CLINICAL STUDIES

The efficacy of TRESIBA administered once-daily either at the same time each day or at any time each day in patients with type 1 diabetes and used in combination with a mealtime insulin was evaluated in three randomized, open-label, treat-to-target, active-controlled trials in adults and one randomized, open-label, treat-to-target, active-controlled trial in pediatric patients 1 year of age and older. The efficacy of TRESIBA administered once-daily either at the same time each day or at any time each day in adult patients with type 2 diabetes and used in combination with a mealtime insulin or in combination with common oral anti-diabetic agents was evaluated in six randomized, open-label, treat-to-target active-controlled trials.

Adult patients treated with TRESIBA achieved levels of glycemic control similar to those achieved with LANTUS (insulin glargine 100 units/mL) and LEVEMIR (insulin detemir) and achieved statistically significant improvements compared to sitagliptin.

Type 1 Diabetes – Adult

TRESIBA Administered at the Same Time Each Day in Combination with a Rapid-Acting Insulin Analog at Mealtimes in Adult Patients

Study A

The efficacy of TRESIBA was evaluated in a 52-week randomized, open-label, multicenter trial in 629 patients with type 1 diabetes mellitus (Study A). Patients were randomized to TRESIBA once-daily with the evening meal or insulin glargine U-100 once-daily according to the approved labeling. Insulin aspart was administered before each meal in both treatment arms.

The mean age of the trial population was 43 years and mean duration of diabetes was 19 years. 59% were male. 93% were White, 2% Black or African American. 5% were Hispanic. 9% of patients had eGFR<60 mL/min/1.73m ². The mean BMI was approximately 26.3 kg/m ².

At week 52, the difference in HbA _1creduction from baseline between TRESIBA and insulin glargine U-100 was -0.01% with a 95% confidence interval of [-0.14%; 0.11%] and met the pre-specified non-inferiority margin (0.4%). See Table 6, Study A.

Study B

The efficacy of TRESIBA was evaluated in a 26-week randomized, open-label, multicenter trial in 455 patients with type 1 diabetes mellitus (Study B). Patients were randomized to TRESIBA or insulin detemir once-daily in the evening. After 8 weeks, insulin detemir could be dosed twice-daily. 67% used insulin detemir once daily at end of trial. 33% used insulin detemir twice daily at end of trial. Insulin aspart was administered before each meal in both treatment arms.

The mean age of the trial population was 41 years and mean duration of diabetes was 14 years. 52% were male. 45% were White, 0.4% Black or African American. 4% were Hispanic. 4% of patients had eGFR<60 mL/min/1.73m ². The mean BMI was approximately 23.9 kg/m ².

At week 26, the difference in HbA _1creduction from baseline between TRESIBA and insulin detemir was -0.09% with a 95% confidence interval of [-0.23%; 0.05%] and met the pre-specified non-inferiority margin (0.4%). See Table 6, Study B.

Table 6: Results at Week 52 in a Trial Comparing TRESIBA to Insulin Glargine U-100 (Study A) and Week 26 in a Trial Comparing TRESIBA to Insulin Detemir (Study B) in Adult Patients with Type 1 Diabetes Mellitus Receiving Insulin Aspart at Mealtimes

Study A		Study B
TRESIBA + Insulin aspart	Insulin glargine U-100 + Insulin aspart	TRESIBA + Insulin aspart	Insulin detemir + Insulin aspart
N	472	157	302	153
HbA _1c(%)
Baseline	7.7	7.7	8.0	8.0
End of trial	7.3	7.3	7.3	7.3
Adjusted mean change from baseline ^•	-0.36	-0.34	-0.71	-0.61
Estimated treatment difference [95%CI] TRESIBA - basal insulin U-100	-0.01 [-0.14;0.11]		-0.09 [-0.23;0.05]
Proportion Achieving HbA _1c< 7% at Trial End	39.8%	42.7%	41.1%	37.3%
FPG (mg/dL)
Baseline	165	174	178	171
End of trial	141	149	131	161
Adjusted mean change from baseline	-27.6	-21.6	-43.3	-13.5
Daily basal insulin dose
Baseline mean	28 U	26 U	22 U	22 U
Mean dose at end of study	29 U ¹	31 U ¹	25 U ²	29 U ²
Daily bolus insulin dose
Baseline mean	29 U	29 U	28 U	31 U
Mean dose at end of study	32 U ¹	35 U ¹	36 U ²	41 U ²

