A Comparison Of Long-Term Outcomes Of Donor Lymphocyte Infusions and Tyrosine Kinase Inhibitors In CML Patients Relapsed After Allogeneic Hemopoietic Stem Cell Transplantation

Abstract 5501

Relapse is a major cause of treatment failure of allogeneic hematopoietic cell transplantation (HCT) for Chronic Myeloid Leukemia (CML). DLI (Donor lymphocyte infusion) and TKI (Tyrosine Kinase inhibitors) are the two standard treatment options in this setting but reports comparing their long-term outcomes are scarce.

Between 1993 and 2012, 28 patients underwent DLI and 18 patients received TKI for chronic or advanced phase relapse of CML at our institution. Chronic hematologic, cytogenetic and molecular relapses were considered as chronic phase relapses. Accelerated phase and blast crisis relapses were considered as advanced phase relapses. Overall survival (OS), failure free survival (FFS) and cumulative incidence of failure (CIF) were analyzed retrospectively and for these analyses, failure was defined as lack of response, relapse or intolerance requiring change in treatment. This study had Institutional Research Ethics Board approval.

Among18 patients treated with TKI, 15 patients received imatinib, 2 received dasatinib and 1 received nilotinib as the first line treatment. In DLI group, number of infusions per patient ranged between 2 and 4. Patients in the DLI group were more likely to be transplanted during the 1st chronic phase of disease (86% vs. 56%, p=0.008), with bone marrow stem cells (100% vs. 56%, p=<0.0001) and less likely to have pre-HCT tyrosine kinase inhibitor treatment (0% vs. 33%, p=0.002). Patients in the TKI group are more likely to be treated during advance phases of relapse (39% vs. 7%, p=0.018). Type of donor, age at stem cell transplant or relapse and prior acute or chronic GVHD were not different between the groups. No patients received reduced intensity conditioning or T-cell depleted grafts.

64% of patients treated with DLI and 83% treated with TKIs achieved complete molecular response (CMR).  21% of patients initially treated with DLI required further treatment with TKIs after a second relapse. In the TKI group, no patients were later treated with DLI, but one required a change from imatinib to dasatinib due to intolerance. The incidences of treatment failures in DLI and TKI groups were 43% and 28% respectively.

At a median follow up of 146 (range, 40-220) and 70 (range, 9-139) months respectively, 9 (32%) in the DLI group and 6 (33%) in TKI group have died. when chronic phase relapses were analyzed separately 7 (21%) in the DLI group and none from the TKI treated group have died.  Mortality was higher in advanced phase relapses with 2 (100%) patients in the DLI group having died without achieving any response, and in the TKI group; 4 (57%) patients achieved CMR but eventually 6 out of 7 (86%) patients have died at a median time of 9 (3-114) months.

In multivariable analysis, advanced phase of relapse was associated with shorter OS (HR=109.7; 95% CI, 12.2-985.3; p= <0.0001), shorter FFS (HR = 82.3; 95% CI, 7.1-956.5; p =0.0004) and higher CIF (HR= 67.9; 95% CI, 6.9-669.8; p=0.0003). Comparing treatment with TKI, treatment with DLI was associated with an inferior OS (HR =37.4; 95% CI, 2.2-625.4; p=0.01), shorter FFS (HR =21.15; 95% CI, 1.8-251; p=0.02) and higher CIF (HR =19.5; 95% CI, 1.6-236.5; p=0.02). Prior acute GVHD grade 2-4 was associated with higher CIF (HR=3.3; 95% CI, 1.15-9.2; p= 0.03), but there was no effect on OS and FFS. Prior chronic GVHD, age at relapse treatment, and time from stem cell transplant to relapse treatment and year of relapse treatment were not found to be significant factors.

New onset or worsening of GVHD was significantly higher in the DLI group (68% vs. 6%, p=0.001). Among the patients treated with DLI, development of post DLI GVHD was associated with better OS (HR=0.26; 95% CI, 0.07-0.99; p=0.048), FFS (HR=0.3, 95% CI, 0.11-0.88; p=0.027) and lower CIF (HR=0.32; 95% CI, 0.1-0.96; p=0.04). Prior history of chronic GVHD was associated with higher FFS in DLI treated patients (HR=0.3; 95% CI, 0.1-0.89; p= 0.03), but did not have significant effect on OS (HR=0.32, 95% CI, 0.08-1.3; p=0.1) or CIF (HR=0.43; 95% CI, 0.143-1.28; p=0.13).

In conclusion, TKIs appears better than DLI in CML patients who had relapsed in chronic phase after myeloablative T-cell-replete HCT. Outcomes were poor in advanced phase relapses irrespective of modality and novel strategies need to be explored in this group of patients.