Diabetes Care Volume 38, July 2015 1181

DIABETES CARE SYMPOSIUM
Arthur T. Sands,1 Brian P. Zambrowicz,1
Sotaglif lozin, a Dual SGLT1 and Julio Rosenstock,2 Pablo Lapuerta,1
Bruce W. Bode,3 Satish K. Garg,4
SGLT2 Inhibitor, as Adjunct John B. Buse,5 Phillip Banks,1
Rubina Heptulla,6 Marc Rendell,7
Therapy to Insulin in Type 1 William T. Cefalu,8 and Paul Strumph1

Diabetes
Diabetes Care 2015;38:1181–1188 | DOI: 10.2337/dc14-2806

OBJECTIVE
To assess the safety and efficacy of dual sodium–glucose cotransporter (SGLT) 1
and SGLT2 inhibition with sotagliflozin as adjunct therapy to insulin in type 1
diabetes.

RESEARCH DESIGN AND METHODS
We treated 33 patients with sotagliflozin, an oral dual SGLT1 and SGLT2 inhibitor,
or placebo in a randomized, double-blind trial assessing safety, insulin dose,
glycemic control, and other metabolic parameters over 29 days of treatment.
RESULTS
In the sotagliflozin-treated group, the percent reduction from baseline in the
primary end point of bolus insulin dose was 32.1% (P = 0.007), accompanied by
lower mean daily glucose measured by continuous glucose monitoring (CGM)
of 148.8 mg/dL (8.3 mmol/L) (P = 0.010) and a reduction of 0.55% (5.9 mmol/mol)
1
(P = 0.002) in HbA1c compared with the placebo group that showed 6.4% reduction in Lexicon Pharmaceuticals, Inc., The Woodlands, TX
2
bolus insulin dose, a mean daily glucose of 170.3 mg/dL (9.5 mmol/L), and a decrease Dallas Diabetes and Endocrine Center, Dallas, TX
3
Atlanta Diabetes Associates, Atlanta, GA
of 0.06% (0.65 mmol/mol) in HbA1c. The percentage of time in target glucose range 4
University of Colorado Denver/Barbara Davis
70–180 mg/dL (3.9–10.0 mmol/L) increased from baseline with sotagliflozin com- Center for Childhood Diabetes, Aurora, CO
5
pared with placebo, to 68.2% vs. 54.0% (P = 0.003), while the percentage of time in University of North Carolina School of Medicine,
hyperglycemic range >180 mg/dL (10.0 mmol/L) decreased from baseline, to 25.0% vs. Chapel Hill, NC
6
Albert Einstein College of Medicine/Montefiore
40.2% (P = 0.002), for sotagliflozin and placebo, respectively. Body weight decreased Medical Center, Bronx, NY
7
(1.7 kg) with sotagliflozin compared with a 0.5 kg gain (P = 0.005) in the placebo group. Creighton Diabetes Center, Omaha, NE
8
Pennington Biomedical Research Center, Baton
CONCLUSIONS Rouge, LA
As adjunct to insulin, dual SGLT1 and SGLT2 inhibition with sotagliflozin improved Corresponding authors: Arthur T. Sands, arthur
glycemic control and the CGM profile with bolus insulin dose reduction, weight .sands@bcm.edu, and Paul Strumph, pstrumph@
lexpharma.com.
loss, and no increased hypoglycemia in type 1 diabetes.
Received 26 November 2014 and accepted 12
March 2015.
Therapy for type 1 diabetes has advanced considerably since the historic publication Clinical trial reg. no. NCT01742208, clinicaltrials
of the Diabetes Control and Complications Trial (1) in 1993 with the introduction of .gov.
new fast-acting and basal insulin analogs, more accurate blood glucose meters, This article contains Supplementary Data online
at http://care.diabetesjournals.org/lookup/
smaller and more technically advanced insulin pumps, and the availability of con-
suppl/doi:10.2337/dc14-2806/-/DC1.
tinuous glucose monitoring (CGM). Despite these advances, sustained improve-
© 2015 by the American Diabetes Association.
ments in glycemic control are still associated with hypoglycemia and severe Readers may use this article as long as the work
hypoglycemia (SH). The 12-month risk of SH associated with seizure or loss of is properly cited, the use is educational and not
consciousness was recently reported at 11.6% in adults and 9.9% in youth, rising for profit, and the work is not altered.

30% of with potent SGLT2 inhibition and partial sulin delivery method: either continuous patients with type 1 diabetes in the U. cent study in patients with type 2 diabetes on an outpatient basis during patients’ tose also exhibit unformed watery stools. trointestinal (GI) tract via SGLT1 inhibi. improve glucose control by reducing delivery regimen: either CSII or MDI. placebo-controlled. a condition significant decrease in PPG excursion supervision of initial insulin dose adjust- characterized by severe diarrhea. most pharmaceutical discov. providing a stark reminder of the tients with galactose malabsorption at bolus insulin. small-molecule dual inhibitor zin on postprandial glucose. thereby avoiding the GI subcutaneous insulin infusion (CSII) or are obese (9. inadequately controlled with metformin. The first treatment dose (day ery programs focused on selective SGLT2 impairment despite the expected reduc. holds promise to type 1 diabetes using their previous insulin transporter primarily responsible for re. This study was a randomized. and moderate to severe renal impairment usual insulin. in in. These data were consistent with a in adults with inadequately controlled type cemic state seeking a “safety margin” from “window” for achieving glycemic efficacy 1 diabetes as adjunct therapy to usual in- hypoglycemia (8). Here. have metabolic syndrome (11). 