“Lilly Announces Withdrawal of Xigris®

Following Recent Clinical Trial Results”

Page 1 of 12
 Table of Content:

Journey of Xigris...............................................................................................................................................3
Why Sepsis?......................................................................................................................................................4
Why Xigris?.......................................................................................................................................................4
Initial Approval – supporting data....................................................................................................................6
PROWESS study:...........................................................................................................................................7
PROWESS Follow-up Study Survival status...................................................................................................7
ENHANCE study:...........................................................................................................................................8
ADDRESS study:............................................................................................................................................8
RESOLVE study:............................................................................................................................................8
Reasons for withdrawal....................................................................................................................................9
Could this be prevented?...............................................................................................................................10
Lessons Learnt................................................................................................................................................11
Bibliography...................................................................................................................................................12

Page 2 of 12
Journey of Xigris
Xigris (Drotrecogin alfa activated) was approved by FDA in November 2001 to be used for
reduction of mortality in adult patients with severe sepsis (sepsis associated with acute organ
dysfunction) who have a high risk of death (as determined by APACHE II) (FDA approval letter,
21Nov2001)

The approval itself was seen somehow as a surprise by some specialists as FDA almost always
follows the recommendations of its committees. FDA Anti-Infective Drugs Advisory Committee
reviewed the information on Xigris in October 2011 and the panel split 10 to 10 not making any
clear decision. At the time of the review only few treatments for sepsis were available and England
Journal of Medicine indicated that Xigris could reduce sepsis deaths by 20%; it was also estimated
that Xigris could reach annual sales of $2 billion dollars or more if it is approved. (Forbes
17Oct2001)

10 years later, in October 2011 Eli Lilly and Company announced withdrawal of its Xigris product in
all markets following results of the PROWESS-SHOCK study (the study did not meet the primary
endpoint of a statistically significant reduction in 28-day all-cause mortality in patients with septic
shock). In the withdrawal letter Timothy Garnett, M.D., Lilly's Senior Vice President and Chief
Medical Officer was saying "We believe the original Xigris approval was appropriate and these
recent results were quite unexpected," Garnett added. "A contributing factor to these study results
could be advances in the standard of care for treating severe sepsis over the past 10 years."

“The PROWESS-SHOCK study was initiated in March of 2008 as a condition for continued market
authorization in Europe. Lilly committed to conduct a new placebo-controlled clinical trial to help
refine appropriate patient identification for treatment with Xigris and to confirm the benefit-risk
profile of the product.” (Lilly Announces Withdrawal of Xigris® Following Recent Clinical Trial
Results, Lilly PRWire, October 2011)

As Xigris was for so many years a presence in the intensive care meetings, was quite aggressively
marketed both in US and Europe leading to being used in many ICU and costed ~ 2 billion USD
the first question coming up is – can that be prevented?

The FDA license issued in 2001 was based on the results of PROWESS study, a phase III trial
suggesting that a non-antimicrobial life-saving drug had finally been found for sepsis. Until then,
other pharmacological therapies had been unsuccessfully tested in randomized controlled trials.
Treatment with drotrecogin alfa provided a significant 6.1% absolute reduction in 28-day mortality
compared with placebo. Analyses of subgroups, however, showed that the efficacy of drotrecogin
alfa might have differed in patients with a high or low risk of death. Mortality was lower in the
drotrecogin alfa group than in the placebo group for patients with high risk of death, defined as
patients with an APACHE II score 25 or more (above the median value) or with multiple organ
Page 3 of 12
failure. However, the treatment seemed to be ineffective in the APACHE II score less than 25
subgroup and in patients with single organ dysfunction. Moreover, treatment with drotrecogin alfa
produced an increase in serious bleeding events in all subgroups, especially during the infusion
period 2.4% of 850 treated patients’ vs 1% of 840 controls (D. Poole, G. Bertolini and S. Garanttin,
Errors in the approval process and post-marketing evaluation of drotrecogin alfa (activated) for the
treatment of severe sepsis, The Lancet Infectious Disease, February 2009)