¹At Week 52
²At Week 26
^•The change from baseline to end of treatment visit in HbA _1cwas analyzed using ANOVA with treatment, region, sex, and anti-diabetic treatment at screening as fixed effects, and age and baseline HbA _1cas covariates.
In Study A, there were 14.8% of subjects in the TRESIBA and 11.5% Insulin glargine arms for whom data was missing at the time of the HbA _1cmeasurement.
In Study B, there were 6.3% of subjects in the TRESIBA and 9.8% Insulin detemir arms for whom data was missing at the time of the HbA _1cmeasurement.

Study C: TRESIBA Administered at the Same Time Each Day or at Any Time Each Day in Combination with a Rapid-Acting Insulin Analog at Mealtimes in Adult Patients

The efficacy of TRESIBA was evaluated in a 26-week randomized, open-label, multicenter trial in 493 patients with type 1 diabetes mellitus. Patients were randomized to TRESIBA injected once-daily at the same time each day (with the main evening meal), to TRESIBA injected once daily at any time each day or to insulin glargine U-100 injected once-daily according to the approved labeling. The any time each day TRESIBA arm was designed to simulate a worst-case scenario injection schedule of alternating short and long, once daily, dosing intervals (i.e., alternating intervals of 8 to 40 hours between doses). TRESIBA in this arm was dosed in the morning on Monday, Wednesday, and Friday and in the evening on Tuesday, Thursday, Saturday, and Sunday. Insulin aspart was administered before each meal in all treatment arms.

The mean age of the trial population was 43.7 years and mean duration of diabetes was 19 years. 58% were male. 98% were White, 2% Black or African American. 3% were Hispanic. 7% of patients had eGFR<60 mL/min/1.73m ². The mean BMI was approximately 26.7 kg/m ².

At week 26, the difference in HbA _1creduction from baseline between TRESIBA administered at alternating times and insulin glargine U-100 was 0.17% with a 95% confidence interval of [0.04%; 0.30%] and met the pre-specified non-inferiority margin (0.4%). See Table 7.

Table 7: Results at Week 26 in a Trial Comparing TRESIBA Dosed Once Daily at the Same and at Alternating Times Each Day to Insulin Glargine U-100 in Adult Patients with Type 1 Diabetes Mellitus Receiving Insulin Aspart at Mealtimes

TRESIBA at the same time each day + Insulin aspart	TRESIBA at alternating times + Insulin aspart	Insulin glargine U-100 + Insulin aspart
N	165	164	164
HbA _1c(%)
Baseline	7.7	7.7	7.7
End of trial	7.3	7.3	7.1
Adjusted mean change from baseline ^•	-0.41	-0.40	-0.57
Estimated treatment difference [95%CI] TRESIBA alternating - Insulin glargine U-100	0.17 [0.04;0.30]
Proportion Achieving HbA _1c< 7% at Trial End	37.0%	37.2%	40.9%
FPG (mg/dL)
Baseline	179	173	175
End of trial	133	149	151
Adjusted mean change from baseline	-41.8	-24.7	-23.9
Daily basal insulin dose
Baseline mean	28 U	29 U	29 U
Mean dose at end of study	32 U	36 U	35 U
Daily bolus insulin dose
Baseline mean	29 U	33 U	32 U
Mean dose at end of study	27 U	30 U	35 U

^•The change from baseline to end of treatment visit in HbA _1cwas analyzed using ANOVA with treatment, region, sex, and anti-diabetic treatment at screening as fixed effects, and age and baseline HbA _1cas covariates.

In Study C, there were 15.8% and 15.9% of subjects in the TRESIBA (same time and alternating times respectively) and 7.9% Insulin glargine arms for whom data was missing at the time of the HbA _1cmeasurement.

