GLP-1 receptor agonists in diabetic kidney disease: from the patient-side to the bench-side

Abstract

Diabetic kidney disease (DKD), one of the most common and severe microvascular complications of diabetes, is the leading cause of chronic kidney disease and end-stage kidney disease worldwide. Since the development of renin-angiotensin system inhibition nearly three decades ago, no new therapeutic agents have received regulatory approval for treatment of DKD. Glucagon-like peptide-1 (GLP-1) receptor agonists, a class of newer antihyperglycemic agents, have shown promise for prevention of DKD onset and progression. This perspective summarizes clinical and experimental observations to give insight into biological mechanisms beyond glycemic control, such as natriuresis and anti-inflammatory actions, for preservation of kidney function in patients with diabetes.

INTRODUCTION

Diabetic kidney disease (DKD) afflicts ~30% of patients with type 1 diabetes and 40% of patients with type 2 diabetes (31). This microvascular complication of diabetes now stands as the leading cause of chronic kidney disease and end-stage kidney disease in the world (5, 38). The relentless increase in DKD prevalence is largely attributable to continued growth of the diabetic population, which is projected to reach a size of 624 million people globally by the year 2040 (13). Almost all of the excess risk for all-cause and cardiovascular-related mortality among patients with diabetes occurs in those with DKD (1). The age-standardized rate of end-stage kidney disease has exhibited a nominal decline, just 28%, compared with ~50–70% declines for other diabetes complications (myocardial infarction, stroke, and lower limb amputation) between the years 1990 and 2012 (12). Notably, the number of deaths attributed to end stage kidney disease from diabetes increased by 94% over the same time period. This dramatic increase in mortality is presumably due to the rapidly expanding number of patients living with diabetes as well as a lack of effective therapies for DKD (19).

New treatments for prevention and treatment of DKD represent a pressing unmet medical need. Since the development of renin-angiotensin system inhibition nearly three decades ago, no new therapeutic agents have received regulatory approval for DKD treatment. Recently, several clinical trials have reported that a class of newer antihyperglycemic agents, the glucagon like peptide-1 (GLP-1) receptor agonists, prevent onset of macroalbuminuria and reduce decline in estimated glomerular filtration rate (eGFR) in patients with type 2 diabetes (21–23, 26). This Perspective describes the clinical trial data that provided these observations, discusses posited mechanisms for protection of the diabetic kidney by GLP-1 receptor agonists, and outlines important unanswered questions (8, 14, 40).

FROM THE PATIENT-SIDE: CLINICAL EFFECTS ON THE DIABETIC KIDNEY

Clinical benefits of the GLP-1 receptor agonists on DKD were identified in data resulting from clinical trials for cardiovascular safety that followed U.S. Food and Drug Administration approvals for treatment of hyperglycemia in type 2 diabetes. These studies showed that significantly fewer participants treated with semaglutide, liraglutide, or lixisenatide developed new or worsening DKD compared with those receiving placebo (22, 23, 26). This effect was largely driven by reducing risk of onset of new macroalbuminuria (urine albumin-to-creatinine ratio >300 mg/g) by 22–36% (21–23, 26). For liraglutide, the reduction in risk of death and major cardiovascular events, primarily atherosclerotic complications, was even greater in the subset of participants with eGFR <60 ml·min⁻¹·1.73 m⁻² than in the overall study population (23). However, kidney disease outcomes in these trials were not stratified by level of kidney function (22, 26).

A recently completed clinical trial of patients with type 2 diabetes and moderate-to-severe chronic kidney disease (CKD) revealed that treatment with dulaglutide, compared with active treatment with insulin glargine as basal therapy, prevented decline in eGFR over 1 yr from a mean of loss of approximately −3 to −1 ml·min⁻¹·1.73 m⁻², respectively (Table 1; Fig. 1A). This difference, if sustained over time, would have a substantial clinical impact to delay kidney disease progression. Notably, all groups achieved similar reductions in hemoglobin A1C (HbA1c) and blood pressure was comparable between treatment groups, supporting the postulate that GLP-1 receptor agonists have a benefit on kidney function in people with diabetes and advanced CKD independent of benefits on glycemia or blood pressure.

Table 1.

