% $0 $2 11,981,675
62.1%

20% $0 $3 12,014,025
62.1%

20% $0 $3 12,046,463
62.1%

20% $0 $4 12,078,988
62.1%

20% $0 $4 12,111,602
62.1%

20% $0 $4 12,144,303
62.1%

20% $0 $4 12,177,093
62.1%

20% $0 $4 12,209,971
62.1%

20% $0 $4 12,242,938
62.1%

20% $0 $4 12,275,994
62.1%

$0 $0 11,757,647
62.1%

$0 $0 11,789,393
62.1%

$0 $0 11,821,224
62.1%

# of women between 18 and 64 years old


% of women between 18 and 64 years old that are obese

7,279,495
20.0%

7,299,150
20.0%

7,318,858
20.0%

7,338,619
20.0%

7,358,433
20.0%

7,378,301
20.0%

7,398,222
20.0%

7,418,197
20.0%

7,438,226
20.0%

7,458,310
20.0%

7,478,447
20.0%

7,498,639
20.0%

7,518,885
20.0%

7,539,186
20.0%

7,559,542
20.0%

7,579,953
20.0%

7,600,419
20.0%

7,620,940
20.0%

# of women between 18 and 64 years old that are obese


BELVIQ penetration in women

1,455,899

1,459,830

1,463,772

1,467,724
0.16%

1,471,687
0.32%

1,475,660
0.44%

1,479,644
0.56%

1,483,639
0.40%

1,487,645
0.20%

1,491,662
0.20%

1,495,689
0.20%

1,499,728
0.20%

1,503,777
0.20%

1,507,837
0.20%

1,511,908
0.20%

1,515,991
0.20%

1,520,084
0.20%

1,524,188
0.20%

# of obese women treated with BELVIQ


BELVIQ/Phentermine penetration in women

2,348

4,709

6,493

8,286
0.00%

5,935
0.80%

2,975
1.20%

2,983
1.60%

2,991
1.60%

2,999
1.80%

3,008
1.80%

3,016
1.80%

3,024
1.80%

3,032
1.80%

3,040
1.80%

3,048
1.80%

# of obese women treated with BELVIQ/Phentermine Average # of prescriptions (Rx)/patient % compliance Average cost/prescription (Rx)
% price increase

6 70% $105
5%

11,869 6 6 70% $127


5%

17,852 6 70% $134


5%

23,867 6 70% $140


5%

23,931 6 70% $147


5%

26,995 6 70% $155


5%

27,068 6 70% $163


5%

27,141 6 70% $171


5%

27,214 6 70% $179


5%

27,288 6 70% $188


0%

27,362 6 70% $188


0%

27,435 6 70% $188


0%

6 70% $110
5%

6 70% $116
5%

70% $121
5%

Discount offered BELVIQ - total Taiwan Sales in women ($MM) BELVIQ/Phentermine - total Taiwan Sales in women ($MM) BELVIQ - total Taiwan Sales ($MM) BELVIQ/Phentermine - total Taiwan Sales ($MM) BELVIQ + BELVIQ/Phentermine- total Taiwan Sales ($MM) % increase Total revenue on Taiwan sales to ARNA $0 $0 $0 $0 $0 $0 $0 $0 $0 $0 $0 $0 $0 $0 $0 $0 $0 $0

22% $1 $0 $1 $0 $1

20% $2 $0 $2 $0 $2 116%

20% $3 $0 $3 $0 $3 45% $1

20% $3 $0 $4 $0 $4 34% $2

20% $3 $5 $3 $6 $10 126% $4

20% $1 $8 $2 $10 $12 23% $5

20% $1 $11 $2 $14 $16 35% $7

20% $1 $12 $2 $15 $17 5% $8

20% $2 $14 $2 $18 $20 17% $9

20% $2 $15 $2 $19 $21 5% $9

20% $2 $16 $2 $20 $22 5% $10

20% $2 $16 $2 $21 $23 5% $10

20% $2 $17 $2 $22 $24 5% $11

20% $2 $17 $2 $22 $24 0% $11

20% $2 $17 $2 $22 $24 0% $11

$0

$1

Source: Cowen and Company

MEMBER: FINRA/SIPC

19

www.cowen.com

Cowen and Company, LLC

EQUITY RESEARCH

Arena: P&L and Balance Sheet


Income statement: For 2012, Arena reported a net loss of $88.3M, or ($0.45) per share, compared to a loss of $111.5M, or ($0.80) per share, in 2011. Total operating expenses in 2012 were $81M, compared to $87.4M in 2011. R&D expenses in 2012 were $54.1M, compared to $58.7M in 2011, while G&A expenses were $26.2M, compared to $26.2M in 2011. In 2Q13, the most recently reported quarter, Arena reported a net profit of $40.1M, or $0.18 per share, compared to a net loss of $22.1M, or ($0.12) per share, in 2Q12. The profit was primarily due to the $65M milestone payment received from Eisai in June 2013 for the delivery of launch supplies and U.S. launch. Total operating expenses for 2Q13 were $27.4M, compared to the $19.5M spent in 2Q12. R&D expenses were $18.8M in 2Q13, compared to the $14.1M spent in 2Q12, while G&A expenses were $8.6M in 2Q13, compared to the $5.2M spent in 2Q12. 2013 financial guidance: Arena expects 2013 R&D expenses to be in the range of $70-$78M, including non-cash expenses of approximately $7M. G&A expenses are expected to be in the range of $28-$34M, including non-cash expenses of approximately $6M. Balance sheet: Arena ended 2Q13 with $178.9M or $0.76/fully diluted share in cash. Share count: As of August 2013, the company had 218.2M common shares, 2M warrants, and 14.6M options outstanding, bringing the fully diluted number of shares to ~234.8M. Options and Warrants Outstanding (MM)
Warrants Outstanding August 2008 Series B Warrants Total Warrants Outstanding Total Options Outstanding 2.0 2.0 14.6 16.5 Weighted AverageExercise Price $4.34 6.23 $4.88 Expiration Date 8/14/2015

Total Options and Warrants outstanding

Source: Cowen and Company, SEC Filings

Manufacturing plant in Switzerland provides 10-year tax-break: In 2008, Arenas Swiss subsidiary, Arena GmbH, purchased a plant in Switzerland from Siegfried, a company which provides custom drug development services, including drug substances and drug manufacturing, for $38.7M ($30.7M in cash and 1.5M shares valued at $8M). Arena plans to manufacture BELVIQ at this facility and sell it to its partners. This subsidiary has been granted a conditional incentive tax holiday for its operations in Switzerland, which Arena expects will exempt it from the majority of potential Swiss income taxes. This tax holiday will continue for a period of up to 10 years, not to extend beyond December 31, 2022. Arena has guided that, as

MEMBER: FINRA/SIPC

20

www.cowen.com

Cowen and Company, LLC

EQUITY RESEARCH

a result of this tax holiday, it expects to pay lower overall taxes during that period, at a tax rate in the range of 15-20%. Arena: Quarterly P&L ($MM)
($MM) Total revenue on US sales to ARNA Total revenue on EU sales to ARNA - probability-adjusted Total revenue on SK sales to ARNA - probability-adjusted Total revenue onTaiwan sales to ARNA - probability-adjusted Total revenue on sales to ARNA Manufacturing services Milestones/License Fees - Eisai/Others Milestones/License Fees - EU partnership Milestones/License Fees - Ildong Milestones/License Fees - CYB Total Revenues Cost of Goods Sold Gross Profit R&D SG&A Total Operating Expenses
% Revenues

2009A 0.0 0.0 0.0 0.0 0.0 6.6 3.8 0.0 0.0 0.0 10.4 6.5 3.9 110.2 25.2 142.2
1369.4%

2010A 0.0 0.0 0.0 0.0 0.0 7.1 9.6 0.0 0.0 0.0 16.6 7.4 9.2 75.5 27.9 105.6
635.4%

2011A 0.0 0.0 0.0 0.0 0.0 5.3 7.4 0.0 0.0 0.0 12.7 8.1 4.6 58.7 24.2 87.4
687.3%

Q1:12A 0.0 0.0 0.0 0.0 0.0 1.3 0.9 0.0 0.0 0.0 2.2 0.8 1.4 14.5 6.4 21.0 (19.6)

Q2:12A 0.0 0.0 0.0 0.0 0.0 1.0 20.9 0.0 0.0 0.0 22.0 0.7 21.3 14.1 5.2 19.5 1.9

Q3:12A 0.0 0.0 0.0 0.0 0.0 0.6 0.9 0.0 0.0 0.0 1.5 1.4 0.1 11.6 7.4 19.2 (19.1)

Q4:12A 0.0 0.0 0.0 0.0 0.0 0.9 0.9 0.0 0.2 0.0 1.9 0.8 1.1 13.9 7.3 21.4 (20.3)

2012A 0.0 0.0 0.0 0.0 0.0 3.8 23.6 0.0 0.2 0.0 27.6 3.7 23.9 54.1 26.2 81.0
293.7%

Q1:13A 0.0 0.0 0.0 0.0 0.0 0.8 1.5 0.0 0.1 0.0 2.4 2.1 0.3 14.0 7.3 21.3 (21.0)

Q2:13A 1.3 0.0 0.0 0.0 1.3 1.0 66.5 0.0 0.2 0.0 68.9 1.6 67.3 18.8 8.6 27.4 39.9

Q3:13E 1.9 0.0 0.0 0.0 1.9 0.8 1.5 0.0 0.2 0.0 4.3 1.3 3.0 20.0 8.7 28.9 (25.9)

Q4:13E 2.6 0.0 0.0 0.0 2.6 0.8 2.0 0.0 0.2 0.0 5.5 1.5 4.0 22.0 8.8 31.0 (27.0)

2013E 5.8 0.0 0.0 0.0 5.8 3.2 71.5 0.0 0.6 0.0 81.1 6.5 74.5 74.8 33.4 108.5
133.9%

Q1:14E 3.2 0.0 0.0 0.0 3.2 0.9 1.5 0.0 0.2 0.0 5.7 1.8 4.0 14.0 8.7 22.9 (18.9)

Q2:14E 3.8 0.0 0.0 0.0 3.8 0.9 1.5 0.0 0.2 0.0 6.3 1.9 4.4 14.0 8.8 23.0 (18.6)

Q3:14E 4.6 0.0 0.0 0.0 4.6 0.9 1.5 0.0 0.2 0.1 7.2 2.2 5.1 14.0 8.7 22.9 (17.8)

Q4:14E 5.5 0.0 0.0 0.0 5.5 0.9 11.5 0.0 0.2 0.1 18.2 2.4 15.8 14.0 8.8 23.0 (7.2)

2014E 17.0 0.0 0.0 0.0 17.0 3.6 16.0 0.0 0.6 0.2 37.5 8.3 29.2 56.0 35.0 91.7
244.7%

2015E 27.6 0.0 0.6 0.4 28.5 0.0 16.0 50.0 3.0 0.2 97.7 5.4 92.3 57.1 35.7 93.5
95.7%

Operating Income
% Revenues

(138.4)

(96.4)

(82.8)

(57.1)

(34.0)
-42.0%

(62.5)
-166.8%

(1.2)
-1.2%

Total Non-Operating Income Pretax Income Income tax expense Tax Rate Net Income - Operations Non-Recurring Gains (Losses) Net Income - Reported Basic EPS Diluted EPS Shares outstanding (basic) Shares outstanding (diluted)

(14.8) (153.2) 0.0 0.0% (153.2) 0.0 (153.2) ($1.82) ($1.82) 84.34 123.52

(28.2) (124.5) 0.0 0.0% (124.5) 0.0 (124.5) ($1.14) ($1.14) 109.6 150.1

(28.7) (111.5) 0.0 0.0% (111.5) 0.0 (111.5) ($0.80) ($0.80) 139.2 181.6

(9.8) (29.4) 0.0 0.0% (29.4) 0.0 (29.4) ($0.18) ($0.18) 164.2 205.0

(24.0) (22.1) 0.0 0.0% (22.1) 0.0 (22.1) ($0.12) ($0.12) 190.3 202.5

3.6 (15.5) 0.0 0.0% (15.5) 0.0 (15.5) ($0.07) ($0.07) 213.9 233.9

(1.0) (21.3) 0.0 0.0% (21.3) 0.0 (21.3) ($0.10) ($0.10) 217.3 234.7

(31.2) (88.3) 0.0 0.0% (88.3) 0.0 (88.3) ($0.45) ($0.45) 196.4 219.1

2.1 (18.9) 0.0 0.0% (18.9) 0.0 (18.9) ($0.09) ($0.09) 217.5 236.3

0.2 40.1 0.0 0.0% 40.1 0.0 40.1 $0.18 $0.18 217.9 224.5

(1.8) (27.7) 0.0 0.0% (27.7) 0.0 (27.7) ($0.13) ($0.13) 219.3 234.8

(1.8) (28.8) 0.0 0.0% (28.8) 0.0 (28.8) ($0.13) ($0.13) 220.4 235.9

(1.3) (35.3) 0.0 0.0% (35.3) 0.0 (35.3) ($0.16) ($0.16) 218.8 232.9

(1.8) (20.7) 0.0 0.0% (20.7) 0.0 (20.7) ($0.08) ($0.08) 246.5 262.1

(1.8) (20.4) 0.0 0.0% (20.4) 0.0 (20.4) ($0.08) ($0.08) 247.8 263.4

(1.8) (19.6) 0.0 0.0% (19.6) 0.0 (19.6) ($0.08) ($0.08) 249.0 264.7

(1.8) (9.0) 0.0 0.0% (9.0) 0.0 (9.0) ($0.04) ($0.04) 250.3 266.1

(7.2) (69.7) 0.0 0.0% (69.7) 0.0 (69.7) ($0.28) ($0.28) 248.4 264.1

(7.2) (8.4) 0.0 0.0% (8.4) 0.0 (8.4) ($0.03) ($0.03) 255.3 271.4

Source: Cowen and Company, Arena Pharmaceuticals

Arena: Annual P&L ($MM)


($MM) Total revenue on US sales to ARNA Total revenue on EU sales to ARNA - probability-adjusted Total revenue on SK sales to ARNA - probability-adjusted Total revenue onTaiwan sales to ARNA - probability-adjusted Total revenue on sales to ARNA Manufacturing services Milestones/License Fees - Eisai/Others Milestones/License Fees - EU partnership Milestones/License Fees - Ildong Milestones/License Fees - CYB Total Revenues Cost of Goods Sold Gross Profit R&D SG&A Total Operating Expenses
% Revenues

2009A 0.0 0.0 0.0 0.0 0.0 6.6 3.8 0.0 0.0 0.0 10.4 6.5 3.9 110.2 25.2 142.2
1369.4%

2010A 0.0 0.0 0.0 0.0 0.0 7.1 9.6 0.0 0.0 0.0 16.6 7.4 9.2 75.5 27.9 105.6
635.4%

2011A 0.0 0.0 0.0 0.0 0.0 5.3 7.4 0.0 0.0 0.0 12.7 8.1 4.6 58.7 24.2 87.4
687.3%

2012A 0.0 0.0 0.0 0.0 0.0 3.8 23.6 0.0 0.2 0.0 27.6 3.7 23.9 54.1 26.2 81.0
293.7%

2013E 5.8 0.0 0.0 0.0 5.8 3.2 71.5 0.0 0.6 0.0 81.1 6.5 74.5 74.8 33.4 108.5
133.9%

2014E 17.0 0.0 0.0 0.0 17.0 3.6 16.0 0.0 0.6 0.2 37.5 8.3 29.2 56.0 35.0 91.7
244.7%

2015E 27.6 0.0 0.6 0.4 28.5 0.0 16.0 50.0 3.0 0.2 97.7 5.4 92.3 57.1 35.7 93.5
95.7%

2016E 38.5 2.8 1.3 0.8 43.4 0.0 71.0 25.0 0.6 0.2 140.2 8.1 132.1 58.2 36.4 95.4
68.0%

2017E 49.9 6.3 1.9 1.1 59.2 0.0 6.0 0.0 0.6 0.2 66.0 11.0 55.0 58.6 37.1 96.5
146.1%

2018E 92.3 9.9 2.6 1.5 106.3 0.0 55.0 25.0 0.0 0.2 186.5 16.4 170.1 59.0 37.9 97.6
52.3%

2019E 120.1 14.0 6.4 3.4 144.0 0.0 0.0 0.0 0.0 0.2 144.2 21.7 122.4 59.4 38.6 98.7
68.5%

2020E 150.9 18.2 8.0 4.2 181.3 0.0 55.0 25.0 0.0 0.2 261.5 27.1 234.4 53.8 39.4 93.9
35.9%

2021E 209.5 23.0 11.1 5.7 249.3 0.0 0.0 0.0 0.0 0.2 249.5 36.9 212.6 50.0 40.2 90.9
36.4%

2022E 269.1 28.2 11.7 6.0 315.0 0.0 100.0 25.0 0.0 0.2 440.2 46.3 393.9 47.5 41.0 89.2
20.3%

2023E 249.7 31.7 13.8 7.1 302.3 0.0 0.0 0.0 0.0 0.2 302.5 44.4 258.1 45.8 41.8 88.3
29.2%

2024E 265.7 35.6 14.6 7.4 323.3 0.0 0.0 37.5 0.0 0.0 360.8 47.3 313.5 44.8 42.7 88.1
24.4%

2025E 282.7 39.7 15.4 7.8 345.7 0.0 0.0 37.5 0.0 0.0 383.2 50.4 332.8 44.2 43.5 88.4
23.1%

2026E 300.7 44.2 16.3 8.2 369.4 0.0 150.0 0.0 0.0 0.0 519.4 53.7 465.8 26.5 44.4 71.6
13.8%

2027E 303.2 4.4 17.2 8.7 333.5 0.0 0.0 0.0 0.0 0.0 333.5 48.6 284.8 15.9 45.3 61.9
18.6%

2028E 305.1 2.2 17.2 8.7 333.2 0.0 0.0 0.0 0.0 0.0 333.2 48.6 284.6 9.5 46.2 55.7
16.7%

2029E 308.0 2.2 17.2 8.7 336.2 0.0 0.0 0.0 0.0 0.0 336.2 49.0 287.2 5.7 47.1 52.8
15.7%

Operating Income
% Revenues

(138.4)

(96.4)

(82.8)

(57.1)

(34.0)
-42.0%

(62.5)
-166.8%

(1.2)
-1.2%

36.7
26.2%

(41.5)
-62.8%

72.6
38.9%

23.7
16.4%

140.5
53.7%

121.7
48.8%

304.8
69.2%

169.7
56.1%

225.4
62.5%

244.4
63.8%

394.2
75.9%

222.9
66.9%

228.9
68.7%

234.3
69.7%

Total Non-Operating Income Pretax Income Income tax expense Tax Rate Net Income - Operations Non-Recurring Gains (Losses) Net Income - Reported Basic EPS Diluted EPS Shares outstanding (basic) Shares outstanding (diluted)

(14.8) (153.2) 0.0 0.0% (153.2) 0.0 (153.2) ($1.82) ($1.82) 84.34 123.52

(28.2) (124.5) 0.0 0.0% (124.5) 0.0 (124.5) ($1.14) ($1.14) 109.6 150.1

(28.7) (111.5) 0.0 0.0% (111.5) 0.0 (111.5) ($0.80) ($0.80) 139.2 181.6

(31.2) (88.3) 0.0 0.0% (88.3) 0.0 (88.3) ($0.45) ($0.45) 196.4 219.1

(1.3) (35.3) 0.0 0.0% (35.3) 0.0 (35.3) ($0.16) ($0.16) 218.8 232.9

(7.2) (69.7) 0.0 0.0% (69.7) 0.0 (69.7) ($0.28) ($0.28) 248.4 264.1

(7.2) (8.4) 0.0 0.0% (8.4) 0.0 (8.4) ($0.03) ($0.03) 255.3 271.4

(7.2) 29.5 0.0 0.0% 29.5 0.0 29.5 $0.11 $0.11 260.4 276.8

(7.2) (48.7) 0.0 0.0% (48.7) 0.0 (48.7) ($0.18) ($0.18) 265.6 282.3

0.0 72.6 3.6 5.0% 68.9 0.0 68.9 $0.25 $0.24 270.9 288.0

0.0 23.7 1.9 8.0% 21.8 0.0 21.8 $0.08 $0.07 276.3 293.7

0.0 140.5 21.1 15.0% 119.4 0.0 119.4 $0.42 $0.40 281.8 299.6

0.0 121.7 18.3 15.0% 103.5 0.0 103.5 $0.36 $0.34 287.5 305.6

0.0 304.8 61.0 20.0% 243.8 0.0 243.8 $0.83 $0.78 293.2 311.7

0.0 169.7 33.9 20.0% 135.8 0.0 135.8 $0.45 $0.43 299.1 318.0

0.0 225.4 56.3 25.0% 169.0 0.0 169.0 $0.55 $0.52 305.1 324.3

0.0 244.4 61.1 25.0% 183.3 0.0 183.3 $0.59 $0.55 311.2 330.8

0.0 394.2 98.5 25.0% 295.6 1.0 296.6 $0.93 $0.88 317.4 337.4

0.0 222.9 55.7 25.0% 167.2 2.0 169.2 $0.52 $0.49 323.7 344.2

0.0 228.9 57.2 25.0% 171.7 3.0 174.7 $0.53 $0.50 330.2 351.1

3.0 237.3 59.3 25.0% 178.0 4.0 182.0 $0.54 $0.51 336.8 358.1

Source: Cowen and Company, Arena Pharmaceuticals

MEMBER: FINRA/SIPC

21

www.cowen.com

Cowen and Company, LLC

EQUITY RESEARCH

Arenas BELVIQ Is Approved and Now Available in U.S.


After completing a three-trial Phase III program, in June 2012, Arenas BELVIQ became the first new weight loss drug to be approved by the FDA in 13 years. BELVIQ is indicated, as an adjunct to low-calorie diet and exercise, for chronic weight management in adults with initial BMI of 30 kg/m2 (obese), or 27 kg/m2 (overweight) in the presence of at least one weightrelated comorbid condition, such as hypertension, dyslipidemia, or type 2 diabetes. The recommended dose of BELVIQ is 10 mg administered orally twice daily, taken with or without food. Response to BELVIQ should be evaluated after 12 weeks, and treatment should be discontinued at that time if the patient has achieved less than 5% weight loss. DEA Schedule and Launch: The FDA recommended that BELVIQ be classified by the DEA as a Schedule IV drug. On May 7, 2013, Arena announced that the DEA designation of BELVIQ as a Schedule IV drug had been filed by the Office of the Federal Register for public inspection. On June 7, 2013 BELVIQ was commercially launched in the U.S. and became available through retail pharmacies on June 11, 2013.

A Look at the U.S. BELVIQ Launch


Eisai launched BELVIQ using a primary care sales force of 200 sales reps who previously detailed Aricept and AcipHex, initially targeting 20,000 to 30,000 physicians (PCPs, cardiologists, and endocrinologists) who actively treat obesity. With this sales force, Eisai generated Aricept and AcipHex U.S. sales of $2.9B, $3B, $2.6B, $844M, and $636M in 2008, 2009, 2010, 2011, and 2012 respectively. In addition, according to Arena, Eisai will utilize approximately 50 dedicated managed markets specialists and 3 health economists to focus on BELVIQs reimbursement. Based on discussions with the company, this is the initial level of detailing effort Eisai is putting behind BELVIQ, and the possibility that it may be increased, if needed, is on the table. In its 2Q13 earnings call, Arena disclosed that there were 12,500 total prescriptions filled in the first six weeks of BELVIQs launch (between 6/11/2013 and 7/19/2013). Eisai has reported that ~3,900 physicians, the majority of whom are primary care physicians, have already started prescribing BELVIQ.

MEMBER: FINRA/SIPC

22

www.cowen.com

Cowen and Company, LLC

EQUITY RESEARCH

BELVIQ Weekly NRx Data (IMS and Symphony Health)

Belviq Weekly NRx


4,500 4,000 3,500 3,000 2,500 2,000 1,500 1,000 500 -

IMS data (NRx) SH data (NRx)

Source: Cowen and Company, IMS and Symphony Health

BELVIQ Weekly TRx Data (IMS and Symphony Health)

Belviq Weekly TRx


5,000 4,500 4,000 3,500 3,000 2,500 2,000 1,500 1,000 500 -

IMS data (TRx) SH data (TRx)

Source: Cowen and Company, IMS and Symphony Health

BELVIQ Sales: For 2Q13, Eisai disclosed gross revenue of $10M from BELVIQ sales, corresponding to ~50,000 bottles of BELVIQ shipped to its distributors (average WAC ~$200). The net sales were ~$4.2M, and thus Arena recognized $1.3M in revenues, given the 31.5% transfer price, from BELVIQ sales. BELVIQ Savings Program: Eisai has a 15-Day free trial voucher program in which patients can receive 15 days of BELVIQ for free. Also, a Savings Card is offered, through which patients receive up to $75 off each month for 12 months ($75 off for cash-pay patients; up to $75 off for patients with insurance coverage having an initial $50 or more co-pay each month). Reimbursement Efforts: According to Arena, Eisai has approximately 50 reimbursement specialists and 3 healthcare economists working on reimbursement for BELVIQ. Eisais stated

MEMBER: FINRA/SIPC

23

www.cowen.com

Cowen and Company, LLC

EQUITY RESEARCH

goal is to improve reimbursement from the current ~ 30% to as much as 50% by the end of its fiscal year (March 2014). Consumer Launch: In September 2013, Arena reported that Eisais US launch campaign for BELVIQ has entered its consumer phase. Full page announcements were placed in major US magazines beginning in September, which are intended to educate consumers about BELVIQ as a weight loss treatment option. Also, Eisai has launched the BELIEVE EVERYDAY SUPPORT program for BELVIQ, which provides free support and savings, and includes an online portal. The program includes: a monthly savings card, one-year membership to the Lose It! app for calorie- and activity-tracking, nutrition information and meal planning tool, and motivational tips to help sustain weight loss efforts.

BELVIQs (Lorcaserin) Mechanism of Action


BELVIQ (lorcaserin) is a selective serotonin (5-HT) 2C receptor agonist. The 5-HT2C receptor subtype appears to be involved in the central regulation of satiety, as validated in both animal models and human studies. It is believed that lorcaserin decreases food consumption and promotes satiety by selectively activating serotonin 2C receptors in the brain. Notably, the serotonergic agonist fenfluramine was historically one of the most successful weight loss drugs ever used, but was withdrawn from the market in 1997 due to cardiac toxicity, considered to result from the compounds lack of selectivity. In ad dition to being an agonist of the 5-HT2C receptor, the action of fenfluramine on 5-HT2B receptors in the heart led to valvular disorders. In contrast, BELVIQ features multi-fold selectivity for the 5-HT2C receptor compared with the 5-HT2B receptor, and also shows relatively low activity at other serotonin receptor subtypes. High-dose studies in animals over several months have not demonstrated appreciable cardiovascular changes with lorcaserin treatment. Lorcaserin Potency (EC50) and Binding Affinity (Ki) to Human 5-HT2A, 5-HT2B, and 5HT2C Receptor Subtypes

Serotonin Receptor Subtype 5-HT 2C 5-HT 2B 5-HT 2A


Source: Cowen and Company, BELVIQ Label

EC50, nM Ki, nM 39 2380 553 13 147 92

MEMBER: FINRA/SIPC

24

www.cowen.com

Cowen and Company, LLC

EQUITY RESEARCH

BELVIQ (Lorcaserin): Clinical Data Summary


BELVIQ Successfully Completed a Three-Trial Phase III Program
BELVIQs Phase III pivotal program consisted of two randomized, placebo-controlled trials, BLOOM and BLOSSOM, which were required for NDA submission. Positive results from BLOOM and BLOSSOM were reported in March 2009 and September 2009, respectively. A third, non-pivotal, Phase III trial (BLOOM-DM) in type 2 diabetics had positive results reported in November 2010. All three trials were conducted in the U.S. BELVIQ has demonstrated modest weight loss with a benign safety profile in these three Phase III trials. The co-primary endpoint for all the studies were: the proportion of patients achieving 5% weight loss from baseline at one year, mean change in body weight from baseline, and the proportion of patients achieving 10% weight loss from baseline at one year. Per FDA guidelines, subjects enrolled in all three trials (BLOOM, BLOSSOM and BLOOM-DM) received diet and exercise counseling, in addition to drug or placebo. FDA guidance has delineated the primary efficacy benchmark for weight management drugs as achievement of either a statistically significant difference in mean weight loss of at least 5% between therapy and placebo groups, or a categorical loss > 5% of baseline body weight in at least 35% of patients on therapy, which proportion should be approximately double that for placebo. The following tables provide an overview of the BELVIQ Phase III program:

MEMBER: FINRA/SIPC

25

www.cowen.com

Cowen and Company, LLC

EQUITY RESEARCH

BELVIQ Phase III Program Summary

BLOOM
Number of patients Duration Treatment groups 3,182 2 years BELVIQ 10 mg BID, Placebo

BLOSSOM
4,008 1 year BELVIQ 10 mg QD, BELVIQ 10 mg BID, Placebo

BLOOM-DM*
604 1 year BELVIQ 10 mg QD, BELVIQ 10 mg BID, Placebo BMI 27-45 kg/m2

Patient demographics

BMI 30-45 kg/m2, or 27-45 kg/m2 with 1 BMI 30-45 kg/m2, or 27-29.9 kg/m2 with co-morbid condition 1 co-morbid condition Required one or more of the following: hypertension, dyslipidemia, cardiovascular disease, impaired glucose tolerance, or sleep apnea 220 36 44 84% Caucasian (67%) African-American (19%) Hispanic (12%) Individual 15-60 min counseling on nutrition and exercise at weeks 2/4 and then monthly 220 36 44 80% Caucasian (67%) African-American (20%) Hispanic (11%) Individual counseling on nutrition and exercise at baseline, weeks 1/2/4 and then monthly

Comorbidities for inclusion Mean baseline weight, lbs Mean BMI, kg/m2 Average age (years) Female Proportion Ethnicity

Type 2 diabetes 228 36 53 54% Caucasian (61%) African-American (21%) Hispanic (14%) Individual 15-60 min counseling on nutrition and exercise at weeks 2/4 and then monthly

Adjunct lifestyle/behavior modification program

Co-primary endpoints Data Announced March 2009

1) proportion of patients 5% weight loss 2) mean change in body weight from baseline 3) proportion of patients 10% weight loss September 2009 November 2010

Source: Cowen and Company, Arena Pharmaceuticals, *BLOOM-DM was not a pivotal trial

BELVIQ Phase III Efficacy Data


Mean Weight Loss (%) Treatment Group Mean Weight Loss (%) Placebo Group Mean Weight Loss (%) >5% Weight Loss >5% Weight Loss Placebo- adjusted Treatment Group Placebo Group >5% Weight Loss PlaceboAdjusted >10% Weight Loss Treatment Group >10% Weight Loss Placebo Group >10% Weight Loss PlaceboAdjusted

BELVIQ 10mg BID (ITT)


BLOOM (009); BELVIQ 10mg BID (n=3,182) BLOOM-DM (010); BELVIQ 10mg BID; (n=604) BLOSSOM (011); BELVIQ 10mg BID; (n=4,008) 5.8% 4.5% 5.8% 2.2% 1.5% 2.8% 3.6% 3.0% 3.0% 47.5% 37.5% 47.2% 20.3% 16.1% 25.0% 27.2% 21.4% 22.2% 22.6% 16.3% 22.6% 7.7% 4.4% 9.7% 14.9% 11.9% 12.9%

BELVIQ 10mg BID (completers)


BLOOM (009); BELVIQ 10mg BID (n=3,182) BLOOM-DM (010); BELVIQ 10mg BID; (n=604) BLOSSOM (011); BELVIQ 10mg BID; (n=4,008) 8.1% 5.5% 7.9% 3.3% 1.7% 4.0% 4.8% 3.8% 3.9% 66.4% 44.6% 63.2% 32.1% 17.9% 34.9% 34.3% 26.7% 28.3% 36.2% 20.8% 35.1% 13.6% 5.8% 16.1% 22.6% 15.0% 19.0%

Source: Cowen and Company, Arena Pharmaceuticals

MEMBER: FINRA/SIPC

26

www.cowen.com

Cowen and Company, LLC

EQUITY RESEARCH

The BLOOM trial included serial 2D echocardiogram monitoring for all subjects, to evaluate for changes consistent with FDA-defined valvulopathy. ECHOs were performed at baseline, 6, 12, 18, and 24 months, and 6- and 12-month results were reviewed by a DSMB. Although the primary efficacy analysis in BLOOM was based on one-year results, the safety evaluation spanned two years. BLOSSOM and BLOOM-DM were one-year trials, which began in December 2007, and the FDA allowed Arena to enroll patients with pre-existing valvulopathy into both trials. ECHOs were collected at baseline, 6, and 12 months for database purposes. Enrollment in the BLOOM-DM trial was initially slow, in part due to a fairly high level of competition for patients among numerous U.S. diabetes studies, but was completed in August 2009, totaling 604 subjects with type 2 diabetes managed on oral medications.

Trial #1: BLOOM meets FDAs efficacy requirements


BLOOM was a 2-year, randomized, placebo-controlled, double-blind, Phase III study which evaluated the efficacy and safety of BELVIQ, administered with a lifestyle modification program. For the first year, 3,182 patients were randomized 1:1 to receive BELVIQ 10 mg BID or placebo. Patients remaining in the trial at the end of year 1 could continue in the study for another year. For year 2, patients in the placebo group continued on placebo, but patients who had received BELVIQ were again randomized 2:1 to continue on BELVIQ or to begin on placebo. Eligibility criteria for the trial included age 18-65 years and BMI 30-45 kg/m2 or 27-45 kg/m2 having at least one of the following co-morbid conditions: hypertension, dyslipidemia, cardiovascular disease, impaired glucose tolerance, or sleep apnea. At each patient visit (weeks 2 and 4 after randomization and then monthly), a trained counselor provided individual standardized nutritional and exercise counseling lasting 15-60 minutes, depending on the study week. Patients were instructed to participate in moderate exercise for 30 minutes per day and to reduce daily caloric intake by 600 kcal. In March 2009, Arena announced that the BLOOM trial achieved all three of its co-primary efficacy endpoints (% of patients losing 5% body weight, average weight loss, and % of patients losing 10% of their weight) with p<0.001. BLOOM Trial: Baseline Patient Characteristics

Number of Patients Average BMI, kg/m Average Weight Average Age, years Female Proportion
Source: Cowen and Company, Arena Pharmaceuticals
2

3,182 36.2 220 lbs 44.1

84%

1) % average weight loss: BELVIQ patients had an average weight loss of 5.8% (12.7 lbs) vs. 2.2% (4.7 lbs) for placebo, for a placebo-adjusted average weight loss of 3.6%, or 8 lbs. This

MEMBER: FINRA/SIPC

27

www.cowen.com

Cowen and Company, LLC

EQUITY RESEARCH

result was statistically significant; it did not, however, meet FDAs efficacy benchmark of a 5% difference in average weight loss between therapy and placebo. 2) % of patients losing at least 5% of their weight: 47.5% of BELVIQ patients lost at least 5% of baseline body weight, compared to 20.3% of placebo patients. With these data, BELVIQ met the threshold specified in FDA's draft guidance for efficacy of obesity drugs. 3) % of patients losing at least 10% of their weight: 22.6% of BELVIQ patients in the trial lost at least 10% of body weight, compared to 7.7% of placebo patients. Arena presented updated data at ADA 2009 from an analysis of those patients who completed one year of treatment per protocol. In this population, the average weight loss was 18 lbs (8%) in the drug arm, compared to 7.3 lbs (3.2%) in the placebo group. 66.4% of BELVIQ completers lost at least 5% of baseline body weight, compared to 32.1% of placebo patients (p<0.0001), while 36.2% of BELVIQ completers lost at least 10% of body weight, compared to 13.6% for placebo (p<0.0001). BLOOM Trial: Efficacy Data

BLOOM (n=3,182) Key Endpoints n (ITT-LOCF) Mean Weight Loss (%) (ITT-LOCF) 5% weight loss (ITT-LOCF) 10% weight loss (ITT-LOCF) n (Completers) Mean Weight Loss (%) (Completers) 5% weight loss (Completers) 10% weight loss (Completers) Completion rate
Source: Cowen and Company, Arena Pharmaceuticals

BELVIQ 10mg BID n=1,538 5.8% 47.5% 22.6% n=583 8.1% 66.4% 36.2% 38%

Placebo n=1,499 2.2% 20.3% 7.7% n=737 3.3% 32.1% 13.6% 49%

Difference

p-value

3.6% 27.2% 14.9%

<0.001 <0.001 <0.001

4.8% 34.3% 22.6% -

<0.001 <0.001 <0.001

BELVIQ also improved a number of secondary parameters to a greater extent than placebo (p<0.001), including waist circumference (-6.8cm vs. -3.9cm), triglycerides (-6.15% vs. 0.14%), hs-CRP (-1.19mg/L vs. -0.17mg/L), hemoglobin A1C (-0.04% vs. 0.03%) and quality of life (as measured by the IWQOL-LITE scale, 12.4 for BELVIQ vs. 10.7 for placebo, p<0.001). With BELVIQ treatment, at 52 weeks, there were also significant reductions from baseline in systolic and diastolic blood pressures, as well as heart rate, compared with placebo.

MEMBER: FINRA/SIPC

28

www.cowen.com

Cowen and Company, LLC

EQUITY RESEARCH

BLOOM Trial: Secondary Endpoints


BELVIQ (n = 1,538) Change in waist circumference Change in triglycerides Change in hs-CRP Change in IWQOL-LITE Change in heart rate Change Systolic BP (mm Hg) Change Diastolic BP (mm Hg) -6.8cm -6.15% -1.19mg/L 12.4 -2.0 bpm -1.4 mm Hg -1.1 mm Hg) Placebo (n = 1,499) -3.9cm -0.14% -0.17mg/L 10.7 -1.6 bpm -0.8 mm Hg -0.6 mm Hg p-value <0.001 <0.001 <0.001 <0.001 0.0499 0.04 0.01

Source: Cowen and Company, Arena Pharmaceuticals

In BLOOM, the primary endpoint at year 2 was the proportion of patients losing 5% of baseline body weight at the end of year 1 who maintained this weight reduction. In BELVIQ patients who had 5% weight loss at 52 weeks, more patients who continued to receive BELVIQ in year 2 maintained this loss, compared to those on placebo (67.9% vs. 50.3%, p<0.001). The mean body weight was lower in patients treated with BELVIQ for two years compared with mean body weight in patients on placebo for both years.

Could Safety be BELVIQs distinguishing feature?


