Raising hemoglobin level with erythropoietin improves cerebral but not muscle tissue oxygenation in patients with sickle cell disease

Abstract Background In sickle cell disease (SCD), chronic anemia impairs tissue oxygen (O2) delivery, causing hypoxia, microvascular occlusion, and organ damage. Hemoglobin (Hb) is routinely used to monitor treatment response, but does not directly monitor tissue oxygenation. Tissue O2saturation (StO₂) from near-infrared spectroscopy (NIRS) offers a non-invasive way to assess the effects of Hb-raising therapies or Hb O2affinity modifiers on microvascular O2delivery. In a substudy of ACHiEvE-SCD (NCT05451940), a phase 1b/2 trial of erythropoietin (EPO) in SCD, we assessed changes in cerebral StO₂ (cStO₂) and muscle StO₂ (mStO₂) with EPO treatment and factors affecting StO₂ in different tissues. Methods Adults ≥18 years with SCD, Hb ≤ 9.0 g/dL, on stable-dose hydroxyurea were enrolled in two centers (US and Nigeria). Patients received subcutaneous, dose-escalated EPO over 12 weeks, followed by 12 weeks of routine care. Continuous-wave NIRS (Portalite, Artinis Medical Systems) was used to measure microvascular StO₂ in the left frontal cortex and forearm muscle at baseline (before EPO), 4, 8, 12, and 24 weeks. Tissue hemodynamics were collected at a sampling frequency of 50 Hz to obtain continuous StO₂ measurements. Observations were excluded for StO2 < 30%, poor data quality, and alternative anemia treatments. Labs and vitals, including SpO2 (by pulse oximetry), were measured at each time point. Paired t-tests were used to compare time points. Correlations were analyzed using Pearson's r. Linear mixed-effects modeling was used to identify factors associated with cStO₂ and mStO₂, and models were selected using standard errors, AIC, BIC, and residual log-likelihood for model performance. Results A total of 17 patients were enrolled; 16 completed 12 weeks of EPO treatment. The mean age was 30 years (range 18–50); 9 (53%) female; all had HbSS genotype and identified as African, Black, or African American. In the 16 patients completing treatment, mean Hb rose from 7.6 ± 1.4 g/dL at baseline to 10.6 ± 1.2 g/dL at week 12 (p<0.001). There were no significant blood pressure (BP) or SpO2 changes at 12 or 24 weeks. At 12 weeks, mean cStO₂ increased from 49.4 ± 8.4% to 55.1 ± 6.7% (p=0.017), while mStO₂ remained unchanged (61.6 ± 4.0% to 61.8 ± 4.1%). At week 24, mean Hb levels declined to 8.5 g/dL (p=0.002), likely due to cessation of EPO after 12 weeks, and no significant changes in cStO2 and mStO2 were observed. We performed univariate linear mixed-effects regression on data through week 24 using pre-selected variables: age, gender, Hb, lactate dehydrogenase (LDH), indirect bilirubin, absolute reticulocyte count (ARC), and systolic and diastolic BP (SBP and DBP). The 17 patients contributed 75 cerebral and 82 mStO₂ observations. cStO₂ was associated with Hb level (β=1.8, p<0.0001) and DBP (β=0.2, p=0.02), with LDH approaching significance (β=-0.01, p=0.06). mStO₂ was associated with female sex (β=2.9, p=0.02), indirect bilirubin (β=-0.7, p=0.04), ARC (β=-0.01, p=0.01), SBP (β=0.09, p=0.01), and DBP (β=0.20, p=0.02), but not Hb. Predictors with p<0.1 were included in multivariate analysis. In a multivariate model, Hb level (β=1.6; p<0.0001) was the only significant predictor of cStO₂ after adjusting for age, gender, LDH, and DBP, suggesting that cStO₂ is predominantly dependent on Hb level. In this model, cStO₂ increases 1.6% for every 1 g/dL increase in Hb. DBP was not associated with cStO₂ after controlling for Hb, suggesting that cStO₂ is less strongly influenced by hemodynamics in the systemic circulation, possibly resulting from cerebral autoregulation. In contrast, DBP (β=0.1, p=0.0007), ARC (β=-0.01, p=0.002), and female sex (β=2.7, p=0.0005) remained independently associated with mStO₂ after adjustment for age, Hb, and indirect bilirubin. Additionally, a positive correlation was observed between SpO₂ and mStO₂ (r=0.2, p=0.05), but not cStO₂ (r=-0.03, p=0.78), further demonstrating the differential regulation of cerebral vs. peripheral tissue hemodynamics. Conclusion ACHiEvE-SCD demonstrates that EPO treatment raises Hb levels in SCD, increasing cStO₂, but not mStO₂. Our data suggest that raising Hb level can enhance cerebral oxygenation, while BP, hemolysis, and sex may influence peripheral microvascular StO₂. These tissue-specific responses might explain observed therapeutic effects and safety considerations related to new treatments in SCD, highlighting the need for further research.