Upcoming Event: Oden Institute Seminar

Personalized mathematical modeling of neoadjuvant chemotherapy response in triple-negative breast cancer using multiparametric MRI

Mahesh Nandyala, Postdoctoral Fellow, Oden Institute

Abstract

Triple-negative breast cancer (TNBC) is considered the most aggressive breast cancer subtype. TNBC patients have highly variable responses to neoadjuvant chemotherapy (NAC), which often makes treatment decisions challenging. By combining tissue-scale reaction-diffusion equations for tumor growth dynamics with cellular-scale drug transport and uptake kinetics, we developed a multiscale mathematical modeling framework that forecasts NAC response for TNBC patients from the ARTEMIS clinical trial (NCT02276443). 
We calibrated the model on a patient-specific basis with (a) dynamic contrast-enhanced and diffusion-weighted MRI data collected at baseline and mid-treatment, and (b) in vitro drug uptake data. The model predicted individual patient responses to the standard doxorubicin-cyclophosphamide protocol with prediction accuracy. We subsequently investigated alternative total dosages, ranging from 180 to 300 mg/m2 for doxorubicin, 1800 to 3000 mg/m2 for cyclophosphamide, and dosage schedules with 7-14 days between two treatment cycles.
Our model achieved excellent predictive accuracy, with concordance correlation coefficients between the measured and predicted data of 0.96 for tumor cellularity and 0.91 for volume changes. Among 50 patients, 23 achieved clinically meaningful tumor shrinkage (≥70%) with alternative regimens while requiring 25% less total drug dose than standard-of-care protocols. Notably, two patients who failed to respond adequately to standard-of-care treatment achieved threshold response under optimized alternatives. However, 25 patients demonstrated poor response across all regimens, indicating potential intrinsic resistance.
These findings demonstrate that integrating imaging data with mathematical modeling offers a promising pathway toward personalized cancer therapy, enabling clinicians to optimize treatment efficacy and identify those patients most likely to benefit from alternative therapeutic approaches early in the course of therapy.

Biography

Mahesh Nandyala received his Ph.D. in Mechanical Engineering from the Indian Institute of Technology Delhi, India, in 2023, where he developed computational models for magnetic nanoparticle hyperthermia in cancer therapy. He also holds an M.Tech. from NIT Jalandhar, India, and a B.Tech. from RGUKT, RK Valley. He is currently a joint postdoctoral fellow at The University of Texas at Austin (Yankeelov Lab) and MD Anderson Cancer Center (Fuentes Lab). At UT Austin, his research focuses on optimizing neoadjuvant therapy (NAT) for breast cancer by integrating MRI-derived data with pharmacokinetic and mechanistic models to predict treatment response. At MD Anderson, he develops bioheat transfer models to simulate magnetic hyperthermia in rodent tumor models, aiming to improve thermal dose delivery through computational prediction. His work bridges computational modeling and clinical translation, contributing to the development of patient-specific cancer treatment strategies.