Type 1 Diabetes – Pediatric Patients 1 Year of Age and Older

Study J: TRESIBA Administered at the Same Time Each Day in Combination with a Rapid-Acting Insulin Analog at Mealtimes in Pediatric Patients 1 Year of Age and Older

The efficacy of TRESIBA was evaluated in a 26-week, randomized, open label, multicenter trial in 350 patients with type 1 diabetes mellitus (Study J). Patients were randomized to TRESIBA once-daily or insulin detemir once or twice-daily. Subjects on a twice-daily insulin detemir regimen were dosed at breakfast and in the evening either with the main evening meal or at bedtime. Insulin aspart was administered before each main meal in both treatment arms. At end of trial, 36% used insulin detemir once daily and 64% used insulin detemir twice daily.

The mean age of the trial population was 10 years; 24% were ages 1-5 years; 39% were ages 6-11 years and 36% were ages 12-17 years. The mean duration of diabetes was 4 years. 55% were male. 75% were White, 3% Black or African American. 3% were Hispanic. The mean z-score for body weight was 0.31.

At week 26, the difference in HbA _1creduction from baseline between TRESIBA and insulin detemir was 0.15% with a 95% confidence interval of [-0.03%; 0.33%] and met the pre-specified non-inferiority margin (0.4%). See Table 8.

Table 8: Results at Week 26 in a Trial Comparing TRESIBA to Insulin Detemir in Pediatric Patients 1 Year of Age and Older with Type 1 Diabetes Mellitus Receiving Insulin Aspart at Mealtimes

TRESIBA+ Insulin aspart	Insulin detemir + Insulin aspart
N	174	176
HbA _1c(%)
Baseline	8.2	8.0
End of 26 weeks	8.0	7.7
Adjusted mean change from baseline after 26 weeks ^±	-0.19	-0.34
Estimated treatment difference [95%CI] TRESIBA v. Insulin detemir	0.15 [ -0.03; 0.33]
FPG (mg/dL)
Baseline	162	151
End of 26 weeks	150	160
Adjusted mean change from baseline after 26 weeks	52.0	59.6
Daily basal insulin dose
Baseline mean	15 U (0.37 U/kg)	16 U (0.41 U/kg)
Mean dose after 26 weeks	16 U (0.37 U/kg)	22 U (0.51 U/kg)
Daily bolus insulin dose
Baseline mean	20 U (0.50 U/kg)	20 U (0.52 U/kg)
Mean dose after 26 weeks	23 U (0.56 U/kg)	22 U (0.57 U/kg)

^±The change from baseline to end of treatment visit in HbA _1cwas analyzed using ANOVA with missing data imputed by multiple imputation carrying forward the baseline value and adding the error term, with treatment, region, sex, and age group as fixed factors, and baseline HbA _1cas covariate.
In Study J, there were 2.9% of subjects in TRESIBA and 6.3% Insulin detemir arms for whom data was missing at the 26-week HbA _1cmeasurement.

14.3 Type 2 Diabetes - Adult

Study D: TRESIBA Administered at the Same Time Each Day as an Add-on to Metformin with or without a DPP-4 Inhibitor in Insulin Naïve Adult Patients

The efficacy of TRESIBA was evaluated in a 52-week randomized, open-label, multicenter trial that enrolled 1030 insulin naïve patients with type 2 diabetes mellitus inadequately controlled on one or more oral antidiabetic agents (OADs). Patients were randomized to TRESIBA once-daily with the evening meal or insulin glargine U-100 once-daily according to the approved labeling. Metformin alone (83%) or in combination with a DPP-4 inhibitor (18%) was used as background therapy in both treatment arms.

The mean age of the trial population was 59 years and mean duration of diabetes was 9 years. 62% were male. 88% were White, 7% Black or African American. 17% were Hispanic. 10% of patients had eGFR<60 mL/min/1.73m ². The mean BMI was approximately 31.1 kg/m ².

At week 52, the difference in HbA _1creduction from baseline between TRESIBA and insulin glargine U-100 was 0.09% with a 95% confidence interval of [-0.04%; 0.22%] and met the pre-specified non-inferiority margin (0.4%); See Table 9.

