Machine learning algorithm improves the detection of NASH (NAS-based) and at-risk NASH: A development and validation study.
Journal
Hepatology (Baltimore, Md.)
ISSN: 1527-3350
Titre abrégé: Hepatology
Pays: United States
ID NLM: 8302946
Informations de publication
Date de publication:
01 07 2023
01 07 2023
Historique:
received:
16
06
2022
accepted:
22
12
2022
medline:
21
6
2023
pubmed:
31
3
2023
entrez:
30
3
2023
Statut:
ppublish
Résumé
Detecting NASH remains challenging, while at-risk NASH (steatohepatitis and F≥ 2) tends to progress and is of interest for drug development and clinical application. We developed prediction models by supervised machine learning techniques, with clinical data and biomarkers to stage and grade patients with NAFLD. Learning data were collected in the Liver Investigation: Testing Marker Utility in Steatohepatitis metacohort (966 biopsy-proven NAFLD adults), staged and graded according to NASH CRN. Conditions of interest were the clinical trial definition of NASH (NAS ≥ 4;53%), at-risk NASH (NASH with F ≥ 2;35%), significant (F ≥ 2;47%), and advanced fibrosis (F ≥ 3;28%). Thirty-five predictors were included. Missing data were handled by multiple imputations. Data were randomly split into training/validation (75/25) sets. A gradient boosting machine was applied to develop 2 models for each condition: clinical versus extended (clinical and biomarkers). Two variants of the NASH and at-risk NASH models were constructed: direct and composite models.Clinical gradient boosting machine models for steatosis/inflammation/ballooning had AUCs of 0.94/0.79/0.72. There were no improvements when biomarkers were included. The direct NASH model produced AUCs (clinical/extended) of 0.61/0.65. The composite NASH model performed significantly better (0.71) for both variants. The composite at-risk NASH model had an AUC of 0.83 (clinical and extended), an improvement over the direct model. Significant fibrosis models had AUCs (clinical/extended) of 0.76/0.78. The extended advanced fibrosis model (0.86) performed significantly better than the clinical version (0.82). Detection of NASH and at-risk NASH can be improved by constructing independent machine learning models for each component, using only clinical predictors. Adding biomarkers only improved the accuracy of fibrosis.
Sections du résumé
BACKGROUND AND AIMS
Detecting NASH remains challenging, while at-risk NASH (steatohepatitis and F≥ 2) tends to progress and is of interest for drug development and clinical application. We developed prediction models by supervised machine learning techniques, with clinical data and biomarkers to stage and grade patients with NAFLD.
APPROACH AND RESULTS
Learning data were collected in the Liver Investigation: Testing Marker Utility in Steatohepatitis metacohort (966 biopsy-proven NAFLD adults), staged and graded according to NASH CRN. Conditions of interest were the clinical trial definition of NASH (NAS ≥ 4;53%), at-risk NASH (NASH with F ≥ 2;35%), significant (F ≥ 2;47%), and advanced fibrosis (F ≥ 3;28%). Thirty-five predictors were included. Missing data were handled by multiple imputations. Data were randomly split into training/validation (75/25) sets. A gradient boosting machine was applied to develop 2 models for each condition: clinical versus extended (clinical and biomarkers). Two variants of the NASH and at-risk NASH models were constructed: direct and composite models.Clinical gradient boosting machine models for steatosis/inflammation/ballooning had AUCs of 0.94/0.79/0.72. There were no improvements when biomarkers were included. The direct NASH model produced AUCs (clinical/extended) of 0.61/0.65. The composite NASH model performed significantly better (0.71) for both variants. The composite at-risk NASH model had an AUC of 0.83 (clinical and extended), an improvement over the direct model. Significant fibrosis models had AUCs (clinical/extended) of 0.76/0.78. The extended advanced fibrosis model (0.86) performed significantly better than the clinical version (0.82).
