Sitagliptin: A Novel Dipeptidyl Peptidase-4 Inhibitor, Improves Glycemic Control in Patients with Type 2 Diabetes Dr Karthik Anantharaman MSD Pharmaceuticals Pvt Ltd (India)

Old Concept of T2DM Insulin Resistance Insulin Deficiency (Beta Cell Dysfunction) Hyperglycemia

Patients with T2DM Have Already Lost Substantial  -Cell Function at Diagnosis *Diet and exercise. N= 376. Adapted from UKPDS 16. Diabetes . 1995;44:1249–1258. Permission required. Diagnosis (%B)

Decrease in Glucose-Stimulated Insulin Secretion in T2DM (Beta Cell glucose sensitivity) Reprinted from Ferrannini E et al. J Clin Endocrinol Metab . 2005;90:493–500. 1000 800 600 400 200 0 5 10 15 20 25 Insulin secretion rate (pmoL·min -1 ·m -2 ) Plasma glucose (mmol/L) Obese NGT tertiles Lean NGT IGT T2DM quartiles

Pancreatic Islet Dysfunction leads to Hyperglycemia in T2DM ↑ Glucose Ohneda A, et al . J Clin Endocrinol Metab. 1978;46:504–510 Gomis R, et al. Diabetes Res Clin Pract . 1989;6:191–198. Fewer  -Cells  -Cells Hypertrophy Insufficient Insulin Excessive Glucagon – + ↓ Glucose uptake ↑ HGO +

Beta Cell Volume in Autopsy Studies A. Butler et al. Diabetes, 2003. Vol 52, 102-110.

Higher α-Cell : β-Cell Ratio in the Islet of Patients with T2DM *P <.05 vs. control groups 1 and 2 Control group 1= free of pancreatic disease; Control group 2= benign or malignant pancreatic tumor or changes in nutritional status Adapted from Yoon KH, et al . J Clin Endocrinol Metab 2003;88: 2300–2308 . * 0.81 0.2 0.3 α -Cell: β -Cell Ratio 0 0.2 0.4 0.6 0.8 1.0 1.2 Control 1 Control 2 DM

In T2DM, β-Cell Mass in Pancreatic Islets is Significantly Reduced 35%  -cells 65% β -cells 52%  -cells 48% β -cells P <0.01 Adapted from Deng S, et al . Diabetes 2004; 53:624–632. T2DM Control

Insulin and Glucagon Dynamics in T2DM -60 0 60 120 180 240 360 330 300 270 240 110 80 120 90 60 30 0 Glucose (mg %) Insulin ( µ U/mL) Glucagon (pg/mL) Meal Time (min) Delayed/depressed insulin response Nonsuppressed glucagon Normal subjects, n=11; Type 2 diabetes, n=12. Adapted from M ü ller WA et al. N Engl J Med . 1970;283:109–115. 140 130 120 110 100 90 Type 2 diabetes Normal subjects

Old Concept – Newer Insights Incretin Defect Insulin Resistance Insulin Deficiency (Beta Cell Dysfunction) Increased HGO Non-suppressed Glucagon (Alpha Cell Dysfunction) Hyperglycemia

Glycemic Profile of Treated Type 2 patients in India The Diabcare-Asia 1998 Study – Outcomes on Control and Complications in Type 1 and Type 2 Diabetic Patients. Nityanant et al. Current Medical Research and Opinion 2002; 18 (5): 317-327

Glycemic Profile of Treated Type 2 patients in India The Diabcare-Asia 1998 Study – Outcomes on Control and Complications in Type 1 and Type 2 Diabetic Patients. Nityanant et al. Current Medical Research and Opinion 2002; 18 (5): 317-327

No Single Class of Oral Antihyperglycemic Monotherapy Targets All Key Pathophysiologies Major Pathophysiologies 1. Glyset [package insert]. New York, NY: Pfizer Inc; 2004. 2. Precose [package insert]. West Haven, Conn: Bayer; 2004. 3. Prandin [package insert]. Princeton, NJ: Novo Nordisk; 2006. 4. Diabeta [package insert]. Bridgewater, NJ: Sanofi-Aventis; 2007. 5. Glucotrol [package insert]. New York, NY: Pfizer Inc; 2006. 6. Actos [package insert]. Lincolnshire, Ill: Takeda Pharmaceuticals; 2004. 7. Avandia [package insert]. Research Triangle Park, NC: GlaxoSmithKline; 2005. 8. Glucophage [package insert]. Princeton, NJ: Bristol-Myers Squibb; 2004.           Excess hepatic glucose output Meglitinides 3 Insulin resistance Insulin deficiency DPP-4 Inhibitors Metformin 8 TZDs 6,7 SUs 4,5 Alpha-Glucosidase Inhibitors 1,2 Intestinal glucose absorption

Challenges of not being able to treat patient to goal Medicines

Old Concept – Newer Insights Incretin Defect Insulin Resistance Insulin Deficiency (Beta Cell Dysfunction) Increased HGO Non-suppressed Glucagon (Alpha Cell Dysfunction) Hyperglycemia

Incretins, DPP-4 inhibition, and glucose homeostasis

GLP-1 Actions Are Glucose Dependent in Patients With Type 2 Diabetes Placebo GLP-1 Time (min) *p<.05 Insulin Glucagon Fasting glucose 250 150 5 250 200 100 50 40 30 20 10 0 mU/L 20 15 10 0 60 120 180 240 15.0 12.5 10.0 7.5 5.0 200 150 100 50 Infusion mmol/L mg/dL pmol/L pmol/L Effect declines as glucose reaches normal n=10. Adapted from Nauck NA et al. Diabetologia . 1993;36:741–744. * * * * * * * * * * * * * * * * * * *

Summary of Trials: GLP-1 and GIP Levels and Actions in Type 2 Diabetes *When corrected for gender and BMI Adapted from Toft-Nielsen M-B et al J Clin Endocrinol Metab 2001;86:3717–3723; Nauck MA et al J Clin Invest 1993;91:301–307.  (p=0.047 vs. NGT) Intact* GIP Intact  (p<0.05 vs. NGT) GLP-1 Incretin actions Incretin levels Patients with type 2 diabetes mellitus

Meal-Induced GLP-1 Secretion is Impaired in Patients with Type 2 Diabetes *p<0.05 between the T2DM and NGT group. T2DM=type 2 diabetes mellitus, NGT=normal glucose tolerant, IGT=impaired glucose tolerant. Reprinted from Toft-Nielsen M-B et al. Clin Endocrinol Metab . 2001:86;3717–3723. 5 15 15 20 0 60 120 180 240 (min) 0 GLP-1 (pmol/l) * * * * * * * NGT IGT T2DM meal

Plasma Concentrations of Glucagon and GIP in Patients With Type 2 Diabetes and Normal Subjects T2DM=type 2 diabetes, n=54; NGT=normal glucose tolerance, n=33 Reprinted from Toft-Nielsen M-B et al. J Clin Endocrinol Metab . 2001;86:3717–3723. * NGT T2DM NGT * * p<0.001 T2DM meal P-GIP (pmol/L) P-Glucagon (pmol/L)

Decreased Postprandial Levels of the Incretin Hormone GLP-1 in Patients With Type 2 Diabetes * P <0.05, Type 2 diabetes vs NGT. Reprinted with permission from Toft-Nielsen MB et al. J Clin Endocrinol Metab . 2001;86:3717–3723. Copyright © 2001, The Endocrine Society. 17 * * * * * * * Meal Started Meal Finished (10–15)

DPP9 DPP8 FAP DPP-4 DPP6 PEP QPP/DPPII APP prolidase DPP-4 Gene Family Other Proline Specific Peptidases Function unknown unknown unknown unknown unknown unknown unknown GLP-1 / GIP cleavage unknown NH 2 -Xaa ~ Pro-COOH --Xaa-Pro ~ Yaa-- NH 2 -Xaa-Pro ~ Yaa-- NH 2 -Xaa ~ Pro-Yaa---- catalytically inactive NH 2 -Xaa - Pro ~ Yaa-- Specificity DPP-4 Is a Member of a Family of Proline Specific Peptidases

Anatomical Relationship Between GLP-1+ L Cells and DPP-4+ Endothelium Cleft Hole Active site Probable entrance to active site Possible exit of cleaved dipeptide Hole

Sitagliptin Is Potent and Highly Selective (>2500x) for the DPP-4 Enzyme Herman et al. ADA . 2004. >100,000 APP >100,000 PEP >100,000 FAP >100,000 DPP-2, DPP-7 >100,000 DPP-9 48,000 DPP-8 18 DPP-4 IC 50 (nM) Enzyme

Selective DPP-4 Inhibitors Are Not Associated With Preclinical Toxicities Observed With Non-Selective Inhibitors 1. Leiting B et al. Abstract 6-OR. 64 th ADA;2004. 2. Lankas GK et al. Diabetes. 2005;54:2988–2994. – + + Decreased Proliferation Study of T-Cell Proliferation 1 2-Week Rat Toxicity Study 2 – + + Bloody diarrhea Acute Dog Toxicity Study 2 – + + Mortality – + + Enlarged spleen – + + Anemia – + + Thrombocytopenia – + + Alopecia Sitagliptin – highly selective DPP-4 inhibitor Selective DPP-8/9 inhibitor Nonselective inhibitor (DPP-8/9 and DPP-4)

Sitagliptin AUC 0-inf vs. creatinine clearance: AUC increases with decreasing creatinine clearance AUC GMR increase < 2-fold when CrCl > 50 mL/min Dose adjustments < 30 mL/min – ¼ dose 30 – 50 mL/min – ½ dose > 50 mL/min – full dose

Single-Dose OGTT Study One Dose of Sitagliptin Inhibited Plasma DPP-4 Activity Hours post-dose ~80% ~50% Trough DPP-4 inhibition Inhibition of plasma DPP-4 activity from baseline (%) 0 1 2 4 8 12 16 20 24 – 10 0 40 50 60 80 100 90 70 30 20 10 6 10 14 18 22 26 OGTT Sitagliptin 25 mg (n=56) Sitagliptin 200 mg (n=56) Placebo (n=56)