1. tion of UGE suggesting that. obtain at least 3 days of blinded CGM data on a diet containing glucose and galac. In a dose-ranging either a total daily dose of 400 mg sotagli- mice given a meal challenge containing study in patients with type 2 diabetes flozin or placebo taken within 15 min prior glucose exhibit decreased blood glucose. and activity regi- decreased food intake. . we present the results of a randomized. tient subgroup with more severe renal samples. multicenter. SGLT1 with SGLT2 half-maximal inhibi. the additional inhibition of SGLT1 by sota- indicating that between 4 and 10% of ally. excretion (UGE) supporting meaningful doses through use of daily diaries. have confirmed the effects of sotagliflo. There is a tion. to record baseline insulin for inhibition of intestinal SGLT1 (18. sotagliflozin produced a ducted in an inpatient setting to allow and galactose malabsorption. the major intestinal tion of sotagliflozin in adult patients with glucose cotransporter (SGLT) 2. However.26).19. tion and renal glucose reabsorption via lowered blood pressure with a low risk to receive. and an increase in GLP-1 secretion (27). that assist in glycemic tablish preliminary safety and to provide primary transporter for absorption of and appetite control (22.24. In a re. Subsequent (16). and reduced (estimated glomerular filtration rate 15–59 men. and normal weight the insulin regimen without weight gain. (20). findings consistent with glucose mL/min/1. multicenter. control while concomitantly simplifying for . orally de. to breakfast for 29 days. The cause of death for mal food intake. and group. investigators tine and cecum and increased GLP-1 We hypothesized that in type 1 diabe. risk of hypoglycemia.23). examined in several longitudinal studies. in a double-blind fashion. SGLT1 is the (PYY) (18. tered. double- worried about hypoglycemia and a similar patible with relatively normal GI function blind evaluation of sotagliflozin treatment proportion purposely maintain a hypergly. proximately one-third of patients are vere reduction in SGLT1 activity is com. The nal glucose reabsorption. sotagliflozin would improve glycemic acting insulin at each meal with guidance . the effects of breakfast with no bolus insulin adminis- knockout mice also exhibit increased de.19). ments and to obtain multiple pharmaco- fants with mutations in SGLT1 (16). study design is presented in Fig. In patients with type enrolled in the placebo-controlled portion to reduce glucose absorption in the gas. SGLT2 inhibition (17). Subsequently. the glucose transporter.50% of patients side effects of complete SGLT1 inhibi. reduced body weight. In contrast to selective SGLT2 inhibitors. double-blind evalua- Highly selective inhibitors of sodium– Inhibiting SGLT1.40 years (2. placebo-controlled. thereby lowering the poten- risks of tight glycemic control with insulin birth could tolerate a normal diet by tial for postprandial hypoglycemia. Additionally. The initial 7 days (days 27 to 21) of the small intestine and cecum. alone.73 m2). multiple dose injection (MDI). centration of 0.6% in adults with type 1 diabetes challenge but have normal stools. Recent data also indicate that ap. intestinal SGLT1 inhibition are main. and using an interactive web response system. adjusted the suggested dosage of short- release after a glucose-containing meal tes. such as The study was initiated with an open- and under exploration in patients with GLP-1 and polypeptide tyrosine tyrosine label pioneer group (n = 3) on CSII to es- type 1 diabetes (13–15). 1) was administered before a mixed- inhibitors to avoid potential GI side ef.S. increased delivery of glucose to the distal HbA1c reduction nearly doubled as sota. nor. suggesting that se. Addition. Sglt1 knockout of hypoglycemia (17. kinetic (PK) and pharmacodynamic (PD) this in mind. SGLT1 inhibition. are approved postprandial glucose peaks and stimu. heterozygous Sglt1 effects of SGLT2 inhibition. of the study and randomly assigned 1:1. dial glycemic excursions and decrease (4–7). patients on either MDI or CSII were of SGLT1 and SGLT2 that was designed PYY (17–19. GLP-1.10) and . to completion of treatment of the pioneer livered.1182 Dual SGLT1 and SGLT2 Inhibition in T1D Diabetes Care Volume 38.25). ing glucose and galactose (18). and increased gliflozin dose increased in the absence of study were used to obtain baseline labo- GLP-1 release indicating a potential utility additional increases in urinary glucose ratory samples.0018 mmol/L Study Design without weight gain or an increase in the and SGLT1 half-maximal inhibitory con. it has been reported that most gliflozin was predicted to lower postpran- deaths can be attributed to hypoglycemia individuals in a large family cohort of pa. tained as kidney function declines. changed. the age of 20 years. unlike the meal tolerance test (MMTT) prior to fects. 2 diabetes.3). Days 1 and 2 of the study were con- weight.21). dietary. individuals with type 1 diabetes has been gain when maintained on a diet contain. Preclinical information on insulin dose adjustment glucose and galactose in the intestine and clinical studies conducted to date during initiation of treatment. With This difference was preserved in the pa. ered HbA1c. sotagliflozin treatment low.036 mmol/L (17). and to Homozygous knockout mice maintained intestinal SGLT1 inhibition (26). Basal insulin was continued un- livery of glucose to the distal small intes. Sotagliflozin fulfilled these criteria clear need for the development of new with 20-fold selectivity for SGLT2 over RESEARCH DESIGN AND adjunct therapies to insulin that can im. Sotagliflozin is a novel. July 2015 to 18. METHODS prove glycemic control in this population tory concentration of 0. for the treatment of type 2 diabetes (12) lating release of GI peptides.