The initial approval was also based on several commitments made by Lilly in 6 letters sent to FDA
after the vote of the advisory committee; according to Lilly commitments and FDA requests Lilly
should submit additional data from ongoing study to assess long term survival outcome by
November 20002, to evaluate the efficacy and safety of Xigris in ~11.350 patients with severe
sepsis and lower risk of death (APACHE II score of 24 or less) by June 2005, to evaluate the
efficacy and safety of Xigris in ~500 pediatric patients with severe sepsis by February 2005, to
evaluate whether low-dose heparin has an effect on mortality in a study of approximately 2000
adult patients with severe sepsis who have a high risk of death and are receiving Xigris, to provide
more complete validation for the existing immunogenicity neutralizing antibody assay by April
2002, to monitor the immunogenicity response of Xigris in patients with severe sepsis post 28
days, to collect additional samples for immunogenicity testing. (FDA approval letter, 21Nov2001)

Why Sepsis?
Severe sepsis was in 2001 one of the important cause of death in Unites States with ~750.000
episodes of severe sepsis per year and about 30%-50% of these being fatal and 33.6% resulting in
organ failure; 70% of patients with three or more organ failures would die. Between 1979 – 2000
there was a 8.7% increase in the annual incidence of sepsis resulting in annual costs estimated at
$17 billion. Sepsis was the 2nd leading cause of death in non-coronary ICUs and the 10th leading
cause of death overall. (Activated protein C in septic shock: A propensity-matched analysis,
Critical Care March 2011).

Why Xigris?
“The development of severe sepsis is associated with a generalized inflammatory and pro-
coagulant response with increased levels of pro-inflammatory cytokines and markers of thrombin
generation. Despite of adequate antimicrobial therapy and removal of the source of infection, an
intensive systemic response can lead to endothelial damage, fibrin deposition in the
microvasculature, organ hypo-perfusion, parenchymal cell dysfunction, multiple organ dysfunction,
and death.

Page 4 of 12
Activated Protein C, an endogenous protein that promotes fibrinolysis and inhibits thrombosis and
inflammation, is an important modulator of the coagulation and inflammation associated with
severe sepsis. Activated Protein C is converted from its inactive precursor, Protein C, by thrombin
coupled to thrombomodulin (Esmon 1989). Reduced levels of Protein C are found in the majority of
patients with sepsis and are associated with an increased risk of death. In addition to low plasma
levels of Protein C, the conversion of Protein C to Activated Protein C may be impaired during
sepsis as a result of the down-regulation of thrombomodulin by inflammatory cytokines.

The above observations lead to the development of recombinant human Activated Protein C
[drotrecogin alfa (activated)] as a therapy for patients with severe sepsis. Data supporting the
safety and efficacy of rAPC for the treatment of adult patients with severe sepsis was derived from
a single, multi-country, placebo controlled Phase 3 study (n=1690 patients) with supporting data
from a single, multi-country, placebo-controlled Phase 2 study (n=131 patients). Analysis of data
from the Phase 2 study indicated that drotrecogin alfa (activated), administered as a 24 μg/kg/hr
constant rate infusion for 96 hours, reduced markers of coagulopathy (D-dimer) and inflammation
(IL-6) compared to placebo. Analysis of data from the Phase 3 study indicated that drotrecogin alfa
(activated), administered as a 24 μg/kg/hr constant rate infusion for 96 hours, significantly reduced
28-day all cause mortality in patients with severe sepsis compared to placebo.

At Study Day 28, the observed mortality rates were 24.71% for the drotrecogin alfa (activated)
group and 30.83% for the placebo group. These data represent an unprecedented breakthrough:
in the pivotal Phase 3 study an additional 6 lives were saved for every 100 patients treated with
drotrecogin alfa (activated). Consistent with its antithrombotic and pro-fibrinolytic effects, the
administration of drotrecogin alfa (activated) was associated with an increase in the percent of
patients experiencing a bleeding event reported as a serious adverse event (3.5% vs. 2.0%).

