Presentations at the 81st scientific sessions of the American Diabetes Association , part 2

The following summarizes presentations pertaining to sodium-glucose cotransporter-2 inhibitors (SGLT2i), glucagon-like peptide-1 receptor agonists (GLP-1RA), and chronic kidney disease (CKD) at the American Diabetes Association meeting, held virtually on 25-29 June 2021. Hamaguchi and coauthors treated 57 people with type 1 diabetes (T1D) with dapagliflozin, instructing those with basal to total daily insulin dose ratio >0.4 to reduce the basal insulin; hypoglycemia increased by ~10%-20%, ketosis occurred approximately 6-fold more often, time in range increased 5%-10%, and glycemic variability decreased (abstract 134-LB). Zhu et al (abstract 776) reported a meta-analysis of 10 trials of SGLT2i, confirming that, in patients with T1D, and in patients with type 2 diabetes (T2D) showing greater reduction in fasting insulin, fasting glucagon levels increased, potentially contributing to greater ketogenesis. In a randomized controlled trial comparing ~100 people with T2D receiving empagliflozin 10 mg daily vs placebo, Tanaka et al found that empagliflozin reduced plasma volume (calculated from the baseline and follow-up hemoglobin and hematocrit levels) by 8% and 6% at 12 and 24 weeks, correlating with reduction in N-terminal pro-brain natriuretic peptide (abstract 6-LB). Matsubayashi et al (abstract 794-P) reported 0.14 and 0.34 g/dL increases in hemoglobin from baseline of 15 g/dL with the SGLT2i tofogliflozin in 774 people with T2D; patients with anemia had mean increase of 0.94 g/dL, whereas those with polycythemia had reduction of 0.29 g/dL at 52 weeks, suggesting that hemoconcentration alone does not explain the phenomenon. Mody et al examined a cohort of nearly 2 000 000 people in the United States with T2D and either cardiovascular disease (CVD) or high CVD risk, finding progressive increase in use of SGLT2i from 6% to 13%, and of GLP-1RA with CVD benefit from 4% to 13% from 2015 to 2019 (abstract 851-P), and Dieuzeide et al (abstract 990-P), in a multinational dataset of nearly 10 000 people with T2D in 2019, found ~16% use of SGLT2i and ~10% use of GLP-1RA. Shin et al (abstract 977), however, in a US dataset, found that although first-line glucose lowering treatment with these agents increased from 2013 to 2019 it remained at 4% or lower. Arnold et al reported that, of 14 576 people with T2D from 37 countries adding a second agent to metformin from 2014 to 2016, just 1268 and 321 added an SGLT2i and a GLP-1RA, respectively; younger age, male sex, obesity, and coronary disease (but not cerebrovascular disease) increased the likelihood of addition of the treatments, whereas people with CKD and peripheral arterial disease had lower likelihood of receiving one of these agents (abstract 324-OR). Pimple et al compared 43 588 propensity score matched patients with T2D initiating empagliflozin or a GLP-1RA, finding that the total medical plus pharmacy cost of the latter was $US 195 more than that of the former (abstract 808-P). DeRemer et al compared 11 911 adults with T2D treated with metformin starting GLP-1RA or SGLT2i between 2012 and 2018. The combined rate of myocardial