Secondary kidney outcomes in clinical trials of glucagon like peptide-1 receptor agonists in patients with type 2 diabetes

Name of the Study	Agent	Intervention	Study Population	Kidney Outcomes	Results
Clinical trial for glycemic control in moderate-to-severe chronic kidney disease
AWARD-7* (n = 576)	Dulaglutide	Dulaglutide 0.75 and 1.5 mg vs. insulin glargine	Type 2 diabetes7.5% ≥ HbA1c ≤ 10.5%15 ≥ eGFR ≤ 6 0 ml·min−1·1.73 m−2Baseline means ± SD eGFR: 35 ± 0.6 ml·min−1·1.73 m−2Baseline median (interquartile range): UACR (mg/g): 209 (39–965)	eGFR and UACR change from baseline	eGFR in dulaglutide 1.5 mg (−1.1 ml·min−1·1.73 m−2, P < 0.05), and 0.75 mg (−1.5 ml·min−1·1.73 m−2 m2, P < 0.05) groups vs. eGFR in insulin glargine group (−2.9 ml·min−1·1.73 m−2 V, P < 0.0001)
					In those with UACR >300 mg/g, eGFR decline was less with dulaglutide (1.5 mg: −0.5 ml·min−1·1.73 m−2, P < 0.05 vs. insulin; 0.75 mg: −0.7 ml·min−1·1.73 m−2, P < 0.05 vs. insulin; insulin: −5.5 ml·min−1·1.73 m−2)
					UACR reduction in dulaglutide 1.5 mg −29% (−34, −11.5), and −12.3% (−29,8.5) in 0.75 mg vs. insulin −13% (27.1, −39); P = 0.020 and P = 0.363, respectively)
Clinical trials for cardiovascular safety
LEADER† (n = 9,340)	Liraglutide	1.8 mg (or the maximum tolerated dose) vs. placebo	Type 2 diabetes	New-onset albuminuria, doubling of sCr and CrCl <45 ml·min−1·1.73 m−2; RRT; death due to kidney disease	1.9 events/100 patient-year in placebo group vs. 1.5 events/100 patient-year in liraglutide group (P = 0.003)
			HbA1c >7%
			Prevalent CVD
			Baseline eGFR: n/N (%):
			eGFR >90 ml·min−1·1.73 m−2:
			placebo [1,655/4,672 (35.4)]; liraglutide [1,620/4,668 (34.7)]
			eGFR 60 ≤ 90 ml·min−1·1.73 m−2: placebo [1,975/4,672 (41.4)]; liraglutide [1,932/4,668 (42.3)]
			eGFR 30 ≤ 60 ml·min−1·1.73 m−2: placebo [935/4,672 (20.0)]; liraglutide [999/4,668 (21.4)]
			eGFR <30 ml·min−1·1.73 m−2:
			placebo [107/4,672 (2.3)];
			liraglutide [117/4,668 (2.5)]
			Microalbuminuria or proteinuria at baseline: n/N (%):
			placebo: 558/4,672 (11.9%);
			liraglutide: 501/4,668 (10.7%)
ELIXA‡ (n = 6,068)	Lixisenatide	10–20 μg of lixisenatide vs. placebo	Type 2 diabetes5.5% ≥ HbA1c ≤ 11%	Proportional change in UACR from baseline to 108 wk	24 vs. 34% reduction in UACR in placebo group vs. lixisenatide group (P = 0.004)
			Recent acute coronary syndrome
			Baseline means ± SD eGFR:
			placebo: 75.2 ± 21.4 ml·min−1·1.73 m−2; lixisenatide: 76.7 ± 21.3 ml·min−1·1.73 m−2
			Baseline median (interquartile range) UACR (mg/g):
			placebo: 10.4 (5.9–32.6); lixisenatide: 10.0 (6.0 −28.0)
SUSTAIN-6¶ (n = 3,297)	Semaglutide	0.5 mg vs. 1.0 mg of semaglutide vs. placebo	Type 2 diabetesHbA1c >7%	New or worsening nephropathy (persistent UACR >300 mg/g; doubling of sCr or eGFR <45 ml·min−1·1.73 m−2; RRT)	6.1% with composite outcome in placebo group vs. 3.8% in semaglutide group (P = 0.005)
			Age >50 yr with established CVD or CKD stages 3–5
			Age >60 yr with CVD risk factors
			Baseline eGFR n/N (%):
			eGFR >90: 990/3,297 (30)
			eGFR 60 ≤ 90: 1,368/3,297 (41.5)
			eGFR 30 ≤ 60: 832/3,297 (25.2)
			eGFR 15 ≤ 30: 95/3,297 (2.9)
			eGFR <15: 12/3,297 (0.4)

UAC, urine albumin-to-creatinine ratio (with albumin measured in mg/g); CKD, chronic kidney disease, CVD, cardiovascular disease; sCr, serum creatinine; CrCl, creatinine clearance; RRT, renal replacement therapy; HbA1c, hemoglobin A1c; eGFR, estimated glomerular filtration rate in ml·min⁻¹·1.73 m⁻².

