(154) Comparative cardiovascular safety evaluation of long- vs short-acting erythropoiesis-stimulating agents (ESA) in chronic kidney disease patients: a meta-analysis and signal detection using the FDA Adverse Event Reporting System

Background: Erythropoiesis-stimulating agent (ESA) is a primary treatment for anemia in chronic kidney disease (CKD) patients, classified as long- or short-acting based on their duration of action. A recent meta-analysis suggests darbepoetin-α, epoetin-α and epoetin-β, may increase hypertension risk compared to placebo. However, the relative cardiovascular (CV) risk among ESAs remains uncertain.

Objectives: This study aims to evaluate the CV risk associated with long-acting ESAs compared to short-acting ESAs.

Methods: We searched PubMed and EMBASE from 2014 to 2023 for English studies comparing long- versus short-acting ESAs in adult CKD patients. Co-primary outcomes were CV events and mortality. Meta-analysis using Review Manager (RevMan) version 5.4. estimated relative risk (RR) or hazard ratio (HR) and assessed heterogeneity with I² statistic, utilizing fixed or random-effects models for pooling effect estimates. For safety signals of CV events associated with ESAs, we analyzed FDA Adverse Events Reporting System (FAERS) data from January 2018 to December 2022. Signals were identified by the number of adverse events ≥3, Chi² value ≥4, the proportional reporting ratio (PRR) ≥2, reporting odds ratio (ROR) ≥2, and lower limit of the 95% confidence intervals (CIs) for the information component (IC) >0.

Results: Initially, 856 studies were identified from the two databases, with 189 duplicates removed, leaving 72 for full-text assessment of eligibility, and ultimately including 12 studies in meta-analysis. 9 randomized controlled trials (RCTs) assessed CV risks, with 920 patients receiving long- and 581 receiving short-acting ESAs, while 7 RCTs with 771 and 478 patients receiving long- and short-acting ESAs and 3 observational studies assessed mortality risks. No significant differences were observed between long- and short-acting ESAs in CV risk (RR: 1.05 [95% CI: 0.82-1.35]) or mortality (RR: 1.19 [95% CI: 0.54-2.65], HR: 1.04 [95% CI: 0.94-1.15]). As a result of the disproportionality analysis, cardiogenic shock (PRR: 15.14, ROR: 15.19, lower 95% CI limit: 0.60), ventricular fibrillation (PRR: 10.49, ROR: 10.51, lower 95% CI limit: 0.54), cardiac arrest (PRR: 2.36, ROR: 2.36, lower 95% CI limit: 0.19), and others were identified as safety signals.

Conclusions: Our meta-analysis found no increased CV risk and mortality with long-acting ESAs compared to short-acting ESAs. However, analysis of the FAERS database detected signals of CV events for long-acting ESAs compared to short-acting ESAs. The limitation of our study is the number of studies included in the meta-analysis is small, and the FAERS data analysis is prone to underreporting. Further well-designed epidemiologic research with sufficient follow-up time to assess CV events is warranted.