% Plasma Inhibition of DPP-4 Activity With Sitagliptin 100 mg in Healthy Adults 16 8 Percent Inhibition From Baseline Hours postdose 100 90 80 70 60 50 40 30 20 10 0 – 10 – 20 0 1 2 4 6 12 24 36 48 Protocol 001. Herman GA et al. Clin Pharmacol Ther . 2005;78:675–688. Sitagliptin 100 mg (N=6) Placebo (N=2)

A Single Dose of Sitagliptin Increased Active GLP-1 and GIP Over 24 Hours OGTT 24 hrs (n=19) Herman et al. Diabetes . PN005, 2005. Active GLP-1 0 5 10 15 20 25 30 35 40 0 2 4 6 24 26 28 Hours Postdose GLP-1 (pg/mL) OGTT 2 hrs (n=55) Crossover study in patients with T2DM Placebo Sitagliptin 25 mg Sitagliptin 200 mg 2-fold increase in active GLP-1 p< 0.001 vs placebo Active GIP 0 10 20 30 40 50 60 70 80 90 0 2 4 6 24 26 28 Hours Postdose GIP (pg/mL) OGTT 24 hrs (n=19) OGTT 2 hrs (n=55) 2-fold increase in active GIP p< 0.001 vs placebo

A Single Dose of Sitagliptin Increased Insulin, Decreased Glucagon, and Reduced Glycemic Excursion After a Glucose Load 0 10 20 30 40 0 1 2 3 4 mcIU/mL 50 55 60 65 70 75 0 1 2 3 4 Time (hours) pg/mL Glucose load Drug Dose 22% ~12% Insulin Glucagon Crossover Study in Patients with T2DM p<0.05 for both dose comparisons to placebo for AUC p<0.05 for both dose comparisons to placebo for AUC Placebo Sitagliptin 25 mg Sitagliptin 200 mg Glucose load Drug Dose 120 160 200 240 280 320 0 1 2 3 4 5 6 Time (hours) Glucose ~26% p<0.001 for both dose comparisons to placebo for AUC

Phase III Clinical Studies of Sitagliptin ● M onotherapy use (P021, P023, A201, P040) ● Combination use with Metformin, a PPAR  agent or SU (P019, P020, P035 and P036) ● Active Sulph comparator trial, added to metformin (P024)

Sitagliptin Provides Significant and Progressively Greater Reductions in A1C with Progressively Higher Baseline A1C Baseline A 1c (%) Mean (%) Reduction in A 1c (%) Inclusion Criteria: 7%–10% Reduction in A 1c (%) <8% 8–9% > 9% 7.37 8.40 9.48 <8% 8–9% > 9% 7.39 8.36 9.58 Reductions are placebo-subtracted Adapted from Raz et al. Diabetologia. 2006;49:2564–2571 ; Aschner et al. Diabetes Care. 2006;29:2632–2637. N=96 N=130 N=70 N=62 N=27 N=37

Sitagliptin Improves the  -Cell Response to Glucose Monotherapy Studies 200 400 600 800 1000 1200 1400 160 180 200 220 240 260 Glucose concentration (mg/dL) Insulin secretion (pmol/min) Pooled monotherapy studies – subset of patients with frequently sampled MTT Model-based assessment of β -cell function Φ s = static component, describes relationship between glucose concentration and insulin secretion Baseline End-Treatment Baseline End-Treatment Sitagliptin 100 mg q.d Placebo

Sitagliptin Improved Markers of Beta-Cell Function 24-Week Monotherapy Study Proinsulin/insulin ratio Aschner P et al. PN021; Abstract presented at: American Diabetes Association; June 10, 2006; Washington, DC. p< 0.001* *P value for change from baseilne compared to placebo Hatched = Baseline Solid = Week 24 ∆ from baseline vs pbo = 0.078 (95% CI -0.114, -0.023) Placebo Sitagliptin 100 mg Ratio (pmol/L / pmol/L) HOMA- β p< 0.001* ∆ from baseline vs pbo = 13.2 (95% CI 3.9, 21.9) Placebo Sitagliptin 100 mg

Indian Clinical Trial PN040 PN040 has been accepted for publication in Diabetes Research and Clinical Practice Exact publication time has not been given as yet Estimate: Feb or March edition

PN040, Comparable Baseline Characteristics BMI = body mass index. 66.6 66.8 Mean weight, kg 24.9 8.75 1.9 25.1 8.74 2.1 Mean BMI, kg/m 2 Mean A1c, % Duration of Diabetes 63 (35.4) 127 (36.1) Indian 33 (18.5) 62 (17.6) Korean 82 (46.1) 163 (46.3) Chinese Race/Ethnicity, n (%) 72 (40.4) 152 (43.2) Female, n (%) 50.9 50.9 Mean age, y Placebo n = 178 Sitagliptin 100 mg n = 352

Placebo Subtracted Change from Baseline in HbA1c Per Country (-1.92, -0.83) -1.38 Korea (-0.92, -0.46) -0.69 China (-1.73, -0.99) -1.36 India 95% Confidence limits Placebo Subtracted % A1c change Country

Sitagliptin Reduces FPG Levels Significantly From Baseline (APT Population) Values represent mean ± SE. 0 6 12 18 – 30 – 20 – 10 0 10 Week LSM Change From Baseline, mg/dL  31.0 p<0.001 Sitagliptin 100 mg Placebo

Four-Point Meal Tolerance Test at Baseline and Week 18 (APT Population) 120 170 220 270 Sitagliptin 100 mg Placebo Minutes After Initiation of Meal Challenge Mean Plasma Glucose, mg/dL Baseline Week 18 0 30 60 120 0 30 60 120

Incidence of Adverse Events AE = adverse event. 1 (0.6) 2 (0.6) Discontinued due to drug-related AE 2 (1.1) 5 (1.4) Discontinued due to AE 1 (0.6) 1 (0.3) Serious drug-related AE 2 (1.1) 6 (1.7) Serious AE 3 (1.7) 10 (2.8) Drug-related AE 27 (15.2) 82 (23.3) One or more AE Placebo n = 178 Sitagliptin 100 mg n = 352 Event, n (%)

Incidence of Laboratory Adverse Events LAE = laboratory adverse event. 0 0 Discontinued due to drug-related LAE 1 (0.6) 1 (0.3) Discontinued due to LAE 0 0 Serious drug-related LAE 0 0 Serious LAE 3 (1.8) 9 (2.6) Drug-related LAE 12 (7.0) 22 (6.5) One or more LAE Placebo n = 178 Sitagliptin 100 mg n = 352 Tolerability, n (%)

Protocol 021 –Completers Phase B results A1c -  from Baseline Phase A Phase B

Phase III Clinical Studies of Sitagliptin ● M onotherapy use (P021, P023, A201, P040) ● Combination use with Metformin, a PPAR  agent or SU (P019, P020, P035 and P036) ● Active Sulph comparator trial, added to metformin (P024)

Sitagliptin Once Daily Significantly Lowers A1C When Added On to Metformin or Pioglitazone  in A1C vs Pbo* = –0.65% (p<0.001)  in A1C vs Pbo* = –0.70% (p<0.001) *Placebo Subtracted Difference in LS Means. Charbonnel et al. Diabetes Care. 2006;29:2638–2643 ; Rosenstock et al. Clin Ther. 2006;28:1556–1568. Placebo (n=224) Sitagliptin 100 mg (n=453) Placebo (n=174) Sitagliptin 100 mg (n=163)

PN020: Extended Treatment With Sitagliptin and Metformin Maintained Lower A1C Levels to 104 Weeks (Completers) Continuation Phase 24-Week Phase 0 6 12 18 24 30 38 46 54 62 70 78 91 104 Week Values represent mean ± SE.

Sitagliptin Added to Ongoing Metformin or Pioglitazone Therapy in Patients With T2DM: Change in Body Weight Over Time LS Mean Change from Baseline in Body Weight (kg) 0.0 -0.4 -0.6 -0.8 -0.2 0 12 24 Study Week -1.0 Charbonnel et al. Diabetes Care. 2006;29:2638–2643 ; Rosenstock et al. Clin Ther. 2006;28:1556–1568. Placebo + Met (n=169) Sita 100 mg qd + Met (n=399) 0.0 0.5 1.0 1.5 2.0 -0.5 -1.0 0 6 12 18 24 Weeks Placebo + pioglitazone (n=174) Sita 100 mg qd + pioglitazone (n=163)

Sitagliptin Once Daily Significantly Increases Proportion of Patients Achieving Goal in Mono- or Combination Therapy Sitagliptin Placebo Monotherapy Study Add-On to Metformin Study Add-On to TZD Study Percentage Percentage Percentage P <0.001 P <0.001 P <0.001 17% 41% 18% 47% 23% 45% Goal A1C < 7% Aschner et al. Diabetes Care. 2006;29:2632–2637 . Charbonnel et al. Diabetes Care. 2006;29:2638–2643 ; Rosenstock et al. Clin Ther. 2006;28:1556–1568.