Northridge. lipid in the randomized. the study between 8 February and 20 The primary outcome of the study was Assessments of safety included vital November 2013. with 3 patients in the the treatment effect on change from signs. and laboratory open-label pioneer group and 33 patients baseline of total daily bolus insulin parameters (chemistry. multiple PK and PD samples. HbA1c as well as changes in GLP-1. The PK population included structed to resume their regular rou. 2. mean amplitude of samples to estimate key parameters for at the Enlite subcutaneous glucose sensor glucose excursion (MAGE). Results for the pioneer period. and urinalysis). and bedtime/overnight: 100– under the curves (AUCs).care. patients dex (HBGI). defined as the least one of the days of sampling. Base- bolus insulin at each meal. comes associated with CGM included patients were discharged and insulin percent time in defined ranges. and all patients provided written mmol/L]. the double-blind portion of the study.org Sands and Associates 1183 Figure 1—Study design. all patients who received at least one dose tines. The secondary outcomes per.0 mmol/L) prised of all randomized patients in the ment of scheduled self-monitoring of and 2) .2 and glucose excursion during the 3-h pe. informed consent. mmol/L) and #180 mg/dL (10. . Blinded CGM data were collected cluding coefficient of variation. and total daily bolus plus study included urogenital infections and shown in Table 1. The safety population in- . as well as yses using this population served as the American Diabetes Association dietary changes in mean daily interstitial glucose. population. outcomes pertaining to glycemic control Study Oversight to-target blood glucose goals consistent included assessing the effect of sotagli. Anal- blood glucose (SMBG). collection and review of adverse events.9 mmol/L) and $70 mg/dL (3. The secondary hypoglycemia. were included as secondary outcomes in. Secondary out. Safety and toler.. postprandial: .9 The intent-to-treat population was com- by the patient and investigator assess. double-blind cohort. primary population for statistical analyses recommendations. patients returned to inpatient for ceeded 1 SD over each 24-h period. Ad. hematology. electrocardiograms. mean SD of of study drug and had sufficient. and received $1 dose of study drug. as de.4–7. and high blood glucose in. 1) . summarized in Supplementary Fig. Patient disposition is total daily basal insulin. from an algorithm based on treat. cokinetic summaries were based on the PK 28. With standard . ing investigative sites approved the pro- mmol/L]. total daily verse events of special interest in this line characteristics of the patients are basal insulin. valid PK on all patients throughout the study with weekly glucose level.70 Statistical Analysis doses were to be adjusted as determined mg/dL (3. Inc. CA).180 mg/dL (10. PYY. and UGE over a 3-h post– A total of 36 patients were enrolled in Study Outcomes breakfast MMTT. The human research committees and/or with current standard of care (fasting flozin on fasting plasma glucose (FPG) institutional review boards of participat- and preprandial: 80–130 mg/dL [4.0 mmol/L) and placebo controlled expansion group. low blood cluded all patients who were randomized (with usual insulin dosing) and to obtain glucose index.36 h for end-of-treatment MMTT scribed by Baghurst (28). Exploratory measures included were then discharged and followed for the effect of sotagliflozin treatment on RESULTS an additional week.9 mmol/L). Additional measures of glucose variability and reporting.180 mg/dL [10 riod after an MMTT as measured by area tocols. placebo-controlled.6–10 mmol/L]).diabetesjournals.250 mg/dL (13. On day 2. patients were in. Pharma- (Medtronic. Patients C-peptide. ability of sotagliflozin were assessed by group were used to evaluate safety and in- tained to specifics of insulin use includ. 180 mg/dL [5. On day average of all glucose excursion that ex. dose during the outpatient treatment profiles. form the insulin-adjustment paradigm for ing change from baseline of total daily which were followed until resolution.

89 (3. $70 mg/dL (3. 0.9) The changes from baseline at lunch and HbA1c (%).24 (3.86) mmol/L) for placebo (P = 0. (Tables 2 and 3). Basal Insulin and Total Daily Insulin while percentage of time spent in the There were no SH events in either group.60 0. a pre.93 Weight (kg).83 (5. placebo (68. breakfast (P = 0. The sotagli.9) mmol/L) vs. zin compared with 0.052) (Table 3).9 baseline for both groups was minimal. 7.6) 74. hyperglycemic range. mean (SD) 76.63 (2. flozin and placebo groups exhibited a minutes of glucose readings .66) 63.9 mmol/L) (6. Of these.1) (P = 0.63 (4.11 (9. 54.8% vs.80) (Table Age (years). 14. n (%) 3). The use of basal insulin was similar be. mean 0.8 mg/dL the day. in hyperglycemic ranges.2%.0) 118. specified subgroup analysis of bolus The effects of sotagliflozin on HbA1c were sotagliflozin therapy resulted in a lower insulin use before major meals was con.2 (55.02) group.70 mg/dL .96) 9.68 (0. P . the mean HBGI was reduced from baseline in the sotagli- flozin group compared with an increase in Outcomes Total daily insulin was lower for the so.5 mg/dL (8. 40.8 (3.1 (3. 13.4 group compared with a numerical in.9 mmol/mol) from baseline placebo group (P = 0. Hypoglycemic events PPD. 42.180 mg/dL Hypoglycemic events per patient per day tween the groups.029.06% (0. Duration of diabetes (years). Sex.8 (7. n (%) White/Caucasian 14 (82) 16 (100) CGM during a 3-h period. Over the outpatient treatment period.94 (0. 107.5 (21. calculated as 50.7% vs.002) (Table 3 and Supplementary Fig.1 (9. P = mmol/L).9 mmol/L) (6.7.2) Notably. mean (SD) 26. In addition.7% for the placebo group There was a significant decrease in HbA1c the sotagliflozin group and 26. 162 requirements were similar whether pa.0%. 