infarction, hospitalization for heart failure (HHF), and stroke (MACE) was similar for the two groups without underlying CVD but was one third lower among those with established CVD starting an SGLT2i, with a 53% lower likelihood of HHF (abstract 220-OR). Other studies suggested different relative benefits. Lyu et al studied hospitalizations among 2492, 1982, and 4762 people with T2D starting SGLT2i, GLP-1RA, or dipeptidyl peptidase 4 inhibitor (DPP4i) between 2015 and 2018 in the Geisinger health system; MACE and HHF hospitalization rates were 30% and 39% lower with SGLT2i than with DPP4i and similarly reduced with GLP-1RA compared with DPP4i (abstract 778-P). In a study by Alaber et al of ~60 000 hospitalized people who received both SGLT2i and GLP-1RA, ~140 000 who received just GLP-1RA, and ~200 000 receiving just SGLT2i, combination treatment was associated with 40%-55% reduction in likelihood of stroke (CVA), myocardial infarction (MI), and heart failure (HF) compared to treatment with just one of the agents, with GLP-1RA monotherapy associated with 14% lower likelihood of CVA and 8% lower likelihood of MI compared with SGLT2i, and similar likelihood of HF (abstract 987-P). The potential for bias related to prescribing decisions explaining some of the outcome differences described must be considered. Lee et al compared a propensity score matched group of 72 689 patients with T2D receiving an SGLT2i and the same number receiving a thiazolidinedione, finding comparable development of atrial fibrillation (5% over a four-year period) (abstract 799-P). In a propensity-score matched study of new users of dapagliflozin vs comparators in US and UK databases the former had ~25% lower likelihood of pyelonephritis or urosepsis and 30% lower likelihood of acute kidney injury (Johannes et al, abstract 805-P and 14-OR). Gautam et al (abstract 764-P) reported that 103 SGLT2i-treated people receiving vs 98 SGLT2i-treated people not receiving cranberry extract had ~25% fewer urinary tract infections; again, the difference might reflect prescribing bias rather than an actual effect of the supplement. Finally, Lee et al (abstract 128-LB, subsequently published1) reported that a study of nearly 50 000 new users of SGLT2i, propensity-score matched with new users of other agents during 2.5-year follow-up, showed retinal vein occlusion occurring 2.19 vs 1.79 per 1000 person-years, a 26% increase; subgroup analysis showed more than threefold increase in risk with estimated glomerular filtration rate (GFR)<60 and 50% greater risk at age ≥60. Gaebe et al (abstract 13-LB) directly measured GFR with plasma iohexol clearance in 53 people with T1D, finding both the creatinine-based CKD Epidemiology Collaboration (CKD-EPI) and modified Schwartz equation for estimating GFR diverged from measured GFR, the CKD-EPI equation overestimating GFR in the lower normal range, suggesting that better approaches to calculation of GFR may be needed. Tang et al (abstract 204-OR) found that participants in the Action to Control Cardiovascular Risk in Diabetes (ACCORD) trial with cardiac autonomic neuropathy diagnosed from analysis of R-R interval trend had greater likelihood of GFR decline during follow-up; the mechanism might