Study Design R Discontinue other AHA Visit 2 Run-in start Visit 10 Week 30 Visit 8 Week 18 Visit 6 Week 6 Visit 7 Week 12 Visit 9 Week 24 30 weeks Visit 1 Screening 1 week Visit 4 Week –2 Visit 3 Up to 12 weeks Visit 5 Day 1 2 weeks Screen patient according to inclusion and exclusion criteria Metformin titration/stable-dose period Placebo Primary end point Glipizide rescue HbA 1c 8%–11% AHA=antihyperglycemic agent; R=randomization. Raz I et al. Curr Med Res Opin. 2008;24:537–550. Single-blind placebo run-in Sitagliptin 100 mg/day Placebo Titrate metformin Continue stable dose of metformin at ≥1500 mg/day

Adding Sitagliptin 100 mg Once Daily to Metformin Reduced HbA 1c LSM=least squares mean; SE=standard error. a Sitagliptin=100 mg/day; b Metformin≥1500 mg/day; c Change from baseline at week 18 was the primary end point. Raz I et al. Curr Med Res Opin. 2008;24:537–550. Week 9.0 8.0 8.5 Mean ± SE Change in HbA 1c , % Between-group difference in LSM was –1.0% at 18 and 30 weeks; P <0.001 Sitagliptin a + metformin b (n=95) Placebo + metformin b (n=92) 6 0 12 18 c 24 30

Greater HbA 1c Reductions in Patients With Higher Baseline HbA 1c at 18 Weeks SE=standard error. a Sitagliptin=100 mg/day; b Metformin≥1500 mg/day. Raz I et al. Curr Med Res Opin. 2008;24:537–550. Sitagliptin a + metformin b Placebo + metformin b Mean ± SE Change in HbA 1c From Baseline, % – 2.0 – 1.5 – 1.0 – 0.5 0 0.5 34 9.4 41 9.4 9%–10% 45 8.4 35 8.4 <9% Baseline HbA 1c Subgroup 13 10.5  10% 19 10.5 n = Baseline mean, %

Placebo Controlled Add-on to Glimperide or Glimepiride/Metformin Study – Design and Patients 035 Placebo Phase B Sitagliptin 100 mg qd Screening Period Single-blind Placebo Stratum 1 Glim (≥ 4 mg/day) alone (~50%, n=212) Stratum 2 Glim + MF ≥1500 mg/d ) (~50%, n=229) Week 24 R A N D O M I Z A T I O N Week 80 Week 0 T2DM, Baseline A1c = 8.34 Age 18-78 yrs Continue/start regimen of glimepiride ± metformin Week -2 eligible if A1c 7.5-10.5% Double-blind Sitagliptin 100 mg qd Pio 30 mg qd

Sitagliptin Improved A1C When Added to Glim *Difference in LS Mean change from baseline Hermansen et al, Diabetes Obesity Metabolism 2007 Δ -0.6 %;p<0.001*

Sitagliptin Improved A1C When Added to Glim + MF 035 Δ -0.9%; p<0.001* *Difference in LS Mean change from baseline Hermansen et al, Diabetes Obesity Metabolism 2007

Sitagliptin Increased Rates of Hypoglycemia in Combination with Sitagliptin ± Metformin 035 Treatment Group N 222 219 4 (1.8) 0 Requiring Non-Medical Assistance and Not Exhibiting Marked Severity ‡ 0 0 Requiring Medical Assistance or Exhibiting Marked Severity ‡ Patients With at Least One Episode † n (%) Total Number of Episodes† Sitagliptin + Glim ± MF Overall n 55 20 9 0 0 0 Placebo + Glim ± MF Placebo + Glim ± MF Sitagliptin + Glim ± MF Overall n (%) 27 (12.2) 4 (1.8)

Sitagliptin + Metformin Factorial Study Design N = 1091 Randomized Mean baseline A1C = 8.8% Screening Period Single-blind Placebo Double-blind Treatment Period Diet/exercise Run-in Period Eligible if A1C 7.5 to 11% If on an OHA, D/C’ed Week- 2 Day 1 Sitagliptin 50/Met 1000 BID Placebo Sitagliptin 100 mg qd Metformin 500 BID Metformin 1000 BID Sitagliptin 50/Met 500 BID Week 24 Duration up to 12 weeks based on prior therapy Open Label Cohort Sitagliptin 50/Met 1000 BID R A N D O M I Z A T I O N Goldstein et al, Diabetes Care: 30; 1979 – 1987, 2007

A1C Results at 24 Weeks Mean A1C = 8.8% Sitagliptin 50 mg + metformin 1,000 mg bid Metformin 1,000 mg bid Sitagliptin 100 mg qd Sitagliptin 50 mg + metformin 500 mg bid Metformin 500 mg bid LSM A1C Change From Baseline, % – 3.5 – 3.0 – 2.5 – 2.0 – 1.5 – 1.0 – 0.5 0.0 0.5 n=178 n=177 n=183 n=178 n=175 – 0.8 a – 1.0 a – 1.3 a – 1.6 a – 2.1 a Open label n=117 – 2.9 b All patients Treated Population a LSM placebo adjusted change b LSM change from baseline without adjustment for placebo. bid=twice a day; qd=once a day. 24-Week Placebo-Adjusted Results Mean A1C = 11.2% Sitagliptin With Metformin Coadministration Initial Therapy Study Goldstein et al. Diabetes Care 2007; 30: 1979-1987

Initial Combination Therapy With Sitagliptin Plus Metformin Study: A1C Results From a Subset of Patients not on Antihyperglycemic Therapy at Study Entry LSM Change From Baseline, % Study 036 – 1.2 n=87 – 2.0 – 1.8 – 1.6 – 1.4 – 1.2 – 1.0 – 0.8 – 0.6 – 0.4 – 0.2 0 Sitagliptin 50 mg + metformin 1,000 mg bid Metformin 1,000 mg bid Sitagliptin 50 mg + metformin 500 mg bid Metformin 500 mg bid Sitagliptin 100 mg qd Placebo LSM=least squares mean change. – 1.1 n=88 – 1.1 n=90 – 1.6 n=100 – 1.9 n=86 – 0.2 n=83

Greater Reductions in HbA 1c Associated With Higher Baseline HbA 1c Without Regard to Therapy at Week 54 (APT) Williams-Herman D et al. Poster presentation at ADA 67th Annual Scientific Sessions in Chicago, Illinois, USA, 22–26 June 2007. Late Breaker (04-LB). Sitagliptin With Metformin Coadministration Initial Therapy Study All patients Treated Population

FPG and PPG Results at 24 Weeks 24-Week Placebo-Adjusted Results All Patients Treated Population FPG=fasting plasma glucose; PPG=postprandial glucose. a LSM adjusted for baseline value. b Difference from placebo. Sitagliptin 50 mg + metformin 1,000 mg bid Metformin 1,000 mg bid Sitagliptin 50 mg + metformin 500 mg bid Metformin 500 mg bid LSM PPG Change, mg/dL b 2-hour PPG Mean baseline level: 283–293 mg/dL – 54 a – 117 a – 125 – 100 – 75 – 50 – 25 0 LSM FPG Change, mg/dL b FPG Mean baseline level: 197–205 mg/dL – 33 a – 70 a – 75 – 50 – 25 0 – 53 a n=183 – 35 a n=179 n=141 – 93 a n=147 – 78 a n=138 n=152 – 23 a n=180 n=178 n=179 – 52 a Sitagliptin 100 mg qd n=136 Sitagliptin With Metformin Coadministration Initial Therapy Study Goldstein et al. Diabetes Care 2007; 30: 1979-1987

Co-administration of Sitagliptin and Metformin in Healthy Adults Increased Active GLP-1 Greater Than Either Agent Alone * Values represent geometric mean±SE. Placebo Metformin 1000 mg Sitagliptin 100 mg Co-administration of sitagliptin 100 mg + metformin 1000 mg Mean AUC ratio* Sita + Met: 4.12 Mean AUC ratios* Sita: 1.95 Met: 1.76 – 2 0 10 20 30 40 50 – 1 0 1 2 3 4 Active GLP-1 Concentrations, pM Meal Morning Dose Day 2 Time (hours pre/post meal) N=16 healthy subjects AUC=area under the curve Migoya EM et al. 67th ADA 2007. Oral Presentation OR-0286. Data available on request from Merck & Co., Inc. Please specify 20752937(1)-JMT. Metformin + Sitagliptin: Effect on Incretin Axis

Metformin + Sitagliptin : Pharmacokinetics of Co-Administration Herman G A Et al. Current Medical Research & Opinion 2006; 22 (10): 1939-1947 Mean (± standard error) metformin plasma concentrations following administration of metformin (1000 mg twice daily) with or without sitagliptin (50 mg twice daily) for 7 days in 13 patients with type 2 diabetes The mean Metformin plasma concentration–time profiles were nearly identical with or without Sitagliptin co-administration

Herman G A Et al. Current Medical Research & Opinion 2006; 22 (10): 1939-1947 Mean (± standard error) sitagliptin plasma concentrations following administration of sitagliptin (50 mg twice daily) with or without metformin (1000 mg twice daily) for 7 days in 13 patients with type 2 diabetes The mean Sitagliptin plasma concentration–time profiles were nearly identical with or without Metformin co-administration Metformin + Sitagliptin : Pharmacokinetics of Co-Administration

Recent efficacy findings from the blinded extension phase for total of 104 weeks Williams-Herman D et al. Poster presentation at ADA 68th Annual Scientific Sessions in USA, 22–26 June 2008. Sitagliptin With Metformin Coadministration Initial Therapy Study Results are presented for patients initially randomized to active therapy The initial combination treatment with Sitagliptin & Metformin was generally well tolerated over 2 years

Sitagliptin and Metformin–Initial Combination Therapy Extension Phase 0 6 12 18 24 30 38 46 54 62 70 78 91 104 6 6.5 7 7.5 8 8.5 9 Sita = sitagliptin; Met = metformin Time (weeks) 24-Week Phase Continuation Phase Extension Phase HbA 1c (LS mean change %) Sita 100 mg q.d. (n=22) Met 500 mg b.i.d. (n=26) Met 1000 mg b.i.d. (n=53) Sita 50 mg b.i.d. + Met 500 mg b.i.d. (n=64) Sita 50 mg b.i.d. + Met 1000 mg b.i.d. (n=77)

Effect of Initial Combination Therapy with Sitagliptin and Metformin on FPG Through Week 104 0 3 6 12 18 24 30 38 46 54 62 70 78 91 104 24-Week (Phase A) Continuation Phase (Phase B) Week FPG (LS Mean mg/dL) Completers Population in the Extension Phase Extension Phase Sita = sitagliptin; Met = metformin

