Prophylactic NAC promoted hematopoietic reconstitution by improving endothelial cells after haploidentical HSCT: a phase 3, open-label randomized trial

BACKGROUND: Poor graft function (PGF) or prolonged isolated thrombocytopenia (PT), which are characterized by pancytopenia or thrombocytopenia, have become serious complications after allogeneic hematopoietic stem cell transplantation (allo-HSCT). Our previous single-arm trial suggests that N-acetyl-L-cysteine (NAC) prophylaxis reduced PGF or PT after allo-HSCT. Therefore, an open-label, randomized, phase 3 trial was performed to investigate the efficacy and tolerability of NAC prophylaxis to reduce PGF or PT after allo-HSCT. METHODS: A phase 3, open-label randomized trial was performed. Based on the percentage of CD34(+)VEGFR2 (CD309)(+) endothelial cells (ECs) in bone marrow (BM) detected by flow cytometry at 14 days before conditioning, patients aged 15 to 60 years with acute leukemia undergoing haploidentical HSCT were categorized as low-risk (EC ≥ 0.1%) or high-risk (EC < 0.1%); patients at high risk were randomly assigned (2:1) to receive NAC prophylaxis or nonprophylaxis. The primary endpoint was PGF and PT incidence at +60 days post-HSCT. RESULTS: Between April 18, 2019, and June 24, 2021, 120 patients with BM EC <0.1% were randomly assigned for NAC (group A, N = 80) or nonprophylaxis (group B, N = 40), and 105 patients with EC≥0.1% (group C) were also analyzed. The +60 days incidence of PGF and PT was 7.5% (95% CI, 1.7 to 13.3%) and 22.5% (95% CI, 9.1 to 35.9%) in group A and group B (hazard ratio, 0.317; 95% CI, 0.113 to 0.890; P = 0.021) and 11.4% (95% CI, 5.2 to 17.6%) in group C (hazard ratio, 0.643; 95% CI, 0.242 to 1.715; P = 0.373). Consistently, NAC prophylaxis gradually improved BM ECs and CD34(+) cells in group A, whereas reduced their reactive oxygen species (ROS) levels post-HSCT. Within 60 days post-HSCT, the most common grade 3 to 5 adverse events for the NAC and control groups were infections (19/80 [24%] vs. 10/40 [25%]) and gastrointestinal adverse events (16/80 [20%] vs. 7/40 [18%]). There were no treatment-related deaths. CONCLUSIONS: N-Acetyl-L-cysteine prophylaxis can prevent the occurrence of poor hematopoietic function and is well tolerated in haploidentical HSCT. It may offer a potential pathogenesis-oriented therapeutic approach for patients with poor hematopoietic function. TRIAL REGISTRATION: This trial was registered at ClinicalTrials.gov as #NCT03967665. SUPPLEMENTARY INFORMATION: The online version contains supplementary material available at 10.1186/s12916-022-02338-9.