Abstract
Introduction Heart failure with preserved ejection fraction (HFpEF) presents a major health burden worldwide. Up to 40% of HFpEF patients present with elevated urine-albumin-creatinine-ratio levels (UACR). In this subgroup, the risk for adverse cardiovascular (CV) outcomes is increased. Prior studies suggest a correlation between albuminuria and incident heart failure and adverse outcomes even at levels below current clinical thresholds and in the presence of preserved renal function and may mark early cardio-renal and endothelial dysfunction. Whether right ventricular (RV) systolic function, as measured by RV free wall strain (RVFWS), relates to albuminuria in heart failure with preserved ejection fraction (HFpEF) has not been characterized. Methods Patients with HFpEF (LVEF ≥50%, HFA-PEFF score ≥5 or invasively elevated filling pressures) and patients of a CKD comparator cohort (CKD ≥G3 or >A1 and ≥1 CV risk factor, without diagnosed HF) were pooled from two cross-sectional cohort studies for this analysis. All patients underwent comprehensive assessment including albuminuria and RVFWS measurement. Albuminuria was categorized by KDIGO stage (A1/A2/A3). RVD was defined by RVFWS ≥ -20%, RV FAC <35% or TAPSE <17 mm. We examined the association between RVFWS and albuminuria using a multivariate linear model with a cohort interaction term and cohort-specific models. Model robustness was assessed by elastic-net regression, bootstrap internal validation, E-value and by adding NT-proBNP, neutrophile count, pulmonary artery systolic pressure and urinary α1-microglobulin individually in separate sensitivity models. Inter- and intra-observer reproducibility of RVFWS was quantified. Results In total, 119 HFpEF and 133 CKD patients were identified. In HFpEF, RVFWS was more impaired (-22.5% vs. -25.7%, p < 0.001) and RVD more prevalent (30% vs. 16%). Median UACR lay within the KDIGO A1 range in both cohorts. More impaired RVFWS was associated with higher UACR in HFpEF (adjusted UACR ratio 1.55 per 5% less-negative strain, 95% CI 1.17-2.04, p=0.002) but not in the CKD cohort (p=0.380), with a significant cohort interaction (p=0.002). The association persisted after adjustment in sensitivity models, and RVFWS was retained by penalized regression. Inter- and intra-observer agreement of RVFWS was moderate-to-good (ICC 0.71 and ICC 0.69, respectively). Discussion In ambulatory HFpEF patients with predominantly low-grade albuminuria, RVFWS is associated with UACR independently of clinical confounders, an association not observed in a CKD cohort with CV risk factors and not attenuated after adjustment for estimated PASP.