Sitagliptin + Metformin: ß Cell function modelling Williams-Herman, Poster # 543, ADA 2008

Sitagliptin Alone and in Combination with Metformin improved both the Proinsulin to Insulin Ratio and HOMA-  HOMA-  Mean Baseline (pmol/L/pmol/L): .455 ± 0.035 *= p<0.01; †= p<0.001 Proinsulin/Insulin Ratio Sita 50 mg b.i.d + Met 1000 mg b.i.d . Sita 50 mg b.i.d + Met 500 mg b.i.d. Met 1000 mg b.i.d. Met 500 mg b.i.d. Sita 100 mg q.d. Placebo * * Mean Baseline 41.8 ± 3.7 *= p<0.001 Proinsulin/Insulin Ratio -0.25 -0.20 -0.15 -0.10 -0.05 0.00 0.05 * * * * †

Summary of Clinical Assessment of Hypoglycemia Extension Phase Results (Weeks 54 – 104) 39 14 (14) 98 1 1 (2.4) 42 Placebo/Met 1000 mg b.i.d. Including Data After Initiation of Glycemic Rescue Therapy ‡ Excluding Data After Initiation of Glycemic Rescue Therapy ‡ Treatment Group Total Number of Episodes † Patients with at Least One Episode † N Total Number of Episodes † Patients with at Least One Episode † N 122 134 121 107 103 107 100 88 65 52 7 5 (4.1) 5 5 (4.7) Sita 50 mg b.i.d. + Met 1000 mg b.i.d. 13 7 (6.5) 1 1 (1.5) Met 500 mg b.i.d. 1 1 (1.0) 0 0 (0.0) Sita 100 mg q.d. n (%) n (%) Met 1000 mg b.i.d. 2 (2.3) 3 6 (5.0) 11 Sita 50 mg b.i.d. + Met 500 mg b.i.d. 2 (2.0) 2 8 (6.0) 14 † Any given episode may belong to multiple categories; ‡ Patients meeting glycemic rescue criteria were treated with open-lable glyburide b.i.d. = twice daily; Met = Metformin; q.d. = once daily; Sita = Sitagliptin

Summary of Clinical Adverse Experiences (AEs) Through 54 Weeks (Phase A and B Combined, cont.) Sita 50 mg + MF 1000 mg b.i.d. N = 182 Sita 50 mg + MF 500 mg b.i.d. N = 190 Sita 100 mg q.d. N = 179 Metformin 1000 mg b.i.d. N = 182 Metformin 500 mg b.i.d N = 182 Number (%) of patients: 48 (26) 36 (19) 56 (31) 26 (14) 18 (10) All Gastrointestinal AEs 5 (3) 4 (2) 2 (1) 2 (1) 2 (1) Hypoglycemia Special AEs of Clinical Interest

Gastrointestinal AEs Through 54 Weeks % 27.7

Change in Body Weight From Baseline at Week 54 (LS mean change ± SE) Body Change From Baseline At Week 54 (kg) – 2.0 – 1.5 – 1.0 – 0.5 0.0 0.5 1.0 Sit 50 mg BID + met 1000 mg BID Sit 50 mg BID + met 500 mg BID Met 1000 mg BID Met 500 mg BID Sit 100 mg QD n=100 n=116 n=132 n=143 n=153 *Change from baseline P < 0.05. * * * *

Proportion of Patients with A1C Goal <7% at Endpoint (Week 54 Analysis) Sita 50 mg BID + Met 1000 mg BID Sita 50 mg BID + Met 500 mg BID Met 1000 mg BID Met 500 mg BID Sita 100 mg QD 58 77 101 106 124 106 117 134 147 153 n = Percent of patients

Sitagliptin Once Daily Shows Similar Glycemic Efficacy to Glipizide When Added to Metformin (52 Weeks) Mean Change in HbA 1c Mean change from baseline (for both groups)*: - 0.67% 6.0 6.2 6.4 6.6 6.8 7.0 7.2 7.4 7.6 7.8 8.0 8.2 8.4 0 12 24 38 52 Time (weeks) *per-protocol analysis; -0.51% and -0.56% for sitagliptin and glipizide in LOCF analysis Nauck et al, Diabetes Obesity Metabolism 9: 194 – 205, 2007 Sitagliptin 100 mg qd + Metformin (n= 382 ) Glipizide + Metformin (n=411)

Progressively Greater Reductions in A1C as Baseline A1C Rises Baseline A1C Category Study inclusion criteria 6.5-10% Change from baseline in A1C (%) Sitagliptin 100 mg q.d. Glipizide N=112 N=167 N=82 N=21 N=117 N=179 N=82 N=33 Per Protocol Population

Sitagliptin Once Daily Shows Better Safety and Tolerability Profile Compared to Glipizide (52 Weeks) Glipizide (n=584) Sitagliptin 100 mg (n=588) p<0.001 Nauck et al, Diabetes Obesity Metabolism 9: 194 – 205, 2007  between groups = –2.5 kg (p<0.001) Hypoglycemia 32% 4.9% 0 10 20 30 40 50 Week 52 Incidence (%) Sitagliptin 100 mg qd (n= 382 ) Glipizide (n=411)

Mean Change from Baseline in Waist Circumference (cm) Over Time All-Patients-as-Treated Population -2 -1 0 1 0 12 24 38 52 Week Mean Change from Baseline (cm) ±SE Sitagliptin 100 mg Glipizide

Mean Change from Baseline in ALT (IU/L) Over Time Week Mean Change from Baseline (IU/L) ±SE -2 -1 0 1 0 12 24 38 52 -3 6 18 30 46 3 Sitagliptin 100 mg Glipizide

Effects of Sitagliptin and Glipizide on Body Weight and Hypoglycemia *(95% CI) LS Means Change in Body Weight (kg) in Between Treatment Groups CSR LS Mean Change in Body Weight (kg) Over Time (APaT population) Overall Number of Episodes of Hypoglycemia: Week 0 Through Week 104 Sitagliptin 100 mg/d (n = 253) Glipizide (n = 261) LS Mean Change From Baseline (kg) 0 12 24 38 52 78 104 – 2 – 1 0 1 2 Week ∆ =−2.3 (−3.0, −1.6)*

HbA1c (%) Change from Baseline CSR Per Protocol Population Week 104 Difference in LS Means HbA1c -0.03; 95% CI -0.13, 0.07 HbA1c (%) Change from Baseline (LS Mean ±SE) (248) (256)

Substantial Proportion of Patients Achieved Goal on Sitagliptin Once-Daily As Add-on to Metformin % A1C < 7% after 52 Weeks Per Protocol Population

JANUVIA TM works consistently well in various patient subtypes

Demographic/Anthropometric Subgroups (and MS) Placebo-subtracted A1C LS mean change from baseline (%) Sex Age Median BMI Metabolic yrs kg/m 2 Syndrome F M <65 ≥65 ≤ 30.8 >30.8 - + p-values for treatment by ALL subgroup interactions are >0.05

Glycemic endpoints analyzed by baseline body mass index (BMI) In Patients with Type 2 Diabetes, Sitagliptin Effectively Lowers A1C Regardless of Patient Age, Gender, or Body Mass Index – pooled analysis of data from 4 phase III placebo controlled studies of Sitagliptin 100 mg monotherapy, involving 1691 patients. Poster by Williams-Herman et al. at ADA 2008. Pooled Results from all Phase III studies of JANUVIA ® 100mg monotherapy Placebo adjusted LS mean change from baseline HbA 1c, %, (95% CI) Placebo adjusted LS mean change from baseline FPG mg/dl, (95% CI) Placebo adjusted LS mean change from baseline 2-h PPG mg/d) (95% CI) Baseline values (mean ± SD) of HbA1c, FPG, and 2-h PPG in subgroups defined by baseline BMI subgroups

Pooled Results from all Phase III studies of JANUVIA ® 100mg monotherapy Glycemic endpoints analyzed by duration of type 2 diabetes Influence of Measures of Beta Cell Function on Efficacy of Sitagliptin in Patients with Type 2 Diabetes – pooled analysis of data from 4 phase III placebo controlled studies of Sitagliptin 100 mg monotherapy, involving 1691 patients. Poster by Harvey L. Katzeff et al. at ADA 2008. Placebo adjusted LS mean change from baseline HbA 1c, %, (95% CI) Placebo adjusted LS mean change from baseline FPG mg/dl, (95% CI) Placebo adjusted LS mean change from baseline 2-h PPG mg/d) (95% CI) Baseline values (mean ± SD) of HbA1c, FPG, and 2-h PPG in subgroups defined by baseline duration of type 2 diabetes

Influence of Measures of Beta Cell Function on Efficacy of Sitagliptin in Patients with Type 2 Diabetes – pooled analysis of data from 4 phase III placebo controlled studies of Sitagliptin 100 mg monotherapy, involving 1691 patients. Poster by Harvey L. Katzeff et al. at ADA 2008. Glycemic endpoints analyzed by baseline HOMA β Pooled Results from all Phase III studies of JANUVIA ® 100mg monotherapy Baseline values (mean ± SD) of HbA1c, FPG, and 2-h PPG in subgroups defined by baseline BMI subgroups

Influence of Measures of Beta Cell Function on Efficacy of Sitagliptin in Patients with Type 2 Diabetes – pooled analysis of data from 4 phase III placebo controlled studies of Sitagliptin 100 mg monotherapy, involving 1691 patients. Poster by Harvey L. Katzeff et al. at ADA 2008. Glycemic endpoints analyzed by baseline Pro-insulin/Insulin ratio Pooled Results from all Phase III studies of JANUVIA ® 100mg monotherapy Baseline values (mean ± SD) of HbA1c, FPG, and 2-h PPG in subgroups defined by baseline BMI subgroups

