IQWiG Reports Commission No. A Dulaglutide

Transcription

1 IQWiG Reports Commission No. A15-07 Dulaglutide Benefit assessment according to 35a Social Code Book V 1 Extract 1 Translation of Sections 2.1 to 2.7 of the dossier assessment Dulaglutid Nutzenbewertung gemäß 35a SGB V (Version 1.0; Status: 29 April 2015). Please note: This translation is provided as a service by IQWiG to Englishlanguage readers. However, solely the German original text is absolutely authoritative and legally binding.

2 Publishing details Publisher: Institute for Quality and Efficiency in Health Care Topic: Dulaglutide Benefit assessment according to 35a Social Code Book V Commissioning agency: Federal Joint Committee Commission awarded on: 2 February 2015 Internal Commission No.: A15-07 Address of publisher: Institute for Quality and Efficiency in Health Care Im Mediapark 8 (KölnTurm) Cologne Germany Tel.: +49 (0) Fax: +49 (0) berichte@iqwig.de Internet: Institute for Quality and Efficiency in Health Care (IQWiG) - i -

3 Medical and scientific advice: Andreas Barthel, Endokrinologikum Ruhr, Bochum, Germany IQWiG thanks the medical and scientific advisor for his contribution to the dossier assessment. However, the advisor was not involved in the actual preparation of the dossier assessment. The responsibility for the contents of the dossier assessment lies solely with IQWiG. IQWiG employees involved in the dossier assessment 2 : Anette ten Haaf Ulrike Mikulić Dorothea Gechter Andreas Gerber-Grote Ulrich Grouven Thomas Kaiser Petra Kohlepp Sarah Mostardt Regine Potthast Keywords: dulaglutide, diabetes mellitus type 2, benefit assessment 2 Due to legal data protection regulations, employees have the right not to be named. Institute for Quality and Efficiency in Health Care (IQWiG) - ii -

4 Table of contents Page List of tables... v List of figures... vii List of abbreviations... viii 2 Benefit assessment Executive summary of the benefit assessment Research questions Research question A: dulaglutide monotherapy Information retrieval and study pool Results on added benefit Extent and probability of added benefit List of included studies Research question B: dulaglutide in dual combination with an oral antidiabetic Information retrieval and study pool Studies included Study characteristics Results on added benefit Outcomes included Risk of bias Results Subgroups and other effect modifiers Extent and probability of added benefit Assessment of added benefit at outcome level Overall conclusion on added benefit List of included studies Research question C: dulaglutide in triple combination therapy with 2 oral antidiabetics Information retrieval and study pool Results on added benefit Extent and probability of added benefit List of included studies Research question D: dulaglutide in combination with insulin with or without oral antidiabetic Institute for Quality and Efficiency in Health Care (IQWiG) - iii -

5 2.6.1 Information retrieval and study pool Studies included Study characteristics Results on added benefit Outcomes included Risk of bias Results Subgroups and other effect modifiers Extent and probability of added benefit Assessment of added benefit at outcome level Overall conclusion on added benefit List of included studies Extent and probability of added benefit summary References for English extract Institute for Quality and Efficiency in Health Care (IQWiG) - iv -

6 List of tables 3 Page Table 2: Subindications, research questions and ACTs for dulaglutide... 1 Table 3: Dulaglutide extent and probability of added benefit Table 4: Subindications, research questions and ACTs for dulaglutide Table 5: Study pool RCT, indirect comparison: dulaglutide + metformin vs. glimepiride + metformin (research question B) Table 6: Characteristics of the studies included RCT, indirect comparison: dulaglutide + metformin vs. glimepiride + metformin (research question B) Table 7: Characteristics of the interventions RCT, indirect comparison: dulaglutide + metformin vs. glimepiride + metformin (research question B) Table 8: Characteristics of the study populations RCT, indirect comparison: dulaglutide + metformin vs. glimepiride + metformin (research question B) Table 9: Risk of bias at study level RCT, indirect comparison: dulaglutide + metformin vs. glimepiride + metformin (research question B) Table 10: Matrix of outcomes RCT, indirect comparison: dulaglutide + metformin vs. glimepiride + metformin (research question B) Table 11: Risk of bias at study and outcome level RCT, indirect comparison: dulaglutide + metformin vs. glimepiride + metformin (research question B) Table 12: Results (mortality, AEs) RCT, indirect comparison: dulaglutide + metformin vs. glimepiride + metformin (research question B) Table 13: Results (morbidity, health-related quality of life, supplementary outcome) RCT, indirect comparison: dulaglutide + metformin vs. glimepiride + metformin (research question B) Table 14: Extent of added benefit at outcome level: dulaglutide + metformin vs. metformin + sulfonylurea (glibenclamide or glimepiride) (research question B) Table 15: Positive and negative effects from the assessment of dulaglutide in comparison with metformin + sulfonylurea (glibenclamide or glimepiride) (research question B) Table 16: Dulaglutide extent and probability of added benefit (research question B) Table 17: Comparison of the titration schedules for insulin glargine in the studies AWARD-2 and LAPTOP (research question C) Table 18: Comparison of the HbA1c values in the common comparator arm (insulin glargine + metformin + glimepiride) of the studies AWARD-2 vs. LAPTOP (research question C) Table 19: Inclusion criteria (AWARD-2 and LAPTOP) and criteria of the sensitivity analyses conducted by the company to increase the comparability of the studies AWARD-2 and LAPTOP (research question C) Table numbers start with 2 as numbering follows that of the full dossier assessment. Institute for Quality and Efficiency in Health Care (IQWiG) - v -

