Diabetes remission off medications is not a suitable endpoint for comparing bariatric/metabolic surgery with pharmacotherapy. Reply to Halpern B, Cercato C, Mancini MC [letter].
Diabetes remission off medications is not a suitable endpoint for comparing bariatric/metabolic surgery with pharmacotherapy. Reply to Halpern B, Cercato C, Mancini MC [letter].
复制标题
药物治疗后的糖尿病缓解并不是减肥/代谢手术与药物治疗比较的合适终点。
DOI：
10.1007/s00125-016-4029-x
复制
发表时间：
2016
期刊：
影响因子：
8.2
通讯作者：
Cummings,DavidE
中科院分区：
文献类型：
--
作者：
Cummings,DavidE
致编辑：我谨代表《卡路里减少或手术：寻求减少肥胖和糖尿病研究》(CROSSROADS)[1] 的作者，衷心感谢 Halpern 等人对我们的试验提出的深思熟虑的批评 [2]。除其他外，这些研究人员对在任何比较减肥/代谢手术与医疗/生活方式干预 (MLI) 的随机临床试验 (RCT) 中使用糖尿病缓解作为主要终点提出了挑战。“糖尿病缓解”在各种 RCT 中被定义为实现 HbA1c 目标 — 通常< 6.0% (42 mmol/mol)、< 6.5% (48) mmol/mol) 或< 7.0%(53 mmol/mol)——所有糖尿病药物均无效。由于在 MLI 群体中实现这一目标的最有力工具是使用药物，Halpern 等人认为要求这些研究参与者停用此类药物以实现主要结果是不公平的。这是一个非常合理的批评：这是最近举行的第二届糖尿病手术峰会上广泛讨论的话题，这是一次共识发展会议，为使用和研究减肥/代谢手术治疗 2 型糖尿病制定了新的指南 [3]。尽管 CROSSROADS 试验的主要终点确实是糖尿病缓解 [1]，但自几年前最初设计该研究以来，我们普遍同意以下观点：在该领域的手术与 MLI 随机对照试验中，血糖控制测量（即在使用或不使用糖尿病药物的情况下达到指定的 HbA1c 阈值）在大多数情况下是比糖尿病缓解更好的主要结局。 案例。然而，值得注意的是，在此类研究中，药物并不是随机分配到非手术治疗组的患者可以用来改善糖尿病的唯一工具，而且 CROSSROADS 中的生活方式干预特别密集，可能比迄今为止任何相关的随机对照试验都要强烈。该干预措施以 Look AHEAD（糖尿病健康行动）试验为蓝本，包括每天 ≥ 45 分钟、每周 ≥ 5 天的有氧运动，主要在设备齐全的专用研究健身房中在私人教练的直接监督下进行。它还涉及密切监督的减肥和降糖饮食，以及面对面的一对一和每周的团体咨询课程。尽管这种严格的生活方式干预在临床实践中很难实施，但 CROSSROADS 的一个主要目标是确定在最好的情况下，积极的饮食、锻炼和最先进的医疗方案是否可以与胃绕道手术对糖尿病状态的影响相匹配。在这个案例中，答案是“不”。重要的是，我们的研究结果与迄今为止已发表的其他 10 项随机对照试验的结果一致，直接比较了几种减肥/代谢手术与各种 MLI 治疗 2 型糖尿病的效果（图 1）[3]。这些试验的一个普遍发现是，就所有血糖结果而言，手术均优于 MLI。这不仅包括可能不公平的终点
To the Editor: On behalf of the authors of the Calorie Reduction Or Surgery: Seeking to Reduce Obesity And Diabetes Study (CROSSROADS)[1], I sincerely thank Halpern et al for their thoughtful critique of our trial [2]. Among other things, these investigators challenge the use of diabetes remission as the primary endpoint in any randomised clinical trial (RCT) comparing bariatric/metabolic surgery vs a medical/lifestyle intervention (MLI).‘Diabetes remission’is defined in various RCTs as achievement of an HbA1c target—typically< 6.0%(42 mmol/mol),< 6.5%(48 mmol/mol) or< 7.0%(53 mmol/mol)—off all diabetes medications. Since the most powerful tool to attain that goal in MLI groups is the use of medications, Halpern et al feel it is unfair to require those study participants to be off such agents in order to achieve the primary outcome. This is a very reasonable critique: the topic of extensive discussion at the recent 2nd Diabetes Surgery Summit, a consensusdevelopment conference that generated new guidelines for the use and study of bariatric/metabolic surgery to treat type 2 diabetes [3]. Although the primary endpoint of the CROSSROADS trial was, indeed, diabetes remission [1], during the time since the study was initially designed several years ago we have generally come to agree with the assertion that in surgery-vs-MLI RCTs in this domain, measurement of glycaemic control—ie achievement of a specified HbA1c threshold with or without diabetes medications—is a better primary outcome than diabetes remission in most cases. It is important to note, however, that medications are not the only instruments that patients randomised into non-surgical treatment arms in such studies can use to improve their diabetes, and the lifestyle intervention in CROSSROADS was particularly intensive, probably more so than in any related RCT to date. Modelled after the Look AHEAD (Action for Health in Diabetes) trial, this intervention included≥ 45 min/day of aerobic exercise,≥ 5 days/week, largely under the direct supervision of a personal trainer in a well-equipped, dedicated research gym. It also involved a closely supervised weight-reducing and glucoselowering diet, with in-person one-on-one and weekly group counselling sessions. Although this rigorous lifestyle intervention would be difficult to implement in clinical practice, a major goal of CROSSROADS was to determine whether, under the best of circumstances, an aggressive diet, exercise and state-ofthe-art medical regimen can match the impact of gastric bypass on diabetes status. The answer in this case was ‘no’. Importantly, our findings are consistent with those of ten other published RCTs to date directly comparing several bariatric/metabolic operations against a variety of MLIs to treat type 2 diabetes (Fig. 1)[3]. A universal finding among these trials is that surgery is superior to MLIs for all glycaemic outcomes. This includes not just the possibly unfair endpoint