While generally considering BELVIQs efficacy to be modest, our consultants were more positive about its side effect profile. BLOOM produced no serious safety signals. There were no seizures, and CNS adverse events (depression, anxiety, and suicidal ideation) occurred at similar rates in both groups. The incidence of depression, depressive symptoms, or depressed mood was 2.5% in the BELVIQ group and 2.2% in the placebo group at 52 weeks, while the incidence of suicidal thoughts was 1.3% in each group. The rates of serious adverse events were similar for placebo and BELVIQ. In BLOOM, the echocardiographic safety endpoint was the proportion of patients developing FDA-defined valvulopathy (mild or greater aortic regurgitation and/or moderate or greater mitral regurgitation) at week 52, and this primarily determined the sample size. A non-inferiority analysis was used to compare the rates of valvulopathy in BELVIQ patients and in placebo patients. Based on a non-inferiority margin equating to a relative risk of valvulopathy with BELVIQ of 1.5, the study was sized to rule out a >50% increase in the incidence of valvulopathy between the groups at 52 weeks, with statistical power of 80% at the 5% significance level. At 52 weeks, FDA-defined valvulopathy had occurred in 2.7% of BELVIQ patients compared with 2.3% of placebo patients (p=0.70, relative risk with BELVIQ: 1.1). At 2 years, the rates for the BELVIQ and placebo groups were 2.6% and 2.7%, respectively. Our consultants believe that BELVIQs clean adverse event profile could be the key driver of its use. For example, other drugs that produce a similar degree of weight loss are sibutramine and orlistat. However, BELVIQ appears to be safer than sibutramine (which was removed from the market for its cardiovascular effects) and better tolerated than orlistat (which can produce urgent bowel movements, oily spotting, and flatulence).

MEMBER: FINRA/SIPC

29

www.cowen.com

Cowen and Company, LLC

EQUITY RESEARCH

BLOOM Trial: Safety Data


BELVIQ 10 mg BID (n=1,593) n(%) 287 (18.0) 235 (14.8) 213 (13.4) 130 (8.2) 119 (7.5) 114 (7.2) 109 (6.8) 106 (6.7) 106 (6.7) 99 (6.2) 95 (6.0) 83 (5.2) 79 (5.0) 73 (4.6) 70 (4.4) 65 (4.1) 59 (3.7) 57 (3.6) 56 (3.5) 53 (3.3) 52 (3.3) 47 (3.0) 41 (2.6) Placebo (n=1,584) n(%) 175 (11.0) 189 (11.9) 190 (12.0) 60 (3.8) 85 (5.4) 130 (8.2) 85 (5.4) 96 (6.1) 64 (4.0) 89 (5.6) 48 (3.0) 37 (2.3) 64 (4.0) 69 (4.4) 75 (4.7) 56 (3.5) 42 (2.7) 43 (2.7) 62 (3.9) 37 (2.3) 49 (3.1) 41 (2.6) 58 (3.7)

Headache Upper respiratory infection Nasopharyngitis Dizziness Nausea Sinusitis Diarrhea Urinary tract infection Constipation Back pain Fatigue Dry mouth Gastroenteritis (viral cause) Influenza Arthralgia Cough Vomiting Pharyngolaryngeal pain Bronchitis Sinus congestion Pain in extremity Procedural pain Insomnia

Source: Cowen and Company, Arena Pharmaceuticals

Trial #2: BLOSSOM


BLOSSOM was a 1-year, randomized, placebo-controlled, double-blind, Phase III study which evaluated the efficacy and safety of BELVIQ, administered with a lifestyle modification program. 4,008 patients were randomized 2:1:2 to receive BELVIQ 10 mg BID, BELVIQ 10 mg daily, or placebo. Eligibility criteria for the trial included age 18-65 years and BMI 30-45 kg/m2 or 27-29.9 kg/m2 having at least one of the following co-morbid conditions: hypertension, dyslipidemia, cardiovascular disease, impaired glucose tolerance, or sleep apnea, as well as ability to participate in a moderate-intensity exercise program. Patients were provided nutritional and physical exercise counseling at baseline, weeks 1/2/4, and subsequently on a monthly basis. Patients were encouraged to moderately exercise for 30 minutes per day and instructed to reduce daily caloric intake by 600 kcal. In September 2009, Arena released top-line results from BELVIQ's BLOSSOM trial. Full data were presented in October 2009 at the Obesity Society Meeting.

MEMBER: FINRA/SIPC

30

www.cowen.com

Cowen and Company, LLC

EQUITY RESEARCH

The trial achieved all three of its co-primary efficacy endpoints (% of patients losing 5% body weight, average weight loss, and % of patients losing 10% of their weight) with p<0.0001. BLOSSOM Trial: Baseline Patient Characteristics

Number Of Patients Average BMI, kg/m2 Average Weight Average Age, years Female Proportion Baseline Valvulopathy
Source: Cowen and Company, Arena Pharmaceuticals

4,008 35.9 220 lbs 44 80% 4-5%

1) % average weight loss: BELVIQ patients (10 mg BID group) had an average weight loss of 5.8% (12.8 lbs) vs. 2.8% (6.4 lbs) for placebo, resulting in a placebo-adjusted average weight loss of 3.0%, or 6.4 lbs. This result was statistically significant; just as in the BLOOM trial though, it did not meet FDAs efficacy benchmark of a 5% difference in average weight loss between therapy and placebo. The 10mg QD group had weaker efficacy, with just a 1.9% difference in placebo-adjusted weight loss. 2) % of patients losing at least 5% of their weight: 47.2% of BELVIQ patients (10 mg BID group) lost at least 5% of baseline body weight, compared to 25% of placebo patients. With these data, BELVIQ met the threshold specified in FDA's draft guidance for efficacy of obesity drugs. The 10mg QD group had weaker efficacy here, too, and the difference between it and placebo was less than the FDA guidance of the active arm being approximately double that of placebo. 3) % of patients losing at least 10% of their weight: 22.6% of BELVIQ patients (10 mg BID group) in the trial lost at least 10% of body weight, compared to 9.7% of placebo patients. Not surprisingly, the 10mg QD group had weaker efficacy here, too. BLOSSOM Trial: Efficacy Data
Key Endpoints n (ITT-LOCF) Mean Weight Loss (%) (ITT-LOCF) 5% weight loss (ITT-LOCF) 10% weight loss (ITT-LOCF) n (Completers) Mean Weight Loss (%) (Completers) 5% weight loss (Completers) 10% weight loss (Completers) Completion rate BELVIQ 10mg BID n=1,602 5.8% 47.2% 22.6% n=846 7.9% 63.2% 35.1% 53% BELVIQ 10mg QD n=801 4.7% 40.2% 17.4% n=418 6.5% 53.1% 26.3% 52% BLOSSOM (n=4,008) Difference from Placebo 10mg BID n=1,601 2.8% 25.0% 9.7% n=764 4.0% 34.9% 16.1% 48% 3.9% 28.3% 19.0% <0.001 <0.001 <0.001 2.5% 18.2% 10.2% <0.001 <0.001 <0.001 3.0% 22.2% 12.9% <0.001 <0.001 <0.001 1.9% 15.2% 7.7% <0.001 <0.001 <0.001 p-value Difference from 10mg QD p-value

Source: Cowen and Company, Arena Pharmaceuticals

MEMBER: FINRA/SIPC

31

www.cowen.com

Cowen and Company, LLC

EQUITY RESEARCH

In addition to the weight loss efficacy, BELVIQ also improved a number of secondary parameters to a greater extent than placebo, including waist circumference (-6.3cm vs. -4.1cm, p<0.001), triglycerides (-4.3% vs. -0.9%, p=0.02), and quality of life (as measured by the IWQOL-LITE scale, 11.8 for BELVIQ vs. 10.0 for placebo, p<0.001). Systolic and diastolic blood pressures, as well as heart rate, decreased from baseline to week 52 in both the BELVIQ and placebo groups, but these differences were not statistically significant. BLOSSOM Trial: Secondary Endpoints
BELVIQ (n = 1,561) -6.3cm -4.3% 11.8 -2.3 bpm -1.9 mm Hg -1.9 mm Hg) Placebo (n = 1,541) -4.1cm -0.9% 10.0 -1.6 bpm -1.2 mm Hg -1.4 mm Hg p-value <0.001 0.02 <0.001 ND NS NS

Change in waist circumference Change in triglycerides Change in IWQOL-LITE Change in heart rate Change Systolic BP (mm Hg) Change Diastolic BP (mm Hg)
ND: not determined; NS: not significant

Source: Cowen and Company, Arena Pharmaceuticals

BELVIQ continues to show clean safety profile in BLOSSOM


Similar to BLOOM, BELVIQ was associated with a favorable safety and tolerability profile in BLOSSOM, with few meaningful imbalances between the arms. Importantly, there was no difference between drug and placebo in rates of depression, suicidal ideation, or FDA-defined valvulopathy. New echocardiographic findings consistent with FDA-defined valvulopathy developed at 52 weeks in 2.0% of BELVIQ BID patients, 1.4% of BELVIQ QD patients, and 2.0% of placebo patients. Among patients who had pre-existing valvulopathy at baseline, the proportion of patients experiencing any increase in mitral or aortic regurgitation was 12.1% in the BELVIQ BID group, 11.1% in the BELVIQ QD group, and 30.6% in the placebo group.

MEMBER: FINRA/SIPC

32

www.cowen.com

Cowen and Company, LLC

EQUITY RESEARCH

BLOSSOM Trial: Safety Data


Adverse Events Patients with any AE Patients with any serious AE Discontinuation due to AE FDA-defined valvulopathy AE with incidence >5% in any group Headache Upper respiratory tract infection Nasopharyngitis Nausea Dizziness Fatigue Sinusitis Urinary tract infection Back pain Diarrhea Dry mouth Constipation Psychiatric AE of interest Depression Depressed mood Suicidal ideation BELVIQ 10 mg BID n(%) 1,323 (82.6) 49 (3.1) 115 (7.2) 24(2.0) 250 (15.6) 204 (12.7) 201 (12.5) 145 (9.1) 140 (8.7) 134 (8.4) 122 (7.6) 107 (6.7) 101 (6.3) 98 (6.1) 87 (5.4) 80 (5.0) 31 (1.9) 10 (0.6) 15 (0.9) BELVIQ 10 mg QD n(%) 653 (81.5) 27 (3.4) 50 (6.2) 9 (1.4) 125 (15.6) 117 (14.6) 95 (11.9) 61 (7.6) 50 (6.2) 53 (6.6) 67 (8.4) 61 (7.6) 55 (6.9) 53 (6.6) 27 (3.4) 41 (5.1) 9 (1.1) 7 (0.9) 5 (0.6) Placebo n(%) 1,205 (75.3) 36 (2.2) 73 (4.6) 23 (2.0) 147 (9.2) 202 (12.6) 192 (12.0) 85 (5.3) 62 (3.9) 66 (4.1) 117 (7.3) 77 (4.8) 91 (5.7) 94 (5.9) 37 (2.3) 61 (3.8) 29 (1.8) 14 (0.9) 11 (0.7)

Source: Cowen and Company, Arena Pharmaceuticals

Trial #3: BLOOM-DM


BLOOM-DM was a 1-year, randomized, placebo-controlled, double-blind, Phase III study which evaluated the efficacy and safety of BELVIQ, administered with a lifestyle modification program, in patients with type 2 diabetes treated with metformin and/or a sulfonylurea (SFU). In the trial, 604 patients were randomized 1:1:1 to receive BELVIQ 10 mg BID, BELVIQ 10 mg daily (QD), or placebo. After approximately 8 months, the protocol was amended, and enrollment was discontinued in the BELVIQ 10 mg QD arm because of slow trial recruitment. Already enrolled patients remained in the study. Eligibility criteria for the trial included age 1865 years and BMI 27-45 kg/m2 with type 2 diabetes treated with metformin a SFU and HbA1c level 7-10% at screening, as well as the ability to participate in a moderate-intensity exercise program. Use of insulin in any form was an exclusion criterion. At each patient visit (weeks 2 and 4 after randomization and then monthly), patients had individual counseling sessions of 15-60 minutes, including advice about exercise, behavior modification, calorie restriction, and food choices. Patients were instructed to moderately exercise for 30 minutes per day and to reduce daily caloric intake by 600 kcal. Arena announced topline results from BLOOM-DM in November 2010. BLOOM-DM met its three co-primary endpoints (% of patients losing 5% body weight, average weight loss, and % of patients losing 10% of their weight) with p < 0.0001. Efficacy was comparable to that in the previous two BELVIQ trials.

MEMBER: FINRA/SIPC

33

www.cowen.com

Cowen and Company, LLC

EQUITY RESEARCH

BLOOM-DM Trial: Baseline Patient Characteristics

Number Of Patients Average BMI, kg/m2 Average Weight Average Age, years Average HbA1c Level Female Proportion
Source: Cowen and Company, Arena Pharmaceuticals

604 36 228 lbs 53 8% 54%

1) % average weight loss: BELVIQ patients had an average weight loss of 4.5% vs. 1.5% for placebo, resulting in a placebo-adjusted average weight loss of 3.0%. This result was statistically significant; once again, as in BELVIQs previous two other Phase III trials, this result did not meet FDAs efficacy benchmark of a 5% difference in average weight loss between therapy and placebo. 2) % of patients losing at least 5% of their weight: 37.5% of BELVIQ patients lost at least 5% of baseline body weight, compared to 16.1% of placebo patients. With these data, BELVIQ met the threshold specified in FDA's draft guidance for efficacy of obesity drugs. 3) % of patients losing at least 10% of their weight: 16.3% of BELVIQ patients in the trial lost at least 10% of body weight, compared to 4.4% of placebo patients. BLOOM-DM Trial: Efficacy Data

BLOOM-DM (n=604) Key Endpoints n (ITT-LOCF) Mean Weight Loss (%) (ITT-LOCF) 5% weight loss (ITT-LOCF) 10% weight loss (ITT-LOCF) n (Completers) Mean Weight Loss (%) (Completers) 5% weight loss (Completers) 10% weight loss (Completers) Completion rate
Source: Cowen and Company, Arena Pharmaceuticals

BELVIQ 10mg BID n=251 4.5% 37.5% 16.3% n=169 5.5% 44.6% 20.8% 67%

Placebo n=248 1.5% 16.1% 4.4% n=157 1.7% 17.9% 5.8% 63%

Difference 3.0% 21.4% 11.9%

p-value <0.001 <0.001 <0.001

3.8% 26.7% 15.0% -

<0.001 <0.001 <0.001 -

Additionally, there were improvements noted in secondary endpoints pertaining to glycemic control, anthropometric measures, and cardiometabolic parameters. Glycemic control was evaluated by HbA1c, fasting glucose, fasting insulin, and calculated insulin resistance using

MEMBER: FINRA/SIPC

34

www.cowen.com

Cowen and Company, LLC

EQUITY RESEARCH

homeostatic model assessment-insulin resistance (HOMA-IR). Mean HbA1c, fasting plasma glucose, and insulin resistance were significantly more decreased in the BELVIQ group compared with placebo. At 52 weeks, 50.4% of BELVIQ patients had HbA1c level 7%, compared with 26.3% of placebo patients. Waist circumference was also reduced significantly with BELVIQ treatment compared to placebo (-5.5 cm vs. -3.3 cm). Systolic and diastolic blood pressures decreased from baseline to week 52 in both the BELVIQ and placebo groups, with no significant differences noted, as was the case with triglyceride levels. Heart rate, however, did decrease significantly with BELVIQ vs. placebo (-2 bpm vs. -0.4 bpm). BLOOM-DM Trial: Secondary Endpoints
BELVIQ 10 mg BID (n = 251) -0.9% 50.4% -27.4 mg/dL -3 IU/ml -0.5 -10.7% -1.3 g/L 11.3 -5.5cm Placebo (n = 248) -0.4% 26.3% -11.9 mg/dL -1.6 IU/ml -0.2 -4.8% -0.6 g/L 10.2 -3.3cm p-value <0.001 <0.001 <0.001 0.12 0.022 0.054 0.15 0.221 0.001

Change in HbA1C % with HbA1C < 7% Change in fasting glucose Change in fasting insulin Change in HOMA-IR Change in triglycerides Change in hs-CRP Change in IWQOL-LITE Change in waist circumference

Source: Cowen and Company, Arena Pharmaceuticals

Safety OK overallbut valvulopathy signal resurfaces


Safety was similar to that observed in the BLOOM and BLOSSOM trials, with headache being the most common unbalanced AE. Hypoglycemia was more frequent with BELVIQ treatment compared with placebo, with 7.4% and 6.3% of patients in each group, respectively, reporting at least one episode of symptomatic hypoglycemia. There were no reports of severe hypoglycemia, which was defined as an episode causing confusion, loss of consciousness, or treatment with intravenous agents, in either group. Also, no patients in either group discontinued because of hypoglycemia.

MEMBER: FINRA/SIPC

35

www.cowen.com

Cowen and Company, LLC

EQUITY RESEARCH

BLOOM-DM Trial: Safety Data

Headache Back pain Nasopharyngitis Nausea Urinary tract infection Cough Symptomatic hypoglycemia Fatigue Gastroenteritis, viral Dizziness Influenza Procedural pain Hypertension Gastroenteritis Depression FDA-defined valvulopathy

BELVIQ 10 mg BID (n=256) n(%) 37 (14.5) 30 (11.7) 29 (11.3) 24 (9.4) 23 (9.0) 21 (8.2) 19 (7.4) 19 (7.4) 18 (7.0) 18 (7.0) 15 (5.9) 13 (5.1) 13 (5.1) 8 (3.1) 6 (2.3) 6(2.9)

Placebo (n=252) n(%) 18 (7.1) 20 (7.9) 25 (9.9) 20 (7.9) 15 (6.0) 11 (4.4) 16 (6.3) 10 (4.0) 11 (4.4) 16 (6.3) 13 (5.2) 5 (2.0) 8 (3.2) 5 (2.0) 5 (2.0) 1 (0.5)

Source: Cowen and Company, Arena Pharmaceuticals

With regard to cardiac valvulopathy, there was a trend toward increased new-onset valvulopathy at 52 weeks in the BELVIQ arm, although the trial was not powered to detect a statistically significant difference. At 52 weeks, one patient (0.5%) in the placebo group and 6 patients (2.9%) in the BELVIQ BID group had echocardiographic FDA-defined valvulopathy which was not present at baseline. The difference in incidence rates between the BELVIQ and placebo groups is higher than differences between the groups in the BLOOM and BLOSSOM studies, but given the small size of the trial, the differences in this trial did not reach statistical significance. Among those with pre-existing FDA-defined valvulopathy at baseline, 11.1% of BELVIQ patients and 12.5% of placebo patients experienced any increase in mitral regurgitation or aortic regurgitation score at week 52. This difference was not statistically significant.

Responder Analysis in BLOOM and BLOSSOM Studies


Updated responder data from a pooled, post hoc analysis of the BLOOM and BLOSSOM studies were presented in June 2013 at the American Diabetes Association meeting. The BLOOM and BLOSSOM study results were combined, representing patients without type 2 diabetes. In the combined analysis, 49% (1,251 of 2,537) of patients treated with BELVIQ were categorized as responders, defined by total body weight loss of at least 5% by week 12. These patients achieved mean weight loss of 10.6 kg (10.7%) in 52 weeks. This amount of

MEMBER: FINRA/SIPC

36

www.cowen.com

Cowen and Company, LLC

EQUITY RESEARCH

weight loss was similar to the weight loss achieved by responders on placebo of 9.5 kg (9.5%). The difference was that only 22.6% of placebo patients (541 of 2,393) met the at least 5% weight loss by week 12 threshold and were categorized as responders, versus the 49.3% of BELVIQ patients meeting that definition. Responder Analysis From the BLOOM and BLOSSOM Studies

Responders Mean baseline weight, kg Mean weight loss, kg Mean weight loss % % responders

BELVIQ 10mg BID (n=1,251) 99.0 10.6 10.7% 49.3%

Placebo (n=541) 100.3 9.5 9.5% 22.6%

Source: Cowen and Company, Arena 2013 ADA Poster

Responder Analysis in BLOOM-DM Study


The June 2013 ADA presentation also had responder analysis from BLOOM-DM, in which patients had type 2 diabetes. In BLOOM-DM, 35.9% (78/217) of patients treated with BELVIQ were categorized as responders, defined by total body weight loss of 5% by week 12. These patients achieved mean weight loss of 9.3 kg (9%) vs. 7.5 kg (7%) for placebo responders. However, only 11.5% of placebo patients met the 5% weight loss by week 12 threshold and were categorized as responders, vs. 35.9% for the patients treated with BELVIQ. Responder Analysis From the BLOOM-DM Study

Responders Mean baseline weight, kg Mean weight loss, kg Mean weight loss % % responders

BELVIQ 10mg BID (n=78) 102 9.3 9.1% 35.9%

Placebo (n=25) 103 7.5 7.3% 11.5%

Source: Cowen and Company, Arena 2013 ADA Poster

Side-Effect Profile Will Carve Out a Niche, But Broader Use Is Likely Only In Combination With Phentermine
Assuming tumorigenesis concerns resulting from preclinical models are viewed by physicians to have no clinical relevance in humans, and that the same goes for the potential valvulopathy signal seen in BLOOM-DM, our consultants believe that BELVIQs otherwise clean side effect profile could carve out a niche in mildly obese people who seek medical treatment. They also suggest that BELVIQs ultimate potential will depend on combinability with phentermine.

MEMBER: FINRA/SIPC

37

www.cowen.com

Cowen and Company, LLC

EQUITY RESEARCH

Consultants are unsure about how quickly a BELVIQ/phentermine combination therapy would be adopted by physicians. Some think that BELVIQ may be combined with phentermine offlabel immediately upon FDA approval. Many others note, however, that most physicians would be hesitant to combine the two because of their experience with Fen-Phen, and that material combination therapy will occur only after a large safety and efficacy trial has been conducted.

Future Studies with BELVIQ


Arena has indicated that, in conjunction with partner Eisai, plans and strategy for further studies of BELVIQ to support additional potential indications are being considered. Discussions with Eisai will be finalized before proceeding to communication with FDA regarding any new potential indication. Management has indicated that potential areas of further investigation include: 1) BELVIQ in combination with phentermine, 2) BELVIQ in combination with olanzapine to address weight gain associated with antipsychotic use, 3) BELVIQ in combination with metformin for diabetic patients, and 4) BELVIQ as a single agent for smoking cessation. BELVIQ-Phentermine Combination: In May 2013, Arena reported that a development plan had been finalized with Eisai for the study of BELVIQ in combination with phentermine. A meeting with the FDA was undertaken in 2Q13 to discuss investigation of this combination. Management commented that the FDA appeared to be amenable to the BELVIQphentermine combination (2Q13 earnings call, 8/1/13). BELVIQ-Phentermine Combination Studies: In August 2013, Arena reported that a PK study evaluating the BELVIQ-phentermine combination was completed, with data being analyzed. This study evaluated BELVIQ 10 mg, phentermine 15 mg, and the BELVIQ-phentermine combination (10 mg/15 mg). Additionally, Arena announced that Eisai will be initiating a 12week pilot study of BELVIQ-phentermine by YE13 or the beginning of 2014. Eisai has not disclosed the study design. Management indicated that the results of both studies will inform the future development plan for the combination. Arena is covering costs of the PK study, and Eisai will cover costs related to the pilot study. The further co-development agreement for the combination is under discussion with Eisai, according to Arena management

MEMBER: FINRA/SIPC

38

www.cowen.com

Cowen and Company, LLC

EQUITY RESEARCH

Summary of BELVIQs Regulatory History


NDA Submitted in December 2009; FDA AdCom in September 2010
In December 2009, Arena submitted BELVIQs NDA to the FDA. In September 2010, the FDA AdCom voted 9-5 against recommending approval. The panel concluded that there was too much uncertainty about BELVIQ's risks, in light of its relatively modest effects on weight loss, to support approval at that time. Committee members were generally comfortable that BELVIQ's efficacy, though modest, met the FDA bar and might be clinically significant. However, the limited efficacy made the panel more critical of risks. Panelists had two primary concerns: 1) an unexplained preclinical carcinogenicity signal, and 2) the patient population used in the pivotal trials was not seen as representative of the presumed real-world patient group. Although they believed Arena made all reasonable efforts to rule out drug-induced valvulopathy, they suggested that programs to monitor this issue should be a condition of any approval.

Rat Carcinogenicity Data a Surprise in Briefing Documents, Made Risk/Benefit Unacceptable To AdCom
In FDA briefing documents prepared for the September 2010 BELVIQ AdCom review, it was revealed that a number of malignant tumors had developed in BELVIQs rat carcinogenicity studies, including tumors of breast, brain, peripheral nerve, skin and subcutis. However, only mammary tumors were increased to a statistically significant degree in both male and female rats, and therefore these were the most concerning to committee members. Moreover, the mammary tumors developed in female rats treated with BELVIQ at doses within sevenfold of the proposed clinical dose of 10mg BID, which was low enough that the FDA was concerned about the margin for safety. The FDAs concerns were further supported by the fact that preclinical studies had not produced similar tumors in mice. The agency suggested that this was likely due to lower levels of drug exposure achieved in mice. Moreover, the FDA did not accept Arenas studies linking increased tumorigenesis to elevated serum prolactin levels, which would suggest little clinical significance in humans. The FDA concluded that given the lack of a safety margin, an unresolved tumorigenic mechanism of action, and a patient population already at increased risk of breast cancer, the relevance of these findings in rats to human risk cannot be dismissed.

MEMBER: FINRA/SIPC

39

www.cowen.com

Cowen and Company, LLC

EQUITY RESEARCH

Incidence of BELVIQ-Induced Tumors In Two-Year Rat Carcinogenicity Study


Male rats Incidence of tumors Brain Mammary Skin, subcutis Skin Nerve Sheath Liver Thyroid Female rats Incidence of tumors Brain Mammary Control 1 0 0 0 3 0 0 1 1 2 0 Control 0 28 20 40 BELVIQ dose, mg/kg/day 10 30 0 4 NS 0 2 1 4 NS 1 6 SS 7 NS 11 SS 0 4 NS 0 2 NS 3 2 1 2 4 4 NS 5 4 NS BELVIQ dose, mg/kg/day 10 30 2 0 34 NS 35 NS 47 SS 53 SS 56 SS 61 SS 100 8 SS 2 NS 6 NS 8 SS 17 SS 5 SS 9 SS 4 6 SS 10 SS 8 SS 100 1 60 SS 45 SS 70 SS

astrocytoma adenocarcinoma fibroadenoma combined benign fibroma squamous carcinoma Schwannoma, all sites hepatocellular carcinoma hepatocellular adenoma combined follicular cell adenoma

astrocytoma adenocarcinoma fibroadenoma combined

SS: Statistically significant, NS: Not statistically significant

Source: Cowen and Company, FDAs BELVIQ briefing documents

CRL Received In October 2010


In October 2010, Arena received a CRL for BELVIQ. The letter requested that Arena complete and submit BELVIQ's BLOOM-DM trial in obese patients with diabetes, as well as additional preclinical rat toxicology studies to better understand BELVIQ's association with mammary and central nervous system (CNS) tumors. Specifically, regarding the rat tumor findings, the FDA requested additional preclinical data and analyses, including a detailed accounting of the female rat slides that contributed to the rat mammary tumor incidence data, a re-adjudication of all mammary and lung tissues from all female rats, a demonstration that the adenocarcinoma findings have no relevance to humans, and additional data on the distribution of BELVIQ to the CNS to better estimate the astrocytoma exposure margins.

Report Shows a Shift in Incidence Numbers of BELVIQ-Induced Tumors


To address the rat tumor findings the company convened a Pathology Working Group (PWG) consisting of 5 pathologists to re-adjudicate the mammary tumors. In August 2011, Arena announced the results from the PWG re-adjudication. The data showed a shift in the numbers of both tumor types, benign and malignant, relative to the initial results included in the NDA. The new analysis showed that the incidences of adenocarcinomas in the BELVIQ low- and mid-dose groups were no longer numerically higher than those in the control group. However, the incidence was statistically higher in the BELVIQ high (100mg/kg/day) dose group. Incidence of fibroadenomas was statistically higher than the control group for all three BELVIQ dose groups.

MEMBER: FINRA/SIPC

40

www.cowen.com

Cowen and Company, LLC

EQUITY RESEARCH

PWG Reported Incidence of BELVIQ-Induced Tumors


Percent of Female Rats with Mammary Adenocarcinoma or Fibroadenoma Dose Control 10 mg/kg/day 30 mg/kg/day 100 mg/kg/day N 65 65 65 75 Mammary Adenocarcinoma (Malignant) Initial Report 43.1% 52.3% 53.9% 80.0% PWG Report 40.0% 32.3% 36.9% 68.0% Mammary Fibroadenoma (Benign) Initial Report 30.8% 72.3% 81.5% 60.0% PWG Report 36.9% 83.1% 84.6% 68.0%
Source: Cowen and Company, Arena Pharmaceuticals

In August 2011, Arena also completed a clinical study designed to measure the concentration of BELVIQ in human cerebrospinal fluid (CSF) and plasma. The study enrolled nine obese but otherwise healthy volunteers who were administered BELVIQ 10 mg BID for seven days. On day seven, lumbar CSF and plasma were serially collected simultaneously over a 12-hour period. Measured human CSF and plasma exposures (mean AUC (standard deviation)) of 9.3 (+/-3.9) hr-ng/mL and 540 (+/-157) hr-ng/mL, respectively, were observed in the study. Arena also calculated the estimated ratio of BELVIQ exposure in the brain relative to plasma in humans using the mean brain: CSF exposure ratio of 101 from preclinical studies, with a range of 75-117. Based on these results, the company estimated that humans concentrate BELVIQ in the brain 14 times less compared to rats, and that plasma concentrations of BELVIQ in the rat carcinogenicity study at a dose with no astrocytoma were five times the anticipated human clinical exposure, providing a higher safety margin.

Arena Resubmitted the NDA in January 2012


In January 2012, Arena submitted its response to the CRL, including data from the Phase III BLOOM-DM trial, which had not been included in the original NDA. Arenas response also included data and analyses from studies that addressed tumors observed in a two-year BELVIQ rat carcinogenicity study, cell culture experiments intended to further refine serotonin subtype 2 receptor activity, and rat studies designed to further assess abuse potential.

Positive FDA AdCom in May 2012 and U.S. Approval in June 2012
In May 2012, the FDA AdCom voted 18-4 in favor of BELVIQs approval. In June 2012, the FDA approved BELVIQ for the treatment of obesity. As per the label, BELVIQ is approved for the treatment of adults with an initial BMI 30 kg/m2 (obese) or 27 kg/m2 (overweight) in the presence of at least one weight-related comorbid condition.

U.S. Post-marketing requirements, CVOT to be completed by YE 2017


As part of BELVIQs approval, the FDA has requested that Arena c onduct six post-marketing studies. The first study is an animal study with juvenile rats, which will be completed prior to initiation of the other studies. Arena will then be conducting a series of studies to evaluate the

MEMBER: FINRA/SIPC

41

www.cowen.com

Cowen and Company, LLC

EQUITY RESEARCH

safety and efficacy of BELVIQ in the obese pediatric patient population. In addition, Arena is also required to conduct a cardiovascular outcomes trial (CVOT) to evaluate the effect of longterm treatment with BELVIQ on CV safety. Echocardiographic assessments will be included in this post-marketing CVOT. Arena is currently in discussions with FDA, is finalizing the protocols for these studies, and expects to initiate these studies sometime in 2013. Arena and Eisai have agreed on a protocol for the CVOT, and a draft was provided to FDA in February 2013. The company is required to complete the CVOT by December 31, 2017 and to submit the final report in 2018. As per the agreement with Eisai, Arena will be responsible for 10% of the CVOT expenses. Furthermore, Arena will be responsible for 50% of certain pediatric development expenses.

E.U. Regulatory Front: MAA Withdrawn


The BELVIQ MAA was submitted to EMA in March 2012, with the UK as the rapporteur and Sweden as the co-rapporteur, and in July 2012, Arena also submitted the MAA in Switzerland. In 4Q12, Arena submitted its response to the CHMPs 120-day assessment report and list of questions regarding the MAA. In January 2013, the company received the Day 180 List of Outstanding Issues from the CHMP. The CHMP requested that the company justify BELVIQs overall benefit-risk profile, taking into account previously raised issues, including tumors in rats, valvulopathy, and psychiatric events. The CHMP requested both oral and written explanations from the company. On May 2, 2013, Arena reported that, after the company submitted its written response as well as oral explanation to the Day 180 List of Outstanding Issues, the CHMP continued to have certain major objections which prevented a recommendation to approve the MAA. The company indicated that the objections were related to results of non-clinical studies and that these could not be resolved before the time of the final CHMP opinion. Therefore, Arena withdrew the BELVIQ MAA for the E.U. and will ask EMA for scientific advice with regard to submission at a later time.

Regulatory Update in Other Territories


Mexico: In April 2013, Arena announced that a marketing authorization application had been submitted in Mexico with the Federal Commission for the Protection Against Sanitary Risk (COFEPRIS). This submission triggered a $500K milestone from Eisai to Arena. Canada: In June 2013, Arena announced that a New Drug Submission (NDS) for BELVIQ had been submitted in Canada, which triggered a $500K milestone from Eisai to Arena. Brazil: In August 2013, Arena announced that regulatory filings will be submitted for BELVIQ around YE13. Switzerland: In August 2013, Arena reported that Swissmedic was not satisfied by the companys Day 120 response to the agencys questions and concerns.

MEMBER: FINRA/SIPC

42

www.cowen.com

Cowen and Company, LLC

EQUITY RESEARCH

BELVIQ Has IP Through mid-2023, Possibly Through mid-2026


As of February 15, 2013, Arena owned issued patents that cover compositions of matter for BELVIQ and related compounds and methods of treatment utilizing BELVIQ and related compounds in 69 jurisdictions, including the United States, Japan, Germany, France, China, Italy, Spain, Canada, the United Kingdom, Russia, Australia, India, and South Korea. Arena holds U.S. Patent # 6,953,787, entitled 5HT2C Receptor Modulators. As stated in the patent abstract, The present invention relates to novel compounds of Formula (I): which act as 5HT2C receptor modulators. These compounds are useful in pharmaceutical compositions whose use includes the treatment of obesity. Arena also holds U.S. Patents # 7,514,422; 7,977,329; 8,207,158; and 8,273,734, which are continuations of U.S. Patent # 6,953,787. These patents cover the composition of matter of lorcaserin for the treatment of obesity, providing protection through mid-2023. Arena has filed an application with FDA for a patent extension under Hatch-Waxman, and believes it should be able to receive an additional 3 years of patent life, resulting in patent protection through mid-2026. BELVIQ Patent Estate
Paten t # 6 ,9 5 3 ,7 8 7 7 ,5 1 4 ,4 2 2 T h e p r e s e n t in v e n tio n r e la te s to n o v e l c o m p o u n d s o f F o r m u la ( I) : w h ic h a c t a s 5 H T 2 C u s e in c lu d e s th e tr e a tm e n t o f o b e s ity T it le F ilin g d a t e 4 /1 0 /2 0 0 3 8 /1 3 /2 0 0 4 E a rlie s t p rio rit y d ate 4 /1 2 /2 0 0 2 4 /1 2 /2 0 0 2 4 /1 2 /2 0 0 2 4 /1 2 /2 0 0 2 4 /1 2 /2 0 0 2 m id - 2 0 2 3 E x p ira t io n d a t e

7 ,9 7 7 ,3 2 9 r e c e p to r m o d u la to r s . T h e s e c o m p o u n d s a r e u s e fu l in p h a r m a c e u tic a l c o m p o s itio n s w h o s e 1 1 /1 4 /2 0 0 6 8 ,2 0 7 ,1 5 8 8 ,2 7 3 ,7 3 4 T h e p r e s e n t in v e n tio n p r o v id e s p r o c e s s e s a n d in te r m e d ia te s fo r th e p r e p a r a tio n o f 3 8 ,3 6 7 ,6 5 7 b e n z a z e p in e s a n d s a lts th e r e o f w h ic h c a n b e u s e fu l a s s e r o to n in ( 5 - H T ) r e c e p to r a g o n is ts fo r th e tr e a tm e n t o f, fo r e x a m p le , c e n tr a l n e r v o u s s y s te m d is o r d e r s s u c h a s o b e s ity . T h e p r e s e n t in v e n tio n is d ir e c te d to c r y s ta llin e fo r m s o f ( R ) - 8 - c h lo r o - 1 - m e th y l- 2 ,3 ,4 ,5 8 ,1 6 8 ,6 2 4 te tr a h y d r o - 1 H - 3 - b e n z a z e p in e , c o m p o s itio n s c o n ta in in g th e s a m e , p r e p a r a tio n s , a n d u s e s th e r e o f. T h e p r e s e n t in v e n tio n r e la te s to a c o m p o s itio n c o m p r is in g p h e n te r m in e a n d a s e le c tiv e 5 H T - 2 C r e c e p to r a g o n is t. In a d d itio n , th e in v e n tio n r e la te s to a c o m p o s itio n c o m p r is in g 8 ,1 5 3 ,6 2 1 p h e n te r m in e a n d a s e le c tiv e 5 H T - 2 C r e c e p to r a g o n is t h a v in g F o r m u la ( I) : o r a p h a r m a c e u tic a lly a c c e p ta b le s a lt, s o lv a te o r h y d r a te th e r e o f. T h e s e c o m p o s itio n s a r e u s e fu l in p h a r m a c e u tic a l c o m p o s itio n s w h o s e u s e in c lu d e s th e tr e a tm e n t o f o b e s ity . 1 2 /2 1 /2 0 0 5 1 2 /2 0 /2 0 0 5 6 /1 4 /2 0 0 4 5 /2 7 /2 0 1 1 6 /1 4 /2 0 1 2

7 /1 7 /2 0 0 3

1 2 /2 1 /2 0 0 4

1 2 /2 3 /2 0 0 4

Source: Cowen and Company, USPTO

MEMBER: FINRA/SIPC

43

www.cowen.com

Cowen and Company, LLC

EQUITY RESEARCH

Obesity: A Brief Overview


Obesity is a chronic condition affecting approximately one-third of the United States population, and its prevalence has doubled over the past two decades. Obesity is a primary risk factor for the development of type 2 diabetes, hyperlipidemia, hypertension, and other cardiovascular disorders, and has been identified as the second most common factor contributing to preventable death behind tobacco use. Increased body weight also plays a role in the development of gallbladder disease, degenerative joint disease, respiratory disorders, and certain types of cancer. As a result, the economic burden of the condition is substantial. Annual healthcare costs associated with obesity have been estimated to be $190 billion. According to CDC data from NHANES surveys (Health, United States 2012, National Center for Health Statistics 2013), among adults aged 20-74 years, the prevalence of obesity, defined as a body mass index (BMI) of 30 kg/m2, has increased from 15% (in the 1976-1980 survey) to 35.7% (in the 2009-2010 survey), a figure equating to over 71 million people. Even more troubling is the fact that, in 2007-2008, the prevalence of overweight and obesity, defined as BMI of 25 kg/m2, was 68% in the adult population. The annual number of obesity-related deaths is estimated by the CDC to surpass 360,000. In 2004, the CDC declared obesity to be the number-one health threat in the U.S. Extrapolating NHANES data, by 2020, 40% of men and 43% of women are projected to be obese. The recent and rapid rise in obesity rates is generally attributed to diet and lifestyle shifts that have occurred since the mid-20th century. Body Mass Index (BMI)

Underweight Healthy weight Overweight Mild Obesity Moderate Obesity Severe Obesity

BMI (kg/m2) Below 18.5 18.5 to 24.9 25.0 to 29.9 30 to 34.9 35 to 39.9 40 or higher

