Dr. Rainer Woker
Every new drug has to be evaluated by a health technology process. Within this process the new compound will be compared to the standard of care therapy in this indication. In Germany a pharmaceutical company has to submit a dossier for this benefit assessment to the Federal Joint Committee (G-BA). That was done by MSD in 2013 for Sitagliptin in combination with Metformin for the treatment of Type-II-Diabetes.
In the dossier Sitagliptin/Metformin was compared with the appropriate comparator Sulfonylurea/Metformin. The G-BA certified that Sitagliptin/Metformin has an additional benefit for the patients. The G-BA conclusion was restricted in a timely manner till July 1st 2016. The reason for the restriction was missing long term data.
Therefore, MSD Germany resubmitted an updated version of the dossier on July 1. All available evidence was provided to the G-BA. The study HARMONY 3 meets the inclusion criteria for this benefit assessment. On October 4 the benefit assessment was published (Link:
https://www.g-ba.de/downloads/92-975-1595/2016-07-01_Nutzenbewertung-IQWiG_Sitagliptin-Metformin-D-246.pdf). It was stated that there is a lack of evidence. Subgroup analyses by metformin dose at baseline for HARMONY 3 are missing.
Therefore we request for data of HARMONY 3 by October 21, 2016. The objective is to show the applicability of the results to the German healthcare context.
[{ "PostingID": 2416, "Title": "GSK-GLP112753", "Description": "A Randomized, Double-Blind, Placebo and Active-Controlled, Parallel-Group, Multicenter Study to Determine the Efficacy and Safety of Albiglutide When Used in Combination With Metformin Compared With Metformin Plus Sitagliptin, Metformin Plus Glimepiride, and Metformin Plus Placebo in Subjects With Type 2 Diabetes Mellitus" }]
1) Summary statistics (n, mean, SD, median, min and max) by treatment arm for daily use of metformin (mg) at baseline in the Safety population. 2) Consider subgroup based on mean baseline metformin use (<1700 mg/>=1700 mg) for the APaT population. Within each of the 2 subgroup levels, the following statistics are needed by treatment arm for each of the below listed endpoints: number of patients in analysis (N), number of patients with event of interest (n). (1) All-cause mortality(2) One or more Adverse Events(3) Serious Adverse Events(4) Discontinuation due to Adverse Events(5) Symptomatic hypoglycemia (blood glucose = 70 mg/dL)(6) Symptomatic hypoglycemia (blood glucose = 54 mg/dL)(7) Hypoglycemic event defined as serious adverse event What Merck would do with the data for each of these 2 items is: 1) No formal statistical comparison of baseline metformin use will be performed. Data will be tabulated descriptively to summarize the metformin use at baseline in the study population.2) The number and proportion of patients experiencing the event of interest will be tabulated by treatment group within each subgroup. In each subgroup, percentages of patients experiencing the event of interest will be compared between the treatment groups using Relative Risk (RR) estimates with associated 95% confidence intervals (CI). Peto-Odds Ratios (and corresponding 95 % CI) instead of RR will be used if the incidence is = 1 % or = 99% for an endpoint in either treatment group. An unconditional exact test (CSZ method) will be used to test for the statistical significance of the treatment difference. In addition to the statistics by subgroup, the consistency of the pairwise treatment effects will be tested using the Cochran's Q-statistic and the I2-statistic. The Q-statistic will be calculated using the logarithm of Relative Risk/Peto Odds-ratio in each subgroup and its standard error. Since the logarithm of the RR/Peto OR is approximately normally distributed and the subgroups are composed of independent sets of patients, the standard Cochran's Q test can be derived and tested using a chi-squared distribution.Data Cut-Offs: The subgroup analyses for all patient relevant endpoints are required for following data cut-offs: · Week 104· Week 164