SGLT2抑制劑之心血管和腎臟結果試驗及嚴重副作用：統合分析

Cardiovascular and Renal Outcomes and Serious Side Effects of SGLT2 Inhibitors: A Meta-analysis

背景：國內使用的3種口服降血糖藥之鈉－葡萄糖協同轉運蛋白2抑制劑（SGLT2i）的大型心血管結果試驗（CVOTs）皆已發表，但尚缺乏其分析比較，本研究目的為探討SGLT2i的CVOTs和腎臟結果試驗及較嚴重副作用，並做一統合分析。方法：以關鍵字“SGLT2 inhibitor”OR“SGLT2 inhibitors”，AND“metaanalysis”OR“systematic review”，搜尋2018年11月24日前之PubMed medline、Cochrane、Web of Science及CEPS中文思博網等資料庫。使用Review Manager 5.3軟體計算風險比（HRs）、95％信賴區間（CIs）及權重，並用MS-Excel及其VBA（Visual Basic for Applications）做森林圖（forest plot），若誤差線（error bar）未交叉到1.0無效線，表示達統計學上顯著的差異。結果：共找到3篇相關文獻做引用，整體而言，SGLT2i可顯著地降低因心衰竭住院（HHF）31％（HR 0.69, 0.61－0.79）和腎臟疾病進展組合（renal composite）45％（0.55, 0.48－0.64），且在3個研究中有相當一致的結果，其HRs皆相差不多。其餘除了心血管死亡（CVM）、非死亡腦中風及全因死亡率（ACM）外之變項亦皆有相當一致的結果。在主要不良心血管事件（MACE）雖只降低11％（0.89, 0.83－ 0.96），但有統計學上顯著差異。Canagliflozin在截肢（1.97, 1.41－2.75）、骨折（1.55, 1.21－1.97）及糖尿病酮酸中毒（DKA）（2.33, 0.76－7.17）皆有較高風險，且其前2項皆達統計學上顯著的差異，三種藥物之三者副作用之統合分析依序上昇26％（1.26, 1.06－1.51）、11％（1.11, 1.00－1.23）及120％（2.20, 1.25－3.87），且皆達統計學上顯著的差異。結論：SGLT2i可降低MACE、HHF和renal composite，除了CVM、非死亡腦中風及ACM外之變項皆有相當一致的結果。截肢、骨折及DKA皆以canagliflozin有較高風險。

Purpose: In spite of the publication of three large-scale cardiovascular outcome trials (CVOTs) of sodium-glucose cotransporters 2 inhibitors (SGLT2i), comparative analysis remains rare. The study accordingly aimed to compare the cardiovascular and renal outcomes and serious side effects of SGLT2i and to conduct a meta-analysis (M-A). Methods: We searched PubMed, Cochrane and Web of Science in EndNote X9 for trials published up to November 24, 2018, using the following keywords: “SGLT2 inhibitor” OR “SGLT2 inhibitors” AND “meta-analysis” OR “systematic review.” The study also searched CEPS with the same keywords in Chinese. Review Manager was used to calculate the hazard ratios (HRs) with 95% CI and weights, and forest plots drawn with MS-Excel with VBA. Significant difference was indicated if the error bars did not cross the no-effect line 1.0. Results: A total of three references were cited. Overall, SGLT2i was able to robustly reduce hospitalization for heart failure (HHF) by 31% (HR 0.69, 95% CI 0.61-0.79) and renal composite by 45% (0.55, 0.48-0.64), and the results appeared to be fairly consistent in three of the four references. Except cardiovascular mortality (CVM), non-fatal stroke, and all-cause mortality (ACM), the results concerning other variables stayed consistent. SGLT2i moderately reduce major adverse cardiovascular events (MACE) by 11.0% with significant difference (0.89, 0.83- 0.96) noted. Compared to the other two agents, Canagliflozin incurred a higher risk in amputation (1.97, 1.41-2.75), fracture (1.55, 1.21-1.97) and diabetic ketoacidosis (DKA) (2.33, 0.76-7.17), while the M-A found all the three side effects reaching significant differences with 26% (1.26, 1.06-1.51), 11% (1.11, 1.00-1.23), and 120% (2.20, 1.25-3.87) respectively. Conclusion: SGLT2i helps reduce MACE, HHF and the progression of renal disease. Except CVM, non-fatal stroke, and ACM, variables are marked with fairly consistent results. Canagliflozin poses greater risks on amputation, fracture and DKA than the other two agents.