Table 9: Results at Week 52 in a Trial Comparing TRESIBA to Insulin Glargine U-100 in Adult Patients with Type 2 Diabetes Mellitus on OAD(s)•

TRESIBA + OAD(s)•		Insulin glargine U-100 + OAD(s)•
N	773		257
HbA _1c(%)
Baseline	8.2		8.2
End of trial	7.1		7.0
Adjusted mean change from baseline ^••	-1.06		-1.15
Estimated treatment difference [95%CI] TRESIBA - Insulin glargine U-100	0.09 [-0.04;0.22]
Proportion Achieving HbA _1c< 7% at Trial End	51.7%	54.1%
FPG (mg/dL)
Baseline	174		174
End of trial	106		115
Adjusted mean change from baseline	-68.0		-60.2
Daily insulin dose
Baseline mean (starting dose)	10 U		10 U
Mean dose after 52 weeks	56 U		58 U

^•OAD: oral antidiabetic agent
^••The change from baseline to end of treatment visit in HbA _1cwas analyzed using ANOVA with treatment, region, sex, and anti-diabetic treatment at screening as fixed effects, and age and baseline HbA _1cas covariates.
In Study D, there were 20.6% of subjects in the TRESIBA and 22.2% Insulin glargine arms for whom data was missing at the time of the HbA _1cmeasurement.

Study E: TRESIBA U-200 Administered at the Same Time Each Day as an Add-on to Metformin with or without a DPP-4 Inhibitor in Insulin Naïve Adult Patients

The efficacy of TRESIBA U-200 was evaluated in a 26-week randomized, open-label, multicenter trial in 457 insulin naïve patients with type 2 diabetes mellitus inadequately controlled on one or more oral antidiabetic agents (OADs) at baseline. Patients were randomized to TRESIBA U-200 once-daily with the evening meal or insulin glargine U-100 once-daily according to the approved labeling. Both treatment arms were receiving metformin alone (84%) or in combination with a DPP-4 inhibitor (16%) as background therapy.

The mean age of the trial population was 58 years and mean duration of diabetes was 8 years. 53% were male. 78% were White, 14% Black or African American. 8% were Hispanic. 8% of patients had eGFR <60 mL/min/1.73m ². The mean BMI was approximately 32.4 kg/m ².

At week 26, the difference in HbA _1creduction from baseline between TRESIBA U-200 and insulin glargine U-100 was 0.04% with a 95% confidence interval of [-0.11%; 0.19%] and met the pre-specified non-inferiority margin (0.4%). See Table 10.

Table 10: Results at Week 26 in a Trial Comparing TRESIBA U-200 to Insulin Glargine U-100 in Adult Patients with Type 2 Diabetes Mellitus on OAD(s)•

TRESIBA U-200 + Met ± DPP-4		Insulin glargine U-100 + Met ± DPP-4
N	228		229
HbA _1c(%)
Baseline	8.3		8.2
End of trial	7.0		6.9
Adjusted mean change from baseline ^••	-1.18		-1.22
Estimated treatment difference [95%CI] TRESIBA - Insulin glargine U-100	0.04 [-0.11;0.19]
Proportion Achieving HbA _1c< 7% at Trial End	52.2%	55.9%
FPG (mg/dL)
Baseline	172		174
End of trial	106		113
Adjusted mean change from baseline	-71.1		-63.5
Daily insulin dose
Baseline mean	10 U		10 U
Mean dose after 26 weeks	59 U		62 U

^•OAD: oral antidiabetic agent
^••The change from baseline to end of treatment visit in HbA _1cwas analyzed using ANOVA with treatment, region, sex, and anti-diabetic treatment at screening as fixed effects, and age and baseline HbA _1cas covariates.
In Study E, there were 12.3% of subjects in the TRESIBA and 12.7% Insulin glargine arms for whom data was missing at the time of the HbA _1cmeasurement.