^*A Randomized, Open-Label, Parallel-Arm Study Comparing the Effect of Once-weekly Dulaglutide With Insulin Glargine on Glycemic Control in Patients With Type 2 Diabetes and Moderate or Severe Chronic Kidney Disease (AWARD-7) (36).

^†Liraglutide Effect and Action in Diabetes: Evaluation of Cardiovascular Outcome Results (LEADER) (21, 22).

^‡Lixisenatide in Patients with Type 2 Diabetes and Acute Coronary Syndrome (ELIXA) (26).

^¶Semaglutide and Cardiovascular Outcomes in Patients with Type 2 Diabetes (SUSTAIN-6) (23).

Fig. 1.

Values estimated by chronic kidney disease epidemiology collaboration equation by cystatin C or creatinine. A: data presented as estimated glomerular filtration rate (eGFR) values by geometric least squares mean from log-transformed analysis. B: data presented as actual untransformed change from baseline in eGFR values [least squares mean, 95% confidence interval (CI)]. C: actual untransformed data macroalbuminuria status at baseline, with values presented as LSM (95% CI). *Versus baseline. †Versus insulin glargine. [Modified from Tuttle et al. (36) with permission from Elsevier.]

Nevertheless, GLP-1 receptor agonists may also provide indirect protection of the kidney by better control of hyperglycemia, hypertension, and excess body weight. First, addition of a GLP-1 receptor agonist to background antihyperglycemic therapy led to greater reductions in HbA1C, ranging from −0.3 to −1.9%, compared with control groups in the clinical trials for cardiovascular safety (20, 22, 23, 26). Second, GLP-1 receptor agonists also produce a reduction in blood pressure. In patients with type 2 diabetes and preserved kidney function, treatment with liraglutide, semaglutide, or dulaglutide lowered systolic blood pressure in the range of 2–5 mmHg compared with placebo or active antihyperglycemic comparators (2, 9, 35). A proposed mechanism for reduction in blood pressure is natriuresis. Greater absolute, fractional, and proximal tubular excretion of sodium have been observed in studies of acute GLP-1 receptor agonist administration in patients with type 2 diabetes and normal kidney function (33, 37). However, in diabetic patients with moderate-to-severe CKD, the blood pressure-lowering effect of GLP-1 receptor agonists appears absent or reduced, perhaps due to impaired natriuretic responsiveness (2, 11, 36, 41). Third, GLP-1 receptor agonists promote weight loss of 3 kg on average compared with other treatments for hyperglycemia in type 2 diabetes (27, 41). Reductions in waist circumference have also been observed with GLP-1 receptor agonist treatment (34). Moreover, the weight loss effect is present in patients with type 2 diabetes and CKD (7, 36). The precise mechanisms for weight loss remain to be determined but likely include central and peripheral neural pathways, reduction of appetite and increase in satiety by GLP-1 receptor activation (39).

TO THE BENCH-SIDE: UNRAVELING BIOLOGICAL MECHANISMS

Induction of natriuresis and anti-inflammatory mediators, effects downstream of GLP-1 receptor activation, may underpin protective mechanisms in the diabetic kidney. GLP-1 receptors are expressed in the kidney as well as in the pancreas, heart, intestine, lung, liver, thyroid, and vascular smooth muscle (28). The exact distribution of GLP-1 receptors in the parenchymal cells of the kidney is not well established and exhibits notable species differences (6, 16, 18, 28). In the Wistar rat, microdissection studies of the kidney have identified GLP-1 receptor mRNA in the glomerulus and proximal convoluted tubule (6). In a separate study of the Wistar rat, GLP-1 receptors were reported, via autoradiography for protein, in the afferent arteriole and in renin-producing cells of the macula densa but not in resident glomerular cells (16). There is immunohistochemical evidence for the GLP-1 receptor protein in the glomerulus of the db/db mouse model (C57BLKS/J-db/db) of type 2 diabetes (25). However, GLP-1 receptor protein in monkey kidneys was observed only in vascular smooth muscle cells and in the macula densa (28). Currently, there are very limited data on GLP-1 receptor distribution in human kidneys, although some early reports from mRNA expression and immunostaining indicate that they may be found in the kidney cortex and arterial walls (18, 28).

Treatment with a GLP-1 receptor agonist, exenatide, has been shown to stimulate production of cAMP and to activate the Na⁺/H⁺ exchanger 3 based on in vitro studies in a proximal tubular cell line (LLC-PK₁), thus inferring a potential mechanism for GLP-1-induced natriuresis (4). In subsequent in vivo studies, GLP-1 administration to spontaneously hypertensive rats led to natriuresis, increases in urinary cAMP (a measure cAMP activation in the kidney), and vasorelaxation (6, 30). Exenatide increased sodium excretion in C57BLKS/J db/db mice but not in a comparator group with GLP-1 receptor knockout (29). As a bridge to clinical translation, natriuretic effects of GLP-1 receptor agonists have been observed in human physiology studies. In healthy men, infusion of a synthetic GLP-1 (7–36) amide increased sodium clearance by ~40% with stable glomerular filtration rate (32). Similarly, liraglutide increased sodium clearance without changing glomerular filtration rate in patients with type 2 diabetes (33).