7 Table 20: Study pool RCT, direct comparison: dulaglutide + insulin lispro with or without metformin vs. insulin glargine + insulin lispro with or without metformin (research question D) Table 21: Characteristics of the studies included RCT, direct comparison: dulaglutide + insulin lispro with or without metformin vs. insulin glargine + insulin lispro with or without metformin (research question D) Table 22: Characteristics of the interventions RCT, direct comparison: dulaglutide + insulin lispro with or without metformin vs. insulin glargine + insulin lispro with or without metformin (research question D) Table 23: Characteristics of the study populations RCT, direct comparison: dulaglutide + insulin lispro with or without metformin vs. insulin glargine + insulin lispro with or without metformin (research question D) Table 24: Risk of bias at study level RCT, direct comparison: dulaglutide + insulin lispro with or without metformin vs. insulin glargine + insulin lispro with or without metformin (research question D) Table 25: Matrix of outcomes RCT, direct comparison: dulaglutide + insulin lispro with or without metformin vs. insulin glargine + insulin lispro with or without metformin (research question D) Table 26: Risk of bias at study and outcome level RCT, direct comparison: dulaglutide + insulin lispro with or without metformin vs. insulin glargine + insulin lispro with or without metformin (research question D) Table 27: Results (mortality, AEs) RCT, direct comparison: dulaglutide + insulin lispro with or without metformin vs. insulin glargine + insulin lispro with or without metformin (research question D) Table 28: Results (morbidity, health-related quality of life, supplementary outcomes) RCT, direct comparison: dulaglutide + insulin lispro with or without metformin vs. insulin glargine + insulin lispro with or without metformin (research question D) Table 29: Subgroups: outcome vomiting by age, dulaglutide + insulin lispro with or without metformin vs. insulin glargine + insulin lispro with or without metformin (research question D) Table 30: Extent of added benefit at outcome level: dulaglutide + insulin lispro with or without metformin vs. metformin + human insulin (research question D) Table 31: Positive and negative effects from the assessment of dulaglutide in comparison with metformin + human insulin (research question D) Table 32: Dulaglutide extent and probability of added benefit (research question D) Institute for Quality and Efficiency in Health Care (IQWiG) - vi -

8 List of figures Page Figure 1: Change in HbA1c value in comparison with the baseline value up to week 104 in the HARMONY 3 study Figure 2: Change in HbA1c value in comparison with the baseline value up to week 104 in the AWARD-5 study (missing values imputed with MMRM) Institute for Quality and Efficiency in Health Care (IQWiG) - vii -

9 List of abbreviations Abbreviation ACT AE APPADL BMI CUI EQ-5D G-BA HbA1c IQWiG IW-SP LBSS LOCF MedDRA MMRM NVL OAD PT RCT SAE SGB SOC SPC VAS Meaning appropriate comparator therapy adverse event Ability to Perform Physical Activities of Daily Living body mass index clinical utility index European Quality of Life-5 Dimensions Gemeinsamer Bundesausschuss (Federal Joint Committee) glycosylated haemoglobin A1c Institut für Qualität und Wirtschaftlichkeit im Gesundheitswesen (Institute for Quality and Efficiency in Health Care) Impact of Weight on Self-Perception Low Blood Sugar Survey last observation carried forward Medical Dictionary for Regulatory Activities mixed-effects model repeated measures Nationale VersorgungsLeitlinie (National Care Guideline) oral antidiabetic Preferred Term randomized controlled trial serious adverse event Sozialgesetzbuch (Social Code Book) System Organ Class Summary of Product Characteristics visual analogue scale Institute for Quality and Efficiency in Health Care (IQWiG) - viii -

10 2 Benefit assessment 2.1 Executive summary of the benefit assessment Background In accordance with 35a Social Code Book (SGB) V, the Federal Joint Committee (G-BA) commissioned the Institute for Quality and Efficiency in Health Care (IQWiG) to assess the benefit of the drug dulaglutide. The assessment was based on a dossier compiled by the pharmaceutical company (hereinafter referred to as the company ). The dossier was sent to IQWiG on 2 February Research question The aim of the present report was to assess the added benefit of dulaglutide for the treatment of adults with type 2 diabetes mellitus in comparison with the appropriate comparator therapy (ACT) specified by the G-BA. Four different research questions result from the different combinations with other bloodglucose lowering drugs. These are shown in Table 2. Table 2: Subindications, research questions and ACTs for dulaglutide Research question A B C D Subindication a Monotherapy when diet and exercise alone do not provide adequate glycaemic control in patients for whom the use of metformin is considered inappropriate due to intolerance or contraindications Dual combination therapy with an OAD when this, together with diet and exercise, does not provide adequate glycaemic control Triple combination therapy with 2 OADs when these, together with diet and exercise, do not provide adequate glycaemic control Combination with insulin, with or without OAD, when this, together with diet and exercise, does not provide adequate glycaemic control Research question of the company Module 4 A dulaglutide as monotherapy Module 4 B dulaglutide + metformin Module 4 C dulaglutide + metformin + glimepiride Module 4 D dulaglutide + insulin lispro with or without metformin ACT specified by the G-BA Sulfonylurea (glibenclamide or glimepiride) Metformin plus sulfonylurea (glibenclamide or glimepiride) (note: if metformin is considered inappropriate according to the SPC, human insulin is to be used as treatment option) Metformin plus human insulin (note: if applicable, treatment only with human insulin if metformin is not tolerated or not sufficiently effective according to the SPC) Metformin plus human insulin (note: if applicable, treatment only with human insulin if metformin is not tolerated or not sufficiently effective according to the SPC) a: Subdivisions of the therapeutic indication according to the G-BA. ACT: appropriate comparator therapy; G-BA: Federal Joint Committee; OAD: oral antidiabetic; SPC: Summary of Product Characteristics Institute for Quality and Efficiency in Health Care (IQWiG) - 1 -

