Impact of pre-transplant bone marrow cellularity and fibrosis on post-transplant outcomes in sickle cell disease

Abstract Introduction: Sickle cell disease (SCD) is an inherited hemoglobinopathy in which erythrocytes sickle during periods of hypoxia and stress, resulting in damage to body organs. Repeated insults to the bone marrow promote a hyper-inflammatory state characterized by altered bone marrow cellularity and fibrosis. SCD is thought to also be associated with clonal hematopoiesis (CH), yet this relationship is not fully understood. Hematopoietic cell transplantation (HCT) offers curative therapy for those with SCD, yet the impact of baseline cellularity and fibrosis on HCT outcomes and CH remains unknown. Alterations in bone marrow fibrosis impact prognosis in other disease states such as AML. We therefore hypothesize that graft failure and falling donor chimerism are seen more frequently in patients with higher levels of baseline bone marrow cellularity and fibrosis, and that those patients with increased cellularity, potentially suggesting increased erythropoietic stress, will demonstrate a higher prevalence of CH. Methods: We conducted a retrospective study to evaluate pre-HCT bone marrow samples for patients with SCD who received non-myeloablative haploidentical HCT at the National Institutes of Health (NIH) between 2010 and 2023 (NCT00977691and NCT03077542). All patients received alemtuzumab and TBI with post-HCT cyclophosphamide; some received additional pre-conditioning with pentostatin and cyclophosphamide. All core and aspirate slides were reviewed by an expert hematopathologist and assigned a numerical cellularity value (0-100%) and a biopsy score consistent with the modified Bauermeister grading system (0-4). Cellularity values were then adjusted for age and primary variables were compared against metrics of HCT outcome, including graft rejection, donor chimerism levels, and the presence and development of CH (variant allele frequency >0.05%). Patients were categorized as either those with sustained engraftment or graft failure; the latter included those with graft rejection and/or low donor myeloid chimerism (<20%) at last follow-up. Results: A total of 39 patients had pre-HCT marrows and were analyzed, of which 34 patients had cellularity data and 26 patients had fibrosis data. 22 (56.4%) patients had sustained engraftment. Median age, African American race, SCD genotype, and baseline laboratory values such as hemoglobin, reticulocyte count, absolute neutrophil count, and %HbS, did not differ significantly between engrafted and failed patients. Neither baseline cellularity nor fibrosis score impacted engraftment status (p=0.33 and p=0.81, respectively). Interestingly, all three patients with a fibrosis score of>2 demonstrated graft failure. We did not observe a significant correlation between cellularity and fibrosis and individuals' average and last reported donor chimerism in follow-up (p=0.47). We observed a significant difference in the development of CH between our two clinical trials following HCT (p=0.01), which was not observed before HCT (p=0.41); 9 (33%) patients had a detectable clone before HCT, while 19 (59%) patients had a detectable clone following HCT. While these differences were not associated with baseline cellularity and fibrosis, baseline cellularity was associated with the presence of DNA-damage response (DDR) mutations before HCT (p=0.04). Conclusions: Our initial data suggest that neither baseline cellularity nor fibrosis significantly impacted HCT outcomes. Interestingly, all patients with fibrosis ratings > 2 demonstrated graft failure. Further, we observe that baseline cellularity is associated with the presence of DDR mutations before HCT. We will expand to our non-myeloablative matched-sibling HCT cohort, increasing our overall study sample size and more definitively assessing the impact of baseline cellularity and fibrosis values on HCT outcomes and CH. Further studies are indicated to evaluate whether increased erythropoietic stress is associated with an increased prevalence of baseline DDR-CH.