The Utility Of Monitoring Chimerism: More Predictive For Nonablative Regimens?

Treatment related mortality and overall survival of allogeneic stem cell transplants have improved over the past 30 years but disease relapse remains the primary cause of death. New tools are sought to detect a window for early intervention. Chimerism studies are one tool that may allow clinicians to recognize impending graft failure or disease recurrence.

Use of chimerism data varies in terms of frequency of monitoring and decision-making based on results. In this single-center study we evaluated the use and utility of chimerism studies in patients undergoing ablative and nonablative stem cell transplants for hematologic malignancies.

Following IRB approval, a retrospective chart review of 133 patients undergoing allogeneic stem cell transplant for leukemia or MDS at our institution during the period from 2007-2012 was performed. Of this group, 90 had chimerism studies performed using the Promega PowerPlex 16 System. Full donor chimerism (FDC) was defined as donor hematopoeisis of greater than or equal to 95% in both CD3 and CD33 lineages. Mixed donor chimerism (MDC) was defined as donor hematopoeisis of less than 95%. Patients with FDC at all assessments were categorized as complete chimerism (CC). Patients with MDC initially that improved on subsequent evaluation were categorized as improving chimerism (IC). Patients with FDC that devolved into stable MDC or worsening MDC were categorized as worsening chimerism (WC). Fifteen patients were unclassifiable due to incomplete data and were excluded. A univariate statistical analysis was performed on the remaining 75 patients and overall survival (OS), disease specific survival (DSS), and relapse free survival (RFS) determined. A sub-analysis using Chi-Square and Fisher's Exact Test was performed on individuals in the nonablative cohort with WC or IC to investigate whether there were higher rates of intervention in the IC group.

Demographic data for the ablative cohort and nonablative cohort, and among patients with CC, IC and WC were similar. There was no statistically significant difference in OSS or DSS between patients with CC, IC and WC. There was a statistically significant hazard ratio favoring both CC and IC versus WC in terms of RFS in the nonablative cohort and in the combined cohort but this was not seen in the ablative cohort (Table 1). Within the nonablative group, there was not a higher rate of withdrawal of immunosuppression (p=0.98), donor lymphocyte infusion (p=0.43), or additional disease-specific therapy (p=0.43) in the IC versus the WC group to explain the improving chimerism, suggesting that interventions based on chimerism results are either too late or are inadequate.

This study suggests that chimerism studies have utility in the setting of nonablative transplant where WC predicts worse RFS. Despite this additional prognostic information, the window for early intervention is narrow and we were unable to demonstrate that improvements in chimerism studies were related to clinician initiated interventions in this small study. The risk of developing GVHD must be weighed in deciding to withdraw immunosuppression or administer DLI given the relatively low success rate of changing MDC into FDC. Finally, despite our institutions' recommendation of monitoring chimerism studies serially, 32% of the charts we reviewed had not had these studies performed. Given the complexity and heterogeneity of patients with hematological malignancies, more evaluation of chimerism studies is needed to fully establish or discredit the potential utility of this test.

No relevant conflicts of interest to declare.