11 Deviating from the company, which only considered part of the possible combinations for research questions B to D, the assessment was conducted for the entire subindication in each case. The assessment was conducted based on patient-relevant outcomes and on the data provided by the company in the dossier. Randomized controlled trials (RCTs) with a minimum duration of 24 weeks were used for the derivation of the added benefit. This concurs with the company s inclusion criteria. Results Research question A: dulaglutide monotherapy The company presented no relevant data for research question A. Hence there is no proof of added benefit of dulaglutide in monotherapy versus the ACT specified by the G-BA. Research question B: dulaglutide in dual combination with an oral antidiabetic Dulaglutide in dual combination with metformin No studies of direct comparison were identified for this research question. The indirect comparison between the studies AWARD-5 and HARMONY 3 was included in the assessment of the added benefit of dulaglutide in combination with metformin. Three further indirect comparisons presented by the company are unsuitable for the assessment, particularly because of incomplete data and lack of similarity between the study populations. Study design and treatment regimen (AWARD-5) The AWARD-5 study was a randomized, active-controlled, double-blind approval study (phase 2/3) sponsored by the company to compare dulaglutide with sitagliptin (each with metformin), which consisted of 2 stages. Stage 1 served as a dose-ranging study to choose the 2 dosages out of 7 dulaglutide dosages, the efficacy of which was to be further investigated in Stage 2. These 2 dosages were 0.75 and 1.5 mg dulaglutide/week. Only the 1.5 mg dulaglutide and the sitagliptin arm of the study are relevant for the present benefit assessment (304 and 315 patients respectively). The study consisted of a lead-in phase of up to 11 weeks, a 104-week treatment phase (Stage 1 and 2) and a 4-week follow-up phase. Patients who had received metformin or another oral antidiabetic (OAD) as monotherapy or a combination therapy of metformin with other OADs, and patients without pretreatment were enrolled. The majority of the patients received metformin as monotherapy or in dual combination (88.2% and 86.3% respectively of the patients in the dulaglutide and the sitagliptin arm). Institute for Quality and Efficiency in Health Care (IQWiG) - 2 -

12 The patients received fixed dulaglutide or sitagliptin dosages (in addition to 1500 mg metformin/day). Escalation of the study medication depending on the patients needs was not envisaged. Study design and treatment regimen (HARMONY 3) The HARMONY 3 study was a randomized, active-controlled, double-blind phase 3 study. Adult patients with type 2 diabetes mellitus were enrolled in whom no sufficient glycaemic control was achieved despite treatment with metformin at a stable dosage of 1500 mg/day (or maximum tolerated dosage < 1500 mg/day) (glycosylated haemoglobin A1c [HbA1c] at the last visit in the stabilization phase between 7% and 10%). Four treatment arms were investigated in the study: albiglutide, glimepiride, sitagliptin and placebo (each with metformin). The glimepiride and the sitagliptin arm are relevant for the present benefit assessment (317 and 313 patients respectively). All patients additionally received 1500 mg/day metformin. The study consisted of a 4-week stabilization phase, a treatment phase of 156 weeks and a follow-up phase of 8 weeks. An interim analysis was planned per protocol after all patients had reached at least week 104. Whereas the sitagliptin dosage in the study was fixed, the glimepiride dosage could be increased from 2 mg/day to 4 mg/day. Doses of 1 mg, 3 mg, 5 mg, and 6 mg were not available. An uncertainty therefore remains regarding the influence of the glimepiride treatment regimen. Nonetheless, the results of the HARMONY 3 study were considered to be interpretable and were used for the indirect comparison, also because of comparable HbA1c courses in the glimepiride and sitagliptin arm. Similarity of the AWARD-5 and HARMONY 3 studies The study populations were comparable both between the 2 studies and between the individual arms of the respective studies. The common comparator (sitagliptin + metformin) was also sufficiently similar for the studies available. On the one hand, the specifications on dosing were identical in both studies, on the other, the decrease in HbA1c value in the sitagliptin arms of both studies was comparable. Overall, the 2 studies AWARD-5 and HARMONY 3 were considered to be sufficiently similar so that the assumption of similarity for an adjusted indirect comparison was not rejected. Risk of bias The risk of bias at study level was rated as low for both studies. For the AWARD-5 study, the risk of bias of all outcomes was also rated as low. For the HARMONY 3 study, the risk of bias was rated as low for all outcomes except for severe and symptomatic confirmed Institute for Quality and Efficiency in Health Care (IQWiG) - 3 -

