PDB21 - COMPARATIVE EFFECTIVENESS OF SGLT2 INHIBITORS IN TYPE II DIABETES DERIVED FROM AN OVERVIEW OF 55 SYSTEMATIC REVIEWS

Abstract

Various SGLT2 inhibitors have been developed and many systematic reviews (SRs) have been published. Some of them are SRs for individual drugs and others are SRs for SGLT2 class drug. We explore the difference between individual drug SRs and the class drug SRs and compare the risk-benefit among SGLT2 inhibitors from network meta-analysis(NMA). Systematic review articles were retrieved by the Medline database. We collected the characteristics of year published, nationality, meta-analysis(MA) including NMA, cardiovascular (CV) events or not, literature search rigorousness, positive finding or not. These factors and their categories were displayed into multi-dimensional configurations by the correspondence analysis. Classification and regression tree(CART) was also applied for discriminating individual drug SRs from the class drug SRs. A total of 55 systematic review articles related to SGLT2 inhibitors were retrieved from 2012 to 2018. By nationality, there were 20 from China, 12 from US, 23 from others. It included 30 conventional MAs, 15 network MAs, and 10 qualitative SR articles. CV events were dealt with in 10 (18%) most of which were academia initiated. Rigorous literature search was conducted in 13 (24%). There were 23 individual drug SRs (dapagliflozin 9, canagliflozin 7, empagliflozin 7) and 32 class drug SRs. Industry sponsor, weak methodology, non-CV, Chinese trials were closely configurated near individual drug SRs by correspondence analysis. The CART also confirmed that individual drug SRs tended to be industry-sponsored, non-CV, positive in findings, weak in methodology. Regarding the risk-benefit among SGLT2 inhibitors, no apparent difference was found from the NMAs although SRs for canagliflozin have been conducted mostly by China. Individual drug SRs were closely related to those with industry sponsor, non-CV endpoints and weak methodological aspects. There seemed to be no apparent difference among individual SGLT2 inhibitors.