Machine Learning to Improve Prognosis Prediction of Early Hepatocellular Carcinoma After Surgical Resection.
artificial intelligence
liver cancer
modelling
prognosis
surgery
Journal
Journal of hepatocellular carcinoma
ISSN: 2253-5969
Titre abrégé: J Hepatocell Carcinoma
Pays: New Zealand
ID NLM: 101674775
Informations de publication
Date de publication:
2021
2021
Historique:
received:
14
05
2021
accepted:
29
07
2021
entrez:
20
8
2021
pubmed:
21
8
2021
medline:
21
8
2021
Statut:
epublish
Résumé
Improved prognostic prediction is needed to stratify patients with early hepatocellular carcinoma (EHCC) to refine selection of adjuvant therapy. We aimed to develop a machine learning (ML)-based model to predict survival after liver resection for EHCC based on readily available clinical data. We analyzed data of surgically resected EHCC (tumor≤5 cm without evidence of extrahepatic disease or major vascular invasion) patients from the Surveillance, Epidemiology, and End Results (SEER) Program to train and internally validate a gradient-boosting ML model to predict disease-specific survival (DSS). We externally tested the ML model using data from 2 Chinese institutions. Patients treated with resection were matched by propensity score to those treated with transplantation in the SEER-Medicare database. A total of 2778 EHCC patients treated with resection were enrolled, divided into 1899 for training/validation (SEER) and 879 for test (Chinese). The ML model consisted of 8 covariates (age, race, alpha-fetoprotein, tumor size, multifocality, vascular invasion, histological grade and fibrosis score) and predicted DSS with C-Statistics >0.72, better than proposed staging systems across study cohorts. The ML model could stratify 10-year DSS ranging from 70% in low-risk subset to 5% in high-risk subset. Compared with low-risk subset, no remarkable survival benefits were observed in EHCC patients receiving transplantation before and after propensity score matching. An ML model trained on a large-scale dataset has good predictive performance at individual scale. Such a model is readily integrated into clinical practice and will be valuable in discussing treatment strategies.
Sections du résumé
BACKGROUND
BACKGROUND
Improved prognostic prediction is needed to stratify patients with early hepatocellular carcinoma (EHCC) to refine selection of adjuvant therapy. We aimed to develop a machine learning (ML)-based model to predict survival after liver resection for EHCC based on readily available clinical data.
METHODS
METHODS
We analyzed data of surgically resected EHCC (tumor≤5 cm without evidence of extrahepatic disease or major vascular invasion) patients from the Surveillance, Epidemiology, and End Results (SEER) Program to train and internally validate a gradient-boosting ML model to predict disease-specific survival (DSS). We externally tested the ML model using data from 2 Chinese institutions. Patients treated with resection were matched by propensity score to those treated with transplantation in the SEER-Medicare database.
RESULTS
RESULTS
A total of 2778 EHCC patients treated with resection were enrolled, divided into 1899 for training/validation (SEER) and 879 for test (Chinese). The ML model consisted of 8 covariates (age, race, alpha-fetoprotein, tumor size, multifocality, vascular invasion, histological grade and fibrosis score) and predicted DSS with C-Statistics >0.72, better than proposed staging systems across study cohorts. The ML model could stratify 10-year DSS ranging from 70% in low-risk subset to 5% in high-risk subset. Compared with low-risk subset, no remarkable survival benefits were observed in EHCC patients receiving transplantation before and after propensity score matching.
CONCLUSION
CONCLUSIONS
An ML model trained on a large-scale dataset has good predictive performance at individual scale. Such a model is readily integrated into clinical practice and will be valuable in discussing treatment strategies.
Identifiants
pubmed: 34414136
doi: 10.2147/JHC.S320172
pii: 320172
pmc: PMC8370036
doi:
Types de publication
Journal Article
Langues
eng
Pagination
913-923Commentaires et corrections
Type : ErratumIn
Informations de copyright
© 2021 Ji et al.
Déclaration de conflit d'intérêts
The authors declare no potential conflicts of interest.
Références
N Engl J Med. 2019 Apr 4;380(14):1347-1358
pubmed: 30943338
Nat Rev Clin Oncol. 2018 Oct;15(10):599-616
pubmed: 30061739
Ann Surg. 2015 May;261(5):939-46
pubmed: 24950276
Circulation. 2015 Nov 17;132(20):1920-30
pubmed: 26572668
Am J Gastroenterol. 2014 Apr;109(4):542-53
pubmed: 24513805
N Engl J Med. 2019 Apr 11;380(15):1450-1462
pubmed: 30970190
Ann Surg Oncol. 2013 Apr;20(4):1223-9
pubmed: 23179993
Radiology. 2020 Mar;294(3):568-579
pubmed: 31934830
Nat Rev Gastroenterol Hepatol. 2019 Oct;16(10):589-604
pubmed: 31439937
BMJ. 2009 Jun 29;338:b2393
pubmed: 19564179
J Hepatol. 2018 Jul;69(1):182-236
pubmed: 29628281
Clin Cancer Res. 2004 Nov 1;10(21):7252-9
pubmed: 15534099
HPB (Oxford). 2020 Mar;22(3):383-390
pubmed: 31416786
Nat Protoc. 2020 Feb;15(2):479-512
pubmed: 31932775
Ann Surg. 2018 Nov;268(5):868-875
pubmed: 30080736
J Hepatol. 2018 Dec;69(6):1284-1293
pubmed: 30236834
JAMA Surg. 2020 Feb 1;155(2):148-158
pubmed: 31825465
Hepatology. 2018 Aug;68(2):723-750
pubmed: 29624699
Ann Surg. 2013 Dec;258(6):1022-7
pubmed: 23299519
Med Decis Making. 2006 Nov-Dec;26(6):565-74
pubmed: 17099194
J Clin Oncol. 2017 Mar;35(7):734-742
pubmed: 28068175
Liver Cancer. 2020 Dec;9(6):682-720
pubmed: 33442540
Hepatology. 2016 Mar;63(3):839-49
pubmed: 26567038
Eur Urol. 2020 Sep;78(3):335-344
pubmed: 31980309
Hepatology. 2020 Jan;71(1):214-224
pubmed: 29742811
Gut. 2021 May;70(5):884-889
pubmed: 32887732