Acute and chronic organ damage are key drivers of early mortality in adults with sickle cell disease (SCD). Chronic kidney disease (CKD) occurs in approximately half of adults with SCD and a rapid decline in estimated glomerular filtration rate (eGFR) portends an increased risk of progression to end-stage kidney disease (ESKD) and mortality.¹ Although renin-angiotensin-aldosterone system (RAAS) blocking agents are commonly used to treat SCD nephropathy, evidence for their effectiveness is limited to short-term studies with mixed results on preserving kidney function in patients with SCD.¹

Glucagon-like peptide-1 (GLP-1) agonists and sodium-glucose cotransporter-2 (SGLT-2) inhibitors are two classes of medications for treating type 2 diabetes mellitus (DM). GLP-1 agonists promote glucose homeostasis by stimulating insulin release from pancreatic beta cells via the incretin effect. SGLT-2 inhibitors improve glycemic control by preventing the reabsorption of filtered glucose by the proximal tubules in the kidneys. In addition to improving glucose control, large-scale clinical trials have demonstrated that these agents may improve adverse cardiovascular and kidney outcomes, such as a reduction in risk of cardiovascular death or progression to ESKD.² The benefits of GLP-1 agonists and SGLT-2 inhibitors on kidney function are independent of their blood glucose lowering effects in patients with DM and CKD, suggesting alternative reno-protective mechanisms for these therapies.²

The efficacy of GLP-1 agonists and SGLT-2 inhibitors on kidney function and their safety on SCD-related complications, such as vaso-occlusive pain episodes (VOEs), have not been studied in adults with SCD and DM. To address this, we retrospectively identified adults with SCD treated at the University of Illinois Chicago, Sickle Cell Center who were prescribed GLP-1 agonists or SGLT-2 inhibitors as an outpatient.

Clinical and laboratory data were obtained from the electronic medical charting system, Epic System (Madison, WI). Data before treatment was defined as prior to the initial start date of GLP-1 agonists or SGLT-2 inhibitors. Data during treatment was defined as from the initial medication start date until date of treatment discontinuation, if applicable. Random serum glucose and creatinine concentrations were obtained from basic metabolic panels drawn as part of outpatient routine care. For each creatinine concentration drawn, we used the 2021 creatinine-based Chronic Kidney Disease Epidemiology Collaboration (CKD-EPI) equation to calculate eGFR. VOEs were defined as pain episodes requiring care in the emergency room, acute care center, or inpatient hospital setting. Baseline eGFR values were calculated by the means of each patient's first 3 available eGFR measurements. For one patient who was treated with both a GLP-1 agonist and SGLT-2 inhibitor, analysis was independently performed for each individual medication's treatment window.

We calculated eGFR slope with linear mixed-effects models.^{3, 4} The fixed effects were population-level baseline eGFR and time. The random effects were intercepts and slopes grouped by individual, representing patient-level variability in baseline eGFR and eGFR slopes, respectively. Separate mixed effects models were generated for patients before and during treatment for GLP-1 agonists and SGLT-2 inhibitors. From these mixed effects models, we used the Best Linear Unbiased Prediction approach to estimate individual patients' eGFR slopes. The treatment effect of the GLP-1 agonists and SGLT-2 inhibitors were compared before versus during treatment using the Wilcoxon rank-sum test. Analyses were conducted using R version 4.3.1.

Between February 2016 and April 2023, seven patients with SCD were treated with either a GLP-1 agonist or SGLT-2 inhibitor at our center (Table 1). The median age at the time of treatment initiation was 55 years (interquartile range [IQR], 51–57 years). Five patients were female, and all seven patients had hemoglobin SS disease. Six patients were concurrently treated with a RAAS blocking agent and three with hydroxyurea. The median baseline eGFR before treatment was 76 (IQR, 63–85) mL/min/1.73 m² and the median urine albumin-to-creatinine ratio was 141 (IQR, 48–190) mg/g creatinine.