Serious bleeding events frequently resulted from injury to a blood vessel (traumatic or iatrogenic)
or following instrumentation of a highly vascular organ, such as the kidney or lung. There were no
other safety concerns associated with the administration of drotrecogin alfa (activated) to patients
with severe sepsis.” (FDA Anti-Infective Drugs Advisory Committee 12 September 2001; Briefing
Document for Xigris for the Treatment of Severe Sepsis; Eli Lilly and Company)

The novel aspect of this product was its action on multiple sites of the inflammatory cascade. The
proposed mechanism of drug action was promotion of fibrinolysis and inhibition of thrombosis,
thereby preventing endovascular injury and multi-organ dysfunction. Subsequent work has shown
that additional benefits of rAPC may include direct cytoprotective effects by alteration in gene,
expression profiles, anti-inflammatory actions and anti-apoptotic activities, and by stabilization of
endothelial barriers (A. Puxty, P. McConnell, S. Crawley, S. McAree, T. Quasim and S. Ramsay,
Recombinant activated protein C usage in Scotland: a comparison with published guidelines and a
survey of attitudes, Anaesthesia 2012, The Association of Anaesthetists of Great Britain and
Ireland)
Page 5 of 12
Initial Approval – supporting data
Despite the initial enthusiasm for rAPC after FDA approval and publication of the PROWESS
study, both the evidence of efficacy and safety of this product have become a point of debate.
Many clinicians expressed concerns based on risk-benefit ratios, drug authority approval and post-
approval marketing activities. European Medicines Evaluation Agency (EMEA) has issued only an
annual license based on current available evidence. (A. Puxty, P. McConnell, S. Crawley, S.
McAree, T. Quasim and S. Ramsay, Recombinant activated protein C usage in Scotland: a
comparison with published guidelines and a survey of attitudes, Anaesthesia 2012, The
Association of Anaesthetists of Great Britain and Ireland)

Xigris was authorized under “Exceptional Circumstances” meaning that for scientific reasons it has
not been possible to obtain complete information on this medicinal product and The European
Medicines Agency (EMEA) would review any new information which may become available every
year and update the SPC as necessary.
Table 1: Summary of Clinical Trials in severe sepsis with drotrecogin alfa (activated): (Clinical trials
in severe sepsis with drotrecogin alfa (activated); Pierre-François Laterre)

Page 6 of 12
As per Summary of Product Characteristics the following clinical efficacies were considered based
on the data from the clinical trials:

PROWESS study: Phase 3 international, multi-center, randomized, double-blind, placebo-

controlled trial in 1690 patients with severe sepsis. Severe sepsis is defined as sepsis associated
with acute organ dysfunction. Exclusion criteria encompassed patients at high risk of bleeding,
patients who were not expected to survive for 28 days due to a pre-existing, non-sepsis related
medical condition, HIV positive patients whose most recent CD4 count was ≤ 50/mm3, patients on
chronic dialysis, and patients who had undergone bone marrow, lung, liver, pancreas or small
bowel transplantation, and patients with acute clinical pancreatitis without a proven source of
infection.

In the PROWESS trial, treatment was initiated within 48 hours of onset of the first sepsis-induced
organ dysfunction. The median duration of organ dysfunction prior to treatment was 18 hours.
Patients were given a 96-hour constant rate infusion of Xigris at 24 μg/kg/hr (n=850) or placebo
(n=840). Xigris was added to best standard care. Best standard care includes adequate antibiotics,
source control and supportive treatment (fluids, inotropes, vasopressors and support of failing
organs, as required). Patients treated with Xigris experienced improved 28-day survival compared
to those treated with placebo.

At 28 days, the overall mortality rates were 24.7% for the Xigris-treated group and 30.8% for the
placebo-treated group (p=0.005).

Significant absolute death reduction was limited to the subgroup of patients with greater disease
severity (baseline APACHE II score ≥25 or at least 2 acute organ dysfunctions at baseline).

In the subgroup of patients with an APACHE II score ≥25 at baseline, the mortality was 31% in the
Xigris group (128 out of 414) and 44% in the placebo group (176 out of 403). No death reduction
was observed in the subgroup of patients with lower disease severity.

In the subgroup of patients with at least 2 acute organ dysfunctions at baseline, the mortality was
26.5% in the Xigris group (168 out of 634) and 33.9% in the placebo group (216 out of 637).

No significant death reduction was observed in the subgroup of patients with less than 2 acute
organ dysfunctions at baseline.

A consistent treatment effect on mortality with Xigris administration was observed across patient
subgroups defined by age, gender and infection type.

PROWESS Follow-up Study Survival status: assessed in a follow-up study of

PROWESS survivors. In-hospital and 3 month survival status was reported for 98% and 94% of
the 1690 PROWESS subjects respectively.