CONCLUSIONS
Detection of NASH and at-risk NASH can be improved by constructing independent machine learning models for each component, using only clinical predictors. Adding biomarkers only improved the accuracy of fibrosis.
Identifiants
pubmed: 36994719
doi: 10.1097/HEP.0000000000000364
pii: 01515467-202307000-00021
doi:
Substances chimiques
nas
64706-31-6
Biomarkers
0
Types de publication
Journal Article
Research Support, Non-U.S. Gov't
Langues
eng
Sous-ensembles de citation
IM
Pagination
258-271Investigateurs
Quentin M Anstee
(QM)
Ann K Daly
(AK)
Olivier Govaere
(O)
Simon Cockell
(S)
Dina Tiniakos
(D)
Pierre Bedossa
(P)
Alastair Burt
(A)
Fiona Oakley
(F)
Heather J Cordell
(HJ)
Christopher P Day
(CP)
Kristy Wonders
(K)
Paolo Missier
(P)
Matthew McTeer
(M)
Luke Vale
(L)
Yemi Oluboyede
(Y)
Matt Breckons
(M)
Patrick M Bossuyt
(PM)
Hadi Zafarmand
(H)
Yasaman Vali
(Y)
Jenny Lee
(J)
Max Nieuwdorp
(M)
Adriaan G Holleboom
(AG)
Joanne Verheij
(J)
Vlad Ratziu
(V)
Karine Clément
(K)
Rafael Patino-Navarrete
(R)
Raluca Pais
(R)
Valerie Paradis
(V)
Detlef Schuppan
(D)
Jörn M Schattenberg
(JM)
Rambabu Surabattula
(R)
Sudha Myneni
(S)
Beate K Straub
(BK)
Toni Vidal-Puig
(T)
Michele Vacca
(M)
Sergio Rodrigues-Cuenca
(S)
Mike Allison
(M)
Ioannis Kamzolas
(I)
Evangelia Petsalaki
(E)
Mark Campbell
(M)
Chris J Lelliott
(CJ)
Susan Davies
(S)
Matej Orešič
(M)
Tuulia Hyötyläinen
(T)
Aiden McGlinchey
(A)
Jose M Mato
(JM)
Óscar Millet
(Ó)
Jean-François Dufour
(JF)
Annalisa Berzigotti
(A)
Mojgan Masoodi
(M)
Michael Pavlides
(M)
Stephen Harrison
(S)
Stefan Neubauer
(S)
Jeremy Cobbold
(J)
Ferenc Mozes
(F)
Salma Akhtar
(S)
Seliat Olodo-Atitebi
(S)
Rajarshi Banerjee
(R)
Matt Kelly
(M)
Elizabeth Shumbayawonda
(E)
Andrea Dennis
(A)
Anneli Andersson
(A)
Ioan Wigley
(I)
Manuel Romero-Gómez
(M)
Emilio Gómez-González
(E)
Javier Ampuero
(J)
Javier Castell
(J)
Rocío Gallego-Durán
(R)
Isabel Fernández
(I)
Rocío Montero-Vallejo
(R)
Morten Karsdal
(M)
Daniel Guldager Kring Rasmussen
(DGK)
Diana Julie Leeming
(DJ)
Antonia Sinisi
(A)
Kishwar Musa
(K)
Estelle Sandt
(E)
Manuela Tonini
(M)
Elisabetta Bugianesi
(E)
Chiara Rosso
(C)
Angelo Armandi
(A)
Fabio Marra
(F)
Amalia Gastaldelli
(A)
Gianluca Svegliati
(G)
Jérôme Boursier
(J)
Sven Francque
(S)
Luisa Vonghia
(L)
Ann Driessen
(A)
Mattias Ekstedt
(M)
Stergios Kechagias
(S)
Hannele Yki-Järvinen
(H)
Kimmo Porthan
(K)
Johanna Arola
(J)
Saskia van Mil
(S)
George Papatheodoridis
(G)
Helena Cortez-Pinto
(H)
Cecilia M P Rodrigues
(CMP)
