Assessing acute toxicity profiles of HLA-identical hematopoietic stem cell transplantation in pediatric patients with sickle cell anemia: A comprehensive analysis on behalf of the SFGM-TC

While emerging therapies such as gene therapy are being explored, allogenic hematopoietic stem cell transplantation (HSCT) remains the most established curative treatment for sickle cell anemia (SCA). Yet its application is limited by toxicity risks, donor availability, and socioeconomic factors. In pediatric cases, myeloablative HSCT from an HLA-matched related donor (HLA-MRD), following either Busulfan or Treosulfan-based conditioning regimen yields SCA-free survival rates surpassing 90%, along with minimal occurrences of graft-versus-host disease (GVHD).^1-4 However, the optimal timing of transplantation remains debated due to concerns about fertility impairment, incidence of GVHD and transplant-related mortality at older ages.² While current literature suggests that outcomes generally improve with younger age (10–16 years), extensive toxicities associated with HSCT in SCA are not well-characterized. This analysis, aims to address this gap by providing a comprehensive assessment of acute toxicities associated with pediatric HSCT from HLA-MRD in the treatment of SCA, with a comparison between two age groups.

This study included 97 SCA patients under 18 years of age who underwent HLA-MRD transplantation following myeloablative conditioning (MAC) regimen with antithymocyte globulin (ATG) between January 2005 and December 2019. Of these, 78 patients were under 13 years old, while 19 were aged 13–18 years. Median age at transplantation was 9 years [6; 12]. Patient and transplant characteristics are detailed in Table 1. The primary indications for HSCT were recurrent vaso-occlusive events, acute chest syndrome, and cerebral vasculopathy. Before transplantation, patients aged 13–18 years exhibited higher incidence of bacterial infections (68% vs. 35%, p = .01). All patients received a MAC regimen, based on IV busulfan (16 mg/kg, adjusted to drug monitoring when available with a targeted AUC of 900–1300 μmol min/L), combined with either cyclophosphamide (n = 91, total dose of 200 mg/kg) or fludarabine (n = 6, total dose of 160 mg/m²). In vivo T-cell depletion was achieved using rabbit ATG (total dose of 20 mg/kg). Cyclosporin A (CSA) was used for GVHD prophylaxis alone (n = 7), or combined with methotrexate (n = 58) or mycophenolate mofetil (n = 28) (conditioning regimens are detailed in Table S1). Bone marrow (BM) was the primary stem cell source (n = 84), while cord blood was exclusively used in patients below 13 years (p = .07), either as a unique source (n = 13, 17%) or in combination with BM (n = 7, 9%). The median infused CD34+ cell count was significantly higher in patients under 13 years (6.0 × 10⁶ CD34⁺/kg [3; 8] vs. 4.5 × 10⁶ CD34⁺/kg [3; 5], p = .034).

All patients achieved engraftment, and no cases of secondary rejection or recurrence of SCA symptoms were observed. However, two patients over 13 years required a secondary infusion of CD34+ selected cells due to poor graft function (p = .037). Both patients exhibited persistent hypoplastic marrow function at 2 months (first patient) and 8 months (second patient) post-HSCT, despite repeated whole blood chimerism assessments confirming complete donor chimerism. The need for additional cell infusions may be attributed to concomitant infections and/or the low quantity of the initial graft. Interestingly, the median duration to achieve platelet recovery was shorter for patients under 13 years (30 vs. 35 days, p = .028). Five-year overall survival (OS) and disease-free survival (DFS) were comparable, reaching 97% (CI95, 94%–100%) in patients aged <13 years and 95% (CI95, 85%–100%) for those aged 13–18 (p = .5), with a median follow-up of 53 months [36; 79]. These excellent results align with published cohorts, consistently demonstrating survival rates exceeding 95% in pediatric SCA recipients undergoing HLA-MRD transplant.^{1, 2} The main outcomes of HSCT are summarized in Table S2 and Figure S1.

Managing infections during immunosuppressive therapy is crucial yet challenging in terms of diagnosis and treatment. Documentation of such outcomes in SCA recipients post-HSCT, especially in younger cohorts, is scarce despite infections being the leading cause of mortality.^{1, 2} Consistent with this observation, all three deaths in our study (2 patients <13 years (3%) and 1 patient aged 13–18 (5%)) were infection-related, occurring at 2-, 5- and 10-months post HSCT, while under immunosuppressive therapy. This underscores the need to carefully managing infection indicators in HSCT for SCA.

Specific treatment for cytomegalovirus (CMV), Epstein Barr Virus (EBV), BK virus and adenovirus replication was required in 28%, 16%, 11% and 4% of cases respectively (Table 1). Antiviral prophylaxis during our study relied on acyclovir or valacyclovir, initiated with the MAC regimen and continued until CD4+ levels reached at least 0.25 × 10⁹/L. Notably, no EBV-related post-transplant lymphoproliferative disease or CMV disease occurred. BK virus and adenovirus replications were successfully treated with cidofovir, administered either systemically or via bladder infusion, and brincidofovir, administered systemically. BK virus replication requiring treatment was significantly more frequent in patients aged 13–18 (26% vs. 8%, p = .037), marking the first instance of such an observation in recipients with SCA. The incidence in our cohort surpassed the rarely described rates found in existing literature⁵ and can be attributed to the tight monitoring and proactive treatment of viruria in our centers. This approach is supported by the high severity of symptoms, as 36% of affected individuals needed an indwelling catheter to prevent acute urinary retention. Additionally, endovascular lesions caused by SCA itself might sustain this phenomenon and warrant consideration.

