Sitagliptin / Metformin (1) – Janumet, Velmetia®, Velmetia®

• Clinical trials
  • Study P803 was a randomised, actively controlled, double-blind study lasting 30 weeks. In the trial, sitagliptin plus metformin was compared with glimepiride plus metformin in patients who had not achieved adequate blood glucose control (HbA1c level ≥ 6.5 % and ≤ 9.0 %) during previous treatment with metformin (at least 1500 mg/day). sitagliptin plus metformin was compared with glimepiride plus metformin.
  • Study P024 was a randomised, actively controlled, double-blind trial lasting 104 weeks, in which the combination of sitagliptin plus metformin was compared with the combination of glipizide plus metformin.
  • To demonstrate the additional benefit of combining sitagliptin with insulin (with and without metformin), the pharmaceutical manufacturer submitted the Hong 2012 study. This was an exploratory, randomised, open-label study which investigated the additional administration of sitagliptin compared with an increase in the insulin dose based on existing insulin therapy.

a) Dual combination of sitagliptin and metformin in addition to diet and exercise to improve blood glucose control in patients in whom monotherapy with metformin at the highest tolerated dose does not sufficiently lower blood glucose

• For the dual combination of sitagliptin and metformin, where diet and exercise plus metformin monotherapy do not sufficiently lower blood glucose levels, there is a hint of a minor additional benefit.
• The certainty of the evidence (probability of additional benefit) is classified as ‘hint’.
• Overall, there is a hint of a minor additional benefit for the dual combination of sitagliptin and metformin, when diet and exercise plus metformin monotherapy do not sufficiently lower blood glucose, compared with the appropriate comparator therapy (glibenclamide or glimepiride in combination with metformin) or compared with treatment with glipizide in combination with metformin, as the results regarding symptomatic, confirmed hypoglycaemic episodes indicate a moderate improvement in treatment-related benefit.
• mortality
  • In study P803, there was no statistically significant difference in overall mortality between the treatment groups.
  • In study P024, there was 1 death (0.2%) in the sitagliptin plus metformin group (588 patients) and 8 deaths (1.4%) in the glipizide plus metformin group (584 patients), which represents a statistically significant result (Peto OR: 0.21 [0.06; 0.77], p = 0.021).
  • However, the mortality data presented here do not possess this high degree of validity, as the mortality findings can only be derived from a post-hoc analysis of the data on adverse events.
  • In light of the uncertainties mentioned above and the fact that these findings were not confirmed in study P803, the findings regarding mortality are to be regarded as insufficiently valid and are therefore not taken into account in the assessment of additional benefit.
• morbidity
  • The results regarding cardiac and cerebrovascular events were not statistically significant in either study.
  • There are therefore no meaningful data available for the morbidity endpoint category – in particular for the cardiovascular and cerebrovascular complications that are generally decisive for the prognosis in type 2 diabetes mellitus – to assess the additional benefit.
  • The primary endpoint of HbA1c selected in studies P803 and P024 (change in HbA1c level compared with the baseline value at the start of the study after week 30) represents a surrogate parameter in the treatment of diabetes mellitus.
• quality of life
  • In study P803, different results were observed for the patient population receiving a metformin dose < 1700 mg (Hedge’s g –0.27 [–0.49; –0.06], p = 0.011) and those receiving ≥ 1700 mg of metformin (Hedge’s g 0.05, [-0.11; 0.21], p = 0.531); however, the differences are not clinically relevant.
  • No data on quality of life were collected in study P024.
• Side effects
  • The results regarding symptomatic, confirmed hypoglycaemia are statistically significant in both studies (Study P803 and Study P024).
  • In study P803, within the relevant target population, 3 symptomatic hypoglycaemic episodes occurred in the sitagliptin plus metformin group (324 patients) and 22 symptomatic hypoglycaemic episodes in the glimepiride plus metformin group (333 patients) (Peto odds ratio [95% CI]: 0.21 [0.10; 0.47]; p-value: < 0.001).
  • In study P024, there were 5 symptomatic hypoglycaemic episodes in the sitagliptin plus metformin group (429 patients) and 34 symptomatic hypoglycaemic episodes in the glipizide plus metformin group (427 patients) up to week 104 (Peto odds ratio [95% CI]: 0.21 [0.11; 0.40]; p-value: < 0.001).
  • Given the largely consistent trends in HbA1c levels across both treatment groups, the present findings on symptomatic, confirmed hypoglycaemic episodes should be interpreted as a significant reduction in side effects and, consequently, as a moderate improvement in treatment-related benefit.
  • Severe hypoglycaemia up to week 104 occurred in the P024 study in 1 out of 429 patients receiving sitagliptin plus metformin, and in 6 out of 427 patients receiving glipizide plus metformin, which represents a statistically significant result (odds ratio [95% CI]: 0.24 [0.05; 1.04]; p-value: 0.011).
  • In the relevant target population, there were 6 therapy discontinuations due to adverse events in the sitagliptin plus metformin group (324 patients) and 2 therapy discontinuations due to adverse events in the glimepiride plus metformin group (333 patients). This result is statistically significant (RR [95% CI]: 3.86 [1.24; 12.05]; p-value = 0.02) to the detriment of sitagliptin in combination with metformin.
  • No statistically significant differences were observed for the other side effect endpoints investigated in the study.
• Overall assessment
  • In the overall assessment of the available findings on side effects, against the background of inconsistent results regarding severe hypoglycaemia and therapy discontinuations in the two studies, and the unclear generalisability to patients who, in line with current knowledge, receive less intensive blood glucose-lowering treatment, a moderate improvement in treatment-related benefit was observed for the dual combination of sitagliptin and metformin—where diet and exercise plus metformin monotherapy do not sufficiently lower blood glucose—and thus a hint of a minor additional benefit compared with the appropriate comparator therapy (glibenclamide or glimepiride in combination with metformin) or compared with treatment with glipizide in combination with metformin.
  • Given the lack of long-term data on cardiovascular endpoints and safety, as well as the partly contradictory results in the studies regarding therapy discontinuations due to adverse events and the occurrence of severe hypoglycaemia, it cannot be concluded overall that there is a relevant reduction in the risk of serious adverse events. A classification as ‘considerable additional benefit’ is therefore not justified.