Obesity Class

I II III

Source: Cowen and Company, Centers for Disease Control

MEMBER: FINRA/SIPC

44

www.cowen.com

Cowen and Company, LLC

EQUITY RESEARCH

U.S. Prevalence Of Obesity

Source: Centers for Disease Control

Because of the diverse mechanisms believed to underlie weight control, a broad array of approaches has been attempted to address this large and growing healthcare issue. Diet, increased physical activity, and behavioral modification are the mainstay strategies for weight loss and maintenance of healthy body weight. While such lifestyle changes can be effective in reducing body fat and increasing lean body mass, and are the typical prescription for most overweight patients, compliance with lower-calorie diets and/or exercise regimens is challenging for most. According to NIH Guidelines on obesity treatment, pharmacotherapy may provide a beneficial adjunct to dietary, activity, and behavioral modifications in selected individuals, including: those with BMI 30 kg/m2 (obese) and no concomitant risk factors, or those with BMI 27 kg/m2 (overweight) and comorbidities such as hypertension, dyslipidemia, type 2 diabetes, coronary artery disease, and sleep apnea. Per NIH recommendations, the initial goal of weight loss therapy is an approximately 10% reduction in body weight from baseline, accomplished within a time frame of 6 months on treatment. The growing view of obesity as a chronic disease that requires targeted treatment has spurred industry-wide efforts to develop novel therapeutics. During the 60-year history of obesity drug marketing, one of the biggest challenges has been to develop a drug that is: (1) effective in promoting durable weight loss, (2) safe for long-term use, and (3) non habit-forming. Until 1996, all drugs approved for weight-loss indications by the FDA were labeled for short-term use only (a few weeks). This was partly because obesity was not considered a chronic condition, so the drugs were viewed as short-term jumpstarts to longer programs of diet and exercise. Also, most of the early drugs were amphetamine derivatives with significant abuse potential. As obesity became more prevalent during the 1980s, the FDA began to recognize it as a chronic condition requiring longer-term treatment. In 1996, the agency issued a draft

MEMBER: FINRA/SIPC

45

www.cowen.com

Cowen and Company, LLC

EQUITY RESEARCH

guidance requiring extensive safety data for one year or longer for new obesity drug applications. Long-term safety is an especially crucial issue for contemporary obesity NDAs since, despite the more stringent 1996 draft guidance, three out of the four obesity drugs subsequently approved were found to be associated with serious safety issues post approval and ultimately withdrawn from one or more major markets. Poor Safety Record of Obesity Drugs

Drug Desoxyn (methamphetamine) Preludin (phenmetrazine) phentermine Tenuate (diethylpropion) Bontril (phendimetrazine) Didrex (benzphetamine) Mazindol Fenfluramine Redux (dexfenfluramine) Meridia (sibutramine) Xenical (orlistat) Acomplia (rimonabant)
Source: Cowen and Company

Year Approved Comment Approved Duration Of Use 1947 Short-term Black box warning for abuse potential 1956 Short-term Banned in Sweden and rarely prescribed elsewhere due to abuse 1959 Short-term Label warns against combination use 1959 Short-term Label contraindicates combination use 1959 Short-term Rarely prescribed due to abuse potential 1960 Short-term Label warns against combination use 1973 Short-term 1973 Short-term Withdrawn in 1997 on valvulopathy risk 1996 Long-term Withdrawn in 1997 on valvulopathy risk 1997 Long-term Withdrawn in 2010 on CV risk 1999 Long-term 2006 (EU) Long-term Withdrawn from EU in 2009 on suicidality risk

In 2007, the FDA revised the weight-loss drug draft guidance. The guidance now requires Phase III placebo-controlled trials, with approximately 1,500 patients on placebo and 3,000 patients on drug for a period of at least one year to assess safety. The 2007 guidance defined two efficacy endpoints, at least one of which should be met for a new drug candidate to be considered effective: (1) mean efficacy (statistically significant difference in mean weight loss of at least 5% between therapy and placebo groups), and (2) categorical efficacy (the proportion of subjects losing 5% of body weight in the active drug group should be greater than 35%, and at least double the proportion in the placebo group, with the difference between groups being statistically significant). Until recently, only a handful of FDA-approved prescription products were marketed, including generic phentermine and Roche/GlaxoSmithKlines Xenical (orlistat). Our clinical consultants use these older medications in about 50% of their obesity patients, primarily because they are only modestly efficacious (resulting in loss of 8-10 pounds, or approximately 5% of baseline body weight) while also associated with poor safety and tolerability profiles. Because phentermine and Xenical have questionable benefit/risk profiles and are not generally reimbursed by payors, only 6MM of 160MM obese patients worldwide are treated. Of those, only 50% are thought to be compliant with therapy. Although bariatric surgery and other invasive procedures have gained wider acceptance and can result in significant weight loss, these options remain reserved for those patients at the severe end of the obesity spectrum. Hence, the overwhelming majority of obese patients are inadequately treated for their

MEMBER: FINRA/SIPC

46

www.cowen.com

Cowen and Company, LLC

EQUITY RESEARCH

condition. In this context, and with a difficult market history of weight-loss drugs stymied by major safety issues, the FDA approved BELVIQ and Qsymia in 2012, after previously rejecting both. Current thinking on obesity reflects understanding that the affected population is not necessarily homogenous. This perspective has been shared in a recent consensus report of the Obesity Drugs Outcome Measures Dialogue Group, composed of academic and clinical key opinion leaders, health advocacy leaders, industry representatives, and government (FDA, CDC, NIH) observers. Beyond characterization based primarily on size (BMI), there is recognition of an obesity spectrum, accounting for variations in health effects, functional limitations, and psychosocial impairment. Patient stratification may be considered in three groups: Obese and Well, Obese with Risk Factors, and Obese and Sick. Obese-Well individuals have no related risks and no significant impairments. The Obese-Sick have obesityrelated comorbidities, with impairments of function or daily feeling. In the middle are those who have no current comorbidities but carry specific risk factors for health and symptomatic effects. Considering disease severity across a spectrum of obesity may help to clarify patient needs, therapeutic goals, and acceptable risk levelsstrategic factors which ultimately impact treatment decisions.

Recent Developments in Obesity: Recognition and Legislation


In a September 2012 position statement (Mechanick JI et al., Endocrine Practice 2012), the American Association of Clinical Endocrinologists (AACE) declared obesity to be a disease with multiple pathophysiological aspects, including genetic, environmental, physiological, and psychological factors. The AACE published a new, comprehensive management algorithm for diabetes (Garber AJ et al., AACE Comprehensive Diabetes Management Algorithm 2013. Endocrine Practice 2013) which incorporates obesity, pre-diabetes, and CV risk factor management in May 2013. In particular, these guidelines specifically incorporate medical therapy with approved weight loss drugs, including BELVIQ and Qsymia, first-line with lifestyle modification for treatment of obese and overweight patients with complications such as diabetes. These therapies are also considered as first-line treatment in the algorithm for prediabetes management. As stated in the AACE Guidelines, lifestyle optimization is essentialHowever, such efforts should not delay pharmacotherapy, which can be initiated simultaneously. In June 2013, the American Medical Association adopted a resolution introduced by the AACE, and formally recognized obesity as a disease, with multiple pathophysiological aspects requiring a range of interventions to advance obesity treatment and prevention. Also in June 2013, a bipartisan group in the U.S. Congress introduced legislation entitled the Treat and Reduce Obesity Act of 2013 (H.R. 2415) for consideration in the House of Representatives. This bill amends title XVIII (Medicare) of the Social Security Act to include information on coverage of intensive behavioral therapy for obesity in patient handbooks. Furthermore, the legislation provides reimbursement for this behavioral therapy. Finally, under the bill, there is coverage with Medicare Part D (voluntary prescription benefit program) for obesity drugs used to treat overweight individuals with one or more comorbidities.

MEMBER: FINRA/SIPC

47

www.cowen.com

Cowen and Company, LLC

EQUITY RESEARCH

BELVIQs Competition, Current and Future


1) The Incumbent: Vivus Qsymia
In July 2012, the FDA approved Qsymia, a controlled-release phentermine/ topiramate combination, as an adjunct to a reduced-calorie diet and increased physical activity for chronic weight management in adult patients with an initial BMI 30 kg/m 2 (obese), or 27 kg/m2 (overweight) in the presence of at least one weight-related comorbidity, such as hypertension, type 2 diabetes mellitus, or dyslipidemia. Per the label, Qsymia treatment is to be started at low-dose (3.75 mg/23 mg phentermine/topiramate) daily for 14 days and then increased to the recommended dose of 7.5 mg/46 mg (mid-dose). After 12 weeks on Qsymia mid-dose treatment, if a patient has not lost at least 3% of baseline body weight, treatment should be discontinued or escalated to Qsymia top-dose (15mg/92mg). After an additional 12 weeks on Qsymia top-dose treatment, if a patient has not lost at least 5% of baseline body weight, treatment should be discontinued. In September 2012, Qsymia was launched in the U.S. In February 2013, the EMA confirmed its original decision of 10/18/12 to decline approval for the MAA for Qsiva, the EU proposed name. EMA indicated that a preapproval CVOT would be necessary to establish Qsivas long term safety.

Qsymia REMS Requirement


In granting approval for Qsymia, the FDA required a Risk Evaluation and Mitigation Strategy (REMS) from Vivus, with the goal of informing prescribers and female patients of childbearing age about potential teratogenic risks of Qsymia and the importance of pregnancy prevention, as well as need to discontinue Qsymia in the event of pregnancy. The components of the Qsymia REMS include a Medication Guide with patient-focused labeling, dispensed with every prescription, as well as Elements to Assure Safe Use (ETASU), with assessments required at 6 months, 12 months, and annually thereafter. The ETASU are comprised by: 1) HCP education and training, with maintenance of database of trained HCPs by Vivus, and 2) certified home delivery pharmacy network, with certified pharmacies and pharmacy staff training on REMS requirements. The ETASU for Qsymia does not include a patient registry, patient enrollment, patient profile review, or pregnancy test requirement, though a test is recommended. REMS amendment approved: In April 2013, Vivus announced the approval by FDA of an amendment and modification to the REMS program, which made Qsymia available through certified retail pharmacies. The other components of the original Qsymia REMS program remain unchanged and in place. Management indicated that the implementation process includes completing wholesale distribution agreements, enrolling and training each pharmacy location, and ensuring REMS-compliant database setup, as well as shipping and stocking of product. On July 1, 2013, Vivus announced the availability of Qsymia in certified retail pharmacies (approximately 8,000 Walgreens, Costco, and Duane Reade retail locations).

MEMBER: FINRA/SIPC

48

www.cowen.com

Cowen and Company, LLC

EQUITY RESEARCH

Qsymia Achieved Strong Efficacy in Its Phase III Program


The Qsymia Phase III program was comprised by four trials: EQUATE, a 28-week factorial study conducted to confirm the contributions of phentermine and topiramate to Qsymia efficacy; EQUIP and CONQUER, two 1-year studies; and SEQUEL, a 1-year extension study of CONQUER. EQUATE and CONQUER were conducted under SPA with FDA. Designs of the two pivotal phase III studies, EQUIP and CONQUER, were determined in consultation with FDA and complied with the FDA Guidance for Industry Developing Products for Weight Management relative to number of participants, enrolled populations, study duration, and endpoints. The Qsymia Phase III trial program evaluated three dose levels (phentermine/topiramate): low-dose (3.75 mg/23 mg), mid/recommended-dose (7.5 mg/46 mg), and top-dose (15 mg/92 mg). Qsymia Phase III Program Summary

EQUATE (OB-301)*
Number of patients Duration 756 28 weeks Qsymia 7.5/46 mg, Qsymia 15/92 mg, Topiramate 46 mg, Topiramate 92 mg, Phentermine 7.5 mg, Phentermine 15 mg, Placebo BMI 30-45; baseline weight 223 pounds 36 46 79% Caucasian (79%)

EQUIP (OB-302)
1,267 56 weeks Qsymia 3.75/23 mg, Qsymia 15/92 mg, Placebo

CONQUER (OB-303)
2,487 56 weeks Qsymia 7.5/46 mg, Qsymia 15/92 mg, Placebo

SEQUEL (OB-305)*
675 108 weeks (extension study) Qsymia 7.5/46 mg, Qsymia 15/92 mg, Placebo

Treatment groups

Patient demographics (BMI, kg/m2) Mean BMI, kg/m2 Average age (years) Female Proportion Ethnicity Adjunct lifestyle/behavior modification program Co-primary endpoints Data Announced

BMI 27-45 and 2 comorbidities BMI 27-45 and 2 comorbidities BMI 35, with no upper limit; (HTN, dyslipidemia, diabetes, (HTN, dyslipidemia, diabetes, baseline weight 256 pounds abdominal obesity; baseline weight abdominal obesity; baseline weight 225 pounds 227 pounds 42 37 36 43 51 51-52 83% 70% 65%-70% Caucasian (80%) Caucasian (70%) Caucasian (83%-87%) African-American (16%-18%) African-American (11%-12%) African-American (11%-15%)

Standardized counseling on nutrition and exercise based on the LEARN weight management program, with monthly progress discussion 1) mean change in body weight from baseline 2) proportion of patients 5% weight loss December 2008 September 2009 September 2009 September 2010

Qsymia Low-Dose: 3.75/23mg (phentermine/topiramate) ; Qsymia Mid-Dose: 7.5/46mg; Qsymia High-Dose: 15/92mg

Source: Cowen and Company, *EQUATE and SEQUEL were not pivotal studies

MEMBER: FINRA/SIPC

49

www.cowen.com

Cowen and Company, LLC

EQUITY RESEARCH

Qsymia Efficacy Data


M e a n W e ig h t L o s s (% ) Tre a tm e n t G ro u p M e a n W e ig h t L o s s (% ) P la c e b o G ro u p M ean W eig h t L o s s (% ) P laceb o a d j u s te d > 5 % W e ig h t L o s s Tre a tm e n t G ro u p > 5 % W e ig h t L o s s P la c e b o G ro u p > 5% W eig h t L o ss P laceb o A d j u s te d > 1 0 % W e ig h t Lo s s Tre a tm e n t G ro u p > 1 0 % W e ig h t Lo s s P la c e b o G ro u p

Q s y m ia 1 5 m g p h e n /9 2 m g to p i (IT T )
E Q U A TE (O B -3 0 1 ) Q s y m ia 1 5 m g p h e n / 9 2 m g to p i; (n = 7 5 6 ) E Q U IP (O B -3 0 2 ) Q s y m ia 1 5 m g p h e n / 9 2 m g to p i; (n = 1 , 2 6 7 ) S E Q U E L (O B -3 0 5 )Q s y m ia 1 5 m g p h e n / 9 2 m g to p i; (n = 2 , 4 8 7 ) C O N Q U E R (O B -3 0 3 ) Q s y m ia 1 5 m g p h e n / 9 2 m g to p i; (n = 6 7 5 ) 9.2% 1.7% 7.5% 66.0% 15.5% 50.5% 38.8% 6.8%

10.9%

1.6%

9.3%

66.7%

17.3%

49.4%

47.2%

7.4%

10.5%

1.8%

8.7%

79.3%

30.0%

49.3%

53.9%

11.5%

10.4%

1.8%

8.6%

70.0%

21.0%

49.0%

48.0%

7.0%

Q s y m ia 7 .5 m g p h e n /4 6 m g to p i (IT T )
E Q U A TE (O B -3 0 1 )Q s y m ia 7 . 5 m g p h e n / 4 6 m g to p ; (n = 7 5 6 ) S E Q U E L (O B -3 0 5 )Q s y m ia 7 . 5 m g p h e n / 4 6 m g to p ; (n = 2 , 4 8 7 ) C O N Q U E R (O B -3 0 3 ) Q s y m ia 7 . 5 m g p h e n / 4 6 m g to p ; (n = 6 7 5 ) 8.5% 1.7% 6.8% 62.1% 15.5% 46.6% 40.8% 6.8%

9.3%

1.8%

7.5%

75.2%

30.0%

45.2%

50.3%

11.5%

8.4%

1.8%

6.6%

62.0%

21.0%

41.0%

37.0%

7.0%

Q s y m ia 1 5 m g p h e n /9 2 m g to p i (c o m p le te r s )
E Q U IP (O B -3 0 2 ) Q s y m ia 1 5 m g p h e n / 9 2 m g to p i; (n = 1 , 2 6 7 ) S E Q U E L (O B -3 0 5 )Q s y m ia 1 5 m g p h e n / 9 2 m g to p i; (n = 2 , 4 8 7 ) C O N Q U E R (O B -3 0 3 ) Q s y m ia 1 5 m g p h e n / 9 2 m g to p i; (n = 6 7 5 ) 14.4% 2.1% 12.2% 83.5% 25.5% 58.0% 67.7% 13.0%

10.7%

2.2%

8.5%

79.3%

30.0%

49.3%

53.9%

11.5%

12.4%

1.6%

10.8%

85.1%

26.2%

58.9%

64.3%

9.7%

7 .5 m g p h e n /4 6 m g to p i (c o m p le te r s )
S E Q U E L (O B -3 0 5 )Q s y m ia 1 5 m g p h e n / 9 2 m g to p i; (n = 2 , 4 8 7 ) C O N Q U E R (O B -3 0 3 ) Q s y m ia 1 5 m g p h e n / 9 2 m g to p i; (n = 6 7 5 ) 9.3% 2.2% 7.1% 75.2% 30.0% 45.2% 50.3% 11.5%

9.6%

1.6%

8.0%

74.6%

26.2%

48.4%

49.1%

9.7%

Q s ym ia L o w - D o s e : 3 .7 5 / 2 3 m g ( p h e n t e r m in e / t o p ir a m a t e ) ; Q s ym ia M id - D o s e : 7 .5 / 4 6 m g ; Q s ym ia T o p - D o s e : 15 / 9 2 m g

Source: Cowen and Company

EQUATE Confirms Efficacy of Qsymia Over Phentermine and Topiramate Alone


EQUATE (OB-301) was a randomized, placebo-controlled, double-blind, multi-center, sixmonth, Phase III factorial trial in 756 obese patients comparing weight loss with two doses of Qsymia (mid-dose, 7.5/46 mg and top-dose, 15/92 mg), single agent phentermine and topiramate at doses corresponding to those in the Qsymia dose combinations, and placebo. Eligibility criteria included age 70 years and BMI 30-45 kg/m2. All participants were provided standardized counseling based on the LEARN weight management program. LEARN is a 16week program incorporating tools for lifestyle, attitude, relationship, nutrition, and exercise changes.

MEMBER: FINRA/SIPC

50

www.cowen.com

Cowen and Company, LLC

EQUITY RESEARCH

Baseline Demographics in EQUATE


Number of Patients Average BMI, Kg/m Average Weight Average Age, years Female Proportion Mean Systolic BP, mm Hg Mean Diastolic BP, mm Hg
2

756 36.2 223 lbs 45.6 79% 122 79

Source: Cowen and Company, Vivus, FDA briefing document

This seven-arm study was conducted to demonstrate the contributions of each drug component to the Qsymia weight loss effect, and to confirm the efficacy of Qsymia over phentermine and topiramate alone. The co-primary endpoints for the trial were mean weight loss from baseline and the proportion of subjects achieving weight loss of 5% or more of body weight at 28 weeks. In order for efficacy of a Qsymia dose to be determined, each of three pairwise comparisons, Qsymia dose vs. each single agent and Qsymia dose vs. placebo, needed to reach the 5% significance level for either co-primary endpoint. EQUATE Trial Schematic

Source: Cowen and Company, FDA Briefing Document

Results from EQUATE were reported in December 2008. The trial demonstrated statistically significant weight loss with both Mid-dose and Top-dose Qsymia compared to placebo. Also,

MEMBER: FINRA/SIPC

51

www.cowen.com

Cowen and Company, LLC

EQUITY RESEARCH

the contributions of phentermine and topiramate individually to Qsymia efficacy were demonstrated. Furthermore, the efficacy of Qsymia over these components was shown. 1) Qsymia mid-dose vs. placebo: Patients treated with mid-dose Qsymia had mean weight loss of 8.5% compared to 1.7% for those treated with placebo. This difference was statistically significant (p<0.0001). 62% of the Qsymia mid-dose group had 5% weight loss compared with 15.5% of the placebo group. This difference was also statistically significant (p<0.0001). For both co-primary endpoints, the benchmark established by the factorial trial design was met to show efficacy of the Qsymia mid-dose over placebo. 2) Qsymia mid-dose vs. phentermine 7.5 mg: Patients treated with mid-dose Qsymia had mean weight loss of 8.5%, while those treated with phentermine 7.5 mg lost an average 5.5% of body weight. This difference was statistically significant (p=0.0003), meeting the significance level set in the factorial design to determine efficacy of the Qsymia mid-dose over phentermine alone. 3) Qsymia mid-dose vs. topiramate 46 mg: Patients treated with mid-dose Qsymia had mean weight loss of 8.5%, while those treated with topiramate 46 mg had average loss of 5.1%. This difference was statistically significant (p<0.0001), also meeting the significance level set in the factorial design to determine efficacy of the Qsymia mid-dose over topiramate alone. 4) Qsymia top-dose vs. placebo: Patients treated with top-dose Qsymia had mean weight loss of 9.2% compared to 1.7% for those treated with placebo. This difference was statistically significant (p<0.0001). 66% of the Qsymia top-dose group had 5% weight loss compared with 15.5% of the placebo group, which was also a statistically significant (p<0.0001) difference. For both co-primary endpoints, the benchmark established by the factorial trial design was met to show efficacy of the Qsymia top-dose over placebo. 5) Qsymia top-dose vs. phentermine 15 mg: Patients treated with top-dose Qsymia had mean weight loss of 9.2%, while those treated with phentermine 15 mg lost 6.1% of body weight on average. This difference was statistically significant (p=0.0001), meeting the significance level set in the factorial design to determine efficacy of the Qsymia top-dose over phentermine alone. 6) Qsymia top-dose vs. topiramate 92 mg: Patients treated with top-dose Qsymia had mean weight loss of 9.2%, while those treated with topiramate 92 mg had average loss of 6.4%. This difference was statistically significant (p=0.0009), and again met the significance level set in the factorial design to determine efficacy of the Qsymia top-dose over topiramate alone.

MEMBER: FINRA/SIPC

52

www.cowen.com

Cowen and Company, LLC

EQUITY RESEARCH

EQUATE: Efficacy Data


Key Endpoints n (ITT-LOCF) Mean Weight Loss (%) (ITT-LOCF) 5% weight loss (ITT-LOCF) 10% weight loss (ITT-LOCF) Qsymia middose n=103 8.5% 62.1% 38.8% Qsymia topdose n=103 9.2% 66.0% 40.8% EQUATE (OB-301) (n=756); (28 weeks) Phentermine Phentermine Placebo p-value 7.5mg 15mg n=103 1.7% 15.5% 6.8% p<0.0001 p<0.0001 p<0.0001 n=104 5.5% n=106 6.1% Topiramate 46mg n=102 5.1% Topiramate 92mg n=105 6.4%

Source: Cowen and Company, Vivus, FDA briefing document

Safety: Both doses were well tolerated. The most common adverse events were: paresthesia (mid-dose: 15%, top-dose: 20%, placebo: 3%), dry mouth (mid-dose: 12%, top-dose: 18%, placebo: 0%), constipation (mid-dose: 6%, top-dose: 11%, placebo: 6%), and altered taste (mid-dose: 8%, top-dose: 15%, placebo: 0%). There were no significant differences in depression and altered mood for either Qsymia dose group compared with placebo (mid-dose: 0.9%, top-dose: 1.9%, placebo: 1.8%).

Qsymias Pivotal Phase III Trials Successful


In September 2009, Vivus reported results from the two 56-week pivotal Phase III trials of Qsymia, EQUIP (OB-302) and CONQUER (OB-303). In both trials, treatment with Qsymia met efficacy benchmarks established by the FDA guidance for weight-loss therapy. FDA guidance has delineated the primary efficacy benchmark for weight management drugs as achievement of either a statistically significant difference in mean weight loss of at least 5% between therapy and placebo groups, or a categorical loss of 5% of baseline body weight in at least 35% of patients on therapy, which proportion should be approximately double that for placebo.

EQUIP: Qsymia Efficacy in Morbidly Obese Patients


EQUIP was a randomized, placebo-controlled, double-blind phase III study evaluating two doses of Qsymia in morbidly obese patients. Patients were randomized 2:1:2 to Qsymia 15/92 mg (top-dose), Qsymia 3.75/23 mg (low-dose), or placebo. Treatment initiation consisted of a 4-week blinded titration period, typically recommended with clinical topiramate use to minimize adverse events, starting with the 3.75/23 mg dose and increasing weekly by 3.75/23 mg increments to the assigned dose. Patients were then maintained at the randomized dose for 52 weeks and evaluated on a monthly basis. Eligibility criteria included: age 18-70 years, BMI 35 kg/m2 with no upper limit, triglycerides 200 mg/dl on 01 lipid lowering medication, BP 140/90 mm Hg on 02 antihypertensive medication(s), and fasting serum glucose level 110 mg/dl. All participants were provided standardized counseling based on the LEARN weight management program. They were advised to reduce daily dietary intake by 500 kcal, to increase water consumption, and to increase physical activity.

MEMBER: FINRA/SIPC

53

www.cowen.com

Cowen and Company, LLC

EQUITY RESEARCH

Baseline Demographics in EQUIP

Number of Patients Average BMI, Kg/m Average Weight Average Age, years Female Proportion Mean Systolic BP, mm Hg Mean Diastolic BP, mm Hg
2

1,267 42 255 lbs 43 83% 122 77

Source: Cowen and Company, Vivus, FDA Briefing Document

The co-primary endpoints for the trial were mean weight loss from baseline and the proportion of subjects achieving weight loss 5% of baseline body weight at 56 weeks. The trial successfully met its co-primary endpoints at 56 weeks: 1) % average weight loss: Patients treated with Qsymia 15/92 mg had an average weight loss of 10.9%, compared with 1.6% in the placebo group, for a placebo-adjusted average weight loss of 9.3%. This result was statistically significant (p<0.0001) and met FDAs efficacy benchmark of at least a 5% difference in average weight loss between therapy and placebo. For the low-dose Qsymia 3.75/23 mg group, patients had mean weight loss of 5.1% and placebo-adjusted loss of 3.5%. This result was also statistically significant (p<0.0001) but did not meet FDAs efficacy benchmark for mean weight loss. 2) % of patients losing at least 5% of their weight: 66.7% of patients in the Qsymia 15/92 mg group lost at least 5% of baseline body weight, compared to 17.3% in the placebo group. This improvement over placebo was statistically significant (p<0.0001), meeting the categorical weight loss threshold specified in FDAs guidance. In the low-dose Qsymia group, 44.9% achieved 5% weight loss. This result, which was statistically significant (p<0.0001) compared to placebo, also met the FDA benchmark for categorical weight loss efficacy. Analysis of completers demonstrated placebo-adjusted average weight losses of 12.3% and 4.6% in the Qsymia top-dose and low-dose groups, respectively, with these differences from placebo being statistically significant (p<0.0001). 83.5% and 59% of completers in the Qsymia top-dose and low-dose groups, respectively, lost 5% of body weight, compared with 25.5% of placebo patients. In EQUIP, the lower limit for baseline BMI was 35 kg/m2, and there was no upper limit. The BMI spectrum ranged from 35 kg/m2 to 79 kg/m2 in the trial. The ITT-LOCF analysis was repeated with categorization of patients by BMI to evaluate any influence of baseline BMI on the efficacy results. There was noted to be no significant interaction (p=0.8056) between BMI category and efficacy result, meaning that weight-loss results did not vary significantly by baseline BMI.

MEMBER: FINRA/SIPC

54

www.cowen.com

Cowen and Company, LLC

EQUITY RESEARCH

EQUIP: Efficacy Data


Key Endpoints n (ITT-LOCF) Mean Weight Loss (%) (ITT-LOCF) 5% weight loss (ITT-LOCF) 10% weight loss (ITT-LOCF) n (completers) Mean Weight Loss (%) (Completers) 5% weight loss (Completers) 10% weight loss (Completers) Completion rate EQUIP (OB-302) (n=1267); (56 weeks) Qsymia lowQsymia topPlacebo p-value dose dose (n=234) 5.1% 44.9% 18.8% n=137 6.7% 59.1% 27.7% 59% (n=498) 10.9% 66.7% 47.2% n=297 14.4% 83.5% 67.7% 57% (n=498) 1.6% 17.3% 7.4% n=239 2.1% 25.5% 13.0% 47% p<0.0001 p<0.0001 p<0.0001

p<0.0001 p<0.0001 p<0.0001

Source: Cowen and Company, Vivus, FDA briefing document

Additionally, weight loss results with Qsymia were associated with general improvements, compared with placebo, in anthropometric and cardiometabolic parameters, such as waist circumference, blood pressure, triglycerides, cholesterol, and fasting glucose. The Qsymia topdose group had statistically significant improvements, while the Qsymia low-dose group had numerically greater, but not always statistically significant, changes in these parameters compared with the placebo group. Blood pressure and heart rate: Mean systolic and diastolic blood pressures were decreased from baseline at week 56 in both Qsymia dose groups, while increased in placebo patients. Mean SBP was reduced by 1.8 mm Hg and 2.9 mm Hg in the Qsymia low-dose and top-dose groups, respectively, compared to an increase of 0.9 mm Hg in the placebo group. Mean DBP was lowered by 0.1 mm Hg and 1.5 mm Hg in the Qsymia low-dose and top-dose groups, respectively, compared to an elevation of 0.4 mm Hg in the placebo group. Heart rate was elevated in the Qsymia top-dose cohort (+1.2 bpm) at 56 weeks, but decreased in the Qsymia low-dose (-0.3 bpm) and placebo (-0.2 bpm) groups. EQUIP: Selected Secondary Endpoints

Waist circumference (cm) Systolic blood pressure (mm Hg) Diastolic blood pressure (mm Hg) Change in heart rate (bpm) Total cholesterol (%) Triglycerides (%) Fasting glucose (mg/dL)
Source: Cowen and Company, Vivus, FDA briefing document

Qsymia lowdose -5.6 -1.8 -0.1 -0.3 -5.4 5.2 0.8

p-value 0.0006 0.0019 0.4257 0.9552 0.0502 0.2639 0.1209

Qsymia topdose -10.9 -2.9 -1.5 +1.2 -6.0 -5.2 -0.6

p-value <0.0001 <0.0001 0.0002 0.083 0.0014 <0.0001 <0.0001

Placebo -3.1 0.9 0.4 -0.2 -3.5 9.1 1.9

Safety: The most common adverse events included paresthesia, dry mouth, constipation, dysgeusia, and insomnia, none of which caused study discontinuation in more than 1% of

MEMBER: FINRA/SIPC

55

www.cowen.com

Cowen and Company, LLC

EQUITY RESEARCH

patients. Assessment of depressive symptoms according to standardized rating scales showed improvement in symptoms over time in all groups, with no significant differences among the groups. Qsymia patients did not have increased suicidality compared with placebo patients. The rates of serious AEs were the same, 2.5%, for all groups. Mood-related AEs, such as depression, anxiety, and irritability, as well as cognition-related AEs, such as disturbance in attention, occurred more frequently in the Qsymia top-dose group. Discontinuations related to AEs were also more common in the Qsymia top-dose group (16%), compared with the lowdose (11%) or placebo (8%) groups. Teratogenicity: Out of 15 pregnancies in women treated with Qsymia, there were 3 spontaneous abortions, 3 elective abortions, and 9 healthy live births. There were no birth defects or congenital malformations in these infants.

MEMBER: FINRA/SIPC

56

www.cowen.com

Cowen and Company, LLC

EQUITY RESEARCH

EQUIP Safety Data: Common Adverse Events


Qsymia Low-dose n=240, n(%) Paresthesia Dry mouth Constipation Upper respiratory tract infection Headache Nasopharyngitis Dysgeusia Insomnia Nausea Sinusitis Dizziness Back pain Bronchitis Cough Influenza Depression Diarrhea Fatigue Irritability Vision blurred Alopecia Anxiety Disturbance in attention Hypoesthesia Dry eye Paresthesia oral Dry skin Anorexia Serum bicarbonate decreased Feeling jittery Amenorrhea Aphasia Back injury Serum potassium decreased Hypogeusia Parosmia Osteoarthritis Rhinitis 10 (4.2) 16 (6.7) 19 (7.9) 38 (15.8) 25 (10.4) 30 (12.5) 3 (1.3) 12 (5.0) 14 (5.8) 18 (7.5) 7 (2.9) 13 (5.4) 16 (6.7) 8 (3.3) 18 (7.5) 8 (3.3) 12 (5.0) 12 (5.0) 4 (1.7) 15 (6.3) 5 (2.1) 7 (2.9) 1 (0.4) 2 (0.8) 2 (0.8) 1 (0.4) 0 (0.0) 3 (1.3) 0 (0.0) 3 (1.3) 0 (0.0) 0 (0.0) 3 (1.3) 1 (0.4) 1 (0.4) 1 (0.4) 4 (1.7) 4 (1.7) Qsymia Top-dose n=511, n(%) 96 (18.8) 87 (17.0) 72 (14.1) 63 (12.3) 61 (11.9) 46 (9.0) 43 (8.4) 40 (7.8) 37 (7.2) 37 (7.2) 29 (5.7) 28 (5.5) 28 (5.5) 26 (5.1) 26 (5.1) 24 (4.7) 24 (4.7) 23 (4.5) 23 (4.5) 23 (4.5) 22 (4.3) 19 (3.7) 18 (3.5) 17 (3.3) 12 (2.3) 11 (2.2) 8 (1.6) 7 (1.4) 7 (1.4) 7 (1.4) 6 (1.2) 6 (1.2) 5 (1.0) 5 (1.0) 5 (1.0) 5 (1.0) 2 (0.4) 1 (0.2) Placebo n=513, n(%) 10 (1.9) 19 (3.7) 35 (6.8) 56 (10.9) 52 (10.1) 37 (7.2) 5 (1.0) 25 (4.9) 24 (4.7) 28 (5.5) 21 (4.1) 26 (5.1) 22 (4.3) 18 (3.5) 24 (4.7) 6 (1.2) 23 (4.5) 17 (3.3) 3 (0.6) 16 (3.1) 5 (1.0) 6 (1.2) 3 (0.6) 4 (0.8) 4 (0.8) 2 (0.4) 1 (0.2) 0 (0.0) 1 (0.2) 1 (0.2) 0 (0.0) 0 (0.0) 0 (0.0) 0 (0.0) 0 (0.0) 0 (0.0) 0 (0.0) 0 (0.0)

Source: Cowen and Company, Vivus, FDA briefing document

MEMBER: FINRA/SIPC

57

www.cowen.com

Cowen and Company, LLC

EQUITY RESEARCH

CONQUER: Qsymia Efficacy in Obese Patients with Comorbidities


CONQUER was a randomized, placebo-controlled, double-blind phase III study evaluating two doses of Qsymia in obese and overweight patients with co-morbidities. Patients were randomized 2:1:2 to Qsymia 15/92 mg (top-dose), Qsymia 7.5/46 mg (mid-dose), or placebo. Treatment initiation consisted of a 4-week blinded titration period, typically recommended with topiramate use to minimize adverse events, starting with the 3.75/23 mg dose and increasing weekly by 3.75/23 mg increments to the assigned dose. Patients were then maintained at the randomized dose for 52 weeks and evaluated on a monthly basis. Eligibility criteria included: age 18-70 years, BMI 27-45 kg/m2 with two or more comorbidities, including hypertension, dyslipidemia, pre-diabetes/diabetes, or abdominal obesity (waist circumference 102 cm in men and 88 cm in women). Type 2 diabetes could be managed with metformin monotherapy or without medication. All participants were provided standardized counseling based on the LEARN weight management program. They were given the LEARN manual at baseline and advised to reduce daily dietary intake by 500 kcal, as well as to implement lifestyle changes. Progress discussions with study staff occurred on monthly visits. Baseline Demographics in CONQUER
Number of Patients Average BMI, Kg/m Average Weight Average Age, years Female Proportion Mean Systolic BP, mm Hg Mean Diastolic BP, mm Hg Proportion with HTN Mean Triglycerides, mg/dL Proportion with dyslipidemia Mean HbA1c, % Proportion with Type 2 diabetes Proportion with > 3 comorbidities
2

2,487 37 227 lbs 51 70% 128 81 52% 162.5 36% 5.9 16% 51%

Source: Cowen and Company, Vivus FDA Briefing Document

The co-primary endpoints of the study were mean percent weight loss and the proportion of subjects achieving weight loss of 5% or more at 56 weeks. The trial successfully met its co-primary endpoints at 56 weeks: 1) % average weight loss: Patients treated with Qsymia 15/92 mg had an average weight loss of 9.8%, compared with 1.2% in the placebo group, for a placebo-adjusted average weight loss of 8.6%. This result was statistically significant (p<0.0001) and met FDAs efficacy benchmark of at least a 5% difference in average weight loss between therapy and placebo. For the mid-dose Qsymia 7.5/46 mg group, patients had mean weight loss of 7.8% and

MEMBER: FINRA/SIPC

58

www.cowen.com

Cowen and Company, LLC

EQUITY RESEARCH

placebo-adjusted loss of 6.6%, which was also statistically significant (p<0.0001) and met FDAs efficacy benchmark for mean weight loss. 2) % of patients losing at least 5% of their weight: 70% of patients in the Qsymia 15/92 mg group lost at least 5% of baseline body weight, compared to 21% in the placebo group. This improvement over placebo was statistically significant (p<0.0001), meeting the categorical weight loss threshold specified in FDAs guidance. In the Qsymia 7.5/46 mg group, 62% achieved 5% weight loss. This result, which was also statistically significant (p<0.0001) compared with placebo, also met the FDA benchmark for categorical weight loss efficacy. Analysis of completers demonstrated placebo-adjusted average weight losses of 10.8% and 8% in the Qsymia top-dose and mid-dose groups, respectively, with these differences from placebo being statistically significant (p<0.0001). 85% and 75% of completers in the Qsymia top-dose and mid-dose groups, respectively, lost 5% of body weight, compared with 26% of placebo patients. CONQUER: Efficacy Data
Key Endpoints n (ITT-LOCF) Mean Weight Loss (%) (ITT-LOCF) 5% weight loss (ITT-LOCF) 10% weight loss (ITT-LOCF) n (completers) Mean Weight Loss (%) (Completers) 5% weight loss (Completers) 10% weight loss (Completers) Completion rate CONQUER (OB-303) (n=2487); (56 weeks) Qsymia midQsymia topPlacebo p-value dose dose (n=488) 7.8% 62.0% 37.0% (n=338) 9.6% 74.6% 49.1% 69% (n=981) 9.8% 70.0% 48.0% (n=625) 12.4% 85.1% 64.3% 64% (n=979) 1.2% 21.0% 7.0% (n=557) 1.6% 26.2% 9.7% 57% p<0.0001 p<0.0001 p<0.0001

p<0.0001 p<0.0001 p<0.0001

Source: Cowen and Company, Vivus, FDA briefing document

Additionally, as in EQUIP, weight loss results with Qsymia were associated with significant improvements in anthropometric and cardiometabolic parameters, such as waist circumference, blood pressure, triglycerides, and total cholesterol. In hypertensive patients, there were greater reductions in systolic blood pressure with Qsymia compared with placebo (top-dose, -9 mm Hg (p<0.0001) and mid-dose, -7 mm Hg (p=0.0475) vs. placebo, -5 mm Hg). Also, more patients in the Qsymia top- and mid-dose groups, 11% and 15%, respectively, than 5% in the placebo group were able to discontinue anti-hypertension medications. For patients with type 2 diabetes, having mean baseline HbA1c level of 6.8% across all groups, there were statistically significant reductions in HbA1c from baseline in the two Qsymia groups (-0.4% in each group) versus placebo (-0.1%). More patients treated with placebo (15%) required an increase in diabetes medications than did Qsymia patients (4% in each of the two dose groups). Blood pressure and heart rate: Mean systolic and diastolic blood pressures were decreased from baseline at week 56 to a greater extent in both Qsymia dose groups than in placebo

MEMBER: FINRA/SIPC

59

www.cowen.com

Cowen and Company, LLC

EQUITY RESEARCH

patients. Mean SBP was reduced by 4.7 mm Hg and 5.6 mm Hg in the Qsymia mid-dose and top-dose groups, respectively, compared to a decrease of 2.4 mm Hg in the placebo group. Mean DBP was lower by 3.4 mm Hg and 3.8 mm Hg in the Qsymia mid-dose and top-dose groups, respectively, compared to a reduction of 2.7 mm Hg in the placebo group. Heart rate was elevated in the Qsymia top-dose cohort (+1.7 bpm) at 56 weeks, and relatively unchanged in the Qsymia mid-dose (+0.1 bpm) and placebo (-0.1 bpm) groups. CONQUER: Selected Secondary Endpoints

Waist circumference (cm) Systolic blood pressure (mm Hg) Diastolic blood pressure (mm Hg) Change in heart rate (bpm) Total cholesterol (%) Triglycerides (%) Fasting glucose (mmol/L) Glycated hemoglobin (%) Change in HbA1c (%) in 388 patients (16%) with baseline T2D (baseline HbA1c 6.8%, all groups) Fasting insulin (pmol/L) HOMA-IR hsCRP (mg/L)
Source: Cowen and Company, Vivus, FDA briefing document

Qsymia middose -7.6 -4.7 -3.4 0.1 -4.9 -8.6 -0.01 0 -0.4 -24 -0.93 -2.49

p-value <0.0001 0.0008 0.1281 0.92 0.0345 <0.0001 0.0047 <0.0001 0.0288 0.0004 0.0007 <0.0001

Qsymia topdose -9.2 -5.6 -3.8 +1.7 -6.3 -10.6 -0.07 -0.1 -0.4 -27.6 -1.07 -2.49

p-value <0.0001 <0.0001 0.0031 <0.0001 <0.0001 <0.0001 <0.0001 <0.0001 0.0043 <0.0001 <0.0001 <0.0001

Placebo -2.4 -2.4 -2.7 -0.1 -3.3 4.7 0.13 0.1 -0.1 5.1 0.46 -0.79

Safety: As seen in EQUIP, the most common adverse events included paresthesia, dry mouth, constipation, dysgeusia, and insomnia. Psychiatric AEs, such as depression, anxiety, and irritability, as well as cognitive AEs, such as disturbance in attention, occurred more frequently in the Qsymia top-dose group. These AEs were noted earlier in treatment and generally resolved with discontinuation of therapy. Assessment of depressive symptoms by standardized rating scales showed no significant differences among the groups. Qsymia patients did not have increased suicidality compared with placebo patients. The rates of serious AEs were similar across groups (5% Qsymia top-dose, 3% Qsymia mid-dose, 4% placebo). Discontinuations related to AEs were more common in the Qsymia top-dose group (19%), compared with the mid-dose (12%) or placebo (9%) groups.