involve glycemia- or insulin resistance-induced development both of nephropathy and neuropathy or a direct effect of autonomic neuropathy on renal function. Two presentations reported analyses of the 5674 people with T2D having baseline micro- or macroalbuminuria enrolled in the Efficacy and Safety of Finerenone in Subjects With Type 2 Diabetes Mellitus and Diabetic Kidney Disease (FIDELIO-DKD) study of the novel mineralocorticoid receptor antagonist finerenone. Rossing et al (abstract 14-LB) compared the 259 people with concomitant SGLT2i use to those not using SGLT2i, finding similar reduction in renal and CVD endpoints and lesser development of hyperkalemia. Burgess et al (abstract 189-OR) reported similar effects of finerenone in reducing CVD and renal endpoints and in leading to hyperkalemia with baseline HbA1c ≤7.5% or >7.5%. In a study of the contribution of glucose-dependent insulinotropic polypeptide (GIP) to the insulin-stimulatory effect of dipeptidyl peptidase 4 (DPP-4) inhibitor treatment, Stensen et al (abstract 85-LB) administered a five-hour infusion of the GIP receptor antagonist GIP(3-30)NH2 vs saline to 12 people with T2D receiving sitagliptin vs placebo, showing that the antagonist lowered insulin secretion by approximately one third with sitagliptin as well as with placebo. Frias et al (abstract 101-OR, subsequently published2) compared 961 patients with T2D and HbA1c 8%-10% treated with semaglutide 1.0 vs 2.0 mg weekly, finding modest although statistically significant benefits: 1.9% vs 2.2% reduction in HbA1c from 8.8% and 8.9%, respectively, and 6.0 vs 6.9 kg weight loss. Baek et al (abstract 102-OR) presented results of a 30 + 26 week study of 406 people with T2D randomized to one of three doses of the exenatide-based once weekly GLP-1RA efpeglenatide or placebo, showing a dose-related 56-week A1c reduction of 1.1-1.3% (vs 0.4% with placebo) from a baseline of 8.1% with placebo-adjusted 2 kg weight loss at the higher doses, although with typical GLP-1RA adverse gastrointestinal events. A recent publication reported cardiovascular and renal outcome benefit with this agent, suggesting an important new member of the GLP-1RA class.3 In a study of the dual GIP/GLP-1 receptor agonist tirzepatide in 1444 obese patients with T2D and baseline HbA1c 8.17%, Ludvik et al (abstract 78-LB) reported that 5, 10, and 15 mg doses of the novel agent were associated both with greater HbA1c reduction than titrated insulin degludec and with weight loss, with lower likelihood of hypoglycemia than degludec, although with mild to moderate gastrointestinal event rates. With the 15 mg dose, 48% and 43% achieved HbA1c <5.7% and ≥15% body weight loss, respectively. Dahl et al (abstract 80-LB) reported a 40-week study of 475 obese insulin glargine-treated patients with T2D and baseline HbA1c 8.31% randomized to one of the three tirzepatide doses vs placebo. All doses led to HbA1c <7% in at least 93% of participants, although 14%-19% experienced severe hypoglycemia, whereas with placebo 34% achieved HbA1c <7% and 13% experienced severe hypoglycemia; at the 15 mg dose, 62% achieved HbA1c <5.7%, and 51% achieved ≥10% weight loss. Frias et al (abstract 84-LB, subsequently published4) compared 