Study F: TRESIBA Administered at the Same Time Each Day in Insulin Naïve Adult Patients as an Add-on to One or More of the Following Oral Agents: Metformin, Sulfonylurea, Glinides or Alpha-Glucosidase Inhibitors

The efficacy of TRESIBA was evaluated in a 26-week randomized, open-label, multicenter trial in Asia in 435 insulin naïve patients with type 2 diabetes mellitus inadequately controlled on one or more oral antidiabetic agents (OADs) at baseline. Patients were randomized to TRESIBA once-daily in the evening or insulin glargine U-100 once-daily according to the approved labeling. Pre-trial oral antidiabetes agents were continued as background therapy except for DPP-4 inhibitors or thiazolidinediones in both treatment arms.

The mean age of the trial population was 59 years and mean duration of diabetes was 12 years. 54% were male. All patients were Asian. 11% of patients had eGFR<60 mL/min/1.73m ². The mean BMI was approximately 25.0 kg/m ².

At week 26, the difference in HbA _1creduction from baseline between TRESIBA and insulin glargine U-100 was 0.11% with a 95% confidence interval of [-0.03%; 0.24%] and met the pre-specified non-inferiority margin (0.4%). See Table 11.

Table 11: Results at Week 26 in a Trial Comparing TRESIBA to Insulin Glargine U-100 in Adult Patients with Type 2 Diabetes Mellitus on OAD(s)•

TRESIBA + OAD(s)•	Insulin glargine U-100 + OAD(s)•
N	289	146
HbA _1c(%)
Baseline	8.4	8.5
End of trial	7.2	7.1
Adjusted mean change from baseline ^•^•	-1.42	-1.52
Estimated treatment difference [95%CI] TRESIBA - Insulin glargine U-100	0.11 [-0.03 ; 0.24]
Proportion Achieving HbA _1c< 7% at Trial End	40.8%	48.6%
FPG (mg/dL)
Baseline	152	156
End of trial	100	102
Adjusted mean change from baseline	-54.6	-53.0
Daily insulin dose
Baseline mean (starting dose)	9 U	9 U
Mean dose after 26 weeks	19 U	24 U

^•OAD: oral antidiabetic agent
^••The change from baseline to end of treatment visit in HbA _1cwas analyzed using ANOVA with treatment, region, sex, and anti-diabetic treatment at screening as fixed effects, and age and baseline HbA _1cas covariates.
In Study F, there were 10% of subjects in the TRESIBA and 6.8% Insulin glargine arms for whom data was missing at the time of the HbA _1cmeasurement.

Study G : TRESIBA Administered at the Same Time Each Day or Any Time Each Day as an Add-on to One and up to Three of the Following Oral Agents: Metformin, Sulfonylurea or Glinides or Pioglitazone in Adult Patients

The efficacy of TRESIBA was evaluated in a 26-week randomized, open-label, multicenter trial in 687 patients with type 2 diabetes mellitus inadequately controlled on basal insulin alone, oral antidiabetic agents (OADs) alone or both basal insulin and OAD. Patients were randomized to TRESIBA injected once-daily at the same time each day (with the main evening meal), to TRESIBA injected once daily at any time each day or to insulin glargine U-100 injected once-daily according to the approved labeling. The any time each day TRESIBA arm was designed to simulate a worst-case scenario injection schedule of alternating short and long, once daily, dosing intervals (i.e., alternating intervals of 8 to 40 hours between doses). TRESIBA in this arm was dosed in the morning on Monday, Wednesday, and Friday and in the evening on Tuesday, Thursday, Saturday, and Sunday. Up to three of the following oral antidiabetes agents (metformin, sulfonylureas, glinides or thiazolidinediones) were administered as background therapy in both treatment arms.

The mean age of the trial population was 56 years and mean duration of diabetes was 11 years. 54% were male. 67% were White, 3% Black or African American. 11% were Hispanic. 6% of patients had eGFR<60 mL/min/1.73m ². The mean BMI was approximately 29.6 kg/m ².

At week 26, the difference in HbA _1creduction from baseline between TRESIBA at alternating times and insulin glargine U-100 was 0.04% with a 95% confidence interval of [-0.12%; 0.20%]. This comparison met the pre-specified non-inferiority margin (0.4%). See Table 12.

