Dataset Information

Machine learning identifies a profile of inadequate responder to methotrexate in rheumatoid arthritis.

ABSTRACT:

Objectives

Around 30% of patients with RA have an inadequate response to MTX. We aimed to use routine clinical and biological data to build machine learning models predicting EULAR inadequate response to MTX and to identify simple predictive biomarkers.

Methods

Models were trained on RA patients fulfilling the 2010 ACR/EULAR criteria from the ESPOIR and Leiden EAC cohorts to predict the EULAR response at 9 months (± 6 months). Several models were compared on the training set using the AUROC. The best model was evaluated on an external validation cohort (tREACH). The model's predictions were explained using Shapley values to extract a biomarker of inadequate response.

Results

We included 493 therapeutic sequences from ESPOIR, 239 from EAC and 138 from tREACH. The model selected DAS28, Lymphocytes, Creatininemia, Leucocytes, AST, ALT, swollen joint count and corticosteroid co-treatment as predictors. The model reached an AUROC of 0.72 [95% CI (0.63, 0.80)] on the external validation set, where 70% of patients were responders to MTX. Patients predicted as inadequate responders had only 38% [95% CI (20%, 58%)] chance to respond and using the algorithm to decide to initiate MTX would decrease inadequate-response rate from 30% to 23% [95% CI: (17%, 29%)]. A biomarker was identified in patients with moderate or high activity (DAS28 > 3.2): patients with a lymphocyte count superior to 2000 cells/mm3 are significantly less likely to respond.

Conclusion

Our study highlights the usefulness of machine learning in unveiling subgroups of inadequate responders to MTX to guide new therapeutic strategies. Further work is needed to validate this approach.

SUBMITTER: Duquesne J

PROVIDER: S-EPMC10321123 | biostudies-literature | 2023 Jul

REPOSITORIES: biostudies-literature

ACCESS DATA

Publications

Machine learning identifies a profile of inadequate responder to methotrexate in rheumatoid arthritis.

Duquesne Julien J Bouget Vincent V Cournède Paul Henry PH Fautrel Bruno B Guillemin Francis F de Jong Pascal H P PHP Heutz Judith W JW Verstappen Marloes M van der Helm-van Mil Annette H M AHM Mariette Xavier X Bitoun Samuel S

Rheumatology (Oxford, England) 20230701 7

<h4>Objectives</h4>Around 30% of patients with RA have an inadequate response to MTX. We aimed to use routine clinical and biological data to build machine learning models predicting EULAR inadequate response to MTX and to identify simple predictive biomarkers.<h4>Methods</h4>Models were trained on RA patients fulfilling the 2010 ACR/EULAR criteria from the ESPOIR and Leiden EAC cohorts to predict the EULAR response at 9 months (± 6 months). Several models were compared on the training set using ...[more]

PMID: 36416134

Similar Datasets

Project description:BackgroundMethotrexate is the preferred initial disease-modifying antirheumatic drug (DMARD) for rheumatoid arthritis (RA). However, clinically useful tools for individualized prediction of response to methotrexate treatment in patients with RA are lacking. We aimed to identify clinical predictors of response to methotrexate in patients with rheumatoid arthritis (RA) using machine learning methods.MethodsRandomized clinical trials (RCT) of patients with RA who were DMARD-naïve and randomized to placebo plus methotrexate were identified and accessed through the Clinical Study Data Request Consortium and Vivli Center for Global Clinical Research Data. Studies with available Disease Activity Score with 28-joint count and erythrocyte sedimentation rate (DAS28-ESR) at baseline and 12 and 24 weeks were included. Latent class modeling of methotrexate response was performed. The least absolute shrinkage and selection operator (LASSO) and random forests methods were used to identify predictors of response.ResultsA total of 775 patients from 4 RCTs were included (mean age 50 years, 80% female). Two distinct classes of patients were identified based on DAS28-ESR change over 24 weeks: "good responders" and "poor responders." Baseline DAS28-ESR, anti-citrullinated protein antibody (ACPA), and Health Assessment Questionnaire (HAQ) score were the top predictors of good response using LASSO (area under the curve [AUC] 0.79) and random forests (AUC 0.68) in the external validation set. DAS28-ESR ≤ 7.4, ACPA positive, and HAQ ≤ 2 provided the highest likelihood of response. Among patients with 12-week DAS28-ESR > 3.2, ≥ 1 point improvement in DAS28-ESR baseline-to-12-week was predictive of achieving DAS28-ESR ≤ 3.2 at 24 weeks.ConclusionsWe have developed and externally validated a prediction model for response to methotrexate within 24 weeks in DMARD-naïve patients with RA, providing variably weighted clinical features and defined cutoffs for clinical decision-making.