Triple combination of sitagliptin and metformin with a sulphonylurea, in addition to diet and exercise, in patients for whom a combination of the highest tolerated dose of metformin and a sulphonylurea is insufficient to lower blood glucose

• For the triple combination of sitagliptin and metformin with a sulphonylurea, in addition to diet and exercise, in patients for whom a combination of the highest tolerated dose of metformin and a sulphonylurea is insufficient to lower blood glucose, The additional benefit is not proven.
• No study has been submitted to assess the additional benefit of a treatment consisting of sitagliptin in combination with a sulphonylurea and metformin, in addition to diet and exercise, in patients in whom a combination of the highest tolerated dose of metformin and a sulphonylurea is insufficient, against the appropriate comparator therapy (human insulin + metformin or human insulin alone).

Triple combination of sitagliptin and metformin with insulin as an adjunct to diet and exercise in patients in whom a stable dose of insulin and metformin alone do not sufficiently lower blood glucose

• For the triple combination of sitagliptin and metformin with insulin as an adjunct to diet and exercise in patients in whom a stable dose of insulin and metformin alone do not sufficiently lower blood glucose levels, the additional benefit is not proven.
• However, the G-BA considers the data presented from the Hong 2012 study to be unsuitable for demonstrating additional benefit for the following reasons. Sitagliptin has marketing authorisation in combination with insulin and metformin, but not in combination with insulin and other oral antidiabetics. In the Hong 2012 study, patients in both treatment arms therefore received concomitant treatment with oral antidiabetic agents (α-glucosidase inhibitors, sulphonylureas, glinides and glitazones) that did not comply with the marketing authorisation. Fewer than half of the patients treated in the study received metformin. Data on a subgroup of study participants treated in accordance with the approved indications (add-on treatment with metformin at an approved dose of at least 1700 mg/day), which is relevant to the question at hand, were not provided by the pharmaceutical manufacturer.
• In view of the methodological shortcomings described in the data submitted, the G-BA concludes that, for the triple combination of sitagliptin and metformin with insulin as an adjunct to diet and exercise in patients in whom a stable dose of insulin and metformin alone do not sufficiently lower blood glucose levels, The additional benefit compared with the appropriate comparator therapy (metformin + human insulin or human insulin alone) is not proven.

Courtesy translation only, please refer to the German original.