Summary Measures of Clinical Adverse Events for Sitagliptin is Similar to Placebo Recommended dose in proposed label: 100 mg q.d. 0.1 0.6 0.8 1.9 0.0 0.1 3.2 10.0 55.5 % Placebo (N=778) 0.0 0.1 Discontinued due to drug-related SAE 0.7 1.3 Discontinued due to SAE 0.0 0.6 Discontinued due to drug-related AE 0.9 2.6 Discontinued due to AE 0.0 0.0 Deaths 0.0 0.3 Drug-related SAEs 3.3 3.2 Serious AEs 9.4 9.5 Drug-related AEs 54.2 55.0 One or more AEs % % % of Patients with Sitagliptin 200 mg (N=456) Sitagliptin 100 mg (N=1082) Pooled Phase III Population

Only Small Differences in Incidence of AEs: Pooled Phase III Population AEs with at least 3% incidence and Numerically Higher in Sitagliptin than Placebo Group Recommended dose in proposed label: 100 mg q.d. Difference vs Pbo (95% CI) 0.1 (-2.3, 2.4) 0 1.2 (-0.7, 3.0) 0.7 (-0.9, 2.2) 0.3 (-1.1, 1.6) 0 3.0 2.3 Diarrhea 3.6 3.6 Headache 6.8 6.7 Upper Respiratory Tract Infection 1.7 1.8 3.3 % Placebo (N = 778) 1.7 Urinary Tract Infection 2.1 Arthralgia 4.5 Nasopharyngitis % Sitagliptin 100 mg (N = 1082)

Effect of Des-F-Sitagliptin on Beta-Cell Mass 1.1% Nondiabetic Control H&E insulin (I) glucagon (G) I/G Diabetic Control Diabetic Mice Treated with Des-F-sitagliptin 0.1% 0.4% Figure 3. HFD/STZ diabetic mice were treated with vehicle or des-fluoro-sitagliptin at indicated dosages for 11 weeks. Whole pancreas from each group was cryopreserved and consecutive sections were stained with H&E, anti-insulin antibody (green), or anti-glucagon antibody (red). Shown are representative islets from each group with single staining and the overlay of the insulin and glucagon staining (I/G).

GLP-1 Preserved Morphology of Human Islet Cells In Vitro Day 1 GLP-1–treated cells Control Day 3 Day 5 Islets treated with GLP-1 in culture were able to maintain their integrity for a longer period of time. Adapted from Farilla L et al. Endocrinology . 2003;144:5149–5158.

Thank You

Back up slides

Islet Cell Dysfunction: Beta-Cell Dysfunction in Type 2 Diabetes

The Beta Cell Micrograph: Lelio Orci, Geneva 10 µm ~ 10,000 granules

The Secretory Granule Micrograph: Lelio Orci, Geneva 100 nm 200,000 molecules of insulin

First-Phase Insulin Response is lost in T2DM Normal Type 2 Diabetes n=9 normal; n=9 type 2 diabetes. Adapted from Pfeifer MA et al. Am J Med . 1981;70:579–588. 0 20 40 60 80 100 120 – 30 0 30 60 90 120 Time (min) 0 20 40 60 80 100 120 – 30 0 30 60 90 120 Time (min) Plasma insulin (µU/mL) Plasma insulin (µU/mL)

Decrease in Glucose-Stimulated Insulin Secretion in T2DM (Beta Cell glucose sensitivity) Reprinted from Ferrannini E et al. J Clin Endocrinol Metab . 2005;90:493–500. 1000 800 600 400 200 0 5 10 15 20 25 Insulin secretion rate (pmoL·min -1 ·m -2 ) Lean NGT Plasma glucose (mmol/L)

Decrease in Glucose-Stimulated Insulin Secretion in T2DM (Beta Cell glucose sensitivity) Reprinted from Ferrannini E et al. J Clin Endocrinol Metab . 2005;90:493–500. 1000 800 600 400 200 0 5 10 15 20 25 Insulin secretion rate (pmoL·min -1 ·m -2 ) Obese NGT tertiles Lean NGT Plasma glucose (mmol/L)

Decrease in Glucose-Stimulated Insulin Secretion in T2DM (Beta Cell glucose sensitivity) Reprinted from Ferrannini E et al. J Clin Endocrinol Metab . 2005;90:493–500. 1000 800 600 400 200 0 5 10 15 20 25 Insulin secretion rate (pmoL·min -1 ·m -2 ) Obese NGT tertiles Lean NGT IGT Plasma glucose (mmol/L)

Decrease in Glucose-Stimulated Insulin Secretion in T2DM (Beta Cell glucose sensitivity) Reprinted from Ferrannini E et al. J Clin Endocrinol Metab . 2005;90:493–500. 1000 800 600 400 200 0 5 10 15 20 25 Insulin secretion rate (pmoL·min -1 ·m -2 ) Obese NGT tertiles Lean NGT IGT T2DM quartiles Plasma glucose (mmol/L)

Decreased Insulin Content in Type 2 Diabetes T2DM=Type 2 diabetes. Data obtained from pancreatic islets isolated from 6 T2DM organ donors and 10 nondiabetic cadaveric organ donors. Marchetti P et al. J Clin Endocrinol Metab . 2004;89:5535–5541. p<0.05

 Cells Contain less Insulin Secretory Granules in T2DM *p<0.05 vs control; T2DM=type 2 diabetes. Data obtained from pancreatic islets isolated from 6 T2DM organ donors and 10 nondiabetic cadaveric organ donors. Marchetti P et al. J Clin Endocrinol Metab . 2004;89:5535–5541. * Secretory granules (No/ 70 µm 2 )

Increased Beta-Cell Apoptosis Occurs in T2DM * *p<0.05. Islet cell death was assessed by an ELISA method, which evaluates the cytoplasmic histone-associated DNA fragments. After incubation absorbance of samples was read spectrophotometrically. Data obtained from pancreatic islets isolated from 6 T2DM organ donors and 10 nondiabetic cadaveric organ donors. Adapted from Marchetti P et al. J Clin Endocrinol Metab . 2004;89:5535–5541.

Adapted from Rhodes CJ. Science . 2005;307:380–384. Fewer Pancreatic Islets in Type 2 Diabetes Normal Compensation More islets Larger islets More beta cells/islet Larger beta cells Nondiabetic Obesity Decompensation Fewer islets Fewer beta cells/islet Amyloidosis Type 2 diabetes

Islet Cell Dysfunction: Excess Glucagon Secretion From Alpha Cells in Type 2 Diabetes

Glucagon Levels Are Elevated in IGT and T2DM p<0.001 Fasting p<0.001 Postprandial NGT=normal glucose tolerance, n=33; IGT=impaired glucose tolerance, n=15; T2DM=type 2 diabetes mellitus, n=54. Toft-Nielsen M-B et al. J Clin Endocrinol Metab . 2001;86:3717–3723. Fasting plasma glucagon (pmol/L) Postprandial glucagon at 240 min (pmol/L)

Postprandial Glucagon is not suppressed in T2DM * 72 108 144 180 216 – 60 0 60 120 180 240 300 360 Time (min) Glucose (mg/dL) Nonsuppressed glucagon Suppressed glucagon *p<0.001; N=9 (7 men, 2 women). Reprinted from Shah P et al. J Clin Endocrinol Metab . 2000;85:4053–4059.

Increase in Alpha-Cell Area in Type 2 Diabetes * *p<0.05; n: immunoperoxidase-stained postmortem pancreatic tissue from T2D=15, control=10. Adapted from Clark A et al. Diabetes Research . 1988;9:151–159.

Dipeptidyl Peptidase-4 (DPP-4) Is Involved in the Inactivation of GLP-1 and GIP XA- t 1/2 ~1 min DPP-4 agonist, active inactive GLP-1 active GIP active

The DPP Protease Family DPP-9 DPP-8 FAP DPP-4 DPP-6 PEP QPP/DPP-II APP prolidase DPP Gene Family Other Proline- Specific Peptidases Function unknown unknown unknown unknown unknown unknown unknown GLP-1/GIP regulation unknown NH 2 -Xaa ~ Pro-COOH --Xaa - Pro ~ Yaa-- NH 2 -Xaa - Pro ~ Yaa-- NH 2 -Xaa ~ Pro-Yaa---- catalytically inactive NH 2 -Xaa - Pro ~ Yaa-- Specificity Šedo A et al. Biochim Biophys Acta. 2001;550:107 – 116.

Inhibition of DPP-4 Leads to Increased Levels of GLP-1 (in Anesthetized Pigs) *Val-pyrrolidide; 300 µmol/kg, a DPP-4 inhibitor. Plasma concentration curves for GLP-1, measured with NH 2 -terminal and COOH-terminal radioimmunoassays. Holst JJ et al. Diabetes . 1998;47:1663–1670. Adapted from Deacon CF et al. Diabetes. 1998;47:764–769. Permission required. Minutes GLP-1, pmol/L GLP-1 infusion DPP-4 inhibitor* GLP-1 infusion Glucose Glucose Intact GLP-1 (NH 2 -terminal) Total GLP-1 (COOH-terminal, intact + NH 2 -terminally degraded peptide) 500 400 300 200 100 0 0 20 40 60 80 100 120 140 160 180 200 220

Increased Active GLP-1 and Improved Glucose Tolerance in DPP-4 Knockout Mice * P <0.05 Conarello SL et al. Proc Natl Acad Sci USA. 2003;100:6825–6830. Marguet D et al. Proc Natl Acad Sci USA. 2000;97:6874–6879. Increased Active GLP-1 Improved Glucose Tolerance Wild type – /– 0 2 4 6 8 * GLP-1 (7-36), pmol/L 20 40 60 80 90 135 180 225 270 315 360 400 * * Minutes Blood Glucose, mg/dL 0 100 120 DPP-4 –/– Wild type

