Optimal Treatment Selection in Sequential Systemic and Locoregional Therapy of Oropharyngeal Squamous Carcinomas: Deep Q-Learning With a Patient-Physician Digital Twin Dyad

Abstract

Background Currently, selection of patients for sequential versus concurrent chemotherapy and radiation regimens lacks evidentiary support and it is based on locally optimal decisions for each step. Objective We aim to optimize the multistep treatment of patients with head and neck cancer and predict multiple patient survival and toxicity outcomes, and we develop, apply, and evaluate a first application of deep Q-learning (DQL) and simulation to this problem. Methods The treatment decision DQL digital twin and the patient’s digital twin were created, trained, and evaluated on a data set of 536 patients with oropharyngeal squamous cell carcinoma with the goal of, respectively, determining the optimal treatment decisions with respect to survival and toxicity metrics and predicting the outcomes of the optimal treatment on the patient. Of the data set of 536 patients, the models were trained on a subset of 402 (75%) patients (split randomly) and evaluated on a separate set of 134 (25%) patients. Training and evaluation of the digital twin dyad was completed in August 2020. The data set includes 3-step sequential treatment decisions and complete relevant history of the patient cohort treated at MD Anderson Cancer Center between 2005 and 2013, with radiomics analysis performed for the segmented primary tumor volumes. Results On the test set, we found mean 87.35% (SD 11.15%) and median 90.85% (IQR 13.56%) accuracies in treatment outcome prediction, matching the clinicians’ outcomes and improving the (predicted) survival rate by +3.73% (95% CI –0.75% to 8.96%) and the dysphagia rate by +0.75% (95% CI –4.48% to 6.72%) when following DQL treatment decisions. Conclusions Given the prediction accuracy and predicted improvement regarding the medically relevant outcomes yielded by this approach, this digital twin dyad of the patient-physician dynamic treatment problem has the potential of aiding physicians in determining the optimal course of treatment and in assessing its outcomes.

A predictive modelling approach to illustrate factors correlating with stunting among children aged 12–23 months: a cluster randomised pre-post study

“…The sensitivity and specificity values, as well as the area under the ROC curve, were used to determine the overall accuracy of the algorithms in predicting childhood stunting. 28 29 …”

“…The same training and test datasets were used for evaluating the performance of all algorithms to ensure consistency in the results. The sensitivity and specificity values, as well as the area under the ROC curve, were used to determine the overall accuracy of the algorithms in predicting childhood stunting 28 29…”

Cause of Death in Patients with Oropharyngeal Carcinoma by Human Papillomavirus Status: Comparative Data Analysis

“…Alcohol and tobacco use have been confirmed as risk factors for HNSCC. HPV-related OPSCC cannot be considered in the same field of cancerization as HPV-negative OPSCC and other HNSCCs [ 2 - 5 ]. However, we observed a quite similar distribution of CODs between the 2 groups, with primary cancer as the major COD.…”

“…However, infection with carcinogenic HPV infection has emerged as an important risk factor that has driven an increase in the incidence of OPSCC. In recent decades, OPSCC has been one of the malignant tumors with rapidly rising incidences in Western countries [2][3][4][5]. In Italy, the incidence of HPV-related OPSCC increased from 16.7% in 2000-2006 to 46.1% in 2013-2018 [6].…”

Digital twins for health: a scoping review

“…In the field of clinical oncology, digital twin blueprints and examples of practical development of digital twin framework focusing on image-guided mechanisms and adaptive radiotherapy for high-grade gliomas have been proposed 60 . In a recent study, patient-physician digital twin dyads were developed to simulate therapy outcomes and determine the optimal treatment selection for oropharyngeal carcinoma, specifically whether patients would benefit from sequential versus concurrent chemotherapy and radiation with high accuracy 61 . In another study with triple-negative breast cancer (TNBC), quantitative MRI and biologically based mathematical modeling were used to successfully predict responses to the neoadjuvant chemotherapy, suggesting digital twins could predict and optimize cancer therapy 62 .…”