13 hypoglycaemias, which were rated as having a high risk of bias due to the uncertainties regarding the use of glimepiride in the study. General note on the presentation of results and types of analysis In contrast to the HARMONY 3 study, rescue medication was not allowed in the AWARD-5 study. Hence for the HARMONY 3 study, the analyses up to the rescue medication, if available, were used for the indirect comparison. This concerns the outcomes on hypoglycaemia and the outcomes diarrhoea and nausea. Since an adjusted indirect comparison according to Bucher with only one study each was used for research question B, and no direct comparison was available, it was not possible to check homogeneity and consistency. Hence at most hints of added benefit or harm were derived from the available data. Mortality All-cause mortality There was no statistically significant difference between the treatment arms for the outcome all-cause mortality. Hence there is no hint of an added benefit of dulaglutide + metformin in comparison with the ACT metformin + sulfonylurea (glibenclamide or glimepiride). An added benefit for this outcome is therefore not proven. Morbidity No sufficient data were available on the outcome cardiovascular morbidity and on further micro- and macrovascular late complications. Health-related quality of life No relevant data were available for the indirect comparison for the outcome health-related quality of life. Adverse events Serious adverse events and discontinuation due to adverse events The indirect comparison showed no statistically significant differences between the treatment groups for the outcomes serious adverse events (SAEs) and discontinuation due to adverse events (AEs). Hence there is no hint of greater or lesser harm of dulaglutide + metformin in comparison with the ACT metformin + sulfonylurea (glibenclamide or glimepiride); an added benefit for these outcomes is therefore not proven. Severe hypoglycaemia No severe hypoglycaemia occurred in the relevant treatment arms of both studies. Hence there is no hint of greater or lesser harm of dulaglutide + metformin in comparison with the ACT metformin + sulfonylurea (glibenclamide or glimepiride). An added benefit for this outcome is therefore not proven. Institute for Quality and Efficiency in Health Care (IQWiG) - 4 -

14 Symptomatic hypoglycaemia (blood glucose 54 mg/dl; blood glucose 70 mg/dl) There were only analyses on hypoglycaemias with the blood-glucose threshold of 70 mg/dl, but no analyses on the threshold of 54 mg/dl. There was a statistically significant difference in favour of dulaglutide + metformin versus glimepiride + metformin for symptomatic hypoglycaemia with a blood glucose threshold of 70 mg/dl. This results in a hint of lesser harm of dulaglutide + metformin in comparison with the ACT metformin + sulfonylurea (glibenclamide or glimepiride). Gastrointestinal disorders There was no statistically significant difference between the treatment groups for the outcome gastrointestinal disorders. Hence there is no hint of greater or lesser harm of dulaglutide + metformin in comparison with the ACT metformin + sulfonylurea (glibenclamide or glimepiride). An added benefit for this outcome is therefore not proven. Nausea, diarrhoea and vomiting There was a statistically significant difference to the disadvantage of dulaglutide + metformin in comparison with glimepiride + metformin for the outcomes nausea, diarrhoea and vomiting. This results in hints of greater harm of dulaglutide + metformin in comparison with the ACT metformin + sulfonylurea (glibenclamide or glimepiride). Pancreatitis No confirmed pancreatitis occurred in any of the patients in both relevant treatment arms. Hence there is no hint of greater or lesser harm of dulaglutide + metformin in comparison with the ACT metformin + sulfonylurea (glibenclamide or glimepiride). An added benefit for this outcome is therefore not proven. Injection site reactions There was no statistically significant difference between the treatment groups for the outcome injection site reactions. Hence there is no hint of greater or lesser harm of dulaglutide + metformin in comparison with the ACT metformin + sulfonylurea (glibenclamide or glimepiride). An added benefit for this outcome is therefore not proven. The fact that the patients in the glimepiride arm received placebo injections because the ACT glimepiride is administered orally has to be taken into account. Due to the form of administration it can be assumed that results for this outcome cannot occur at all under the use of glimepiride. The actual difference between the interventions is underestimated as a result. Since the proportion of events in the dulaglutide was low (1.3%) and also below the proportion of events in the glimepiride study (7.8%), this has no consequences for the present benefit assessment. Institute for Quality and Efficiency in Health Care (IQWiG) - 5 -

15 Dulaglutide in dual combination with an OAD other than metformin The company presented no relevant studies on dual combinations of dulaglutide with an OAD other than metformin. Hence there is no proof of an added benefit of dulaglutide with an OAD other than metformin in comparison with the ACT metformin + sulfonylurea (glibenclamide or glimepiride). Research question C: dulaglutide in triple combination therapy with 2 oral antidiabetics The company presented no relevant studies of direct comparisons for research question C. The indirect comparison presented by the company is unsuitable for the present benefit assessment because both studies included by the company, AWARD-2 and LAPTOP, are not comparable regarding the common comparator (insulin glargine + metformin + glimepiride) and the study population. The sensitivity analyses conducted by the company are unsuitable to remove the lack of comparability of the common comparator or of the study populations of the studies AWARD-2 and LAPTOP. In addition, the suitability of the LAPTOP study for the indirect comparison is doubtful because it is not guaranteed that the majority of the patients included in the LAPTOP study correspond to the target population (inadequate glycaemic control under maximum tolerated dose of metformin). Hence there is no proof of an added benefit of dulaglutide in the triple combination with 2 OADs in comparison with the ACT metformin + human insulin (or treatment only with human insulin if metformin is not tolerated or not sufficiently effective according to the Summary of Product Characteristics [SPC]). Research question D: dulaglutide in combination with insulin with or without oral antidiabetic Dulaglutide in combination with a short-acting insulin with or without metformin The H9X-MC-GBDD study (hereinafter referred to as AWARD-4 ) was included in the assessment of dulaglutide in combination with a short-acting insulin with or without metformin. The AWARD-4 study was a randomized, active-controlled study sponsored by the company with a treatment phase of 52 weeks. Adult patients ( 18 years) with type 2 diabetes mellitus with inadequate glycaemic control under optimized and stable insulin dosage in conventional insulin treatment alone or in combination with OAD treatment together with diet and exercise were enrolled. A total of 884 patients were randomly assigned in a ratio of 1:1:1 to 3 treatment arms: dulaglutide 0.75 mg daily (293 patients), dulaglutide 1.5 mg daily (295 patients) und insulin glargine (296 patients), each + insulin lispro with or without metformin. Of the 2 dulaglutide arms, only the arm with a dosage of 1.5 mg/week was relevant for the present benefit assessment. Institute for Quality and Efficiency in Health Care (IQWiG) - 6 -

