Evaluating the efficacy and safety of gene therapy in transfusion-dependent β-thalassemia: A focus on hemolysis improvement and clonal hematopoiesis monitoring

Abstract Background. Gene therapy (GT) has emerged as a transformative treatment for transfusion-dependent β-thalassemia (TDT) and sickle cell disease (SCD). Lentiviral vector (LV)-based therapies and CRISPR-Cas9–based gene editing have demonstrated curative potential. However, reports of myeloid malignancies in SCD patients treated with LV-based therapy have raised safety concerns, particularly regarding the role of pre-existing clonal hematopoiesis (CH) and the impact of residual hemolysis. Methods. The present study proposes an approach to evaluate the efficacy and safety of TDT patients treated by GT with focus in in hemolysis improvement and CH. Real-world experience with data from the first TDT patient in France treated with GT (gene editing approach after approval by the European Medicines Agency), is presented here. Plasma oxyhemoglobin (HbO₂), heme, and hemopexin (Hpx) were measured by an in-house spectrophotometry technique (BET 18P3417). A targeted high-throughput capture panel covering 43 genes involved in CH or myeloid malignancies, and the BCL11A targeted sequence was used as part of the pre-GT workup to identify rare hematopoietic cells carrying somatic mutations (validated LOD = 0.3% on hotspots and 0.4% on all covered regions) and to follow gene editing rate and clonal evolution post-GT. All variants with an allelic frequency (VAF) above the LOD and below 40% were considered clonal. BCL11A editing was analyzed using a custom python tool suite to count and characterize InDels. Structural rearrangement involving BCL11A targeted locus was evaluated with HmnFusion. Results. Intensified therapy by GT was proposed in a 35-year-old TDT patient required regular transfusions since the first year of life. Genetically, the patient is homozygous for a β⁰ mutation (Codon 39 CAGTAG) in the HBB gene, associated with a severe phenotype. Hemolysis at baseline was represented by a significant splenomegaly (20 cm), a plasma heme level at 13,20µM, plasma HbO₂ at 17µM and undetectable Hpx at 0,0µM. Hb level was 9g/dL before transfusion (3 RBC units every 15 days), and MCHbF (MCHbF pg = MCH × %HbF) at a steady base was 1.39pg. The patient underwent autologous GT with CRISPR-Cas9–based gene editing (exagamglogene autotemcel, Vertex/CRISPR Therapeutics) on March 17, 2025, following a splenectomy to optimize engraftment. Conditioning with IV busulfan was well tolerated and no veno-occlusive disease observed. The patient received 3.48 × 10⁶ CD34+ cells/kg. Neutropenia lasted 14 days, and the last RBC unit was transfused on Day +28. Since then, the patient has remained transfusion-free with Hb reached at 15,5 g/dL and MCHbF at 31,05 pg at last follow-up (16 weeks). A high rate of edited alleles was observed over time: 38% at the first time point, 66.7% at the second, and 76.9% at the third at 16 weeks of follow-up. This progressive increase may reflects efficient and sustained gene editing indicating successful homozygous editing of both BCL11A alleles. The number of distinct edited clones identified increased from 86 at the first time point to 109 at the second, and remained stable at 108 at the third, supporting the absence of clonal dominance or selection. The most frequent repair outcome was insA, and no structural rearrangement involving the BCL11A-edited site was detected. Concerning hemolysis biomarkers at last follow up (16 weeks), Hpx was detected at 9,21µM, plasma heme drastically decreased at 0,19µM (-98,5% from baseline) and HbO₂ at 3,45µM, indicating significant hemolysis improvement. A DNMT3ACys562Gly mutation with a low variant allele frequency (1.5%) was detected before GT. Although it does not contraindicate GT, it warrants long-term hematologic monitoring due to the potential for clonal expansion, particularly following myeloablative conditioning. Following GT, the DNMT3A-mutated clone remained stable at a low frequency: 0.6% at the first time point, 0.5% at the second, and 1.1% after 16 weeks, with no evidence of clonal expansion, suggesting that it was not directly affected by the CRISPR-based intervention. Conclusion. We propose here a tailored approach to evaluate the efficacy and safety of TDT patients treated by GT, with a focus on residual hemolysis and clonal hematopoiesis. The risks observed in SCD are contrasted, and the importance of CH screening in optimizing patient selection is proposed. Long-term outcomes related to improvement of hemolysis and the potential for clonal expansion are also highlighted.