55) 3 and Supplementary Fig.75) 137. mean (SD) 41.0 mmol/L) compared with ber of events was 304.99) 4. was lower compared (PPD).04) flozin than placebo but did not reach FPG (mmol/L).5 mmol/L) (P = Total hypoglycemic events defined as patients treated with sotagliflozin com.65 mmol/mol) Glucose Levels by CGM tered once daily before breakfast. median (range) 34. mean (SD) 8. (Table 2). 5. Reductions of bolus insulin from (8. 1). July 2015 Table 1—Demographic characteristics (intent-to-treat population) similar percentage of time spent in the Patient characteristics Placebo (N = 17) Sotagliflozin (N = 16) hypoglycemic ranges of . patients in SMBG #70 mg/dL (3.022) (Table 3).08).82) mmol/L) for the sotagliflozin treatment Serum BUN (mmol/L). mean (SD) 138. sotagliflozin treatment Female 9 (53) 8 (50) Male 8 (47) 8 (50) produced favorable effects on postpran- dial glycemic excursions as assessed by Race.87) 41. 0. there was less glucose variabil- Serum sodium (mmol/L). declined significantly from base- There was a numerical decrease from 0. and baseline of 2.002) (Table 2). Of these. blinded CGM (defined as $10 continuous cebo group (P = 0.1 Hematocrit. .60 gliflozin treatment group compared with Daily insulin.0) 27.38 (6.3% the target glycemic range defined as (33%) were asymptomatic.1% at spent a greater percentage of time in 185 (52%) were symptomatic and 117 lunch (P = 0. n (%) After meals. HbA1c tal daily bolus insulin use was 232. defined as SMBG #70 mg/dL (3.3% (P = 0.55% (5.7 mg/dL (20. after 29 days of treatment with sotagliflo- Given that sotagliflozin was adminis.041) (Table Insulin therapy.23 mmol/L) BMI (kg/m2). 24. 225.3 mg/dL (9. mean (SD) 4. Con- sistent with the percentage of time spent BP.45 0.7% vs. mean (SD) 62. 1) of 148. In the sotagli- Asian 2 (12) 0 flozin group. P = 0.9 mmol/L) in the pared with placebo: 228. mean (SD) 119.98 (0.6. baseline of 15. the placebo group (P = 0.0% vs.2 (3. Tables 2 and 3). as was percentage of time spent in both groups hypoglycemia PPD was 0.3 mmol/L) compared with a placebo Hypoglycemia baseline before each meal were noted in value of 170. 1).3.8) 16.4% in the (P = 0.2% vs.0% in duction of 0. blood pressure. P = 0.94 (4. CGM SD was also lower on sotagliflozin MDI 5 (29) 6 (38) CSII 12 (71) 10 (63) treatment. similar whether patients were on MDI or mean daily glucose as measured by CGM ducted to detect differences throughout CSII.9 mmol/L) and #180 gliflozin treatment group.55) dinner were numerically lower for sotagli- HbA1c (mmol/mol).250 mg/dL (13.002) and a re.006) (Table 3).8%.8 mg/dL (6. Seated systolic BP (mmHg).49 58.86) 76. The ef. median (range) 72.7 (55.45 (3.96 (1.0 mmol/L). .51) 7.4% vs.010) (Table 3). ratio of bolus/total 0.2% in the pla.1184 Dual SGLT1 and SGLT2 Inhibition in T1D Diabetes Care Volume 38.046). calculated as 120.9% vs. 104.003) events (53%) were symptomatic and tients were on MDI or CSII.45) ity on sotagliflozin treatment as measured by MAGE.7 Serum creatinine (mmol/L).6% at the sotagliflozin treatment group placebo group numbered 354.007) (Tables 2 and 3). mean change from baseline Other 1 (6) 0 at breakfast was 216. (Supplementary Fig. by crease from baseline of 0. the total num- fects of sotagliflozin on bolus insulin mg/dL (10.4% for the sotagliflozin 1).8 for placebo (P = 0. In the sota- at dinner (P = 0. Bolus Insulin tagliflozin group. median (range) 18.2 mg/dL (20. mean (SD) 7. difference between groups) of 0.0 (21. 80 events (26%) were asymptomatic.0 for the sota- Total daily insulin (IU/kg). with a reduction from The percent change from baseline in to. 39.008) (Table 3).4. for the placebo group (P = 0. compared with 145. (Table 3 and Supplementary Fig. 57) 45. with placebo (25. and 223.1%.53.45) statistical significance.034) (Supplementary Table 2). 40. and the change from (10.94 (11.70 mg/dL (3.5 (4. line during treatment in both groups.

P values are calculated from two-sample t tests using the observed means.9 mol/L).41) N/A Hematocrit. The mean 3-h verse events led to discontinuation ing the study period. unless otherwise specified. and the last day of therapy (day 29).50 (1. In the sotagliflozin group. and uric placebo (P = 0.45 Safety Patients with any TEAE (%) 12 (71) 14 (88) N/A Patients with SAE (both with DKA‡) 0 2 N/A Hypoglycemic events (SMBG #70 mg/dL.018) (Table 2). change from baseline was assessed at day 29. Change from 36.7 pmol/L z h over 3 h with C-reactive protein. ment with 400 mg sotagliflozin given of 3.53 Total daily insulin change from baseline assessed at days 3–27 (%) 20. PPD) change from baseline assessed at days 3–27 20. P = 0.09 0.7 6.41 (0. (day 36).1 (1. N/A. which returned toward baseline at day treated with sotagliflozin.9 0.4* 20. change from baseline at day 29 (day 36) 21.15 20. patients continued UGE was higher in the sotagliflozin group from the study.05. Urinary Glucose. change from baseline at day 29 (day 36) 21.9 mmHg in the disorders.7 215. P = 0.9 24. PYY.0 pmol/L z h over meaningful changes in other exploratory to the study drug.5 21. No ad.77 Hypoglycemia (CGM $10 continuous min .029 Mean body weight change from baseline assessed at day 29 (kg) 0.1 g/3 h vs. therefore.63) N/A Serum BUN (mmol/L).11) 1. 