GLP-1 Secretion is Reduced in Obesity Reprinted from Verdich C et al. Int J Obes . 2001;25:1206 – 1214. meal 0 10 20 30 0 20 40 60 80 100 120 140 160 180 Time (min) Plasma GLP-1 (pmol/L) Lean (BMI 23 kg/m 2 ; n=12) Obese (BMI 38.7 kg/m 2 ; n=19) Reduced obese (BMI 33.0 kg/m 2 n=19)

Meal-Stimulated Incretin Hormone Concentrations Correlate Positively with Insulin Sensitivity in Non-Diabetic Men Rask et al. Diabetes Care (2001) GLP-1 GIP meal meal Lowest insulin sensitivity tertile (n=11) Middle insulin sensitivity tertile (n=11) Highest insulin sensitivity tertile (n=11) 0 30 60 90 120 180 Time (min) 80 60 40 20 0 Plasma GIP (pmol/l) # * 0 30 60 90 120 180 Time (min) 20 15 10 5 0 Plasma GLP-1 (pmol/l) *

Meal-Stimulated Incretin Secretion is Impaired in Patients with Type 2 Diabetes T2DM=type 2 diabetes; NGT=normal glucose tolerance Reprinted from Toft-Nielsen M-B et al. J Clin Endocrinol Metab . 2001;86:3717–3723. meal GIP (pmol/L) * * * * * * * 20 15 10 5 0 GLP-1 (pmol/l) NGT (n=33) T2DM (n=54) AUC, p<0.05 AUC, p<0.05

Native GLP-1 is Rapidly Degraded by DPP IV Plasma T ½ =1-2 minutes (i.v.) MCR = 5-10 l/min DPP IV (red) and GLP-1 (green) in human small intestine DPP IV=dipeptidyl peptidase IV Hansen L et al, Endocrinology 1999; 140:5356-5363 MCR= metabolic clearance rate. Vilsb ø ll T et al. J Clin Endocrinol Metab . 2003;88:220 – 224 . DPP-IV 7 36 9 -NH 2 His Ala Ala Ala Ala Glu Glu Gly Gly Gly Glu Thr Phe Thr Phe Ser Ser Ser Asp Val Tyr Leu Leu Val Gln Lys Lys Ile Trp Ala

Intact GLP-1 + metabolite 1.5 nmol/kg sc Only ~10% GLP-1 survives intact after sc injection ** ** ** ** ** ** * Intact GLP-1 Survival of sc GLP-1 in Type 2 Diabetes 0 30 60 90 120 150 180 210 240 Time (min) 0 100 200 300 400 GLP-1 (pmol/l) Deacon CF et al, Diabetes 1995; 44:1126-1131

Plasma Concentrations of Active GLP-1 are Decreased in Type 2 Diabetes Total GLP-1, type 2 diabetic patients Total GLP-1, control group Intact GLP-1, type 2 diabetic patients Intact GLP-1, control group Time (min) GLP-1 (pmol/L) 50 0 100 150 10 20 30 0 * * * *p<0.05 Adapted from Vilsb ø ll T et al. Diabetes . 2001;50:609 – 613.

Impaired Second-Phase Insulin Responses to Hyperglycaemic Clamp During IV GIP in T2DM All subjects were obese (BMI 29 kg/m 2 ); patients with type 2 diabetes (n=8); control subjects (n=6). IV=intravenous. Blood glucose was clamped at 15 mmol/l Adapted from Vilsbøll Tl et al. Diabetologia . 2002;45:1111 – 1119. 240 120 Glucose Obese Diabetic Patients Obese Control Subjects 0 1000 2000 3000 4000 5000 0 60 180 Time (min) Insulin (pmol/L) Glucose Low GIP High GIP Low GIP GLP-1

Insulin Responses (cont) High GIP, Patients Low GIP, Patients Glucose, Patients All patients were obese with type 2 diabetes (n=8). Insulin responses to the glucose clamp alone and with a low (4 pmol/kg/min) and high (16 pmol/kg/min) dose of GIP. Vilsbøll T et al. Diabetologia . 2002;45:1111 – 1119 . 0 200 400 600 -20 0 20 40 60 80 100 120 140 160 180 200 220 240 Time (min) Insulin (pmol/L)

Glucagon Responses to Hyperglycaemic Clamp During IV GLP-1 and GIP Reprinted from Vilsbøll T et al. Diabetologia . 2002;45:1111–1119. All subjects were obese (BMI 29 kg/m2); patients with type 2 diabetes (n=8); control subjects (n=6). 240 Glucose Obese Diabetic Patients Obese Control Subjects 0 5 10 15 0 60 120 180 Time (min) Glucagon (pmol/L) Glucose Low GIP (4 pmol/kg/min) High GIP (16 pmol/kg/min) Low GIP (4 pmol/kg/min) GLP-1 (1 pmol/kg/min)

Effect of GLP-1 on β -Cell Glucose Responsiveness in Type 2 Diabetes ISR=insulin secretion rate. Reprinted from Kjems LL et al. Diabetes . 2003;52:380–386 . Relationship between average glucose concentrations and ISR in Control subjects Patients with Type 2 diabetes Saline infusion GLP-1 (0.5 pmol/kg/min) GLP-1 (1.0 pmol/kg/min) GLP-1 (2.0 pmol/kg/min) ISR (pmol/kg/min) 4 9 15 20 4 9 15 20 4 9 15 20 Glucose (mmol/l) 4 9 15 20 0 20 30 10 0 20 30 10 0 20 30 10 0 20 30 10 Glucose (mmol/l) ISR (pmol/kg/min) 0 20 30 10 0 20 30 10 4 9 15 20 4 9 15 20 0 20 30 10 4 9 15 20 0 20 30 10 4 9 15 20

Effect of GLP-1 on β -Cell Responsiveness to Glucose Adapted from Kjems LL et al. Diabetes . 2003;52:380–386. Glucose (mmol/l) 4 9 15 20 12 0 4 8 Type 2 diabetic: GLP-1 (0.5 pmol/kg/min) + Controls: Saline infusion ISR (pmol/kg/min) 4 9 15 20 0 4 8 12 Type 2 diabetic: Saline infusion Type 2 diabetic: GLP-1 (0.5 pmol/kg/min) 4 9 15 20 0 4 8 12

β -Cell Responsiveness to Glucose ISR=insulin secretion rate. β -cell responsiveness to glucose expressed as the slope of the linear relation between ISR and glucose concentration. Reprinted from Kjems LL et al. Diabetes . 2003;52:380 – 386 . GLP-1 (pmol • kg -1 • min -1 ) ISR vs glucose (pmol • kg -1 • min -1 /mmol/L) 0 2 4 6 8 0 0.5 1 1.5 2 Patients with type 2 diabetes Control subjects

Insulinotropic Effects of GIP and GLP-1 in Diabetes of Different Etiology n=6 in each group. Modified from from Vilsbøll T et al. J Clin Endocrinol Metab . 2003;88:4897–4903. Chronic pancreatitis 0 5 10 15 20 Time (min) Plasma glucose (mmol/l) -20 -5 10 25 40 55 70 85 100 115 LADA -20 -5 10 25 40 55 70 85 100 115 Type 1 DM -20 -5 10 25 40 55 70 85 100 115 Lean, Type 2 DM 0 100 200 300 400 500 600 700 -20 -5 10 25 40 55 70 85 100 115 Plasma insulin (pmol/l) Chronic pancreatitis Lean, Type 2 DM LADA Type 1 DM GLP-1 (1 pmol/kg/min) GIP (4 pmol/kg/min) Glucose

How Important Is the Gut as an Endocrine Organ? Apelin, Amylin Bombesin Calcitonin Gene-Related Peptide Cholecystokinin Galanin Gastric Inhibitory Polypeptide Gastrin Gastrin-releasing Peptide Ghrelin Glucagon, Glicentin, GLP-1, GLP-2, Oxyntomodulin Motilin Neuropeptide Y Neurotensin, Neuromedins, Neurokinins Peptide YY Pancreatic Polypeptide Pituitary Adenylate Cyclase Activating Peptide Secretin Somatostatin Tachykinins Thyrotropin-releasing Hormone Vasoactive Intestinal Peptide Ahrén B. Curr Diab Rep. 2003;3:356–372.

Measurement of the Incretin Effect: OGTT and Matched IV Infusion 400 210 Minutes 0 100 200 300 – 30 0 30 60 90 120 150 180 Glucose (mg/dL) Insulin (pmol/L) Nauck MA et al. J Clin Endocrinol Metab. 1986;63:492–498. Copyright © 1986. The Endocrine Society. Glucose Insulin 200 Minutes 0 50 100 150 – 30 0 30 60 90 120 150 180 210 Oral IV

The Incretins Y A E G T F I S D Y S I A M D K I H Q Q D F V N W L L A Q K G K K N D W K H N Q T I GIP: Glucose-dependent Insulinotropic Peptide H A E G T F T S D V S S Y L E G Q A A K E F I A W L V K G R G GLP-1: Glucagon-like Peptide–1 Amino acids shown in blue are homologous with the structure of glucagon. Drucker DJ. Diabetes Care. 2003;26:2929–2940.

Synthesis and Secretion of GLP-1 and GIP L cell (ileum) Proglucagon GLP-1 [7–37] GLP-1 [7–36NH 2 ] K cell (jejunum) ProGIP GIP [1–42] Drucker DJ. Diabetes Care. 2003;26:2929–2940.

Incretin Secretion Is Proportional to the Amount of Food Ingested in Healthy Individuals GLP-1 GIP 0 40 80 120 160 – 30 0 30 60 90 120 150 180 210 Minutes Total GIP (pmol/L) 0 10 20 30 40 50 – 30 0 30 60 90 120 150 180 210 Minutes Total GLP-1 (pmol/L) kcal 260 kcal 520 kcal 260 kcal 520 Vilsbøll T et al. J Clin Endocrinol Metab. 2003;88:2706–2713. Copyright © 2003. The Endocrine Society.