Project description:BackgroundSarilumab is a human monoclonal antibody directed against the alpha subunit of the interleukin-6 receptor complex. In the MOBILITY phase III randomized controlled trial (RCT), sarilumab + methotrexate (MTX) treatment resulted in clinical improvements at 24 weeks that were maintained at 52 weeks in adults with rheumatoid arthritis (RA), who have inadequate response to MTX (MTX-IR). These analyses indicate the effects of sarilumab + MTX versus placebo on patient-reported outcomes (PROs) in this RCT.MethodsPatients (n = 1197) were randomized to receive placebo, sarilumab 150 or 200 mg subcutaneously + MTX every 2 weeks for 52 weeks; after 16 weeks, patients without ≥20 % improvement from baseline in swollen or tender joint counts on two consecutive assessments were offered open-label treatment. PROs included patient global assessment of disease activity (PtGA), pain, health assessment questionnaire disability index (HAQ-DI), Short Form-36 Health Survey (SF-36), and functional assessment of chronic illness therapy-fatigue (FACIT-F). Changes from baseline at weeks 24 and 52 were analyzed using a mixed model for repeated measures. Post hoc analyses included percentages of patients reporting improvements equal to or greater than minimal clinically important differences (MCID) and normative values in the FACIT-F and SF-36. Pearson correlation between observed PRO scores and clinical measures of disease activity was tested at week 24.ResultsBoth doses of sarilumab + MTX vs placebo + MTX resulted in improvement from baseline by week 24 in PtGA, pain, HAQ-DI, SF-36 and FACIT-F scores (p < 0.0001) that was clinically meaningful, and persisted until week 52. In post hoc analyses, the percentages of patients with improvement equal to or greater than the MCID across all PROs were greater with sarilumab than placebo (p < 0.05), with differences ranging from 11.6 to 26.2 %, as were those reporting equal to or greater than normative scores.ConclusionsIn this RCT in patients with MTX-IR RA, sarilumab + MTX resulted in sustained improvement in PROs that were clinically meaningful, greater than placebo + MTX, and complement the previously reported clinical efficacy and safety of sarilumab.Trial registrationClinicalTrials.gov. NCT01061736 . February 2, 2010.

