Beiqun Zhao1,2, Rodney A Gabriel3, Florin Vaida4, Nicole E Lopez5, Samuel Eisenstein5, Bryan M Clary5. 1. Department of Surgery, University of California San Diego, San Diego, CA, USA. beiqunmzhao@gmail.com. 2. , San Diego, USA. beiqunmzhao@gmail.com. 3. Department of Anesthesiology, University of California San Diego, San Diego, CA, USA. 4. Department of Family Medicine and Public Health, University of California San Diego, San Diego, CA, USA. 5. Department of Surgery, University of California San Diego, San Diego, CA, USA.
Abstract
BACKGROUND: A significant proportion of patients with rectal cancer will present with synchronous metastasis at the time of diagnosis. Overall survival (OS) for these patients are highly variable and previous attempts to build predictive models often have low predictive power, with concordance indexes (c-index) less than 0.70. METHODS: Using the National Cancer Database (2010-2014), we identified patients with synchronous metastatic rectal cancer. The data was split into a training dataset (diagnosis years 2010-2012), which was used to build the machine learning model, and a testing dataset (diagnosis years 2013-2014), which was used to externally validate the model. A nomogram predicting 3-year OS was created using Cox proportional hazard regression with lasso penalization. Predictors were selected based on clinical significance and availability in NCDB. Performance of the machine learning model was assessed by c-index. RESULTS: A total of 4098 and 3107 patients were used to construct and validate the nomogram, respectively. Internally validated c-indexes at 1, 2, and 3 years were 0.816 (95% CI 0.813-0.818), 0.789 (95% CI 0.786-0.790), and 0.778 (95% CI 0.775-0.780), respectively. External validated c-indexes at 1, 2, and 3 years were 0.811, 0.779, and 0.778, respectively. CONCLUSIONS: There is wide variability in the OS for patients with metastatic rectal cancer, making accurate predictions difficult. However, using machine learning techniques, more accurate models can be built. This will aid patients and clinicians in setting expectations and making clinical decisions in this group of challenging patients.
BACKGROUND: A significant proportion of patients with rectal cancer will present with synchronous metastasis at the time of diagnosis. Overall survival (OS) for these patients are highly variable and previous attempts to build predictive models often have low predictive power, with concordance indexes (c-index) less than 0.70. METHODS: Using the National Cancer Database (2010-2014), we identified patients with synchronous metastatic rectal cancer. The data was split into a training dataset (diagnosis years 2010-2012), which was used to build the machine learning model, and a testing dataset (diagnosis years 2013-2014), which was used to externally validate the model. A nomogram predicting 3-year OS was created using Cox proportional hazard regression with lasso penalization. Predictors were selected based on clinical significance and availability in NCDB. Performance of the machine learning model was assessed by c-index. RESULTS: A total of 4098 and 3107 patients were used to construct and validate the nomogram, respectively. Internally validated c-indexes at 1, 2, and 3 years were 0.816 (95% CI 0.813-0.818), 0.789 (95% CI 0.786-0.790), and 0.778 (95% CI 0.775-0.780), respectively. External validated c-indexes at 1, 2, and 3 years were 0.811, 0.779, and 0.778, respectively. CONCLUSIONS: There is wide variability in the OS for patients with metastatic rectal cancer, making accurate predictions difficult. However, using machine learning techniques, more accurate models can be built. This will aid patients and clinicians in setting expectations and making clinical decisions in this group of challenging patients.
Entities:
Keywords:
Lasso; Machine learning; NCDB; Nomograms; Rectal cancer
Authors: Eddie K Abdalla; René Adam; Anton J Bilchik; Daniel Jaeck; Jean-Nicolas Vauthey; David Mahvi Journal: Ann Surg Oncol Date: 2006-09-06 Impact factor: 5.344
Authors: Kenneth L Kehl; Mary Beth Landrum; Neeraj K Arora; Patricia A Ganz; Michelle van Ryn; Jennifer W Mack; Nancy L Keating Journal: JAMA Oncol Date: 2015-04 Impact factor: 31.777
Authors: Timothy M Pawlik; Charles R Scoggins; Daria Zorzi; Eddie K Abdalla; Axel Andres; Cathy Eng; Steven A Curley; Evelyne M Loyer; Andrea Muratore; Gilles Mentha; Lorenzo Capussotti; Jean-Nicolas Vauthey Journal: Ann Surg Date: 2005-05 Impact factor: 12.969
Authors: Michalis Zacharakis; Ioannis D Xynos; Andreas Lazaris; Tsaousi Smaro; Christos Kosmas; Anna Dokou; Evangelos Felekouras; Efstathios Antoniou; Aris Polyzos; John Sarantonis; John Syrios; George Zografos; Alexandros Papalambros; Nikolas Tsavaris Journal: Anticancer Res Date: 2010-02 Impact factor: 2.480
Authors: Michael W Kattan; Mithat Gönen; William R Jarnagin; Ronald DeMatteo; Michael D'Angelica; Martin Weiser; Leslie H Blumgart; Yuman Fong Journal: Ann Surg Date: 2008-02 Impact factor: 12.969
Authors: Christina Hackl; Peter Neumann; Michael Gerken; Martin Loss; Monika Klinkhammer-Schalke; Hans J Schlitt Journal: BMC Cancer Date: 2014-11-04 Impact factor: 4.430
Authors: B Zhao; R A Gabriel; F Vaida; S Eisenstein; G T Schnickel; J K Sicklick; B M Clary Journal: Colorectal Dis Date: 2020-02-16 Impact factor: 3.788
Authors: Michael F Gensheimer; Sonya Aggarwal; Kathryn R K Benson; Justin N Carter; A Solomon Henry; Douglas J Wood; Scott G Soltys; Steven Hancock; Erqi Pollom; Nigam H Shah; Daniel T Chang Journal: J Am Med Inform Assoc Date: 2021-06-12 Impact factor: 4.497
Authors: Paula Dhiman; Jie Ma; Constanza L Andaur Navarro; Benjamin Speich; Garrett Bullock; Johanna A A Damen; Lotty Hooft; Shona Kirtley; Richard D Riley; Ben Van Calster; Karel G M Moons; Gary S Collins Journal: BMC Med Res Methodol Date: 2022-04-08 Impact factor: 4.615