MEMBER: FINRA/SIPC

60

www.cowen.com

Cowen and Company, LLC

EQUITY RESEARCH

CONQUER Safety Data: Common Adverse Events


Qsymia Mid-dose n=498, n(%) Dry mouth Paraesthesia Constipation Upper respiratory tract infection Nasopharyngitis Dysgeusia Insomnia Headache Dizziness Sinusitis Back pain Nausea Fatigue Diarrhoea Blurred vision Urinary tract infection Arthralgia Bronchitis Depression Anxiety Irritability Time to onset (days; median, IQR) Duration (days; median, IQR) Resolution among patients discontinuing drug Disturbance in attention Time to onset (days; median, IQR) Duration (days; median, IQR) Resolution among subjects discontinuing drug 67 (13%) 68 (14%) 75 (15%) 61 (12%) 53 (11%) 37 (7%) 29 (6%) 35 (7%) 36 (7%) 34 (7%) 28 (6%) 18 (4%) 22 (4%) 32 (6%) 20 (4%) 26 (5%) 23 (5%) 22 (4%) Psychiatric adverse events 14 (3%) 9 (2%) 13 (3%) 36 (8138) 35 (1181) 10/10 (100%) 10 (2%) 23 (10100) 51 (8149) 2/2 (100%) Cognitive adverse events 35 (4%) 25 (1151) 36 (1881) 21/21 (100%) 7 (<1%) 22 (8119) 39 (1376) 3/3 (100%) 39 (4%) 41 (4%) 34 (3%) 29 (17118) 29 (1263) 33/37 (89%) 29 (3%) 21 (2%) 8 (<1%) 92 (26164) 44 (17121) 4/5 (80%) Qsymia Top-dose n=994, n(%) 207 (21%) 204 (21%) 173 (17%) 133 (13%) 98 (10%) 103 (10%) 102 (10%) 101 (10%) 99 (10%) 85 (9%) 72 (7%) 68 (7%) 67 (7%) 58 (6%) 60 (6%) 54 (5%) 44 (4%) 52 (5%) Placebo n=993, n(%) 24 (2%) 20 (2%) 59 (6%) 128 (13%) 86 (9%) 11 (1%) 47 (5%) 90 (9%) 31 (3%) 67 (7%) 49 (5%) 42 (4%) 50 (5%) 48 (5%) 36 (4%) 37 (4%) 54 (5%) 43 (4%)

Source: Cowen and Company, Vivus, FDA briefing document

SEQUEL: Qsymia Efficacy Results Maintained in Extension of CONQUER


SEQUEL (OB-305) was a placebo-controlled, double-blind, 52-week extension of the CONQUER study. In this extension, patients continued in their randomized arms from CONQUER (2:1:2 Qsymia 15/92 mg, Qsymia 7.5/46 mg, or placebo). Patients were not permitted to continue in SEQUEL if: 1) their BMI was 22 kg/m2 at the end of CONQUER, 2) they were not continuously taking study drug for > 4 weeks at the end of CONQUER, or 3) they had developed a condition which would interfere with compliance and further participation. All participants continued to receive standardized counseling based on the LEARN weight management program. Progress discussions with study staff occurred on monthly visits.

MEMBER: FINRA/SIPC

61

www.cowen.com

Cowen and Company, LLC

EQUITY RESEARCH

Baseline Demographics in SEQUEL

Number of Patients Average BMI, Kg/m Average Weight Average Age, years Female Proportion Mean Systolic BP, mm Hg Mean Diastolic BP, mm Hg Proportion with HTN Mean Triglycerides, mg/dL Proportion with dyslipidemia Mean HbA1c, % Proportion with Type 2 diabetes
Source: Cowen and Company, Vivus, FDA briefing document
2

676 36 224 lbs 52 66% 128 80 50% 156.6 34% 6 22%

The co-primary endpoints of the SEQUEL study, retained from CONQUER, were mean percentage weight loss and the proportion of subjects achieving weight loss of 5% or more from baseline (week 0 of CONQUER) to 108 weeks. The trial successfully met its co-primary endpoints at 108 weeks: 1) % average weight loss: Patients treated with Qsymia 15/92 mg had an average weight loss of 10.5%, compared with 1.8% in the placebo group, for a placebo-adjusted average weight loss of 8.7%. This improvement was statistically significant (p<0.0001). For the middose Qsymia 7.5/46 mg group, patients had mean weight loss of 9.3% and placebo-adjusted loss of 7.5%, which was also statistically significant (p<0.0001). 2) % of patients losing at least 5% of their weight: 79% of patients in the Qsymia 15/92 mg group lost at least 5% of baseline body weight, compared to 30% in the placebo group. This improvement over placebo was statistically significant (p<0.0001). In the Qsymia 7.5/46 mg group, 75% achieved 5% weight loss. This result was also statistically significant (p<0.0001). Analysis of completers demonstrated placebo-adjusted average weight losses of 8.5% and 7% in the Qsymia top-dose and mid-dose groups, respectively, with these differences from placebo again being statistically significant (p<0.0001).

MEMBER: FINRA/SIPC

62

www.cowen.com

Cowen and Company, LLC

EQUITY RESEARCH

SEQUEL: Efficacy Data


Key Endpoints n (ITT-LOCF) Mean Weight Loss (%) (ITT-LOCF) 5% weight loss (ITT-LOCF) 10% weight loss (ITT-LOCF) n (completers) Mean Weight Loss (%) (Completers) Completion rate SEQUEL (OB-305) (n=675); (108 weeks) Qsymia midQsymia topPlacebo p-value dose dose (n=153) 9.3% 75.2% 50.3% (n=127) 9.3% 83% (n=295) 10.5% 79.3% 53.9% (n=245) 10.7% 83% (n=227) 1.8% 30.0% 11.5% (n=196) 2.2% 86% p<0.0001 p<0.0001 p<0.0001

p<0.0001

Source: Cowen and Company, Vivus, FDA briefing document

In SEQUEL, weight loss results with Qsymia continued to be associated with significant improvements in anthropometric and cardiometabolic parameters, such as waist circumference, triglycerides, and fasting insulin. For patients with type 2 diabetes, having mean baseline HbA1c levels of 6.9% in the placebo and Qsymia top-dose groups and 7.3% in the mid-dose group, there were reductions in HbA1c levels from baseline of 0.2% and 0.4% in the two Qsymia groups, respectively, versus the placebo group (-0.04%), in which HbA1c was essentially unchanged. Blood pressure and heart rate: At week 108, mean systolic and diastolic BPs were decreased from baseline by 3-5 mm Hg across all groups. The degree of BP reduction did not differ significantly among the treatment arms. However, more Qsymia patients (16% top-dose and 13% mid-dose) decreased use of anti-hypertensive medications compared with placebo patients (8%). Heart rate was elevated in the Qsymia top-dose group (+1.7 bpm) at 108 weeks to a slightly greater extent than in the Qsymia mid-dose (+1.3 bpm) and placebo (+0.4 bpm) groups. There were no reported AEs associated with heart rate elevations, and no clinically significant effects resulting from the elevations were observed. SEQUEL: Selected Secondary Endpoints

Waist circumference (cm) Systolic blood pressure (mm Hg) Diastolic blood pressure (mm Hg) Change in heart rate (bpm) Fasting insulin (lIU/mL) Triglycerides (%) Change in HbA1c (%) Change in HbA1c (%) in 145 patients (22%) with baseline T2D (baseline HbA1c 6.9-7.3%)
Source: Cowen and Company, Vivus, FDA briefing document

Qsymia middose -9.8 -4.7 -3.7 +1.3 -5.3 -12.5 0.01 -0.4

p-value <0.0001 NS NS 0.0051 <0.01 0.0042 NS

Qsymia topdose -10.6 -4.3 -3.5 +1.7 -5.2 -13.7 0 -0.2

p-value <0.0001 NS NS 0.0012 <0.0001 0.0003 NS

Placebo -3.6 -3.2 -3.9 +0.4 -2.6 0.4 0.2 -0.04

MEMBER: FINRA/SIPC

63

www.cowen.com

Cowen and Company, LLC

EQUITY RESEARCH

In April 2011, Vivus presented additional data from SEQUEL at the 60th Annual Scientific Meeting of the American College of Cardiology (ACC). The ACC poster highlighted SEQUEL data in two subgroups of patients, those with dyslipidemia and hypertension: 1) Hypertensive patients (SBP 140-160 mmHg or DBP 90-100 mmHg in non-diabetic patients, and SBP 130-160 mmHg or DBP 85-100mmHg in diabetic patients) across all groups had BP reductions over 108 weeks. Mean SBP decreased by 8.6 mm Hg and 6.9 mm Hg in the Qsymia top-dose and mid-dose groups, respectively, with reduction of 6.7 mm Hg in placebo patients. Mean DBP decreased by 5.8 mm Hg, 4.8 mm Hg, and 5.7 mm Hg in the Qsymia topdose, Qsymia mid-dose, and placebo groups, respectively. BP reductions in the Qsymia group were not statistically significant compared to those in the placebo group. 2) Patients with dyslipidemia (triglycerides 200-400mg/dL or patients taking 2 lipid-lowering agents) on Qsymia mid-dose and top-dose had reductions in triglyceride levels of 25.9% and 26.3%, respectively, compared with a 14.3% reduction in those on placebo (p=0.055 and p=0.015, respectively). 3) Patients with dyslipidemia on Qsymia mid-dose and top-Dose had improvements in HDL cholesterol levels of +11.4% and +16.7%, respectively, compared with a +9.1% increase in placebo patients (NS and p<0.015, respectively). Safety: The AEs occurring during the SEQUEL second year extension were similar to those observed in CONQUER, but the incidence of individual AEs was lower in the second year compared with the first year. The most common AEs included paresthesia, dry mouth, constipation, dysgeusia, and upper respiratory tract infection. Rates of depression-related AEs were comparable in placebo and Qsymia top-dose patients (8% each) and lower in the middose group (4%). Qsymia patients did not have increased suicidality compared with placebo patients. The rates of serious AEs were similar across groups (8% Qsymia top-dose, 6% Qsymia mid-dose, 6% placebo). Discontinuations related to AEs by week 108 were also similar across the groups: Qsymia top-dose group (4.4%), mid-dose (4.5%), and placebo (3%). Teratogenicity: During the 108 weeks of CONQUER and SEQUEL, there were 2 pregnancies, with one pregnancy carried to term by a patient treated with Qsymia 15/92 mg. The infant was born healthy, with no congenital malformations observed.

MEMBER: FINRA/SIPC

64

www.cowen.com

Cowen and Company, LLC

EQUITY RESEARCH

SEQUEL Safety Data: Common Adverse Events


Qsymia Mid-dose (n=153) Qsymia Top-dose (n=295) Weeks 056; n(%) Constipation Paraesthesia Dry mouth Upper respiratory tract infection Nasopharyngitis Dysgeusia Sinusitis Headache Insomnia Diarrhea Back pain Dizziness Nausea Bronchitis Fatigue Procedural pain Arthralgia Influenza Urinary tract infection Gastroenteritis 25 (16.3) 21 (13.7) 21 (13.7) 23 (15.0) 20 (13.1) 18 (11.8) 17 (11.1) 8 (5.2) 12 (7.8) 14 (9.2) 11 (7.2) 9 (5.9) 5 (3.3) 9 (5.9) 7 (4.6) 7 (4.6) 13 (8.5) 11 (7.2) 8 (5.2) 3 (2.0) 62 (21.0) 62 (21.0) 59 (20.0) 55 (18.6) 39 (13.2) 39 (13.2) 39 (13.2) 28 (9.5) 24 (8.1) 21 (7.1) 21 (7.1) 20 (6.8) 19 (6.4) 17 (5.8) 17 (5.8) 17 (5.8) 13 (4.4) 13 (4.4) 13 (4.4) 12 (4.1) 16 (7.1) 6 (2.6) 5 (2.2) 47 (20.7) 35 (15.4) 4 (1.8) 19 (8.4) 21 (9.3) 15 (6.6) 12 (5.3) 19 (8.4) 6 (2.6) 13 (5.7) 8 (3.5) 11 (4.9) 6 (2.6) 20 (8.8) 11 (4.9) 11 (4.9) 12 (5.3) 11 (7.2) 1 (0.7) 1 (0.7) 26 (17.0) 13 (8.5) 1 (0.7) 12 (7.8) 4 (2.6) 9 (5.9) 3 (2.0) 9 (5.9) 2 (1.3) 10 (6.5) 8 (5.2) 2 (1.3) 8 (5.2) 7 (4.6) 10 (6.5) 14 (9.2) 2 (1.3) Placebo (n=227) Qsymia Mid-dose (n=153) Qsymia Top-dose (n=295) Placebo (n=227)

Weeks 56-108; n(%) 12 (4.1) 10 (3.4) 4 (1.4) 45 (15.3) 26 (8.8) 3 (1.0) 28 (9.5) 12 (4.1) 11 (3.7) 11 (3.7) 15 (5.1) 1 (0.3) 4 (1.4) 10 (3.4) 4 (1.4) 14 (4.7) 16 (5.4) 19 (6.4) 18 (6.1) 9 (3.1) 7 (3.1) 0 (0.0) 1 (0.4) 42 (18.5) 26 (11.5) 0 (0.0) 18 (7.9) 6 (2.6) 8 (3.5) 3 (1.3) 7 (3.1) 2 (0.9) 4 (1.8) 7 (3.1) 2 (0.9) 4 (1.8) 14 (6.2) 8 (3.5) 13 (5.7) 6 (2.6)

Source: Cowen and Company, Vivus, FDA briefing document

Qsymia Safety
Across the Qsymia clinical trial program, AEs generally occurred as expected and were consistent with side effects of the individual component drugs. In the 1-year cohort of patients, the most common AEs across Qsymia mid-dose and top-dose patients were paresthesias, dry mouth, and constipation. Discontinuations resulting from treatment-emergent AEs were more frequent in the Qsymia groups, 17% (top-dose) and 12% (mid-dose), compared with placebo (8%). The rate of serious AEs was similar across the groups: 4% (top-dose), 3% (mid-dose), 3% (placebo). There were no serious psychiatric or cognitive AEs with Qsymia treatment. AE profiles in the second year were consistent with those during the first year, but with overall lower rates. There were no new or unexpected AEs observed during the second year.

MEMBER: FINRA/SIPC

65

www.cowen.com

Cowen and Company, LLC

EQUITY RESEARCH

Qsymia Phase III Safety Data (1-year Cohort)


Qsymia Low- Qsymia Middose dose (n=240) (n=498) Nervous System Disorders Paresthesia 1.9% 4.2% 13.7% Headache 9.3% 10.4% 7.0% Dizziness 3.4% 2.9% 7.2% Dysgeusia 1.1% 1.3% 7.4% Hypoesthesia 1.2% 0.8% 3.6% Disturbance in Attention 0.6% 0.4% 2.0% Psychiatric Disorders Insomnia 4.7% 5.0% 5.8% Depression 2.2% 3.3% 2.8% Anxiety 1.9% 2.9% 1.8% Gastrointestinal Disorders Constipation 6.1% 7.9% 15.1% Dry Mouth 2.8% 6.7% 13.5% Nausea 4.4% 5.8% 3.6% Diarrhea 4.9% 5.0% 6.4% Dyspepsia 1.7% 2.1% 2.2% Gastroesophageal Reflux Disease 1.3% 0.8% 3.2% Paresthesia Oral 0.3% 0.4% 0.6% General Disorders and Administration Site Conditions Fatigue 4.3% 5.0% 4.4% Irritability 0.7% 1.7% 2.6% Thirst 0.7% 2.1% 1.8% Chest Discomfort 0.4% 2.1% 0.2% Cardiac Disorders Palpitations 0.8% 0.8% 2.4% Placebo (n=1,561)
Source: Cowen and Company, Vivus, FDA briefing document

Qsymia Topdose (n=1,580) 19.9% 10.6% 8.6% 9.4% 3.7% 3.5% 9.4% 4.3% 4.1% 16.1% 19.1% 7.2% 5.6% 2.8% 2.6% 2.2% 5.9% 3.7% 2.0% 0.9% 1.7%

Both the Initial Ad Com Panel and the FDA Rejected Qsymia in 2010...
On July 15, 2010, the FDA's Endocrinologic and Metabolic Drugs Advisory Committee voted 10-6 against recommending Qsymia for approval. The advisory committee concluded that Qsymias one-year data did not adequately represent the drug's long-term use, and that Qsymia had not been sufficiently examined in patients with cardiovascular disease. The FDA and the advisory committee focused almost exclusively on Qsymia's safety profile, particularly cardiovascular risk, teratogenicity, cognitive AEs, neuropsychiatric AEs, and metabolic acidosis. Vivus presented data adequate to assuage the panel on the psychiatric, cognitive, and metabolic acidosis issues, but the cardiovascular data presented were insufficient to judge CV outcomes and more teratogenicity data were deemed to be needed. On October 28, 2010, Vivus announced receipt of a Complete Response Letter from the FDA. The CRL requested: (1) a comprehensive assessment of topiramates and Qsymias

MEMBER: FINRA/SIPC

66

www.cowen.com

Cowen and Company, LLC

EQUITY RESEARCH

teratogenic potential, (2) evidence that the elevation in heart rate observed in Qsymias trials does not increase the risk for MACE, and (3) the results of the two-year SEQUEL trial. At a January 2011 meeting with the FDA, Vivus was asked to perform an analysis of existing healthcare databases (the FORTRESS study) to determine the incidence of oral clefts in offspring of women treated with topiramate for migraine prophylaxis.

Qsymia Post Hoc CV Risk Analysis


In November 2010, at the American Heart Association meeting, Vivus presented a poster entitled Low-Dose, Controlled-Release Phentermine/Topiramate Significantly Improves Reynolds 10-Year Risk Score In Obese Women And Men. The Reynolds Risk Score is designed as a rough predictive measure of heart attack, stroke, or major heart disease over the next 10 years. The risk score is calculated by taking into account gender, age, blood pressure, smoking history, cholesterol levels, high-sensitivity C-reactive protein (hsCRP) level as a measure of inflammation, and family history. The Reynolds test was validated in both female and male populations over a 10-year period to evaluate the development factors for heart attack, stroke, angioplasty, coronary artery bypass surgery, or mortality due to heart disease. The AHA presentation highlighted 56-week modified Reynolds Risk Score data pooled from the Phase III EQUIP and CONQUER studies of Qsymia in a total of 3,652 patients (2,711 women, 941 men). Reynolds Risk Score data were collected for patients at baseline and at the conclusion of the trials (56 weeks), including gender, age, smoking status, systolic BP, family history of early CVD, total and HDL cholesterol measurements, hsCRP levels, and diabetic status (for women only). The Reynolds Risk Score calculator was used to measure the change in relative risk of a CV event over 10 years between baseline and week 56. In 3,652 patients analyzed, weight loss with Qsymia in all groups, except for the low-dose Qsymia group in men, was associated with statistically significant reductions in 10-year Reynolds Risk Score from baseline to week 56 (p<0.05). Least squares mean percent change analysis showed risk decline of 5.4%-16.3% in the Qsymia-treated groups, and risk increase by 4.0%-6.1% in the placebo group. Because the modified Reynolds Risk Score calculator used in this analysis is driven by systolic BP and total and HDL cholesterol measurements, which improve with weight loss, this was an expected outcome.

MEMBER: FINRA/SIPC

67

www.cowen.com

Cowen and Company, LLC

EQUITY RESEARCH

Qsymia Reynolds Risk Score (RRS) Data Summary (EQUIP/CONQUER)

RRS Analysis baseline mean RRS mean change RRS (SD) least squares mean % change RRS RRS Analysis baseline mean RRS mean change RRS (SD) least squares mean % change RRS
*p<0.05 vs baseline; p<0.05 vs placebo

low-dose Qnexa n=193 1.24 -0.12* (0.64) -6.80*


,

Women (n=2,711) mid-dose Qnexa high-dose Qnexa n=341 n=1,084 3.06 2.51 -0.65* (1.86) -13.64*
,

placebo n=1,093 2.65 -0.13 (1.76) 3.97* placebo n=372 6.70 -0.20 (3.00) 6.06*

-0.55* (2.09) -16.27*


,

low-dose Qnexa n=39 5.12 -0.36 (1.65) -8.88

Men (n=941) mid-dose Qnexa high-dose Qnexa n=147 n=383 7.67 6.54 -1.15* (3.01) -5.40*, -1.23* (3.24) -11.78*,

Source: Cowen and Company, AHA 2010 Company Data Presentation

While the 2010 panel members were satisfied with Qsymia's efficacy profile, they were concerned about the lack of CV outcomes data and the relatively short safety database (12 months) available for a drug that would be used by high CV risk patients for many years. Even with the addition of the two-year SEQUEL data, Qsymias dataset was considered lacking in long-term safety data for patients with elevated CV risk.

Pooled Analysis of Blood Pressure and Heart Rate Data


In a pooled analysis of CV data from the Qsymia clinical trial program, mean systolic and diastolic BPs were decreased from baseline at week 56 in all three Qsymia dose groups as well as in the placebo group. Mean SBP was reduced by 3.3 mm Hg, 5.2 mm Hg, and 5.2 mm Hg in the Qsymia low-dose, mid-dose, and top-dose groups, respectively, compared to a decrease of 2.1 mm Hg in the placebo group. Mean DBP was lower by 0.9 mm Hg, 3.3 mm Hg, and 2.9 mm Hg in the Qsymia low-dose, mid-dose, and top-dose groups, respectively, compared to a reduction of 1.9 mm Hg in the placebo group. Heart rate was elevated in all three Qsymia dose cohorts by an average of +1.3 bpm (low-dose), +0.6 bpm (mid-dose), and +1.6 bpm (top-dose) at 56 weeks, while there was no overall change in heart rate for the placebo group.

MEMBER: FINRA/SIPC

68

www.cowen.com

Cowen and Company, LLC

EQUITY RESEARCH

Qsymia: Pooled CV Safety Data


Qsymia Lowdose (n=234) Qsymia Middose (n=488) Qsymia Topdose (n=1,553) Placebo (n=1,532)

Systolic blood pressure, mm Hg Mean change Difference from placebo (mmHg) Diastolic blood pressure, mm Hg Mean change Difference from placebo (mmHg) Heart rate, bpm Mean change Difference from placebo (bpm)

-3.3 -1.20

-5.2 -3.10

-5.2 -3.10

-2.1

-0.9 1.00

-3.3 -1.40

-2.9 -1.00

-1.9

+1.3 +1.30

+0.6 +0.60

+1.6 +1.60

0.0

Source: Cowen and Company, Vivus, FDA briefing document

What About Teratogenicity? Top-Line FORTRESS Data Look OK


While the July 15, 2010 EMDAC panel was satisfied with Qsymia's efficacy profile, the panel discussion focused on five key safety issues: psychiatric AEs, cognitive AEs, metabolic acidosis, teratogenicity, and cardiovascular AEs. The committee appeared to conclude that the psychiatric AEs, the cognitive AEs, and the metabolic acidosis risks of Qsymia were manageable and could be handled with label warnings. The committee was split on teratogenicity, but it appeared that a Pregnancy Category X designation and a strict REMS program might be sufficient. In January 2011, the FDA requested that Vivus provide additional analyses of the historical teratogenicity issues associated with topiramate, particularly oral clefts in the offspring of pregnant women treated with topiramate for migraine prophylaxis. Upon agreement with the FDA, Vivus is conducting a retrospective study of medical claims databases (the FORTRESS study) to measure the teratogenic risk for the topiramate component of Qsymia. Top-line data from the FORTRESS study were reported in December 2011. Below, we have summarized the preliminary data from FORTRESS and the smaller Wolter Kluwer studies conducted by Vivus, along with listing oral cleft prevalence rates and risk ratios from the North American and the UK Epilepsy and Pregnancy Registries. We view the impact of the top-line FORTRESS data as neutral to incrementally positive for Qsymia, since they point to a risk ratio (1.88-5.44X) within the known range for topiramate. We point investors to the small subset of the Total Topiramate Cohort that was composed of women exposed to topiramate polytherapy; in that group, prevalence of oral clefts was much higher, as was the associated increased risk ratio (6-14X).

MEMBER: FINRA/SIPC

69

www.cowen.com

Cowen and Company, LLC

EQUITY RESEARCH

February 2012: FDA AdCom Recommends Approval With Post-Approval CVOT Requirement
At the February 2012 FDA advisory committee, a number of panel members expressed serious concerns about Qsymias CV safety profile. However, given the lack of available treatment options and the impact of obesity and its comorbidities in patients, the panel recommended that a post-approval CVOT would be sufficient to rule out CV risk, voting 20-2 in favor of Qsymias approval.

2012 Saw a Change of Heart at the FDA for Obesity Drugs


In July 2012, the FDA approved Qsymia for the treatment of obesity. As part of its postmarketing commitment, Vivus is required to complete 10 post-marketing requirements, including a CVOT. According to the companys latest guidance, Vivus will conduct a post-approval CVOT, known as AQCLAIM (A Qsymia CardiovascuLAr morbIdity and Mortality), which will enroll approximately 11,000 patients with documented CV risk (1/3 high-risk and 2/3 low-risk patients) at 300 sites worldwide. This randomized, double-blind, placebo-controlled, multicenter trial will evaluate the effect of long-term Qsymia treatment on incidence of MACE, and will be designed to show superiority of Qsymia over placebo (16-20% risk reduction). Approximately 630 events will be needed to complete the trial, and based on a 2% event rate per year, the company guided that the trial could be completed in approximately 4.5 years. In September 2013, Vivus indicated that patient enrollment in AQCLAIM is anticipated to begin in 1Q14.

What about Europe?


In December 2010, Vivus filed a MAA with the EMA for Qsiva, the proposed name for Qsymia in Europe. In May 2011, the company received the EMAs 120-day list of questions from the CHMP, which included issues that were similar to those raised by the FDA. Vivus submitted its response to the 120-day questions in the fourth quarter of 2011. In January 2012, Vivus received the 180-day List of Outstanding Issues from the CHMP and submitted its response to these issues in April 2012. In September 2012, Vivus announced that, based on preliminary feedback from EMA's CHMP, it expects a negative decision on the Qsiva MAA. In October 2012, Vivus announced that it had received the formal opinion from EMAs CHMP, recommending against approval of Qsiva. According to the company, the rejection was due to concerns about potential CV and CNS effects following long-term use, teratogenic potential, and concerns about inappropriate use of the drug. In February 2013, Vivus announced that the EMA confirmed its original decision of 10/18/12 to decline approval for the MAA for Qsiva. EMA indicated that a pre-approval CVOT would be necessary to establish Qsivas long-term safety.

MEMBER: FINRA/SIPC

70

www.cowen.com

Cowen and Company, LLC

EQUITY RESEARCH

EU approval pushed back a minimum of 2-3 years: Management previously indicated that the most likely scenario would involve proceeding with the CVOT required for the US, and reengaging EMA with discussions about possibly re-filing based on interim CVOT data, hoping for an outcome similar to the Contrave situation (Orexigen) with the LIGHT study in the US. In September 2013, Vivus announced the submission of a request to EMA for scientific advice regarding use of a pre-specified interim analysis of AQCLAIM in support of the Qsiva MAA resubmission. The company indicated that the anticipated 1Q14 date for initiation of enrollment in AQCLAIM has been chosen to accommodate advice from EMA.

Qsymias IP Estate Controversial


Qsymias intellectual property estate has been the subject of recent controversy. Qsymia is covered by patents and patent filings in the U.S., Europe, and international markets. The key U.S. patent (#7,056,890 B2, Najarian patent) covering Qsymias use was issued in June 2006. This patent, along with other patent applications, claims a method-of-use combining a sympathomimetic agent (i.e., phentermine) and an anticonvulsant (i.e., topiramate) for the treatment of obesity. This patent also covers phentermine doses of 5-15mg and topiramate doses of 100-200mg.

Our Consultants Believe That JNJs Shank Patent Could Be an IssueBut a Settlement Is the Most Likely Outcome
J&J owns a method-of-use patent for Topamax (topiramate) (U.S. patent #6,071,537, the Shank patent), which makes broad claims regarding treatment of obesity comprising the use of topiramate. It also specifically claims therapeutically effective doses of topiramate from 50 to 400mg. The patent claiming composition-of-matter of Topamax expired in September 2008 and thus does not present a commercialization barrier for Qsymia. According to our consultants, the Qsymia Najarian patent estate could potentially be seen as infringing on the J&J Shank patent, which expires in 2017. Given that J&J is not marketing topiramate for the treatment of obesity, the company is unlikely to be able to claim irreparable harm. We see a royalty settlement between Vivus and J&J as the most likely eventual outcome.

MEMBER: FINRA/SIPC

71

www.cowen.com

Cowen and Company, LLC

EQUITY RESEARCH

2) On Deck: Orexigens Contrave


Contrave consists of a fixed-dose combination of sustained-release (SR) bupropion and a proprietary SR version of naltrexone designed for improved tolerability. These individual drugs that have been approved separately for non-obesity indications and are sold as generics in the US. The arcuate nucleus of the hypothalamus plays a significant role in the regulation of body weight. In the arcuate nucleus, neurons which produce pro-opiomelanocortin (POMC) release both -melanocyte stimulating hormone (-MSH) and -endorphin. POMC activation results in an anorectic effect mediated by -MSH at melanocortin receptors. A negative autoinhibitory feedback loop is provided by -endorphin acting at opioid receptors on POMC neurons. Bupropion, a chemical aminoketone, is a non-tricyclic antidepressant that is widely used for depression (approved in the US in 1985) and smoking cessation (approved in the US in 1997). The primary action of bupropion is considered to be re-uptake inhibition of dopamine and norepinephrine. Bupropion therapy is known to be associated with modest weight reduction. Similar to endogenous signals causing weight loss, bupropion appears to increase POMC firing, ultimately leading to decreased appetite and increase in energy expenditure. Naltrexone is a non-selective opioid receptor antagonist used in the treatment of opioid dependence (approved in the US in 1984) and alcohol addiction (approved in the US in 1995). It is believed that naltrexone inhibits the reinforcement effects of addictive substances by blocking neural reward pathways. It is possible that -endorphin may be one compensatory mechanism limiting long term efficacy of weight loss therapy. By blocking opioid receptors, naltrexone interferes with the inhibitory action of -endorphin, which allows POMC firing to be unchecked. Therefore, naltrexone should augment the effect of bupropion on appetite suppression. The synergistic combination of bupropion and naltrexone in Contrave is meant to generate more robust and durable weight loss. In particular, Orexigen anticipates that, in addition to causing weight loss, this combination may be especially effective for reduction of food cravings, which are believed to be governed by central reward systems.

Contrave: Proposed Indication


The proposed indication for Contrave is the management of obesity, including weight loss and maintenance of weight loss, in conjunction with lifestyle modification. Obese adults with initial body mass index (BMI) 30 kg/m2 and overweight adults (BMI 27 kg/m2) who have one or more concomitant risk factors, such as diabetes, dyslipidemia, or hypertension, comprise the target population. The recommended daily dose of Contrave is 32 mg naltrexone/360 mg bupropion (NB32), administered as 16 mg naltrexone/180 mg bupropion twice daily. .