Project description:BACKGROUND:Sarilumab is a human immunoglobulin G1 anti-interleukin-6 (IL-6) receptor monoclonal antibody that blocks IL-6 from binding to membrane-bound and soluble IL-6 receptor α. This bridging study assessed the efficacy and safety of sarilumab + methotrexate (MTX) in Japanese patients with active rheumatoid arthritis (RA) and inadequate response to MTX (MTX-IR). METHODS:In this phase III study, 243 patients were randomized 2:2:1:1 to receive subcutaneous sarilumab 150 mg every 2 weeks (q2w), sarilumab 200 mg q2w, placebo switching to sarilumab 150 mg q2w + MTX at 24 weeks, or placebo switching to sarilumab 200 mg q2w at 24 weeks, all in combination with MTX, for a total of 52 weeks (double-blind, placebo-controlled 24-week period followed by a single-blind 28-week extension). The primary endpoint was the proportion of patients achieving American College of Rheumatology 20% improvement criteria (ACR20) responses at week 24. RESULTS:ACR20 response rates at week 24 were 67.9%, 57.5%, and 14.8% for sarilumab 150 mg, sarilumab 200 mg, and placebo, respectively. Serious treatment-emergent adverse events were reported by 9.9%, 6.3%, 0%, and 13.3% of patients in the sarilumab 150 mg, sarilumab 200 mg, placebo to sarilumab 150 mg, and placebo to sarilumab 200 mg groups, respectively. No deaths occurred. The incidence of infections ranged from 52.5 to 67.9%, with five serious infections for the sarilumab 150 mg group and one for the group switched from placebo to 200 mg sarilumab. Absolute neutrophil count < 1.0 Giga/l occurred in 13.6% and 7.5% of patients in the sarilumab 150 and 200 mg groups, respectively, and was not associated with infection. CONCLUSIONS:In Japanese MTX-IR RA patients treated with sarilumab (150 and 200 mg q2w) in combination with MTX, sustained clinical efficacy was shown by significant improvement in signs, symptoms, and physical function; bridging between this and a previous global study was achieved. At week 52, the safety profiles of both doses of sarilumab were generally similar, as previously observed and as expected based on IL-6 class. TRIAL REGISTRATION:ClinicalTrials.gov, NCT02293902 . Registered on 19 November 2014.

Project description:IntroductionOur aim was to evaluate protein biomarker changes related to the administration of filgotinib, a Janus kinase (JAK) 1 preferential inhibitor, in patients with moderately to severely active rheumatoid arthritis (RA) with inadequate response to methotrexate.MethodsPlasma and serum samples were collected from patients enrolled in FINCH 1 (NCT02889796), a Phase 3 trial. Patients with stable backgrounds of methotrexate were randomly assigned once-daily oral filgotinib 200 or 100 mg, subcutaneous adalimumab 40 mg every 2 weeks (W), or placebo. Up to 35 biomarkers were analyzed at baseline, W4, and W12 with enzyme-linked immunosorbent assays and chemiluminescence and electrochemiluminescence assays.ResultsAt baseline, four distinct biomarker clusters were identified. The strongest intragroup correlations were in bone-cartilage resorption/inflammation and JAK/signal transducer and activator of transcription (STAT) signaling activity. At baseline, significant positive correlations were identified for cytokines with patient-reported pain and with patient measures of fatigue. Filgotinib reduced levels of cytokines associated with inflammation and cell migration as early as W4 and through W12. Compared to adalimumab, filgotinib induced significant reductions in bone-related turnover biomarkers, N-telopeptide of type 1 collagen and C-telopeptide 1, as well as biomarkers associated with baseline disease activity. No baseline predictors of therapeutic response to filgotinib were identified.ConclusionsFilgotinib reduced peripheral protein biomarkers associated with JAK/STAT signaling, inflammatory signaling, immune cell migration, and bone resorption as soon as W4 in FINCH 1. Effects were dose-dependent and consistent with the clinical efficacy of filgotinib observed in FINCH 1. The changes in peripheral biomarkers associated with filgotinib treatment in methotrexate-experienced patients are consistent with changes observed in both methotrexate-naïve and biologic disease-modifying antirheumatic drug-experienced RA populations. These data demonstrate dose-dependent effects of preferential JAK1 inhibition by filgotinib on peripheral blood protein biomarkers in methotrexate-experienced patients with RA.Trial registrationClinicalTrials.gov, NCT02889796.

Dataset Information

Machine learning identifies a profile of inadequate responder to methotrexate in rheumatoid arthritis.

Objectives

Methods

Results

Conclusion

Publications

Machine learning identifies a profile of inadequate responder to methotrexate in rheumatoid arthritis.

Similar Datasets

OmicsDI is part of the ELIXIR infrastructure

Tweets