Page 7 of 12
In the overall population, the in-hospital mortality was significantly lower in patients on Xigris than
in patients on placebo (29.4% vs. 34.6%; p=0.023).

Survival through 3 months was also better in the Xigris group compared to placebo.

These data confirmed that the benefit of Xigris is limited to the more severely affected sepsis
patients such as patients with multiple organ failure and shock.

ENHANCE study: Further Clinical Experience In a Phase 3b international, single-arm, open-

label clinical trial with 2378 adult patients with severe sepsis received drotrecogin alfa (activated).
The entry criteria were similar to those employed in PROWESS. Patients received drotrecogin alfa
activated) within 48 hours of onset of the first sepsis induced organ dysfunction. The median
duration of organ dysfunction prior to treatment was 25 hours.

At 28 days, the mortality rate in the Phase 3b study was 25.3%. The mortality rate was lower for
patients treated within 24 hours of organ dysfunction compared to those treated after 24 hours,
even after adjustment for differences in disease severity.

ADDRESS study: A total of 2640 adult patients with severe sepsis who were at low risk of
death (e.g. patients with APACHE II<25 or with only one sepsis-induced organ failure) were
enrolled in a randomized, double-blind, placebo-controlled trial.

The trial was stopped for futility after an interim analysis. No benefit of drotrecogin alfa (activated)
was observed in the subgroup of 872 patients at low risk of death with multiple organ dysfunction,
so ADDRESS did not confirm the efficacy results of the PROWESS study.

In the multiple organ dysfunction subgroup of ADDRESS the 28-day placebo mortality was 21.9%,
similar to the single organ dysfunction subgroup of PROWESS (21.2%), confirming the lack of
efficacy in patients with severe sepsis who are at low risk of death.

RESOLVE study: Data from a placebo-controlled clinical trial did not establish efficacy of Xigris
in pediatric patients suffering from severe sepsis, acute infection, systemic inflammation and
respiratory and cardiovascular organ dysfunction.

This trial was stopped for futility after 477 patients had received the study drug (out of 600 patients
intended). A planned interim analysis (with 400 patients enrolled) showed a low likelihood of
demonstrating a significant difference in the primary endpoint of “Composite Time to Complete
Organ Failure Resolution”.

There was also no difference in 28-day mortality (17.1% versus 17.3% in the Xigris and placebo
groups, respectively).

Investigators attributed 2 deaths in the Xigris group and 5 deaths in the placebo group to bleeding
events.

Page 8 of 12
There was a higher rate of central nervous system (CNS) bleeding in the drotrecogin alfa
(activated) versus the placebo group. Over the infusion period (study days 0-6) the number of
patients experiencing CNS bleeding was 5 versus 1 (2.1% versus 0.4%) for the overall population
(drotrecogin alfa (activated) versus placebo), with 4 of the 5 events in the drotrecogin alfa
(activated) group occurring in patients ≤ 60 days old or ≤ 3.5 kg bodyweight. Fatal CNS bleeding
events, serious bleeding events (over the infusion period and over the 28-day study period),
serious adverse events, and major amputations were similar in the drotrecogin alfa (activated) and
placebo groups. In placebo controlled clinical trials, the treatment effect was most evident at sites
enrolling larger numbers of patients.

