Sara R Alcorn1, Jacob Fiksel2, Jean L Wright3, Christen R Elledge3, Thomas J Smith4, Powell Perng3, Sarah Saleemi3, Todd R McNutt3, Theodore L DeWeese3, Scott Zeger2. 1. Department of Radiation Oncology and Molecular Radiation Sciences, Johns Hopkins School of Medicine, Baltimore, MD. Electronic address: salcorn2@jhmi.edu. 2. Department of Biostatistics, Johns Hopkins Bloomberg School of Public Health, Baltimore, MD. 3. Department of Radiation Oncology and Molecular Radiation Sciences, Johns Hopkins School of Medicine, Baltimore, MD. 4. Department of Oncology, Johns Hopkins School of Medicine, Baltimore, MD.
Abstract
PURPOSE: To determine whether a machine learning approach optimizes survival estimation for patients with symptomatic bone metastases (SBM), we developed the Bone Metastases Ensemble Trees for Survival (BMETS) to predict survival using 27 prognostic covariates. To establish its relative clinical utility, we compared BMETS with 2 simpler Cox regression models used in this setting. METHODS AND MATERIALS: For 492 bone sites in 397 patients evaluated for palliative radiation therapy (RT) for SBM from January 2007 to January 2013, data for 27 clinical variables were collected. These covariates and the primary outcome of time from consultation to death were used to build BMETS using random survival forests. We then performed Cox regressions as per 2 validated models: Chow's 3-item (C-3) and Westhoff's 2-item (W-2) tools. Model performance was assessed using cross-validation procedures and measured by time-dependent area under the curve (tAUC) for all 3 models. For temporal validation, a separate data set comprised of 104 bone sites treated in 85 patients in 2018 was used to estimate tAUC from BMETS. RESULTS: Median survival was 6.4 months. Variable importance was greatest for performance status, blood cell counts, recent systemic therapy type, and receipt of concurrent nonbone palliative RT. tAUC at 3, 6, and 12 months was 0.83, 0.81, and 0.81, respectively, suggesting excellent discrimination of BMETS across postconsultation time points. BMETS outperformed simpler models at each time, with respective tAUC at each time of 0.78, 0.76, and 0.74 for the C-3 model and 0.80, 0.78, and 0.77 for the W-2 model. For the temporal validation set, respective tAUC was similarly high at 0.86, 0.82, and 0.78. CONCLUSIONS: For patients with SBM, BMETS improved survival predictions versus simpler traditional models. Model performance was maintained when applied to a temporal validation set. To facilitate clinical use, we developed a web platform for data entry and display of BMETS-predicted survival probabilities.
PURPOSE: To determine whether a machine learning approach optimizes survival estimation for patients with symptomatic bone metastases (SBM), we developed the Bone Metastases Ensemble Trees for Survival (BMETS) to predict survival using 27 prognostic covariates. To establish its relative clinical utility, we compared BMETS with 2 simpler Cox regression models used in this setting. METHODS AND MATERIALS: For 492 bone sites in 397 patients evaluated for palliative radiation therapy (RT) for SBM from January 2007 to January 2013, data for 27 clinical variables were collected. These covariates and the primary outcome of time from consultation to death were used to build BMETS using random survival forests. We then performed Cox regressions as per 2 validated models: Chow's 3-item (C-3) and Westhoff's 2-item (W-2) tools. Model performance was assessed using cross-validation procedures and measured by time-dependent area under the curve (tAUC) for all 3 models. For temporal validation, a separate data set comprised of 104 bone sites treated in 85 patients in 2018 was used to estimate tAUC from BMETS. RESULTS: Median survival was 6.4 months. Variable importance was greatest for performance status, blood cell counts, recent systemic therapy type, and receipt of concurrent nonbone palliative RT. tAUC at 3, 6, and 12 months was 0.83, 0.81, and 0.81, respectively, suggesting excellent discrimination of BMETS across postconsultation time points. BMETS outperformed simpler models at each time, with respective tAUC at each time of 0.78, 0.76, and 0.74 for the C-3 model and 0.80, 0.78, and 0.77 for the W-2 model. For the temporal validation set, respective tAUC was similarly high at 0.86, 0.82, and 0.78. CONCLUSIONS: For patients with SBM, BMETS improved survival predictions versus simpler traditional models. Model performance was maintained when applied to a temporal validation set. To facilitate clinical use, we developed a web platform for data entry and display of BMETS-predicted survival probabilities.
Authors: Paulien G Westhoff; Alexander de Graeff; Evelyn M Monninkhof; Laurens Bollen; Sander P Dijkstra; Elzbieta M van der Steen-Banasik; Marco van Vulpen; Jan Willem H Leer; Corrie A Marijnen; Yvette M van der Linden Journal: Int J Radiat Oncol Biol Phys Date: 2014-09-24 Impact factor: 7.038
Authors: Samuel T Chao; Shlomo A Koyfman; Neil Woody; Lilyana Angelov; Sherry L Soeder; Chandana A Reddy; Lisa A Rybicki; Toufik Djemil; John H Suh Journal: Int J Radiat Oncol Biol Phys Date: 2011-04-12 Impact factor: 7.038
Authors: Frances A Shepherd; José Rodrigues Pereira; Tudor Ciuleanu; Eng Huat Tan; Vera Hirsh; Sumitra Thongprasert; Daniel Campos; Savitree Maoleekoonpiroj; Michael Smylie; Renato Martins; Maximiliano van Kooten; Mircea Dediu; Brian Findlay; Dongsheng Tu; Dianne Johnston; Andrea Bezjak; Gary Clark; Pedro Santabárbara; Lesley Seymour Journal: N Engl J Med Date: 2005-07-14 Impact factor: 91.245
Authors: Laurence Collette; George van Andel; Andrew Bottomley; Gosse O N Oosterhof; Walter Albrecht; Theo M de Reijke; Sophie D Fossà Journal: J Clin Oncol Date: 2004-10-01 Impact factor: 44.544
Authors: Susan Halabi; Nicholas J Vogelzang; Alice B Kornblith; San-San Ou; Philip W Kantoff; Nancy A Dawson; Eric J Small Journal: J Clin Oncol Date: 2008-05-20 Impact factor: 44.544