Luca Valenti
(L)
Serena Pelusi
(S)
Salvatore Petta
(S)
Grazia Pennisi
(G)
Luca Miele
(L)
Andreas Geier
(A)
Christian Trautwein
(C)
Guruprasad P Aithal
(GP)
Susan Francis
(S)
Paul Hockings
(P)
Moritz Schneider
(M)
Philip Newsome
(P)
Stefan Hübscher
(S)
David Wenn
(D)
Christian Rosenquist
(C)
Aldo Trylesinski
(A)
Rebeca Mayo
(R)
Cristina Alonso
(C)
Kevin Duffin
(K)
James W Perfield
(JW)
Yu Chen
(Y)
Carla Yunis
(C)
Theresa Tuthill
(T)
Magdalena Alicia Harrington
(MA)
Melissa Miller
(M)
Yan Chen
(Y)
Euan James McLeod
(EJ)
Trenton Ross
(T)
Barbara Bernardo
(B)
Corinna Schölch
(C)
Judith Ertle
(J)
Ramy Younes
(R)
Anouk Oldenburger
(A)
Rachel Ostroff
(R)
Leigh Alexander
(L)
Hannah Biegel
(H)
Mette Skalshøi Kjær
(MS)
Lea Mørch Harder
(LM)
Peter Davidsen
(P)
Lars Friis Mikkelsen
(LF)
Maria-Magdalena Balp
(MM)
Clifford Brass
(C)
Lori Jennings
(L)
Miljen Martic
(M)
Jürgen Löffler
(J)
Douglas Applegate
(D)
Sudha Shankar
(S)
Richard Torstenson
(R)
Céline Fournier-Poizat
(C)
Anne Llorca
(A)
Michael Kalutkiewicz
(M)
Kay Pepin
(K)
Richard Ehman
(R)
Gerald Horan
(G)
Gideon Ho
(G)
Dean Tai
(D)
Elaine Chng
(E)
Scott D Patterson
(SD)
Andrew Billin
(A)
Lynda Doward
(L)
James Twiss
(J)
Paresh Thakker
(P)
Henrik Landgren
(H)
Carolin Lackner
(C)
Annette Gouw
(A)
Prodromos Hytiroglou
(P)
Informations de copyright
Copyright © 2023 The Author(s). Published by Wolters Kluwer Health, Inc.
Références
Younossi Z, Henry L. The Burden of NAFLD Worldwide. Non-Alcoholic Fatty Liver Disease: Springer; 2020:15–24.
Noureddin M, Truong E, Gornbein JA, Saouaf R, Guindi M, Todo T, et al. MRI-based (MAST) score accurately identifies patients with NASH and significant fibrosis. J Hepatol. 2022;76:781–787.
Satapathy SK, Sanyal AJ. Epidemiology and natural history of nonalcoholic fatty liver disease Seminars in liver disease. Thieme Medical Publishers; 2015.
Younossi Z, Anstee QM, Marietti M, Hardy T, Henry L, Eslam M, et al. Global burden of NAFLD and NASH: Trends, predictions, risk factors and prevention. Nat Rev Gastroenterol Hepatol. 2018;15:11–20.
Younossi Z, Stepanova M, Ong JP, Jacobson IM, Bugianesi E, Duseja A, et al. Nonalcoholic steatohepatitis is the fastest growing cause of hepatocellular carcinoma in liver transplant candidates. Clin Gastroenterol Hepatol. 2019;17:748–55. e3.
Ascha MS, Hanouneh IA, Lopez R, Tamimi TAR, Feldstein AF, Zein NN. The incidence and risk factors of hepatocellular carcinoma in patients with nonalcoholic steatohepatitis. Hepatology. 2010;51:1972–1978.
