TY - JOUR
T1 - Prostate cancer therapy personalization via multi-modal deep learning on randomized phase III clinical trials
AU - NRG Prostate Cancer AI Consortium
AU - Esteva, Andre
AU - Feng, Jean
AU - van der Wal, Douwe
AU - Huang, Shih Cheng
AU - Simko, Jeffry P.
AU - DeVries, Sandy
AU - Chen, Emmalyn
AU - Schaeffer, Edward M.
AU - Morgan, Todd M.
AU - Sun, Yilun
AU - Ghorbani, Amirata
AU - Naik, Nikhil
AU - Nathawani, Dhruv
AU - Socher, Richard
AU - Michalski, Jeff M.
AU - Roach, Mack
AU - Pisansky, Thomas M.
AU - Monson, Jedidiah M.
AU - Naz, Farah
AU - Wallace, James
AU - Ferguson, Michelle J.
AU - Bahary, Jean Paul
AU - Zou, James
AU - Lungren, Matthew
AU - Yeung, Serena
AU - Ross, Ashley E.
AU - Kucharczyk, Michael
AU - Souhami, Luis
AU - Ballas, Leslie
AU - Peters, Christopher A.
AU - Liu, Sandy
AU - Balogh, Alexander G.
AU - Randolph-Jackson, Pamela D.
AU - Schwartz, David L.
AU - Girvigian, Michael R.
AU - Saito, Naoyuki G.
AU - Raben, Adam
AU - Rabinovitch, Rachel A.
AU - Katato, Khalil
AU - Sandler, Howard M.
AU - Tran, Phuoc T.
AU - Spratt, Daniel E.
AU - Pugh, Stephanie
AU - Feng, Felix Y.
AU - Mohamad, Osama
N1 - Publisher Copyright:
© 2022, The Author(s).
PY - 2022/12
Y1 - 2022/12
N2 - Prostate cancer is the most frequent cancer in men and a leading cause of cancer death. Determining a patient’s optimal therapy is a challenge, where oncologists must select a therapy with the highest likelihood of success and the lowest likelihood of toxicity. International standards for prognostication rely on non-specific and semi-quantitative tools, commonly leading to over- and under-treatment. Tissue-based molecular biomarkers have attempted to address this, but most have limited validation in prospective randomized trials and expensive processing costs, posing substantial barriers to widespread adoption. There remains a significant need for accurate and scalable tools to support therapy personalization. Here we demonstrate prostate cancer therapy personalization by predicting long-term, clinically relevant outcomes using a multimodal deep learning architecture and train models using clinical data and digital histopathology from prostate biopsies. We train and validate models using five phase III randomized trials conducted across hundreds of clinical centers. Histopathological data was available for 5654 of 7764 randomized patients (71%) with a median follow-up of 11.4 years. Compared to the most common risk-stratification tool—risk groups developed by the National Cancer Center Network (NCCN)—our models have superior discriminatory performance across all endpoints, ranging from 9.2% to 14.6% relative improvement in a held-out validation set. This artificial intelligence-based tool improves prognostication over standard tools and allows oncologists to computationally predict the likeliest outcomes of specific patients to determine optimal treatment. Outfitted with digital scanners and internet access, any clinic could offer such capabilities, enabling global access to therapy personalization.
AB - Prostate cancer is the most frequent cancer in men and a leading cause of cancer death. Determining a patient’s optimal therapy is a challenge, where oncologists must select a therapy with the highest likelihood of success and the lowest likelihood of toxicity. International standards for prognostication rely on non-specific and semi-quantitative tools, commonly leading to over- and under-treatment. Tissue-based molecular biomarkers have attempted to address this, but most have limited validation in prospective randomized trials and expensive processing costs, posing substantial barriers to widespread adoption. There remains a significant need for accurate and scalable tools to support therapy personalization. Here we demonstrate prostate cancer therapy personalization by predicting long-term, clinically relevant outcomes using a multimodal deep learning architecture and train models using clinical data and digital histopathology from prostate biopsies. We train and validate models using five phase III randomized trials conducted across hundreds of clinical centers. Histopathological data was available for 5654 of 7764 randomized patients (71%) with a median follow-up of 11.4 years. Compared to the most common risk-stratification tool—risk groups developed by the National Cancer Center Network (NCCN)—our models have superior discriminatory performance across all endpoints, ranging from 9.2% to 14.6% relative improvement in a held-out validation set. This artificial intelligence-based tool improves prognostication over standard tools and allows oncologists to computationally predict the likeliest outcomes of specific patients to determine optimal treatment. Outfitted with digital scanners and internet access, any clinic could offer such capabilities, enabling global access to therapy personalization.
UR - http://www.scopus.com/inward/record.url?scp=85134004361&partnerID=8YFLogxK
U2 - 10.1038/s41746-022-00613-w
DO - 10.1038/s41746-022-00613-w
M3 - Article
C2 - 35676445
AN - SCOPUS:85134004361
SN - 2398-6352
VL - 5
JO - npj Digital Medicine
JF - npj Digital Medicine
IS - 1
M1 - 71
ER -