MEMBER: FINRA/SIPC

72

www.cowen.com

Cowen and Company, LLC

EQUITY RESEARCH

Contrave: Clinical Data Summary COR: Contraves Successful Phase III Program
In July 2009, Orexigen successfully completed four Phase III trials with Contrave, a development program termed COR (Contrave Obesity Research). This program met the requirements detailed in the FDAs weight management guidance for one-year studies, adequate patient exposure, efficacy endpoints, and appropriate target populations consisting of obese and overweight patients with comorbidities, including type 2 diabetes. Four 56-week trials evaluated a 360 mg daily dose of bupropion SR paired with 16 mg or 32 mg naltrexone SR doses. The co-primary endpoints of each trial were mean percentage change in body weight from baseline, and proportion of subjects with categorical 5% weight loss from baseline. Secondary endpoints included proportion of subjects with categorical 10% weight loss, weight-related cardiometabolic parameters, and impact of weight on QOL. The modified intent-to-treat (mITT) population, comprised by all randomized patients with at least one postbaseline measurement while on study drug, was analyzed for primary efficacy. Contraves Phase III Program: COR (Contrave Obesity Research)
COR-BMOD (formerly NB-302)
Number of patients Duration Treatment groups Patient demographics Comorbidities for inclusion Mean baseline weight, lbs Mean BMI, kg/m
2

COR-I (formerly NB-301)


1,742 1 year Contrave32, Contrave16, Placebo

COR-II (formerly NB-303)


1,496 1 year Contrave32, Placebo

COR-Diabetes (formerly NB-304)


505 1 year Contrave32, Placebo

793
1 year Contrave32, Placebo

BMI 30-45 kg/m2 or BMI 27-45 kg/m2 with at least 1 comorbidity Required controlled hypertension, dyslipidemia, or both 220 36-37 36 44 85% Caucasian (75%) African-American (19%) 36 44 85% Caucasian (84%) African-American (14%)

BMI 27-45 kg/m2 Type 2 diabetes 232 36 54 56% Caucasian (79%) African-American (16%)

Average age (years) Female Proportion Ethnicity

46 90%
Caucasian (70%) African-American (24%) Intensive 90 min group meetings weekly for 16 weeks, then every other week for 12 weeks, then monthly (28 total sessions)

Adjunct lifestyle/behavior modification program

Instruction on diet and advice on lifestyle/behavior modification, including exercise instruction, at baseline and weeks 12/24/36/48 1) mean weight loss from baseline 2) proportion of patients > 5% weight loss

Co-primary endpoints Data Announced

January 2009

July 2009

July 2009

July 2009

Source: Cowen and Company, Orexigen Therapeutics

MEMBER: FINRA/SIPC

73

www.cowen.com

Cowen and Company, LLC

EQUITY RESEARCH

Efficacy Data From Contrave's Four Phase III Trials


Mean Weight Mean Weight Mean Weight Loss Loss (%) Loss (%) (%) Treatment Group Placebo Group Placebo- adjusted >5% Weight Loss Treatment Group >5% Weight >5% Weight Loss >10% Weight >10% Weight Loss PlaceboLoss Loss Placebo Group Adjusted Treatment Group Placebo Group >10% Weight Loss PlaceboAdjusted

Contrave 32 mg naltrexone/360 mg bupropion (modified-ITT)


COR-BMOD (NB-302); Contrave 32; (n=793) COR-I (NB-301); Contrave 32; (n=1,742) COR-II (NB-303); Contrave 32; (n=1,496) COR-Diabetes (NB-304); Contrave 32; (n=505) 9.3% 6.1% 6.5% 5.0% 5.1% 1.3% 1.9% 1.8% 4.2% 4.8% 4.6% 3.2% 66.4% 48.0% 55.6% 44.5% 42.5% 16.4% 17.5% 18.9% 23.9% 31.6% 38.1% 25.6% 41.5% 24.6% 27.3% 18.5% 20.2% 7.4% 7.0% 5.7% 21.3% 17.2% 20.3% 12.8%

Contrave 32 mg naltrexone/360 mg bupropion (completers)


COR-BMOD (NB-302); Contrave 32; (n=793) COR-I (NB-301); Contrave 32; (n=1,742) COR-II (NB-303); Contrave 32; (n=1,496) COR-Diabetes (NB-304); Contrave 32; (n=505) 11.5% 8.1% 7.8% 5.9% 7.3% 1.8% 2.4% 2.2% 4.2% 6.3% 5.4% 3.7% 80.4% 61.8% 68.8% 53.1% 60.4% 23.1% 22.3% 24.0% 20.0% 38.7% 46.5% 29.1% 55.2% 34.5% 35.7% 26.3% 30.2% 10.7% 9.4% 8.0% 25.0% 23.8% 26.3% 18.3%

Source: Cowen and Company, Orexigen Therapeutics

COR-BMOD: First Phase III Trial Meets One of Two FDA Efficacy Benchmarks
In January 2009, Orexigen reported data from the first of four pivotal trials of Contrave, the NB302 (COR-BMOD) study. In this 56-week study, both Contrave treatment and placebo were combined with intensive group behavior modification (BMOD), a program of diet, exercise, and behavior therapy. Patients were randomized 1:3 to placebo or treatment with 32 mg sustainedrelease (SR) naltrexone + 360 mg SR bupropion (Contrave32). The study included a 3-week dose escalation, starting with one-quarter of the full dose, which was increased weekly until the full dose was reached by week 4. Eligibility criteria in the trial included: age 18-65 years, and BMI 30-45 kg/m2 or 27-45 kg/m2 with controlled hypertension and/or dyslipidemia. Exclusion criteria included diabetes mellitus as well as current smoking, or tobacco/nicotine use within 6 months before screening. The intensive behavior modification program was delivered to groups of 10-20 participants by registered dieticians, behavioral psychologists, or exercise specialists. Group meetings of 90 minutes, including a weigh-in, occurred weekly for 16 weeks, then every other week for 12 weeks, and then monthly, for a total of 28 sessions. Patients were prescribed a diet of 1,200 kcal/day to 2,000 kcal/day, depending on body weight, and were encouraged to gradually increase to 180 min/week of moderate physical activity in the first 6 months of the trial. The trial successfully met its co-primary endpoints:

MEMBER: FINRA/SIPC

74

www.cowen.com

Cowen and Company, LLC

EQUITY RESEARCH

1) % average weight loss: Contrave32 patients had an average weight loss of 9.3%, compared with 5.1% in the placebo group, for a placebo-adjusted average weight loss of 4.2%. This result was statistically significant (p<0.001). However, it did not meet FDAs efficacy benchmark of a 5% difference in average weight loss between therapy and placebo. 2) % of patients losing at least 5% of their weight: 66.4% of the patients in the Contrave32 group lost at least 5% of baseline body weight, compared to 42.5% in the in the placebo group. This result on the one hand seems to satisfy the 35% FDA guidance benchmark; on the other hand, the agency could, at least theoretically, question whether the trial satisfied the efficacy requirement, since the proportion of subjects on treatment achieving at least 5% weight loss was approximately 1.5 times that of the number of subjects on placebo, versus the FDA guidance for approximately double. COR-BMOD: Efficacy Data
Key Endpoints n (mITT-LOCF) Mean Weight Loss (%) (mITT-LOCF) 5% weight loss (mITT-LOCF) 10% weight loss (mITT-LOCF) n (Completers) Mean Weight Loss (%) (Completers) 5% weight loss (Completers) 10% weight loss (Completers) Completion rate Contrave32 n=482 9.3% 66.4% 41.5% n=301 11.5% 80.4% 55.2% 62% COR-BMOD (NB-302) (n=793) Placebo Difference n=193 5.1% 4.2% 42.5% 23.9% 20.2% 21.3% n=106 7.3% 60.4% 30.2% 55% 4.2% 20.0% 25.0% p-value p<0.001 p<0.001 p<0.001 p<0.001 p<0.001 p<0.001 -

Source: Cowen and Company, Orexigen Therapeutics

Secondary Endpoints: Contrave treatment improved other secondary anthropometric and cardiometabolic measures: Waist circumference was reduced by 10.0 cm vs. 6.8 cm for placebo, HDL increased by 4.1mg/dL, and triglycerides reduced by 22.2 mg/dL (16.6%). By comparison, patients in the placebo arm increased HDL by 0.9mg/dL and reduced triglycerides by 11.3mg/dL (8.5%). Blood pressure and heart rate: At week 56, systolic blood pressure (SBP) was reduced by 3.9 mm Hg on average in placebo patients, compared with a decrease of 1.3 mm Hg in patients treated with Contrave (p=0.002). Diastolic blood pressure (DBP) was also reduced to different degrees in placebo and Contrave patients, by 2.8 mm Hg and 1.4 mm Hg, respectively (p=0.017). At 56 weeks, there were no significant differences in heart rate between groups (+0.2 bpm for placebo vs. +1.1 bpm for Contrave); however, placebo-adjusted heart rate increases as large as +3.4 bpm were noted in the first 20 weeks of treatment with Contrave.

MEMBER: FINRA/SIPC

75

www.cowen.com

Cowen and Company, LLC

EQUITY RESEARCH

COR-BMOD: Secondary Endpoints


Change in blood fasting glucose Change in insulin Change in HOMA-IR Change in triglycerides Change in hs-CRP Change in IWQOL-LITE Change in waist circumference Change in Systolic blood pressure Change in Diastolic blood pressure Change in Heart rate (beats per minute)
Source: Cowen and Company, Orexigen Therapeutics

Contrave32 -1.5% -28.0% -29.9% -16.6% -25.9% 13.4 -10 cm -1.3 mm Hg -1.4 mm Hg +1.1

Placebo 0.0% -15.5% -16.6% -8.5% -16.9% 10.3 -6.8cm -3.9 mm Hg -2.8 mm Hg +0.2

p-value 0.185 0.003 0.003 0.004 0.165 <0.001 <0.001 0.002 0.017 0.139

Safety & Tolerability: Nausea, mostly mild to moderate in intensity, was the most common AE and more frequently associated with Contrave treatment. Other common AEs included constipation, dizziness, dry mouth, tremor, upper abdominal pain, and tinnitus. There were no differences between groups in the most common psychiatric AEs, such as anxiety, sleep disorder, depressed mood, and insomnia. Depression was more frequent in placebo patients.

MEMBER: FINRA/SIPC

76

www.cowen.com

Cowen and Company, LLC

EQUITY RESEARCH

COR-BMOD: Safety Data

Nausea Headache Constipation Dizziness Vomiting Insomnia Dry mouth Anxiety Tremor Abdominal pain, upper Tinnitus Insomnia Anxiety Sleep disorder Depressed mood Abnormal dreams Middle insomnia Tension Depression Stress Dissociation Nausea Urticaria Anxiety Disturbance in attention Headache Blood pressure increased Dizziness Vomiting Depressed mood Feeling abnormal Abdominal pain Abdominal pain upper Disorientation Dissociation Feeling jittery Insomnia Rash

Contrave32 (n=584) n(%) 199 (34.1) 139 (23.8) 141 (24.1) 85 (14.6) 64 (11.0) 51 (8.7) 47 (8.0) 30 (5.1) 34 (5.8) 32 (5.5) 31 (5.3) Psychiatric AEs 51 (8.7) 30 (5.1) 14 (2.4) 11 (1.9) 8 (1.4) 6 (1.0) 7 (1.2) 2 (0.3) 3 (0.5) 6 (1.0) AEs resulting in discontinuation 27 (4.6) 10 (1.7) 7 (1.2) 6 (1.0) 5 (0.9) 4 (0.7) 4 (0.7) 4 (0.7) 3 (0.5) 3 (0.5) 3 (0.5) 3 (0.5) 3 (0.5) 3 (0.5) 3 (0.5) 3 (0.5) 3 (0.5)

Placebo (n=200) n(%) 21 (10.5) 35 (17.5) 28 (14.0) 9 (4.5) 13 (6.5) 12 (6.0) 6 (3.0) 7 (3.5) 2 (1.0) 3 (1.5) 1 (0.5) 12 (6.0) 7 (3.5) 6 (3.0) 8 (4.0) 4 (2.0) 2 (1.0) 1 (0.5) 5 (2.5) 4 (2.0) 0 (0) 0 (0) 1 (0.5) 3 (1.5) 0 (0) 1 (0.5) 0 (0) 0 (0) 0 (0) 1 (0.5) 1 (0.5) 0 (0) 0 (0) 0 (0) 0 (0) 0 (0) 0 (0) 0 (0)

Source: Cowen and Company, Orexigen Therapeutics

MEMBER: FINRA/SIPC

77

www.cowen.com

Cowen and Company, LLC

EQUITY RESEARCH

Contraves Final Three Phase III Trials Meet the Mark


In July 2009, Orexigen announced that Contraves three remaining Phase III trials met their co primary endpoints, with 48.0%, 50.5%, and 44.5% of patients on Contrave losing 5% of body weight after 56 weeks in the NB-301, NB-303, and NB-304 trials, respectively, compared to 16.4%, 17.1%, and 18.9% of placebo patients (p<0.001 for all). These data exceeded the FDA primary efficacy benchmark for categorical weight loss.

COR-I (NB-301) Trial


This trial was the largest of the four trials in the Phase III Contrave Obesity Research (COR) program. COR-I was a 56-week trial which evaluated two doses of Contrave in overweight and obese patients. Patients were randomized 1:1:1 to receive daily 32 mg SR naltrexone + 360 mg SR bupropion (Contrave32), 16 mg SR naltrexone + 360 mg SR bupropion (Contrave16), or placebo. The study included a 3-week dose escalation, starting with one-quarter of the full dose, which was increased weekly until the full dose was reached by week 4. Eligibility criteria in the trial included: age 18-65 years, and BMI 30-45 kg/m2 or 27-45 kg/m2 with controlled hypertension and/or dyslipidemia. Participants were instructed on diet and provided advice on lifestyle modification, including exercise instruction, at baseline and weeks 12, 24, 36, and 48. The trial met both its co-primary endpoints: 1) % average weight loss: Contrave patients had an average weight loss of 6.1% in the Contrave32 group, 5.0% in the Contrave16 group, and 1.3% in the placebo group, for placeboadjusted average weight losses of 4.8% and 3.7% in the Contrave32 group and Contrave16 groups, respectively. These results were statistically significant. However, they did not meet FDAs efficacy benchmark of a 5% difference in average weight loss between therapy and placebo. 2) % of patients losing at least 5% of their weight: 48% of the patients in the Contrave32 group lost at least 5% of baseline body weight, compared to 39% in the Contrave16 group and 16% in the placebo group. With these data, Contrave met the threshold specified in FDA's draft guidance for efficacy of obesity drugs.

MEMBER: FINRA/SIPC

78

www.cowen.com

Cowen and Company, LLC

EQUITY RESEARCH

COR-I: Primary efficacy endpoints


COR-I (NB-301) (n=1,742) Key Endpoints n (mITT-LOCF) Mean Weight Loss (%) (mITT-LOCF) 5% weight loss (mITT-LOCF) 10% weight loss (mITT-LOCF) n (Completers) Mean Weight Loss (%) (Completers) 5% weight loss (Completers) 10% weight loss (Completers) Completion rate
Source: Cowen and Company, Orexigen Therapeutics

Contrave16 n=471 5.0% 39.5% 20.2% n=284 6.7% 54.6% 29.9% 60%

Contrave32 n=471 6.1% 48.0% 24.6% n=296 8.1% 61.8% 34.5% 63%

Placebo n=511 1.3% 16.4% 7.4% n=290 1.8% 23.1% 10.7% 57%

Difference (Contrave16) 3.7% 23.1% 12.8% 4.9% 31.5% 19.2% -

p-value p<0.0001 p<0.0001 p<0.0001 p<0.0001 p<0.0001 p<0.0001 -

Difference (Contrave32) 4.8% 31.6% 17.2% 6.3% 38.7% 23.8% -

p=value p=0.0079 p=0.0099 p<0.0001 p<0.0001 p<0.0001 p<0.0001 -

Secondary Endpoints: The Contrave treatment groups also had significant improvements in a number of secondary anthropometric and cardiometabolic parameters compared with the placebo group, including (Contrave32/Contrave16/placebo, p value for Contrave 32 vs. placebo): waist circumference (-6.2 cm vs. -5.0 cm vs. -2.5 cm, p<0.0001); triglycerides (12.7% vs. -8% vs. -3.1%, p<0.0001); hs-CRP (-29% vs. -28% vs. -16.7%, p=0.0076); fasting insulin (-17.1% vs. -11.8% vs. -4.6%, p=0.0007). Patients in the Contrave groups also had improved quality of life scores, as measured by the IWQOL-LITE scale (12.7 vs. 11.7 vs. 8.6, p<0.0001). Blood pressure and heart rate: Mean blood pressure decreased from baseline to 56 weeks in placebo patients. However, both systolic and diastolic blood pressures were unchanged or slightly changed in the Contrave groups. SBP decreased by 0.1 mm Hg and increased by 0.3 mm Hg in the Contrave32 and Contrave16 groups, respectively, compared with a reduction of 1.9 mm Hg in the placebo group. DBP did not change from baseline in the Contrave32 group and increased by 0.1 mm Hg in the Contrave16 group, while placebo patients had a decrease of 0.9 mm Hg. Mean systolic and diastolic blood pressures were noted to increase by approximately 1.5 mm Hg from baseline in the first 8 weeks of Contrave treatment, followed by return to baseline levels after week 12, with subsequent 1 mm Hg decrease for the remaining time in the study. These blood pressure changes were less than those experienced by the placebo group, in which there was a decrease in systolic and diastolic blood pressures of approximately 1.5 mm Hg from baseline during the first 12 weeks and subsequently continued reductions between 1.5 and 3 mm Hg. Heart rates increased up to 1.5 beats per minute from baseline in Contrave patients, while placebo patients maintained baseline heart rates.

MEMBER: FINRA/SIPC

79

www.cowen.com

Cowen and Company, LLC

EQUITY RESEARCH

COR-I: Key Secondary Endpoints

Contrave16 Change in blood fasting glucose Change in insulin Change in HOMA-IR Change in triglycerides Change in hs-CRP Change in IWQOL-LITE Change in waist circumference Change in Systolic blood pressure Change in Diastolic blood pressure Change in Heart rate (beats per minute)
Source: Cowen and Company, Orexigen Therapeutics

Contrave32 -2.6% -17.1% -20.2% -12.7% -29.0% 12.7 -6.2 cm -0.1 mm Hg 0.0 mm Hg +1.0

Placebo -0.7% -4.6% -5.9% -3.1% -16.7% 8.6 -2.5 cm -1.9 mm Hg -0.9 mm Hg -0.1

-1.9% -11.8% -14.3% -8.0% -28.0% 11.7 -5.0 cm 0.3 mm Hg 0.1 mm Hg +1.5

p-value Contrave16 0.1584 0.0628 0.0442 0.0461 0.0159 <0.0001 <0.0001 <0.0001 0.0150 <0.05

p-value Contrave32 0.0104 0.0007 0.0003 <0.0001 0.0076 <0.0001 <0.0001 0.0008 0.0217 <0.05

Safety and Tolerability: The most common adverse event in the Contrave groups was nausea. There was no increase in treatment-emergent psychiatric adverse events, such as depression, suicidality, or other mood-related issues. The rate of serious adverse events did not differ significantly between groups (1.6% for both Contrave groups and 1.4% for placebo). There were no seizures or serious hypertension events. There were two serious cardiovascular adverse events in the Contrave groups, one cardiac failure and one death resulting from MI, but investigators did not consider these events to be treatment-related. COR-I: Safety Data
Contrave16 (n=569) n(%) 1.60% 455 (800%) 155 (272%) 91 (160%) 90 (158%) 49 (86%) 44 (77%) 36 (63%) 36 (63%) 34 (60%) 42 (74%) 32 (56%) 31 (54%) 13 (23%) 76 (134%) 36 (63%) 12 (21%) 9 (16%) Contrave32 (n=573) n(%) 1.60% 476 (831%) 171 (298%) 79 (138%) 90 (157%) 57 (99%) 54 (94%) 43 (75%) 56 (98%) 30 (52%) 43 (75%) 29 (51%) 26 (45%) 30 (52%) 85 (148%) 43 (75%) 9 (16%) 3 (05%) Placebo (n=569) n(%) 1.40% 390 (685%) 30 (53%) 53 (93%) 32 (56%) 64 (112%) 15 (26%) 29 (51%) 14 (25%) 34 (60%) 11 (19%) 31 (54%) 28 (49%) 7 (12%) 62 (109%) 29 (51%) 12 (21%) 6 (11%)

Serious adverse events Participants reporting any adverse event Nausea Headache Constipation Upper respiratory tract infection Dizziness Insomnia Vomiting Sinusitis Dry mouth Nasopharyngitis Diarrhoea Hot fl ush Participants reporting any psychiatric adverse event Insomnia Anxiety Depression

Source: Cowen and Company, Orexigen Therapeutics

COR-II (NB-303) Trial


This was a 56-week study in obese and overweight patients with co-morbidities. Patients were randomized 2:1 to daily 32 mg SR naltrexone + 360 mg SR bupropion (Contrave32) or

MEMBER: FINRA/SIPC

80

www.cowen.com

Cowen and Company, LLC

EQUITY RESEARCH

placebo. The Contrave dose was escalated weekly over the first 3-4 weeks, with full dose reached by the beginning of week 5. To evaluate efficacy and safety of a dose increase in patients with sub-optimal response, Contrave32 patients with < 5% weight loss between weeks 28 and 44 were randomized again double-blind, 1:1 to continue on Contrave32 or escalate to 48 mg SR naltrexone + 360 mg SR bupropion (Contrave48). The two co-primary endpoints of the study were mean weight loss at week 28, and proportion of patients with 5% weight loss at week 28. Eligibility criteria in the trial included: age 18-65 years, and BMI 30-45 kg/m2 or 2745 kg/m2 with controlled hypertension and/or dyslipidemia. Participants were instructed on diet and provided advice on lifestyle/behavioral modification, including exercise instruction, at baseline and weeks 12, 24, 36, and 48. The trial met both its co-primary endpoints: 1) % average weight loss at week 28: At week 28, Contrave32 patients had an average weight loss of 6.5%, compared with 1.9% in the placebo group. This significant difference was maintained at 56 weeks, with Contrave patients losing 6.4% of body weight and placebo patients losing 1.2%, for a placebo-adjusted average weight loss of 5.2%. These results were statistically significant. The 56-week result did meet FDAs efficacy benchmark of a 5% difference in average weight loss between therapy and placebo. 2) % of patients losing at least 5% of their weight: At week 28, 55.6% of the patients in the Contrave32 group lost at least 5% of baseline body weight, compared to 17.5% in the placebo group. This significant difference was also maintained at 56 weeks, with 50.5% of Contrave patients losing 5% of body weight, compared with 17.1% of placebo patients. With these data, Contrave met the categorical weight loss threshold specified in FDA's draft guidance for efficacy of obesity drugs. COR-II: Efficacy Data
COR-II (NB-303) (n=1,496) Key Endpoints n (mITT-LOCF) Mean Weight Loss (%) (mITT-LOCF) 5% weight loss (mITT-LOCF) 10% weight loss (mITT-LOCF) n (Completers) Mean Weight Loss (%) (Completers) 5% weight loss (Completers) 10% weight loss (Completers) Completion rate Contrave32 n=825 6.5% 55.6% 27.3% n=619 7.8% 68.8% 35.7% 75% Placebo n=456 1.9% 17.5% 7.0% n=319 2.4% 22.3% 9.4% 70% Difference 4.6% 38.1% 20.3% 5.4% 46.5% 26.3% p-value p<0.001 p<0.001 p<0.001 p<0.001 p<0.001 p<0.001 Contrave32 n=702 6.4% 50.5% 28.3% n=434 8.2% 64.9% 39.4% 62% Placebo 56 weeks n=456 1.2% 17.1% 5.7% n=267 1.4% 21.7% 7.9% 59% 5.2% 33.4% 22.6% 6.8% 43.2% 31.5% p<0.001 p<0.001 p<0.001 p<0.001 p<0.001 p<0.001 Difference p-value

28 weeks (timepoint for primary endpoint analysis)

Source: Cowen and Company, Orexigen Therapeutics

Secondary Endpoints: As in the other trials in the COR program, Contrave treatment also resulted in improvements of a number of secondary anthropometric and cardiometabolic parameters, such as waist circumference, triglycerides, fasting insulin, and HOMA-IR. There

MEMBER: FINRA/SIPC

81

www.cowen.com

Cowen and Company, LLC

EQUITY RESEARCH

was similarly also improvement in quality of life with Contrave treatment versus placebo, as measured by the IWQOL-Lite score, including physical function self-esteem, and sexual life. Blood pressure and heart rate: At week 56, mean SBP was elevated by 0.6 mm Hg in the Contrave group, but decreased by 0.5 mm Hg in the placebo group. Mean DBP was elevated in both groups, by 0.4 mm Hg and 0.3 mm Hg, respectively. Heart rate remained unchanged with placebo, and a +1 bpm increased rate was noted in the Contrave group. COR-II: Key Secondary Endpoints

Contrave32 Change in blood fasting glucose Change in insulin Change in HOMA-IR Change in triglycerides Change in hs-CRP Change in IWQOL-LITE Change in waist circumference Change in Systolic blood pressure Change in Diastolic blood pressure Change in Heart rate (beats per minute)
Source: Cowen and Company, Orexigen Therapeutics

-2.1 mg/dL -14.1% -16.4% -7.3% -9.4% 9.9 -6.2 cm -0.9 (mm Hg) +0.2 (mm Hg) -

Placebo 28 weeks -1.7 mg/dL -0.5% -4.2% -1.4% -1.1% 6.2 -2.7 cm -1.2 (mm Hg) -0.7 (mm Hg) -

p-value 0.0544 <0.001 <0.001 0.007 0.091 <0.001 <0.001 0.556 0.017 -

Placebo p-value 56 weeks -2.8 mg/dL -1.3 mg/dL 0.051 -11.4% 3.5% <0.001 -13.8% 1.2% <0.001 -9.8% -0.5% <0.001 -28.8% -8.3% <0.001 10.9 6.4 <0.001 -6.7 cm -2.1 cm <0.001 +0.6 (mm Hg) -0.5 (mm Hg) 0.039 +0.4 (mm Hg) +0.3 (mm Hg) 0.847 +0.8 -0.3 <0.05

Contrave32

Safety and tolerability: Consistent with the other COR trials, the most common treatmentemergent AEs were nausea, headache, and constipation, which were usually mild to moderate in intensity. There were no differences between the Contrave and placebo groups in psychiatric measures, such as sadness, irritability, tension, and suicidality. Contrave was not associated with increased depression or other mood-related AEs. Serious AE rates were similar between the Contrave (2.1%) and placebo (1.4%) groups. In the Contrave group, there was one MI in a patient with CV history and one seizure.

MEMBER: FINRA/SIPC

82

www.cowen.com

Cowen and Company, LLC

EQUITY RESEARCH

COR-II: Safety Data


Contrave32 n=992 85.9% 29.2% 19.1% 17.5% 9.8% 9.1% 8.7% 8.5% 8.3% 6.9% 5.5% 5.1% 3.8% 1.4% 20.7% 9.8% 4.8% 1.3% 1.1% 24.3% 6.0% 2.6% 0.5% +0.20 +00 +0.8 Placebo n=492 75.2% 6.9% 7.1% 8.7% 6.7% 2.6% 11.2% 2.0% 8.1% 3.7% 3.7% 7.1% 5.7% 5.1% 15.2% 6.7% 4.3% 1.6% 0.8% 13.8% 0.2% 0.8% 1.2% -0.40 +0.10 -0.30

Participants (%) reporting any adverse event Nausea Constipation Headache Insomnia Dry mouth Upper respiratory Vomiting Nasopharyngitis Dizziness Diarrhea Sinusitis Arthralgia Bronchitis Participants (%) reporting any psychiatric AEs Insomnia Anxiety Depression Sleep disorder Participants (%) reporting any adverse event leading to discontinuation Nausea Headache Depression Cardiovascular endpoints Systolic blood pressure, change from baseline (mm Hg) Diastolic blood pressure, change from baseline (mm Hg) Heart rate, change from baseline to Week 56 (bpm)

Source: Cowen and Company, Orexigen Therapeutics

COR-Diabetes (NB-304) Trial


This was a 56-week trial in obese and overweight patients with type 2 diabetes. Patients were randomized 2:1 to daily 32 mg SR naltrexone + 360 mg SR bupropion (Contrave32) or placebo. Eligibility criteria in the trial included: age 18-70 years, and BMI 27-45 kg/m2 with type 2 diabetes mellitus, on no injectable or inhaled insulin for more than 3 months, and with HbA1c levels between 7%-10%. Patients were permitted to be on oral single or combination diabetes medications, stable for at least 3 months before randomization, or on no medications. Participants were instructed on diet and provided advice on lifestyle/behavioral modification, including exercise instruction, at baseline and weeks 12, 24, 36, and 48. The trial met both its co-primary endpoints: 1) % average weight loss: At week 56, Contrave patients had an average weight loss of 5%, compared with 1.8% in the placebo group, for a placebo-adjusted average weight loss of 3.2%. While statistically significant, this result did not meet FDAs efficacy benchmark of a 5% difference in average weight loss between therapy and placebo. 2) % of patients losing at least 5% of their weight: 44.5% of the patients in the Contrave32 group lost at least 5% of baseline body weight, compared to 18.9% in the placebo group. With

MEMBER: FINRA/SIPC

83

www.cowen.com

Cowen and Company, LLC

EQUITY RESEARCH

these data, Contrave met the categorical weight loss threshold specified in FDA's draft guidance for efficacy of obesity drugs. COR-Diabetes: Efficacy Data
Key Endpoints n (mITT-LOCF) Mean Weight Loss (%) (mITT-LOCF) 5% weight loss (mITT-LOCF) 10% weight loss (mITT-LOCF) n (Completers) Mean Weight Loss (%) (Completers) 5% weight loss (Completers) 10% weight loss (Completers) Completion rate Contrave32 n=265 5.0% 44.5% 18.5% n=175 5.9% 53.1% 26.3% 66% COR-Diabetes (NB-304) (n=505) Placebo Difference n=159 1.8% 3.2% 18.9% 25.6% 5.7% 12.8% n=100 2.2% 24.0% 8.0% 62% 3.7% 29.1% 18.3% p-value p<0.001 p<0.001 p<0.001 p<0.001 p<0.001 p<0.001 -

Source: Cowen and Company, Orexigen Therapeutics

Secondary Endpoints: Anthropometric and cardiometabolic parameters were generally improved with Contrave, similar to the other COR trials. In this diabetic population, glycemic control was improved with Contrave treatment. Contrave patients with baseline HbA1c > 8% showed a reduction of 1.1%, compared with 0.5% reduction in placebo patients (p<0.01). More than 44% of Contrave patients reached the American Diabetes Association treatment target of HbA1c level < 7%, compared with 26% of placebo patients (p<0.001). COR-Diabetes: Key Secondary Endpoints
Change in HbA1c Proportion of subjects with HbA1c <7% Change in triglycerides Change in hs-CRP Change in IWQOL-LITE Change in waist circumference
Source: Cowen and Company, Orexigen Therapeutics

Contrave32 -0.6% 44.1% -11.2% -20.9% 9.3 -5 cm

Placebo -0.1% 26.3% -8.0% -13.3% 7.9 -2.9 cm

p-value <0.001 <0.001 0.007 0.312 0.208 <0.006

Safety and tolerability: Consistent with the other COR trials, the most common treatmentemergent AEs included nausea and constipation. The overall safety profile was consistent with that demonstrated in other COR trials. The incidence of hypoglycemia was similar between the groups.

MEMBER: FINRA/SIPC

84

www.cowen.com

Cowen and Company, LLC

EQUITY RESEARCH

COR-Diabetes: Safety Data

Participants (%) reporting any AEs Nausea Vomiting Constipation Diarrhea Headache Dizziness Insomnia Hypertension Hypoglycemia Tremor Dry mouth Anxiety Upper respiratory
Source: Cowen and Company, Orexigen Therapeutics

Contrave 32 n=333 90.4% 42.3% 18.3% 17.7% 15.6% 13.8% 11.7% 11.1% 9.9% 7.5% 6.6% 6.0% 5.4% 5.1%

Placebo n=169 85.2% 7.1% 3.6% 7.1% 9.5% 8.9% 5.3% 5.3% 4.1% 7.1% 2.4% 3.0% 1.2% 1.8%

Responders Maintained Meaningful Weight Loss


Approximately 51% and 67% of the patients in the COR program and COR-BMOD trial, respectively, met the definition of responder by achieving >5% weight loss at week 16. Responders achieved meaningful weight losses of 11.3% and 13.4%, respectively, after one year of treatment. In the COR program, 19% of placebo patients met the definition of responder, and these patients had an average loss of 8.6% at one year. COR Program responders

COR Program (n=1,038) >5% weight loss at 16 weeks % Meeting responder definition % Weight loss at 1 year
Source: Cowen and Company, Orexigen

Placebo Responder (n=254) 19.0% 8.6%

51.0% 11.3%

Similarly, approximately 62% and 73% of the patients in the COR program and COR-BMOD trial, respectively, met the responder definition of achieving >3% weight loss at week 12. This patient population achieved meaningful weight loss of 10.3% and 12.7%, respectively, in the trials after one year of treatment. We believe that only responders will continue treatment with Contrave after 16 weeks. The representative weight loss in the Contrave responder patient population is similar to that achieved by treatment with either Qsymia or BELVIQ.

MEMBER: FINRA/SIPC

85

www.cowen.com

Cowen and Company, LLC

EQUITY RESEARCH

COR-I, COR-II, COR-Diabetes And COR-BMOD: Responder Analysis

COR Program (n=1,038) >5% weight loss at 16 weeks % Meeting responder definition % Weight loss at 1 year 51.0% 11.3%

COR BMOD (n=322) 67.0% 13.4% COR BMOD (n=352) 73.0% 12.7%

>3% weight loss at 12 weeks COR Program (n=1,268) % Meeting responder definition 62.0% % Weight loss at 1 year
Source: Cowen and Company, Orexigen

10.3%

We View Contraves Tolerability in the COR Phase III Program as Acceptable


The discontinuation rates due to adverse events across the Phase III COR program ranged from 19% to 29% for patients treated with Contrave, compared with 10% to 15% for patients treated with placebo. In the double-blind treatment phase, the discontinuation rates related to adverse events were 23.8% for the Contrave arm vs. 11.9% for the placebo arm. The most common adverse events leading to treatment discontinuation were nausea (6.3% Contrave vs. 0.2% placebo), headache (1.7% Contrave vs. 0.6% placebo), dizziness (0.9% Contrave vs. 0.3% placebo), and vomiting (1.1% Contrave vs. < 0.1% placebo). Our consultants were satisfied with these data. They described the dropout rate as fair and in line with that seen in many obesity trials. They consider the major adverse events nausea, dizziness, and vomiting to be tolerability issues and not safety problems, of which the rates and severity are acceptable. In fact, our consultants feel it would be possible that the rates could decrease if patients up-titrated their Contrave dose over time. Therefore, they are optimistic that Contrave would become more tolerable as patients and physicians better learn how to use it.

MEMBER: FINRA/SIPC

86

www.cowen.com

Cowen and Company, LLC

EQUITY RESEARCH

Safety and Tolerability Data From Contrave's Four Phase III COR Trials

Treatment-Emergent Adverse Events Discontinue treatment due to any AE Nausea Headache Dizziness Vomiting Insomnia Anxiety Urticaria Depression Blood pressure increased Rash Hypertension Fatigue Palpitations Abdominal pain Tremor
Source: Cowen and Company, Orexigen

Placebo (N=1,515) 181 (11.9%) 3 (0.2%) 9 (0.6%) 5 (0.3%) 1 (<0.1%) 7 (0.5%) 10 (0.7%) 4 (0.3%) 13 (0.9%) 3 (0.2%) 2 (0.1%) 0 3 (0.2%) 0 1 (<0.1%) 1 (<0.1%)

Contrave Contrave (naltrexone 16mg) (naltrexone 32mg) (N=633) (N=2,545) 139 (22.0%) 32 (5.1%) 10 (1.6%) 15 (2.4%) 4 (0.6%) 5 (0.8%) 1 (0.2%) 3 (0.5%) 6 (0.9%) 3 (0.5%) 1 (0.2%) 3 (0.5%) 3 (0.5%) 5 (0.8%) 5 (0.8%) 3 (0.5%) 612 (24.0%) 160 (6.3%) 43 (1.7%) 23 (0.9%) 28 (1.1%) 17 (0.7%) 19 (0.7%) 16 (0.6%) 10 (0.4%) 10 (0.4%) 12 (0.5%) 7 (0.3%) 7 (0.3%) 4 (0.2%) 4 (0.2%) 6 (0.2%)

Contrave (combined 16/32 data) (N=3,239) 771 (23.8%) 203 (6.3%) 55 (1.7%) 42 (1.3%) 35 (1.1%) 23 (0.7%) 21 (0.6%) 19 (0.6%) 16 (0.5%) 13 (0.4%) 13 (0.4%) 11 (0.3%) 10 (0.3%) 10 (0.3%) 9 (0.3%) 9 (0.3%)

p-value <0.001 <0.001 0.002 0.001 <0.001 0.290 0.843 0.198 0.144 0.276 0.125 0.021 0.518 0.031 0.157 0.141

Summary of Contraves Regulatory History The AdCom gives Contrave a surprise 13-7 positive nod
The EMDAC (Endocrinologic and Metabolic Drugs Advisory Committee) of the FDA met to consider the Contrave NDA on December 7, 2010. Given that the AdCom had previously voted against the approvals of both BELVIQ and Qsymia, there was low investor expectation for a positive vote. Thus, the 13-7 vote in favor of Contrave approval for treatment of obesity came as a surprise. While panel members viewed Contraves efficacy as modest, they considered it to meet FDA requirements for weight loss efficacy. In view of the modest benefit, a key issue was Contraves safety, with CV safety being the greatest concern to advisory panel members. In the COR Phase III program, overall mean elevations in BP and HR were associated with Contrave treatment relative to placebo. Both the FDA and Orexigen agreed that a cardiovascular outcomes trial (CVOT) would be required. However, the key question revolved around the timing requirement of the study, specifically before or after potential approval. There was debate between panel members considering the CV risk as low and definable in a quickly conducted trial and those worried about the potential need to withdraw another obesity drug from the market after approval, with resulting loss of FDA credibility. The panel ultimately voted 11-8 for CVOT completion post-approval.

MEMBER: FINRA/SIPC

87

www.cowen.com

Cowen and Company, LLC

EQUITY RESEARCH

Contraves CV Safety Data


C o n tr a v e ( n a ltr e x o n e 1 6 m g ) (n=276) C o n tr a v e ( n a ltr e x o n e 3 2 m g ) ( n = 1 ,3 1 2 ) C o n tr a v e ( c o m b in e d 1 6 /3 2 d a ta ) ( n = 1 ,5 8 8 ) P la c e b o (n=751)

S y s to l i c b l o o d p re s s u re , m m H g M e an c hang e D i ffe re n c e fro m p l a c e b o (m m H g ) D i a s to l i c b l o o d p re s s u re , m m H g M e an c hang e D i ffe re n c e fro m p l a c e b o (m m H g ) -0 . 3 7 + 1.13 -1 . 1 4 + 0.36 -1 . 0 1 + 0.51 -1 . 5 -0 . 2 4 + 2.09 -0 . 9 4 + 1.39 -0 . 8 2 + 1.54 -2 . 3 3

H e a rt ra te , b p m M e an c hang e D i ffe re n c e fro m p l a c e b o (b p m ) + 1.3 + 2.28 + 0.1 + 1.07 + 0.3 + 1.35 -0 . 9 8

Source: Cowen and Company, Orexigen

but FDA goes 3-for-3 in rejecting these new obesity drugs


Despite the positive opinion from the AdCom, in February 2011, Orexigen announced that it had received a Complete Response Letter (CRL) from the FDA, noting " concern about the cardiovascular safety profile of naltrexone/bupropion when used long term in a population of overweight and obese subjects." The CRL requested the completion of a pre-approval CV safety study by Orexigen and specifically stated, "before your application can be approved, you must conduct a randomized, double-blind placebo-controlled trial of sufficient size and duration to demonstrate that the risk of major adverse cardiovascular events in overweight and obese subjects treated with naltrexone/bupropion does not adversely affect the drug's adverse event profile." In May 2011, Orexigen met with FDAs Division of Metabolic and Endocrinologic Products (DMEP) to gain more clarity on the CRL. Orexigen proposed Contrave approval with a narrow indication, in patients with lower cardiovascular risk, until data from a proposed CVOT could be reviewed for label expansion. The DMEP rejected this request and advised Orexigen that its proposed CVOT would not adequately address cardiovascular issues. The DMEP further stated that it would not consider approval of Contrave for a narrow indication without first reviewing CVOT data. Based on this feedback, Orexigen decided to put further clinical development of Contrave on hold in the United States. After a dispute resolution process with the FDA, Orexigen announced in September 2011 that an agreement had been reached on the design requirement for the CVOT, addressing concerns specified in the CRL. According to the company, the FDA stated that if the interim analysis meets the specified criteria to exclude an unacceptable increased cardiovascular (CV) risk, the drug could be approved. This signaled a reversal from the previous FDA stance that CV risk be ruled out, which would have required a much larger trial. In agreeing with the terms, Orexigen considered the design requirements to be reasonable and feasible.