1879 obese metformin-treated patients with T2D and baseline HbA1c 8.28% randomized to one of the three tirzepatide doses vs semaglutide 1 mg for 40 weeks, finding significantly greater HbA1c-lowering and weight loss with all tirzepatide doses. Serious adverse events occurred more frequently with tirzepatide than with semaglutide, including four deaths with each tirzepatide dose vs one death with semaglutide, one to two acute cholecystitis episodes with each tirzepatide dose vs none with semaglutide, and six myocardial infarctions or strokes in the 1409 people receiving tirzepatide vs none with semaglutide. All three studies showed dose-response in the degree of HbA1c reduction and weight loss. Several reports addressed other potential incretin-related treatments. Tham et al (abstract 106-OR) administered healthy subjects placebo vs single doses of LY3305677, an analog of mammalian oxyntomodulin activating both the GLP-1 and glucagon receptors, with weekly dosing, showing increases in fasting insulin, reductions in fasting glucagon and triglyceride levels, and weight loss sustained over a 29-day period of observation. Golubic et al (abstract 105-OR) administered daily doses of cotadutide, another GLP1/glucagon receptor dual agonist, to overweight adults with T2D, finding weight loss in association with reduced energy intake, and Urva et al (abstract 104-OR) reported a phase 1 study of a triple GIP/GLP-1/glucagon receptor agonist LY3437943 administered as a single dose to normal volunteers, showing a seven-day half-life and dose-dependent weight loss. Pettus et al (abstract 236-OR) reported two 12-week studies of weekly administration of placebo vs volagidemab, a human glucagon receptor antagonist, to 153 people with T1D, showing 0.3% and 0.6% placebo-adjusted HbA1c reduction with modest reduction in insulin dosage. Meiffren et al (abstract 197-OR) administered ADO09, a coformulation of an insulin analog and pramlintide, to 16 people with T1D, showing reduction in gastric emptying and in postmeal glucagon, 1.6 kg weight loss, and increased time in range on continuous glucose monitoring, although with an increase in hypoglycemia and in gastrointestinal events. The authors declare none. 下面总结了在2021年6月25-29日召开的美国糖尿病协会会议上关于钠-葡萄糖共转运体-2抑制剂(SGLT2i)、胰高血糖素样肽-1受体激动剂(GLP-1RA)和慢性肾脏病(CKD)的内容。 1|钠-葡萄糖共转运蛋白-2抑制剂 Hamaguchi及其同事使用达格列净治疗57名1型糖尿病(T1D)患者，指导那些基础胰岛素与每日总胰岛素剂量比>0.4的患者减少基础胰岛素；低血糖增加10%-20%，酮症发生率增加约6倍，范围内达标时间增加5%-10%，血糖变异性降低。Zhu等人报告了对SGLT2i的10项试验的meta分析，证实在T1D患者和空腹胰岛素下降的2型糖尿病患者中，空腹胰岛素下降幅度较大，空腹胰高血糖素水平升高，这可能有助于更多的酮类生成。在一项随机对照试验中，对100名T2D患者每天服用10毫克恩格列净与服用安慰剂进行了比较，Tanaka等人发现，在12周和24周，恩格列净使血浆容量(根据基线和后续的血红蛋白和红细胞压积水平计算)分别减少了8%和6%，这与N-末端前脑利钠肽的减少有关。Matsubayashi等人报道，在服用SGLT2i 托格列净的774名T2D患者中，血红蛋白从基线的15g/dL增加了0.14和0.34g/dL；贫血患者在52周时平均增加了0.94g/dL，而红细胞增多症患者减少了0.29g/dL，这表明仅靠血液浓度不能解释这一现象。 Mody等人研究了美国近两百万人的T2D和心血管疾病(CVD)或高CVD风险队列，发现从2015年到2019年，SGLT2i的使用从6%逐渐增加到13%，有益于CVD的GLP-1RA使用从4%增加至13%。Dieuzeide等在2019年近10 000名T2D患者的跨国数据集中发现SGLT2i的使用率约为16%，GLP-1RA的使用率约为10%。然而，Shin等人在美国的一个数据集中发现，尽管这些药物的一线降糖治疗从2013年到2019年有所增加，但仍保持在4%或更低。Arnold等报道，2014年至2016年，来自37个国家的14 576名T2D患者，在二甲双胍基础上添加了第二种药物，分别仅有1268和321名患者添加了SGLT2i和GLP-1RA；年龄较小、男性、肥胖和冠心病(但不是脑血管疾病)增加了添加其他治疗药物的可能性，而CKD和外周动脉疾病患者接受这些药物之一的可能性较低。