Renin-angiotensin-aldosterone system inhibitors exacerbate anemia in sickle cell disease

Abstract Introduction: Renin-angiotensin-aldosterone system inhibitors (RAASi) are recommended by American Society of Hematology and National Institutes of Health guidelines to treat albuminuria in patients with sickle cell disease (SCD). RAASi reduce angiotensin II activity and are associated with worsening anemia in non-SCD diabetic patients. This may be due to angiotensin II's role as a growth factor for erythroid progenitors and as an erythropoietin (Epo) secretagogue [PMID: 20400218]. The effect of RAASi on hemoglobin (Hb) concentration in SCD is unclear but important to understand in this chronic hemolytic disorder. Methods: We investigated the association between RAASi with anemia in a cross-sectional and 2 longitudinal SCD cohorts as well as in SCD mice. Cross-sectional analyses were performed using baseline data from the multicenter Walk-PHaSST cohort (n=656). Hb, Epo concentrations, and reticulocyte % were compared by RAASi using ANOVA adjusting for age, sex, SCD genotype, estimated glomerular filtration rate (eGFR), and hydroxyurea (HU). For longitudinal analyses, rates of change in Hb concentration and reticulocyte % pre- versus on RAASi therapy were compared in 24 SCD patients enrolled in a longitudinal registry with ≥1 year follow up both before and on RAASi therapy. Hb and reticulocyte % slopes were calculated by linear regression by time. Changes in Hb and reticulocyte % were also compared in the multi-center, phase 2 losartan clinical trial before (average of screening and Visit 1 values) versus at 6 months of losartan therapy in all 15 participants with data available. Paired t-test was used to compare pre- versus on-treatment values in both longitudinal analyses. Transgenic SCD mice (Townes model; 6 male, 6 female) were treated with losartan (0.2mg/mL in drinking water) for 14 weeks starting at 12-weeks of age. Hb, absolute reticulocyte counts, and Epo concentrations were compared to 12 untreated SCD mice (6 male, 6 female) after 6- and 14-weeks of treatment by ANOVA adjusting for sex. Results: Cross-sectional: The median age of the Walk-PHaSST cohort was 36 (interquartile range [IQR] 25–47) years, 346 (53%) were female, 501 ](76%) were Hb SS/Sβ0 genotype, 245 (37%) were on HU, and the median eGFR was 117 (IQR, 93–129) mL/min/1.73m2. Those on RAASi (n=58) were older (42 vs 35 years; P=0.002) and had a lower eGFR (117 vs 100 mL/min/1.73m2; P<0.0001). RAASi use was independently associated with lower Hb concentrations (β -0.44 ±0.21; P=0.039) after adjusting for age, sex, HU, eGFR, and SCD genotype; there were no significant differences in reticulocyte %, Epo, or Epo-to-Hb concentrations by RAASi use (P≥0.3). Longitudinal: The median age of the longitudinal SCD cohort treated with RAASi was 42 (IQR, 34–48) years, 14 (58%) were female, 21 (88%) were Hgb SS/Sβ0 genotype, 13 (54%) were on HU, and the median eGFR was 109 (83–128) mL/min/1.73m2. The rate of Hb decline accelerated on versus pre-RAASi therapy (mean difference -0.31 g/dL per year; 95% confidence interval [CI] -0.55 to -0.06 g/dL per year; P=0.016); no significant difference in reticulocyte % change was observed (P=0.7). Baseline demographics of the losartan clinical trial were previously described (PMID 28589652). Compared to pre-RAASi use, the mean Hb declined by -0.36 g/dL (95% CI: -0.76 to 0.04 g/dL; P=0.077) on RAASi therapy. No difference in reticulocyte % was observed (P=0.3). Transgenic SCD mice: Losartan treatment led to significantly lower Hb concentrations after 6-weeks (β -1.9 ±0.41g/dL; P=0.0002) and 14-weeks (β -2.0 ±0.34g/dL; P<0.0001) compared to untreated mice, adjusting for sex. There was a trend at 6-weeks (β -489 ±242 x103/µL; P=0.058) and significantly lower absolute reticulocyte counts after 14-weeks (β -460 ±192 x103/µL; P=0.027) of RAASi treatment compared to untreated mice. No significant difference in Epo or Epo-to-Hb concentration was observed by RAASi treatment (P≥0.1). Discussion: Our findings demonstrate RAASi use is associated with lower Hb concentrations in SCD. RAASi were not associated with lower Epo levels suggesting that our observations are due to reduced erythroid progenitor function. This is currently being investigated in bone marrow cells of SCD mice treated with RAASi. While a similar decline in Hb has been observed in non-SCD patients on RAASi, a drop in Hb in an anemic patient may not be as well tolerated. In conclusion, RAASi should be used with caution in SCD with close monitoring of Hb concentrations.