GLP-1 receptor agonists reduce markers of kidney-level inflammation in rodent models of diabetes and systemic inflammation in humans. In the kidney of the C57BKS-db/db type 2 diabetic mouse, treatment with exenatide for 4–8 wk reduced glomerular macrophage infiltration and produced a dose-related reduction in 24-h urine albumin excretion (25). In a streptozotocin-induced Sprague-Dawley rat model of type 1 diabetes, exendin-4 reduced albuminuria, mesangial matrix expansion, and expression of mRNA of inflammatory markers including cluster of differentiation 14 and intracellular adhesion molecule (17). As a clinical correlate to systemic inflammation, dulaglutide lowered serum levels of C-reactive protein (CRP) by ~1 mg/l in patients with type 2 diabetes (9). Addition of exenatide to metformin reduced levels of serum CRP by a mean of 0.5 mg/l in another study of type 2 diabetes (3). A meta-analysis of randomized controlled trials with GLP-1 receptor agonist treatment in patients with type 2 diabetes found an ~2 mg/l reduction in serum CRP (24).

The precise mechanisms underlying the anti-inflammatory effects of GLP-1 receptor agonists remain to be elucidated. One candidate mechanism is via reduction of oxidative stress. In the KK/Ta-Akita mouse model of type 1 diabetes, liraglutide treatment reduced albuminuria and nicotinamide adenine dinucleotide phosphate oxidase activity in the kidney without altering glycemic control (10). Protein kinase A activity and cAMP were also elevated in the kidney in response to liraglutide therapy. In a streptozotocin model of type 1 diabetes in rats, liraglutide attenuated oxidative stress, expression of transforming growth factor-β and fibronectin in the kidney, and albuminuria via protein kinase A-mediated inhibition of renal nicotinamide adenine dinucleotide phosphate oxidases (15). Corresponding in vitro experiments in mesangial cell culture suggested that this effect was mediated by increased production of protein kinase A and cAMP (15). A conceptual model of hypothesized mechanisms for activation of natriuretic and antioxidant mechanisms by GLP-1 receptor agonists is shown in Fig. 2.

Fig. 2.

Conceptual model of hypothesized mechanisms for activation of natriuretic and antioxidant mechanisms by glucagon-like peptide 1 (GLP-1) receptor agonists. GLP-1 receptor agonists inhibit Na⁺/H⁺ exchanger 3 (NHE3) via a protein kinase A (PKA) and cyclic adenosine monophosphate (cAMP) dependent pathway to induce natriuresis. Antioxidant effects of GLP-1 receptor agonists occur through activation of protein kinase A and inhibition of nicotinamide adenine dinucleotide phosphate (NADPH) oxidase.

CONCLUSIONS

DKD poses an accelerating public health threat worldwide. Progress toward better treatments to prevent development and progression of DKD are urgently needed. In patients with, or at risk of DKD, emerging evidence from both clinical trials and experimental models points to GLP-1 receptor agonists as promising agents with clinical efficacy and a favorable safety profile. However, to optimize therapeutic application of GLP-1 receptor agonists there are pivotal questions that remain to be addressed. Specific human cell types and kidney structures that contain GLP-1 receptors, as well as associated anti-inflammatory pathways and antioxidant mechanisms, need to be delineated for a more robust understanding of underlying mechanisms for albuminuria-lowering and preservation of kidney function. Other key, yet unanswered, questions about GLP-1 receptor agonists include precise mechanisms for blood pressure reduction, weight loss, and natriuretic effects across various stages of kidney disease severity.

GRANTS

This study was supported by National Institute of Diabetes and Digestive and Kidney Diseases Grants 5UM1-DK-100846-03, 1U54-DK-083912, and U2C-DK-114886 (to K. R. Tuttle); National Center for Advancing Translational Sciences Grant 4UL1TR000423-10/WESC8883 (to K. R. Tuttle); and Janssen Research and Development Grant.

DICLOSURES

No conflicts of interest, financial or otherwise, are declared by the authors.

AUTHOR CONTRIBUTIONS

B.P.D., R.Z.A., and K.R.T. prepared figures, drafted manuscript, edited and revised manuscript, and approved final version of manuscript.