Development of a Universal Assay for Minimal Residual Disease in Acute Myeloid Leukemia using Duplex Sequencing

Problem: With a 5-year survival of ~30%, AML is the 6th deadliest cancer, and >20,000 new patients are
diagnosed each year in the United States. Although most patients achieve Complete Remission (CR) with
aggressive therapy, most will eventually relapse. The criteria for CR, however, is based on historical diagnostic
methods, and patients in CR may carry up to 1010 residual leukemia blasts. Significant effort has gone into
developing tools to detect Minimal amounts of Residual Disease (MRD), including multi-parametric flow
cytometry (MFC) and polymerase chain reaction (PCR). MRD is the strongest predictor of relapse, and several
AML trials have demonstrated that survival was significantly better when MRDpos patients were subjected to
intensified therapy. Yet, lack of sensitivity is a clear problem with MFC and PCR. For example, relapses occur
in ~40% of patients who are MRDneg by MFC after chemotherapy. Next Generation Sequencing (NGS) holds the
promise to identify MRD by detecting mutations associated with residual AML cells. Yet, the sensitivity of
conventional NGS is limited by a relatively high error rate, which makes it difficult to differentiate sequencing
errors from true low-frequency mutations. Solution: Duplex Sequencing (DS) is an ultra-sensitive NGS
technology which uses specialized biochemistry and informatics to improve the accuracy of standard DNA
sequencers by more than ten-thousand-fold. In our prior Phase I SBIR study, we developed a broadly-applicable
DS-based AML MRD assay that overcomes the above limitations. The DS assay targets many MRD-relevant
genes simultaneously, with sensitivity and specificity rivaling or exceeding single-gene PCR assays. In Phase II,
we demonstrated excellent reproducibility across labs, applied the optimized assay to retrospectively banked
cohorts, and showed superior clinical performance vs. MFC. However, based on many customer interviews, the
most substantial barrier to widespread commercial adoption we face is the lack of large-scale prospective clinical
trial data. Specific Aims: In the present Phase 2b application we propose generating this comprehensive clinical
utility data set through collaboration with the world-class US National Marrow Donor Program (NMDP) and other
top AML MRD key opinion leaders. The primary goals of our proposed study include: Aim 1: Prospectively
validate the prognostic value of pre-transplantation DS AML MRD testing for predicting post-transplant relapse
and survival; Aim 2: Prospectively generate data supporting the ability of post-transplantation DS AML MRD
testing to assess relapse risk for patients treated with different transplant conditioning intensities; and Aim 3:
Demonstrate the potential of using tumor-informed DS AML MRD testing for patients without a driver gene variant
targeted by our fixed gene panel. Impact: This combined approach may bring NGS MRD testing up to 100%
patient applicability. Blood MRD comparison may present an equally sensitive but less invasive alternative to
marrow. Our anticipated data will also support the use of the technology in assigning high-risk patients to
intensified therapies and as an early marker of anti-leukemic efficacy for novel drug development efforts.

Grant Number: 5R44CA233381-06
NIH Institute/Center: NIH
Principal Investigator: Kellie Bickel