Sanyal AJ, Van Natta ML, Clark J, Neuschwander-Tetri BA, Diehl A, Dasarathy S, et al. Prospective study of outcomes in adults with nonalcoholic fatty liver disease. N Engl J Med. 2021;385:1559–69.
Ratziu V, Charlotte F, Heurtier A, Gombert S, Giral P, Bruckert E, et al. Sampling variability of liver biopsy in nonalcoholic fatty liver disease. Gastroenterology. 2005;128:1898–1906.
Brunt EM, Clouston AD, Goodman Z, Guy C, Kleiner DE, Lackner C, et al. Complexity of ballooned hepatocyte feature recognition: Defining a training atlas for artificial intelligence-based imaging in NAFLD. J Hepatol. 2022;76:1030–41.
Berzigotti A, Tsochatzis E, Boursier J, Castera L, Cazzagon N, Friedrich-Rust M, et al. EASL Clinical Practice Guidelines on non-invasive tests for evaluation of liver disease severity and prognosis–2021 update. J Hepatol. 2021;75:659–689.
Mózes FE, Lee JA, Selvaraj EA, Jayaswal ANA, Trauner M, Boursier J, et al. Diagnostic accuracy of non-invasive tests for advanced fibrosis in patients with NAFLD: An individual patient data meta-analysis. Gut. 2022;71:1006–19.
Lee J, Vali Y, Boursier J, Duffin K, Verheij J, Brosnan MJ, et al. Accuracy of cytokeratin 18 (M30 and M65) in detecting non-alcoholic steatohepatitis and fibrosis: A systematic review and meta-analysis. PLoS ONE. 2020;15:e0238717.
Tada T, Kumada T, Toyoda H, Saibara T, Ono M, Kage M. New scoring system combining the FIB-4 index and cytokeratin-18 fragments for predicting steatohepatitis and liver fibrosis in patients with nonalcoholic fatty liver disease. Biomarkers. 2018;23:328–34.
Younossi ZM, Page S, Rafiq N, Birerdinc A, Stepanova M, Hossain N, et al. A biomarker panel for non-alcoholic steatohepatitis (NASH) and NASH-related fibrosis. Obes Surg. 2011;21:431–439.
Anty R, Iannelli A, Patouraux S, Bonnafous S, Lavallard V, Senni‐Buratti M, et al. A new composite model including metabolic syndrome, alanine aminotransferase and cytokeratin‐18 for the diagnosis of non‐alcoholic steatohepatitis in morbidly obese patients. Aliment Pharmacol Ther. 2010;32(11‐12):1315–1322.
Boursier J, Anty R, Vonghia L, Moal V, Vanwolleghem T, Canivet C, et al. Screening for therapeutic trials and treatment indication in clinical practice: MACK‐3, a new blood test for the diagnosis of fibrotic NASH. Aliment Pharmacol Ther. 2018;47:1387–96.
Newsome PN, Sasso M, Deeks JJ, Paredes A, Boursier J, Chan W-K, et al. FibroScan-AST (FAST) score for the non-invasive identification of patients with non-alcoholic steatohepatitis with significant activity and fibrosis: A prospective derivation and global validation study. Lancet Gastroenterol Hepatol. 2020;5:362–73.
Harrison SA, Ratziu V, Boursier J, Francque S, Bedossa P, Majd Z, et al. A blood-based biomarker panel (NIS4) for non-invasive diagnosis of non-alcoholic steatohepatitis and liver fibrosis: A prospective derivation and global validation study. Lancet Gastroenterol Hepatol. 2020;5:970–985.
Adamichou C, Genitsaridi I, Nikolopoulos D, Nikoloudaki M, Repa A, Bortoluzzi A, et al. Lupus or not? SLE Risk Probability Index (SLERPI): A simple, clinician-friendly machine learning-based model to assist the diagnosis of systemic lupus erythematosus. Ann Rheum Dis. 2021;80:758–766.
