Etienne Audureau1, Fabrice Carrat2, Richard Layese1, Carole Cagnot3, Tarik Asselah4, Dominique Guyader5, Dominique Larrey6, Victor De Lédinghen7, Denis Ouzan8, Fabien Zoulim9, Dominique Roulot10, Albert Tran11, Jean-Pierre Bronowicki12, Jean-Pierre Zarski13, Ghassan Riachi14, Paul Calès15, Jean-Marie Péron16, Laurent Alric17, Marc Bourlière18, Philippe Mathurin19, Jean-Frédéric Blanc20, Armand Abergel21, Olivier Chazouillères22, Ariane Mallat23, Jean-Didier Grangé24, Pierre Attali25, Louis d'Alteroche26, Claire Wartelle27, Thông Dao28, Dominique Thabut29, Christophe Pilette30, Christine Silvain31, Christos Christidis32, Eric Nguyen-Khac33, Brigitte Bernard-Chabert34, David Zucman35, Vincent Di Martino36, Angela Sutton37, Stanislas Pol38, Pierre Nahon39. 1. AP-HP, Hôpital Henri Mondor, Département de Santé Publique, and Université Paris-Est, A-TVB DHU, CEpiA (Clinical Epidemiology and Ageing) Unit EA7376, UPEC, F-94000, Créteil, France. 2. Sorbonne Université, Inserm, Institut Pierre Louis d'Epidémiologie et de Santé Publique, Paris, France; AP-HP, Hôpital Saint-Antoine, Unité de Santé Publique, Paris, France. 3. Unit for Basic and Clinical research on Viral Hepatitis, ANRS (France REcherche Nord & sud Sida-HIV Hépatites-FRENSH). 4. AP-HP, Hôpital Beaujon, Service d'Hépatologie, Clichy. 5. CHU Pontchaillou, Service d'Hépatologie, Rennes. 6. Hôpital Saint Eloi, Service d'Hépatologie, Montpellier. 7. Hôpital Haut-Lévêque, Service d'Hépatologie, Bordeaux. 8. Institut Arnaud Tzanck, Service d'Hépatologie, St Laurent du Var. 9. Hospices Civils de Lyon, Service d'Hépatologie; INSERM U1052 - CRCL; Université de Lyon, Lyon. 10. AP-HP, Hôpital Avicenne, Service d'Hépatologie, Bobigny. 11. CHU de Nice, Service d'Hépatologie, F-06202, Cedex 3, Nice; Inserm U1065, C3M, Team 8, "Hepatic Complications in Obesity", F-06204, Cedex 3, Nice. 12. Inserm 954, CHU de Nancy, Université de Lorraine, Vandoeuvre-les-Nancy. 13. Hôpital Michallon, Service d'Hépatologie, Grenoble. 14. Hôpital Charles-Nicolle, Service d'Hépatologie, Rouen. 15. CHU d'Angers, Service d'Hépato-Gastroentérologie, Angers. 16. Hôpital Purpan, Service d'Hépatologie, Toulouse. 17. CHU Toulouse, Service de Médecine Interne-Pôle Digestif UMR 152, Toulouse. 18. Hôpital Saint Joseph, Service d'Hépatologie, Marseille. 19. Hôpital Claude Huriez, Service d'Hépatologie, Lille. 20. Hôpital St André, Service d'Hépatologie, Bordeaux et Hôpital Haut-Lévêque, CHU Bordeaux, 33604 Pessac. 21. Hôpital Hôtel Dieu, Service d'Hépatologie, Clermont-Ferrand. 22. AP-HP, Hôpital Saint-Antoine, Service d'Hépatologie, and Sorbonne Université, INSERM, Centre de Recherche Saint-Antoine, Paris. 23. AP-HP, Hôpital Henri Mondor, Service d'Hépatologie, Créteil. 24. AP-HP, Hôpital Tenon, Service d'Hépatologie, Paris. 25. AP-HP, Hôpital Paul Brousse, Service d'Hépatologie, Villejuif. 26. Hôpital Trousseau, Unité d'Hépatologie, CHRU de Tours. 27. Hôpital d'Aix-En-Provence, Service d'Hépatologie, Aix-En-Provence. 28. Hôpital de la Côte de Nacre, Service d'Hépatologie, Caen. 29. AP-HP, Groupe Hospitalier de La Pitié-Salpêtrière, Service d'Hépatologie, Paris. 30. CHU Le Mans, Service d'Hépatologie, Le Mans. 31. CHU de Poitiers, Service d'Hépatologie, Poitiers. 32. Institut Mutualiste Montsouris, Service d'Hépatologie, Paris. 33. CHU Amiens Picardie Hôpital Sud, Service d'Hépatologie, Amiens. 34. Hôpital Robert Debré, Service d'Hépatologie, Reims. 35. Hôpital Foch, Service de Médecine Interne, Suresnes. 36. Hôpital Jean Minjoz, Service d'Hépatologie, Besançon. 37. CRB (liver disease biobank) Groupe Hospitalier Paris Seine-Saint-Denis BB-0033-00027; AP-HP, Hôpital Jean Verdier, Service de Biochimie, Bondy; Inserm U1148, Université Paris 13, Bobigny. 38. AP-HP, Hôpital Cochin, Département d'Hépatologie; Inserm UMS20 et U1223, Institut Pasteur, Université Paris Descartes, Paris. 39. AP-HP, Hôpital Jean Verdier, Service d'Hépatologie, Bondy; Université Paris 13, Sorbonne Paris Cité, "Equipe labellisée Ligue Contre le Cancer", F-93206 Saint-Denis; Inserm, UMR-1162, "Génomique fonctionnelle des tumeur solides", F-75000, Paris, France. Electronic address: pierre.nahon@aphp.fr.
