Gil M, Romio S, Scotti L, Leal I, Masclee GMC, Schmedt N, Varas-Lorenzo C, Pladevall-Vila M, Smits MM, Bezemer I, Seeger JD, de Beradis G, Pecchioli S, Trifiro G, McGrogan A, Rijnbeek P, Sturkenboom M. Risk of hemorrhagic stroke associated with non-insulin blood glucose lowering drugs. Results from the SAFEGUARD project. Poster presented at the 32nd ICPE International Conference on Pharmacoepidemiology & Therapeutic Risk Management; August 27, 2016. Dublin, Ireland. [abstract] Pharmacoepidemiol Drug Saf. 2016 Aug; 25(Suppl 3):428-9.

BACKGROUND: The SAFEGUARD project aims to assess the cardio/cerebrovascular and pancreatic safety of non-insulin blood glucose lowering drugs (NIGBLDs) in type 2 diabetes mellitus patients (T2DM). To date, no studies addressing the risk of Hemorrhagic stroke (HS) in patients exposed to NIGBLDs have been published.

OBJECTIVES: To estimate the risk of HS associated with individual NIBGLDs in the multidatabase network of the SAFEGUARD project.

METHODS: A case-control design, nested in a cohort of new NIBGLD users was performed. Incident HS cases (intracerebral/subarachnoid hemorrhage) were matched with up to 5 controls on database (DB), sex, cohort entry(± 3 months) and date of birth (± 1 year) using risk set sampling. Data were retrieved from 9 electronic healthcare data sources (PHARMO, IPCI (Netherlands); BIFAP (Spain); GePaRD (Germany); Health Search, Regional DBs of Lombardy and Puglia (Italy); CPRD (United Kingdom); Medicare (US)). DB-specific adjusted odds ratios (ORs) and corresponding 95% confidence intervals (95%CI) were estimated by comparing current use of metformin in combination with sulfonylureas (SUs) (reference) with each NIGBLD monotherapy, dual therapy of metformin plus another NIGBLD (not SUs) and other NIGBLD combinations. Metaanalyses of data source specific estimates (ORmeta with 95% CI) and conditional logistic regression based on the individual pooled data (adjOR pool with 95% CI) were performed.

RESULTS: A total of 3,817 cases of HS were matched to 18,656 controls T2DM. As compared with users of metformin in combination with SUs, current use of repaglinide monotherapy (adjOR pool=l.68 (1.28- 2.20); 0Rmeta=2.12 (1.11-4.05)). The metaanalysis also showed an increased risk of HS in current users of metformin in combination with rosiglitazone (ORmeta=l.85 (1.04-3.29)).

CONCLUSIONS: Our findings suggest that the risk of HS is higher in current users of repaglinide monotherapy and in those using the combination of metformin and rosiglitazone as compared with current users of metformin and SUs.

Share on: