Baseline Hemoglobin and Outcomes in Advanced Chronic Kidney Disease with Moderate-to-Severe Myocardial Ischemia: A Secondary Analysis of the ISCHEMIA-CKD Trial

Background/Objectives: Hemoglobin may identify differences in prognosis among patients with advanced chronic kidney disease (CKD) and documented myocardial ischemia. We examined the shape of its association with clinical outcomes in ISCHEMIA-CKD. Methods: This exploratory observational secondary analysis included 758 randomized participants with baseline hemoglobin. The primary endpoint was the parent-trial composite of all-cause death or myocardial infarction (MI). Adjusted Cox regression with restricted cubic splines characterized the hemoglobin association. Mortality and pooled nonfatal MI were examined using Cox and competing-risk models. Results: Median hemoglobin was 11.1 g/dL; 247 participants experienced the primary endpoint. The adjusted association departed from log-linearity (p = 0.015), with higher hazard at lower hemoglobin and an uncertain upper limb. Relative to 11.1 g/dL, the hazard ratio (HR) at 8 g/dL was 1.98 (95% confidence interval [CI] 1.26–3.10). As a linear summary, the HR per 1 g/dL lower hemoglobin was 1.13 (95% CI 1.04–1.23) for the primary endpoint. Corresponding mortality associations were present, whereas pooled nonfatal MI showed no clear association (subdistribution HR 1.03, 95% CI 0.89–1.21). Exploratory treatment interaction tests provided no convincing evidence of effect modification (linear p = 0.059; spline p = 0.124). Conclusions: Among selected, clinically stable participants with advanced CKD and moderate-to-severe ischemia, lower baseline hemoglobin was associated with adverse prognosis. Nonlinearity and uncertainty at higher values argue against a single summary estimate or an inferred optimal target. These potentially confounded findings neither establish benefit from hemoglobin correction nor support hemoglobin-based selection of invasive versus conservative management.