Machine learning to predict early recurrence after oesophageal cancer surgery
Saqib A Rahman,
Robert C Walker,
Megan A Lloyd,
Ben L Grace,
Gijs I van Boxel,
Jelle P. Ruurda,
Richard van Hillegersberg,
Ewen A Griffiths,
Rebecca C Fitzgerald,
Timothy J Underwood,
On behalf of the OCCAMS Consortium, the full list of contributors is displayed in acknowledgements
Posted 03 Jul 2019
medRxiv DOI: 10.1101/19001073
Posted 03 Jul 2019
ObjectiveTo develop a predictive model for early recurrence after surgery for oesophageal adenocarcinoma using a large multi-national cohort. Summary Background DataEarly cancer recurrence after oesophagectomy is a common problem with an incidence of 20-30% despite the widespread use of neoadjuvant treatment. Quantification of this risk is difficult and existing models perform poorly. Machine learning techniques potentially allow more accurate prognostication and have been applied in this study. MethodsConsecutive patients who underwent oesophagectomy for adenocarcinoma and had neoadjuvant treatment in 6 UK and 1 Dutch oesophago-gastric units were analysed. Using clinical characteristics and post-operative histopathology, models were generated using elastic net regression (ELR) and the machine learning methods random forest (RF) and XG boost (XGB). Finally, a combined (Ensemble) model of these was generated. The relative importance of factors to outcome was calculated as a percentage contribution to the model. ResultsIn total 812 patients were included. The recurrence rate at less than 1 year was 29.1%. All of the models demonstrated good discrimination. Internally validated AUCs were similar, with the Ensemble model performing best (ELR=0.785, RF=0.789, XGB=0.794, Ensemble=0.806). Performance was similar when using internal-external validation (validation across sites, Ensemble AUC=0.804). In the final model the most important variables were number of positive lymph nodes (25.7%) and vascular invasion (16.9%). ConclusionsThe derived model using machine learning approaches and an international dataset provided excellent performance in quantifying the risk of early recurrence after surgery and will be useful in prognostication for clinicians and patients. DRAFT VISUAL ABSTRACT O_FIG O_LINKSMALLFIG WIDTH=200 HEIGHT=110 SRC="FIGDIR/small/19001073v1_ufig1.gif" ALT="Figure 1"> View larger version (26K): org.highwire.dtl.DTLVardef@2f60b7org.highwire.dtl.DTLVardef@76bfb6org.highwire.dtl.DTLVardef@2469deorg.highwire.dtl.DTLVardef@a27d47_HPS_FORMAT_FIGEXP M_FIG C_FIG Icons taken from www.flaticon.com, made by Freepik, smashicons, and prettycons. Reproduced under creative commons attribution license MINI-ABSTRACTEarly recurrence after surgery for adenocarcinoma of the oesophagus is common. We derived a risk prediction model using modern machine learning methods that accurately predicts risk of early recurrence using post-operative pathology
- Downloaded 912 times
- Download rankings, all-time:
- Site-wide: 26,339
- In surgery: 5
- Year to date:
- Site-wide: 51,563
- Since beginning of last month:
- Site-wide: 36,064
Downloads over time
Distribution of downloads per paper, site-wide
- 27 Nov 2020: The website and API now include results pulled from medRxiv as well as bioRxiv.
- 18 Dec 2019: We're pleased to announce PanLingua, a new tool that enables you to search for machine-translated bioRxiv preprints using more than 100 different languages.
- 21 May 2019: PLOS Biology has published a community page about Rxivist.org and its design.
- 10 May 2019: The paper analyzing the Rxivist dataset has been published at eLife.
- 1 Mar 2019: We now have summary statistics about bioRxiv downloads and submissions.
- 8 Feb 2019: Data from Altmetric is now available on the Rxivist details page for every preprint. Look for the "donut" under the download metrics.
- 30 Jan 2019: preLights has featured the Rxivist preprint and written about our findings.
- 22 Jan 2019: Nature just published an article about Rxivist and our data.
- 13 Jan 2019: The Rxivist preprint is live!