Sodium–glucose co-transporter-2 inhibitors (SGLT2i) are a novel class of glucose-lowering agents that inhibit glucose reabsorption by renal proximal tubules, thereby improving blood glucose levels in association with reduction of blood pressure and body weight in patients with diabetes in an insulin-independent manner.^{1, 2} Recent randomized clinical trials have revealed that SGLT2i significantly reduce the risk of composite cardiovascular endpoints, including death, major cardiovascular events and hospitalization of heart failure in type 2 diabetic patients.^{3, 4} Furthermore, SGLT2i have also been shown to decrease the risk of renal events, regardless of the presence or absence of chronic kidney disease.^{5, 6} Therefore, many clinical practice guidelines currently have recommended SGLT2i as an oral hypoglycaemic agent for the treatment of high-risk patients with type 2 diabetes, such as those with a history of cardiovascular disease.⁷ Furthermore, meta-analysis showed that combination therapy of SGLT2i with incretin-based agents exhibited more favourable effects on glycated haemoglobin (HbA1c) and systolic blood pressure in patients with type 2 diabetes.⁸

Now SGLT2i are one of the most widely used drugs for the treatment of type 2 diabetic patients.^7-9 In addition, some types of SGLT2i, such as dapagliflozin and ipragliflozin, are approved in Europe and Japan as an adjunct therapy to insulin in adults with type 1 diabetes, which could also improve glycaemic control in association with reduction of insulin requirements and body weight in these patients.⁹ However, there is some safety concern associated with their use in diabetic patients.^{9, 10} Indeed, SGLT2i have been reported to increase the risk of euglycaemic ketoacidosis, a rare but serious life-threatening complication, especially in patients with type 1 diabetes.^9-11 SGLT2i increase plasma glucagon levels and simultaneously reduce blood glucose and insulin levels in type 2 diabetes.^{10, 12} Dapagliflozin have been shown to stimulate glucagon secretion from pancreatic alpha cells as well.¹³ Since glucagon promotes hepatic ketogenesis under fasting conditions and acts as an antagonist to insulin,¹⁴ the increased ratio of glucagon to insulin has been supposed to contribute to the risk of euglycaemic ketoacidosis in SGLT2i-treated diabetic patients. However, some paper showed that interruption of glucagon signalling did not affect SGLT2i-induced ketosis in mice.¹⁵ Therefore, the role of glucagon for euglycaemic ketoacidosis in SGLT2i-treated type 1 diabetic patients remains unclear.

Moreover, glucagon exerts various metabolic effects in humans; it not only promotes gluconeogenesis, but also affects lipid metabolism and energy expenditure.¹⁶ Glucagon increases glucogenic amino acids, such as alanine, which in turn stimulates glucagon secretion.¹⁷ These potential biological effects of glucagon could offset the clinical effectiveness of SGLT2i in diabetic patients by stimulating gluconeogenesis, or augment their cardiometabolic effects via the lipolytic and thermogenic properties.¹⁴ However, it remains unclear whether plasma glucagon levels are actually increased by the treatment with SGLT2i in type 1 diabetic patients, and if so, which clinical parameters are associated with the increase in glucagon. Therefore, in this commentary, we would like to discuss the pros and cons of glucagon in SGLT2i-treated type 1 diabetic patients on the basis of our recent clinical study.

Nine type 1 diabetic patients (two males and seven females; mean age of 50.6 ± 7.3 years old), who had already received intensive insulin therapy in Showa University Hospital, were enrolled in the present study. The exclusion criteria of this study were as follows: (i) patients treated with hypoglycaemic agents other than insulin, (ii) low-carbohydrate diets, (iii) a history of diabetic ketoacidosis or severe hypoglycaemia that requires another person's help during the previous 6 months, (iv) pregnancy and (v) chronic kidney disease (estimated glomerular filtration rate (eGFR) < 30 mL/min/1.73 m²). During the study period, patients were instructed not to change their life habits and to continue to take the same dose of any concomitant drugs, except for insulin. Insulin doses were adjusted as needed to avoid hypoglycaemia. At baseline and 24 weeks after treatment with ipragliflozin, blood was drawn after fasting for determination of blood biochemistry, and body composition was simultaneously evaluated. All participants gave their informed consent to participate in the present study. The study protocol was approved by the Ethics Committee of the Showa University School of Medicine, Tokyo, Japan (Permission number: 3152).