Of five patients treated with a GLP-1 agonist, 2 received subcutaneous injections of liraglutide (both 1.8 mg weekly) and 3 received subcutaneous injections of dulaglutide (2 with 1.0 mg weekly, 1 with 1.5 mg weekly). The median duration of treatment was 21.8 (IQR, 12.0–33.7) months. The serum glucose concentrations improved from a median of 213 (IQR, 137–249) mg/dL before treatment to 152 (IQR, 140–169) mg/dL during treatment. The rate of VOE increased while on therapy from a median of 1.0 (IQR, 0.5–3.25) VOE/year prior to treatment to 2.1 (IQR, 1.8–4.5) VOE/year during treatment. The eGFR slope improved from a median of −2.37 (IQR, −3.43 to −1.18) mL/min/1.73 m² per year before treatment to +0.49 (IQR, −0.74 to +1.43) mL/min/1.73 m² per year during treatment. Four of five patients had an improvement in eGFR slope with a median change of +2.36 (IQR, +1.04 to +4.39) mL/min/1.73 m² per year while on GLP-1 agonist treatment (Supplementary Figure 1).

Three patients were treated with the SGLT-2 inhibitor, oral empagliflozin (2 with 25 mg daily, 1 with 10 mg daily). The median duration of treatment was 12.6 (IQR, 9.8–25.7) months. The serum glucose concentrations were relatively stable from a median of 124 (IQR, 109–179) mg/dL pre-treatment to 119 (IQR, 109–130) mg/dL during treatment. The rate of VOE while on the SGLT-2 inhibitor increased from a median of 3.0 (IQR, 1.5–6.0) VOE/year prior to treatment to 6.2 (IQR, 2.6–10.6) VOE/year during treatment. The eGFR slope improved from a median of −0.55 (IQR, −1.06 to +0.53) mL/min/1.73 m² per year before treatment to +4.70 (IQR, +4.45 to +4.74) mL/min/1.73 m² per year during treatment. All three patients had an improvement in eGFR slope, with a median change of +5.30 (IQR, +3.92 to +5.77) mL/min/1.73 m² per year while on SGLT-2 inhibitor treatment (Supplementary Figure 1).

DM has been increasingly recognized as a comorbidity in SCD, with some studies demonstrating similar prevalence of DM in people with SCD (16.5%) compared to African Americans from the National Health and Nutritional Examination Survey (18.2%).⁵ Furthermore, CKD is a common complication that leads to increased morbidity and early mortality, and therapies that preserve kidney function are urgently needed. We observed improvements in glucose control and in the rate of eGFR decline in 4 of 5 patients during treatment with a GLP-1 agonist. Although the SGLT-2 inhibitors led to relatively stable glucose values, the rate of eGFR decline improved in all 3 patients while on treatment. Based upon our results in a small cohort of SCD patients with DM, the effects of these newer agents on kidney function should be further investigated.

Notably, the rate of VOE increased while on therapy with GLP-1 agonists or SGLT-2 inhibitors. One potential explanation for this observation may be due to dehydration from reduced water intake with GLP-1 agonists and from natriuresis and glucosuria with SGLT-2 inhibitors. The impact of these therapies on VOE will need to be closely studied and it may be important to educate patients on hydration and avoiding concurrent use of diuretics when starting GLP-1 agonists or SGLT-2 inhibitors.

