AUTHORS
Xianfu Sheng, MD, Nanxi Dong, MD2, Yingying Shen3, Wenbin Liu, MD,PhD4, Baodong Ye, MD, PhD5, Yu Zhang Jr, MD6
AFFILIATIONS
1 The First Affiliated Hospital of Zhejiang Chinese Medical University (Zhejiang Provincial Hospital of Chinese Medicine), Hangzhou, Zhejiang, China
2 The First Affiliated Hospital of Zhejiang Chinese Medical University (Zhejiang Provincial Hospital of Chinese Medicine), Hangzhou, Zhejiang, China
3 The First Affiliated Hospital of Zhejiang Chinese Medical University (Zhejiang Provincial Hospital of Chinese Medicine), Hangzhou, Zhejiang, China
4 The First Affiliated Hospital of Zhejiang Chinese Medical University (Zhejiang Provincial Hospital of Chinese Medicine), Hangzhou, Zhejiang, China
5 The First Affiliated Hospital of Zhejiang Chinese Medical University (Zhejiang Provincial Hospital of Chinese Medicine), Hangzhou, Zhejiang, China
6 The First Affiliated Hospital of Zhejiang Chinese Medical University (Zhejiang Provincial Hospital of Chinese Medicine), Hangzhou, Zhejiang, China
ABSTRACT
Severe aplastic anemia (SAA) is a life-threatening bone marrow failure syndrome for which allogeneic hematopoietic stem cell transplantation (HSCT) is curative. Haploidentical HSCT (haplo-HSCT) provides rapid donor availability, but graft rejection, graft-versus-host disease (GVHD), and regimen-related toxicity remain major concerns. Within the granulocyte colony-stimulating factor (G-CSF)/antithymocyte globulin (ATG)-based Beijing platform, busulfan/cyclophosphamide/ATG (Bu/Cy/ATG) is commonly recommended to reduce rejection, whereas fludarabine/cyclophosphamide/ATG (Flu/Cy/ATG) is often used in practice to mitigate busulfan-related toxicity. Direct comparative data between these conditioning strategies in haplo-HSCT for SAA are limited.
We retrospectively analyzed 76 consecutive patients with SAA who underwent first haplo-HSCT at our institution from January 2019 through December 2022. This study was registered at Chinese Clinical Trial Registry (ChiCTR2600128918). Forty-six patients received Flu/Cy/ATG and 30 received Bu/Cy/ATG. Flu/Cy/ATG used fludarabine (total 150 mg/m2), cyclophosphamide (total 120 mg/kg), and rabbit ATG (total 10mg/kg); Bu/Cy/ATG used busulfan (total 6.4mg/kg), cyclophosphamide (total 200 mg/kg), and rabbit ATG (total 10mg/kg). GVHD prophylaxis included cyclosporine, mycophenolate mofetil, and short-course methotrexate. Primary endpoints were overall survival (OS) and GVHD-free, failure-free survival (GFFS); secondary endpoints included engraftment, GVHD, infections, organ toxicity, and other safety events.
Baseline characteristics were generally balanced between groups. Median age was 30.5 years in the Flu/Cy/ATG group and 30 years in the Bu/Cy/ATG group (P=0.941). Disease severity, prior therapy, HLA mismatch, donor-specific antibody positivity, HCT-CI, graft source, mononuclear cell dose, and CD34+ cell dose were also comparable.
Engraftment was rapid and similar between groups. Median neutrophil recovery was 11 days in both (ranges 9–23 vs 8–29; P=0.962), and platelet recovery was 13 days in both (ranges 10–72 vs 10–70; P=0.997). Cumulative day+30 neutrophil engraftment (97.8% vs 96.7%, P=0.396) and day+100 platelet engraftment (95.7% vs 96.7%, P=0.758) were comparable. No primary graft failure occurred; secondary failure occurred in 2 (4.4%) vs 1 (3.4%) patient (P=1.000).
GVHD outcomes did not differ significantly between conditioning groups. The cumulative incidences of grade II-IV acute GVHD were 26.1% after Flu/Cy/ATG and 23.3% after Bu/Cy/ATG (P=0.789), and the corresponding incidences of grade III-IV acute GVHD were 10.9% and 16.7% (P=0.485). The 5-year cumulative incidences of chronic GVHD were 24.4% and 21.4% (P=0.739), respectively, and moderate-severe chronic GVHD occurred in 7.3% and 3.6% (P=0.502).
With a median follow-up of approximately 5 years among survivors, the estimated 5-year OS was 77.3% after Flu/Cy/ATG and 90.0% after Bu/Cy/ATG (P=0.196); 5-year GFFS was 62.3% and 76.7%, respectively (P=0.246). In multivariable analysis, conditioning regimen was not independently associated with OS or GFFS. Graft failure was associated with OS (HR 33.545, 95% CI 6.545-171.919; P<0.001) and GFFS (HR 10.572, 95% CI 2.815-39.704; P<0.001), while grade II-IV acute GVHD and moderate-severe chronic GVHD were associated with loss of GFFS.
Despite comparable efficacy endpoints, toxicity profiles differed. Flu/Cy/ATG was associated with more pneumonia (34.8% vs 10.0%; P=0.015) and EBV reactivation requiring rituximab (26.1% vs 6.7%; P=0.033). CMV reactivation was numerically higher after Flu/Cy/ATG (34.8% vs 16.7%; P=0.084), but CMV disease was rare. In contrast, Bu/Cy/ATG was associated with more thrombotic microangiopathy (16.7% vs 2.2%; P=0.027) and numerically higher cardiac dysfunction (16.7% vs 4.3%; P=0.103) and hemorrhagic cystitis (36.7% vs 19.6%; P=0.097). Ten deaths occurred after Flu/Cy/ATG and 3 after Bu/Cy/ATG, with infection-related mortality predominating after Flu/Cy/ATG.
In this single-center real-world cohort, Flu/Cy/ATG and Bu/Cy/ATG conditioning produced comparable engraftment, GVHD, OS, and GFFS after haplo-HSCT for SAA, but with distinct toxicity patterns. These findings support risk-adapted conditioning selection: Flu/Cy/ATG may be reasonable when busulfan-related toxicity or fertility preservation is a major concern, provided infection monitoring is stringent, whereas Bu/Cy/ATG may be favored when rejection risk is paramount or infection risk is high.
DISCLOSURES
Xianfu Sheng, MD: No relevant conflicts of interest to declare.