ATLANTA – Clinicians may be able to get a jump on identifying risk factors for relapse among adults with acute myeloid leukemia (AML) in first complete remission through the use of next-generation DNA sequencing, investigators reported.
Among 430 adults with AML with somatic driver mutations persistent in bone marrow during morphological complete remission (CR) following induction therapy, the presence of minimal residual disease (MRD) bearing specific disease-related mutations on next-generation sequencing (NGS) was significantly associated with both the cumulative incidence of relapse and with overall survival. Tim Grob, MD, of Erasmus University Medical Center in Rotterdam, the Netherlands, reported the findings during a late-breaking abstract session at the annual meeting of the American Society of Hematology.
Neil Osterweil/Frontline Medical News
Dr. Tim Grob
By excluding three common AML mutations in genes commonly associated with clonal hematopoiesis – DNMT3A, TET2, and ASXL1 (collectively, DTA) – Dr. Grob and his colleagues at centers in the Netherlands, Belgium, and Switzerland were able to demonstrate that non-DTA mutations present in the marrow of patients in CR are highly predictive for relapse within 5 years and for worse overall survival.
They also showed that mutations associated with clonal hematopoiesis (the presence of small, preleukemic clones) in CR is not significantly associated with risk of relapse.
More than 80% of patients with AML are able to have a CR after induction therapy, but a significant proportion of patients will also experience relapse. Investigators have yet to nail down which leukemia-specific mutations that linger in patients with CR may be responsible for subsequent relapses, Dr. Grob said.
To get a better handle on which residual mutations may signal the need for extra vigilance or additional therapy in patients in CR after two cycles of induction therapy, the investigators used targeted next-generation (high-throughput) sequencing at the time of diagnosis and first CR in 430 patients enrolled in joint Dutch/Swiss clinical trials. The median patient age was 51.
The investigators screened marrow samples using a commercially available gene panel (Illumina) covering 54 genes that are commonly mutated in myeloid malignancies.
They divided the patients into a training cohort (283 patients) and a validation set (147) for confirmation of results.
About half of all patients in the training cohort (51.4%) had persistent mutations in bone marrow that occurred with highly variable variant allele frequencies. The most common mutations were in the DTA group with the most frequently mutated gene being DNMT3A (78.7% variant allele frequency), followed by TET2 (54.2%) and ASXL1 (51.6%).
Mutations in DTA genes in this cohort were not associated with the incidence of relapse at any variant allele frequency cut-off point used, which indicated that these mutations represented a stage of clonal hematopoiesis rather than early relapse signals.
However, among patients who had persistent DTA mutations, there was significant correlation with a risk for relapse when they also had persistence of any other non-DTA mutations. The cumulative 5-year incidence of relapse in patients with both persistent DTA and non-DTA mutations was 76.4%, compared with 39.4% for those without other, non-DTA mutations (P = .002).
Also in the training cohort, persistent non-DTA mutations (NGS MRD) were found to be highly associated with the risk of relapse with a subdistribution hazard ratio (SHR) of 1.85 (P = .001). In the validation set the effect was even stronger, with an SHR or 2.81 (P less than .001).
When data from the training and validation cohort were combined, the 5-year cumulative incidence of relapse was 58.3% for non-DTA mutation, vs. 33.9% (P less than .001).
NGS MRD was also predictive of overall survival, with a hazard ratio in the training cohort of 1.64 (P = .012) and an HR in the validation cohort of 3.08 (P less than .001).
In multivariable analysis of all 430 patients, adjusted for age, white blood cell count, 2017 European LeukemiaNet risk category, and number of induction cycles need to achieve CR, NGS MRD was an independent prognostic factor for both relapse (SHR 1.89, P less than .001) and overall survival (HR 1.64 P = .003).
When the investigators conducted a sensitivity analysis with time-dependent correction for allogeneic stem cell transplantation, they found that NGS MRD was still significantly prognostic for both relapse and survival.
The study was supported by the Dutch Cancer Society, the Haemato-Oncology Foundation for Adults in the Netherlands, the Swiss Group for Clinical Cancer Research, and The Netherlands Organization for Health Research and Development. Dr. Grob reported having no relevant disclosures.
SOURCE: Jongen-Lavrencic M et al. ASH 2017 Abstract LBA 5.