New publication "Predicting Time to Relapse in Acute Myeloid Leukemia through Stochastic Modeling of Minimal Residual Disease Based on Clonality Data" in Computational and Systems Oncology, by Khanh N. Dinh, current IICD postdoctoral researcher in the laboratory of Professor Simon Tavaré.
Minimal residual disease (MRD), also called measurable residual disease, refers to the small number of leukemic cells that remain after treatment. Detected by flow cytometry or RT-qPCR, MRD can be used to predict risks of relapse and decreased survival rates for patients with acute myeloid leukemia (AML). However, MRD is poorly understood in AML, due to the latter unpredictability. In this study led by Marek Kimmel (Rice University) and Seth Corey (Cleveland Clinic), the authors developed a stochastic model that can track and predict the evolution of the clonal landscape of AML as it undergoes chemotherapy treatment and eventually relapses. Using independent training and validation data sets from two previously published studies, they demonstrate that the inter-patient heterogeneity of the clonal identity of relapse in AML could be explained by stochasticity. Indeed, clonal populations are diminished during treatment and therefore are subject to random events that may decide their fate. The model predicts a quantitative relationship between MRD and time to relapse (TTR) in AML, which would be of high relevance to clinicians. Comparison with data from clinical experiments shows agreement with the predicted MRD-TTR relationship. Current limitations with this model are the lack of appropriate datasets (experimental studies that track time-series genomic landscapes in AML). Application of the model with larger datasets could help develop a personalized tool to predict the chance and timing of relapse based on patients’ MRD, informing better treatment strategies.