16 After the screening phase, the study consisted of 3 study phases: a 9-week lead-in phase, a 52- week treatment phase and a 4-week follow-up phase. Treatment regimen and dose adjustments The AWARD-4 study was a study with intensive insulin therapy targeted at near-normal blood-glucose levels in both treatment groups. Insulin treatment with insulin glargine and insulin lispro was optimized with defined algorithms in the course of the study. The insulin glargine dose was adapted on the basis of 3 previous fasting plasma glucose levels, aiming at a target value between 71 and 99 mg/dl. The insulin lispro doses for administration before breakfast, before lunch and before the evening meal was also adapted (for all treatment groups equally) according to a prespecified algorithm based on the 3 last fasting plasma glucose levels before lunch, before the evening meal and before bedtime. The target values were between 71 and 100 mg/dl (before lunch, before the evening meal) and between 71 and 130 mg/dl (before bedtime). Risk of bias The risk of bias at study level for the AWARD-4 study was rated as low. Except for the outcomes all-cause mortality and SAEs, the risk of bias of the outcomes was rated as high because of the subjective component in the open-label study design. General note on the presentation of results and types of analysis Analyses on several time periods were partly available for the outcomes included in the assessment. For the present assessment, the analysis of the longest available time period was used for each outcome (also after administration of rescue medication). Hence analyses at the time point 52 weeks were included in the assessment for most outcomes. Mortality All-cause mortality There was no statistically significant difference between the treatment groups regarding deaths. Hence there is no hint of an added benefit of dulaglutide + insulin lispro with or without metformin in comparison with the ACT (metformin + human insulin). An added benefit for this outcome is therefore not proven. Morbidity Health status (European Quality of Life-5 Dimensions visual analogue scale [EQ-5D VAS]) There was no statistically significant difference between the treatment groups for the outcome health status. Hence there is no hint of an added benefit of dulaglutide + insulin lispro with or without metformin in comparison with the ACT (metformin + human insulin). An added benefit for this outcome is therefore not proven. Institute for Quality and Efficiency in Health Care (IQWiG) - 7 -

17 Micro- and macrovascular late complications No evaluable data were available on the outcome cardiovascular morbidity and on further micro- and macrovascular late complications. Health-related quality of life No relevant data were available in the AWARD-4 study for the outcome health-related quality of life. Adverse events Serious adverse events For the outcome SAEs, there was a statistically significant difference in favour of dulaglutide + insulin lispro with or without metformin in comparison with insulin glargine + insulin lispro with or without metformin for the time period of up to week 52. Events occurred across all organ classes without increase in any area. Overall, there is therefore an indication of lesser harm of the combination of dulaglutide + insulin lispro with or without metformin in comparison with the ACT (metformin + human insulin) for this outcome. Discontinuation due to AEs Treatment with dulaglutide + insulin lispro with or without metformin in comparison with the combination therapy with insulin glargine + insulin lispro with or without metformin resulted in a statistically significantly greater proportion of patients with discontinuation due to AEs for the time period up to week 52. Overall, there is therefore a hint of greater harm of dulaglutide + insulin lispro with or without metformin in comparison with the ACT (metformin + human insulin) for the outcome discontinuation due to AEs. It can be inferred from the recordings of the most common AEs in the AWARD-4 study that the majority of events that led to discontinuation can be classified as non-serious. Severe hypoglycaemia There were no evaluable data for the outcome severe hypoglycaemia. Symptomatic hypoglycaemia (blood glucose < 54 mg/dl and blood glucose 70 mg/dl) There was no statistically significant difference between the treatment groups for the outcomes of symptomatic hypoglycaemia (blood glucose < 54 mg/dl and blood glucose 70 mg/dl). Hence there is no hint of greater or lesser harm of dulaglutide + insulin lispro with or without metformin in comparison with the ACT (metformin + human insulin). An added benefit for these outcomes is therefore not proven. Gastrointestinal disorders, nausea, diarrhoea and dyspepsia There was a statistically significant difference to the disadvantage of treatment with dulaglutide + insulin lispro with or without metformin in comparison with insulin glargine + insulin lispro with or without metformin for the outcomes gastrointestinal disorders, Institute for Quality and Efficiency in Health Care (IQWiG) - 8 -