Reasons for withdrawal

In April 2009 FDA has already issued an early communication about ongoing safety review for
Xigris; although the communication was not intended as advice for discontinuation for prescribing
Xigris as there was no conclusion on the causal relationship between the drug and the emergency
safety issue it expressed the FDA analysis on the available data concerning Xigris. The information
was provided from a retrospective study (Gentry et al.: Adverse outcomes associated with the use
of drotrecogin alfa (activated) in patients with severe sepsis and baseline bleeding precaution,
Critical Care Med 2009), which reported an increased risk of serious bleeding events and of death
in patients with sepsis (a severe illness related to a bloodstream infection) and baseline bleeding
risk factors who received drotrecogin alfa (activated). The study was a retrospective medical
record review of 73 patients who received Xigris. Serious bleeding events occurred in 7 of 20
patients (35%) who had a bleeding risk factor vs. only 2 of 53 (3.8%) patients without any bleeding
risk factors. More patients with baseline bleeding risk factors died (13/20; 65%) compared to
patients without any bleeding risk factors (13/53; 24.5%). The authors acknowledge that there are
limitations to this study, such as its retrospective design and the small size of the patient
population, that limit the ability to draw definitive conclusions from the data. The current (at that
time) prescribing information for Xigris described the increased risk of bleeding, and included a
statement in the Warnings and Precautions section that bleeding is the most common serious
adverse reaction experienced by patients who received the drug. The Warnings and Precautions
section also lists a number of risk factors for the increased risk of bleeding that should be taken
into account when considering use of Xigris therapy. The Contraindications section states that
Xigris is not to be used in the clinical situations where bleeding could lead to significant morbidity
or death. Overall, the finding by Gentry et al. of an increased risk of death and serious bleeding
events in patients treated with Xigris who also have baseline bleeding risk factors is consistent with
the information in the current product label. Prescribers should refer to the product label for the
specific contraindications, warnings, and, precautions and carefully weigh the increased risk of
bleeding against the benefits of Xigris. FDA is working with the manufacturer to further evaluate
Page 9 of 12
the incidence of serious bleeding events and mortality in patients who received Xigris. FDA will
communicate its conclusions and any resulting recommendations to the public when our review is
completed, which may take several months. This communication is in keeping with FDA’s
commitment to inform the public about its ongoing safety reviews of drugs. (FDA Early
communication about Ongoing Safety review Xigris, April 2009).

EMA requested that Lilly to conduct an additional multi-center, placebo-controlled, phase III trial
(PROWESS-SHOCK) in severe sepsis patients at high risk of death. This study reported a 28-day
all-cause mortality rate of 26.4% in patients treated with drotrecogin alfa compared with 24.2% in
the placebo group and this was not considered to be statistically or clinically significant. The severe
bleeding rate was about 1% in both arms.

Since the PROWESS-SHOCK trial found no significant reduction in mortality in patients treated
with Xigris compared with placebo, Eli Lilly decided to take action. The company discontinued all
other ongoing clinical trials, and announced a voluntary withdrawal of Xigris from the market.
Patients should not receive Xigris, and all remaining Xigris products was to be returned to
wholesalers / drug suppliers. (FDA Drug Safety Communication: Voluntary market withdrawal of
Xigris due to failure to show a survival benefit, 25Oct2011)

Could this be prevented?

The initial FDA approval process was in 2001 under the questions of conflicting opinions during the
Anti-Infective Drugs Advisory Committee meeting. Several concerns with regard to the Protein C in
Severe Sepsis (PROWESS) trial were raised such as: severity scores such as APACHE II were
not created for making treatment decisions on individual patients and even more important
(although not clearly stated) the “fundamental question of whether analyses of subgroups,
although derived from an RCT with an overall positive result, should be considered a reliable
methodology for drug approval and labelling”. (D. Poole, G. Bertolini and S. Garanttin, Errors in the
approval process and post-marketing evaluation of drotrecogin alfa (activated) for the treatment of
severe sepsis, The Lancet Infectious Disease, February 2009)

Both regulatory agencies recognized efficacy and safety issues - the drug increased the number of
severe bleedings and required post-marketing studies to support the licensing. All studies failed in
confirming the first positive results but their findings were not considered conclusive and the
license was anyway granted.

In 2004, the surviving sepsis campaign (SSC) guidelines, funded mainly by Eli- Lilly the
manufacturer of Xigris, provided a high-grade recommendation for the use of the drug. However,
the scientific community was split between the supporters and the antagonists of Xigris, and the
upcoming negative findings convinced the EMA in requiring a confirmatory trial on patients with
septic shock (the PROWESS-Shock trial) in 2007. Furthermore, the 2008 revision of the SSC

Page 10 of 12
guidelines downgraded the evidence provided by the original PROWESS trial to those of an
upgraded observational study, recognizing its limits. However, the drug was not withdrawn from
the market and it was only voluntary withdrawn by the manufacturer based on the results of
PROWESS-Shock trial. (D. Poole, G. Bertolini and S. Garanttini in “Withdrawl of Xigris from the
market: old and new lessons” , Journal of epidemiology and community health in March 2012)

