BACKGROUND Immune checkpoint inhibitors (ICI) exert their therapeutic effect by modulating the immune system and potentiating antitumor immunity. ICI have been associated with several immune-related adverse events, such as diabetes. However, no formal metaanalysis with this respect has been conducted so far. Aim of the present metaanalysis of randomized trials is to assess the effects of ICI on incident diabetes and hyperglycemia. METHODS A MEDLINE, Scopus, ISI-WOS, and Cochrane database search was performed to identify trials, enrolling patients with any form of cancer, up to April 23rd, 2019 in which ICI have been compared either with placebo or active comparators. Data were extracted from published reports or, if not available, from clinicaltrials.gov. The principal endpoints were the incidence of diabetes and cases of hyperglycemia, reported as adverse events. Mantel-Haenszel Odds Ratio with 95% Confidence Interval (MH-OR) was calculated for all outcomes. The study has been registered on PROSPERO website (CDR133927). FINDINGS Out of 42 trials retrieved, 40 reported information on incident diabetes or hyperglycemia. No association of ICI with incident diabetes (MH-OR 1.27 [0.66, 2.43], p = 0.47) was observed; whereas there was a trend toward an increased risk of hyperglycemia (MH-OR 1.45 [0.99, 2.13], p = 0.060), which reached statistical significance in sensitivity analyses and when analyzing separately placebo-controlled trials (MH-OR 1.95 [1.10, 3.49], p = 0.020). I2 statistics did not suggest any relevant heterogeneity for all the principal analyses performed. INTERPRETATION ICI treatment is associated with an increased risk of hyperglycemia, and an increase in the risk of diabetes cannot be excluded.

中文翻译：

---

免疫检查点抑制剂和高血糖：一项随机对照试验的荟萃分析。

背景技术免疫检查点抑制剂（ICI）通过调节免疫系统和增强抗肿瘤免疫来发挥其治疗作用。ICI与多种免疫相关的不良事件（例如糖尿病）有关。但是，到目前为止，尚未进行这方面的正式荟萃分析。本随机试验的荟萃分析的目的是评估ICI对糖尿病和高血糖事件的影响。方法截至2019年4月23日，进行MEDLINE，Scopus，ISI-WOS和Cochrane数据库搜索以鉴定试验，招募患有任何形式癌症的患者，其中将ICI与安慰剂或活性比较剂进行了比较。数据是从已发表的报告中提取的，或者如果没有的话，则从Clinicaltrials.gov中提取。主要终点是糖尿病的发生率和高血糖病例，报告为不良事件。计算所有结果的置信区间为95％（MH-OR）的Mantel-Haenszel赔率。该研究已在PROSPERO网站（CDR133927）上进行了注册。发现的42项试验中，有40项报告了有关糖尿病或高血糖的信息。没有观察到ICI与糖尿病的关联（MH-OR 1.27 [0.66，2.43]，p = 0.47）；而存在高血糖风险增加的趋势（MH-OR 1.45 [0.99，2.13]，p = 0.060），在敏感性分析和单独分析安慰剂对照试验中（MH-OR 1.95 [1.10， 3.49]，p = 0.020）。I2统计数据并未表明所有执行的主要分析均具有任何相关的异质性。解释ICI治疗与高血糖风险增加有关，