1 week off therapy.2 29. creased by 0.63 (0. baseline–day 36) 117 80 N/A SH 0 0 N/A Hypoglycemia (SMBG #70 mg/dL.1 0. not applicable. tients (18%). acid (data not shown). tients using insulin infusion pumps and crease in the placebo group (21. There were no as pump related and were not related PYY was increased by 6.53 (1.care.025) (Table 2). the last day of therapy.75 Laboratory values associated with volume status Serum sodium (mmol/L).7 kg vs. ‡Both were assessed as due to insulin pump and deemed not drug related.009) (Table 2). declined numerically during status (serum sodium.55* 0. are provided in Supplementary Table 1. once daily before breakfast resulted in Laboratory values associated with volume balance being three reports of nausea significant reductions in bolus insulin .005 PYY postmeal AUC change from baseline assessed at day 29 (pmol/L z h over 3 h) 20. †Day 1 is not a true “baseline”. with a decline of 4.00 (20.018 Seated systolic blood pressure change from baseline assessed at day 29 (mmHg) 23.002 FPG change from baseline assessed at day 29 (mg/dL) 39.5) N/A For laboratory values. SAE.org Sands and Associates 1185 Table 2—Overall summary of results Placebo (N = 17) Sotagliflozin (N = 16) P Efficacy HbA1c change from baseline (%) 20. and Treatment differences in incidence of hematocrit) were assessed at baseline. eight sotagliflozin patients their usual activity levels and diet. events (TEAEs) in the sotagliflozin group Diabetes Association–recommended Systolic blood pressure decreased in both by MedDRA System Organ Class were GI glucose targets goals and maintaining groups. TEAEs Body Weight. serum BUN [blood urea nitrogen].5 kg) (P = 0. baseline–day 36) 354 304 N/A Documented symptomatic hypoglycemia (SMBG #70 mg/dL.06 20.53) 2. cebo-controlled trial over 29 days. The most frequently re. CONCLUSIONS An MMTT performed on the last We evaluated dual inhibition of SGLT1 treatment day of the study showed a Adverse Events and SGLT2 using sotagliflozin as adjunct lower serum glucose 3-h AUC in the so.70 mg/dL. serum creatinine.3* 0.4 0. (3. Details of the events 3 h in the sotagliflozin group and de.7* 0. Both cases of 0. baseline–day 36) 185 162 N/A Asymptomatic hypoglycemia (SMBG #70 mg/dL. Other Outcomes week after last dose of study medication Two serious adverse events of diabetic Mean body weight decreased with sota.0* 0. 12 (71%) patients on placebo. TEAEs were not statistically tested.50) N/A Serum creatinine (mmol/L). respectively.0* 0. 0. Fourteen (88%) patients on sotagliflozin to insulin in inadequately controlled tagliflozin group compared with placebo reported adverse events compared with type 1 diabetes in a double-blind. consistent with reversible perfusion DKA were assessed by the investigators baseline AUC after a meal challenge for and volume effects.2 22.02 (20. Dur- tively.15 Daily bolus insulin change from baseline assessed at days 3–27 (%) 26. pla- (595 mg z h/dL vs. PPD) change from baseline at days 3–27 20. there ketoacidosis were reported in two pa- gliflozin treatment compared with an in. their usual insulin delivery regimens compared with placebo (29.6 0.7* 0.025 Postmeal plasma glucose AUC (mg z h/dL over 3 h) at day 29† 761 595 0. end points including stimulated C-peptide. ported treatment-emergent adverse while attempting to achieve American 9.007 Daily basal insulin change from baseline assessed at days 3–27 (%) 0. with the most notable im.4 (0) 2.53) 20.4 232. Treat- sotagliflozin-treated group and a decline (50%) compared with three placebo pa.005) (Table 2). were numeric increases in most values. and day 36. serious adverse event. *P . change from baseline. respec. treatment in both groups.45) (Table 2). 761 mg z h/dL. Bold values are statistically significant. on sotagliflozin versus one on placebo.0 218. change from baseline at day 29 (day 36) 0.005 Postmeal urinary glucose (g/3 h) at day 29† 9.2 g/3 h. triglycerides. and 1 are summarized in Supplementary Table 1.diabetesjournals. change from baseline at day 29 (day 36) 20.9 with placebo (P = 0.

3) 20. Patients treated with sotagliflozin numerically less hypoglycemic events flozin versus differences in trial design.2) 22. Arithmetic change from baseline is shown.5) 21.007 Total daily basal 26. dose while improving glycemic control were accompanied by a 0.2)* 210. and titis in the posttreatment follow-up period.8 (18.2 (13.4) 40.006 LBGI 2.5 [NC] 0.9 (12.7) 2.0] 0. where compared with placebo.5) 5.7) 148.9 [NC] 0.01) 0.2 (4.6 (3.7 (18.1)* 20.6] 5.53 Total daily (basal + bolus) 45. less time study in patients with type 1 diabetes. July 2015 Table 3—Summary of CGM and CGM-derived results and prespecified insulin dose analysis Placebo (N = 17) Sotagliflozin (N = 16) Change from Change from Baseline Treatment baseline value [%] Baseline Treatment baseline value [%] P CGM mean daily glucose (mg/dL) 160.4 (2. SGLT1 inhibition in the GI tract and MMTT.55% reduction blood pressure decrease in the sotagli- by multiple measures including lower of HbA1c after 29 days of treatment with flozin group (24. and of short duration tive SGLT2 inhibitors. the favorable effects on compared with a weight gain (0.09 (1.7) 5.6 (25.80 70–180 55. These cose levels by CGM.8 (4.010 CGM hypoglycemia events/ 1. The systolic the sotagliflozin group reported an event .29–31) increased hypoglycemic events.75 patient/day ($10 continuous min .5)* 220.9) d [13. demonstrated weight loss (21.6 (14.6 (8.9 (7.7 kg) PPD in the sotagliflozin-treated group Importantly. change from baseline.0 [NC] 0.0) 22.2) 20.0 (11.6) 68.7) 22.2 [NC] 1.3) 27.3 (14.0 (9. The baseline analysis period consists of days 26 to 22.6 (4.9 (1.0)* 214.7) 9.41) 20. Bold values are statistically significant.8) 35.70 8. Con- of time spent in target range.052 Data are mean (SD) unless otherwise indicated. pared with one patient in the placebo mary effect of SGLT1 inhibition is reduc.0) 5.4 (3.041 HBGI 8.4] 0. low blood glucose index.7 (6.6 [215.4] 23.2) 27.8) 5.5 kg) compared with placebo.9) d [7.5) 1.3 [232.5) 6. range. One case and.9) 37. The three cases that occurred sotagliflozin.2 (12. tions in bolus insulin use (15.4 (15.250 12. the 24-h SD.0) 21.7] 47. not calculated.1) 5. Fur. line for both treatment groups when SGLT2 inhibition in the kidney. extent to which this discrepancy is MAGE. NC.70 mg/dL) CGM % time in ranges (mg/dL) .3) 14.3) 6.and postmeal improvements in glu. of sotagliflozin.9) 58. P values are from least squares mean analyses of change from baseline scores (absolute and % change).0) 18. There were no cases of SH reported.9) 1.3] 0.7 (5.008 CGM variability measures SD (mg/dL) 57.1 (9.41 MAGE 135. Based on these findings. provement in glycemic control without