Meal-Stimulated Levels of GLP-1 Are Decreased in Type 2 Diabetes 0 10 20 30 0 50 100 150 Minutes GLP-1 (pmol/L) * * * * P <0.05 for differences between patients with type 2 diabetes and healthy subjects. Vilsbøll T et al. Diabetes. 2001;50:609–613. Total GLP-1, Controls Total GLP-1, Patients Intact GLP-1, Controls Intact GLP-1, Patients

GLP-1 Stimulates Insulin Secretion in Patients With Type 2 Diabetes 0 2000 4000 6000 8000 Minutes C-peptide (pmol/L) GLP-1 GIP Saline Hyperglycemic Clamp Saline or GIP or GLP-1 – 15 –10 0 5 10 15 20 30 45 60 75 90 105 120 150 Adapted from Vilsbøll T et al. Diabetologia. 2002;45:1111–1119.

GLP-1 Infusion Normalized Blood Glucose in Patients With Diabetes 0 36 72 108 144 180 216 252 288 00:00 04:00 08:00 12:00 16:00 Diabetic-saline Diabetic-GLP-1 Nondiabetic Glucose (mg/dL) Time of day Breakfast Lunch Snack Rachman J et al. Diabetologia. 1997;40:205–211.

Effect of Exogenous GLP-1 on Gastric Emptying 500 400 300 200 100 0 360 270 180 90 0 Plasma glucose (mg/dL) * – 30 0 30 60 90 120 150 180 210 240 GLP-1 [7–36 amide] sc Liquid meal P <0.0001 * * * * * * Gastric volume (mL) * GLP-1 [7–36 amide] sc Liquid meal P <0.0001 * Placebo GLP-1 * * * – 30 0 30 60 90 120 150 180 210 240 Minutes 350 300 250 200 150 100 50 0 Insulin (pmol/L) * GLP-1 [7–36 amide] sc Liquid meal P =0.0002 – 30 0 30 60 90 120 150 180 210 240 * * Minutes P values indicate significance for the interaction of experiment (GLP-1 [7–36 amide] vs placebo) and time. Asterisks indicate differences at specific time points (t test, P<0.05). Nauck MA et al. Diabetologia. 1996;39:1546–1553. Copyright © 1996 Springer. Reprinted with permission.

Effect of Exogenous GLP-1 on Food Intake in Healthy Men * * † 0 200 400 600 800 0 0.375 0.75 1.5 GLP-1 (pmol/kg/min) Food intake (g) * P <0.05; † P <0.001 vs control. Gutzwiller JP et al. Gut. 1999;44:81–86. Copyright © 1999. Reprinted with permission.

Incretin Secretion and DPP-4–Mediated Inactivation Intestinal GLP-1 Release GLP-1 (9–36) Inactive (~80% of pool) GLP-1 (7–36) Active DPP-4 t 1/2 = 1 to 2 min Increased insulin secretion Intestinal GIP Release GIP (1–42) Active Effect on gastric emptying, food intake, and glucagon secretion Mixed Meal Drucker DJ. Diabetes Care. 2003;26:2929–2940. DPP-4i

The Concept of Incretin Degradation: Role of DPP-4 Plasma profile of total vs intact GLP-1 following exogenous GLP-1 administration in diabetic subjects 0 30 60 90 120 150 180 210 240 Minutes 0 100 200 300 400 GLP-1 (pmol/L) ** ** ** ** ** ** * Total GLP-1(7–36) + (9–36)amide Intact GLP-1(7–36)amide N = 8 **P<0.05; *P<0.01; paired t test. Deacon CF et al. Diabetes. 1995;44:1126–1131.

DPP-4 Inhibition Prevents N-Terminal Degradation of GLP-1 in Anesthetized Pigs 0 20 40 60 80 100 120 140 160 180 200 220 Minutes 0 100 200 300 400 500 GLP-1 (pmol/L) GLP-1 infusion Glucose GLP-1 infusion Glucose Val-pyr DPP-4 Inhibition Deacon CF et al. Diabetes. 1998;47:764–769.

Pharmacological or Genetic Inactivation of DPP-4 Potentiates Incretin Action and Glucose Clearance In Vivo in Knockout Mice – 30 0 30 60 120 180 0 90 180 270 Blood glucose (mg/dL) ‡ † +/+ – /– Lower glucose in DPP-4 –/– mice +/+ -/- 0 180 360 * Plasma glucose (mg/dL) +/+ -/- 0 50 100 150 * Plasma insulin (pM) +/+ -/- 0 1 2 3 * Plasma GLP-1 (pM) * P <0.05 † P <0.01 ‡ P <0.001 Increased levels of insulin and intact GLP-1 and GIP in DPP-4 –/– mice Minutes Marguet D et al. Proc Natl Acad Sci. 2000;97:6874–6879.

Potentiation of Insulinotropic Activity of Exogenous DPP-4 Substrates Following Administration of Val-Pyr in Mice Minutes 0 10 20 30 40 50 Insulin (pmol/L) 0 3000 6000 9000 12000 Time (minutes) 0 10 20 30 40 50 Glucose (mmol/L) 90 270 375 525 GLP-1 Minutes 0 10 20 30 40 50 Insulin (pmol/L) 0 3000 6000 9000 12000 Glucose (mmol/L) PACAP38 Minutes 0 10 20 30 40 50 Insulin (pmol/L) 0 3000 6000 Glucose (mmol/L) GIP Minutes 0 10 20 30 40 50 Insulin (pmol/L) 0 3000 6000 Glucose (mmol/L) GRP Ahrén B, Hughes TE. Endocrinology. 2005;146:2055–2059. Val-pyr + Peptide Peptide alone Time (minutes) 0 10 20 30 40 50 90 270 375 525 Time (minutes) 0 10 20 30 40 50 90 270 375 525 Time (minutes) 0 10 20 30 40 50 90 270 375 525

DPP-4 Inhibitors Acutely Lower Blood Glucose and Stimulate Insulin Secretion in Single Incretin Receptor –/– but Not in DIRKO Mice – 30 0.3 0.9 Insulin * – 30 0 30 60 90 120 0 90 180 270 360 ** * * Minutes 0.6 – 30 0 30 60 90 120 0 90 180 270 360 *** *** Minutes Blood Glucose (mg/dL) 0.3 Insulin 0.6 – 30 0 30 60 90 120 0 *** Minutes 0.3 0.9 Insulin 0.6 0 30 60 90 120 0 Minutes 0.3 Insulin 0.6 Wild-type GIPR –/– GLP-1R –/– DIRKO Blood Glucose (mg/dL) GLP-1 and GIP receptors are essential for DPP-4 inhibitor action ** *** *** * P <0.05; ** P <0.01; *** P <0.001 vehicle vs Val-pyr–treated mice Blood Glucose (mg/dL) Blood Glucose (mg/dL) Vehicle DPP-4 Inhibitors 90 180 270 360 90 180 270 360 Hansotia T et al. Diabetes. 2004;53:1326–1335.

Peripheral but Not Portal Glucose Infusion Increases Blood Glucose in Mice 36 73 108 144 0 Blood glucose (mg/dL) Saline Portal glucose infusion Femoral glucose infusion 0 20 60 100 140 180 Minutes – 400 – 200 0 200 400 600 800 AUC (mmol/L · min) S P F † * * † S= Saline P= Portal glucose F= Femoral glucose 180 *Statistically different from portal-vein mice; † Statistically different from saline-infused mice; P <0.05. Burcelin R et al. Diabetes. 2001;50:1720–1728.

GLP-1 and GIP Modulate Insulin and GLP-1 Modulates Glucagon to Decrease Blood Glucose Levels During Hyperglycemia Glucose output Glucose uptake Glucagon (alpha cells) Insulin (beta cells) Pancreas GLP-1=glucagon-like peptide-1; GIP=glucose-dependent insulinotropic polypeptide. Porte D Jr, Kahn SE. Clin Invest Med. 1995;18:247–254. Drucker DJ. Diabetes Care. 2003;26:2929–2940. Liver Decreased blood glucose GLP-1 GIP Muscle Adipose tissue

Question Will 4 weeks of near-normalisation of blood glucose normalise incretin hormone secretion and improve β -cell sensitivity in type 2 diabetes patients?

Meal-Induced GLP-1 and GIP Secretion in Type 2 Diabetic Patients Before and After 4 Weeks of Normoglycaemia Blood glucose was normalised using insulin HbA1c = 7.9 ±0.4% before treatment and 6.7±0.3% at week 4; n=9 Højberg PV et al. Diabetes 2006; 55:(suppl 1): A85 -50 0 50 100 150 200 250 300 Time (min) 50 40 30 20 10 0 -50 0 50 100 150 200 250 300 Time (min) 0 40 20 60 100 80 GLP-1 (pmol/l) GIP (pmol/l) Matched control Patients, baseline Patients, after 4 wk normoglycaemia Patients, after 24 hr normoglycaemia

 -Cell Responsiveness to GLP-1 Improves After 4 Weeks of Normoglycaemia in Patients with Type 2 Diabetes Units are mol.min - 1 .kg -1 /(mmol/l), evaluated from the slope of the linear regression between insulin secretion rate and concomittant plasma glucose during graded glucose infusions and infusions of saline or GLP-1 (1 pmol/kg/min) Højberg PV et al. Diabetes 2005; 54:(suppl 1): A362 Unpublished data: β -cell responsiveness to GIP improves after near-normalisation of glycaemia in T2DM ( submitted ADA ) P<0.0001 P< 0.001 P< 0.0001 P-value Saline vs GLP-1 P =0.40 P <0.02 0.39  0.04 1.73  0.24 0.33  0.04 1.27  0.24 1.01  0.14 4.79  0.53 Saline GLP-1 P-value Diabetics, Before vs After Type 2 patients ” After” Type 2 patients ” Before” Control subjects

If the impaired incretin response contributes significantly to the defective insulin secretion in type 2 diabetes, will restoration of incretin action improve metabolism? Question