Glucagon was measured with sandwich enzyme-linked immunosorbent assay (Mercodia AB), while ketone bodies with enzyme-amplified assay (KAINOS Laboratories Inc.). N-terminal pro-brain natriuretic peptide (NT-pro-BNP) was evaluated with electrochemiluminescence immunoassay (Roche, Basel, Switzerland). Other blood chemistry was measured as described previously.¹⁸ Whole-body muscle and fat masses were evaluated with bioelectrical impedance analysis using an InBody770 (InBody770; Biospace). Data were presented as mean values ± standard deviation. Comparisons between groups were performed using Student's t-test or the non-parametric Mann–Whitney U test. Non-parametric correlations were identified using the Spearman's rank correlation coefficient.

Clinical characteristics of the participants and changes in clinical parameters at baseline and 24 weeks after treatment with ipragliflozin are shown in Table 1. After 24-week treatment, HbA1c, body weight, body mass index, body fat, waist circumstances and alkaline phosphatase were significantly decreased, while haemoglobin and haematocrit levels were increased. There was a trend for increase of total ketone bodies. Total, basal and bolus insulin doses were modestly, but not significantly, reduced by ipragliflozin treatment. Mean glucagon levels were significantly increased from 25.8 ± 9.1 to 42.1 ± 13.4 pg/ml (p = 0.0046). Change in glucagon after ipragliflozin treatment (Δglucagon) was inversely correlated with Δlow-density lipoprotein cholesterol (LDL-C) and ΔeGFR (r = −0.72, p = 0.03 and r = −0.75, p = 0.02, respectively) (Table 2). Furthermore, baseline body mass index and body fat mass were significantly associated with Δglucagon (Table 3). Although some patients experienced mild hypoglycaemia, there were no cases of urinary tract and genital infections, severe hypoglycaemia, or diabetic ketoacidosis.

As far as we know, this is the first report to demonstrate that 24-week treatment with ipragliflozin significantly increased plasma glucagon levels in association with reduction of HbA1c, body mass index and body fat mass in type 1 diabetic patients. Furthermore, although ipragliflozin treatment did not alert mean LDL-C levels, Δglucagon was inversely associated with ΔLDL-C levels, which was significantly correlated with baseline body mass index and body fat mass. The present findings suggest that SGLT2i-induced increase in plasma glucagon levels may be involved in reduction in LDL-C in type 1 diabetic patients, especially those with high body mass index and increased body fat mass. Therefore, the present findings suggest that glucagon may be a friend for type 1 diabetic patients receiving SGLT2i.

Glucagon has been shown to alter lipid metabolism and decrease LDL-C levels in both animal models and humans via various mechanisms. Indeed, administration of glucagon to rats has been reported to decrease LDL-C levels.^{19, 20} Some glucagon receptor antagonists have been shown to dose-dependently increase LDL-C levels in type 2 diabetic patients partly by increasing cholesterol absorption.^21-23 Moreover, Spolitu et al. recently found that inhibition of hepatic glucagon signalling increased proprotein convertase subtilisin/kexin type 9 (PCSK9) expression and resultantly increased LDL-C levels in mice by reducing the LDL receptor.²⁴ The facts that obesity is associated with increased PCSK9 expression^{25, 26} could partly explain why ipragliflozin-induced decrease in LDL-C levels was observed in type 1 diabetic patients with high body mass index and increased body fat mass. In other words, SGLT2i-induced elevation of glucagon may lower LDL-C levels partly through the suppression of PCSK9 in these subjects.

No cases with diabetic ketoacidosis were observed in the present study. Furthermore, although circulating total ketone body levels were modestly increased by the treatment with ipragliflozin, there was no correlation between Δglucagon and Δtotal ketone bodies. In this study, prandial (bolus), but not basal insulin dose, was cautiously reduced when initiating the SGLT2i therapy to avoid hypoglycaemia. Patients following a low carbohydrate diet were excluded from the present study. Since reduced basal insulin by more than 10%–20% and/or low carbohydrate diet have been shown to be risk factors for ketoacidosis associated with SGLT2i therapy in diabetes,²⁷ strategies for reducing the risk factors for ketoacidosis could enhance the efficacy and safety of this life-saving drug in type 1 diabetes. In any case, further clinical studies should be needed to clarify the pathophysiological role of glucagon in type 1 diabetic patients receiving SGLT2i. We are now planning to explore the role of glucagon in metabolic parameters, including glycaemic variability and LDL-C in type 1 diabetic patients receiving ipragliflozin in a multicentre collaborative trial, which was already registered with the UMIN Clinical Trials Registry (University Medical Information Network 000020156).

