Early T-Lymphocyte Chimerism Kinetics Is Influenced By Conditioning Regimen in Reduced Intensity Allogeneic Stem Cell Transplantation

Introduction: Disease relapse remains one of the major causes of treatment failure in hematopoietic stem cell transplant (HSCT) following reduced-intensity conditioning (RIC) regimens.  Donor chimerism assessment early after HSCT may help identify patients at risk for relapse who may benefit from additional therapeutic interventions but its precise role in this setting is unclear. In addition, the differential impact of conditioning regimen is not well studied. 

Aim:

1.      Study the impact of early full T-lymphocyte chimerism (T-lymphocyte chimerism ≥ 90% on day 30 to 60) on relapse incidence (RI) and acute graft versus host disease (aGVHD) in patients with acute myeloid leukemia (AML) or myelodysplastic syndrome (MDS) following RIC-HSCT.

2.      Looking at the tempo of achievement of complete donor T-lymphocyte chimerism based on the two most common RIC regimens, Fludarabine/Melphalan (FM) and Fludarabine/Busulfan (FB).

Method:

After IRB approval, patients with AML or MDS who underwent RIC-HSCT with FB or FM between 2008 and 2014 at our institution were identified. Post-transplantation CD3 (lymphocyte) and CD33 (myeloid) donor chimerism were measured per protocol using PCR-based analysis of short tandem repeats. Categorical and continuous variables were analyzed using Pearson’s chi-squared and Wilcoxon / Kruskal-Wallis, respectively. Survival estimates were calculated using the Kaplan-Meier and log-rank test. RI was estimated accounting NRM on a competing event analysis using Grey’s model. Cox proportional regression was used to compute hazard ratios (HR).

Results:

Baseline characteristics: Out of 134 patients who received FM or FB, 75 patients had evaluable early sorted chimerism data.  42 and 33 patients received FB and FM, respectively. Patient, disease and transplant variables were balanced in both FM and FB (Table 1). The median follow up of the cohort was 37 months (2.4-43) and at last follow up 29.1% of patients relapsed and 59% died. All patients achieved early complete myeloid chimerism whereas T-lymphocyte chimerism was complete in 50 patients and mixed in 25 patients. Characteristics of early lymphocyte chimerism by status are shown in table 1. Notably, likelihood of achieving complete T-lymphocyte chimerism ≥ 90% at 60 days was 100% versus 47% in FM and FB respectively (p <0.0001).

Acute graft versus host disease:  There was no significant difference in incidence of grade II-IV aGVHD between patients with full early donor T-lymphocyte chimerism in comparison to those with early mixed donor chimerism (46% vs. 48%, p =0.87).

Impact of full chimerism and conditioning regimen on RI  and overall survival (OS): Among the entire cohort, patients achieving early full T-lymphocyte chimerism had a significantly lower RI compared to those with mixed chimerism with a HR of 0.38 (CI 0.14-0.97; p=0.044) but OS was not significantly different with HR 0.69 (CI 0.33-1.51; p=0.34). In patients conditioned with FB, there was no difference in 2-year RI between complete or mixed early T-lymphocyte chimerism at 38.2% and 40.7% respectively (p = 0.9). After FM conditioning, all patients achieved early T-lymphocyte chimerism and the 2-year RI was 6.9%.  Achievement of early complete lymphocyte chimerism post FB in comparison to FM was associated with significantly higher risk of relapse (HR 8.6 CI 2.07-57, p=0.0025). Furthermore, there was a trend towards worse OS with FB vs. FM with a median of 23 vs. 44 months, respectively (p = 0.058).

Conclusion:

Similar to prior reports, we observed that full T-lymphocyte chimerism is associated with decreased relapse risk without an increase in aGVHD. Our findings also suggest that RIC with FM yields a higher incidence of early complete T-lymphocyte chimerism and is significantly associated with decreased risk of relapse compared to FB. There was a greater risk of relapse in the FB group, even among those achieving early complete donor lymphoid chimerism. This needs further validation given the small sample size.