TY - JOUR
T1 - Spatially-aware clustering improves AJCC-8 risk stratification performance in oropharyngeal carcinomas
AU - Canahuate, Guadalupe
AU - Wentzel, Andrew
AU - Mohamed, Abdallah S.R.
AU - van Dijk, Lisanne V.
AU - Vock, David M.
AU - Elgohari, Baher
AU - Elhalawani, Hesham
AU - Fuller, Clifton D.
AU - Marai, G. Elisabeta
N1 - Publisher Copyright:
© 2023 Elsevier Ltd
PY - 2023/9
Y1 - 2023/9
N2 - Objective: Evaluate the effectiveness of machine learning tools that incorporate spatial information such as disease location and lymph node metastatic patterns-of-spread, for prediction of survival and toxicity in HPV+ oropharyngeal cancer (OPC). Materials & methods: 675 HPV+ OPC patients that were treated at MD Anderson Cancer Center between 2005 and 2013 with curative intent IMRT were retrospectively collected under IRB approval. Risk stratifications incorporating patient radiometric data and lymph node metastasis patterns via an anatomically-adjacent representation with hierarchical clustering were identified. These clusterings were combined into a 3-level patient stratification and included along with other known clinical features in a Cox model for predicting survival outcomes, and logistic regression for toxicity, using independent subsets for training and validation. Results: Four groups were identified and combined into a 3-level stratification. The inclusion of patient stratifications in predictive models for 5-yr Overall survival (OS), 5-year recurrence free survival, (RFS) and Radiation-associated dysphagia (RAD) consistently improved model performance measured using the area under the curve (AUC). Test set AUC improvements over models with clinical covariates, was 9 % for predicting OS, and 18 % for predicting RFS, and 7 % for predicting RAD. For models with both clinical and AJCC covariates, AUC improvement was 7 %, 9 %, and 2 % for OS, RFS, and RAD, respectively. Conclusion: Including data-driven patient stratifications considerably improve prognosis for survival and toxicity outcomes over the performance achieved by clinical staging and clinical covariates alone. These stratifications generalize well to across cohorts, and sufficient information for reproducing these clusters is included.
AB - Objective: Evaluate the effectiveness of machine learning tools that incorporate spatial information such as disease location and lymph node metastatic patterns-of-spread, for prediction of survival and toxicity in HPV+ oropharyngeal cancer (OPC). Materials & methods: 675 HPV+ OPC patients that were treated at MD Anderson Cancer Center between 2005 and 2013 with curative intent IMRT were retrospectively collected under IRB approval. Risk stratifications incorporating patient radiometric data and lymph node metastasis patterns via an anatomically-adjacent representation with hierarchical clustering were identified. These clusterings were combined into a 3-level patient stratification and included along with other known clinical features in a Cox model for predicting survival outcomes, and logistic regression for toxicity, using independent subsets for training and validation. Results: Four groups were identified and combined into a 3-level stratification. The inclusion of patient stratifications in predictive models for 5-yr Overall survival (OS), 5-year recurrence free survival, (RFS) and Radiation-associated dysphagia (RAD) consistently improved model performance measured using the area under the curve (AUC). Test set AUC improvements over models with clinical covariates, was 9 % for predicting OS, and 18 % for predicting RFS, and 7 % for predicting RAD. For models with both clinical and AJCC covariates, AUC improvement was 7 %, 9 %, and 2 % for OS, RFS, and RAD, respectively. Conclusion: Including data-driven patient stratifications considerably improve prognosis for survival and toxicity outcomes over the performance achieved by clinical staging and clinical covariates alone. These stratifications generalize well to across cohorts, and sufficient information for reproducing these clusters is included.
KW - Clustering
KW - Head & Neck cancer
KW - Health Informatics
KW - Imaging
KW - Machine learning
KW - Oropharyngeal cancer
KW - Radiation oncology
KW - Radiomics
KW - Risk stratification
KW - Statistical and research methods
UR - http://www.scopus.com/inward/record.url?scp=85163514845&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=85163514845&partnerID=8YFLogxK
U2 - 10.1016/j.oraloncology.2023.106460
DO - 10.1016/j.oraloncology.2023.106460
M3 - Article
C2 - 37390759
AN - SCOPUS:85163514845
SN - 1368-8375
VL - 144
JO - Oral Oncology
JF - Oral Oncology
M1 - 106460
ER -