0 2 4 6 8 10 12 14 16 00:00 04:00 08:00 12:00 16:00 Snack Lunch Breakfast Diabetic - saline Non-diabetic Glucose (mmol/L) Time of day Rachman J et al., Diabetologia 1997;40:205-211 Proof of Hypothesis: Glucose Tolerance can be Restored by iv GLP-1 in T2DM Diabetic - GLP-1 (1.2 pmol/kg/min)

Continuous s.c. Infusion of GLP-1 Reduces Blood Glucose and Improves β -cell Function in Patients with Type 2 Diabetes Zander M et al, Lancet 2002;359:824-830 Week 0 Week 1 GLP-1 Week 6 GLP-1 Plasma glucose (mmol/l) 0 5 10 15 20 25 0 1 2 3 4 5 6 7 8 Hours post-injection 8-hour BG profiles Minutes 0 1000 2000 3000 4000 5000 6000 7000 10 30 50 70 90 arginine Hyperglycaemic clamp C-peptide (pmol/l)

Type 2 diabetic phenotype Actions of GLP-1 • ↑ insulin secretion and biosynthesis • Improves β -cell function • Impaired β -cell function ( glucose sensitivity, proinsulin/insulin ratio ) • Upregulates other genes essential for β -cell function ( eg. GLUT 2, glucokinase ) • Reduced β -cell mass • ↑ β -cell proliferation/differentiation animal studies • ↓ β -cell apoptosis + in vitro • Glucagon hypersecretion • ↓ glucagon secretion • Accelerated gastric emptying • ↓ gastric emptying • Overeating, obesity • ↑ satiety, ↓ appetite  weight loss • Macrovascular complications • Beneficial cardiovascular effects Actions which may be secondary to improved metabolic control • Insulin resistance • Improvements in insulin sensitivity GLP-1: Therapeutic Potential in Type 2 Diabetes

Data from elderly patients

Elderly: Effect of Sitagliptin and Placebo on HbA 1c Through Week 24 HbA 1c (%) LS Mean Full Analysis Set Time (weeks) BARDS data Table 4 7.30 8.00

Elderly Monotherapy : HbA 1c Change from Baseline by Age at Baseline Age <75 years ≥75 years Full Analysis Set at Week 24 HbA 1c (%) Placebo-subtracted Change from Baseline (LS MeanI) Data on file. Safety & Tolerability consistent with overall profile. No Hypoglycemia. (71) (30) ( Sitagliptin

Elderly: HbA 1c Change from Baseline by HbA1c at Baseline HbA 1c <8% ≥8% and <9% ≥9% Full Analysis Set at Week 24 BARDS data Table 30 HbA 1c (%) Placebo-subtracted Change from Baseline (LS MeanI) (68) (20) (13) ( Sitagliptin

Effect of JANUVIA TM on fasting blood glucose

By effectively inhibiting DPP-4 enzyme for a full 24 hours, JANUVIA TM provides substantial HbA1c lowering through effect both on FPG & PPG. Levels of incretins increase substantially following a meal though incretins are also released in the body throughout the day. Incretins affect both insulin & glucagon, which help regulate hepatic glucose production (primary mechanism for maintaining glucose homeostasis in fasting state).

Steady-state 24-hour glucose measures over 24 hours after 4 weeks of treatment with JANUVIA TM (50 mg twice daily) plus metformin (≥1500 mg daily) vs. placebo plus metformin JANUVIA TM 50 mg BD provides no additional glycemic efficacy compared to 100 mg OD. Period 1 (4 weeks) in a double-blind, randomized, placebo-controlled, 8-week crossover study to assess the effect on glycemic control of adding JANUVIA TM to ongoing metformin therapy in patients inadequately controlled on ≥1500 mg/day of metformin. Dose 1 of JANUVIA TM administered at 7:30, dose 2 at 18:30. Brazg R et al. Diabetes Obesity & Metabolism 2007; 9: 186–193.

A Multicenter, Randomized, Double-Blind, Placebo-Controlled Study to Evaluate the Safety and Efficacy of the Addition of Sitagliptin for Patients With Type 2 Diabetes Mellitus Who Have Inadequate Glycemic Control on Glimepiride, Alone or in Combination With Metformin Study 035 – Continuation Phase

Study Design Placebo Sitagliptin 100 mg QD Single-blind Placebo Week 24 Continue/start regimen of glimepiride ± metformin Single-blind eligible if A1C 7.5% to 10.5% 24-Week Phase R Stratum 1: Glimepiride (≥4 mg/d) Stratum 2: Glimepiride + Metformin (≥1500 mg/d) Continuation Phase Week 54 Patients not requiring rescue medication in 24-week phase could continue through 54 weeks. Active Treatment* *=Pioglitazone 30 mg QD Week 0 Screening Period Patients with treated or untreated T2DM, ages 18 to 78 years

Sitagliptin Improved A1C When Added to Glim ± MF 035 *Difference in LS Mean change from baseline Δ -0.7 %;p<0.001*

Sitagliptin Improved A1C When Added to Glim 035 *Difference in LS Mean change from baseline Δ -0.6 %;p<0.001*

Sitagliptin Improved A1C When Added to Glim + MF 035 *Difference in LS Mean change from baseline Δ -0.9%; p<0.001*

Disposition of Patients Screened: N = 1098 Randomized: N = 441 Glimepiride n = 106 Completed Week 24 n = 83 Glimepiride + Metformin n = 116 Completed Week 24 n = 87 Glimepiride n = 106 Completed Week 24 n = 102 Glimepiride+ Metformin n = 113 Completed Week 24 n = 92 Placebo n = 219 Sitagliptin n = 222 Entered Continuation Phase Sitagliptin + Glimepiride + Metformin n = 93 Entered Continuation Phase Pioglitazone + Glimepiride n = 58 Excluded: n = 657 Entered Continuation Phase Sitagliptin + Glimepiride n = 67 Entered Continuation Phase Pioglitazone + Glimepiride + Metformin n = 64

BMI = body mass index. Comparable Baseline Characteristics of Patients Entering Continuation Phase 10 (8.2) 13 (8.1) Other 82.2 85.6 Mean weight, kg 29.8 30.6 Mean BMI, kg/m 2 19 (15.6) 21 (13.1) Asian 20 (16.4) 28 (17.5) Hispanic 8 (6.6) 7 (4.4) Black 65 (53.3) 91 (56.9) Caucasian Race/Ethnicity, n (%) 58 (47.5) 70 (43.8) Female, n (%) 56.6 55.5 Mean age, y Placebo/pioglitazone n = 122 Sitagliptin 100 mg n = 160

FPG = fasting plasma glucose. Comparable Baseline Disease Characteristics of Patients Entering Continuation Phase 75 (61.5) 109 (68.1) Combination therapy 44 (36.1) 45 (28.1) Monotherapy 3 (2.5) 6 (3.8) None 9.4 8.0 Mean duration of diabetes, y Antihyperglycemic therapy, n (%) 11.3 14.0 Mean fasting insulin, µIU/mL 167.5 173.3 Mean FPG, mg/dL 18 (15.0) 28 (17.5) ≥ 9% 52 (43.3) 66 (41.3) ≥ 8% and <9% 50 (41.7) 66 (41.3) <8% Distribution of A1C, n (%) 8.2 8.2 Mean A1C, % Placebo/pioglitazone n = 122 Sitagliptin 100 mg n = 160

Extended Treatment With Sitagliptin + Glimepiride, With or Without Metformin, Maintained Lower A1C Levels to 54 Weeks ( Entire Cohort , APT ) Values represent mean ± SE. APT = all patients treated. 24-Week Phase Continuation Phase 0 6 12 18 24 30 38 46 54 Week Sitagliptin 100 mg (n = 158)

Extended Treatment With Sitagliptin + Glimepiride, With or Without Metformin, Maintained Lower A1C Levels to 54 Weeks ( Entire Cohort , Completers ) Values represent mean ± SE. 24-Week Phase Continuation Phase 0 6 12 18 24 30 38 46 54 Week Sitagliptin 100 mg (n = 93)

Mean FPG Through 54 Weeks (Entire Cohort, Completers) Values represent mean ± SE. 24-Week Phase Continuation Phase Sitagliptin 100 mg (n = 92)

FPG Approached Baseline With Extended Treatment With Sitagliptin + Glimepiride, With or Without Metformin, at 54 Weeks (Entire Cohort, APT) Values represent mean ± SE. 24-Week Phase Continuation Phase 0 6 12 18 24 30 38 46 54 Week 2 Sitagliptin 100 mg (n = 159)

Values represent mean ± SE. 24-Week Phase Continuation Phase 0 6 12 18 24 30 38 46 54 Week Extended Treatment With Sitagliptin + Glimepiride Maintained Lower A1C Levels to 54 Weeks (Stratum 1, APT) Sitagliptin 100 mg (n = 66)

Values represent mean ± SE. 24-Week Phase Continuation Phase 0 6 12 18 24 30 38 46 54 Week Extended Treatment With Sitagliptin + Glimepiride Maintained Lower A1C Levels to 54 Weeks (Stratum 1, Completers ) Sitagliptin 100 mg (n = 38)

Mean FPG Through 54 Weeks (Stratum 1, Completers) Values represent mean ± SE. 24-Week Phase Continuation Phase Sitagliptin 100 mg (n = 37)

FPG Approached Baseline With Extended Treatment With Sitagliptin + Glimepiride at 54 Weeks (Stratum 1, APT) Values represent mean ± SE. 24-Week Phase Continuation Phase 0 6 12 18 24 30 38 46 54 Week 2 Sitagliptin 100 mg (n = 67)

Extended Treatment With Sitagliptin + Glimepiride With Metformin Maintained Lower A1C Levels to 54 Weeks (Stratum 2, APT ) Values represent mean ± SE. 24-Week Phase Continuation Phase 0 6 12 18 24 30 38 46 54 Week Sitagliptin 100 mg (n = 92)

Extended Treatment With Sitagliptin