Table 12: Results at Week 26 in a Trial Comparing TRESIBA at Same and Alternating Times to Insulin Glargine U-100 in Adult Patients with Type 2 Diabetes Mellitus on OAD(s)•

TRESIBA at the same time each day ± OAD(s)•	TRESIBA at alternating times ± OAD(s)•	Insulin glargine U-100 ± OAD(s)•
N	228	229	230
HbA _1c(%)
Baseline	8.4	8.5	8.4
End of trial	7.3	7.2	7.1
Adjusted mean change from baseline ^••	-1.03	-1.17	-1.21
Estimated treatment difference [95%CI] TRESIBA alternating- Insulin glargine U-100	0.04 [-0.12;0.20]
Estimated treatment difference TRESIBA alternating – TRESIBA same	-0.13
Proportion Achieving HbA _1c< 7% at Trial End	40.8%	38.9%	43.9%
FPG (mg/dL)
Baseline	158	162	163
End of trial	105	105	112
Adjusted mean change from baseline	-54.2	-55.0	-47.5
Daily insulin dose
Baseline mean	21 U	19 U	19 U
Mean dose after 26 weeks	45 U	46 U	44 U

^•OAD: oral antidiabetic agent
^••The change from baseline to end of treatment visit in HbA _1cwas analyzed using ANOVA with treatment, region, sex, and anti-diabetic treatment at screening as fixed effects, and age and baseline HbA _1cas covariates.
In Study G, there were 11.4% subjects for TRESIBA (both same time and alternating times) and 11.7% Insulin glargine arms for whom data was missing at the time of the HbA _1cmeasurement.

Study H: TRESIBA Administered at the Same Time Each Day in Combination with a Rapid-Acting Insulin Analog at Mealtimes in Adult Patients

The efficacy of TRESIBA was evaluated in a 52-week randomized, open-label, multicenter trial in 992 patients with type 2 diabetes mellitus inadequately controlled on premix insulin, bolus insulin alone, basal insulin alone, oral antidiabetic agents (OADs) alone or any combination thereof. Patients were randomized to TRESIBA once-daily with the main evening meal or insulin glargine U-100 once-daily according to the approved labeling. Insulin aspart was administered before each meal in both treatment arms. Up to two of the following oral antidiabetes agents (metformin or pioglitazone) were used as background therapy in both treatment arms.

The mean age of the trial population was 59 years and mean duration of diabetes was 14 years. 54% were male. 83% were White, 10% Black or African American. 12% were Hispanic. 12% of patients had eGFR<60 mL/min/1.73m ². The mean BMI was approximately 32.2 kg/m ².

At week 52, the difference in HbA _1creduction from baseline between TRESIBA and insulin glargine U-100 was 0.08% with a 95% confidence interval of [-0.05%; 0.21%] and met the pre-specified non-inferiority margin (0.4%). See Table 13.

Table 13: Results at Week 52 in a Trial Comparing TRESIBA to Insulin Glargine U-100 in Adult Patients with Type 2 Diabetes Mellitus Receiving Insulin Aspart at Mealtimes and OADs•

TRESIBA + Insulin aspart ± OAD(s)•	Insulin glargine U-100 + Insulin aspart ± OAD(s)•
N	744	248
HbA _1c(%)
Baseline	8.3	8.4
End of trial	7.1	7.1
Adjusted mean change from baseline ^••	-1.10	-1.18
Estimated treatment difference [95%CI] TRESIBA - Insulin glargine U-100	0.08 [-0.05;0.21]
Proportion Achieving HbA _1c< 7% at Trial End	49.5%	50.0%
FPG (mg/dL)
Baseline	166	166
End of trial	122	127
Adjusted mean change from baseline	-40.6	-35.3
Daily basal insulin dose
Baseline mean	42 U	41 U
Mean dose after 52 weeks	74 U	67 U
Daily bolus insulin dose
Baseline mean	33 U	33 U
Mean dose after 52 weeks	70 U	73 U

^•OAD: oral antidiabetic agent
^••The change from baseline to end of treatment visit in HbA _1cwas analyzed using ANOVA with treatment, region, sex, and anti-diabetic treatment at screening as fixed effects, and age and baseline HbA _1cas covariates.
In Study H, there were 16.1% of subjects in the TRESIBA and 14.5% Insulin glargine arms for whom data was missing at the time of the HbA _1cmeasurement.