Pimple等人比较了43 588例倾向评分匹配的开始使用恩格列净或GLP-1RA的T2D患者，发现后者的医疗和药物总费用比前者高195美元。 DeRemer等人比较了使用二甲双胍基础上，在2012年至2018年期间开始GLP-1RA或SGLT2i治疗的11911名T2D成年人。心肌梗死、因心力衰竭住院(HHF)和卒中(MACE)的合并率在两组没有潜在心血管疾病的患者中相似，但在那些已确诊的心血管疾病患者中，开始接受SGLT2i治疗的患者的心肌梗死、心力衰竭住院(HHF)和卒中(MACE)的综合发生率低三分之一，发生HHF的可能性低53%。其他研究也表明不同的相对益处。Lyu等人研究了盖辛格卫生系统中在2015年至2018年期间，2492名服用SGLT2i、1982名服用GLP-1RA和4762名服用二肽基肽酶4抑制剂(DPP4i)T2D的患者的住院率；SGLT2i的MACE和HHF住院率分别比DPP4i低30%和39%，GLP-1RA与DPP4i的住院率相似。Alaber等人对60000名同时接受SGLT2i和GLP-1RA治疗的住院患者、14万名只接受GLP-1RA治疗的住院患者和20万名只接受SGLT2i治疗的住院患者进行了一项研究，结果表明，与仅使用其中一种药物治疗相比，联合治疗时，卒中(CVA)、心肌梗死(MI)和心力衰竭(HF)的可能性降低40%-55%，而GLP-1RA单一治疗使CVA发生的可能性降低14%，发生心力衰竭(HF)的可能性降低8%。需要注意的是，医生开具药物处方的潜在偏见可以解释一部分的差异结果。 Lee等将72689名接受SGLT2i治疗的T2D患者与接受噻唑烷二酮治疗的72689名T2D患者进行了比较，发现心房颤动的发生率相似(四年内5%)。在一项倾向-分数匹配的研究中，新使用者与美国和英国数据库中的对照者进行了比较，前者发生肾盂肾炎或尿毒症的可能性降低了25%，急性肾损伤的可能性降低了30%。Gautam等人报道，103名接受SGLT2i治疗的患者与98名未接受蔓越莓提取物治疗的患者相比，尿路感染减少了25%；同样，这种差异可能反映了处方的偏见，而不是补充剂的实际效果。最后，Lee等人报告了一项对近5万名新使用SGLT2i的患者进行的研究，发现在2.5年的随访中，与其他药物的新使用者的倾向得分相匹配后，每1000人年发生视网膜静脉阻塞的概率为2.19 vs 1.79，增加了26%；亚组分析显示，估计肾小球滤过率<60ml/mL/1.73m2的风险增加三倍以上，在≥60岁时风险增加50%。 2|慢性肾病 Gaebe等人直接用血浆碘清除率测定53例T1D患者的GFR，发现基于肌酐的CKD流行病学协作(CKD-EPI)和修正的估计GFR的Schwartz方程都偏离了测量的GFR，CKD-EPI方程高估了在正常范围低限的GFR，这表明可能需要更好的方法来计算GFR。Tang等人发现，通过分析R-R间期趋势诊断出心脏自主神经病变的糖尿病心血管风险控制行动(ACCORD)试验的参与者在随访期间GFR下降的可能性更大；其机制可能涉及血糖或胰岛素抵抗导致的肾病和神经病变的发展，或者自主神经病变对肾功能的直接影响。两篇报告报道了对5674名有微量或大量蛋白尿的T2D患者的分析，这些患者参加了新型盐皮质激素受体拮抗剂非奈利酮对2型糖尿病和糖尿病肾病(FIDELIO-DKD)受试者的有效性和安全性研究。Rossing等人比较了259名使用SGLT2i和未使用SGLT2i的患者，发现肾脏和心血管终点的减少相似，高钾血症的发生较少。Burgess等人报道，在基线糖化血红蛋白(HbA1c)≤为7.5%或>7.5%的情况下，非奈利酮在减少心血管疾病和肾脏终点以及导致高钾血症方面有类似的效果。 3|胰高血糖素样肽-1受体激动剂(GLP-1RA)及相关治疗途径 在一项关于葡萄糖依赖型促胰岛素多肽(GIP)对二肽基肽酶4(DPP-4)抑制剂的胰岛素刺激效应研究中，Stensen等人对12名接受西格列汀和安慰剂治疗的T2D患者进行了5小时的GIP受体拮抗剂GIP(3-30)NH2 vs. 生理盐水的输注，结果显示该拮抗剂使西格列汀和安慰剂的胰岛素分泌减少了约三分之一。 Frias等人对961名接受司美格鲁肽治疗的HbA1c在8%-10%的T2D患者进行比较，发现每周注射司美格鲁肽1.0和2.0 mg的患者，HbA1c分别从8.8%和8.9%下降1.9%和2.2%，体重减轻6.0 kg和6.9 kg，差异有统计学意义。Baek等人公布了一项对406名T2D患者进行的30+26周研究的结果，这些患者被随机分成三种剂量的GLP-1RA，每周一次的GLP-1RA 依培那肽或安慰剂，结果显示，与剂量相关的糖化血红蛋白从8.1%下降了1.1-1.3%(与安慰剂0.4%相比)，尽管有典型的GLP-1RA不良胃肠道反应，但可以使体重减轻2公斤。最近的一篇文章报道了这种药物对心血管和肾脏结果的益处，这表明它是GLP-1RA类别的一个重要的新成员。在一项对1444名患有T2D和基线HbA1c 8.17%的肥胖症患者使用GIP/GLP-1受体激动剂tirzepatide的研究中，Ludvik等人报告说，5、10和15毫克剂量的新型药物与滴定的德谷胰岛素相比，糖化血红蛋白的降低幅度更大，体重减轻，低血糖的可能性比德谷胰岛素低，尽管有轻度到中度的胃肠道事件发生率。当剂量为15 mg时，43%受试者糖化血红蛋白(HbA1c)<5.7%，43%体重下降≥15%。Dahl等人报道了一项为期40周的研究，研究对象为475名接受甘精胰岛素治疗的肥胖T2D患者，基线HbA1c为8.31%，他们被随机分配到三种tirzepatide剂量中的一种，与安慰剂相比，所有剂量的参与者中至少93%的人HbA1c<7%，尽管14%-19%经历了严重低血糖，而安慰剂的患者中，34%的人HbA1c<7%，13%的人经历了严重低血糖；Frias等人比较1879名肥胖T2D患者，基线HbA1c为8.28%，他们随机接受三种剂量tirzepatide vs. 索马鲁肽 1 mg，治疗40周，发现所有剂量的tirzepatide 均显著降低HbA1c并减轻体重。接受tirzepatide治疗的患者比用索马鲁肽的患者发生严重不良事件的频率更高，包括每次注射tirzepatide剂量的患者死亡4例，索马鲁肽的患者死亡1例，注射tirzepatide的患者发生急性胆囊炎1-2次，注射tirzepatide的患者发生6次心肌梗死或卒中，注射索马鲁肽的1409人中没有发生心肌梗死或中风。这三项研究在糖化血红蛋白降低和体重减轻的程度上都显示出剂量效应。 一些报告提到了其他潜在的与肠促胰岛素相关的治疗方法。Tham等人给予健康受试者安慰剂与单剂量LY3305677，一种激活GLP-1和胰高血糖素受体的哺乳动物泌酸调节素类似物，每周给药，结果显示空腹胰岛素增加，空腹胰高血糖素和甘油三酯水平减少，并在29天的观察期内减轻体重。Golubic等人每天给超重的T2D成年人另一种GLP1/胰高血糖素受体双重激动剂cotadutide，发现体重减轻与能量摄入减少有关；Urva等人报道了一项对正常志愿者单次使用三重GIP/GLP-1/胰高血糖素受体激动剂LY3437943的1期研究，显示7天半衰期和剂量依赖性体重减轻。Pettus等报道了153例T1D患者每周给予安慰剂 vs 人胰高血糖素受体拮抗剂volagidemab的两项为期12周的研究，结果显示胰岛素使用剂量降低0.3%和0.6%。Meiffren等对16例T1D患者给予ADO09(一种胰岛素类似物和普兰林肽的联合制剂)，显示胃排空和餐后胰高血糖素减少，体重减轻1.6 kg，血糖范围内时间延长，尽管低血糖和胃肠道事件增加。