[Effects of chemotherapy combined with donor lymphocyte infusion on chronic graft-versus-host disease and prognosis in minimal residual disease positive patients after allogeneic hematopoietic stem cell transplantation].

Objective: To explore clinical features and severity of chronic graft- versus- host disease (cGVHD) after chemotherapy plus donor lymphocyte infusion (Chemo-DLI) in a consecutive cohort of acute leukemia patients who were minimal residual disease (MRD) positive after allogeneic hematopoietic stem cell transplantation (allo-HSCT) . Methods: The global scoring system proposed by National Institutes of Health (NIH) Consensus Conference was used to identify the characteristics and severity of cGVHD in patients who MRD positive after Chemo-DLI. Results: 54 (59.3%) patients were diagnosed with cGVHD after Chemo-DLI, with the median time of onset of 70 (13-504) days. There were 6 cases (6.6%) of mild cGVHD, 21 cases (23.1%) of moderate cGVHD and 27 cases (29.7%) of severe cGVHD.The 5-year cumulative incidence of relapse after Chemo-DLI was 61.9% (95% CI 45.3%-78.5%) , 15.1% (95% CI 1.1%-29.1%) , and 26.6% (95% CI 9.2%-44.0%) ( χ (2)=18.901, P <0.001) in non-cGVHD, mild to moderate cGVHD, and severe cGVHD groups, respectively. The 5-year cumulative incidence of relapse after Chemo-DLI was 61.9% (95% CI 45.3%-78.5%) , 19.9% (95% CI 8.1%-31.7%) , and 28.6% (95% CI 0.0%-65.0%) ( χ (2)=18.307, P <0.001) in non-cGVHD, classical cGVHD, and overlap syndrome groups, respectively. cGVHD was not associated with non-relapse morality after Chemo-DLI. Probabilities of 5-year leukemia-free survival (LFS) after Chemo-DLI were 24.0% (95% CI 9.1%-38.9%) , 77.2% (95% CI 60.8%-93.6%) , and 64.9% (95% CI 45.7%-84.1%) ( χ (2)=24.447, P <0.001) in non-cGVHD, mild to moderate cGVHD, and severe cGVHD groups, respectively. Probabilities of 5-year LFS after Chemo-DLI were 24.0% (95% CI 9.1%-38.9%) , 75.5% (95% CI 62.7%-88.3%) , and 42.9% (95% CI 1.8%-84.0%) ( χ (2)=25.665, P <0.001) in non-cGVHD, classical cGVHD, and overlap syndrome groups, respectively. Probabilities of 5-year overall survival (OS) after Chemo-DLI were 50.0% (95% CI 31.1%-68.9%) , 87.9% (95% CI 74.7%-100.0%) , and 71.0% (95% CI 52.0%-90.0%) ( χ (2)=9.517, P =0.009) in non-cGVHD, mild to moderate cGVHD, and severe cGVHD groups, respectively. Probabilities of 5-year OS after Chemo-DLI were 50.0% (95% CI 31.1%-68.9%) , 83.9% (95% CI 72.8%-95.0%) , and 51.4% (95% CI 6.2%-96.6%) ( χ (2)=10.673, P =0.005) in non-cGVHD, classical cGVHD, and overlap syndrome groups, respectively. In multivariate analysis, patients receiving allo-HSCT in first complete remission stage and classical cGVHD after Chemo-DLI were associated with lower relapse risk and better survival. Conclusions: These findings highlight the close relation between cGVHD and the graft-versus-leukemia effect in patients who were MRD positive and received Chemo-DLI after allo-HSCT. However, overlap syndrome could not improve the clinical outcomes of these patients.