Linlin Wang1, Taotao Dong2, Bowen Xin3, Chongrui Xu3, Meiying Guo1,4, Huaqi Zhang1,5, Dagan Feng3, Xiuying Wang6, Jinming Yu7. 1. Department of Radiation Oncology, Shandong Cancer Hospital Affiliated to Shandong University, Shandong Academy of Medical Science, No. 440, Ji Yan Road, Jinan, 250017, China. 2. Department of Gynecology and Obstetrics, Qilu Hospital of Shandong University, Jinan, China. 3. School of Information Technologies, the University of Sydney, Building J12, Sydney, NSW, 2006, Australia. 4. Medical College of Shandong University, Jinan, China. 5. Tianjin Medical University, Tianjin, China. 6. School of Information Technologies, the University of Sydney, Building J12, Sydney, NSW, 2006, Australia. xiu.wang@sydney.edu.au. 7. Department of Radiation Oncology, Shandong Cancer Hospital Affiliated to Shandong University, Shandong Academy of Medical Science, No. 440, Ji Yan Road, Jinan, 250017, China. sdyujinming@163.com.
Abstract
OBJECTIVES: To determine the integrative value of clinical, hematological, and computed tomography (CT) radiomic features in survival prediction for locally advanced non-small cell lung cancer (LA-NSCLC) patients. METHODS: Radiomic features and clinical and hematological features of 118 LA-NSCLC cases were firstly extracted and analyzed in this study. Then, stable and prognostic radiomic features were automatically selected using the consensus clustering method with either Cox proportional hazard (CPH) model or random survival forest (RSF) analysis. Predictive radiomic, clinical, and hematological parameters were subsequently fitted into a final prognostic model using both the CPH model and the RSF model. A multimodality nomogram was then established from the fitting model and was cross-validated. Finally, calibration curves were generated with the predicted versus actual survival status. RESULTS: Radiomic features selected by clustering combined with CPH were found to be more predictive, with a C-index of 0.699 in comparison to 0.648 by clustering combined with RSF. Based on multivariate CPH model, our integrative nomogram achieved a C-index of 0.792 and retained 0.743 in the cross-validation analysis, outperforming radiomic, clinical, or hematological model alone. The calibration curve showed agreement between predicted and actual values for the 1-year and 2-year survival prediction. Interestingly, the selected important radiomic features were significantly correlated with levels of platelet, platelet/lymphocyte ratio (PLR), and lymphocyte/monocyte ratio (LMR) (p values all < 0.05). CONCLUSIONS: The integrative nomogram incorporated CT radiomic, clinical, and hematological features improved survival prediction in LA-NSCLC patients, which would offer a feasible and practical reference for individualized management of these patients. KEY POINTS: • An integrative nomogram incorporated CT radiomic, clinical, and hematological features was constructed and cross-validated to predict prognosis of LA-NSCLC patients. • The integrative nomogram outperformed radiomic, clinical, or hematological model alone. • This nomogram has value to permit non-invasive, comprehensive, and dynamical evaluation of the phenotypes of LA-NSCLC and can provide a feasible and practical reference for individualized management of LA-NSCLC patients.
OBJECTIVES: To determine the integrative value of clinical, hematological, and computed tomography (CT) radiomic features in survival prediction for locally advanced non-small cell lung cancer (LA-NSCLC) patients. METHODS: Radiomic features and clinical and hematological features of 118 LA-NSCLC cases were firstly extracted and analyzed in this study. Then, stable and prognostic radiomic features were automatically selected using the consensus clustering method with either Cox proportional hazard (CPH) model or random survival forest (RSF) analysis. Predictive radiomic, clinical, and hematological parameters were subsequently fitted into a final prognostic model using both the CPH model and the RSF model. A multimodality nomogram was then established from the fitting model and was cross-validated. Finally, calibration curves were generated with the predicted versus actual survival status. RESULTS: Radiomic features selected by clustering combined with CPH were found to be more predictive, with a C-index of 0.699 in comparison to 0.648 by clustering combined with RSF. Based on multivariate CPH model, our integrative nomogram achieved a C-index of 0.792 and retained 0.743 in the cross-validation analysis, outperforming radiomic, clinical, or hematological model alone. The calibration curve showed agreement between predicted and actual values for the 1-year and 2-year survival prediction. Interestingly, the selected important radiomic features were significantly correlated with levels of platelet, platelet/lymphocyte ratio (PLR), and lymphocyte/monocyte ratio (LMR) (p values all < 0.05). CONCLUSIONS: The integrative nomogram incorporated CT radiomic, clinical, and hematological features improved survival prediction in LA-NSCLCpatients, which would offer a feasible and practical reference for individualized management of these patients. KEY POINTS: • An integrative nomogram incorporated CT radiomic, clinical, and hematological features was constructed and cross-validated to predict prognosis of LA-NSCLCpatients. • The integrative nomogram outperformed radiomic, clinical, or hematological model alone. • This nomogram has value to permit non-invasive, comprehensive, and dynamical evaluation of the phenotypes of LA-NSCLC and can provide a feasible and practical reference for individualized management of LA-NSCLCpatients.
Authors: Ping Yang; Mark S Allen; Marie C Aubry; Jason A Wampfler; Randolph S Marks; Eric S Edell; Stephen Thibodeau; Alex A Adjei; James Jett; Claude Deschamps Journal: Chest Date: 2005-07 Impact factor: 9.410
Authors: Everett E Vokes; James E Herndon; Michael J Kelley; M Giulia Cicchetti; Nithya Ramnath; Harvey Neill; James N Atkins; Dorothy M Watson; Wallace Akerley; Mark R Green Journal: J Clin Oncol Date: 2007-04-02 Impact factor: 44.544
Authors: Ahmedin Jemal; Freddie Bray; Melissa M Center; Jacques Ferlay; Elizabeth Ward; David Forman Journal: CA Cancer J Clin Date: 2011-02-04 Impact factor: 508.702
Authors: Nasser Hanna; Marcus Neubauer; Constantin Yiannoutsos; Ronald McGarry; James Arseneau; Rafat Ansari; Craig Reynolds; Ramaswamy Govindan; Anton Melnyk; William Fisher; Donald Richards; Daniel Bruetman; Thomas Anderson; Naveed Chowhan; Sreenivasa Nattam; Prasad Mantravadi; Cynthia Johnson; Tim Breen; Angela White; Lawrence Einhorn Journal: J Clin Oncol Date: 2008-11-10 Impact factor: 44.544