AUTHORS
Jordan Nunnelee, MD, Yoshito Nishimura, M.D., PhD, MPH2, Richard C. Godby, MD3
AFFILIATIONS
1 Mayo Clinic, Rochester, Minnesota, United States
2 Mayo Clinic, Rochester, Minnesota, United States
3 Mayo Clinic, Rochester, Minnesota, United States
ABSTRACT
Introduction
Anemia of chronic kidney disease (CKD) is a common complication of CKD that requires significant healthcare utilization, often including erythropoiesis-stimulating agent (ESA) prescriptions, parenteral iron infusions, and packed red blood cell (PRBC) transfusions. Sodium–glucose cotransporter-2 inhibitors (SGLT2i) are FDA-approved for CKD with albuminuria, shown to slow progression of kidney disease. SGLT2i have demonstrated pleiotropic anti-inflammatory, immunomodulatory, and renal and other organ-protective effects beyond glycemic control. Moreover, SGLT2i increase endogenous erythropoietin (EPO) synthesis and hemoglobin levels without increased risk of myocardial infarction or venous thromboembolism. However, the potential impact of SGLT2i on survival, thrombosis rates, and anemia-specific factors in patients with advanced CKD with anemia has not been elucidated.
Methods
We conducted a retrospective cohort analysis using an institutional de-identified data exploration tool, identifying adult patients diagnosed with anemia of CKD (ICD-10 code D63.1) from March 2013 (first approval of SGLT2i) to February 2026. All patients included had exposure to an angiotensin-converting enzyme (ACE) inhibitor or angiotensin II receptor blocker (ARB). Patients with end-stage renal disease (ESRD) and dialysis dependence (ICD-10 codes N18.6 and Z99.2, respectively) were excluded. Patients were then stratified based on exposure to SGLT2i. The prespecified primary endpoint was ESA exposure rate, and the secondary endpoint was five-year overall survival (OS). Propensity-score matching (PSM) was performed for the following covariates: age at diagnosis, race, sex, and diagnoses of type 2 diabetes mellitus and congestive heart failure. Mutually exclusive 1:1 matching was used with a caliper of 0.2. Survival was estimated by the Kaplan–Meier method and compared using the log-rank test. Hazard ratios (HRs) were derived from Cox proportional-hazards models. Categorical outcomes were compared using the χ² test. A two-sided P value less than .05 was considered statistically significant.
Results
5,039 (12.5%) patients were exposed to SGLT2i and 35,426 (87.5%) were not exposed. There were 5,039 patients in each group after PSM. Baseline demographics in the groups were comparable with a median age of 74.1 years (interquartile range (IQR) 66.5-80.1) at first diagnosis match, and male and white race predominance (64.3% and 88.8%, respectively).
There were differences among CKD stage diagnoses among patients in the SGLT2i exposed vs. unexposed groups: 31.5% vs.18.8% diagnosed as CKD 3a; 32.5% vs. 17.9% as CKD 3b; 18.7% vs. 23.4% as CKD 4; and 1.3% vs. 8.4% as CKD 5 (P < .001 for all diagnoses).
Five-year OS was significantly improved among patients exposed to SGLT2i compared to those not exposed (69.7% vs 50.2%, HR 0.55, 95% CI: 0.51-0.57, P < .001). SGLT2i exposure was associated with significantly less ESA exposure (2.5% vs. 11.9%, P <.001), parenteral iron replacement (9.0% vs. 10.7%, P = .003), and PRBC transfusion (9.5% vs. 12.9%, P < .001). There were no differences in occurrence of pulmonary embolism (3.0% vs. 3.7%, P = .77), deep venous thrombosis (0.7% vs. 0.9%, P = .23), or ischemic stroke (1.8% vs. 1.7%, P = .59) between groups. Median hemoglobin was 9.9 g/dL (IQR 8.7-11.3 g/dL) vs. 9.4 g/dL (IQR 8.4-10.7 g/dL) between groups (Weighted Cohen’s D effect size of .91).
A sensitivity analysis was performed to investigate five-year OS among patients with stage 3 CKD and stage 4/5 CKD exposed and not exposed to SGLT2i. OS was significantly increased in patients with stage 3 CKD exposed to SGLT2i (69.6% vs. 60.2%, HR 0.74, 95% CI: 0.67-0.81, P < .001) and in patients with stage 4/5 CKD (70.4% vs. 50.3%, HR 0.53, 95% CI: 0.47-0.57, P < .001).
Conclusions
In this large real-world cohort of patients diagnosed with anemia of non-dialysis CKD, SGLT2i exposure was associated with improved five-year OS, reduced ESA, parenteral iron, and PRBC usage with similar rates of venous and arterial thrombosis. Median hemoglobin was increased in the SGLT2i-exposed group. These findings suggest anemia-specific benefits of SGLT2i therapy that mitigate ESA reliance and transfusion burdens. While causality cannot be inferred from this retrospective analysis, further investigations and prospective evaluations are warranted to determine the potential benefits of SGLT2i therapy as an anemia-targeted adjunctive treatment in anemia of CKD and beyond.
DISCLOSURES
Jordan Nunnelee, MD: No relevant conflicts of interest to declare.