Population-Scale Precision Safety in Oncology Reveals Clinical and Genetic Determinants of Systemic Therapy Toxicity
- 1Department of Epidemiology and Biostatistics, Memorial Sloan Kettering Cancer Center, New York, NY, USA
Abstract
Treatment toxicity constrains the use of effective cancer therapies, but its clinical and genetic determinants remain poorly defined. We developed a large language model-based approach to produce MSK-Tox, a pan-cancer resource capturing the incidence, temporality, and grade of toxicity to anti-cancer therapy across more than 50,000 patients. Analysis of six representative adverse events--pneumonitis, adrenal insufficiency, liver toxicity, colitis, hyperthyroidism and hypothyroidism--revealed distinct toxicity landscapes shaped by cancer type, treatment regimen, and clinical context. Pretreatment clinical features enabled individualized prediction of toxicity risk across adverse events, supporting risk assessment before therapy initiation. Beyond clinical predictors, we identified two modes of germline susceptibility to treatment toxicity: an organintrinsic mode, in which germline variation confers risk across systemic therapies, exemplified by a regulatory variant near FOXE1 associated with hypothyroidism; and an immune-mediated mode, confined to immune checkpoint inhibitor-treated patients, in which HLA-DRB1*15 was a major determinant of adrenal insufficiency. Notably, the same allele predisposes to multiple sclerosis in individuals without cancer, indicating that immune checkpoint inhibition unmasks a latent autoimmune predisposition. These findings provide an empirical basis for a new precision safety paradigm for predicting who will be harmed by a therapy on the same principles that guide prediction of therapeutic benefit.
Related
- Unravelling the biological divergence of CD30+ lymphoproliferative disorders with skin involvement: from clinical risk to precision oncology
- Clinical Translation of Artificial Intelligence in Oncology: Real-World Validation, Workflow Integration, and Precision Medicine Applications
- Digital Twins in Oncology: From Multimodal Data Integration to Precision Clinical Practice.
- Clinical Artificial Intelligence in Non-Small Cell Lung Cancer: From Early Detection and Computational Pathology to Precision Oncology
- Patient-derived gastric cancer organoids for functional precision oncology: disease-specific bottlenecks and clinical readiness.
BibTeX
@misc{bakouny2026populationscale,
title = {{Population-Scale Precision Safety in Oncology Reveals Clinical and Genetic Determinants of Systemic Therapy Toxicity}},
author = {Z. Bakouny and X. A. Guo and F. Huang and S. Mohan and R. Walser and Z. Lu and T. Perea-Chamblee and L. Khan and N. Ahmed and A. Ocejo and A. A. Hakimi and N. Shah and M. H. Voss and K. C. Arbour and Y.-M. M. Cheung and H. Azhari and D. M. Faleck and R. Niec and M. T. A. Donoghue and J. J. Orgera and A. Syed and M. F. Berger and M. Waters and K. Pichotta and C. Fong and J. Jee and N. Schultz and D. Schrag and L. Yarmus and A. J. Schoenfeld and A. Gusev and R. R. Kotecha and R. J. Motzer and C. B. Thompson and W. Tansey and J. Carrot-Zhang and E. Reznik},
howpublished = {medRxiv},
year = {2026},
month = sep,
doi = {10.64898/2026.09.16.26363259},
url = {https://www.medrxiv.org/content/10.64898/2026.09.16.26363259v1},
note = {CC BY-NC-ND},
}
RIS
TY - UNPB
TI - Population-Scale Precision Safety in Oncology Reveals Clinical and Genetic Determinants of Systemic Therapy Toxicity
AU - Z. Bakouny
AU - X. A. Guo
AU - F. Huang
AU - S. Mohan
AU - R. Walser
AU - Z. Lu
AU - T. Perea-Chamblee
AU - L. Khan
AU - N. Ahmed
AU - A. Ocejo
AU - A. A. Hakimi
AU - N. Shah
AU - M. H. Voss
AU - K. C. Arbour
AU - Y.-M. M. Cheung
AU - H. Azhari
AU - D. M. Faleck
AU - R. Niec
AU - M. T. A. Donoghue
AU - J. J. Orgera
AU - A. Syed
AU - M. F. Berger
AU - M. Waters
AU - K. Pichotta
AU - C. Fong
AU - J. Jee
AU - N. Schultz
AU - D. Schrag
AU - L. Yarmus
AU - A. J. Schoenfeld
AU - A. Gusev
AU - R. R. Kotecha
AU - R. J. Motzer
AU - C. B. Thompson
AU - W. Tansey
AU - J. Carrot-Zhang
AU - E. Reznik
PY - 2026
DA - 2026/09/16
PB - medRxiv
DO - 10.64898/2026.09.16.26363259
UR - https://www.medrxiv.org/content/10.64898/2026.09.16.26363259v1
AB - Treatment toxicity constrains the use of effective cancer therapies, but its clinical and genetic determinants remain poorly defined. We developed a large language model-based approach to produce MSK-Tox, a pan-cancer resource capturing the incidence, temporality, and grade of toxicity to anti-cancer therapy across more than 50,000 patients. Analysis of six representative adverse events--pneumonitis, adrenal insufficiency, liver toxicity, colitis, hyperthyroidism and hypothyroidism--revealed distinct toxicity landscapes shaped by cancer type, treatment regimen, and clinical context. Pretreatment clinical features enabled individualized prediction of toxicity risk across adverse events, supporting risk assessment before therapy initiation. Beyond clinical predictors, we identified two modes of germline susceptibility to treatment toxicity: an organintrinsic mode, in which germline variation confers risk across systemic therapies, exemplified by a regulatory variant near FOXE1 associated with hypothyroidism; and an immune-mediated mode, confined to immune checkpoint inhibitor-treated patients, in which HLA-DRB1*15 was a major determinant of adrenal insufficiency. Notably, the same allele predisposes to multiple sclerosis in individuals without cancer, indicating that immune checkpoint inhibition unmasks a latent autoimmune predisposition. These findings provide an empirical basis for a new precision safety paradigm for predicting who will be harmed by a therapy on the same principles that guide prediction of therapeutic benefit.
ER -