18 nausea, diarrhoea and dyspepsia. In each case this resulted in a hint of greater harm of dulaglutide + insulin lispro with or without metformin in comparison with the ACT (metformin + human insulin). Vomiting There was a statistically significant difference to the disadvantage of treatment with dulaglutide + insulin lispro with or without metformin in comparison with insulin glargine + insulin lispro with or without metformin for the outcome vomiting. In addition, there was proof of an effect modification by the characteristic age for this outcome. For patients < 65 years, this resulted in a hint of greater harm of dulaglutide + insulin lispro with or without metformin in comparison with the ACT (metformin + human insulin). For patients 65 years, there is no hint of greater harm of dulaglutide + insulin lispro with or without metformin in comparison with the ACT (metformin + human insulin). Appetite loss There was a statistically significant difference to the disadvantage of treatment with dulaglutide + insulin lispro with or without metformin in comparison with insulin glargine + insulin lispro with or without metformin for the outcome appetite loss. This resulted in a hint of greater harm of dulaglutide + insulin lispro with or without metformin in comparison with the ACT (metformin + human insulin). Pancreatitis Regarding the proportion of patients with pancreatitis, no events occurred under treatment with dulaglutide + insulin lispro with or without metformin or under treatment with insulin glargine + insulin lispro with or without metformin. Hence there is no hint of greater or lesser harm of dulaglutide + insulin lispro with or without metformin in comparison with the ACT (metformin + human insulin). An added benefit for this outcome is therefore not proven. Injection site reactions There was no statistically significant difference between the treatment groups for the outcome injection site reactions. Hence there is no hint of greater or lesser harm of dulaglutide + insulin lispro with or without metformin in comparison with the ACT (metformin + human insulin). An added benefit for this outcome is therefore not proven. Dulaglutide in combination with a long-acting insulin with or without oral antidiabetic The company presented no relevant data on the combination of dulaglutide with a long-acting insulin with or without OAD. Hence there is no proof of an added benefit of dulaglutide in these combinations in comparison with the ACT metformin + human insulin (or treatment only with human insulin if metformin is not tolerated or not sufficiently effective according to the SPC). Institute for Quality and Efficiency in Health Care (IQWiG) - 9 -

19 Extent and probability of added benefit, patient groups with therapeutically important added benefit 4 On the basis of the results presented, the extent and probability of the added benefit of the drug dulaglutide compared with the ACT is assessed as follows: Research question A: dulaglutide monotherapy Since no relevant study was presented for research question A, there is no proof of an added benefit of dulaglutide in monotherapy in comparison with the ACT specified by the G-BA (sulfonylurea [glibenclamide or glimepiride]). Hence there are also no patient groups for whom a therapeutically important added benefit can be derived. Research question B: dulaglutide in dual combination with an oral antidiabetic Dulaglutide in dual combination with metformin Overall, one positive effect and several negative effects with the same certainty of results and the same extent remain. The positive effect was shown in the outcome category non-serious/non-severe AEs for confirmed symptomatic hypoglycaemia (blood glucose 70mg/dL) with a hint of lesser harm (extent: considerable ). Negative effects were shown in the outcome category non-serious/non-severe AEs for the 3 outcomes nausea, vomiting and diarrhoea with hints of greater harm (extent: in each case considerable ). Furthermore, no sufficient data were available on micro- and macrovascular late complications. In summary, there is therefore no proof of added benefit of dulaglutide in the dual combination with metformin versus the ACT metformin + sulfonylurea (glibenclamide or glimepiride) for patients with type 2 diabetes mellitus. Dulaglutide in dual combination with an OAD other than metformin There is no proof of added benefit of the dual combination of dulaglutide with oral bloodglucose lowering drugs other than metformin in comparison with the ACT metformin + sulfonylurea (glibenclamide or glimepiride). 4 On the basis of the scientific data analysed, IQWiG draws conclusions on the (added) benefit or harm of an intervention for each patient-relevant outcome. Depending on the number of studies analysed, the certainty of their results, and the direction and statistical significance of treatment effects, conclusions on the probability of (added) benefit or harm are graded into 4 categories: (1) proof, (2) indication, (3) hint, or (4) none of the first 3 categories applies (i.e., no data available or conclusions 1 to 3 cannot be drawn from the available data), see [1]. The extent of added benefit or harm is graded into 3 categories: (1) major, (2) considerable, (3) minor (in addition, 3 further categories may apply: non-quantifiable extent of added benefit, no added benefit, or less benefit), see [2]. Institute for Quality and Efficiency in Health Care (IQWiG)

20 Research question C: dulaglutide in triple combination therapy with 2 oral antidiabetics Since no relevant study was presented for the benefit assessment, there is no proof of an added benefit of dulaglutide in the triple combination with 2 OADs in comparison with the ACT metformin + human insulin (or treatment only with human insulin if metformin is not tolerated or not sufficiently effective according to the SPC). Hence there are also no patient groups for whom a therapeutically important added benefit can be derived. Research question D: dulaglutide in combination with insulin with or without oral antidiabetic Dulaglutide in combination with a short-acting insulin with or without metformin Overall, one positive effect and several negative effects with different certainty of results, but the same extent, remain. The positive effect was shown in the outcome category serious/severe AEs for the outcome SAEs with an indication of lesser harm under dulaglutide + short-acting insulin with or without metformin (extent: considerable ). There were negative effects in the outcome category non-serious/non-severe AEs for the following outcomes: discontinuation due to AEs, gastrointestinal disorders, nausea, diarrhoea, vomiting, dyspepsia and decreased appetite, in each case with a hint of greater harm under dulaglutide + short-acting insulin with or without metformin (extent: considerable ). For the outcome vomiting, the negative effect only applies to patients < 65 years. No sufficient data were available on micro- and macrovascular late complications. In the balancing of the results, the negative effects for the outcomes discontinuation due to AEs, gastrointestinal disorders, nausea, diarrhoea, vomiting, dyspepsia and appetite loss do not fully outweigh the advantage of dulaglutide regarding SAEs. However, they resulted in a weakening of the advantage so that, overall, there is a hint of a minor added benefit of dulaglutide + short-acting insulin with or without metformin in comparison with the ACT metformin + human insulin. Dulaglutide in combination with a long-acting insulin with or without oral antidiabetic The company presented no data on other combinations of dulaglutide with a long-acting insulin with or without OAD. Overall, this resulted in no proof of added benefit of dulaglutide + long-acting insulin with or without OAD versus the ACT metformin + human insulin. Summary Table 3 presents a summary of the extent and probability of the added benefit of dulaglutide. Institute for Quality and Efficiency in Health Care (IQWiG)

