在二甲双胍的基础上加用胰岛素可能会增加CVD风险

《JAMA》6月11日在线发表的一项研究表明，在接受二甲双胍治疗的2型糖尿病患者中，加用胰岛素而非磺脲类药物与心血管事件风险和全因死亡率增加相关。

美国退伍军人健康管理局-田纳西谷医疗系统老年病学研究教育临床中心的Christianne L. Roumie及其同事称，目前的推荐意见是当二甲双胍单药治疗不能降低HbA1c时，胰岛素作为添加治疗等同于磺脲类药物，而我们的研究结果对此提出了质疑。

Christianne L. Roumie博士

目前，对于应该首选哪种药物作为添加治疗尚未达成共识：胰岛素（长效、预混型、速效）、磺脲类药物（格列本脲、格列吡嗪或格列美脲）、噻唑烷二酮类药物、胰高血糖素样肽-1受体激动剂还是二肽基肽酶-4抑制剂。研究者在开展这项对照研究之前原本预计采用胰岛素而非磺脲类药物进行强化治疗可以改善心血管结局，因为胰岛素在达到血糖控制目标方面更具优势。

研究者评估了42,938例于2001~2008年期间开始服用二甲双胍单药的患者，中位随访期为50个月。共有2,948例患者添加了胰岛素治疗，39,990例添加了一种磺脲类药物。主要复合结局指标是急性心肌梗死（MI）或卒中致入院治疗或者任何原因导致的死亡。

Roumie博士及其同事报告称，在胰岛素组中主要复合结局的发生率为42.7例事件/1,000人-年，而磺脲类药物组为32.8例事件/1,000人-年；校正风险比为1.30（JAMA 2014;311:2288-96）。

研究者指出，敏感性分析以及按患者CVD病史和年龄分层的亚组分析也都得出了类似的结果；我们的研究结果也与近期开展的2项大规模临床试验以及多项观察性研究的结果相吻合，所有这些研究一致表明胰岛素添加治疗较之磺脲类药物添加治疗并无优势，部分研究还提示胰岛素反而会增加心血管风险。

这项研究由美国医疗保健研究与质量局、退伍军人事务部临床科学研究与发展基金以及糖尿病转化研究中心共同资助。Roumie博士及其同事声明无相关经济利益冲突。

随刊评论

潜在危害‘令人担忧’

Monika M. Safford博士说，Roumie等人开展了一项高水平的疗效对照研究，经仔细且全面的分析之后发现胰岛素治疗的“潜在危害信号确实令人担忧。”

她指出：“应该尽快采用其他数据库（比如Medicare、Kaiser或Group Health Cooperative of Puget Sound）进行验证性研究以确定胰岛素添加治疗的潜在危害。”

Safford博士来自美国阿拉巴马大学伯明翰分校医学院。她声明无相关利益冲突。上述文字摘自她对Roumie博士研究报告的随刊评论（JAMA 2014;311:2275-6）。

By: MARY ANN MOON, Cardiology News Digital Network

 

Among patients receiving metformin for type 2 diabetes, adding insulin rather than a sulfonylurea is associated with an increased risk of cardiovascular events and all-cause mortality, according to a report published online June 11 in JAMA.

This finding calls into question the recommendation that insulin is equivalent to sulfonylureas as an add-on agent when metformin alone fails to reduce hemoglobin A_1c, said Dr. Christianne L. Roumie of the Veterans Health Administration–Tennessee Valley Heathcare System Geriatric Research Education Clinical Center, Nashville, and her associates.

They assessed 42,938 patients who began taking metformin alone during 2001-2008 and were followed for a median of 50 months. A total of 2,948 added on insulin therapy and 39,990 added on a sulfonylurea. The rate of the primary composite outcome – acute MI or stroke hospitalization or death from any cause – was 42.7 events per 1,000 person-years with insulin, compared with 32.8 events per 1,000 person-years with sulfonylureas. The adjusted hazard ratio was 1.30, Dr. Roumie and her colleagues reported (JAMA 2014;311:2288-96).At present, there is no consensus as to which add-on medication is preferred: insulin (long acting, premixed, or fast acting), sulfonylureas (glyburide, glipizide, or glimepiride), thiazolidinediones, glucagon-like peptide-1 receptor agonists, or dipeptidyl peptidase 4 inhibitors. The investigators performed this comparative study expecting that intensifying therapy with insulin rather than sulfonylureas would improve cardiovascular outcomes because of insulin’s superiority in achieving glycemic control.

These findings remained robust in a sensitivity analysis and in subgroup analyses that stratified patients by CVD history and age. They are consistent with the results of two recent large clinical trials and several observational studies, all of which found no advantage of add-on insulin over add-on sulfonylureas and some of which reported increased cardiovascular risks with insulin, the investigators said.

This study was funded by the Agency for Healthcare Research and Quality, a Veterans Affairs Clinical Science Research and Development grant, and the Center for Diabetes Translation Research. Dr. Roumie and her associates reported no potential financial conflicts of interest.

View on the News

 ‘Concerning’ potential for harm

Roumie et al. performed a state-of-the-art comparative effectiveness study, with careful and comprehensive analysis that reveals a "signal for potential harm that is certainly concerning," said Dr. Monika M. Safford.

Confirmatory studies addressing the potential harm of add-on insulin therapy "should be conducted relatively quickly, using other databases (such as Medicare, Kaiser, or Group Health Cooperative of Puget Sound)," she said.

Dr. Safford is in the department of medicine at the University of Alabama at Birmingham. She reported no conflicts of interest. These remarks were taken from her editorial accompanying Dr. Roumie’s report (JAMA 2014;311:2275-6).