钠-葡萄糖协同转运蛋白 2 抑制剂 (SGLT2i) 是一类新型降糖药，可抑制肾近端小管对葡萄糖的重吸收，从而改善血糖水平，同时降低糖尿病患者的血压和体重。一种不依赖胰岛素​​的方式。^{1, 2}最近的随机临床试验表明，SGLT2i 显着降低了复合心血管终点的风险，包括 2 型糖尿病患者的死亡、主要心血管事件和心力衰竭住院治疗。^{3, 4}此外，SGLT2i 还被证明可以降低肾脏事件的风险，无论是否存在慢性肾脏疾病。^5、6因此，目前许多临床实践指南都推荐SGLT2i作为口服降糖药，用于治疗高危2型糖尿病患者，例如有心血管病史的患者。⁷此外，荟萃分析表明，SGLT2i 与基于肠促胰岛素的药物联合治疗对 2 型糖尿病患者的糖化血红蛋白 (HbA1c) 和收缩压表现出更有利的影响。⁸

现在SGLT2i是治疗2型糖尿病患者最广泛使用的药物之一。^7-9此外，某些类型的 SGLT2i，如达格列净和伊格列净，在欧洲和日本被批准作为成人 1 型糖尿病患者胰岛素的辅助治疗，这也可以通过减少胰岛素需求来改善血糖控制和这些患者的体重。⁹然而，在糖尿病患者中使用它们存在一些安全问题。^{9, 10}事实上，据报道，SGLT2i 会增加正常血糖酮症酸中毒的风险，这是一种罕见但严重的危及生命的并发症，尤其是在 1 型糖尿病患者中。^9-11SGLT2i 增加血浆胰高血糖素水平，同时降低 2 型糖尿病患者的血糖和胰岛素水平。^{10, 12} Dapagliflozin 已被证明也能刺激胰腺 α 细胞分泌胰高血糖素。¹³由于胰高血糖素在禁食条件下促进肝脏生酮并作为胰岛素的拮抗剂，¹⁴胰高血糖素与胰岛素的比例增加被认为会增加 SGLT2i 治疗的糖尿病患者发生正常血糖酮症酸中毒的风险。然而，一些论文表明，胰高血糖素信号传导的中断不会影响 SGLT2i 诱导的小鼠酮症。¹⁵因此，胰高血糖素对 SGLT2i 治疗的 1 型糖尿病患者正常血糖酮症酸中毒的作用仍不清楚。

此外，胰高血糖素对人体有多种代谢作用；它不仅促进糖异生，而且影响脂质代谢和能量消耗。¹⁶胰高血糖素增加生糖氨基酸，如丙氨酸，进而刺激胰高血糖素分泌。¹⁷胰高血糖素的这些潜在生物学效应可以通过刺激糖异生来抵消 SGLT2i 在糖尿病患者中的临床有效性，或通过脂肪分解和产热特性增强其心脏代谢作用。¹⁴然而，尚不清楚在 1 型糖尿病患者中使用 SGLT2i 治疗是否真的增加了血浆胰高血糖素水平，如果是，哪些临床参数与胰高血糖素的增加有关。因此，在这篇评论中，我们想根据我们最近的临床研究讨论胰高血糖素对 SGLT2i 治疗的 1 型糖尿病患者的利弊。

已在昭和大学医院接受强化胰岛素治疗的 9 名 1 型糖尿病患者（2 男 7 女；平均年龄 50.6 ± 7.3 岁）被纳入本研究。本研究的排除标准如下：(i) 接受除胰岛素以外的降糖药治疗的患者，(ii) 低碳水化合物饮食，(iii) 有糖尿病酮症酸中毒史或在过去 6 年中需要他人帮助的严重低血糖病史。个月，(iv) 怀孕和 (v) 慢性肾病（估计肾小球滤过率 (eGFR) < 30 mL/min/1.73 m ²）。在研究期间，患者被告知不要改变他们的生活习惯，并继续服用相同剂量的任何伴随药物，胰岛素除外。根据需要调整胰岛素剂量以避免低血糖。在基线和 ipragliflozin 治疗后 24 周，禁食后抽血测定血液生化，同时评估身体成分。所有参与者均知情同意参与本研究。研究方案经日本东京昭和大学医学院伦理委员会批准（许可编号：3152）。