Abstract
BACKGROUND & AIMS: Refining hepatocellular carcinoma (HCC) surveillance programs requires improved individual risk prediction. Thus, we aimed to develop algorithms based on machine learning approaches to predict the risk of HCC more accurately in patients with HCV-related cirrhosis, according to their virological status. METHODS: Patients with compensated biopsy-proven HCV-related cirrhosis from the French ANRS CO12 CirVir cohort were included in a semi-annual HCC surveillance program. Three prognostic models for HCC occurrence were built, using (i) Fine-Gray regression as a benchmark, (ii) single decision tree (DT), and (iii) random survival forest for competing risks survival (RSF). Model performance was evaluated from C-indexes validated externally in the ANRS CO22 Hepather cohort (n = 668 enrolled between 08/2012-01/2014). RESULTS: Out of 836 patients analyzed, 156 (19%) developed HCC and 434 (52%) achieved sustained virological response (SVR) (median follow-up 63 months). Fine-Gray regression models identified 6 independent predictors of HCC occurrence in patients before SVR (past excessive alcohol intake, genotype 1, elevated AFP and GGT, low platelet count and albuminemia) and 3 in patients after SVR (elevated AST, low platelet count and shorter prothrombin time). DT analysis confirmed these associations but revealed more complex interactions, yielding 8 patient groups with varying cancer risks and predictors depending on SVR achievement. On RSF analysis, the most important predictors of HCC varied by SVR status (non-SVR: platelet count, GGT, AFP and albuminemia; SVR: prothrombin time, ALT, age and platelet count). Externally validated C-indexes before/after SVR were 0.64/0.64 [Fine-Gray], 0.60/62 [DT] and 0.71/0.70 [RSF]. CONCLUSIONS: Risk factors for hepatocarcinogenesis differ according to SVR status. Machine learning algorithms can refine HCC risk assessment by revealing complex interactions between cancer predictors. Such approaches could be used to develop more cost-effective tailored surveillance programs. LAY SUMMARY: Patients with HCV-related cirrhosis must be included in liver cancer surveillance programs, which rely on ultrasound examination every 6 months. Hepatocellular carcinoma (HCC) screening is hampered by sensitivity issues, leading to late cancer diagnoses in a substantial number of patients. Refining surveillance periodicity and modality using more sophisticated imaging techniques such as MRI may only be cost-effective in patients with the highest HCC incidence. Herein, we demonstrate how machine learning algorithms (i.e. data-driven mathematical models to make predictions or decisions), can refine individualized risk prediction.
BACKGROUND & AIMS: Refining hepatocellular carcinoma (HCC) surveillance programs requires improved individual risk prediction. Thus, we aimed to develop algorithms based on machine learning approaches to predict the risk of HCC more accurately in patients with HCV-related cirrhosis, according to their virological status. METHODS:Patients with compensated biopsy-proven HCV-related cirrhosis from the French ANRS CO12 CirVir cohort were included in a semi-annual HCC surveillance program. Three prognostic models for HCC occurrence were built, using (i) Fine-Gray regression as a benchmark, (ii) single decision tree (DT), and (iii) random survival forest for competing risks survival (RSF). Model performance was evaluated from C-indexes validated externally in the ANRS CO22 Hepather cohort (n = 668 enrolled between 08/2012-01/2014). RESULTS: Out of 836 patients analyzed, 156 (19%) developed HCC and 434 (52%) achieved sustained virological response (SVR) (median follow-up 63 months). Fine-Gray regression models identified 6 independent predictors of HCC occurrence in patients before SVR (past excessive alcohol intake, genotype 1, elevated AFP and GGT, low platelet count and albuminemia) and 3 in patients after SVR (elevated AST, low platelet count and shorter prothrombin time). DT analysis confirmed these associations but revealed more complex interactions, yielding 8 patient groups with varying cancer risks and predictors depending on SVR achievement. On RSF analysis, the most important predictors of HCC varied by SVR status (non-SVR: platelet count, GGT, AFP and albuminemia; SVR: prothrombin time, ALT, age and platelet count). Externally validated C-indexes before/after SVR were 0.64/0.64 [Fine-Gray], 0.60/62 [DT] and 0.71/0.70 [RSF]. CONCLUSIONS: Risk factors for hepatocarcinogenesis differ according to SVR status. Machine learning algorithms can refine HCC risk assessment by revealing complex interactions between cancer predictors. Such approaches could be used to develop more cost-effective tailored surveillance programs. LAY SUMMARY:Patients with HCV-related cirrhosis must be included in liver cancer surveillance programs, which rely on ultrasound examination every 6 months. Hepatocellular carcinoma (HCC) screening is hampered by sensitivity issues, leading to late cancer diagnoses in a substantial number of patients. Refining surveillance periodicity and modality using more sophisticated imaging techniques such as MRI may only be cost-effective in patients with the highest HCC incidence. Herein, we demonstrate how machine learning algorithms (i.e. data-driven mathematical models to make predictions or decisions), can refine individualized risk prediction.
Authors: Jérémy Dana; Aïna Venkatasamy; Antonio Saviano; Joachim Lupberger; Yujin Hoshida; Valérie Vilgrain; Pierre Nahon; Caroline Reinhold; Benoit Gallix; Thomas F Baumert Journal: Hepatol Int Date: 2022-02-09 Impact factor: 9.029
Authors: Hamish Innes; Peter Jepsen; Scott McDonald; John Dillon; Victoria Hamill; Alan Yeung; Jennifer Benselin; April Went; Andrew Fraser; Andrew Bathgate; M Azim Ansari; Stephen T Barclay; David Goldberg; Peter C Hayes; Philip Johnson; Eleanor Barnes; William Irving; Sharon Hutchinson; Indra Neil Guha Journal: JHEP Rep Date: 2021-10-07