The GLP-1 agonists and SGLT-2 inhibitors have demonstrated consistent benefits in improving glycemic control and cardiorenal outcomes in the general population, leading to a paradigm shift in the management of DM. In large randomized clinical trials, GLP-1 agonists reduced the rate of annual eGFR decline from −3.3 mL/min/1.73 m² pre-treatment to −0.7 mL/min/1.73 m² on treatment and reduced the risk of CKD progression by 22%, defined as a composite of persistent new-onset macroalbuminuria, persistent doubling of serum creatinine, ESKD, or kidney-related mortality.² Similarly, the SGLT-2 inhibitors reduce the risk of a doubling of serum creatinine by 44% and the risk of progressing to ESKD by 55%.² The benefits of the SGLT-2 inhibitors on kidney function appear to be independent of glycemic control since subgroup analyses do not demonstrate differences by baseline hemoglobin A1c or changes in hemoglobin A1c. Furthermore, in a randomized study of 6609 subjects with a wide range of etiologies for CKD (54% without DM), SGLT-2 inhibitors reduced the risk of CKD progression by 29% and improved the annual rate of eGFR decline by +0.75 mL/min/1.73 m².⁶ Potential reno-protective mechanisms for the effects of GLP-1 agonists and SGLT-2 inhibitors beyond better glycemic control include reducing tubular work and oxygen requirements, decreasing cAMP and protein kinase A-mediated oxidative stress and renal inflammation, and activating the tubulo-glomerular feedback pathway.² Renal tubular stress and tubulo-glomerular feedback pathways have also been implicated in SCD-related CKD; GLP-1 agonists and SGLT-2 inhibitors may target these pathways to preserve kidney function in sickle cell nephropathy.¹

There are several limitations to our study including the small sample size, the use of different drugs with varied routes of delivery, and the retrospective nature of our observations limiting our ability to infer causality or differentiate the effects of these agents from concurrent RAAS blocking agents or hydroxyurea on eGFR decline. The study period included the years 2020–2021, representing the height of the COVID-19 pandemic when VOE frequency may have changed.

Based on our preliminary findings of an improvement in eGFR slope in 4 of 5 patients during treatment with GLP-1 agonists and all 3 patients during treatment with SGLT-2 inhibitors, these new classes of therapies may provide reno-protective benefits in those with sickle cell nephropathy. Future prospective studies will be needed to confirm the benefits on kidney function as well as the potential risks on SCD-related complications such as VOE.

急性和慢性器官损伤是镰状细胞病 (SCD) 成人早期死亡的关键驱动因素。大约一半患有 SCD 的成年人患有慢性肾病 (CKD)，估计肾小球滤过率 (eGFR) 的快速下降预示着进展为终末期肾病 (ESKD) 和死亡的风险增加。¹虽然肾素-血管紧张素-醛固酮系统 (RAAS) 阻断剂常用于治疗 SCD 肾病，但其有效性的证据仅限于短期研究，在保护 SCD 患者肾功能方面的结果不一。¹

胰高血糖素样肽-1 (GLP-1) 激动剂和钠-葡萄糖协同转运蛋白-2 (SGLT-2) 抑制剂是治疗 2 型糖尿病 (DM) 的两类药物。 GLP-1 激动剂通过肠促胰素效应刺激胰腺 β 细胞释放胰岛素，从而促进葡萄糖稳态。 SGLT-2 抑制剂通过防止肾脏近端小管重吸收滤过的葡萄糖来改善血糖控制。除了改善血糖控制外，大规模临床试验还表明，这些药物可以改善心血管和肾脏的不良结局，例如降低心血管死亡或进展为 ESKD 的风险。² GLP-1 激动剂和 SGLT-2 抑制剂对肾功能的益处与其对 DM 和 CKD 患者的血糖降低作用无关，这表明这些疗法有替代的肾脏保护机制。²

GLP-1 激动剂和 SGLT-2 抑制剂对肾功能的疗效及其对 SCD 相关并发症（如血管闭塞性疼痛发作 (VOE)）的安全性尚未在成人 SCD 和 DM 中进行研究。为了解决这个问题，我们回顾性地确定了在伊利诺伊大学芝加哥分校镰状细胞中心接受 GLP-1 激动剂或 SGLT-2 抑制剂门诊治疗的成人 SCD。