group.33).7 (3.90 (0. and empagliflozin produced a 0.05.9) 170.9 mmHg).0 [NC] 0.9) 145. No patient on sotagliflozin that did not show any significant reduc. selec.5) 120.7 (3.7 (6. zin provided clinically meaningful im. 24-h glucose interquartile reported any genitourinary infections.002 .8 (4.0) d [228.6)* 27.3 (24.1)* 11.4 (15.2)* 28.4% reduc. on 24-h CGM analysis during the treat.1)* 22.8 (9.2 [0.046 Lunch bolus 5.6 (38.5 (39.1 [26.2) d [39.1) 4.47) 20.5 [NC] 9.3) 25.2 (5.1] 6.6) 3.022 Coefficient of variation 35. sotagliflo.8 (1. and driven by the SGLT1 inhibition of sotagli.6) 15.1 [NC] 0.5 [22. tion in HbA1c (32).9] 0.9 [NC] 0.32.0 (12. Two patients in daily glycemic control and insulin use for the placebo group.4) d [223.6 [NC] 0.1 (7.2) 1.0) 20.9) 4.2 (13.4 (2.4 (5.8 (30.4 (9. parameters provide confirmation of so- postprandial glucose was also demon. sotagliflozin produced a statistically sig.6 (15.2] 27.4 (5. This contrasts with trials of measures of glycemic variability based during treatment were early in onset.2 [NC] 0.1 [NC] 15.4] 0. mean daily sensor glucose. Improvement in prandial hypoglycemia. In an 8-week open-label the placebo group (23. of note. mild in intensity.9 [NC] 0.0 (11.9 (9.2 (2. occurred with significantly treated with sotagliflozin also demon. LBGI.9 mmHg) was similar to mean daily glucose.1 (7. was reflected by increased UGE.0 (17. could contribute to a lower risk for post.5 (25.0 (12.5 [NC] 35.1) 26.8 (11.5 (9. group reported an AE of nausea com- AUC at the end of treatment.5) 44. sistent with the SGLT1 inhibitory effects spent in hyperglycemic ranges.9 [NC] 163.1 [NC] 0.4) 25.2 [NC] 60.6) 1. and it is inter.5 (1.3] 7.6) 7.5 (34.7) d [225. empagliflozin and dapagliflozin.1) 54.2 [NC] 56.029 Breakfast bolus 4. 0.2 (3. a higher percentage sotagliflozin.5 (16.9 (14.61 Insulin dose data Total daily bolus (primary) 20.0 (3. occurred 3 days after cessation of treat- lower bolus insulin used by patients on strated significant improvements in ment.3 [NC] 37.9 (17.4 (2. esting to note that hypoglycemic events tagliflozin’s dual mechanism of action of strated by favorable effects during the PPD were numerically lower than base. The pri.9 [NC] 0. *P . Patients with increased GLP-1 activity.5 [20.2 (6. mone PYY was significantly increased Sotagliflozin also produced significant The reduction in bolus insulin use and sotagliflozin’s SGLT2 inhibitor effect pre. and the treatment analysis period consists of days 3–27.59) 0. in type 1 diabetes ment period.3 [NC] 7.06 (0. HBGI (a predictor for hyperglyce.6 (8.5) 50.8] 0. an effect possibly associated tion in postprandial glucose (19. while one patient on placebo reported cys- ther work is required to clarify the mia).2) 0.95 (0.08 Dinner bolus 6. postprandial GI hor- lower HbA1c.7) 0.180 35.003 . These measures included (2 days or less).7) 20.6 [NC] 1.2) 33.5 [NC] 145. measured by either SMBG or blinded Four patients in the sotagliflozin nificant decrease in 3-h plasma glucose CGM.8 (4.1186 Dual SGLT1 and SGLT2 Inhibition in T1D Diabetes Care Volume 38.8) 6.3 (3.

B.50:2439–2442 tion with insulin to improve glycemic Dance Pharmaceuticals.S. Merck. S. N Engl J Med dual SGLT1 and SGLT2 inhibitor.B. has received speaking/advisory board try. Lexicon. Harjutsalo V. Nonetheless. and diabetes mellitus.R. Astellas.S.Z. ment. T1D Exchange.180 mg/dL. contributed to the manuscript blood glucose levels lower than typically monitoring.H. a next-generation Medtronic. MAGE. a (mean SD.B... Transtech et al. Inc. The Diabetes Control and Complications Trial the study. P.L. Veritas. and immunomod. Pfizer. AstraZeneca. MA. Halozyme. are the guarantors of this gestive and Kidney Diseases and JDRF and has work and. Dahlquist G. contrib- Funding. Boston. contributed to the ceived speaking/advisory board consulting fees patient safety in an effort to lower the manuscript and reviewed and edited the manu- from the National Institute of Diabetes and Di- theoretical risk for episodes of SH. efforts is an investigator and/or consultant. bolus insulin Pharmaceuticals.. tween 70 and 180 mg/dL. R.W. wrote the manu- type 1 diabetes receiving the selective script. script and reviewed and edited the manuscript. contributed to the manuscript and the known effects of sotagliflozin on re. but in Acknowledgments. work he has personally received stock options from the Yorkshire Register of diabetes in chil- sotagliflozin significantly improved time and payments. Johnson & Johnson. has had grants/ References they were effectively unblinded during research support. 6. and contributed to the manuscript and reviewed and This initial study of sotagliflozin in type edited the manuscript. Inc. MannKind. also funded the study. A.98:3411–3419 apy alone remain daunting. Diseases.P.). Janssen.. and Verva and is a consultant important causes of death for children and to PhaseBio Pharmaceuticals. et al. contributed to the manuscript and reviewed was conservative with an emphasis on Jaeb Center for Health Research. This study was presented Duality of Interest. GlaxoSmithKline. This script. highly signifi. without any 4. of Diabetes Interventions and Complications Study Lexicon Pharmaceuticals. Sangeeta Sawhney for data review and reviewed and edited the manuscript.R.T. R. Diabe- control. J. Bristol-Myers Squibb. and change Clinic Network.W. J. Osiris.L. Xing D. dosing upon first dosing with sotagliflozin has received research and grant support from the W. Houston.G. analyzed ducing glucose absorption. Elcelyx. DKA will be closely monitored in all future variability. compared with basal over the outpatient Prior Presentation. and glucose variability tor on both basic research and clinical research mg/dL [19. Novo Nordisk. Notably. Ernest Wang for study management. contributed to the manuscript. GI 5. and his employer has received research and diabetic ketoacidosis in adults with type 1 di- betes.. Larger studies of a longer