MEMBER: FINRA/SIPC

88

www.cowen.com

Cowen and Company, LLC

EQUITY RESEARCH

Contraves CVOT: LIGHT Study


The LIGHT Study is the randomized, double-blind, placebo-controlled cardiovascular outcomes trial (CVOT) for Contrave. The trial is being conducted under a Special Protocol Assessment (SPA) with the FDA. FDA requirements for the trial design include powering based on an intent-to-treat (ITT) analysis. The trial must enroll an overweight and obese population which has an estimated annual risk of 1% to 1.5% for major adverse cardiovascular events (MACE), defined as: 1) cardiovascular death, 2) non-fatal myocardial infarction, and 3) non-fatal stroke. LIGHT Study Statistical Criteria: Specific statistical criteria have been issued for the interim and final analyses of CV risk in this population: The upper bound of the 95% confidence interval (CI) should exclude a hazard ratio (HR) of 2.0 at the interim analysis, and of 1.4 at the final analysis . With this design, there is an estimate of at least 87 total required MACE events at the interim analysis, and of at least 371 MACE events at the final analysis. The company has previously indicated that, at interim analysis, in order to exclude a HR of 2.0 from the confidence interval, the observed HR in the trial must be <1.32, and at final analysis, the observed HR must be <1.14. LIGHT Study Design: Patients are randomized 1:1 to treatment with Contrave, administered in conjunction with a comprehensive weight management program, or to placebo administered with the same program. The dose of Contrave is escalated weekly over 4 weeks up to the full dose. Lead-in Period: There is a two-week lead-in period after screening and prior to the treatment period. For the lead-in, patients are randomized 1:1 to receive either placebo during the first week and the lowest daily dose of Contrave (8 mg naltrexone/90 mg bupropion) during the second week, or the lowest Contrave dose during the first week and placebo during the second week. Patients will be trained on using food diaries, which they are expected to consistently complete during this time. Orexigen has guided that the purpose of the lead-in period is two-fold: 1) to provide subjects with a small dose of Contrave, in order to stimulate nausea and allow patients unable to tolerate this known side-effect of naltrexone to drop out, and 2) to evaluate subjects commitment to a clinical trial by assessing compliance with food diaries. The company has noted its intention to avoid discontinuations resulting from nausea during LIGHTs treatment period, which occurred with a 15-20% incidence and a usual duration of ~2 weeks, in the Phase III Contrave trial program. Subjects who discontinue during the lead-in period will not be counted as part of the ITT population. Evaluation to Continue Treatment: Patients will be evaluated at Week 16 to determine continuation in the study. Patients who 1) have not lost 2% of body weight by week 16 and/or 2) experience a sustained increase in either systolic or diastolic BP, defined as multiple observations of at least 10 mm Hg increases in BP on two successive readings, any time

MEMBER: FINRA/SIPC

89

www.cowen.com

Cowen and Company, LLC

EQUITY RESEARCH

during the 16-week period will discontinue treatment. However, as part of the ITT population, their MACE events, if any, will be included in the final study analysis . LIGHT Study Design

Source: Orexigen Presentation

Orexigen expects that, after Week 16, approximately 50% of patients in both trial arms will have discontinued study treatment, either resulting from standard drop-outs or from meeting criteria for discontinuation at 16 weeks. These patients will be counted in the ITT population. Primary Endpoint: The primary endpoint of the trial is time from treatment period randomization to the first confirmed occurrence of MACE events. The time frame extends from Day 1 to first confirmed occurrence, assessed up to 4 years. LIGHT Study Population: Inclusion and exclusion criteria for the trial reflect the target background annual 1%-1.5% CV event risk. Using U.S. population-based data sets, for example NHANES, Orexigen targeted enrollment of a patient population with a modeled background annual MACE rate of > 2%, expecting to observe a MACE rate of about 1.5%. The trial population included females 50 years old and males 45 years old, with BMI range 27 50 kg/m2. Subjects were required to have 1) cardiovascular disease; or 2) diabetes mellitus, with at least two of four additional risk factors; or 3) both: CV disease was identified by at least one of the following: history of MI >3 months prior to screening, history of coronary/carotid/peripheral revascularization procedures, angina with ischemia, evidence of peripheral vascular disease, or 50% arterial stenosis (coronary/carotid/lower extremity vessel) within prior 2 years. Otherwise, instead of, or in addition to, CV disease, patients had type 2 diabetes mellitus accompanied by at least two of the following conditions: controlled hypertension with BP<145/95mmHg, dyslipidemia requiring pharmacotherapy, low HDL cholesterol (< 50 mg/dL in women or < 40 mg/dL in men), or current smoking history.

MEMBER: FINRA/SIPC

90

www.cowen.com

Cowen and Company, LLC

EQUITY RESEARCH

Particularly high-risk individuals, such as those with MI within 3 months, unstable angina, stroke history, tachyarrhythmia history, planned bariatric surgery/cardiac surgery/coronary angioplasty, or life expectancy < 4 years, were excluded. The following table illustrates the characteristics of 8,911 patients enrolled in LIGHT. Patient Characteristics in LIGHT Study

Characteristics Age Younger age group Gender Minority CV disease Diabetes Smoking status

Target 64 <25% 50:50 (M:F) 15% 30% >70% >5%

Randomized subjects (n=8,911) 61 14% 45:55 (M:F) 22% 34% 85% 9%

Source: Cowen and Company, Orexigen Presentation August 2013

LIGHT Study Enrollment: The trial began enrolling patients on June 6, 2012. On September 5, 2012, Orexigen announced that the LIGHT study was enrolling patients more quickly than anticipated, with > 4,500 patients enrolled as of August 31, 2012. On December 18, 2012, the company announced the completion of screening and enrollment, accomplished within 6.5 months of initiation, indicating that 13,192 patients had been screened for the study, of which approximately 10,400 patients met eligibility criteria, were enrolled, and entered the lead-in period. Ultimately 8,911 patients were then randomized to the treatment period. The company has stated that the final population in the trial is sufficient for and reflective of the targeted 11.5% annual MACE rate. LIGHT Trial Enrollment Faster Than Expected

Source: Orexigen Presentation

MEMBER: FINRA/SIPC

91

www.cowen.com

Cowen and Company, LLC

EQUITY RESEARCH

Faster Path to Resubmission of the Contrave NDA


On January 7, 2013, Orexigen announced further progress with regard to the potential resubmission of the Contrave NDA. The FDA will allow Orexigen to resubmit the NDA based on the Data Monitoring Committee (DMC) summary report of the LIGHT interim analysis, and to submit the complete clinical study report within 60 days. Interim analysis from the study is expected by early December 2013, with plan for NDA resubmission by YE13. 2Q13 Earnings Call Update: On August 6, 2013, Orexigen management reiterated guidance that the LIGHT trial interim analysis and resubmission of the Contrave NDA (Class 2 resubmission, with a 6-month review process expected) were expected in 2H13. According to management, the LIGHT trial timeline was on track, and the 87th MACE event was expected within the next few months. More information on this, with a more specific time frame, was anticipated after the next DMC meeting, which was expected to occur pretty soon. The DMC would notify Orexigen when the number of events was close to the target, so that the company could finalize preparations for the interim analysis. August 2013 Update: On August 27, 2013, Orexigen announced the DMCs confirmation that sufficient MACE events for interim analysis in LIGHT are expected to occur within two months, meaning that the DMC expects the 87th MACE event will occur within 8 weeks. Therefore, the company expects the interim analysis of the LIGHT study to have been conducted by early December 2013. Management guided that the Contrave NDA will then be resubmitted by YE13 (recall that the FDA is allowing resubmission based on the DMC summary report of the LIGHT interim analysis, and will accept the complete clinical study report within 60 days). Additionally, the company reiterated that the MAA for Contrave will be submitted prior to the interim analysis of LIGHT, with the expectation that data will be ready for the CHMP Day 120 List of Questions.

Preparing For the Launch


The Ignite Study: In preparation for the launch of Contrave, Orexigen has initiated a multicenter, randomized, open-label, controlled Phase IIIb method-of-use study designed to provide additional information regarding the real world weight loss potential of Contrave. This trial will assess the effects of Contrave used in conjunction with a commercially available, comprehensive lifestyle intervention (CLI) program versus Usual Care, which is defined as a self-directed, minimal lifestyle intervention program. The CLI program is a telephone-based, interactive system with counseling, education, goal setting, and tracking tools. The study is evaluating Contrave utilization in the manner intended after approval. In February 2013, the trial completed enrollment (in 3 days) of approximately 240 adult patients with BMI 30-45 kg/m2 or BMI 27-45 kg/m2 with dyslipidemia and/or controlled hypertension. Diabetic patients are excluded. At week 26, patients originally assigned to the Usual Care arm will switch to the Contrave/CLI arm. Those in the Contrave/CLI arm continue on treatment for the duration of the study, which is 78 weeks. The primary endpoint is change in body weight after 26 weeks. Secondary endpoints include categorical weight loss and changes in

MEMBER: FINRA/SIPC

92

www.cowen.com

Cowen and Company, LLC

EQUITY RESEARCH

anthropometric and cardiometabolic parameters. Results from the study are anticipated by YE13.

E.U. Regulatory Front


In 4Q12, Orexigen submitted a MAA Letter of Intent (LOI) with the EMA. Orexigen met with the EMA in 1H13 for pre-submission discussions. On May 8, 2013, the company reported that the rapporteurs had been assigned. Status Update: On August 6, 2013, in its 2Q13 Earnings Call, Orexigen reported that, having met with the rapporteur (Denmark) and co-rapporteur (U.K.) for discussions about filing, it planned to submit the MAA for Contrave to the EMA using the centralized procedure in 2H13 (in the next few months), before the LIGHT trial interim analysis. The company plans to have CVOT data from LIGHT available for the CHMP Day 120 List of Questions. According to management, the rapporteurs are supportive of this plan for MAA submission. The company noted that the pediatric investigational plan (PIP) has been accepted, and the risk management plan is currently being refined before submission (as per management, a key part of this involves LIGHT, but may also include commitments for other components, such as a registry program). Orexigen anticipates potential E.U. approval in Fall 2014, should there be a positive CHMP opinion.

Orexigen Partnered with Takeda in North America and Mexico


In September 2010, Orexigen and Takeda announced a partnership to commercialize Contrave in the U.S., Canada, and Mexico. Orexigen received an upfront payment of $50M and is eligible for up to $1B in additional milestone payments, as well as tiered royalties of 20%-35% on net sales. Orexigen will be responsible for all pre-approval development costs, the first $60M of post-approval development costs, and 25% of most post-approval development costs. Takeda will be responsible for all post-approval manufacturing and commercialization costs, plus 75% of post-approval development costs. Orexigen retains the right to co-promote Contrave in North America and Mexico and intends to out-license ex-North American rights for Contrave to a pharma partner.

MEMBER: FINRA/SIPC

93

www.cowen.com

Cowen and Company, LLC

EQUITY RESEARCH

Orexigen/Takeda partnership summary


Partner Geographies Partnership date Upfront payment Royalty rate Next expected milestone Total potential milestones Patent life
Source: Cowen and Company, SEC Filings

Takeda Takeda owns rights to U.S., Canada and Mexico; Orexigen owns rights to rest of the wold Partnered with Takeda in September 2010 $50M 20-35% $20M on U.S. approval, $10M upon the delivery of launch supplies to Takeda and $70M on first commercial sales. up to $1B US patents expire 2024-2025

Sales and Marketing Efforts


Orexigen has not yet disclosed the exact details of the sales and marketing effort that Takeda will implement for Contrave. However, according to Orexigen, the companies have a multi-year contractual commercial commitment, outlining specific agreements on marketing spend, sales force compensation structure, and the minimum number of prescribers that Takeda will be targeting. Orexigen has guided that Takeda expects to reach between 50,000+ prescribers. In December 2012, at an Analyst Day, Orexigen discussed a theoretical scenario in which a sales force of 500 representatives, having an average frequency of 12 calls a year per physician, could target a total of 60,000 physicians in any given year. The figure below illustrates a potential reaching frequency of a representative primary care launch. We believe Takedas marketing strategy with its commitment of a sizeable primary care sales force provides Contrave with a significant commercial advantage, enabling comprehensive outreach to potential physician prescribers. The projected sales force for Contrave is substantial, generating a potential advantage relative to other products in the market.

Orexigens IP Estate
Contrave is covered by an intellectual property estate for which Orexigen has exclusive rights. The compositions of matter for Contrave as well as its use for obesity, will be protected in the U.S. and abroad until at least 2024-25. Orexigen holds U.S. Patents # 7,375,111 and 7,462,626, entitled Compositions For Affecting Weight Loss. As stated in the patent abstracts, Disclosed are compositions for affecting weight loss comprising a first compound and a second compound, where the first compound is an opioid antagonist and the second compound causes increased agonism of a melanocortin 3 receptor (MC3-R) or a melanocortin 4 receptor (MC4-R) compared to normal physiological conditions. Also disclosed are methods of affecting weight loss, increasing energy expenditure, increasing satiety in an individual, or suppressing the appetite of an individual, comprising

MEMBER: FINRA/SIPC

94

www.cowen.com

Cowen and Company, LLC

EQUITY RESEARCH

identifying an individual in need thereof and treating that individual to antagonize opioid receptor activity and to enhance -MSH activity. The company refers to these patents as the Weber/Cowley composition patent and the Weber/Cowley methods patent, respectively. Collectively, these patents are known as the Weber/Cowley patents and cover the current composition of Contrave and methods of administering Contrave to treat obesity. The composition patent expires in March 2025, and the methods patent expires in July 2024. The European Patent Office has granted the European version of the Weber/Cowley patents, published as EP1617832 B1. This patent has been issued in numerous countries throughout the European Union, providing coverage for Contrave until at least 2024. Orexigen has also filed a number of international patent applications in foreign countries.

MEMBER: FINRA/SIPC

95

www.cowen.com

Cowen and Company, LLC

EQUITY RESEARCH

3) Big Pharma: Novo Nordisks Victoza (liraglutide)


Novo Nordisks Victoza (liraglutide) has been approved by the FDA (with a black box warning about possible increased risk of thyroid cancer) and by the EMA for the treatment of diabetes at doses of 1.2 mg and 1.8 mg (and 0.9 mg in Japan) once daily. Liraglutide is a long-acting analog of glucagon-like peptide-1 (GLP-1), which increases insulin secretion and decreases glucagon secretion. It also slows gastric emptying. The drug is in Phase III development for the treatment of obesity in the non-diabetic setting. Like other GLP-1 analogs, such as Amylins Byetta, Victoza has demonstrated an ability to generate weight loss in its trials. In a 165-subject dose-ranging Phase IIb monotherapy study in type 2 diabetics, patients at the highest dose of Victoza (1.8 mg) lost approximately 3 kg from baseline vs. 1.2 kg on placebo after 14 weeks. Data from the 533-patient LEAD-4 trial, part of Victozas Phase III diabetes program, demonstrated that patients treated with a combinatio n of Victoza, metformin, and Avandia experienced 2.5 kg weight loss at 26 weeks, compared to patients administered metformin and Avandia. In 2007, Victoza initiated a 550-patient, dose-finding, Phase II trial in obese subjects without diabetes. This 20-week, double-blind study randomized participants to either Victoza, placebo, or an open-label Xenical arm. Baseline BMI was 30-40 kg/m2, and the protocol included a lowfat diet and exercise program. In November 2007, Novo Nordisk released positive results from the study, demonstrating that patients who received Victoza at the highest dose had weight loss from baseline of >7 kg versus weight loss of <3 kg in the placebo group and weight loss of >4 kg in the Xenical treated group. All doses of Victoza reduced body weight. More than 75% of the people treated with the highest dose experienced a weight loss greater than 5%, and more than 25% experienced weight loss greater than 10% relative to their body weight at randomization. The study revealed a beneficial effect on systolic blood pressure after treatment with Victoza. In addition, 30% of the participants showed signs of pre-diabetes at randomization. After 20 weeks of treatment with any dose of Victoza, 80-90% of patients no longer had any signs of pre-diabetes, compared with approximately 40% in the placebo- and Xenical-treated groups. Victoza was generally well tolerated. The overall withdrawal rate across the study was around 20%, and no more than 10% of patients treated with Victoza withdrew from the trial due to adverse events. In a 32-week open label extension study, patients who were treated at the highest dose experienced mean weight loss from baseline of 7.5-8.0 kg (5.5-6.0 kg as adjusted for placebo). Approximately 75% of all treated subjects achieved >5% weight loss, and >35% achieved >10% weight loss after 52 weeks of treatment vs. 25% and 10% of placebo subjects, respectively. Novo Nordisk completed enrollment for one of Victozas Phase III obesity studies in non diabetic patients with a history of hypertension or dyslipidemia. However, in mid-2009, Novo Nordisk announced that two other Phase III trials would not be initiated until more clarity on the U.S. regulatory process for liraglutide for the treatment of type 2 diabetes was obtained.

MEMBER: FINRA/SIPC

96

www.cowen.com

Cowen and Company, LLC

EQUITY RESEARCH

In June 2010, the company announced that, based on feedback from the FDA, it decided to reinitiate the global Phase III liraglutide obesity program. The clinical trial program evaluating safety and efficacy of liraglutide 3 mg for weight management is known as the SCALE (Satiety and Clinical Adiposity-Liraglutide Evidence in Non-diabetic and Diabetic People) program. It is comprised by four Phase III trials with approximately 5,000 obese (BMI 30 kg/m 2) and overweight (BMI 27 kg/m2) patients with comorbidities, such as hypertension, dyslipidemia, or type 2 diabetes: SCALE Maintenance: This was a 56-week, randomized, double-blind, placebo-controlled evaluating weight loss maintenance with liraglutide 3 mg vs. placebo in 422 obese and overweight patients after successfully achieving 5% weight loss during a 12-week run-in period consisting of a lifestyle modification program with low calorie diet and exercise alone. After the run-in period, patients were randomized for the 56-week main trial period to daily liraglutide 3 mg or placebo. Patients then discontinued treatment and were followed for an additional observation period of 12 weeks. Diabetic patients were excluded from the trial. The 3 co-primary endpoints were: 1) mean change in body weight from baseline; 2) proportion of patients maintaining full run-in weight loss; and 3) proportion of patients losing 5% of body weight from week 0 (i.e., achieving further weight loss after randomization). SCALE Diabetes: This was a 56-week, randomized, double-blind, placebo-controlled trial in 846 obese and overweight patients with type 2 diabetes, evaluating safety and efficacy of liraglutide 3 mg vs. liraglutide 1.8 mg and placebo to induce and maintain weight loss in diabetic patients over 56 weeks, when added to background diabetes treatment, consisting of metformin, a sulfonylurea, or glitazone as single-agents or in combination. Patients were randomized 2:1:1 to treatment with daily liraglutide 3 mg, liraglutide 1.8 mg, or placebo. Treatment was discontinued at 56 weeks, and patients were followed during an additional 12week observation period. The co-primary endpoints were: 1) change in mean body weight from baseline; 2) proportion of patients losing 5% body weight from baseline; and 3) proportion of patients losing 10% body weight from baseline. SCALE Obesity and Pre-Diabetes: This is a 56-week/160-week randomized, double-blind, placebo-controlled trial in 3,731 obese and overweight patients with comorbidities, evaluating safety and efficacy of liraglutide 3 mg vs. placebo to induce and maintain weight loss, as well as to delay onset of type 2 diabetes in pre-diabetic patients. Patients with known diabetes (HbA1c 6.5%) are excluded from the trial. Patients without pre -diabetes are randomized 2:1 to daily liraglutide 3 mg or placebo for 56 weeks. After 56 weeks, patients are re-randomized to receive liraglutide or placebo for an additional 12 weeks. Pre-diabetic patients are randomized to liraglutide 3 mg or placebo for 160 weeks. The co-primary endpoints are: 1) change in mean body weight from baseline to week 56; 2) proportion of patients losing 5% body weight from baseline at week 56; 3) proportion of patients losing 10% body weight from baseline at week 56; and 4) proportion of patients with onset of type 2 diabetes at 160 weeks. SCALE Sleep Apnea: This is a 32-week randomized, double-blind, placebo-controlled trial in approximately 340 obese patients with moderate or severe obstructive sleep apnea (OSA), evaluating efficacy of liraglutide 3 mg in combination with diet and exercise vs. placebo to

MEMBER: FINRA/SIPC

97

www.cowen.com

Cowen and Company, LLC

EQUITY RESEARCH

reduce the severity of OSA. Patients must be unwilling or unable to use continuous positive airway pressure (CPAP) treatment and must not have had CPAP treatment for at least 4 weeks prior to screening. Diabetic patients (HbA1c 6.5%) ar e excluded from the trial. Patients are randomized to daily liraglutide 3 mg or placebo, both in combination with diet and exercise, for 32 weeks. The primary endpoint of the trial is change in apnea-hypopnea index (AHI) from baseline. SCALE Phase III Trial Program
Study Name Study Type SCALE Maintenance SCALE Diabetes SCALE Obesity and Pre-Diabetes SCALE Sleep Apnea

Phase III, randomized, double-blind, placebo-controlled 12-week run-in period, followed by a 56-week treatment 56-week treatment period followed period, followed by 12-week observation period by 12-week observation period Obese/overweight patients who have achieved 5% weight loss after 12-week run-in period of lifestyle modification program with low calorie diet and exercise alone BMI 30 kg/m2 and BMI 27 kg/m2 with comorbidities, such as hypertension or dyslipidemia 422 liragultide 3.0 mg vs. placebo Without pre-diabetes: 56-week treatment period, followed 12-week re-randomization period With pre-diabetes: 160-week treatment period Obese/overweight patients with comorbidities, including pre-diabetes BMI 30 kg/m2 and BMI 27 kg/m2 with comorbidities, such as hypertension or dyslipidemia 3,731 liragultide 3.0 mg vs. placebo

Study Duration

32-week treatment period

Disease Setting

Obese/overweight patients with type 2 diabetes BMI 27 kg/m2 with type 2 diabetes (HbA1c 7.0-10.0%) 846 liragultide 3.0 mg vs. liragultide 1.8 mg vs. placebo (2:1:1 randomization) 1) Mean change in body weight from baseline 2) Proportion of patients losing 5% body weight 3) Proportion of patients losing 10% body weight June 2011 March 2013

Obese patients with moderate or severe obstructive sleep apnea (OSA) BMI 30 kg/m2 with diagnosis of moderate or severe OSA 340 liragultide 3.0 mg vs. placebo

Demographics Estimated Enrollment

Treatment Regimen

Primary Endpoint

1) Mean change in body weight from baseline 2) Proportion of patients maintaining full run-in weight loss 3) Proportion of patients losing 5% of body weight from week 0 (i.e., achieving further weight loss after randomization). October 2008 2010

1) Mean change in body weight from baseline to week 56 2) Proportion of patients losing 5% body weight at week 56 Change in apnea-hypopnea index 3) Proportion of patients losing 10% (AHI) from baseline body weight at week 56 4) Proportion of patients with onset of type 2 diabetes at 160 weeks June 2011 May 2013 June 2012 4Q13

Study Started Data Announced/Expected

Source: Cowen and Company, www.clinicaltrials.gov

Data from SCALE Maintenance: Results were presented at the ADA and EASD meetings in 2011. Patients in the trial had average age of 46 years and average BMI of 38 kg/m 2, with an average run-in weight loss of 6%. At 56 weeks after randomization, 81% of liraglutide patients maintained this run-in weight loss, compared with 49% of placebo patients. Liraglutide treatment also resulted in an additional mean weight loss of 6% after randomization, while there was minimal additional weight loss (0.1%) in the placebo group. 51% of liraglutide patients lost 5% of body weight after randomization, compared with 22% of placebo patients. The most common adverse event was nausea, which occurred in 48% of patients. Data from SCALE Diabetes: Topline data were announced in March 2013. Patients had baseline mean weight of 233 pounds and baseline average BMI of 27 kg/m 2. The trial

MEMBER: FINRA/SIPC

98

www.cowen.com

Cowen and Company, LLC

EQUITY RESEARCH

successfully met all three endpoints, with all differences for both liraglutide doses being statistically significant. At 56 weeks, the mean weight loss in the liraglutide 3 mg and 1.8 mg groups was 6% and 5%, respectively, compared with 2% in the placebo group. 50% (3 mg) and 22% (1.8 mg) of liraglutide-treated patients lost 5% of body weight from baseline, compared with 13% in the placebo group, while 35% (3 mg) and 13% (1.8 mg) of liraglutidetreated patients lost 10% of body weight from baseline, versus 4% for placebo. During the 12 week follow-up period, patients from both liraglutide treatment groups had modest re-gain of weight. Starting from baseline HbA1c of 8%, more patients in the liraglutide 3 mg and 1.8 mg groups, 69% and 67%, respectively, achieved HbA1c level 7% (ADA target recommendation), compared with 27% of placebo patients. The most common adverse events were reported as gastrointestinal in nature. SCALE Diabetes Efficacy Data

liraglutide 3 mg Mean Weight Loss (%) 6% >5% Weight Loss 50% >10% Weight Loss 22% Completion rate 77% Patients achieving HbA1c target of 7% 69%
Source: Cowen and Company, Novo Nordisk

liraglutide 1.8 mg 5% 35% 13% 78% 67%

Placebo 2% 13% 4% 66% 27%

Data from SCALE Obesity and Pre-Diabetes: Topline data for all patients at 56 weeks were announced in May 2013. The trial succeeded in all three co-primary endpoints at 56 weeks. Patients had mean baseline BMI of 38 kg/m2, and 61% had pre-diabetes at randomization. At 56 weeks, the average weight loss with liraglutide 3 mg was 8% vs. 2.6% on placebo. Liraglutide patients had greater categorical weight loss of 5% (64% vs. 27%) and 10% (33% vs. 10%) compared to placebo. At 56 weeks, 69% of the pre-diabetes subgroup receiving liraglutide eliminated signs of pre-diabetes, compared to 33% of the placebo group. Of the 39% of patients without baseline pre-diabetes, 7% of liraglutide patients and 21% of placebo patients eventually developed pre-diabetes. The company stated that all differences between liraglutide and placebo were statistically significant, though specific statistical values were not provided. The most common AEs were reported to be GI-related and diminishing over time. Whats next? The Phase III SCALE Sleep Apnea trial was completed in 3Q13. The company expects to file for approval of liraglutide 3 mg as an obesity treatment in the U.S. and E.U. around YE13 (around the turn of the year).

MEMBER: FINRA/SIPC

99

www.cowen.com

Cowen and Company, LLC

EQUITY RESEARCH

4) Orexigens Empatic
Orexigens Empatic is a combination of bupropion and zonisamide, for the treatment of obese patients with high BMI (high 30 kg/m2 range and beyond). Although zonisamide is rarely used for weight loss, it does have appetite suppressant effects, which may be related to enhancement of dopamine and serotonin neurotransmission. It is generally well-tolerated, although headache and somnolence are reported to occur with some frequency. Zonisamide went generic in the U.S. in 2005. As with naltrexone in Contrave, Orexigen has developed a proprietary, sustained-release form of zonisamide, which has shown improved tolerability compared to the generic IR form, with significant reduction in the incidence of spontaneous adverse events. Empatic has been evaluated in two Phase IIb trials (Study ZB-201 and ZB202).

ZB-201: Empatic succeeds in its first Phase IIb trial


In study ZB-201, 623 otherwise healthy subjects with BMI 29-45 kg/m2 (average weight, 220 lbs.) were enrolled at 15 sites in the U.S. Patients were given minimal diet/exercise interventions. Six different bupropion/zonisamide combinations were tested. Empatic treatment resulted in 24-week, placebo-adjusted weight loss in the ITT group of up to 8.6%, and up to 10.3% in completers. Patients who remained in the Phase IIb trial at 24 weeks were allowed to continue through 48 weeks of double-blind treatment. Those failing to achieve 5% weight loss at 24 weeks in any group were permitted to switch to the open-label Z360/B360 dose during this extension. At 48 weeks, Empatic treatment resulted in placebo-adjusted weight reductions of up to 12.9% in the ITT group, and of up to 13.9% among completers. We view these levels of weight loss as promising, and the observation of robust weight loss continuing after six months is a notable feature of this combination treatment. With evidence of a continued trend of weight loss at the end of the 48-week period, it is likely that additional reductions may occur with longer-term therapy. As this extension was conducted under double-blind conditions, the data are likely to better represent Empatics efficacy, compared with typical open-label designs.

ZB-202: Empatic goes 2-for-2 in second successful Phase IIb study


In September 2009, Orexigen released positive data from a 24-week, Phase IIb trial of Empatic in 729 obese and overweight patients (BMI range 27-45 kg/m2). Patients were randomized to one of six arms: two Empatic groups (Bup360/Zon360, Bup360/Zon120), three singletreatment groups (Bup360/placebo, Zon360/placebo, Zon120/placebo), and placebo. Patients treated with Empatic360 and Empatic120 had significantly greater average weight loss from baseline, 7.5% and 6.1%, respectively, compared with placebo (1.4%, p0.001). Weight loss in the Empatic groups was also greater than that achieved with zonisamide (3.2% and 5.3%) or bupropion (2.3%) alone. There was no plateau in weight loss observed at 24 weeks, and this may indicate the potential for greater weight loss with treatment over a longer period.

MEMBER: FINRA/SIPC

100

www.cowen.com

Cowen and Company, LLC

EQUITY RESEARCH

The most common side effects were nausea, headache, and insomnia. The most common AEs resulting in discontinuation were insomnia, headache and urticarial (hives). According to Orexigen, no patients experienced serious AEs related to Empatic, and there were no statistically or clinically meaningful difference in incidence of depression, anxiety, suicidality, and impaired cognition with Empatic treatment compared to placebo. Our consultants have noted these potential CNS effects as causing some hesitation to use zonisamide, and they are interested to see results in a larger patient population receiving Empatic. Whats next for Empatic? In March 2013, Orexigen announced that, in a series of discussions about continued development of Empatic, FDA indicated Phase III data for Empatic may be sufficient to support a NDA submission without additional data from a CVOT. Also, the company guided that, should there be no CV signals in the Empatic development program and if placebo-adjusted body weight, blood pressure, and heart rate changes are similar to or more favorable than those observed with Contrave, FDA has stated reassuring results of a CVOT with Contrave will be sufficient. Furthermore, Orexigen reported that the FDA will allow Phase III studies of Empatic to include women of childbearing age, despite teratogenicity concerns resulting from the zonisamide component. Orexigen has indicated its thinking at this point that, prior to initiating Phase III development of Empatic, it seeks a collaboration partner to help fund clinical development and future commercialization.

MEMBER: FINRA/SIPC

101

www.cowen.com

Cowen and Company, LLC

EQUITY RESEARCH

5) The Novel Mechanism: Zafgens Beloranib


The private company Zafgen is developing beloranib, a novel injectable agent that inhibits methionine aminopeptidase 2 (MetAP2), a key enzyme in the control of the production and utilization of fatty acids. Beloranib is a highly selective MetAP2 inhibitor, which is being developed as a twice-weekly, subcutaneous injection for severe obesity. Beloranib has shown very promising activity in its Phase Ib trials, and is currently in a Phase IIa trial for the treatment of severely obese patients. Data from two Phase Ib trials, 2012 Obesity Society meeting: Data from two Phase Ib trials were presented at the 2012 Obesity Society meeting in San Antonio. Both trials were randomized, double-blind, placebo-controlled studies, designed to evaluate the safety, tolerability, and metabolic effects of twice-weekly administered beloranib in severely obese women. In one trial, beloranib was administered by IV, and in the second trial, beloranib was administered subcutaneously. Patients were allowed to eat normally and were not counseled to change their exercise habits. 1) Phase Ib trial, IV administration: The trial enrolled 16 severely obese women (BMI 39.5 kg/m2) who were randomized into three arms: 3.0 mg (n=6), 6.0 mg (n=5), and placebo (n=5). 14 patients (6 patients in 3.0 mg arm, 3 patients in the 6.0 mg arm, and 6 patients in the placebo arm) completed treatment, and data for these patients were reported. Efficacy Data: After four weeks, patients in the 3.0 mg arm lost an average of 4.7 kg from baseline (p=0.0008), and patients in the 6.0 mg arm lost an average of 6.7 kg (p=0.0013), while patients in the placebo arm gained 0.2 kg. Hunger tended to be reduced on beloranib (28% for 3.0 mg arm, -52% for 6.0 mg arm, compared to -2% for placebo). Body composition measurements were consistent with reduced adipose tissue mass. Safety Data: The most frequent AEs were mild diarrhea, nausea, headache, dizziness, infusion site injury, and mild-to-moderate sleep disturbance, which resulted in two drop-outs from the 6.0 mg arm. There were no clinically significant, abnormal laboratory or ECG findings. Doses less than 6.0 mg appeared to have clinical utility in balancing effectiveness and tolerability. 2) Phase Ib trial, Subcutaneous administration: This trial enrolled 25 obese women (BMI, 34 to 36.4 kg/m2) who were randomized into four arms: beloranib (twice-weekly for 4 weeks) at 1.0 mg (n=6), 2.0 mg (n=6), 4.0 mg (n=7), or placebo (n=6). 21 patients (6 patients in the 1.0 mg arm, 5 patients in the 2.0 mg arm, 4 patients in the 4.0 mg arm, and 6 patients on placebo) completed treatment, and data for these patients were reported. Efficacy Data: After four weeks, patients in the 1.0 mg arm lost an average of 4.3 kg from baseline; patients in the 2.0 mg arm lost an average of 4.2 kg, and patients in the 4.0 mg arm lost an average of 6.1 kg (all p<0.001). Patients receiving placebo lost 1.2 kg from baseline. Hunger tended to be reduced with beloranib (-42% for the 1.0 mg arm, -45% for the 2.0 mg arm, -46% for the 4.0 mg arm, compared to -22% for placebo arm).

MEMBER: FINRA/SIPC

102

www.cowen.com

Cowen and Company, LLC

EQUITY RESEARCH

Safety Data: The most frequent AEs were decreased appetite, vivid dreams, and sleep disturbance, which resulted in two drop-outs from the 4.0 mg arm. There were no severe AEs, serious AEs, or deaths. There were no clinically significant, abnormal laboratory or ECG findings. Subcutaneous doses of beloranib less than 4.0 mg appeared to have clinical utility in balancing effectiveness and tolerability. Current Status: Based on the activity observed in the two Phase Ib trials, Zafgen advanced beloranib into Phase II development. In November 2012, a Phase IIa clinical trial testing beloranib for the treatment of severe obesity was initiated. This is a randomized, double-blind, placebo-controlled, dose ranging, Phase IIa trial evaluating weight loss, safety, and PK with beloranib in ~150 obese patients, both with and without type 2 diabetes. Patients will be randomized to receive either placebo or beloranib as a subcutaneous injection for 12 weeks. They are allowed to eat normally and are not counseled to change their exercise habits. The primary endpoints of the trial are safety and change in body weight from baseline. Phase IIa Trial Interim Data, ADA 2013: Interim data for 19 patients who completed the 12week treatment period were presented at the ADA meeting in June 2013. Enrolled patients have a mean age of 40 years, with an average body weight of 101.2 kg and BMI of 37.9 kg/m 2. There were 148 patients randomized to treatment with beloranib 0.6 mg (n=37), 1.2 mg (n=37), or 2.4 mg (n=36), or placebo (n=38). The 19 patients for which data were presented had been randomized as follows: beloranib 0.6 mg (n=5), 1.2 mg (n=6), or 2.4 mg (n=3), and placebo (n=5). After 12 weeks of treatment, patients treated with 0.6 mg, 1.2 mg, or 2.4 mg of beloranib had average weight losses of 3.8 kg, 6.1 kg, and 9.9kg, respectively, compared with a weight gain of 1.8 kg in the placebo group (all p<0.005 vs. placebo). The most common AEs with beloranib included nausea, vomiting, and sleep disturbance. There were no serious AEs or deaths. Full data from the trial are expected in 2013.

6) Another Novel Mechanism: Rhythms RM-493


Private company Rhythm Pharmaceuticals is developing RM-493, a selective, small-peptide melanocortin-4 receptor (MC4R) agonist, for the treatment of obesity and diabetes. The MC4 receptor is a G-protein-coupled receptor which is expressed in cells of the hypothalamus, and is the main melanocortin receptor involved in control of food intake. MC4R modulates both an agonist signal from the pro-opiomelanocortin (POMC) derivative -melanocyte stimulating hormone (-MSH) which is anorexigenic (appetite suppressing), and an antagonist signal from Agouti-related peptide (AgRP) which is orexigenic (appetite stimulating). The balance in these neuropeptide effects on MC4R activation regulates food intake, energy expenditure, and body weight. It is estimated that MC4R mutations account for up to 6% of severe obesity cases, and they are the most common cause of monogenic (resulting from mutations in a single gene) genetic obesity. In preclinical animal studies, treatment with RM-493 was associated with reduction in food intake and decreases in body weight. RM-493 is currently in Phase II clinical development.

MEMBER: FINRA/SIPC

103

www.cowen.com

Cowen and Company, LLC

EQUITY RESEARCH

Preclinical Studies: Results from preclinical evaluation of RM-493 in a non-human primate (monkeys) model of diet-induced obesity were published in 2012 (Kievit P et al., Diabetes 2012). A dose response study, evaluating doses of 0.17, 0.5, and 1.5 mg/kg/day) in lean animals demonstrated that a RM-493 dose of 0.5 mg/kg/day maximally reduced food intake, and this dose was used for further study. Twelve adult (age 9-11 years) male rhesus macaque monkeys, with body weights of 9-19 kg, were studied. The monkeys had been provided a highfat diet (32% of calories from fat) and daily calorie-rich treats for 1.5 years. Nine monkeys were obese, insulin-resistant, and hypertensive; they were classified as diet-sensitive. Three monkeys had normal body weight and blood pressure, and were classified as lean or diet resistant. The animals were treated with delivery vehicle a lone for 4 weeks to obtain baseline values. Then, they received 0.5 mg/kg/day of RM-493 for a total of 8 weeks. Efficacy Results in Monkeys: Food intake decreased significantly, by 35%, in the first week of treatment with RM-493. However, food intake then normalized by weeks 4 through 7 and increased by 8 weeks, compared to levels during the previous vehicle delivery alone period. Both the obese and lean monkey groups showed comparable decreases in food intake. The monkeys lost an average of 1 kg of body weight during the first 4 weeks of treatment. They also continued to lose weight through the treatment period, with a total average weight loss of 1.5 kg (~10%) at 8 weeks. The average peak weight loss was 13.5% of body weight, with 8/12 of the monkeys continuing to lose weight for 2 weeks after treatment was discontinued. Daytime activity was noted to increase during the treatment period, nearly doubling by Week 8, through there was no significant relationship between weight loss amount and activity increase on statistical analysis. Energy expenditure (measured by carbon dioxide production) was also increased by 14% at the end of treatment. CV Safety Results in Monkeys: Potential cardiovascular side effects of increased heart rate and blood pressure have been associated with MC4R agonist treatment in humans. The monkeys were implanted with telemetry devices in order to measure blood pressure and heart rate during the delivery vehicle and treatment periods. Throughout the RM-493 treatment period, there were no increases in heart rate or blood pressure measured in the animals. Heart rate significantly decreased in the treatment period, compared to the delivery vehicle period (average morning HR 128 bpm vs. 141 bpm; average evening HR 98 bpm vs. 111 bpm; p=0.008 for the course of the treatment). There were also decreases in DBP during treatment compared to the delivery vehicle period (79-80 mm Hg vs. 83-84 mm Hg), whereas SBP did not significantly differ (129-130 mm Hg vs. 129-133 mm Hg). Phase II Trial in Obese Patients: This is a randomized, double-blind, placebo-controlled, Phase II trial evaluating RM-493 versus placebo in adult (ages 18-65 years) obese patients, with BMI range of 35-50 kg/m2. Patients must have had stable body weight ( 5 kg) in the prior 6 months, and blood pressure must be < 140/90 mm Hg, with use of up to 2 anti-hypertensive medications allowed. Patients with uncontrolled hypertension, diabetes (fasting blood glucose > 140 mg/dL and HbA1c 6.5%), and psychiatric disorders (including major depression) are excluded. Treatment will be administered with RM-493 (1 mg/14 hrs via subcutaneous infusion) for 90 days or placebo (subcutaneous infusion) for 90 days. The primary endpoint of the trial is mean percentage change in weight loss. Secondary endpoints include the proportion

MEMBER: FINRA/SIPC

104

www.cowen.com

Cowen and Company, LLC

EQUITY RESEARCH

with loss of 5% baseline body weight, PK, effect on hunger/satiety, quality of life, and safety. The trial has an estimated enrollment of 70 patients and an estimated completion date in October 2013. Phase Ib Trial in Patients with MC4R Gene Deficiency: In September 2013, Rhythm announced initiation of the first in a series of RM-493 trials for treatment of obesity in patients with a genetic loss-of-function deficiency in the MC4R pathway. This is a Phase Ib trial which will evaluate efficacy and safety of RM-493 in obese patients with a loss-of-function mutation in the MC4R gene. Patients will receive up to 4 weeks of treatment with RM-493. According to Rhythm, this study expands the ongoing Phase II program.