Lessons Learnt
The 10 years history of Xigris comes with a series of considerations for the future as summarized
by D. Poole, G. Bertolini and S. Garanttini in “Withdrawl of Xigris from the market: old and new
lessons” (published in Journal of epidemiology and community health in March 2012):

 Looking at the different (opposite) responses to Xigris in terms of efficacy and safety in 2
similar studies with similar population questions the generalization of the results of a single
randomized clinical trial
 The low mortality rate in both studies (PROWESS and PROWESS-Shock) (between 24%-
34%) is significantly different with the rate from several epidemiological studies (reporting
mortality of 50% due to septic shock in ICU) meaning that eligibility criteria have limited the
standard population and so the use of medication is also limited
 Neither FDA nor EMA have taken into consideration the criticism raised during and after
approval and both have interpreted the results of post marketing studies in favor of the
manufacturer
 Keeping the drug on the market during the conduct of the confirmatory trial suggested that
the regulatory agencies have trusted (by maintain the license) and distrusted (by asking for
the confirmatory trial) at the same time the drug
 Trying to address some of the criticisms, the approval was restricted to a subgroup of the
study population that was not predefined, so violating well-known methodological rules and
guidelines established by the FDA itself. EMA approved the drug following a similar biased
approach.

In the end the story of Xigris questions the balance between industry and public health interest and
moreover questions the standards applied by the regulatory agencies for the sensitive areas of
unmet medical need (such as severe sepsis and oncology). Lowering the standards ends up not
only is increased healthcare costs but in exposing patients to avoidable and unnecessary adverse
events which should be avoided for the future.

(D. Poole, G. Bertolini and S. Garanttini in “Withdrawl of Xigris from the market: old and new
lessons” , Journal of epidemiology and community health in March 2012)

Page 11 of 12
Bibliography

 FDA approval letter, 21Nov2001

 Forbes 17Oct2001
 Lilly Announces Withdrawal of Xigris® Following Recent Clinical Trial Results, Lilly
PRWire, October 2011
 D. Poole, G. Bertolini and S. Garanttin, Errors in the approval process and post-marketing
evaluation of drotrecogin alfa (activated) for the treatment of severe sepsis, The Lancet
Infectious Disease, February 2009
 Activated protein C in septic shock: A propensity-matched analysis, Critical Care March
2011
 FDA Anti-Infective Drugs Advisory Committee 12 September 2001; Briefing Document for
Xigris for the Treatment of Severe Sepsis; Eli Lilly and Company
 Puxty, P. McConnell, S. Crawley, S. McAree, T. Quasim and S. Ramsay, Recombinant
activated protein C usage in Scotland: a comparison with published guidelines and a
survey of attitudes, Anaesthesia 2012, The Association of Anaesthetists of Great Britain
and Ireland
 Clinical trials in severe sepsis with drotrecogin alfa (activated); Pierre-François Laterre)
 Summary of product characteristics for Xigris
 FDA Early communication about Ongoing Safety review Xigris, April 2009
 FDA Drug Safety Communication: Voluntary market withdrawal of Xigris due to failure to
show a survival benefit, 25Oct2011
 D. Poole, G. Bertolini and S. Garanttini in “Withdrawl of Xigris from the market: old and new
lessons” , Journal of epidemiology and community health in March 2012
 Drotrecogin Alfa (Activated) in Adults with Septic Shock, V. Marco Ranieri, M.D., B. Taylor
Thompson, M.D., Philip S. Barie, M.D., M.B.A., Jean-François Dhainaut, M.D., Ivor S.
Douglas, M.D., Simon Finfer, F.R.C.P., Bengt Gårdlund, M.D., John C. Marshall, M.D.,
Andrew Rhodes, M.D., Antonio Artigas, M.D., Ph.D., Didier Payen, M.D., Ph.D., Jyrki
Tenhunen, M.D., Ph.D., Hussein R. Al-Khalidi, Ph.D., Vivian Thompson, M.P.H., Jonathan
Janes, M.B., B.Ch., William L. Macias, M.D., Ph.D., Burkhard Vangerow, M.D.,and Mark D.
Williams, M.D., for the PROWESS-SHOCK Study Group*, The New England Journal of
Medicine, 31May2012)

Page 12 of 12