胰高血糖素用夹心酶联免疫吸附试验（Mercodia AB）测量，而酮体用酶放大试验（KAINOS Laboratories Inc.）测量。用电化学发光免疫测定法（Roche，Basel，Switzerland）评估 N 端脑钠肽前体（NT-pro-BNP）。如前所述测量其他血液化学。¹⁸全身肌肉和脂肪块使用 InBody770（InBody770；Biospace）通过生物电阻抗分析进行评估。数据表示为平均值±标准偏差。使用Student's t检验或非参数Mann-Whitney U检验进行组间比较。使用 Spearman 等级相关系数确定非参数相关性。

参加者的临床特征及基线和依格列净治疗后24周临床参数变化见表1。治疗24周后，HbA1c、体重、体重指数、体脂、腰部情况和碱性磷酸酶均显着降低，而血红蛋白和血细胞比容水平升高。总酮体有增加的趋势。依格列净治疗使总胰岛素、基础胰岛素和推注胰岛素剂量适度但不显着降低。平均胰高血糖素水平从 25.8 ± 9.1 显着增加到 42.1 ± 13.4 pg/ml ( p = 0.0046)。依格列净治疗后胰高血糖素的变化（Δglucagon）与Δ低密度脂蛋白胆固醇（LDL-C）和ΔeGFR呈负相关（r = -0.72，p = 0.03 和r = -0.75，p = 0.02，分别）（表 2）。此外，基线体重指数和体脂肪量与 Δglucagon 显着相关（表 3）。虽然部分患者出现轻度低血糖，但未出现尿路和生殖器感染、严重低血糖或糖尿病酮症酸中毒的病例。

据我们所知，这是第一份证明 24 周 ipragliflozin 治疗显着增加血浆胰高血糖素水平的报告，与 1 型糖尿病患者的 HbA1c、体重指数和体脂肪量降低有关。此外，尽管 ipragliflozin 治疗并未提醒平均 LDL-C 水平，但 Δglucagon 与 ΔLDL-C 水平呈负相关，后者与基线体重指数和体脂肪量显着相关。目前的研究结果表明，SGLT2i 诱导的血浆胰高血糖素水平升高可能与 1 型糖尿病患者的 LDL-C 降低有关，尤其是那些体重指数高和体脂量增加的患者。因此，目前的研究结果表明，胰高血糖素可能是接受 SGLT2i 的 1 型糖尿病患者的朋友。

在动物模型和人类中，胰高血糖素已被证明可以通过各种机制改变脂质代谢并降低 LDL-C 水平。事实上，据报道给大鼠施用胰高血糖素可降低 LDL-C 水平。^{19, 20}一些胰高血糖素受体拮抗剂已显示出剂量依赖性地增加 2 型糖尿病患者的 LDL-C 水平，部分原因是增加胆固醇吸收。^21-23此外，Spolitu 等人。最近发现抑制肝胰高血糖素信号会增加前蛋白转化酶枯草杆菌蛋白酶 / kexin 9 型 (PCSK9) 的表达，从而通过减少 LDL 受体增加小鼠的 LDL-C 水平。²⁴肥胖与 PCSK9 表达增加相关的事实^{25, 26}可以部分解释为什么在具有高体重指数和增加体脂量的 1 型糖尿病患者中观察到 ipragliflozin 诱导的 LDL-C 水平降低。换句话说，SGLT2i 诱导的胰高血糖素升高可能部分通过抑制这些受试者中的 PCSK9 来降低 LDL-C 水平。

在本研究中未观察到糖尿病酮症酸中毒病例。此外，尽管使用 ipragliflozin 治疗后循环总酮体水平适度增加，但 Δglucagon 和 Δtotal 酮体之间没有相关性。在这项研究中，当开始 SGLT2i 治疗时，谨慎减少餐食（推注）而不是基础胰岛素剂量以避免低血糖。本研究排除了遵循低碳水化合物饮食的患者。由于基础胰岛素减少超过 10%–20% 和/或低碳水化合物饮食已被证明是与糖尿病 SGLT2i 治疗相关的酮症酸中毒的危险因素，²⁷减少酮症酸中毒危险因素的策略可以提高这种救命药物在 1 型糖尿病中的疗效和安全性。在任何情况下，都需要进一步的临床研究来阐明胰高血糖素在接受 SGLT2i 的 1 型糖尿病患者中的病理生理作用。我们现在计划在一项多中心合作试验中探索胰高血糖素在代谢参数中的作用，包括血糖变异性和 LDL-C 在接受 ipragliflozin 的 1 型糖尿病患者中的作用，该试验已在 UMIN 临床试验注册中心（大学医学信息网络 000020156）注册）。