Karaglani M, Gourlia K, Tsamardinos I, Chatzaki E. Accurate blood-based diagnostic biosignatures for Alzheimer’s disease via automated machine learning. J Clin Med. 2020;9:3016.
Docherty M, Regnier SA, Capkun G, Balp M-M, Ye Q, Janssens N, et al. Development of a novel machine learning model to predict presence of nonalcoholic steatohepatitis. J Am Med Inform Assoc. 2021;28:1235–1241.
Moons KG, Altman DG, Reitsma JB, Ioannidis JP, Macaskill P, Steyerberg EW, et al. Transparent Reporting of a multivariable prediction model for Individual Prognosis or Diagnosis (TRIPOD): Explanation and elaboration. Ann Intern Med. 2015;162:W1–W73.
Vali Y, Lee J, Boursier J, Petta S, Wonders K, Tiniakos D, et al. Liver Investigation: Testing Marker Utility in Steatohepatitis (LITMUS) consortium investigators. Biomarkers for staging fibrosis and non-alcoholic steatohepatitis in non-alcoholic fatty liver disease (the LITMUS project): a comparative diagnostic accuracy study. Lancet Gastroenterol Hepatol. doi:10.1016/S2468-1253(23)00017-1. Epub ahead of print. PMID: 36958367.
doi: 10.1016/S2468-1253(23)00017-1
Hardy T, Wonders K, Younes R, Aithal GP, Aller R, Allison M, et al. The European NAFLD Registry: A real-world longitudinal cohort study of nonalcoholic fatty liver disease. Contemp Clin Trials. 2020;98:106175.
Kleiner DE, Brunt EM, Van Natta M, Behling C, Contos MJ, Cummings OW, et al. Design and validation of a histological scoring system for nonalcoholic fatty liver disease. Hepatology. 2005;41:1313–1321.
Bedossa P. Diagnosis of non-alcoholic fatty liver disease/non-alcoholic steatohepatitis: Why liver biopsy is essential. Liver Inter. 2018;38(S1):64–66.
Anania FA, Dimick-Santos L, Mehta R, Toerner J, Beitz J. Nonalcoholic steatohepatitis: Current thinking from the Division of Hepatology and Nutrition at the Food and Drug Administration. Hepatology. 2021;73:2023–2027.
Ratziu V, Harrison SA, Francque S, Bedossa P, Lehert P, Serfaty L, et al. Elafibranor, an agonist of the peroxisome proliferator−activated receptor−α and −δ, induces resolution of nonalcoholic steatohepatitis without fibrosis worsening. Gastroenterology. 2016;150:1147–59.e5.
Friedman JH. Stochastic gradient boosting. Comput Stat Data Anal. 2002;38:367–378.
Van Buuren S, Groothuis-Oudshoorn K. mice: Multivariate imputation by chained equations in R. J Stat Softw. 2011;45:1–67.
Hoogland J, van Barreveld M, Debray TPA, Reitsma JB, Verstraelen TE, Dijkgraaf MGW, et al. Handling missing predictor values when validating and applying a prediction model to new patients. Stat Med. 2020;39:3591–607.
Musoro JZ, Zwinderman AH, Puhan MA, ter Riet G, Geskus RB. Validation of prediction models based on lasso regression with multiply imputed data. BMC Med Res Methodol. 2014;14:1–13.
Sanyal AJ, Brunt EM, Kleiner DE, Kowdley KV, Chalasani N, Lavine JE, et al. Endpoints and clinical trial design for nonalcoholic steatohepatitis. Hepatology. 2011;54:344–53.
Noncirrhotic nonalcoholic steatohepatitis with liver fibrosis: Developing drugs for treatment. In: Services DoHaH, (CDER) CfDEaR, editors.
Marshall A, Altman DG, Holder RL, Royston P. Combining estimates of interest in prognostic modelling studies after multiple imputation: Current practice and guidelines. BMC Med Res Methodol. 2009;9:57.
