Predicting survival in prospective clinical trials using weakly-supervised QSP

Abstract Quantitative systems pharmacology (QSP) models of cancer immunity provide mechanistic insights into cellular dynamics and drug effects that are difficult to study clinically. However, their inability to predict patient survival mechanistically limits their utility in anti-cancer drug develo...

Full description

Saved in:
Bibliographic Details
Main Authors: Matthew West, Kenta Yoshida, Jiajie Yu, Vincent Lemaire
Format: Article
Language:English
Published: Nature Portfolio 2025-04-01
Series:npj Precision Oncology
Online Access:https://doi.org/10.1038/s41698-025-00898-6
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1850156655706112000
author Matthew West
Kenta Yoshida
Jiajie Yu
Vincent Lemaire
author_facet Matthew West
Kenta Yoshida
Jiajie Yu
Vincent Lemaire
author_sort Matthew West
collection DOAJ
description Abstract Quantitative systems pharmacology (QSP) models of cancer immunity provide mechanistic insights into cellular dynamics and drug effects that are difficult to study clinically. However, their inability to predict patient survival mechanistically limits their utility in anti-cancer drug development. To overcome this, we link virtual patients from a QSP model to real clinical trial patients. Using data from atezolizumab trials in non-small cell lung cancer, we show that tumor-based linkage effectively captures survival outcomes. By treating linked survival and censoring as weak supervision labels, we trained survival models using only QSP model covariates, without clinical covariates. Our approach also predicts survival for treatments not included in training data. Specifically, we accurately estimated survival hazard ratios (HR) for chemotherapy monotherapy and atezolizumab plus chemotherapy combination. The predicted HR of 0.70 (95% prediction interval [PI] 0.55–0.86) closely matches the observed HR of 0.79 (95% PI 0.64–0.98) from the IMpower130 trial.
format Article
id doaj-art-d6cd18e2502d4afb9866f24cd1757e38
institution OA Journals
issn 2397-768X
language English
publishDate 2025-04-01
publisher Nature Portfolio
record_format Article
series npj Precision Oncology
spelling doaj-art-d6cd18e2502d4afb9866f24cd1757e382025-08-20T02:24:26ZengNature Portfolionpj Precision Oncology2397-768X2025-04-01911810.1038/s41698-025-00898-6Predicting survival in prospective clinical trials using weakly-supervised QSPMatthew West0Kenta Yoshida1Jiajie Yu2Vincent Lemaire3Department of Biostatistics, Harvard T.H. Chan School of Public HealthModeling and Simulation, Clinical Pharmacology, Genentech Inc.Clinical Pharmacology, Genentech Inc.Modeling and Simulation, Clinical Pharmacology, Genentech Inc.Abstract Quantitative systems pharmacology (QSP) models of cancer immunity provide mechanistic insights into cellular dynamics and drug effects that are difficult to study clinically. However, their inability to predict patient survival mechanistically limits their utility in anti-cancer drug development. To overcome this, we link virtual patients from a QSP model to real clinical trial patients. Using data from atezolizumab trials in non-small cell lung cancer, we show that tumor-based linkage effectively captures survival outcomes. By treating linked survival and censoring as weak supervision labels, we trained survival models using only QSP model covariates, without clinical covariates. Our approach also predicts survival for treatments not included in training data. Specifically, we accurately estimated survival hazard ratios (HR) for chemotherapy monotherapy and atezolizumab plus chemotherapy combination. The predicted HR of 0.70 (95% prediction interval [PI] 0.55–0.86) closely matches the observed HR of 0.79 (95% PI 0.64–0.98) from the IMpower130 trial.https://doi.org/10.1038/s41698-025-00898-6
spellingShingle Matthew West
Kenta Yoshida
Jiajie Yu
Vincent Lemaire
Predicting survival in prospective clinical trials using weakly-supervised QSP
npj Precision Oncology
title Predicting survival in prospective clinical trials using weakly-supervised QSP
title_full Predicting survival in prospective clinical trials using weakly-supervised QSP
title_fullStr Predicting survival in prospective clinical trials using weakly-supervised QSP
title_full_unstemmed Predicting survival in prospective clinical trials using weakly-supervised QSP
title_short Predicting survival in prospective clinical trials using weakly-supervised QSP
title_sort predicting survival in prospective clinical trials using weakly supervised qsp
url https://doi.org/10.1038/s41698-025-00898-6
work_keys_str_mv AT matthewwest predictingsurvivalinprospectiveclinicaltrialsusingweaklysupervisedqsp
AT kentayoshida predictingsurvivalinprospectiveclinicaltrialsusingweaklysupervisedqsp
AT jiajieyu predictingsurvivalinprospectiveclinicaltrialsusingweaklysupervisedqsp
AT vincentlemaire predictingsurvivalinprospectiveclinicaltrialsusingweaklysupervisedqsp