Study I: TRESIBA Administered at Any Time Each Day as an Add-on to One or Two of the Following Oral Agents: Metformin, Sulfonylurea, or Pioglitazone in Adult Patients

The efficacy of TRESIBA was evaluated in a 26-week randomized, open-label, multicenter trial in 447 patients with type 2 diabetes mellitus inadequately controlled on one or more oral antidiabetic agent (OADs) at baseline. Patients were randomized to TRESIBA once-daily at any time of day or sitagliptin once-daily according to the approved labeling. One or two of the following oral antidiabetes agents (metformin, sulfonylurea or pioglitazone) were also administered in both treatment arms.

The mean age of the trial population was 56 years and mean duration of diabetes was 8 years. 59% were male. 61% were White, 8% Black or African American. 21% were Hispanic. 6% of patients had eGFR<60 mL/min/1.73m ². The mean BMI was approximately 30.4 kg/m ².

At the end of 26 weeks, TRESIBA provided greater reduction in mean HbA _1ccompared to sitagliptin (p < 0.001). See Table 14.

Table 14: Results at Week 26 in a Trial Comparing TRESIBA to Sitagliptin in Adult Patients with Type 2 Diabetes Mellitus on OADs•

TRESIBA + OAD(s)•	Sitagliptin + OAD(s)•
N	225	222
HbA _1c(%)
Baseline	8.8	9.0
End of trial	7.2	7.7
Adjusted mean change from baseline ^••	-1.52	-1.09
Estimated treatment difference [95%CI] TRESIBA - Sitagliptin	-0.43 [-0.61;-0.24] ¹
Proportion Achieving HbA _1c< 7% at Trial End	40.9%	27.9%
FPG (mg/dL)
Baseline	170	179
End of trial	112	154
Adjusted mean change from baseline	-61.4	-22.3
Daily insulin dose
Baseline mean	10 U	N/A
Mean dose after 26 weeks	43 U	N/A

^•OAD: oral antidiabetic agent
^••The change from baseline to end of treatment visit in HbA _1cwas analyzed using ANOVA with treatment, region, sex, and anti-diabetic treatment at screening as fixed effects, and age and baseline HbA _1cas covariates.
In Study I, there were 20.9% of subjects in the TRESIBA and 22.5% Sitagliptin arms for whom data was missing at the time of the HbA _1cmeasurement.
¹p <0.001; 1-sided p-value evaluated at 2.5% level for superiority

Safety Outcomes Trial

DEVOTE (NCT01959529) Cardiovascular Outcomes Trial of TRESIBA Administered Once-Daily Between Dinner and Bedtime in Combination with Standard of Care in Subjects with Type 2 Diabetes and Atherosclerotic Cardiovascular Disease

DEVOTE was a multi-center, multi-national, randomized, double-blinded, active-controlled, treat-to-target, event-driven trial. 7,637 patients with inadequately controlled type 2 diabetes and atherosclerotic cardiovascular disease were randomized to either TRESIBA or insulin glargine U-100. Each was administered once-daily between dinner and bedtime in addition to standard of care for diabetes and cardiovascular disease for a median duration of 2 years.

Patients eligible to enter the trial were; 50 years of age or older and had established, stable, cardiovascular, cerebrovascular, peripheral artery disease, chronic kidney disease or NYHA class II and III heart failure (85% of the enrolled population) or were 60 years of age or older and had other specified risk factors for cardiovascular disease (15% of the enrolled population).

At baseline, demographic and disease characteristics were balanced between treatment groups. The mean age of the trial population was 65 years and the mean duration of diabetes was 16 years. The population was 63% male, 76% White 11% Black or African American, 10% Asian. 15% had Hispanic ethnicity. The mean HbA _1cwas 8.4% and the mean BMI was 33.6 kg/m ². The baseline mean estimated glomerular filtration rate (eGFR) was 68 mL/min/1.73m ². 41% of patients had eGFR 60-90 mL/min/1.73m ²; 35% of patients had eGFR 30 to 60 mL/min/1.73 m ²and 3% of patients had eGFR <30 mL/min/1.73 m ². Previous history of severe hypoglycemia was not captured in the trial.

