Minh Vu Chuong Nguyen1, Athan Baillet2, Xavier Romand3, Candice Trocmé4, Anaïs Courtier5, Hubert Marotte6, Thierry Thomas7, Martin Soubrier8, Pierre Miossec9, Jacques Tébib10, Laurent Grange11, Bertrand Toussaint12, Thierry Lequerré10, Olivier Vittecoq10, Philippe Gaudin3. 1. EA 7408, University Grenoble Alpes, GREPI, 38400 Saint-Martin-d'Hères, France; Sinnovial, 38000 Grenoble, France. 2. EA 7408, University Grenoble Alpes, GREPI, 38400 Saint-Martin-d'Hères, France; Rheumatology Department, centre hospitalier universitaire Grenoble Alpes, hôpital Sud Echirolles, 38130 Echirolles, France. Electronic address: ABaillet@chu-grenoble.fr. 3. EA 7408, University Grenoble Alpes, GREPI, 38400 Saint-Martin-d'Hères, France; Rheumatology Department, centre hospitalier universitaire Grenoble Alpes, hôpital Sud Echirolles, 38130 Echirolles, France. 4. Pôle biologie, hôpital Michallon, centre hospitalier universitaire Grenoble Alpes, 38700 La Tronche, France. 5. Sinnovial, 38000 Grenoble, France. 6. Department of Rheumatology, Saint-Etienne University Hospital, 42270 Saint-Etienne, France; Inserm 1059, SAINBIOSE University of Saint-Etienne, 42270 Saint-Etienne, France. 7. Service de rhumatologie, hôpital G.-Montpied, centre hospitalier universitaire Clermont-Ferrand, 63003 Clermont-Ferrand, France; Department of Immunology and Rheumatology, Immunogenomics and Inflammation Research Unit EA 4130, University of Lyon, Edouard Herriot Hospital, 69003 Lyon, France. 8. Rheumatology Department, centre hospitalier Lyon-Sud, 69310 Pierre-Bénite, France. 9. TIMC-IMAG Laboratory UMR 5525, CNRS, université Grenoble Alpes, 38041 Grenoble, France. 10. Inserm CIC/CRB1404, Normandie Univ, UNIROUEN, Inserm U 1234, Rouen University Hospital, Department of Rheumatology, 76031 Rouen cedex, France. 11. Rheumatology Department, centre hospitalier universitaire Grenoble Alpes, hôpital Sud Echirolles, 38130 Echirolles, France. 12. Pôle biologie, hôpital Michallon, centre hospitalier universitaire Grenoble Alpes, 38700 La Tronche, France; TIMC-IMAG Laboratory UMR 5525, CNRS, université Grenoble Alpes, 38041 Grenoble, France.
Abstract
OBJECTIVES: Tumour necrosis factor-alpha inhibitors (TNFi) are effective treatments for Rheumatoid Arthritis (RA). Responses to treatment are barely predictable. As these treatments are costly and may induce a number of side effects, we aimed at identifying a panel of protein biomarkers that could be used to predict clinical response to TNFi for RA patients. METHODS: Baseline blood levels of C-reactive protein, platelet factor 4, apolipoprotein A1, prealbumin, α1-antitrypsin, haptoglobin, S100A8/A9 and S100A12 proteins in bDMARD naive patients at the time of TNFi treatment initiation were assessed in a multicentric prospective French cohort. Patients fulfilling good EULAR response at 6 months were considered as responders. Logistic regression was used to determine best biomarker set that could predict good clinical response to TNFi. RESULTS: A combination of biomarkers (prealbumin, platelet factor 4 and S100A12) was identified and could predict response to TNFi in RA with sensitivity of 78%, specificity of 77%, positive predictive values (PPV) of 72%, negative predictive values (NPV) of 82%, positive likelihood ratio (LR+) of 3.35 and negative likelihood ratio (LR-) of 0.28. Lower levels of prealbumin and S100A12 and higher level of platelet factor 4 than the determined cutoff at baseline in RA patients are good predictors for response to TNFi treatment globally as well as to Infliximab, Etanercept and Adalimumab individually. CONCLUSION: A multivariate model combining 3 biomarkers (prealbumin, platelet factor 4 and S100A12) accurately predicted response of RA patients to TNFi and has potential in a daily practice personalized treatment.
OBJECTIVES:Tumour necrosis factor-alpha inhibitors (TNFi) are effective treatments for Rheumatoid Arthritis (RA). Responses to treatment are barely predictable. As these treatments are costly and may induce a number of side effects, we aimed at identifying a panel of protein biomarkers that could be used to predict clinical response to TNFi for RApatients. METHODS: Baseline blood levels of C-reactive protein, platelet factor 4, apolipoprotein A1, prealbumin, α1-antitrypsin, haptoglobin, S100A8/A9 and S100A12 proteins in bDMARD naive patients at the time of TNFi treatment initiation were assessed in a multicentric prospective French cohort. Patients fulfilling good EULAR response at 6 months were considered as responders. Logistic regression was used to determine best biomarker set that could predict good clinical response to TNFi. RESULTS: A combination of biomarkers (prealbumin, platelet factor 4 and S100A12) was identified and could predict response to TNFi in RA with sensitivity of 78%, specificity of 77%, positive predictive values (PPV) of 72%, negative predictive values (NPV) of 82%, positive likelihood ratio (LR+) of 3.35 and negative likelihood ratio (LR-) of 0.28. Lower levels of prealbumin and S100A12 and higher level of platelet factor 4 than the determined cutoff at baseline in RApatients are good predictors for response to TNFi treatment globally as well as to Infliximab, Etanercept and Adalimumab individually. CONCLUSION: A multivariate model combining 3 biomarkers (prealbumin, platelet factor 4 and S100A12) accurately predicted response of RApatients to TNFi and has potential in a daily practice personalized treatment.