npj Digital Medicine (Jun 2022)

Prostate cancer therapy personalization via multi-modal deep learning on randomized phase III clinical trials

  • Andre Esteva,
  • Jean Feng,
  • Douwe van der Wal,
  • Shih-Cheng Huang,
  • Jeffry P. Simko,
  • Sandy DeVries,
  • Emmalyn Chen,
  • Edward M. Schaeffer,
  • Todd M. Morgan,
  • Yilun Sun,
  • Amirata Ghorbani,
  • Nikhil Naik,
  • Dhruv Nathawani,
  • Richard Socher,
  • Jeff M. Michalski,
  • Mack Roach,
  • Thomas M. Pisansky,
  • Jedidiah M. Monson,
  • Farah Naz,
  • James Wallace,
  • Michelle J. Ferguson,
  • Jean-Paul Bahary,
  • James Zou,
  • Matthew Lungren,
  • Serena Yeung,
  • Ashley E. Ross,
  • NRG Prostate Cancer AI Consortium,
  • Howard M. Sandler,
  • Phuoc T. Tran,
  • Daniel E. Spratt,
  • Stephanie Pugh,
  • Felix Y. Feng,
  • Osama Mohamad

DOI
https://doi.org/10.1038/s41746-022-00613-w
Journal volume & issue
Vol. 5, no. 1
pp. 1 – 8

Abstract

Read online

Abstract 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.