Abstract 227: Association between treatment response classification and resistance in metastatic prostate cancer patients treated with enzalutamide

• Published in: Cancer research (Chicago, Ill.) Ill.), 2021-07, Vol.81 (13_Supplement), p.227-227
• Main Authors: Turk, Maruša, Valentinuzzi, Damijan, Jeraj, Robert
• Format: Article
• Language: English
• ISSN: 0008-5472; 1538-7445
• DOI: 10.1158/1538-7445.AM2021-227

Introduction: Treatment resistance contributes importantly to clinical response of metastatic prostate cancer (mPC) patients. We hypothesize that patients with more resistant lesions would lead to more unfavorable treatment response. The goal of this study was to use a computational model to investigate the association between treatment response and individual lesion resistance to enzalutamide therapy. Materials and Methods: A deterministic population model was used to simulate dynamics of drug-sensitive and drug-resistant cells in individual lesions. Intrinsic and acquired resistance were taken into account. Model parametrization, assuming patient and lesions specific parameters, was performed by benchmarking the model to lesion-level burden of 365 lesions extracted from 18F-NaF PET of 7 mPC patients. Distributions of patient and lesion characteristics and model parameters were estimated using a maximum likelihood estimation. With Monte Carlo random sampling of a sample population of 1000 patients lesion-level characteristics were chosen. Afterwards, model simulated lesion-level burden. Lesions were classified as completely responding, partial responding (iPR), stable disease (iSD), progression disease (iPD) and new disease (iND) based on simulated lesion-level burden and measured uncertainties. The proportion of visible lesions (iPR, iSD, iPD and iND) consisting of only drug-resistant cells (NL) was calculated for individual patient. The percentage of patients having higher NL in comparison to the proportion of lesions classified as iSD, iPD, and iND was calculated. All the calculations were performed at 0.5, 1 and 2 years. Results: With Monte Carlo random sampling dynamics of a total of 36185 lesion burden was simulated. Median proportion of lesions classified as iSD, iPD, and iND at 0.5, 1 and 2 years was 0.56, 0.50, 0.54, respectively. Median NL at 0.5, 1 and 2 years was 0.09 (range: 0-1), 0.57 (range: 0.1-1) and 0.82 (range: 0.27-1), respectively. 1%, 59% and 90% of patients had higher NL in comparison to the occupancy of iSD, iPD, and iND at year 0.5, 1 and 2, meaning that some percentage of lesions classified as iPR disease are completely resistant. Median proportion of lesions classified as iPR disease and being completely resistant is 0.18, 0.17 and 0.3 at year 0.5, 1 and 2. Conclusions This work presents a computational model as a useful complementary tool in standard treatment response and therapy efficiency assessment. Contrary to our hypothes