MEMBER: FINRA/SIPC

105

www.cowen.com

Cowen and Company, LLC

EQUITY RESEARCH

Beyond BELVIQ: Arenas Four Pipeline Candidates


In addition to BELVIQ, Arena is developing additional product candidates for the treatment of autoimmune diseases, pulmonary arterial hypertension (PAH), and multiple sclerosis. Arenas early stage pipeline includes temanogrel, APD811, APD334, and APD371. Temanogrel: Temanogrel, an inverse agonist of the serotonin 2A receptor, is being codeveloped by Arena and Ildong Pharmaceutical for the treatment of thrombotic disease. Platelet activation is an initial event in thrombus formation, with resultant release of factors, including serotonin, which cause aggregation and adherence. Serotonin promotes platelet aggregation and also causes vasoconstriction, and these effects are mediated by the 5-HT2A receptors. Inverse agonists of the 5-HT2A receptor may inhibit these effects. In preventing serotonin-mediated platelet aggregation and reversing serotonin-mediated vasoconstriction, temanogrel has a potential dual mechanism of action which may be relevant for treatment or prevention of thrombotic diseases. Arena has completed two Phase I trials of temanogrel. Ildong is responsible for funding and conducting the next two planned clinical trials of temanogrel, which include a Phase I multipledose trial in healthy volunteers and a Phase IIa proof-of-concept trial. According to Arena, Ildong plans to initiate the Phase I trial by YE13. Phase Ia trial: The first Phase Ia trial was initiated in July 2007. This was a randomized, double-blind, placebo-controlled, single-ascending dose trial designed to evaluate temanogrel in 90 healthy volunteers. Initially the intended doses for this trial ranged from 1 mg to 160 mg, but the maximum dose was increased to 320 mg because of favorable tolerability. Dosedependent inhibition of serotonin-mediated platelet aggregation was demonstrated. The MTD could not be defined, despite achieving high blood concentrations of the drug. Phase Ib trial: The Phase Ib trial was initiated in January 2008. This was a randomized, double-blind, placebo-controlled, multiple-ascending dose trial in 50 healthy volunteers who received total daily doses ranging from 15 mg to 80 mg. The goal of the study was to evaluate safety, tolerability, pharmacokinetics and pharmacodynamics of multiple-ascending doses of temanogrel over a period of one week. Dose-dependent inhibition of serotonin-mediated platelet aggregation was demonstrated starting at the 15 mg dose. This may permit identification of exposure ranges producing minimal, moderate and near-complete inhibition of serotonin-mediated platelet aggregation. The most frequent AE was headache, which was more common in the placebo group than in any temanogrel group. No adverse events were dose-related, with the exception of epistaxis, which occurred in two volunteers receiving the 80 mg dose, which is above the anticipated therapeutic range. APD811: This agent is an orally available agonist of the prostacyclin (IP) receptor, intended for the treatment of pulmonary arterial hypertension. In October 2012, Arena initiated a randomized, double-blind, Phase Ib multiple ascending dose titration trial that enrolled 55 healthy volunteers (40 received APD811 and 15 received placebo). The trial evaluated safety, tolerability, pharmacokinetics, and the optimal titration schedule for APD811.

MEMBER: FINRA/SIPC

106

www.cowen.com

Cowen and Company, LLC

EQUITY RESEARCH

In August 2013, Arena announced the completion of the Phase Ib trial. There were no serious adverse events observed. The most frequent treatment-emergent AEs were headache, nausea, and jaw pain. There was one serious AE, transient atrial fibrillation, which occurred in a single patient and was considered possibly treatment related. Arena announced that a Phase II trial of APD811 will be initiated in 1Q14. APD334: This agent is a S1P1 (sphingosine 1-phosphate subtype 1) receptor agonist, currently in Phase I development for the potential treatment of autoimmune diseases, including multiple sclerosis and rheumatoid arthritis. S1P1 receptors are involved in the modulation of several biological responses, including lymphocyte trafficking from lymph nodes to peripheral blood. Five S1P receptors have been identified. A non-selective oral S1P agonist, fingolomod, (Novartiss Gilenya) has demonstrated lowering of lymphocyte counts in blood and is approved for the treatment of multiple sclerosis. Arena has optimized potent and selective small molecule S1P1 receptor agonists, such as APD334, that reduce severity of disease in preclinical autoimmune disease models of multiple sclerosis, such as the experimental autoimmune encephalomyelitis (EAE) model and the collagen-induced arthritis (CIA) model. Phase I trial: In April 2013, Arena announced the advancement of APD334 into Phase I clinical development with the initiation of dosing in a Phase I randomized, double-blind, placebo-controlled trial, evaluating safety, tolerability and PK of single-ascending doses of APD334 in up to 64 healthy adult volunteers. In August 2013, Arena reported that dosing was completed in the trial. Preliminary results demonstrated reduction in blood lymphocyte count (the desired benefit), but a reduction in heart rate (an AE observed with other S1P1-targeting drugs). Further data evaluation is ongoing. APD371: This agent is a CB2 receptor agonist currently in preclinical development for the treatment of pain. Arena has identified several novel, potent, CB2-selective, orally available compounds that are intended to retain the analgesic activity of CB receptor agonists while avoiding psychotropic side effects.

MEMBER: FINRA/SIPC

107

www.cowen.com

Cowen and Company, LLC

EQUITY RESEARCH

So, What Do Physicians Think About All This? Our Two Obesity Surveys
Obesity remains, and according to most experts and projections, will continue to be a major public health issue in the United States and a significant portion of the developed world. The most recent national data from the CDC on obesity prevalence indicate that more than onethird of U.S. adults (35.7%) are obese, equating to over 71 million people. According to NIH guidelines (Clinical Guidelines on the Identification, Evaluation, and Treatment of Overweight and Obesity in Adults: The Evidence Report, NIH Publication 98-4083) on obesity treatment, pharmacotherapy may provide a beneficial adjunct to dietary, activity, and behavioral modifications in selected individuals, including: those with BMI 30 kg/m 2 (obese) and no concomitant risk factors, or those with BMI 27 kg/m2 (overweight) and comorbidities such as hypertension, dyslipidemia, type 2 diabetes, coronary artery disease (CAD), and sleep apnea. There has been a difficult market history for weight-loss drugs, stymied by major safety issues. After more than a decade without the approval of a new weight-loss drug, and after handing out an initial round of rejections to each of the three most recent anti-obesity agents under review, the FDA approved BELVIQ (lorcaserin, Arena Pharmaceuticals) in late June 2012. Three weeks later, in July 2012, Qsymia (phentermine/topiramate, Vivus) was also approved. Qsymia became available in the U.S. market first, in mid-September 2012, and BELVIQ was commercially launched in June 2013. The third agent, Contrave (naltrexone/bupropion, Orexigen Therapeutics) was also denied approval by the FDA in 2011, based on cardiovascular safety concerns, with a cited requirement for a cardiovascular outcomes trial (CVOT). The FDA has notified Orexigen that it supports a faster path to NDA resubmission for Contrave and provided guidance on design of the LIGHT study, the Contrave CVOT, which has completed enrollment and is currently ongoing. Interim analysis of the LIGHT Study is expected by December 2013, and NDA resubmission is anticipated by YE13. In order to gain further insight into the evolving obesity therapeutics market and to gauge physician perceptions of these new agents, we have conducted two surveys with a total of 100 U.S. physicians. The first survey polled 50 primary care physicians (PCPs) who, while not specializing in obesity management, have familiarity with the clinical profiles of the new obesity drugs and would be open to using these agents in their practices (The PCP Survey). The second survey polled 50 physicians who specialize in obesity management and currently prescribe weight-loss medications (The Obesity Specialist Survey). We report the findings from both of our surveys. Our key conclusions follow: 1) Qsymia is considered to have the strongest weight-loss efficacy and is projected to be the obesity drug market leader. Despite the majority of PCPs in our survey having no (62%) or very little (28%) experience using Qsymia so far, more than half considered Qsymia to be the most efficacious weight-loss agent. Our specialists concurred with this opinion and were even more positive, with 68% considering Qsymia to produce the best weight loss among

MEMBER: FINRA/SIPC

108

www.cowen.com

Cowen and Company, LLC

EQUITY RESEARCH

obesity therapies. A clear majority (near three fourths) of our specialists projected Qsymia to have a 60% market share relative to BELVIQ. 2) All three new obesity drugs will potentially share the obesity drug market. Assuming approval of Contrave and its availability along with Qsymia and BELVIQ, most PCP respondents in our survey feel that all three drugs will be used, though with an uneven allocation in the market. (And here, more expect Qsymia to have the advantage). Our specialists agree with this outlook. Like PCPs, specialists expect all three drugs to have some use in patients, while they too think that Qsymia will dominate overall. 3) Safety issues are not likely to stand in the way of obesity drug uptake. While physicians are familiar with potential safety issues of these new therapies, safety does not appear to present a major barrier to utilization. PCPs generally expressed moderate concern about specific risks potentially associated with Qsymia, BELVIQ, or Contrave, such as teratogenicity, serotonin syndrome, and cardiovascular concerns. However, only 30% viewed safety and tolerability issues as significant enough to prevent adoption. Specialists tended to be more even more comfortable, with fewer than 15% of respondents identifying a specific concern which would present a barrier to uptake of these respective drugs. 4) Cost and reimbursement are the key concerns weighing on physicians minds about these drugs. Both PCPs and specialists most frequently cited drug costs and lack of reimbursement as the most significant hurdles to wide use of the new anti-obesity agents. Most physicians in both groups remain relatively pessimistic about insurance coverage for these agents. Given cost concerns, when possible, both PCPs and specialists appear willing to use generic substitutions for branded agents. In our surveys, PCPs (76%) seemed more willing in this regard, but a significant proportion of specialists (58% and 54% regarding Qsymia and Contrave, respectively) also indicated comfort with prescribing generics.

MEMBER: FINRA/SIPC

109

www.cowen.com

Cowen and Company, LLC

EQUITY RESEARCH

Survey Methodology
1) The Primary Care Physician (PCP) Survey
We surveyed 50 primary care physicians (PCP) across the United States. Given our view that the commercial success or failure of these new anti-obesity drugs will be determined by how well received and how widely prescribed they will be by the PCP community, we specifically wanted to gain insights into these non-specialist physicians views on these drugs. Therefore, in this survey, we sought the opinion of PCPs who did not consider themselves to be specialists in the treatment of obesity, and we pre-screened respondents using this criterion (see full questionnaire, including the four screening questions, at the back of this report). However, we did seek survey participants who were knowledgeable about the therapeutic area. Therefore, we also required respondents to be familiar with the efficacy and safety data from the clinical trials of Qsymia and BELVIQ. In addition, physicians needed to be willing to treat patients with these agents in their clinical practices. We received a total of 50 responses from PCPs meeting the screening criteria. The physicians were from diverse regions across the U.S., representing 28 states. The majority of respondents were either solo practitioners or based in group practices.

2) The Obesity Specialist Survey


For our second survey, we polled 50 physicians who specialize in the management of obesity across the United States. We sought physicians whose clinical practices primarily focus on the care of obese patients. Respondents were prescreened according to this criterion. We also screened for physicians who currently prescribe anti-obesity drugs in their practices. We collected a total of 50 responses from obesity specialist physicians who met the screening criteria. These physicians were geographically diverse, representing 22 institutions, in 19 states across the U.S. All respondents were affiliated with either tertiary academic medical centers or specialized clinical centers dedicated to weight management.

MEMBER: FINRA/SIPC

110

www.cowen.com

Cowen and Company, LLC

EQUITY RESEARCH

#1) The PCP Survey Results


Primary care physicians (PCPs) are the front-line providers of care for obese and overweight patients. In order to understand the perspective of this physician segment on the new obesity therapies, we conducted a survey limited to self-identified PCPs who did not consider themselves specialists in obesity treatment. We selected for PCPs with 1) knowledge of the efficacy and safety data for Qsymia and BELVIQ, and 2) willingness to treat patients in their practices with these agents.

PCP experience with Qsymia remains limited in our survey


In the months after the commercial launch of Qsymia in the U.S., clinical experience with the drug by PCPs remains fairly low. A significant majority of our surveyed physicians (62%) reported having no experience with Qsymia use in their patients, while 90% of the respondents had treated only up to 5 patients each. Looking at these numbers in another way, only 10% of our responding PCPs had treated more than 5 patients using Qsymia, and among that 10%, about half had treated between 6 and 10 patients, and the other half had treated between 1620 patients. No physician among the 50 PCPs we surveyed had treated more than 20 patients. In our survey, PCP experience with Qsymia to date is limited
How many patients have you treated with Qsymia since it became commercially available in September 2012?
70% 60% 50% 40% 30% 20% 10% 0% None 1-5 6-10 28%

62%

6%
0% 11-15

4%
16-20

0% 21-30

0% 31-49

0% >50

Source: Cowen and Company, Epocrates

We asked the physicians who had not yet treated any patients with Qsymia to elaborate further and answer the Why not? question. A number of them expressed general cautiousness about the use of new drugs and tendency for slow uptake of medications with which they are not familiar. Others delineated issues with cost and availability of Qsymia. We have included all the individual answers we received in response to the Why not question in the next figure.

MEMBER: FINRA/SIPC

111

www.cowen.com

Cowen and Company, LLC

EQUITY RESEARCH

PCPs list multiple reasons for lack of Qsymia use


No rep...haven't researched it enough The amount of weight loss is not worth the expense or risk of potential side effects Nervous about about any product with phentermine Paperwork at this time Just have not had new patient requests for medication treatment of obesity I am still gaining familiarity with this product Hesitant to Rx new medication Although open to idea to use , am still cautious I use the generic versions broken into two meds rather than expensive branded Drug cost prohibitive for my patient population Waiting a year or two for post launch adverse event data to be gathered I am always slow to use new drugs No one has requested it, and I haven't volunteered to treat since diet and exercise is the best way to lose weight Not on formulary Waiting for a patient who wants to pay the Brand price and has failed phentermine alone (as most wish to do first) Not covered by insurance It is not in local pharmacies Usually wait 6 months before using new product Just discovered it, was briefed on it for first time...last week

Source: Cowen and Company, Epocrates

Nevertheless, Qsymia is considered the most efficacious weight loss agent...


Despite limited clinical experience with the drug, our survey results show that many PCPs maintain a very positive view of Qsymia and consider it to be superior to BELVIQ and other weight-loss drugs. When asked to choose the agent with best weight-loss efficacy, 52% of our respondents selected Qsymia from a list which included other approved and investigational anti-obesity agents. The next agent from the list considered most efficacious was phentermine, notably a component of Qsymia, selected by 18% of physicians, while BELVIQ was selected by just 12% of respondents.

MEMBER: FINRA/SIPC

112

www.cowen.com

Cowen and Company, LLC

EQUITY RESEARCH

Half of surveyed PCPs selected Qsymia as the most efficacious weight-loss drug

In my opinion, the anti-obesity drug that results in the most robust weight loss is:
60% 50% 52%

40%
30% 20% 18%

12%
0% Qsymia Belviq Contrave 2% Liraglutide 0% Empatic 2% Orlistat Phentermine 2% Other

12%

10%
0%

None of these drugs result in what I consider to be meaningful weight loss

Source: Cowen and Company, Epocrates

The PCPs in our survey favor Qsymia, but we also note that they generally consider weightloss drugs to potentially be efficacious for their patients. In response to our question about the anti-obesity drug with the best efficacy, only 12% of respondents indicated that none of the approved and investigational agents listed would result in meaningful weight loss.

...and PCPs expect it to lead the market over BELVIQ.


Furthermore, focusing on the two most recently approved weight-loss therapies, we asked physicians to project the market shares for Qsymia and BELVIQ at 1-year after BELVIQs launch. A 62% majority of our physicians forecasted Qsymia to be the dominant leader, with 60% share of the market. Fewer than 30% of physicians indicated that the market would be split evenly between the two drugs, and only 10% of surveyed PCPs believed that BELVIQ would have greater uptake than Qsymia.

MEMBER: FINRA/SIPC

113

www.cowen.com

Cowen and Company, LLC

EQUITY RESEARCH

Most surveyed PCPs expect Qsymia to lead the market over BELVIQ: 62% for Qsymia lead vs. 10% for BELVIQ lead vs. 28% for a market split
One year after Belviqs launch, I would expect that Qsymia and Belviq will have split the anti -obesity market in the following way:
Qsymia 10%, Belviq 90% Qsymia 25%, Belviq 75% Qsymia 40%, Belviq 60% Qsymia 50%, Belviq 50% 2% 28% 0% 8%

Qsymia 60%, Belviq 40%


Qsymia 75%, Belviq 25% Qsymia 90%, Belviq 10% 0% 5% 10% 10% 15% 20%

24% 28%

25%

30%

Source: Cowen and Company, Epocrates

Our PCPs will also use Contrave, if approved; however, Qsymia seems to be the agent of choice in our group of PCPs
With the potential approval of a third new weight-loss agent, Contrave, on the horizon, possibly sometime in 2014, we were interested to find out the PCP perspective on use of all three of these new drugs. We asked physicians to project to a time with full availability of all three drugs and to consider multiple market-split scenarios. 12% voted for an even split among the three drugs One-fourth of the PCPs voted for the one-drug-really-dominates-the-market scenario, or the 80/10/10 split, and Qsymia dominated the vote there with 18%, vs. just 4% and 2% for BELVIQ and Contrave, respectively. Similarly, approximately half of the PCPs (52%) voted for scenarios of one dominant drug in the market, i.e. the 60/20/20 or the 80/10/10 splits. Again, Qsymia was the clear winner here, with 38% vs. 10% and 4% for BELVIQ and Contrave, respectively. However, a third of our PCPs expect two agents to share the lead position in the market (40% each). Again, Qsymia was more often the drug included as one of the two leads (26% of votes). That does show that our physicians believe there is room for multiple drugs to be used in this space, possibly reflecting the fact that there are several patient types that may benefit from different characteristics of each agent.

MEMBER: FINRA/SIPC

114

www.cowen.com

Cowen and Company, LLC

EQUITY RESEARCH

Surveyed PCPs view Qsymia as the dominant one, but expect to use of all three new weight-loss agents
Assuming Contrave is approved, 18-24 months from its launch, I believe the obesity drug market will be split in the following way:
Other market split Qsymia 33%, Belviq 33%, Contrave 33% Contrave 40%, Qsymia 40%, Belviq 20% 10% 8% 16% 2% 12%

Belviq 40%, Contrave 40%, Qsymia 20% Qsymia 40%, Belviq 40%, Contrave 20%
Contrave 60%, Qsymia 20%, Belviq 20% Belviq 60%, Qsymia 20%, Contrave 20% Qsymia 60%, Belviq 20%, Contrave 20% Contrave 80%, Qsymia 10%, Belviq 10% Belviq 80%, Qsymia 10%, Contrave 10% Qsymia 80%, Belviq 10%, Contrave 10% 0% 5% 2% 4%

2% 6%
20%

18% 10% 15% 20% 25%

Source: Cowen and Company, Epocrates

What about treatment duration? PCPs much more optimistic (out of touch?) than in recent history
Again focusing on the two newly approved weight-loss drugs, we asked our PCPs to indicate the anticipated average duration of therapy on either Qsymia or BELVIQ for their patients. 78% of the physicians expected the treatment duration to be 5 months or longer, including 38% who expected an average patient would remain on therapy for close to a year (10-12 months) or longer than a year. This assessment by PCPs may be a too optimistic outlook and expectation for these drugs, in the context of a therapy category with poor patient adherence in the longterm historically. In recent investor calls, Vivus management has indicated that Qsymias persistence rate, which could be a surrogate for treatment duration, was approximately 3.4 months. In our survey, only 22% of the PCPs expected the average treatment duration to be between 0-4 months.

MEMBER: FINRA/SIPC

115

www.cowen.com

Cowen and Company, LLC

EQUITY RESEARCH

Most surveyed PCPs anticipate significant Qsymia/BELVIQ treatment duration


How long would you expect the average Qsymia and Belviq patient to remain on therapy?
50% 45% 40%

36%

35%
30% 25% 20% 20% 20% 18%

15%
10% 5% 0% 0-2 months 3-4 months 5-6 months 7-9 months 10-12 months >12 months 2% 4% 0% Other

Source: Cowen and Company, Epocrates

What about safety? They all have their issues, and none of them really stand out
Safety issues have caused a troubled history for weight-loss agents and resulted in the withdrawal of multiple drugs from the U.S. market. Prior to the approvals of BELVIQ and Qsymia in June and July 2012, respectively, these withdrawals had left orlistat as the only prescription drug approved for the long-term treatment of obesity. We sought to ascertain how safety concerns might impact our physicians utilization of the new anti -obesity drugs. In our survey, we asked respondents to choose the drug with the best safety profile from a list which included approved agents (Qsymia, BELVIQ, orlistat, phentermine) and agents under investigation (Contrave, Empatic, liraglutide). The answers we received show that no agent was singled out by physicians with regard to safety, and indeed the responses were fairly evenly distributed among the approved agents. Interestingly, Contrave and Empatic received zero votes from our PCPs as the safest choice, while one-fifth of our respondents indicated that safety and tolerability concerns of the listed weight-loss drugs would prevent clinical use in patients.

MEMBER: FINRA/SIPC

116

www.cowen.com

Cowen and Company, LLC

EQUITY RESEARCH

PCPs in our survey didnt differentiate any obesity drug as safer than the others
The obesity drug with the cleanest safety and tolerability profile is:
25% 20% 15% 10% 5% 4% 0% Qsymia Belviq Contrave Liraglutide 0% Empatic Orlistat Phentermine 2% Other I believe that the safety and tolerability profile of these agents would likely prevent me from using them in most of my patients 20% 16% 16% 22% 20%

0%

Source: Cowen and Company, Epocrates

but BELVIQ fared better in terms of comfort than Qsymia


To investigate in further detail, we focused on some specific safety issues which have been associated with Qsymia, BELVIQ, and Contrave. These potential concerns were highlighted by crafting specific patient types for each agent. We then asked our PCPs to rate their comfort level with prescribing on a scale of 1 (very comfortable) to 5 (not comfortable). Qsymia: The most highlighted concerns with Qsymia have been teratogenicity and heart rate elevations. In particular, data have suggested that women exposed to topiramate during pregnancy have an increased risk of orofacial cleft formation in offspring. Qsymia approval required a risk evaluation and mitigation strategy (REMS) program, including educational efforts on teratogenic risk for patients and physicians. With regard to Qsymia, we specified women of child-bearing age and patients at increased CV risk as patient types to focus on these concerns. BELVIQ: Over the course of BELVIQs development, identified safety concerns have included potential tumorigenicity, numerical imbalances in occurrence of cardiac valvulopathy, as well as risks of psychiatric, cognitive, and serotonergic adverse effects. We therefore specified patient types with potential risk factors, such as increased malignancy risk, increased CV risk, and those taking selective serotonin reuptake inhibitors (SSRIs) , respectively, and queried about prescribing of BELVIQ. Contrave: Contrave is currently being further evaluated in a Cardiovascular Outcomes Trial (CVOT), the LIGHT study, as a pre-approval requirement. In particular, across its clinical development program, Contrave therapy has been associated with blood pressure elevations. We specified patients at increased CV risk in asking about PCPs comfort level in prescribing Contrave.

MEMBER: FINRA/SIPC

117

www.cowen.com

Cowen and Company, LLC

EQUITY RESEARCH

The mean comfort level for 5 out of the 6 patient types we asked about ranged from 3.1 to 3.4. Thus, our PCPs generally had fairly neutral comfort levels (with a trend toward less comfort) for prescribing the respective agents in the patient types which highlighted specific safety concerns. The highest level of discomfort was noted for prescribing Qsymia to women of childbearing age, with a mean 3.7 comfort level rating. This is not surprising, given that teratogenic risk is a main focus of the current Qsymia REMS program. Surveyed PCPs had generally neutral comfort levels with potential safety concerns
Level of comfort around the following factors pertaining to use of anti-obesity agents on a scale of 1-5 (1=very comfortable, 5=not comfortable)
Prescribing Belviq to patients with increased malignancy risk Prescribing Belviq to patients at increased CV risk Prescribing Belviq to patients taking SSRIs Prescribing Contrave to patients at increased CV risk 3.12 3.36 3.36 3.42 3.44

Prescribing Qsymia to patients at increased CV risk Prescribing Qsymia to women of child-bearing age 0 1 2 3

3.7 4 5

Source: Cowen and Company, Epocrates

While the obesity pharmacotherapy space has had a very difficult history from a safety standpoint, when looking at these data and in combination with the answer to the previous question on safety, in which for 80% of our surveyed PCPs, safety and tolerability concerns would not prevent clinical use of various weight-loss drugs, our conclusion from this specific survey would have to be that the safety profiles of the new anti-obesity therapies would not present major barriers to physician uptake among PCPs.

Cost is an issue that has to be dealt with, according to our PCPs


In order to identify the most significant potential barrier to widespread adoption of Qsymia, BELVIQ, and eventually Contrave, we asked our PCPs to choose one reason from the following: 1) safety/tolerability concerns; 2) cost/lack of reimbursement; 3) patient disappointment with lack of efficacy and consequent lack of motivation; 4) competition from offlabel use of generic phentermine/topiramate and bupropion/naltrexone; 5) no major hurdles preventing widespread prescribing; 6) other issues. The most commonly identified hurdle to adoption was cost or lack of reimbursement, chosen by 54% of respondents. Not too surprisingly, the second most commonly identified issue was safety, selected by 30% of the PCPs in our survey. Regarding other adoption hurdles, only a minority (10%) indicated patient disappointment from lack of efficacy; 4% of our physicians felt that there would not be any substantial impediments to prescribing these therapies, 2% (i.e. one physician) selected competition from off-label use

MEMBER: FINRA/SIPC

118

www.cowen.com

Cowen and Company, LLC

EQUITY RESEARCH

of generic components of the branded products (Qsymia and Contrave), and none of the physicians in our survey selected reasons other than those listed. In our survey, PCPs selected cost/lack of reimbursement as most significant hurdle to adoption of new weight-loss drugs
In my view, the biggest hurdle to Qsymia, Belviq, and Contraves wide adoption will be:
2%
0% 4% 10% 30% Patient disappointment with the small amount of weight loss and the resulting lack of motivation to continue treatment Competition from off-label use of generic phentermine/topiramate and bupropion/naltrexone I dont expect any big hurdles; I think Qsymia, Belviq, and Contrave will be widely prescribed 54% Cost/lack of reimbursement Concerns over safety and/or tolerability

Other hurdles

Source: Cowen and Company, Epocrates

With third-party payers historically reluctant to pay for obesity pharmacotherapy, lack of reimbursement for obesity drugs has been a common issue, until the progress achieved recently. We polled our PCPs on their expectations for insurance coverage of the approved agents Qsymia and BELVIQ in the future, two to three years following their market entries, asking them to estimate what percentage of insurers will provide prescription coverage. Most of our physicians were not optimistic about the prospects for coverage of these drugs. Only 20% of the group believed that 50% of insurers would cover obesity drugs in that time frame. In fact, a 62% majority of PCPs in our survey forecasted that less than 30% of insurance companies would reimburse these drugs approximately 2-3 years after their launches. Majority of surveyed PCPs pessimistic about insurance coverage
What percentage of insurers do you expect will cover Qsymia and Belviq approximately 23 years after their launch:
100% 90-99% 80-89% 0% 2% 0% 14% 2% 16% 18% 20% 24% 0% 4%

70-79%
60-69% 50-59% 40-49% 30-39% 20-29% 10-19% Less than 10%

0%

5%

10%

15%

20%

25%

30%

Source: Cowen and Company, Epocrates

MEMBER: FINRA/SIPC

119

www.cowen.com

Cowen and Company, LLC

EQUITY RESEARCH

PCPs appear willing to use generics off-label...


Given potential patient concerns about the costs associated with new obesity treatments, we surveyed our PCPs regarding the possibility of prescribing the individual drugs that make up a drug combination, in the cases with generic availability. In this question, we focused on Qsymia and Contrave, both of which are combinations of individual drugs with generic availability, and framed our question in terms of a patient request derived from cost concern. Querying our PCPs on their willingness to potentially use generic combinations of phentermine/topiramate or bupropion/naltrexone in such a scenario, we noted that a 3:1 majority (76%) was amenable to this type of substitution for the branded products. In our survey, PCPs appear very willing to use the individual generic drugs instead of the branded ones
For a motivated patient concerned about drug cost and asking for treatment with generic phentermine/topiramate or bupropion/naltrexone:

24% Yes, I will prescribe the generic combination

No, I will not prescribe the generic combination because. 76%

Source: Cowen and Company, Epocrates

...but this answer possibly contradicts an earlier answer.


We note that our PCPs apparent willingness to use off-label generics could be viewed as contradictory to their prior answer to the question on what they view as the biggest hurdle to these drugs adoption, since only one of our 50 respondents had indicated that competition from these types of generic substitutions might interfere with adoption of the branded products. One possible explanation to this apparent contradiction could be that PCPs would not expect many patients to request two separate prescriptions for the two generics, and thus would not consider this a major hurdle to branded drug adoption. We next asked physicians who answered that they would not prescribe the two individual generics instead of the branded drug to elaborate on why not. Most reasons given were related

MEMBER: FINRA/SIPC

120

www.cowen.com

Cowen and Company, LLC

EQUITY RESEARCH

to concerns about dosing equivalence and off-label prescribing, as we would expect. We have provided the answers we received to this question in the next table. Surveyed PCPs unwilling to use generics cite off-label and equivalence concerns
PCP Reasons for Not Using Generics I don't prescribe phentermine; I also suspect the 2 meds separate will still be expensive anyway It's not approved Dosages may be different Not FDA approved Not clear on the long-term side-effects of these combination products To my knowledge, the generic phentermine does not come in the equivalent dose alone Dosing/strength The studies showing safety and efficacy were done with the combination [branded product]
Source: Cowen and Company, Epocrates

MEMBER: FINRA/SIPC

121

www.cowen.com

Cowen and Company, LLC

EQUITY RESEARCH

#2) The Obesity Specialist Survey Results


In order to gain further perspective on the treatment of obesity using this new generation of pharmacotherapies from a group of physicians that exclusively treat obesity, we conducted a second survey and limited participation to physicians who are obesity specialists. All survey participants were current prescribers of anti-obesity pharmacotherapies in their clinical practices.

Two thirds of obesity specialists view Qsymia as the most efficacious agent
By a very wide margin, 68% vs. 4% for the next highest drug, our specialists selected Qsymia as the drug resulting in the greatest weight loss. Notably, none of the obesity specialists we surveyed selected BELVIQ. All other agents on our list received a very small number of votes, ranging from 2% to 4%. Thus, it is very clear that among this specialist group of prescribers, Qsymia is considered to be, by far, the most efficacious agent. In our survey, most obesity specialists consider Qsymia as the most efficacious agent
The obesity drug that results in the most robust weight loss is:
80% 70% 60% 50% 40% 30% 20% 10% 0% 68%

18% 0%
Qsymia Belviq

4%
Contrave

4%
Liraglutide

4%
Empatic

2%
Orlistat None of these drugs result in what I consider meaningful weight loss

Source: Cowen and Company, SurveyMonkey

We further asked our specialists to predict the relative market share of the two newly approved weight-loss agents, Qsymia and BELVIQ, in order to further gauge perceptions on relative efficacy and measure relative potential utilization. When asked to project to 18-24 months after their launches, 3/4 of our specialists (74%) viewed Qsymia as the dominant drug in the market over BELVIQ, with 60% or greater share. In addition, only 4% of the respondents in our survey selected BELVIQ as the market leader over Qsymia, with 60% or higher market share. Finally, 22% of the physicians we surveyed projected a 50-50 split between the two agents.

MEMBER: FINRA/SIPC

122

www.cowen.com

Cowen and Company, LLC

EQUITY RESEARCH

74% of obesity specialists pick Qsymia to be the market leader; only 4% pick BELVIQ
18-24 months after their launch, I believe that Qsymia and Belviq will have split the market in the following way:
Qsymia 10%, Belviq 90%
Qsymia 25%, Belviq 75% Qsymia 40%, Belviq 60%

0% 2% 2% 22% 34% 36% 4%

Qsymia 50%, Belviq 50%


Qsymia 60%, Belviq 40% Qsymia 75%, Belviq 25%

Qsymia 90%, Belviq 10%

0%

5%

10%

15%

20%

25%

30%

35%

40%

Source: Cowen and Company, SurveyMonkey

While our specialists were clearly not impressed with the weight-loss efficacy of BELVIQ, it is worth examining whether they may view it as a safer replacement for fenfluramine. Given the notorious history of the fenfluramine-phentermine combination, the popular and effective Fenphen, eventually withdrawn from the market because of cardiac valvulopathy risk, and the greater serotonin receptor selectivity of BELVIQ, some clinicians may view one potential use of BELVIQ to be in combination with phentermine. We asked our specialists about their possible use of this combination. Only one fourth of the obesity specialists in our survey answered that they would use the BELVIQ-Phentermine combination right now. The remaining three fourths stated that they would not utilize it, most often (54%) because of lack of clinical trial data, with the next most common reason (14%) being lack of FDA approval for this combination. It is therefore clear that, until clinical trial data and/or FDA approval of this combination, the vast majority of obesity specialists would be unwilling to use this potentially promising combination off-label. Majority of specialists in our survey would not use BELVIQ-Phentermine now
I would be willing to treat my patients with Belviq in combination with phentermine to help them achieve more robust weight loss:
60% 50% 40% 30% 20% 10% 0%
Yes No, I dont believe the addition of No, because we have not yet No, because the combination is not FDA-approved phentermine can significantly seen clinical trial data with the add to the weight loss seen with Belviq + phentermine combination Belviq alone

54%

26%
14%

6%

Source: Cowen and Company, SurveyMonkey

MEMBER: FINRA/SIPC

123

www.cowen.com

Cowen and Company, LLC

EQUITY RESEARCH

Specialists are open to trying multiple agentsbut it seems like Qsymia will get the first script
Not surprisingly, the majority of our obesity specialists believe in the use of pharmacotherapy for the treatment of obesity. In our earlier question concerning the most efficacious obesity therapy, only 9 respondents (18%) stated than none of the listed drugs result in meaningful weight loss. Despite their clear distinction of Qsymia as the most efficacious anti-obesity therapy, most of our specialists appear willing to embrace multiple options. We asked our physicians about how they might use Qsymia and BELVIQ in practice. Half of the group indicated willingness to try both agents in sequence, should a patient lack response to the first agent, without specifying which agent would be used upfront. Still, Qsymia seems to be the agent of choice, when a drug is selected by name: 38% expect to use Qsymia first, while only 6% expect to start first with BELVIQ. Finally, a 10% minority of specialists expected to utilize only one agent, and only 4% indicated that neither drug would be used. Vast majority of surveyed specialists expect to use both Qsymia and BELVIQbut Qsymia is the preferred agent when docs do select one

The following statement best describes my expectation for my use of Qsymia and Belviq:
I wont use either of these drugs I will try one of the two first, if the patient doesnt respond, I wont try the other one

4% 2% 50% 0% 8%

I will try one of the two first, if the patient doesnt respond, I will then try the other one I will try Belviq first, if the patient doesnt respond, I wont try Qsymia
I will try Qsymia first, if the patient doesnt respond, I wont try Belviq I will try Belviq first, if the patient doesnt respond, I will then try Qsymia

6%
30% 0% 10% 20% 30% 40% 50% 60%

I will try Qsymia first, if the patient doesnt respond, I will then try Belviq

Source: Cowen and Company, SurveyMonkey

We further polled our specialists about their views on two drugs under investigation for obesity treatment, Novo Nordisks injectable GLP-1 receptor agonist Victoza (liraglutide), already approved for the treatment of Type 2 diabetes in the EU and the US, and Orexigens Empatic (bupropion/zonisamide). Concerning both agents, the physicians in our survey tended to have a fairly positive outlook. The majority of respondents for questions on both agents (76% and 66%, respectively) considered the drugs to be either incremental improvements compared to currently available therapies, or very promising anti-obesity agents. For both drugs, 20% of respondents voted that they had low expectations for these agents. In Victozas case, a third of respondents (32%) viewed the drug as very promising, while 44% viewed it as an incremental improvement over existing therapies.

MEMBER: FINRA/SIPC

124

www.cowen.com

Cowen and Company, LLC

EQUITY RESEARCH

Specialists view Victoza as very promising (32%) or as incremental (44%)


What is your view of Novo Nordisks Victoza (liraglutide), which is already approved for diabetes and is currently in Phase III development for the treatment of obesity?
50% 45% 40% 35% 30% 25% 20% 15% 10% 5% 0% 44% 32% 20%

4%
A very promising anti-obesity An incremental improvement I have very low expectations agent over currently available for this agent therapies I dont have an opinion

Source: Cowen and Company, SurveyMonkey

In Empatics case, a fourth of respondents (26%) viewed the drug as very promising, while 40% viewed it as an incremental improvement over existing therapies. When looking at the survey results for these two agents, it is worth remembering a number of differences, including that 1) Victoza has already produced Phase III data, while Empatic has only been tested in two Phase II studies, 2) Victoza is an injectable, while Empatic is an oral, and 3) physicians have clinical experience with Victoza, since it has been on the US market since its 2009 approval for diabetes, while Empatic is still an investigational agent, which may partially account for the lower percentage of votes for it as a very promising agent, compared with Victoza. On the other hand, even if a physician in our survey was not familiar with the Empatic Phase II dataset, Empatics individual drug components, bupropion and zonisamide, are used by US physicians, (bupropion a lot more often than zonisamide, of course), and thus the idea of using that drug combination should not be completely unfamiliar to obesity specialists. Specialists view Empatic as very promising (26%) or as incremental (40%)
What is your view of Orexigens Empatic (buprorion/zonisamide) which was tested in two Phase II trials for the treatment of obesity?
45% 40% 35% 30% 25% 20% 15% 10% 5% 0% 40% 26% 20% 14%

A very promising anti-obesity An incremental improvement I have very low expectations agent over currently available for this agent therapies

I dont have an opinion

Source: Cowen and Company, SurveyMonkey

MEMBER: FINRA/SIPC

125

www.cowen.com

Cowen and Company, LLC

EQUITY RESEARCH

Focusing on more near-term development, we asked our obesity specialists to evaluate a future market with availability of Qsymia, BELVIQ, and an approved Contrave, in order to project respective market shares for these products. The answers to this question, once again, pointed to two conclusions: 1) physicians will utilize more than one of these anti-obesity agents, and 2) Qsymia seems, once again, to be the agent of choice. Here is a summary of the data: 90% of the votes went to scenarios in which there is room for all three drugs to gain at least 20% market share; only 10% of the physicians voted for the scenario that one drug (in that case, it was Qsymia) gains 80% market share, while the other two have only 10% each. All other votes went to scenarios in which there is a dominant player (40-60% market share), but there was 40-60% of the market left for the other two, indicating that physicians believe that the market will accommodate multiple drugs. 40% of the votes went to scenarios in which Qsymia is the dominant drug in the market, i.e. having 80% market share (10% of the vote) or 60% market share (30% of the vote). The corresponding scenarios for BELVIQ and Contrave only received 4% of the votes each. In addition, 54% of the votes went to scenarios in which Qsymia would get 40% or higher market share. The corresponding scenarios for BELVIQ and Contrave received 6% and 24% of the votes, respectively. Eight out of the fifty respondents (16%) indicated that the market would be split equally among the three drugs.