21 Table 3: Dulaglutide extent and probability of added benefit Research question A Subindication ACT Extent and probability of added benefit Dulaglutide monotherapy when diet and exercise alone do not provide adequate glycaemic control in patients for whom the use of metformin is considered inappropriate due to intolerance or contraindications Sulfonylurea (glibenclamide or glimepiride) B Dulaglutide + metformin Metformin + sulfonylurea (glibenclamide or glimepiride a ) C D Dulaglutide + OAD other than metformin when this, together with diet and exercise, does not provide adequate glycaemic control Dulaglutide + 2 OADs when these, together with diet and exercise, do not provide adequate glycaemic control Dulaglutide + short-acting insulin with or without metformin Dulaglutide + long-acting insulin with or without OAD when this, together with diet and exercise, does not provide adequate glycaemic control (note: if metformin is considered inappropriate according to the SPC, human insulin is to be used as treatment option) Metformin + human insulin (note: if applicable, treatment only with human insulin if metformin is not tolerated or not sufficiently effective according to the SPC) Metformin + human insulin (note: if applicable, treatment only with human insulin if metformin is not tolerated or not sufficiently effective according to the SPC) Added benefit not proven Added benefit not proven Added benefit not proven Added benefit not proven Hint of a minor added benefit Added benefit not proven a: The company chose no option, but presented studies versus glimepiride. Hence glimepiride is the ACT and is printed in bold. ACT: appropriate comparator therapy; G-BA: Federal Joint Committee; OAD: oral antidiabetic; SPC: Summary of Product Characteristics The approach for deriving an overall conclusion on added benefit is a proposal by IQWiG. The G-BA decides on the added benefit. Institute for Quality and Efficiency in Health Care (IQWiG)

22 2.2 Research questions The aim of the present report was to assess the added benefit of dulaglutide for the treatment of adults with type 2 diabetes mellitus in comparison with the ACT. Four different research questions within the therapeutic indication result from the different combinations with other blood-glucose lowering drugs. These are presented in Table 4 together with the respective ACTs specified by the G-BA. Table 4: Subindications, research questions and ACTs for dulaglutide Research question A B Subindication a Monotherapy when diet and exercise alone do not provide adequate glycaemic control in patients for whom the use of metformin is considered inappropriate due to intolerance or contraindications Dual combination therapy with an OAD when this, together with diet and exercise, does not provide adequate glycaemic control C Triple combination therapy with 2 OADs when these, together with diet and exercise, do not provide adequate glycaemic control D Combination with insulin, with or without OAD when this, together with diet and exercise, does not provide adequate glycaemic control Research question of the company Module 4 A dulaglutide as monotherapy Module 4 B dulaglutide + metformin Module 4 C dulaglutide + metformin + glimepiride Module 4 D dulaglutide + insulin lispro with or without metformin ACT specified by the G-BA Sulfonylurea (glibenclamide or glimepiride) Metformin plus sulfonylurea (glibenclamide or glimepiride) (note: if metformin is considered inappropriate according to the SPC, human insulin is to be used as treatment option) Metformin plus human insulin (note: if applicable, treatment only with human insulin if metformin is not tolerated or not sufficiently effective according to the SPC) Metformin plus human insulin (note: if applicable, treatment only with human insulin if metformin is not tolerated or not sufficiently effective according to the SPC) a: Subdivisions of the therapeutic indication according to the G-BA. ACT: appropriate comparator therapy; G-BA: Federal Joint Committee; OAD: oral antidiabetic; SPC: Summary of Product Characteristics The company did not completely investigate the research questions B to D because it only considered part of the possible combinations in each case. It based its research questions on the studies with dulaglutide it had conducted and did not address the missing combinations. Regarding the ACT, the company followed the G-BA s specifications for all research questions. The company additionally considered studies with glipizide for research question B. This had no consequences for the present assessment, however, because the corresponding indirect comparison is unsuitable for the benefit assessment. Institute for Quality and Efficiency in Health Care (IQWiG)

23 The assessment was conducted based on patient-relevant outcomes and on the data provided by the company in the dossier. RCTs with a minimum duration of 24 weeks were used for the derivation of the added benefit. This concurs with the company s inclusion criteria. Institute for Quality and Efficiency in Health Care (IQWiG)

