Effects of vaso-occlusive episodes, acute kidney injury, and APOL1 status on urine albumin concentration in patients with sickle cell disease

Abstract Introduction: Chronic kidney disease (CKD) is a common complication associated with substantial morbidity and early mortality in people with sickle cell disease (SCD). The NHLBI and ASH guidelines recommend screening for those at risk for advanced CKD with the urine albumin-to-creatinine ratio (UACR). Although an elevated UACR predicts the risk of CKD progression in some studies there is very high intrapatient variability. This leads to challenges in establishing thresholds to initiate therapy and to assess response to agents, such as renin-angiotensin-aldosterone system inhibitors (RAASi). Vaso-occlusive episodes (VOE) may lead to sudden increases in intravascular hemolysis and acute vascular damage that may exacerbate organ function, such as to the kidneys, both in the acute and long-term. The impact of a VOE or a VOE with acute kidney injury (AKI) on UACR are unknown. The APOL1 G1 and G2 kidney risk variants, associated with a higher risk of developing albuminuria in children and CKD progression in adults with SCD, have recently been implicated in AKI-risk in non-SCD. The effect of APOL1 on AKI risk during a VOE is unclear but may also contribute to UACR variability in this clinical setting. Methods: To address these issues, we analyzed predictors of UACR in a longitudinal registry of patients with SCD followed at our institution. Patients with UACR values between 1/2010-7/2024 and with at least 6 months of follow up were selected. Generalized estimated equations were applied to investigate the relationship between a VOE requiring emergency room care or inpatient hospitalization or a VOE complicated by AKI, as determined by ICD-9 or ICD-10 codes, with UACR. Days from a VOE or VOE with AKI event to the next UACR measurement were evaluated as linear variables and as a categorical variable based on a 30-day threshold. APOL1 G1 and G2 risk variants were genotyped by PCR and high risk was defined as compound heterozygosity or homozygosity for G1 and/or G2. Univariate models and multivariate models, adjusting for age, sex, SCD genotype, hydroxyurea [HU] or RAASi therapy, were performed. Beta estimates ± standard errors from the generalized estimated equations are provided. Results: In the 378 SCD patients analyzed, 42% were male, mean age at enrolment was 40 ± 12 years, 75% were hemoglobin SS/Sβ0-thalassemia genotype, 56% were on HU, and 30% were on RAASi therapy. The mean UACR value was 308 ± 718 mg/g, demonstrating wide variability. Over a mean follow up of 8.1 ± 3.6 years, a VOE requiring emergency room care or inpatient hospitalization was observed in 325 (86%) SCD patients. No associations were observed between UACR concentrations and a history of VOE, proximity of a UACR measurement to VOE in days, or if the UACR was measured within 30-days of a VOE (P ≥ 0.5). A VOE complicated by AKI was observed in 148 (39%) of SCD patients. UACR concentrations were higher in SCD patients who had an AKI event (β 0.31 ± 0.11; P = 0.0057) or if the UACR was assessed within 30-days of the AKI event (β 0.25 ± 0.13; P = 0.05) on univariate analysis. High-risk APOL1 was co-inherited in 13% of SCD patients in our cohort (48/370 evaluable). A similar proportion of SCD patients had an AKI event with (38%, 18/48) versus without (39%, 126/322) high-risk APOL1. VOE complicated by AKI (β 0.34 ± 0.11; P = 0.0025) and high-risk APOL1 (β 0.71 ± 0.26; P = 0.007) were independently associated with UACR after adjusting for age, sex, SCD genotype, HU & RAASi use. A trend for higher UACR when measured within 30-days of AKI (β 0.24 ± 0.13; P = 0.067) was observed after similar adjustment along with APOL1 high-risk status (β 1.03 ± 0.39; P = 0.008) being associated with higher UACR in the multivariable model. We did not observe an interaction between AKI and APOL1 high-risk status on UACR (P = 0.9). Discussion: Our findings demonstrate that VOE complicated by an AKI increases UACR values and testing UACR within 30 days of the AKI event may lead to acute rises in UACR. This suggests that a steady-state UACR value should be defined by a measurement > 30 days from a VOE with AKI. Furthermore, while APOL1 high risk was associated with higher UACR values over time, we did not observe an association between high-risk APOL1 status and AKI or an interaction for the association between UACR and AKI by APOL1 status. This may indicate separate and independent mechanisms for APOL1- versus AKI-mediated kidney damage in the context of SCD.