At baseline, patients treated their diabetes with oral antidiabetic drugs (72%) and with an insulin regimen (84%). Types of insulins included long acting insulin (60%), intermediate acting insulin (14%) short acting insulin (37%) and premixed insulin (10%). 16% of patients were insulin naive. The most common background oral antidiabetic drugs used at baseline were metformin (60%), sulfonylureas (29%) and DPP-4 inhibitors (12%).

During the trial, investigators could modify anti-diabetic and cardiovascular medications to achieve local standard of care treatment targets for lipids and blood pressure.

Cardiovascular Outcomes - Patients with T2DM and Atherosclerotic CVD

The incidence of major cardiovascular events with TRESIBA was evaluated in DEVOTE. Subjects treated with TRESIBA had a similar incidence of major adverse cardiovascular events (MACE) when compared to those treated with insulin glargine U-100.

The primary endpoint in DEVOTE was time from randomization to the first occurrence of a 3-component major adverse cardiovascular event (MACE): cardiovascular death, non-fatal myocardial infarction, or non-fatal stroke. The trial was designed to exclude a pre-specified risk margin of 1.3 for the hazard ratio of MACE comparing TRESIBA to insulin glargine U-100. The primary outcome at end of trial was available for 98.2% of participants in each treatment group.

The time to first occurrence of MACE with TRESIBA as compared to insulin glargine U-100 was non-inferior (HR: 0.91; 95% CI [0.78;1.06]; see Figure 3). The results of the primary composite MACE endpoint and a summary of its individual components are shown in Table 15.

Table 15: Analysis of the Composite 3-point MACE and Individual Cardiovascular Endpoints in DEVOTE

TRESIBA		Insulin glargine U-100
N	3,818		3,819
Number of Patients (%)	Rate per 100 PYO•	Number of Patients (%)	Rate per 100 PYO•	Hazard Ratio (95% CI)
Composite of first event of CV death, non-fatal MI, or non-fatal stroke (3-Point MACE)	325 (8.5)	4.41	356 (9.3)	4.86	0.91 [0.78; 1.06]
CV death	136 (3.6)	1.85	142 (3.7)	1.94
Non-fatal MI	144 (3.8)	1.95	169 (4.4)	2.31
Non-fatal stroke	71 (1.9)	0.96	79 (2.1)	1.08

• PYO = patient-years of observation until first MACE, death, or trial discontinuation

Figure 3: Cumulative Event Probability for Time to First MACE in DEVOTE

Hypoglycemia Outcomes - Patients with T2DM and Atherosclerotic CVD

The pre-specified secondary endpoints of event and incidence rates of severe hypoglycemia were sequentially tested.

Severe hypoglycemia was defined as an episode requiring assistance of another person to actively administer carbohydrate, glucagon, or other resuscitative actions and during which plasma glucose concentration may not have been available, but where neurological recovery following the return of plasma glucose to normal was considered sufficient evidence that the event was induced by a low plasma glucose concentration.

The incidence of severe hypoglycemia was lower in the TRESIBA group as compared to the insulin glargine U-100 group (Table 16). Glycemic control between the two groups was similar at baseline and throughout the trial.

Table 16: Severe Hypoglycemic Episodes in Patients Treated with TRESIBA or Insulin Glargine U-100 in DEVOTE

TRESIBA	Insulin glargine U-100
N	3,818	3,819
Severe Hypoglycemia
Percent of patients with events	4.9%	6.6%
Estimated odds ratio [95%CI] TRESIBA/Insulin glargine U-100	0.73 [0.60; 0.89]•
Events per 100 Patient Years of Observation	3.70	6.25
Estimated rate ratio [95%CI] TRESIBA/Insulin glargine U-100	0.60 [0.48; 0.76]•

• Test for superiority evaluated at 5% level for significance, (2-sided p<0.001)

Tresiba | Renown Pharmacy