Antihyperglycemic Medication Use Among U.S. Adults With Hemodialysis-Dependent Kidney Disease and Comorbid Diabetes

The 558,000 individuals in the U.S. receiving maintenance dialysis experience tremendous mortality, with a median survival of ∼48 months (1). Comorbid diabetes is present in >60% of incident cases of dialysis-dependent kidney failure and is associated with higher mortality rates, primarily driven by cardiovascular causes (1,2). Renal impairment limits the use of many antihyperglycemic agents, rendering insulin the mainstay diabetes treatment for most patients receiving dialysis. However, insulin-induced hypoglycemia is common due to autonomic neuropathy, hypoglycemia unawareness, and impaired renal gluconeogenesis (3). Management of glycemia is therefore challenging, and frequent oscillations between hyper- and hypoglycemia contribute to poor cardiovascular outcomes in patients with dialysis-dependent kidney failure (2,4).

Glucagon-like peptide 1 receptor agonists (GLP-1 RAs) improve glycemic control without causing hypoglycemia and confer cardiovascular risk reduction on people with type 2 diabetes. Some GLP-1 RAs do not require dose adjustments for kidney insufficiency; however, no clinical trials in the U.S. have evaluated GLP-1 RA safety, tolerability, or cardiovascular efficacy in the hemodialysis population, a population with a cardiovascular mortality rate exceeding that of the general population five- to sevenfold (1). We undertook this study to describe the utilization patterns of antihyperglycemic agents, including GLP-1 RAs, among U.S. patients with hemodialysis-dependent kidney failure.

Using 2012–2017 data from the U.S. Renal Data System, a claims-based national registry that captures the vast majority of U.S. dialysis patients, we identified adults with diabetes receiving in-center hemodialysis at the start of each calendar quarter who had continuous Medicare coverage during the preceding 180 days. During each calendar quarter, we used Medicare Part D prescription drug claims to determine the percentage of patients with prescription fills for antihyperglycemic agents recommended for use in dialysis-dependent kidney failure by medication class and, separately, stratified by concomitant insulin fill status. We also described the clinical characteristics of the most contemporary (October 2017) quarterly cohort. This study was approved by the University of North Carolina at Chapel Hill Institutional Review Board (no. 18-0297). A waiver of consent was granted.

Insulin was the most prescribed antihyperglycemic agent, followed by sulfonylureas (SUs), dipeptidyl peptidase 4 inhibitors (DPP-4is), and thiazolidinediones, in October 2017 (Fig. 1A). From 2012 to 2017, DPP-4i use rose, GLP-1 RA use rose minimally, and SU use fell. Less than 5% of patients received combination therapy of insulin plus a second agent. DPP-4is and SUs were the most common agents used with insulin (Fig. 1B). Only 752 individuals (0.6%) filled a prescription for GLP-1 RAs in October 2017; 66% of those individuals also filled a prescription for insulin.

Disregarding kidney disease as an atherosclerotic cardiovascular disease (ASCVD) risk factor, 64% of study patients met American Diabetes Association (ADA) ASCVD criteria, and 98% were at high risk for ASCVD regardless of antihyperglycemic treatment class (Table 1). Compared with people prescribed insulin alone, individuals prescribed GLP-1 RAs without insulin were younger, more likely to be obese and non-White, and had fewer claims for hypoglycemia and cardiovascular hospitalizations (Table 1). Of the full patient cohort, only 9.5% of patients had an endocrinology claim in the prior 6 months.

While our study does have limitations (e.g., lack of data beyond 2017 and use of Medicare claims prescription data, which lack definitive adherence information), our findings show that despite increasing awareness of the untoward effects of hypoglycemia, insulin and SUs were the antihyperglycemic agents most prescribed to hemodialysis patients in 2017. Prescription of combination therapy with SU and insulin was three times more common than prescription of GLP-1 RAs, agents with no hypoglycemia risk in the general population (5). Although GLP-1 RAs have yet to be rigorously tested among individuals receiving dialysis, GLP-1 RAs have the potential to confer glycemic control without increasing hypoglycemia risk in these vulnerable patients who are at risk for hypoglycemia independent of diabetes status (2). Finally, according to the ADA Standards of Medical Care in Diabetes, 100% of people with type 2 diabetes treated with dialysis should receive GLP-1 RA therapy (5). Lack of clinical trials of GLP-1 RAs in the dialysis-dependent kidney disease population is a barrier to implementation of this potentially high-impact intervention.

In addition, our study emphasizes the need for collaboration between endocrinologists and nephrologists. We found that few individuals with dialysis-dependent kidney failure and comorbid diabetes had a recent endocrinology claim. This finding may relate to patient reluctance to engage in care outside of their burdensome thrice-weekly hemodialysis treatments and/or nephrologist discomfort with newer antihyperglycemic agents and may contribute to low use of GLP-1 RAs among hemodialysis patients. Optimizing diabetes care with newer agents, like GLP-1 RAs, that confer benefits beyond glycemia could improve both patient-reported and biomedical outcomes for individuals with dialysis-dependent kidney failure and represents an opportunity for interdisciplinary collaborations.