dura. under contracts between his Early mortality in EURODIAB population-based must also be made to identify adjunct employer and Andromeda. Paul Tubbs for project manage- associated with DKA (14. uted to the manuscript and reviewed and edited the manuscript. After 29 days’ La Roche. Bodansky HJ. investigators) to insulin pump therapy. T1D Ex- Medtronic. Boehringer Ingelheim. Roche. Pfizer. treatment period. Musen G.K. and guidance for insulin has received grant support from the NIH.org Sands and Associates 1187 of DKA.C. J Clin Endocrinol Metab 2013. Patterson CC. Diabetologia 2007.B. the challenges and burden of and grant support from Lexicon Pharmaceuticals. Eli Lilly. editing. type 1 diabetes in the T1D Exchange clinic reg- b-cell transplantation. Jacobson AM. and QC. daily bolus Novartis. percent time spent be. abetes: results from the T1D Exchange clinic regis- managing the disease with insulin ther.P.S. contrib- SGLT2 inhibitor empagliflozin.O.32). Novo Nordisk.356:1842–1852 cantly reduced HbA1c levels.250 mg/dL.Z. Pediatr Diabetes 2013. GlaxoSmithKline. has served as principal investiga- oratory blood glucose readings (. W. et al.B. Ryan CM. No other potential conflicts of interest relevant to type 1 diabetes trials. 2007. Inc. As work consulting fees from Eisai. participants or caregivers to believe P. Maahs DM. A. In addition. A. tes Control and Complications Trial/Epidemiology improve quality of life. Novo Nordisk.T. Inc. mean daily glucose. M. and own stock. Johnson & Johnson. AstraZeneca.G. J. 5–9 June employees of Lexicon Pharmaceuticals. has received 922–926 . opment Company. Weinstock RS. had full access to all the data in could have introduced bias in the results received research grants from the National In- the study and take responsibility for the integrity favoring a reduction of bolus insulin use stitute of Diabetes and Digestive and Kidney of the data and the accuracy of the data analysis. postprandial Rhythm. R..T. at the 2015.B. et al.4 mmol/L]) at presentation. et al.R. B. N Engl J Med 1993. Medtronic. proved both glucose control and glycemic and Lexicon Pharmaceuticals. Cengiz E.B.329:977–986 Lexicon. at the 75th Scientific Sessions of the American Inc. has re- and edited the manuscript.31: spent in the glucose range as measured and Lexicon Pharmaceuticals.. Bristol-Myers Squibb. TX. and received honorarium or consulting Research Group. lower the burden of disease. and body weight. All parties providing script.B.. for which young adults with type 1 diabetes: results increase in hypoglycemia risk. were cant reductions in insulin doses achieved Diabetes Association. contributed to the manu- of the study. and has served as a consultant for Intarcia Therapeutics and Sanofi.W. Inc. which was attributed (by the by the following CGM glucose indices: grant support from Lexicon Pharmaceuticals. two cases tion are needed to confirm these findings. administration was closely monitored in reviewed and edited the manuscript.K. Johnson & Johnson. Long-term effect of diabetes and treatment. P. Lexicon. 3. its treatment on cognitive function. contributed to the Kristi Boehm for assistance with manuscript manuscript and reviewed and edited the manu- 1 diabetes had several limitations. B. T1D Although insulin provides a lifesaving Associates and is a partner with Atlanta Diabetes Exchange Clinic Network. and has an investment interest in Thermalin Di- Both cases were associated with high lab. Feltbower RG. Sotagliflozin grants awarded to his institutions from AstraZeneca. Orexigen. finding expected in DKA. is a consultant for MannKind dren and young adults..B. Inc. Inc. contributed to the manuscript and reviewed and edited the manuscript. J.K. and analysis.S. and B. assistance currently are employees of Lexicon reviewed and edited the manuscript. introducing behavioral biases boards for..T. JDRF. Halozyme. is employed by Atlanta Diabetes 2. Inc. Xing D. since 1989.B. signifi. of DKA were reported in patients with Author Contributions. as such. The Robert and Janice McNair Founda- an inpatient setting during the first 48 h tion. Sanofi. Acute complications and drug misuse are blood glucose. and Sanofi.H. served on scientific advisory 1.care.. Diabetes Care 2008. The authors thank Ike uted to the manuscript and reviewed and edited both cases the patients presented with Ogbaa for protocol development and medical the manuscript. Tolerex. . P. Severe hypoglycemia therapy for individuals with type 1 dia. or were employees at the the data. and Dynamics. istry. Eli Lilly.350 and . LipoScience. Boehringer Ingelheim. and HBGI). with no Pharma. Merck. P. Intarcia Therapeutics. Wong JC. Patterson CC.S. Inc.C. Research Group. Severe hypoglycemia continues to develop disease-modifying Sanofi and has received research grants from Eli and diabetic ketoacidosis among youth with treatments such as the artificial pancreas. as an adjunctive treatment to insulin im. and T1D Exchange. The effect of intensive treatment that could have impacted the results fees from manufacturers of SGLT2 inhibitors of diabetes on the development and progression given the small numbers of patients en. Sanofi. and time the study was conducted (I.. Takeda. S. and P. Lilly.H. are employees of Lexicon Pharma- ceuticals. cohorts of type 1 diabetes diagnosed in childhood therapies that could be used in combina.G. B. given the serious nature of such events. does contract research with Medtronic..diabetesjournals. sotagliflozin. Merck. partly funded this study. Hoffmann. Metabolon.14:447–454 ulatory therapy to protect b-cells. direct financial benefit. of long-term complications in insulin-dependent rolled in this study. Finally. abetes. Lexicon Pharmaceuticals. With writing.T. Metavention.. S. Associates.S. and total daily insulin dose. this article were reported. and P. Metabolic Solutions Devel. in some patients may have led certain time the study was conducted and own stock.