临床和实验室数据从电子医疗图表系统 Epic System（麦迪逊，威斯康星州）获得。治疗前的数据定义为 GLP-1 激动剂或 SGLT-2 抑制剂初始开始日期之前。治疗期间的数据定义为从初始用药开始日期到治疗停止日期（如果适用）。随机血清葡萄糖和肌酐浓度是从作为门诊常规护理的一部分抽取的基本代谢组中获得的。对于每个肌酐浓度，我们使用 2021 年基于肌酐的慢性肾脏病流行病学协作 (CKD-EPI) 方程来计算 eGFR。 VOE 被定义为需要在急诊室、急症护理中心或住院医院进行护理的疼痛发作。基线 eGFR 值通过每位患者的前 3 次可用 eGFR 测量值计算得出。对于同时接受 GLP-1 激动剂和 SGLT-2 抑制剂治疗的一名患者，对每种药物的治疗窗口进行了独立分析。

我们使用线性混合效应模型计算了 eGFR 斜率。^{3, 4}固定效应是人群水平基线 eGFR 和时间。随机效应是按个体分组的截距和斜率，分别代表基线 eGFR 和 eGFR 斜率的患者水平变异性。在 GLP-1 激动剂和 SGLT-2 抑制剂治疗之前和治疗期间为患者生成单独的混合效应模型。根据这些混合效应模型，我们使用最佳线性无偏预测方法来估计个体患者的 eGFR 斜率。使用 Wilcoxon 秩和检验比较治疗前和治疗期间 GLP-1 激动剂和 SGLT-2 抑制剂的治疗效果。使用 R 版本 4.3.1 进行分析。

2016 年 2 月至 2023 年 4 月期间，7 名 SCD 患者在我们中心接受了 GLP-1 激动剂或 SGLT-2 抑制剂治疗（表 1）。开始治疗时的中位年龄为 55 岁（四分位数范围 [IQR]，51-57 岁）。 5 名患者为女性，7 名患者均患有血红蛋白 SS 病。六名患者同时接受 RAAS 阻断剂治疗，三名患者接受羟基脲治疗。治疗前的中位基线 eGFR 为 76 (IQR，63–85) mL/min/1.73 m ²，中位尿白蛋白与肌酐比值为 141 (IQR，48–190) mg/g 肌酐。

在接受 GLP-1 激动剂治疗的 5 名患者中，2 名接受利拉鲁肽皮下注射（均为每周 1.8 mg），3 名接受度拉鲁肽皮下注射（2 名每周 1.0 mg，1 名每周 1.5 mg）。中位治疗持续时间为 21.8（IQR，12.0-33.7）个月。血清葡萄糖浓度从治疗前的中位数 213（IQR，137-249）mg/dL 改善至治疗期间的 152（IQR，140-169）mg/dL。治疗期间 VOE 发生率从治疗前中位数 1.0（IQR，0.5-3.25）VOE/年增加至治疗期间 2.1（IQR，1.8-4.5）VOE/年。 eGFR 斜率从治疗前每年中位数 -2.37（IQR，-3.43 至 -1.18）mL/min/1.73 m ²改善至每年+0.49（IQR，-0.74 至 +1.43）mL/min/1.73 m ²治疗期间一年。在接受 GLP-1 激动剂治疗时，五名患者中有四名 eGFR 斜率有所改善，中位变化为每年 +2.36（IQR，+1.04 至 +4.39）mL/min/1.73 m ^{2 （补充图 1）。}

3 名患者接受 SGLT-2 抑制剂口服恩格列净治疗（2 名患者每天 25 mg，1 名患者每天 10 mg）。中位治疗持续时间为 12.6（IQR，9.8-25.7）个月。血清葡萄糖浓度相对稳定，从治疗前的中位数 124（IQR，109-179）mg/dL 到治疗期间的 119（IQR，109-130）mg/dL。服用 SGLT-2 抑制剂时的 VOE 发生率从治疗前的中位值 3.0（IQR，1.5-6.0）VOE/年增加到治疗期间的 6.2（IQR，2.6-10.6）VOE/年。 eGFR 斜率从治疗前每年中位数 -0.55（IQR，-1.06 至 +0.53）mL/min/1.73 m ²改善至每年+4.70（IQR，+4.45 至 +4.74）mL/min/1.73 m ²治疗期间一年。所有三名患者的 eGFR 斜率均有所改善，在接受 SGLT-2 抑制剂治疗时，每年的中位变化为 +5.30（IQR，+3.92 至 +5.77）mL/min/1.73 m ^{2 （补充图 1）。}

