Skip to Main Content
An official website of the United States government
CDAS has a New Look: On December 9th, the CDAS website was updated with a new design! The update incorporates all of the existing CDAS functionality with a more modern and user friendly interface.

Building Digital Twin Framework for Personalization of Prostate Cancer Management in Kidney Transplant Patients

Principal Investigator

Name
Naoru Koizumi

Degrees
Ph.D.

Institution
George Mason University

Position Title
Professor / Associate Dean of Research

Email
nkoizumi@gmu.edu

About this CDAS Project

Study
PLCO (Learn more about this study)

Project ID
PLCO-1997

Initial CDAS Request Approval
Nov 17, 2025

Title
Building Digital Twin Framework for Personalization of Prostate Cancer Management in Kidney Transplant Patients

Summary
Background
Active malignancy has traditionally been considered a contraindication for organ transplantation (1-2). Accordingly, most transplant centers require patients with active prostate cancer to undergo definitive treatment and remain cancer-free for a specified period before being placed on the kidney transplant waiting list. This practice persists despite the 2020 Kidney Disease: Improving Global Outcomes (KDIGO) guideline, which recommends that patients with indolent or low-grade prostate cancer may be waitlisted without definitive treatment (3). With the recent evidence demonstrating that patients with low- or intermediate-risk prostate cancer have excellent long-term survival (4-6), the current practice not only delays access to transplantation but also exposes patients to unnecessary risks such as urinary incontinence, erectile dysfunction, infection, and nerve damage (7). To improve transplant access to active prostate cancer patients, this project develops a data-driven digital twin framework used to personalize prostate cancer management in transplant candidates and recipients.

Significance
Kidney transplantation offers superior survival and quality-of-life outcomes for patients with end-stage renal disease (ESRD). Nonetheless, current practices exclude many ESRD patients with low-risk prostate cancer from timely transplantation, representing a major barrier to equitable care. This proposal directly addresses a critical clinical and policy gap by developing evidence-based, personalized approaches to prostate cancer management in transplant candidates, aligning transplant eligibility criteria based on the recent oncologic and nephrologic evidence.

Innovation
This project develops a data-driven digital twin framework that integrates Discrete Event Simulation (DES) with Artificial Intelligence (AI) and Machine Learning (ML) methodologies, augmented with novel approaches to risk modeling. Unlike static risk assessments, this framework will dynamically simulate individualized patient trajectories across cancer management, ESRD treatment as well as transplantation, providing both personalized and system-level insights. The integration of stochastic simulation and AI in the transplant-oncology field will facilitate a paradigm shift from one-size-fits-all eligibility criteria towards data-informed, patient-specific decision support.

Approach
The framework will leverage three publicly available datasets:
1) Prostate, Lung, Colorectal, and Ovarian (PLCO) Cancer Screening Trial dataset (NIH/NCI) – to model prostate cancer management and outcomes;
2) United States Renal Data System (USRDS) (NIH/NIDDK) – to model ESRD progression, treatment dynamics, and costs; and
3) United Network for Organ Sharing (UNOS) registry – to model transplant-related states and outcomes, including post-transplant recurrence and de novo prostate cancer incidence.

PLCO and UNOS datasets will inform complementary components of the model. Modeling ESRD and prostate cancer events independently is justified by robust evidence indicating that immunosuppressive therapy does not increase the progression risk of low- or intermediate-grade prostate cancer to metastatic disease.

The final product will generate:
• Patient-level outputs: individualized, optimal prostate cancer management protocols pre- and post-transplant; and
• System-level outputs: generalizable guidelines that transplant centers can adopt to improve equity and efficiency in waitlisting ESRD patients with active prostate cancer.

This research aims to replace uniform treatment protocols with personalized, data-driven recommendations—thereby improving both patient outcomes and system performance.

Aims

Aim 1: Assess the individual patients’ health outcomes under two scenarios: 1) follow the current guideline of definitive treatment for prostate cancer before being waitlisted; and 2) follow the personalized protocols for prostate cancer management pre- and post-transplant. Hypotheses: Individual health outcomes measured by Quality-Adjusted Life Year (QALY) are, overall, higher under the second scenario (H1.1); and patient and graft survival rates are, overall, higher under the second scenario (H1.2).

Aim 2: Assess equity and efficiency of the ESRD care system under two scenarios: 1) follow the current guideline of active treatment before being waitlisted; and 2) follow the personalized protocols for prostate cancer management. Hypotheses: fairness/equity in access to transplant will improve under the second scenario (H2.1); while efficiency in the ESRD care system measured by the total cost of treating ESRD patients is less for the second scenario (H2.2).

Aim 3: Establish the prostate cancer management protocols including: a) criteria for wait listing, i.e., PSA and Gleason scores, and remission duration; and b) post-transplant surveillance protocols including frequencies for PSA tests, digital rectal exams (DREs), prostate MRIs as well as biopsies.

Collaborators

Naoru Koizumi George Mason University
Meng-Hao Li George Mason University
Xiaoyu Chen SUNY
Michael Fu UMD
Obi Ekwenna Toledo
Ali Andalibi Charles R. Drew