Hao Yu1,2, Ka-On Lam3,4, Huanmei Wu2, Michael Green5,6, Weili Wang7, Jian-Yue Jin7, Chen Hu8, Shruti Jolly6, Yang Wang1, Feng-Ming Spring Kong3,4,7. 1. Biomedical Engineering, Shenzhen Polytechnic, Shenzhen, China. 2. BioHealth Informatics, School of Informatics and Computing, Indiana University - Purdue University Indianapolis (IUPUI), Indianapolis, IN, United States. 3. Department of Clinical Oncology, Li Ka Shing (LKS) Faculty of Medicine, The University of Hong Kong, Hong Kong, Hong Kong. 4. Clinical Oncology Center, The University of Hong Kong-Shenzhen Hospital, Shenzhen, China. 5. Radiation Oncology, Ann Arbor VA Health Care, Ann Arbor, MI, United States. 6. Radiation Oncology, University of Michigan, Ann Arbor, MI, United States. 7. University Hospitals, Cleveland Medical Center, Seidman Cancer Center and Case Comprehensive Cancer Center, Case Western Reserve University, Cleveland, OH, United States. 8. Sidney Kimmel Comprehensive Cancer Center, Johns Hopkins University School of Medicine, Baltimore, MD, United States.
Abstract
BACKGROUND: Radiation-induced lung fibrosis (RILF) is an important late toxicity in patients with non-small-cell lung cancer (NSCLC) after radiotherapy (RT). Clinically significant RILF can impact quality of life and/or cause non-cancer related death. This study aimed to determine whether pre-treatment plasma cytokine levels have a significant effect on the risk of RILF and investigate the abilities of machine learning algorithms for risk prediction. METHODS: This is a secondary analysis of prospective studies from two academic cancer centers. The primary endpoint was grade≥2 (RILF2), classified according to a system consistent with the consensus recommendation of an expert panel of the AAPM task for normal tissue toxicity. Eligible patients must have at least 6 months' follow-up after radiotherapy commencement. Baseline levels of 30 cytokines, dosimetric, and clinical characteristics were analyzed. Support vector machine (SVM) algorithm was applied for model development. Data from one center was used for model training and development; and data of another center was applied as an independent external validation. RESULTS: There were 57 and 37 eligible patients in training and validation datasets, with 14 and 16.2% RILF2, respectively. Of the 30 plasma cytokines evaluated, SVM identified baseline circulating CCL4 as the most significant cytokine associated with RILF2 risk in both datasets (P = 0.003 and 0.07, for training and test sets, respectively). An SVM classifier predictive of RILF2 was generated in Cohort 1 with CCL4, mean lung dose (MLD) and chemotherapy as key model features. This classifier was validated in Cohort 2 with accuracy of 0.757 and area under the curve (AUC) of 0.855. CONCLUSIONS: Using machine learning, this study constructed and validated a weighted-SVM classifier incorporating circulating CCL4 levels with significant dosimetric and clinical parameters which predicts RILF2 risk with a reasonable accuracy. Further study with larger sample size is needed to validate the role of CCL4, and this SVM classifier in RILF2.
BACKGROUND: Radiation-induced lung fibrosis (RILF) is an important late toxicity in patients with non-small-cell lung cancer (NSCLC) after radiotherapy (RT). Clinically significant RILF can impact quality of life and/or cause non-cancer related death. This study aimed to determine whether pre-treatment plasma cytokine levels have a significant effect on the risk of RILF and investigate the abilities of machine learning algorithms for risk prediction. METHODS: This is a secondary analysis of prospective studies from two academic cancer centers. The primary endpoint was grade≥2 (RILF2), classified according to a system consistent with the consensus recommendation of an expert panel of the AAPM task for normal tissue toxicity. Eligible patients must have at least 6 months' follow-up after radiotherapy commencement. Baseline levels of 30 cytokines, dosimetric, and clinical characteristics were analyzed. Support vector machine (SVM) algorithm was applied for model development. Data from one center was used for model training and development; and data of another center was applied as an independent external validation. RESULTS: There were 57 and 37 eligible patients in training and validation datasets, with 14 and 16.2% RILF2, respectively. Of the 30 plasma cytokines evaluated, SVM identified baseline circulating CCL4 as the most significant cytokine associated with RILF2 risk in both datasets (P = 0.003 and 0.07, for training and test sets, respectively). An SVM classifier predictive of RILF2 was generated in Cohort 1 with CCL4, mean lung dose (MLD) and chemotherapy as key model features. This classifier was validated in Cohort 2 with accuracy of 0.757 and area under the curve (AUC) of 0.855. CONCLUSIONS: Using machine learning, this study constructed and validated a weighted-SVM classifier incorporating circulating CCL4 levels with significant dosimetric and clinical parameters which predicts RILF2 risk with a reasonable accuracy. Further study with larger sample size is needed to validate the role of CCL4, and this SVM classifier in RILF2.
Authors: José Marcio Luna; Hann-Hsiang Chao; Eric S Diffenderfer; Gilmer Valdes; Chidambaram Chinniah; Grace Ma; Keith A Cengel; Timothy D Solberg; Abigail T Berman; Charles B Simone Journal: Radiother Oncol Date: 2019-01-23 Impact factor: 6.280
Authors: Angela M Groves; Jacqueline P Williams; Eric Hernady; Christina Reed; Bruce Fenton; Tanzy Love; Jacob N Finkelstein; Carl J Johnston Journal: Radiat Res Date: 2018-08-17 Impact factor: 2.841
Authors: Jing Zhao; Ellen D Yorke; Ling Li; Brian D Kavanagh; X Allen Li; Shiva Das; Moyed Miften; Andreas Rimner; Jeffrey Campbell; Jinyu Xue; Andrew Jackson; Jimm Grimm; Michael T Milano; Feng-Ming Spring Kong Journal: Int J Radiat Oncol Biol Phys Date: 2016-03-25 Impact factor: 7.038
Authors: Deborah E Citrin; Pataje G S Prasanna; Amanda J Walker; Michael L Freeman; Iris Eke; Mary Helen Barcellos-Hoff; Molykutty J Arankalayil; Eric P Cohen; Ruth C Wilkins; Mansoor M Ahmed; Mitchell S Anscher; Benjamin Movsas; Jeffrey C Buchsbaum; Marc S Mendonca; Thomas A Wynn; C Norman Coleman Journal: Radiat Res Date: 2017-05-10 Impact factor: 2.841
Authors: Feng-Ming Spring Kong; Vitali Moiseenko; Jing Zhao; Michael T Milano; Ling Li; Andreas Rimner; Shiva Das; X Allen Li; Moyed Miften; ZhongXing Liao; Mary Martel; Soren M Bentzen; Andrew Jackson; Jimm Grimm; Lawrence B Marks; Ellen Yorke Journal: Int J Radiat Oncol Biol Phys Date: 2018-11-26 Impact factor: 8.013
Authors: Tiago Carvalheiro; Sara Horta; Joel A G van Roon; Mariana Santiago; Maria J Salvador; Hélder Trindade; Timothy R D J Radstake; José A P da Silva; Artur Paiva Journal: Inflamm Res Date: 2017-11-10 Impact factor: 4.575