糖尿病越来越被认为是 SCD 的合并症，一些研究表明，国家健康和营养检查调查显示，SCD 患者的 DM 患病率 (16.5%) 与非裔美国人 (18.2%) 相似。⁵此外，CKD 是一种常见并发症，会导致发病率和早期死亡率增加，迫切需要保留肾功能的治疗方法。我们观察到，在接受 GLP-1 激动剂治疗期间，5 名患者中有 4 名的血糖控制有所改善，eGFR 下降率有所改善。尽管 SGLT-2 抑制剂导致血糖值相对稳定，但所有 3 名患者在治疗期间 eGFR 下降率均有所改善。根据我们对一小群 SCD 合并 DM 患者的研究结果，应进一步研究这些新药对肾功能的影响。

值得注意的是，在接受 GLP-1 激动剂或 SGLT-2 抑制剂治疗时，VOE 发生率有所增加。对于这一观察结果的一个可能的解释可能是由于 GLP-1 激动剂导致的水摄入量减少以及 SGLT-2 抑制剂引起的尿钠排泄和糖尿导致的脱水。需要仔细研究这些疗法对 VOE 的影响，并且在开始使用 GLP-1 激动剂或 SGLT-2 抑制剂时，对患者进行水合教育并避免同时使用利尿剂可能很重要。

GLP-1 激动剂和 SGLT-2 抑制剂在改善普通人群的血糖控制和心肾结局方面表现出一致的益处，从而导致糖尿病治疗的范式转变。在大型随机临床试验中，GLP-1激动剂将年eGFR下降率从治疗前的-3.3 mL/min/1.73 m ^{2降低至治疗后的-0.7 mL/min/1.73 m 2}，并降低了CKD进展的风险22%，定义为持续新发大量白蛋白尿、血清肌酐持续加倍、ESKD 或肾脏相关死亡率的综合结果。²同样，SGLT-2 抑制剂可将血清肌酐加倍的风险降低 44%，并将进展为 ESKD 的风险降低 55%。² SGLT-2 抑制剂对肾功能的益处似乎与血糖控制无关，因为亚组分析未显示基线血红蛋白 A1c 或血红蛋白 A1c 变化的差异。此外，在一项针对 6609 名具有多种 CKD 病因的受试者（54% 无糖尿病）的随机研究中，SGLT-2 抑制剂将 CKD 进展的风险降低了 29%，并将 eGFR 年下降率提高了 +0.75 mL/分钟/1.73 m ²。⁶除了更好的血糖控制之外，GLP-1 激动剂和 SGLT-2 抑制剂的潜在肾脏保护机制包括减少肾小管工作和氧气需求、减少 cAMP 和蛋白激酶 A 介导的氧化应激和肾脏炎症，以及激活肾小管-肾小球反馈通路。²肾小管应激和肾小管-肾小球反馈通路也与 SCD 相关的 CKD 有关； GLP-1 激动剂和 SGLT-2 抑制剂可能会针对这些途径来保护镰状细胞肾病的肾功能。¹

我们的研究存在一些局限性，包括样本量小、使用不同给药途径的不同药物以及我们观察的回顾性，限制了我们推断因果关系或区分这些药物与同时使用的 RAAS 阻断剂或药物的作用的能力。羟基脲对 eGFR 下降的影响。研究期间包括 2020 年至 2021 年，这段时间代表了 COVID-19 大流行的高峰期，此时 VOE 频率可能会发生变化。

根据我们的初步发现，5 名患者中的 4 名在使用 GLP-1 激动剂治疗期间以及所有 3 名患者在使用 SGLT-2 抑制剂治疗期间 eGFR 斜率有所改善，这些新类别的疗法可能为镰状细胞病患者提供肾保护益处细胞肾病。未来需要进行前瞻性研究来确认对肾功能的益处以及 SCD 相关并发症（如 VOE）的潜在风险。