MEMBER: FINRA/SIPC

126

www.cowen.com

Cowen and Company, LLC

EQUITY RESEARCH

Surveyed specialists expect use of Qsymia, BELVIQ, and Contrave in the market

Assuming Contrave is approved, 18-24 months from its launch, I believe the obesity drug market will be split in the following way:
Contrave 80%, Qsymia 10%, Belviq 10% Belviq 80%, Qsymia 10%, Contrave 10% Qsymia 80%, Belviq 10%, Contrave 10% Contrave 60%, Qsymia 20%, Belviq 20% Belviq 60%, Qsymia 20%, Contrave 20% Qsymia 60%, Belviq 20%, Contrave 20% Contrave 40%, Qsymia 40%, Belviq 20% Belviq 40%, Contrave 40%, Qsymia 20% Qsymia 40%, Belviq 40%, Contrave 20% Qsymia 33%, Belviq 33%, Contrave 33%

0%

0%
10%

4%
4% 30% 20% 2% 14% 16%

0%
Source: Cowen and Company, SurveyMonkey

5%

10%

15%

20%

25%

30%

35%

Another way to look at these data is that more than one-third of the respondents forecasted a market with two agents having equal lead shares of 40% each, most often with Qsymia being one of the two leaders (34% of votes). Otherwise, the most votes (30%) were allocated to a scenario with Qsymia holding 60% of the market and the two other agents dividing the remaining share equally. With these market allocations, physicians in our survey have accounted for some degree of utilization of all three agents, albeit often disproportionate. We believe our specialists are recognizing the heterogeneous nature of obesity and considering the potential variability of therapeutic responses with different drugs in individual patients. These results reiterate for us that 1) Qsymia seems to be the dominant drug in the space, and 2) there is indeed room for multiple anti-obesity agents in clinical practice.

Specialists seem comfortable dealing with these drugs potential safety issues
We asked our specialists to select the obesity agent with the best safety and tolerability profile. The results did not distinguish any one specific agent as the safest, and votes were almost evenly distributed among currently approved obesity agents and liraglutide, which is approved for Type 2 diabetes treatment. We believe this fairly uniform split in votes demonstrates that our physicians recognize potential safety issues with all the weight-loss therapies which they may use. Notably, only 3 out of the 50 respondents (6%) indicated that safety concerns associated with these agents would prevent their clinical use. Clearly, the vast majority of obesity specialists in our survey seem comfortable with management of any potential safety or tolerability issues, and have concluded that the benefits provided by these therapies outweigh possible risks.

MEMBER: FINRA/SIPC

127

www.cowen.com

Cowen and Company, LLC

EQUITY RESEARCH

Specialists in our survey did not distinguish any one weight-loss drug as the safest
The obesity drug with the best safety and tolerability profile is:
30% 25% 20% 15% 20% 22%

24%
18%

10%
5% 0%
Qsymia Belviq

10%
6% 0%
Contrave Liraglutide Empatic Orlistat I believe that the safety and tolerability profile of these agents is such that I would probably not use them in most of my patients

Source: Cowen and Company, SurveyMonkey

In our survey, we took the opportunity to gain insight into our obesity specialists perspectives on specific safety concerns that have been associated with the approved agents Qsymia and BELVIQ. We referred to these concerns and rated comfort with prescribing of a particular agent in a clinical scenario, according to four levels: 1) very comfortable; 2) comfortable, but with precautions; 3) somewhat comfortable, limiting prescribing to a small group of patients; and 4) not comfortable, with no prescribing. For Qsymia, we asked about prescribing to women of child-bearing age and to patients at increased cardiovascular risk, in order to elicit reactions to the risks of teratogenicity and elevated heart rate, respectively. In both Qsymia scenarios, the majority of our respondents (66% with regard to teratogenic risk and 62% for CV risk) rated themselves in one of the two highest comfort levels. Specialists comfortable prescribing Qsymia to younger women
How comfortable would you be prescribing Qsymia to women of child-bearing age?
60% 50% 40% 30% 20% 10% 0%
Very comfortable, I trust my patients to not take Qsymia once they realize theyre pregnant Comfortable, but I would Somewhat comfortable, I Not comfortable, I would not closely monitor and remind would only prescribe it to a prescribe Qsymia to women of the patient every month small fraction of these patients child-bearing age

54%

26%
12% 8%

Source: Cowen and Company, SurveyMonkey

MEMBER: FINRA/SIPC

128

www.cowen.com

Cowen and Company, LLC

EQUITY RESEARCH

Specialists comfortable prescribing Qsymia to CV risk patients


How comfortable would you be prescribing Qsymia to patients at increased CV risk?
40% 35% 30% 25% 20% 15% 10% 5% 0% 36% 26%

22%
16%

Very comfortable

Comfortable, but would Somewhat comfortable, I closely monitor and warn the would only prescribe it in a patient and would ask them to small fraction of these discuss with their cardiologist patients before or after starting on Qsymia

Not comfortable, I would not prescribe to this group of patients

Source: Cowen and Company, SurveyMonkey

For BELVIQ, we queried on prescribing to patients taking SSRIs and to patients at increased cardiovascular risk, referring to serotonergic risk/serotonin syndrome and cardiac valvulopathy, respectively. In the case of patients taking SSRIs, 56% of our specialists indicated the two lowest comfort levels for prescribing BELVIQ, with approximately one-third signifying intention to limit prescribing for such patients, and about one-quarter not prescribing in this scenario. Another approximately one-third of the group was comfortable prescribing BELVIQ to patients taking SSRIs, with close monitoring and discussion. On the other hand, with regard to CV risk, most physicians (64%) indicated the two highest comfort levels for prescribing BELVIQ in this case. Some discomfort prescribing BELVIQ to patients on SSRIs
How comfortable would you be prescribing Belviq to patients taking SSRIs?
Not comfortable, I would not prescribe Belviq to this group of patients Somewhat comfortable, I would only prescribe to a small fraction of these patients Comfortable, but I would closely monitor and warn the patient

24%

32%

32%

Very comfortable, I dont think this is an issue for Belviq

12% 0% 5% 10% 15% 20% 25% 30% 35%

Source: Cowen and Company, SurveyMonkey

MEMBER: FINRA/SIPC

129

www.cowen.com

Cowen and Company, LLC

EQUITY RESEARCH

Specialists comfortable prescribing BELVIQ to patients with CV risk


How comfortable would you be prescribing Belviq to patients at increased CV risk?
40% 35% 30% 25% 20% 15% 10% 5% 0% 36% 28% 28%

8%

Very comfortable

Comfortable, but would closely Somewhat comfortable, I would monitor and warn the patient and only prescribe in a small fraction would ask them to discuss this of these patients with their cardiologist before or after starting on Belviq

Not comfortable, I would not prescribe to this group of patients

Source: Cowen and Company, SurveyMonkey

We also were interested in gauging reaction to signals of tumorigenicity and cardiac valvulopathy in BELVIQs preclinical dataset. Our specialists were asked to rate their levels of concern resulting from these data. 62% of physicians indicated some concern, but no impediment to use of BELVIQ, based on risk-benefit assessment, with another 16% of the group expressing no concern. Therefore, it appears that more than three fourths of our physicians would not limit their use of BELVIQ based on these issues. Most specialists not significantly concerned by BELVIQ preclinical safety signals
How concerned are you about the rat carcinogenicity and potential valvulopathy data observed in Belviqs preclinical studies?
70%
60% 50% 40%

62%

30%
20% 10% 0% 4%

18%

16%

Very concerned, I will not Concerned, and I will only Somewhat concerned, but the prescribe Belviq to my patients prescribe Belviq to a small % of risk/reward makes sense to me my patients and I will use Belviq in quite a few of my patients

Not concerned at all, Ill use Belviq to many of my patients

Source: Cowen and Company, SurveyMonkey

With respect to impressions of Contraves cardiovascular safety, we polled the obesity specialist group on the probability of success for the Light Study, Contraves Cardiovascular Outcomes Study (CVOT). The trials primary endpoint is time to the first confirmed occurrence of major adverse cardiovascular events (MACE), defined as cardiovascular death, non-fatal myocardial infarction, and non-fatal stroke. Ultimately, the goal of the study is to demonstrate non-inferior CV safety of Contrave compared to placebo in patients with moderate CV risk. Sixty-two percent of our specialists estimated the probability of success to be 50% or greater,

MEMBER: FINRA/SIPC

130

www.cowen.com

Cowen and Company, LLC

EQUITY RESEARCH

with most within this segment (40%) being very optimistic and indicating a 75% or greater chance of success. Majority of specialists in our survey predict 50% probability of success for Contraves CVOT

The probability that Contraves cardiovascular outcomes LIGHT trial succeeds is approximately:
30% 25% 20% 16% 22% 28%

20%
15% 10% 5% 0%
10%

12%

2%
20% 30% 50% 75% 90+%

Source: Cowen and Company, SurveyMonkey

Therefore, with the exception of serotonergic risk concerns regarding the use of BELVIQ in patients taking SSRIs, most obesity specialists surveyed appear to have a relatively high comfort level with possible safety issues of Qsymia and BELVIQ, as well as optimism about CV safety with Contrave. Given these data, it appears that most obesity specialists are comfortable with the safety profiles, and satisfied with the risk-benefit ratios, of these new antiobesity agents.

Obesity specialists see cost/reimbursement as the biggest issue for these drugs
In order to better understand the potential barriers to adoption of Qsymia, BELVIQ, and Contrave from the point of view of obesity specialists, we surveyed our physicians on the issues which may prevent uptake for each of the three drugs individually. Only a small minority of the group (2%-8% across the questions about the three agents) indicated that there would be no potential hurdles to widespread prescribing for any of the drugs. In the case of each new anti-obesity agent, the greatest number of respondents (56% for Qsymia, 52% for BELVIQ, and 48% for Contrave) voted for cost/lack of reimbursement as the primary issue preventing wide adoption. This concern outweighed the others listed, including: 1) specific safety concerns; 2) competition from the other new anti-obesity agents; 3) lack of efficacy and subsequent patient disappointment; and 4) competition from generic substitution, in the cases of Qsymia and Contrave. For Qsymia, the next most identified barrier (20%) was competition from off-label generic substitution, while for BELVIQ, competition from Qsymia was the second greatest barrier (22%). In the case of Contrave, the next most identified barrier (18%) was patient disappointment with amount of weight loss.

MEMBER: FINRA/SIPC

131

www.cowen.com

Cowen and Company, LLC

EQUITY RESEARCH

56% of specialists identify cost/lack of reimbursement as biggest hurdle to Qsymias wide adoption
In my view, the biggest hurdle to Qsymias wide adoption will be:

10%

6%

I dont expect any big hurdles; I think Qsymia will be very widely prescribed Cost/lack of reimbursement

20%

Concerns about teratogenicity Concerns about heart rate increase Competition from Belviq

2% 0% 6% 56%

Competition from off-label use of generic phentermine/topiramate Patients wont lose much weight, will be disappointed and discontinue treatment

Source: Cowen and Company, SurveyMonkey

52% of specialists identify cost/lack of reimbursement as biggest hurdle to BELVIQs wide adoption
In my view, the biggest hurdle to Belviqs wide adoption will be:

8% 22% I dont expect any big hurdles; I think Belviq will be very widely prescribed Cost/lack of reimbursement Concerns about valvulopathy

4% 0% Concerns about tumorigenicity


Concerns about serotonin syndrome 14% 52% Competition from Qsymia

Source: Cowen and Company, SurveyMonkey

MEMBER: FINRA/SIPC

132

www.cowen.com

Cowen and Company, LLC

EQUITY RESEARCH

48% of specialists identify cost/lack of reimbursement as biggest hurdle to Contraves wide adoption
Assuming it is approved, the biggest hurdle to Contraves wide adoption will be:
2% 18% I dont expect any big hurdles; I think Contrave will be very widely prescribed

Cost/lack of reimbursement

Competition from Qsymia


8% 48% Competition from Belviq

12%

Competition from off-label use of bupropion and naltrexone 12% Safety concerns

0%

Source: Cowen and Company, SurveyMonkey

We also asked our specialists about their expectations for future insurance coverage of Qsymia and BELVIQ. One-third of the specialists we surveyed predicted that half of all insurers would provide prescription coverage for these drugs, while another 18% was even more optimistic, expecting 70% or higher coverage. However, almost 50% of our specialists expected only one-third or less of insurers to reimburse these drugs. Our specialists slightly pessimistic about future insurance coverage

I believe that 2-3 years after their launch, _____ of insurers will cover Qsymia and Belviq:
40%

35% 30%
25% 20% 15% 10% 5% 16.0% 18.0% 14.0%

34.0%

8.0%

6.0%

4.0% 100%

0% 10% 20% 33% 50% 70% 85%

Source: Cowen and Company, SurveyMonkey

MEMBER: FINRA/SIPC

133

www.cowen.com

Cowen and Company, LLC

EQUITY RESEARCH

In light of the potential cost and reimbursement issues associated with new weight-loss drugs, we set out to assess our specialists perspectives on substituting the branded drug with its individual components, which are available as generics, focusing on Qsymia and Contrave. We asked about their willingness to use generic phentermine/topiramate or bupropion/naltrexone, in the context of patient requests driven by cost concerns. Respondents unwilling to prescribe generic combinations indicated whether legal liability issues or concerns about accurate generic titration motivated this decision. In both the Qsymia and Contrave scenarios, the majority of respondents, 58% and 54%, respectively, indicated willingness to utilize the offlabel generic combinations. A higher proportion of specialists had legal liability concerns with generic substitution for Contrave (30%) compared to Qsymia (20%). This difference may result from the fact that phentermine is an approved weight-loss drug, although for short term use. Most obesity specialists may likely have more experience using phentermine compared with bupropion/naltrexone in the context of weight-loss therapy, and this may account for more comfort, from a liability standpoint, using a phentermine-based generic combination. More physicians were concerned about the accuracy of generic titration and subsequent efficacy/tolerability effects with generic replacement of Qsymia (22%) compared to Contrave (16%). More than half the specialists in our survey willing to use the individual generic components of Qsymia
To a motivated patient that I think could benefit from Qsymia, but could not afford it and would ask if I could please help them by prescribing generic phentermine and topiramate instead, my answer would be:
70% 60% 50% 40% 30% 20% 10% 0% 58.0%

More than half the specialists in our survey willing to use the individual generic components of Contrave
To a motivated patient that I think could benefit from Contrave (when/if it is approved), but could not afford it and would ask if I could please help them by prescribing generic bupropion and naltrexone instead, my answer would be:
60% 50% 54.0% 30.0%

20.0%

22.0%

OK, will do it

No, Im afraid I cant, due to the legal No, Im afraid I cant, its too liability of using these two generics complicated to accurately titrate the off-label when theres a branded two generics and still get the same drug on the market efficacy and tolerability that youd get with Qsymia

40% 30% 20% 10% 0%


OK, will do it

16.0%

No, Im afraid I cant, due to the legal No, Im afraid I cant, because its liability of using these two generics too complicated to accurately titrate off-label when theres a branded the two generics and still get the drug on the market same efficacy and tolerability that youd get with Contrave

Source: Cowen and Company, SurveyMonkey

Source: Cowen and Company, SurveyMonkey

PCPs & Obesity Specialists Aligned on Key Conclusions


Through our two surveys, we have gained insights on the evolving obesity therapeutics market from two separate physician groups, each of which manages obese and overweight patients. Overall, we have found that the perspectives of PCPs and obesity specialists on weight-loss drugs tend to be aligned. There were no material disagreements between these physician groups with respect to our key findings: 1. Qsymia has the best efficacy and is expected to be used widely. PCPs may have had less hands-on experience with Qsymia and its individual components pre-

MEMBER: FINRA/SIPC

134

www.cowen.com

Cowen and Company, LLC

EQUITY RESEARCH

approval in clinical practice, but both groups consider Qsymia to produce the most robust weight loss and project Qsymia to be the obesity pharmacotherapy market leader. Obesity specialists are even more positive about Qsymia than PCPs. 2. There is room for multiple weight-loss drugs in clinical practice. PCPs and obesity specialists are very aligned on the fact that one treatment does not fit all obese patients. Assuming Contrave approval, they agree that Qsymia, BELVIQ, and Contrave will all be used to some extent for obesity treatment. Safety concerns will not significantly prevent use of anti-obesity agents. Obesity specialists are generally more comfortable with the specific safety issues that may be associated with these drugs. However, both groups do not consider safety to be the major barrier preventing utilization of these therapies. Drug cost and lack of reimbursement are viewed as the major barriers to adoption. PCPs and obesity specialists identify therapy cost concerns and lack of insurance coverage as the primary hurdles to widespread uptake of anti-obesity drugs. Both groups appear relatively willing to use the individual generic components of branded drugs off-label in order to help patients with cost concerns. Interestingly, and somewhat surprisingly, PCPs appear even more willing than specialists to prescribe such generic replacements.

3.

4.

MEMBER: FINRA/SIPC

135

www.cowen.com

Cowen and Company, LLC

EQUITY RESEARCH

Cowen and Company PCP Obesity Survey Questionnaire


n=50 for all questions 1. How many patients have you treated with Qsymia since it became commercially available in September 2012? 1. None (62%) 2. 1-5 (28%) 3. 6-10 (6%) 4. 11-15 (0%) 5. 16-20 (4%) 6. 21-30 (0%) 7. 31-49 (0%) 8. >50 (0%) One year after BELVIQs launch, I would expect that Qsymia and BELVIQ will have split the anti-obesity market in the following way: 1. Qsymia 90%, BELVIQ 10% (10%) 2. Qsymia 75%, BELVIQ 25% (28%) 3. Qsymia 60%, BELVIQ 40% (24%) 4. Qsymia 50%, BELVIQ 50% (28%) 5. Qsymia 40%, BELVIQ 60% (2%) 6. Qsymia 25%, BELVIQ 75% (8%) 7. Qsymia 10%, BELVIQ 90% (0%) In my opinion, the anti-obesity drug that results in the most robust weight loss is: 1. Qsymia (52%) 2. BELVIQ (12%) 3. Contrave (0%) 4. Liraglutide (2%) 5. Empatic (0%) 6. Orlistat (2%) 7. Phentermine (18%) 8. Other (2%) 9. None of these drugs result in what I consider to be meaningful weight loss (12%) The obesity drug with the cleanest safety and tolerability profile is: 1. Qsymia (16%) 2. BELVIQ (16%) 3. Contrave (0%) 4. Liraglutide (4%) 5. Empatic (0%) 6. Orlistat (20%) 7. Phentermine (22%) 8. Other (2%) 9. I believe that the safety and tolerability profile of these agents would likely prevent me from using them in most of my patients (20%) How long would you expect the average Qsymia and BELVIQ patient to remain on therapy? 1. 0-2 months (2%) 2. 3-4 months (20%) 3. 5-6 months (4%) 4. 7-9 months (4%) 5. 10-12 months (20%) 6. >12 months (18%) 7. Other (0%)

2.

3.

4.

5.

MEMBER: FINRA/SIPC

136

www.cowen.com

Cowen and Company, LLC

EQUITY RESEARCH

6.

Please rate your level of comfort around the following factors pertaining to your use of anti-obesity agents on a scale of 1-5 (1=very comfortable, 5=not comfortable) 1. Prescribing Qsymia to women of child-bearing age (mean 3.7) 2. Prescribing Qsymia to patients at increased CV risk (mean 3.44) 3. Prescribing BELVIQ to patients taking SSRIs (mean 3.36) 4. Prescribing BELVIQ to patients at increased CV risk (mean 3.36) 5. Prescribing BELVIQ to patients with increased malignancy risk (mean 3.12) 6. Prescribing Contrave to patients at increased CV risk (mean 3.42) For a motivated patient who may benefit from Qsymia or Contrave but is concerned about drug cost and is asking for treatment with generic phentermine/topiramate or bupropion/naltrexone instead, my answer would be: 1. Yes, I will prescribe the generic combination (76%) 2. No, I will not prescribe the generic combination because. (24%) Please provide your reasoning In my view, the biggest hurdle to Qsymia, BELVIQ, and Contraves wide adoption will be: 1. Concerns over safety and/or tolerability (30%) 2. Cost/lack of reimbursement (54%) 3. Patient disappointment with the small amount of weight loss and the resulting lack of motivation to continue treatment (10%) 4. Competition from off-label use of generic phentermine/topiramate and bupropion/naltrexone (2%) 5. I dont expect any big hurdles; I think Qsymia, BELVIQ, and Contrave will be widely prescribed (4%) 6. Other hurdles (0%) Assuming Contrave is approved, 18-24 months from its launch, I believe the obesity drug market will be split in the following way: 1. Qsymia 80%, BELVIQ 10%, Contrave 10% (18%) 2. BELVIQ 80%, Qsymia 10%, Contrave 10% (4%) 3. Contrave 80%, Qsymia 10%, BELVIQ 10% (2%) 4. Qsymia 60%, BELVIQ 20%, Contrave 20% (20%) 5. BELVIQ 60%, Qsymia 20%, Contrave 20% (6%) 6. Contrave 60%, Qsymia 20%, BELVIQ 20% (2%) 7. Qsymia 40%, BELVIQ 40%, Contrave 20% (16%) 8. BELVIQ 40%, Contrave 40%, Qsymia 20% (8%) 9. Contrave 40%, Qsymia 40%, BELVIQ 20% (10%) 10. Qsymia 33%, BELVIQ 33%, Contrave 33% (12%) 11. Other market split (2%)

7.

8.

9.

10. What percentage of insurers do you expect will cover Qsymia and BELVIQ approximately 2-3 years after their launch:

1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11.

Less than 10% (24%) 10-19% (20%) 20-29% (18%) 30-39% (16%) 40-49% (2%) 50-59% (14%) 60-69% (0%) 70-79% (2%) 80-89% (0%) 90-99% (4%) 100% (0%)

MEMBER: FINRA/SIPC

137

www.cowen.com

Cowen and Company, LLC

EQUITY RESEARCH

Cowen and Company Obesity Specialist Survey Questionnaire


n=50 for all questions 1. I believe that 2-3 years after their launch, _____ of insurers will cover Qsymia and BELVIQ: 1. 10% (16%) 2. 20% (14%) 3. 33% (18%) 4. 50% (34%) 5. 70% (8%) 6. 85% (6%) 7. 100% (4%) In my view, the biggest hurdle to Qsymias wide adoption will be: 1. I dont expect any big hurdles; I think Qsymia will be very widely prescribed (6%) 2. Cost/lack of reimbursement (56%) 3. Concerns about teratogenicity (6%) 4. Concerns about heart rate increase (0%) 5. Competition from BELVIQ (2%) 6. Competition from off-label use of generic phentermine/topiramate (20%) 7. Patients wont lose much weight, will be disappointed and discontinue treatment (10%) How comfortable would you be prescribing Qsymia to women of child-bearing age? 1. Very comfortable, I trust my patients to not take Qsymia once they realize theyre pregnant (12%) 2. Comfortable, but I would closely monitor and remind the patient every month (54%) 3. Somewhat comfortable, I would only prescribe it to a small fraction of these patients (26%) 4. Not comfortable, I would not prescribe Qsymia to women of child-bearing age (8%) How comfortable would you be prescribing Qsymia to patients at increased CV risk? 1. Very comfortable (26%) 2. Comfortable, but would closely monitor and warn the patient and would ask them to discuss with their cardiologist before or after starting on Qsymia (36%) 3. Somewhat comfortable, I would only prescribe it in a small fraction of these patients (22%) 4. Not comfortable, I would not prescribe to this group of patients (16%) To a motivated patient that I think could benefit from Qsymia, but coul d not afford it and would ask if I could please help them by prescribing generic phentermine and topiramate instead, my answer would be: 1. OK, will do it (58%) 2. No, Im afraid I cant, due to the legal liability of using these two generics off -label when theres a branded drug on the market (20%) 3. No, Im afraid I cant, its too complicated to accurately titrate the two generics and still get the same efficacy and tolerability that youd get with Qsymia (22%) In my view, the biggest hurdle to BELVIQs wide adoption will be: 1. I dont expect any big hurdles; I think BELVIQ will be very widely prescribed (8%) 2. Cost/lack of reimbursement (52%) 3. Concerns about valvulopathy (14%) 4. Concerns about tumorigenicity (0%) 5. Concerns about serotonin syndrome (4%) 6. Competition from Qsymia (22%) How comfortable would you be prescribing BELVIQ to patients taking SSRIs? 1. Very comfortable, I dont think this is an issue for BELVIQ (12%)

2.

3.

4.

5.

6.

7.

MEMBER: FINRA/SIPC

138

www.cowen.com

Cowen and Company, LLC

EQUITY RESEARCH

2. 3. 4. 8.

Comfortable, but I would closely monitor and warn the patient (32%) Somewhat comfortable, I would only prescribe to a small fraction of these patients (32%) Not comfortable, I would not prescribe BELVIQ to this group of patients (24%)

How comfortable would you be prescribing BELVIQ to patients at increased CV risk? 1. Very comfortable (36%) 2. Comfortable, but would closely monitor and warn the patient and would ask them to discuss this with their cardiologist before or after starting on BELVIQ (28%) 3. Somewhat comfortable, I would only prescribe in a small fraction of these patients (28%) 4. Not comfortable, I would not prescribe to this group of patients (8%) How concerned are you about the rat carcinogenicity and potential valvulopathy data observed in BELVIQs preclinical studies? 1. Very concerned, I will not prescribe BELVIQ to my patients (4%) 2. Concerned, and I will only prescribe BELVIQ to a small % of my patients (18%) 3. Somewhat concerned, but the risk/reward makes sense to me and I will use BELVIQ in quite a few of my patients (62%) 4. Not concerned at all, Ill use BELVIQ to many of my patients (16%)

9.

10. I would be willing to treat my patients with BELVIQ in combination with phentermine to help them achieve more robust weight loss: 1. Yes (26%) 2. No, I dont believe the addition of phentermine can significantly add to the weight loss seen with BELVIQ alone (6%) 3. No, because we have not yet seen clinical trial data with the BELVIQ + phentermine combination (54%) 4. No, because the combination is not FDA-approved (14%) 11. The following statement best describes my expectation for my use of Qsymia and BELVIQ: 1. I will try Qsymia first, if the patient doesnt respond, I will then try BELVIQ (30%) 2. I will try BELVIQ first, if the patient doesnt respond, I will then try Qsymia (6%) 3. I will try Qsymia first, if the patient doesnt respond, I wont try BELVIQ (8%) 4. I will try BELVIQ first, if the patient doesnt respond, I wont try Qsymia (0%) 5. I will try one of the two first, if the patient doesnt respond, I will then try the other one (50%) 6. I will try one of the two first, if the patient doesnt respond, I wont try the other one (2%) 7. I wont use either of these drugs (4%) 12. 18-24 months after their launch, I believe that Qsymia and BELVIQ will have split the market in the following way: 1. Qsymia 90%, BELVIQ 10% (4%) 2. Qsymia 75%, BELVIQ 25% (36%) 3. Qsymia 60%, BELVIQ 40% (34%) 4. Qsymia 50%, BELVIQ 50% (22%) 5. Qsymia 40%, BELVIQ 60% (2%) 6. Qsymia 25%, BELVIQ 75% (2%) 7. Qsymia 10%, BELVIQ 90% (0%) 13. The obesity drug that results in the most robust weight loss is: 1. Qsymia (68%) 2. BELVIQ (0%) 3. Contrave (4%) 4. Liraglutide (4%) 5. Empatic (4%) 6. Orlistat (2%) 7. None of these drugs result in what I consider meaningful weight loss (18%)

MEMBER: FINRA/SIPC

139

www.cowen.com

Cowen and Company, LLC

EQUITY RESEARCH

14. The obesity drug with the best safety and tolerability profile is: 1. Qsymia (20%) 2. BELVIQ (22%) 3. Contrave (10%) 4. Liraglutide (24%) 5. Empatic (0%) 6. Orlistat (18%) 7. I believe that the safety and tolerability profile of these agents is such that I would probably not use them in most of my patients (6%) 15. The probability that Contraves cardiovascular outcomes LIGHT trial succeeds is approximately: 1. 10% (2%) 2. 20% (20%) 3. 30% (16%) 4. 50% (22%) 5. 75% (28%) 6. 90+% (12%) 16. Assuming it is approved, the biggest hurdle to Contraves wide adoption will be: 1. I dont expect any big hurdles; I think Contrave will be very widely prescribed (2%) 2. Cost/lack of reimbursement (48%) 3. Competition from Qsymia (12%) 4. Competition from BELVIQ (0%) 5. Competition from off-label use of bupropion and naltrexone (12%) 6. Safety concerns (8%) 7. Patients wont lose much weight, will be disappointed and discontinue treatment (18%) 17. To a motivated patient that I think could benefit from Contrave (when/if it is approved), but could not afford it and would ask if I could please help them by prescribing generic bupropion and naltrexone instead, my answer would be: 1. OK, will do it (54%) 2. No, Im afraid I cant, due to the legal liability of using these two generics off-label when theres a branded drug on the market (30%) 3. No, Im afraid I cant, because its too complicated to accurately titrate the two generics and still get the same efficacy and tolerability that youd get with Contrave (16%) 18. Assuming Contrave is approved, 18-24 months from its launch, I believe the obesity drug market will be split in the following way: 1. Qsymia 33%, BELVIQ 33%, Contrave 33% (16%) 2. Qsymia 40%, BELVIQ 40%, Contrave 20% (14%) 3. BELVIQ 40%, Contrave 40%, Qsymia 20% (2%) 4. Contrave 40%, Qsymia 40%, BELVIQ 20% (20%) 5. Qsymia 60%, BELVIQ 20%, Contrave 20% (30%) 6. BELVIQ 60%, Qsymia 20%, Contrave 20% (4%) 7. Contrave 60%, Qsymia 20%, BELVIQ 20% (4%) 8. Qsymia 80%, BELVIQ 10%, Contrave 10% (10%) 9. BELVIQ 80%, Qsymia 10%, Contrave 10% (0%) 10. Contrave 80%, Qsymia 10%, BELVIQ 10% (0%) 19. What is your view of Novo Nordisks Victoza (liraglutide), which is already approved for diabetes and is currently in Phase III development for the treatment of obesity?

MEMBER: FINRA/SIPC

140

www.cowen.com

Cowen and Company, LLC

EQUITY RESEARCH

1. 2. 3. 4.

A very promising anti-obesity agent (32%) An incremental improvement over currently available therapies (44%) I have very low expectations for this agent (20%) I dont have an opinion (4%)

20. What is your view of Orexigens Empatic (buprorion/zonisamide) which was tested in two Phase II trials for t he treatment of obesity? 1. A very promising anti-obesity agent (26%) 2. An incremental improvement over currently available therapies (40%) 3. I have very low expectations for this agent (20%) 4. I dont have an opinion (14%)

MEMBER: FINRA/SIPC

141

www.cowen.com

Cowen and Company, LLC

EQUITY RESEARCH

Valuation Methodology & Investment Risks


Valuation Methodology
Biotechnology: In calculating our 12-month target price, we employ one or more valuation methodologies, which include a discounted earnings analysis, discounted cash flow analysis, net present value analysis and/or a comparable company analysis. These analyses may or may not require the use of objective measures such as price-to-earnings or price-to-sales multiples as well as subjective measures such as discount rates. We make investment recommendations on early stage (pre-commercial) biotechnology companies based upon an assessment of their technology, the probability of pipeline success, and the potential market opportunity in the event of success. However, because these companies lack traditional financial metrics, we do not believe there are any good methodologies for assigning a specific target price to such stocks.

Investment Risks
Biotechnology: There are multiple risks that are inherent with an investment in the biotechnology sector. Beyond systemic risk, there is also clinical, regulatory, and commercial risk. Additionally, biotechnology companies require significant amounts of capital in order to develop their clinical programs. The capitalraising environment is always changing and there is risk that necessary capital to complete development may not be readily available.

Company Specific Risks


Risks to our Market Perform rating on ARNA include: 1) significant changes in the uptake in BELVIQ scripts, 2) competitive pressure from Qsymia, 3) positive (or negative) data from Orexigen's LIGHT CVOT, 4) change of EU regulatory plans for BELVIQ, and 5) acceleration of development plans and/or regulatory pathway news for combination product with phentermine.

MEMBER: FINRA/SIPC

142

www.cowen.com

Cowen and Company, LLC

EQUITY RESEARCH

Addendum
STOCKS MENTIONED IN IMPORTANT DISCLOSURES
Ticker ARNA OREX VVUS Company Name Arena Pharmaceuticals Orexigen Therapeutics Vivus

Analyst Certification
Each author of this research report hereby certifies that (i) the views expressed in the research report accurately reflect his or her personal views about any and all of the subject securities or issuers, and (ii) no part of his or her compensation was, is, or will be related, directly or indirectly, to the specific recommendations or views expressed in this report.

Important Disclosures
Cowen and Company, LLC and or its affiliates make a market in the stock of Arena Pharmaceuticals, Orexigen Therapeutics and Vivus securities. Cowen and Company, LLC compensates research analysts for activities and services intended to benefit the firm's investor clients. Individual compensation determinations for research analysts, including the author(s) of this report, are based on a variety of factors, including the overall profitability of the firm and the total revenue derived from all sources, including revenues from investment banking. Cowen and Company, LLC does not compensate research analysts based on specific investment banking transactions.

Disclaimer
This research is for our clients only. Our research is disseminated primarily electronically and, in some cases, in printed form. Research distributed electronically is available simultaneously to all Cowen and Company, LLC clients. All published research can be obtained on the Firm's client website, https://cowenlibrary.bluematrix.com/client/library.jsp. Further information on any of the above securities may be obtained from our offices. This report is published solely for information purposes, and is not to be construed as an offer to sell or the solicitation of an offer to buy any security in any state where such an offer or solicitation would be illegal. Other than disclosures relating to Cowen and Company, LLC, the information herein is based on sources we believe to be reliable but is not guaranteed by us and does not purport to be a complete statement or summary of the available data. Any opinions expressed herein are statements of our judgment on this date and are subject to change without notice. For important disclosures regarding the companies that are the subject of this research report, please contact Compliance Department, Cowen and Company, LLC, 599 Lexington Avenue, 20th Floor, New York, NY 10022. In addition, the same important disclosures, with the exception of the valuation methods and risks, are available on the Firm's disclosure website at https://cowen.bluematrix.com/sellside/Disclosures.action. Price Targets: Cowen and Company, LLC assigns price targets on all covered companies unless noted otherwise. The price target for an issuer's stock represents the value that the analyst reasonably expects the stock to reach over a performance period of twelve months. The price targets in this report should be considered in the context of all prior published Cowen and Company, LLC research reports (including the disclosures in any such report or on the Firm's disclosure website), which may or may not include price targets, as well as developments relating to the issuer, its industry and the financial markets. For price target valuation methodology and risks associated with the achievement of any given price target, please see the analyst's research report publishing such targets. Notice to UK Investors: This publication is produced by Cowen and Company, LLC which is regulated in the United States by FINRA. It is to be communicated only to persons of a kind described in Articles 19 and 49 of the Financial Services and Markets Act 2000 (Financial Promotion) Order 2005. It must not be further transmitted to any other person without our consent. Copyright, User Agreement and other general information related to this report 2013 Cowen and Company, LLC. Member NYSE, FINRA and SIPC. All rights reserved. This research report is prepared for the exclusive use of Cowen clients and may not be reproduced, displayed, modified, distributed, transmitted or disclosed, in whole or in part, or in any form or manner, to others outside your organization without the express prior written consent of Cowen. Cowen research reports are distributed simultaneously to all clients eligible to receive such research reports. Any unauthorized use or disclosure is prohibited. Receipt and/or review of this research constitutes your agreement not to reproduce, display, modify, distribute, transmit, or disclose to others outside your organization the contents, opinions, conclusion, or information contained in this report (including any investment recommendations, estimates or price targets). All Cowen trademarks displayed in this report are owned by Cowen and may not be used without its prior written consent.

MEMBER: FINRA/SIPC

143

www.cowen.com

Cowen and Company, LLC

EQUITY RESEARCH

Cowen and Company, LLC. New York (646) 562-1000 Boston (617) 946-3700 San Francisco (415) 646-7200 Chicago (312) 577-2240 Cleveland (440) 331-3531 Atlanta (866) 544-7009 London (affiliate) 44-207-071-7500 COWEN AND COMPANY RATING DEFINITIONS Cowen and Company Rating System effective May 25, 2013 Outperform (1): The stock is expected to achieve a total positive return of at least 15% over the next 12 months Market Perform (2): The stock is expected to have a total return that falls between the parameters of an Outperform and Underperform over the next 12 months Underperform (3): Stock is expected to achieve a total negative return of at least 10% over the next 12 months Assumption: The expected total return calculation includes anticipated dividend yield Cowen and Company Rating System until May 25, 2013 Outperform (1): Stock expected to outperform the S&P 500 Neutral (2): Stock expected to perform in line with the S&P 500 Underperform (3): Stock expected to underperform the S&P 500 Assumptions: Time horizon is 12 months; S&P 500 is flat over forecast period Cowen Securities, formerly known as Dahlman Rose & Company, Rating System until May 25, 2013 Buy The fundamentals/valuations of the subject company are improving and the investment return is expected to be 5 to 15 percentage points higher than the general market return Sell The fundamentals/valuations of the subject company are deteriorating and the investment return is expected to be 5 to 15 percentage points lower than the general market return Hold The fundamentals/valuations of the subject company are neither improving nor deteriorating and the investment return is expected to be in line with the general market return COWEN AND COMPANY RATING ALLOCATION
Distribution of Ratings/Investment Banking Services (IB) as of 09/30/13 Rating Count Buy (a) 394 Hold (b) 255 Sell (c) 22

Ratings Distribution 58.72% 38.00% 3.28%

Count 54 5 1

IB Services/Past 12 Months 13.71% 1.96% 4.55%

(a) Corresponds to "Outperform" rated stocks as defined in Cowen and Company, LLC's rating definitions. (b) Corresponds to "Market Perform" as defined in Cowen and Company, LLC's ratings definitions. (c) Corresponds to "Underperform" as defined in Cowen and Company, LLC's ratings definitions. Note: "Buy", "Hold" and "Sell" are not terms that Cowen and Company, LLC uses in its ratings system and should not be construed as investment options. Rather, these ratings terms are used illustratively to comply with FINRA and NYSE regulations.

MEMBER: FINRA/SIPC

144

www.cowen.com

Cowen and Company, LLC

EQUITY RESEARCH

MEMBER: FINRA/SIPC

145

www.cowen.com

Cowen and Company, LLC

EQUITY RESEARCH

Legend for Price Chart: I = Initation | 1 = Outperform | 2 = Market Perform | 3 = Underperform | T = Terminated Coverage | $xx = Price Target | NA = Not Available

MEMBER: FINRA/SIPC

146

www.cowen.com