Heymans M, Eekhout I. Applied missing data analysis with SPSS and (R) Studio. Amsterdam, The Netherlands: Heymans and Eekhout; 2019:20; Available online: https://bookdownorg/mwheymans/bookmi/[accessed 23 May 2020].
Daniels SJ, Leeming DJ, Eslam M, Hashem AM, Nielsen MJ, Krag A, et al. ADAPT: An Algorithm Incorporating PRO-C3 Accurately Identifies Patients With NAFLD and Advanced Fibrosis. Hepatology. 2019;69:1075–86.
Vallet‐Pichard A, Mallet V, Nalpas B, Verkarre V, Nalpas A, Dhalluin‐Venier V, et al. FIB‐4: An inexpensive and accurate marker of fibrosis in HCV infection. comparison with liver biopsy and fibrotest. Hepatology. 2007;46:32–36.
Day JW, Rosenberg WM. The enhanced liver fibrosis (ELF) test in diagnosis and management of liver fibrosis. Br J Hosp Med (Lond). 2018;79:694–699.
Kuhn M. Building predictive models in R using the caret package. J Stat Softw. 2008;28:1–26.
Davison BA, Harrison SA, Cotter G, Alkhouri N, Sanyal A, Edwards C, et al. Suboptimal reliability of liver biopsy evaluation has implications for randomized clinical trials. J Hepatol. 2020;73:1322–32.
Van Buuren S, Boshuizen HC, Knook DL. Multiple imputation of missing blood pressure covariates in survival analysis. Stat Med. 1999;18:681–694.
Fialoke S, Malarstig A, Miller MR, Dumitriu A. Application of machine learning methods to predict non-alcoholic steatohepatitis (NASH) in non-alcoholic fatty liver (NAFL) patients. AMIA Annu Symp Proc. 2018;2018:430–439.
Vali Y, Lee J, Schattenberg J, Gomez MR, Tiniakos D, Bedossa P, et al. Comparative diagnostic accuracy of blood-based biomarkers for diagnosing NASH: Phase 1 results of the LITMUS project. International Liver Congress; 2021.
Sanyal AJ, Williams YA, Lavine JE, Neuschwander-Tetri BA, Alexander L, Ostroff R, et al. Defining the serum proteomic signature of hepatic steatosis, inflammation, ballooning and fibrosis in non-alcoholic fatty liver disease. J Hepatol. 2023;78:693–703.
Tanwar S, Trembling PM, Guha IN, Parkes J, Kaye P, Burt AD, et al. Validation of terminal peptide of procollagen III for the detection and assessment of nonalcoholic steatohepatitis in patients with nonalcoholic fatty liver disease. Hepatology. 2013;57:103–111.
Darweesh SK, AbdElAziz RA, Abd-ElFatah DS, AbdElazim NA, Fathi SA, Attia D, et al. Serum cytokeratin-18 and its relation to liver fibrosis and steatosis diagnosed by FibroScan and controlled attenuation parameter in nonalcoholic fatty liver disease and hepatitis C virus patients. Eur J Gastroenterol Hepatol. 2019;31:633–641.
Feldstein AE, Alkhouri N, De Vito R, Alisi A, Lopez R, Nobili V. Serum cytokeratin-18 fragment levels are useful biomarkers for nonalcoholic steatohepatitis in children. Am J Gastroenterol. 2013;108:1526–1531.
Feng G, Zheng KI, Li Y-Y, Rios RS, Zhu P-W, Pan X-Y, et al. Machine learning algorithm outperforms fibrosis markers in predicting significant fibrosis in biopsy-confirmed NAFLD. J Hepatobiliary Pancreat Sci. 2021;28:593–603.
Christodoulou E, Ma J, Collins GS, Steyerberg EW, Verbakel JY, Van Calster B. A systematic review shows no performance benefit of machine learning over logistic regression for clinical prediction models. J Clin Epidemiol. 2019;110:12–22.