Centers for Disease Control and Preven- cular disease in type 1 diabetes mellitus. 33. KGA-2727. Frazier K. Madsbad S. Xin B. Głowi nska. Ding ZM. Rosenstock RR. et al. Eur J Pharm Biopharm 2010. fects of LX4211. Im.6:453–467 22. Hirayama BA. Rosenwasser RF. Willis WD. Choksi (SGLT1). Diabetes Care 2011. a dual SGLT1/SGLT2 inhibitor. Cherney DZI. Lapuerta P. Kuhadiya ND.418:650–654 abetes 2012. Department of Health and Human Serv- Lera´ rio AC. im. Metab Syndr Obes 2013. Lapuerta P.37:71– (DCCT/EDIC) Research Group. Am J Physiol Endocrinol Metab mellitus and renal impairment despite low uri- of Diabetes Interventions and Complications 2013. Epstein BJ.38:412–419 in healthy subjects. Glucose Four weeks of treatment with liraglutide re. Teraoka Y. Diabetologia of human sodium glucose transporters. Expert Rev Pharmacoe. dapagliflozin in type 1 diabetes: a randomized.170:891–898 glucose-galactose malabsorption in a large co. a novel selective inhibitor tion. on glucosuria with LX4211. Henry RR. glucagon-like peptide 1. Chen L. Rosenstock J. glycemic control in type 2 diabetes. Edelman S. Diabetes Care treated patients with diabetes: results from a 2 diabetes in a randomized.14:912–923 of a high-affinity sodium glucose cotransporter 30. LX4211 82.75:366–374 13.38:431–438 European online survey. Wright EM. Zambrowicz B. Disparity in manage. Ef. Smith M. Cefalu W. 23. et al. Dia- transporter 2 inhibition in patients with type 1 di.345:250–259 29. 19. Baghurst PA. Biology 1173. exhibits antidiabetic efficacy in rodent boxyl group-terminated polyamidoamine den- R. a dual inhibitor of Aslam A. Cowley MA. 1):A475–A476 residual b-cell function. and features of metabolic syndrome in children 20.342:288– drimers bearing glucosides inhibit intestinal tial in the treatment of diabetes. Gastroenterology 2007. et al. Glycemia and cardiovas. Partridge H. Medical resource use. DaCosta CM. et al. placebo-controlled pilot study. Madec-Hily A. ´ of intestinal glucose. Centers for Disease Control and Prevention. Brown PM. e12 ment of diabetes and coronary heart disease risk increases serum glucagon-like peptide 1 and 28. obesity 2013. transport via SGLT1 is critical for post-prandial duces insulin dose without loss of glycemic con. LX4211 therapy reduces postprandial glu- 9. Skrivarhaug T. Diabetes Care 2015. Malik R. 1463–1468 fects of LX4211. hort of Old Order Amish.61(Suppl.91:733–794 Greater dose-ranging effects on A1C levels than 8. 2011 Pract 2008. Clin Ther 8-week open-label proof-of-concept trial. Biology of incretins: GLP..129:587–597 25. Perkins BA. Greer J. et al. Sultan S. et al. Clin Genet 2011. Di- trol in type 1 diabetic patients with and without food intake. Batterham RL. and Prediabetes in the United States. et al. Diabetes 296 hexose transporter-mediated D-glucose uptake. placebo-controlled 2015. Shibazaki T. Larkin ME. Drucker DJ. Hanssen KF. Szypowska A. Loo DD.e8 type 1 diabetic patients in Norway. Liu YJ. et al. double-blind. Physiol 26. Bossowski A. trial.79:86–91 GA. Krarup T. Powell DR. Bangstad H-J. Bellini NJ.304:E117–E130 nary glucose excretion.S. plus sitagliptin on postprandial active GLP-1 and glycemic control in type 1 diabetes: results of an Renal hemodynamic effect of sodium-glucose co. Overweight. Sutton D. peptide tyrosine tyrosine in a dose-timing study Sandvik L. Long-term mor. 1 and GIP. Backlund J-Y. July 2015 7. Sakuma S. SGLT1 and SGLT2. Ef. J Pharmacol Exp Ther Diabetes Technol Ther 2011. et al. Stene LC. Joner G. models. J Pharmacol Exp Ther 2012. Sagawa T. Freiman J. Atlanta.35:273–285.1188 Dual SGLT1 and SGLT2 Inhibition in T1D Diabetes Care Volume 38. a dual SGLT1/SGLT2 inhibitor. in patients with type 2 diabe- daily life and burden of hypoglycemia in insulin. Zambrowicz B. Clin Ther 2015. Tomae M. Powell DR. 2006. National Diabetes Fact Sheet: National Esti- with diabetes treated with insulin pump ther. Small CJ. Perkins BA. U. Rassi N. et al. Cleary P. 2013. on postprandial obese patients with type 1 diabetes mellitus. et al. Dobbins RL. insulin. Ogbaa I. Ogbaa I.e7 betes Care 2014. Clin Ther 2013. Eur J Pediatr 2011. proved glycemic control in mice lacking Sglt1 cose levels in patients with type 2 diabetes tes Control and Complications Trial/Epidemiology and Sglt2. Soleymanlou N. disturbance of LX4211. Diabet Med 2010. con Outcomes Res 2013.132:2131–2157 31. Diabe. Multiple sequwnce vatia. Holst JJ.19:963–967 . 21. Zambrowicz B. Łuczy nski ´ W. 11. proved glycemic control in patients with type tes on metformin monotherapy. et al. 32. Calculating the mean ampli- factors by sex in DCCT/EDIC. glucose. Gay J. Circulation 2014. Zambrowicz B. tions in SLC5A1 gene are associated with mates and General Information on Diabetes apy. and Endocr Pract 2013.27: peptide YY levels by reducing sodium/glucose tude of glycemic excursion from continuous glu- 451–458 cotransporter 1 (SGLT1)-mediated absorption cose monitoring data: an automated algorithm. Baggio LL.13:296–302 Olszewska B.37:1480–1483 abetes mellitus. Wajchenberg BL. Sodium-glucose cotransporter 2 inhibition and 14. 10. Clin Pharmacol Ther 2012. et al.13:123–130 18. Diago-Cabezudo JI. Strumph P. Wang H. Feitosa AC. Endocr Y. Kielgast U.49:298–305 Rev 2011. Car- 12. SGLT-2 inhibitors and their poten. Nature 2002. et al. Gut hormone PYY(3-36) physiologically inhibits GIP secretion in rats and humans (Abstract). Betti RTB. Lir- 15. Cherney DZI. 17. a dual sodium-dependent glucose aglutide as additional treatment to insulin in Exploring the potential of the SGLT2 inhibitor cotransporters 1 and 2 inhibitor. et al.35:1162– tality in a nationwide cohort of childhood-onset 16. Ishikawa-Takemura ices.34: 24. et al.92:158–169 27.