A Risk Score System for Stratifying the Risk of Relapse in B Cell Acute Lymphocytic Leukemia Patients After Allogenic Stem Cell Transplantation

B cell acute lymphocytic leukemia (B-ALL) remains a challenging hematologic malignancy, with relapse being the primary cause of treatment failure even after allogeneic stem cell transplantation (allo-SCT). Despite advancements in therapeutic strategies, including chemotherapy and transplantation, a significant proportion of patients experience disease recurrence, underscoring the need for robust prognostic tools to guide post-transplant management. This study aimed to develop a risk score system integrating key clinical and laboratory variables to stratify relapse risk in B-ALL patients following allo-SCT, enabling personalized therapeutic interventions.

The research cohort comprised 477 B-ALL patients who underwent allo-SCT at Peking University People’s Hospital between December 2010 and December 2015. The cohort included both pediatric and adult patients (median age: 26 years; range: 2.5–63 years), with 34.2% having Philadelphia chromosome-positive (Ph+) disease and 65.8% Ph-negative (Ph−) disease. Most patients (88.7%) were transplanted in first complete remission (CR1), while 11.3% were in ≥CR2. Transplant types included haploidentical blood and marrow transplantation (HBMT; 71.3%), HLA-matched sibling donor transplantation (MSDT; 26.6%), and HLA-matched unrelated donor transplantation (MUDT; 2.1%). All patients received myeloablative conditioning regimens, with variations in protocols depending on donor type. For instance, HBMT recipients received anti-thymocyte globulin (ATG) as part of their conditioning, while MSDT patients did not.

Post-transplant outcomes were evaluated with a median follow-up of 1,570 days (range: 24–3,107 days). Neutrophil engraftment was universal, and 95.4% achieved platelet engraftment. The cumulative incidence of grade III–IV acute graft-versus-host disease (aGVHD) at 100 days was 5.9%, while chronic GVHD (cGVHD) occurred in 48% of patients by 5 years. The 5-year cumulative incidence of relapse (CIR), overall survival (OS), leukemia-free survival (LFS), and non-relapse mortality (NRM) were 20.7%, 70.4%, 65.6%, and 13.9%, respectively. These outcomes highlighted the persistent challenge of relapse despite successful engraftment and GVHD management.

Minimal residual disease (MRD) status, assessed via multiparametric flow cytometry (MFC), emerged as a critical prognostic factor. Pre-transplant MRD (pre-MRD) positivity (≥0.001% leukemic cells) was observed in 24.7% of patients (118/477), while post-transplant MRD (post-MRD) positivity occurred in 16.1% (77/477). Post-MRD positivity was strongly associated with adverse outcomes: patients with post-MRD+ had a 5-year CIR of 66.6% compared to 11.9% in post-MRD− patients (P < 0.001). Similarly, 5-year OS and LFS were significantly lower in post-MRD+ patients (41.5% vs. 76.0% and 28.2% vs. 72.8%, respectively; P < 0.001). Landmark analyses at days +30, +60, +90, +120, and +180 post-transplant consistently demonstrated that post-MRD positivity at any time point correlated with higher relapse rates and inferior survival.

Disease status at transplantation also significantly influenced outcomes. Patients transplanted in ≥CR2 had a 5-year CIR of 25.3% versus 10.5% for those in CR1 (P = 0.007). Additionally, the absence of cGVHD was linked to higher relapse risk. Patients without cGVHD experienced a 5-year CIR of 16.3% compared to 6.3% in those with cGVHD (P < 0.001). These findings underscored the dual role of cGVHD in mediating graft-versus-leukemia (GVL) effects while balancing risks of morbidity.

Multivariate Cox analysis identified three independent predictors of relapse: post-MRD positivity (hazard ratio [HR] = 7.84, P < 0.001), transplantation in ≥CR2 (HR = 2.21, P = 0.004), and absence of cGVHD (HR = 3.12, P < 0.001). These variables were integrated into a novel risk score system, assigning one point each for post-MRD+, ≥CR2 status, and lack of cGVHD. Patients were stratified into four risk groups: score 0 (no risk factors), score 1 (one factor), score 2 (two factors), and score 3 (all three factors).

The risk score system demonstrated remarkable discriminative power. The 5-year CIR escalated sharply across groups: 6.3% (score 0), 16.6% (score 1), 55.9% (score 2), and 81.8% (score 3) (P < 0.001). Survival outcomes mirrored this trend, with 5-year OS declining from 85.7% (score 0) to 27.3% (score 3) and LFS from 83.2% to 18.2%. Subgroup analyses further validated the model. For instance, patients with post-MRD+ and ≥CR2 had a 5-year CIR of 80.0%, whereas those with post-MRD− and CR1 had only 10.5% relapse. Similarly, cGVHD mitigated relapse risk even in high-risk subgroups, reducing CIR from 69.2% to 61.3% in post-MRD+ patients.

The study also explored interventions for MRD+ patients. Among 77 post-MRD+ cases, 60 received pre-emptive therapies, including immunosuppression tapering, tyrosine kinase inhibitors, donor lymphocyte infusion (DLI), or chimeric antigen receptor T-cell (CAR-T) therapy. Although the intervention group showed a non-significant trend toward lower CIR (68.1% vs. 75.2%), their OS (41.4% vs. 11.8%, P = 0.002) and LFS (31.9% vs. 15.7%, P = 0.039) were significantly better, suggesting clinical benefit from early MRD-directed therapy.

This risk score system offers a practical tool for post-transplant surveillance and decision-making. Low-risk patients (score 0–1) may require less intensive monitoring, while high-risk patients (score 2–3) could benefit from aggressive interventions, such as DLI or CAR-T therapy, to preempt relapse. The model’s reliance on readily available clinical variables (disease status, MRD, cGVHD) enhances its applicability across diverse transplant settings.

However, limitations include the retrospective design and single-center cohort, which may affect generalizability. The study did not account for molecular markers like BCR-ABL1, which could refine risk stratification. Additionally, the timing of cGVHD onset relative to MRD detection warrants further investigation, as earlier cGVHD may exert stronger GVL effects.

In conclusion, this study establishes a validated risk score system integrating post-MRD status, disease stage, and cGVHD to predict relapse in B-ALL patients post-allo-SCT. By identifying high-risk cohorts, this model facilitates tailored therapeutic strategies, potentially improving survival outcomes. Future prospective studies should validate these findings and explore integrating molecular biomarkers for enhanced precision.

doi.org/10.1097/CM9.0000000000001402

Was this helpful?

YesNo

0 / 0