24 2.3 Research question A: dulaglutide monotherapy Information retrieval and study pool The study pool of the assessment was compiled on the basis of the following information: Sources of the company in the dossier: study list on dulaglutide (studies completed up to 1 December 2014) bibliographical literature search on dulaglutide (last search on 1 December 2014) search in trial registries for studies on dulaglutide (last search on 1 December 2014) To check the completeness of the study pool: search in trial registries for studies on dulaglutide (last search on 21 January 2015) No relevant study on the comparison of dulaglutide with the ACT specified by the G-BA (sulfonylurea [glibenclamide or glimepiride]) was identified for the present research question. This concurs with the company s assessment. Alternatively, the company used a study on the comparison of dulaglutide with metformin for its assessment (Study AWARD-3; other study name: H9X-MC-GBDC). This study is unsuitable for the derivation of conclusions on the added benefit of dulaglutide versus the ACT specified by the G-BA (sulfonylurea [glibenclamide or glimepiride]) Results on added benefit The company presented no relevant data for research question A. Hence there is no proof of added benefit of dulaglutide in monotherapy versus the ACT specified by the G-BA Extent and probability of added benefit Since no relevant study was presented for research question A, there is no proof of an added benefit of dulaglutide in monotherapy in comparison with the ACT specified by the G-BA (sulfonylurea [glibenclamide or glimepiride]). Hence there are also no patient groups for whom a therapeutically important added benefit can be derived. The company also claimed no added benefit for the comparison versus the ACT specified by the G-BA, but derived an indication of minor added benefit in comparison with the alternative comparator therapy (metformin as monotherapy) it had chosen List of included studies Not applicable as the company in its assessment did not present any relevant studies on the comparison with the ACT specified by the G-BA. Institute for Quality and Efficiency in Health Care (IQWiG)

25 2.4 Research question B: dulaglutide in dual combination with an oral antidiabetic Information retrieval and study pool The study pool of the assessment was compiled on the basis of the following information: Sources of the company in the dossier: study list on dulaglutide (studies completed up to 1 December 2014) bibliographical literature search on dulaglutide (last search on 1 December 2014) search in trial registries for studies on dulaglutide (last search on 1 December 2014) bibliographical literature search on the ACT (last search on 21 November 2014) search in trial registries for studies on the ACT (last search on 1 December 2014) To check the completeness of the study pool: search in trial registries for studies on dulaglutide (last search on 21 January 2015) search in trial registries for studies on sitagliptin (last search on 4 March 2015) No relevant study of direct comparison was identified from the check Studies included Dulaglutide in dual combination with metformin In Module 4 B of the dossier, the company presented a total of 4 indirect comparison, 3 with the common comparator sitagliptin, and one with the common comparator liraglutide: Common comparator: sitagliptin + metformin studies AWARD-5 (dulaglutide + metformin) and Arechavaleta 2011 (glimepiride + metformin) studies AWARD-5 (dulaglutide + metformin) and Nauck 2007/Seck 2010 (glipizide + metformin) studies AWARD-5 (dulaglutide + metformin) and HARMONY 3 (glimepiride + metformin) Common comparator: liraglutide + metformin studies AWARD-6 (dulaglutide + metformin) and LEAD-2 (glimepiride + metformin) Except for the comparison between the studies AWARD-5 and HARMONY 3, the indirect comparisons considered by the company are unsuitable for the assessment of the added benefit of dulaglutide. This is due to the following reasons: Institute for Quality and Efficiency in Health Care (IQWiG)

26 The comparisons between the AWARD-5 study and the Arechavaleta 2011 (30 weeks) and Nauck 2007/Seck 2010 (104 weeks) studies are unsuitable for the benefit assessment because in both cases the data presented on AEs were incomplete and did not allow to draw an appropriate overall conclusion on added benefit. Moreover, the study populations in the AWARD-5 and the Arechavaleta 2011 studies were not sufficiently similar (see Section of the full dossier assessment). The indirect comparison between the studies AWARD-6 and LEAD-2 (26 weeks) is unsuitable for the benefit assessment because, on the one hand, the populations were not sufficiently similar with regard to baseline HbA1c, and, on the other, the HbA1c decrease in the respective liraglutide arms of the 2 studies was also not comparable. In addition, it was not guaranteed in the LEAD-2 study that the patients had received their maximum tolerated dose of metformin. A detailed explanation of this can be found in Section of the full dossier assessment. Hence only the indirect comparison between the studies AWARD-5 and HARMONY 3 was included in the assessment of the added benefit of dulaglutide in combination with metformin (see Table 5). Data from studies with a duration of 104 weeks were therefore available for research question B. Table 5: Study pool RCT, indirect comparison: dulaglutide + metformin vs. glimepiride + metformin (research question B) Study Study for approval of the drug to be assessed (yes/no) Study category Sponsored study a Third-party study (yes/no) (yes/no) Study with dulaglutide H9X-MC-GBCF Yes Yes No (AWARD-5) b Study with glimepiride HARMONY 3 No No Yes a: Study for which the company was sponsor, or in which the company was otherwise financially involved. b: Hereinafter referred to as AWARD-5. RCT: randomized controlled trial; vs.: versus Section contains a reference list for the studies included. Dulaglutide in dual combination with an OAD other than metformin No relevant studies on dual combinations of dulaglutide with an OAD other than metformin were identified. This concurs with the company s assessment Study characteristics Table 6 and Table 7 describe the studies used for the benefit assessment. Institute for Quality and Efficiency in Health Care (IQWiG)