In our cohort, 31% of patients developed bacterial infections requiring systemic antibiotics, with no significant association with prior splenectomy (p = .6). Gram-positive bacteria were the predominant pathogens, accounting for 50% of these infections. Notably, adolescents aged 13–18 had a significantly higher infection incidence than younger patients (53% vs. 26%, p = .029). Yet neutrophil reconstitution was similar across age groups, suggesting that factors such as bacterial carriage or antibiotic resistance from previous exposures might contribute to this disparity and need further investigations. Additionally, 10 patients (10%) were treated for probable or proven fungal infection (Candida sp. (n = 7), Aspergillus (n = 3)), with similar incidence between age groups. In univariate analysis, patients with infections requiring systemic treatment were older (10 [7; 12] vs. 8 [5; 9], p = .008) and received lower stem cell doses (4 × 10⁶ CD34⁺/kg [3; 6] vs. 7 × 10⁶ CD34⁺/kg [5; 9], p < .001) (Table S3). Interestingly, viral replication and bacteremia rates seem to be less frequent in SCA patients receiving reduced-intensity regimens,⁵ highlighting the need for further prospective studies on conditioning regimen intensity, especially in adolescents.

Older patients in our cohort demonstrated a greater prevalence of acute Grade III–IV skin toxicity (10% vs. 0%, p = .037), all linked to busulfan administration, as diffuse toxic erythema associated with pruritus and/or oral or genital erosion. These symptoms resolved spontaneously without steroid treatment, underscoring their distinctive nature compared to acute GVHD. Following mucositis (53% incidence), neurological impairment was the second most prevalent Grade III–IV acute organ toxicity in our study, affecting 12% of patients. Occurrences were consistent across age groups and not influenced by prior cerebral vasculopathy, with all cases resolving favorably. This was primarily characterized by acute encephalitis in a septic context (n = 2) or seizures and posterior reversible encephalopathy syndrome (PRES; n = 10). PRES occurred early, with a median time of 17 days post-HSCT (5–28), and were associated with CSA treatment. Switching to mycophenolate mofetil and maintaining strict blood pressure control led to positive outcomes. The incidence of PRES in our cohort was notably higher than reported in the literature.³ All patients received anticonvulsant prophylaxis during busulfan administration and CSA therapy (Table S1). However recent years have seen enhanced management strategies, focusing on precise control of arterial hypertension, prompt correction of magnesium deficiency, and maintenance of hemoglobin levels above 9 g/dL and platelet counts above 50 × 10⁹/L. Additionally, eight patients (8%) experienced significant acute endothelial toxicity, manifesting as thrombotic microangiopathy and capillary leak syndrome. Three patients (3%) developed Grade III–IV acute cardiac toxicity, all presenting with pericardial effusion. Concurrent infections (Adenovirus or Candida sp.) were identified in two cases without definitive evidence for infectious causality.

The incidence of veno-occlusive disease (VOD) in the cohort was 12% (n = 12), displaying comparability between age groups (Table 1). Individuals aged 13–18 years exclusively manifested mild forms of VOD, while younger patients experienced moderate and severe VOD in 5% and 3% of cases, respectively. The median time to onset was 13.5 days (9–21). Importantly, no cases of very severe VOD were observed, consistent with previous literature.

Grade III–IV acute GVHD occurred in 11% of the cohort (n = 11), aligning with previous findings in comparable cohorts,^{1, 2} and all cases resolved favorably. There was no significant difference between the age groups. The skin was the predominantly affected organ (88%), before gut (19%) and liver (11%). Of note, no risk factors of Grade III/IV acute GVHD nor acute organ toxicity were identified.

This study corroborates HSCT excellent outcomes in young HLA-MRD recipients in SCA. While large adult and pediatric series have demonstrated that older transplant age is associated with reduced OS/DFS or increased GVHD incidence, our findings align with Bernaudin et al.³ and indicate that this trend does not persist in younger cohorts. Yet, our study emphasizes a notable association between older age at transplantation and increased infection rates, suggesting interests of reduced-intensity conditioning in adolescents. Future prospective studies examining alternative therapies (gene therapy or exagamglogene autotemcel),⁶ as well as other conditioning regimens and transplantation approaches are essential to improve quality of life, and reduce both acute and long-term toxicity in SCA patients transplanted at ages older than 13.

MD, AG, and JHD contributed to the study's conception, data interpretation, and manuscript writing. GPA conducted statistical data generation. BK, CP, and BN aided in assembling patient data. All authors approved the final submitted version.

The authors declare no conflicts of interest.