Measurable residual disease in AML: Reduced-intensity conditioning associated with poorer post-transplant outcomes
“While allogeneic hematopoietic cell transplantation [allo-HCT] is curative for many patients with AML, many others relapse. By far the main reason for transplant failure is the primary disease itself, regardless of donor type and source,” said Dr Christopher Hourigan of the Laboratory of Myeloid Malignancies, National Heart, Lung, and Blood Institute, Bethesda, US. “One adjustable intervention prior to allo-HCT is the amount of conditioning: reduced-intensity conditioning [RIC] or high-intensity myeloablative conditioning [MAC],” he continued.
A retrospective study in over 5,000 patients with AML and myelodysplastic syndrome (MDS) previously demonstrated higher treatment-related mortality (TRM) with MAC than with RIC, while finding no difference in DFS and OS between the regimens. [Bone Marrow Transplant 2012;47:203-211]
However, the more recent prospective phase III 0901 clinical trial reported a substantially higher relapse rate among AML patients who received RIC prior to allo-HCT. [J Clin Oncol 2017;35:1154-1161] The study was stopped early, owing to a clear survival benefit with MAC.
“Strikingly, over half of AML patients receiving RIC relapsed within 18 months after transplant,” commented Hourigan.
In the current study, Hourigan and colleagues assessed MRD using a next-generation ultra-deep error-corrected genomic sequencing technology. Blood samples collected prior to conditioning from 188 patients who participated in the 0901 trial were tested for genomic variants across 13 commonly mutated genes in AML. The patients were well matched for baseline characteristics, including age, gender, comorbidity, disease risk, disease duration, cytogenetics, donor type and match, graft type, and antithymocyte globulin use. Presence of at least one genomic variant was interpreted as evidence of MRD. [Hourigan C, et al, EHA 2019, abstract LB2600]
Among patients with detectable variants prior to conditioning, the 3-year OS rate was significantly higher in the MAC vs RIC group (61 percent vs 44 percent; p=0.02). Patients with detectable variants randomized to receive RIC had a six-fold likelihood of relapse (hazard ratio [HR], 5.98; 95 percent confidence interval [CI], 3.19 to 11.26; p<0.001) and a nearly three-fold risk of decreased DFS (HR, 2.80; 95 percent CI, 1.76 to 4.44; p<0.001) compared with those who received MAC. Overall, of patients who experienced relapse, 76 percent had at least one detectable variant prior to conditioning.
“As we use higher sensitivity tools rather than morphology alone to assess residual disease burden, we get a better understanding of the underlying clinical reality,” commented Hourigan.
“However, the survival rates of patients with no genomic evidence of residual disease pretransplant did not differ, regardless of whether they were assigned to RIC or MAC,” said Hourigan.
Thirty-one percent of patients in the MAC group and 33 percent in the RIC group had no genomic variants indicative of MRD. Subsequent 3-year OS rate was 58 percent in the MAC group and 65 percent in the RIC group (p=0.98) – a finding that may be used in the future to drive therapeutic decisions and minimize TRM among suitable, MRD-free patients.