Predicting Response to Methotrexate Therapy in New-Onset RA: No Value of Adding Protein Biomarkers to Clinical Predictors

Xavier M Teitsma¹, Johannes W. G. Jacobs², Pascal HP de Jong³, JMW Hazes³, Angelique EAM Weel^3,4, Paco MJ Welsing¹, Attila Pethö-Schramm⁵, Michelle EA Borm⁶, Jacob van Laar¹, Johannes W. J. Bijlsma⁷ and Floris PJG Lafeber⁸, ¹Department of Rheumatology & Clinical Immunology, University Medical Center Utrecht, Utrecht, the Netherlands, Utrecht, Netherlands, ²Rheumatolgy & Clinical Immunology, University Medical Center Utrecht, Utrecht, Netherlands, ³Rheumatology, Erasmus Medical Center Rotterdam, Rotterdam, Netherlands, ⁴Rheumatology, Maastad Hospital, Rotterdam, Netherlands, ⁵F. Hoffmann-La Roche, Basel, Switzerland, ⁶Roche Nederland BV, Woerden, Netherlands, ⁷Department of Rheumatology & Clinical Immunology, University Medical Center Utrecht, Utrecht, Netherlands, ⁸Rheumatology & Clinical Immunology, University Medical Center Utrecht, Utrecht, Netherlands

Meeting: ACR/ARHP Annual Meeting 2018

Keywords: methotrexate (MTX), prognostic factors, rheumatoid arthritis (RA) and treatment

Session Information

Date: Monday, October 22, 2018

Title: Rheumatoid Arthritis – Diagnosis, Manifestations, and Outcomes Poster II: Diagnosis and Prognosis

Session Type: ACR Poster Session B

Session Time: 9:00AM-11:00AM

Background/Purpose: We previously identified higher disease activity score assessing 28 joints (DAS28) at baseline, current smoking and no alcohol consumption as clinical predictors for inadequate response (IR) to a methotrexate (MTX) strategy.¹ In addition, we showed the potential of protein biomarkers to improve the prediction of clinical response in patients likely to show IR to MTX.² In this analysis, we aimed to validate predictive protein biomarkers with additive value to the clinical predictors mentioned above in predicting IR to MTX.

Methods: Data was used from patients initiating MTX therapy in the U-Act-Early trial for developing the prediction model.³ If remission (DAS28) <2.6 with ≤4 swollen joints for ≥24 weeks) was not achieved, hydroxychloroquine was added and thereafter replaced by tocilizumab if remission still was not achieved. Serum was collected at baseline and analyzed using multi-analyte (n=85) profiling; relevant proteins were identified using partial least square discriminant analyses. Data of patients treated with MTX plus glucocorticoid (GC) in the Rotterdam Early Arthritis Cohort trial were used for validation.⁴ These patients switched to MTX plus etanercept, if remission (DAS44 <2.4) was not achieved. IR to MTX was defined for both studies as the need of adding a biological DMARD within the first year to MTX, including additional conventional synthetic disease modifying anti-rheumatic drug (csDMARD) or GC therapy (here designated as ‘MTX+’).

Results: None of the potentially relevant proteins (n=7) in the development sample were found to be significant (IR vs. no IR) in the validation sample. Although not statistically significant, the best performing biomarker was vascular cell adhesion protein 1 (VCAM-1, odds ratio (OR) 1.36, 95%-confidence interval (CI), 0.84-2.32; p=0.19), and was further investigated. In the development sample, the area under the receiver operator characteristics curve (AUROC) of the combined model (i.e. clinical predictors plus VCAM-1) was 0.80 (95%-CI 0.71-0.88): no significantly improved discriminative ability compared to clinical model (AUROC 0.75, 95%-CI 0.65-0.84; p=0.051). When applying the linear predictor to the validation sample, an AUROC of 0.67 (95%-CI 0.55-0.80) was found for the combined model: neither significantly different from clinical model (AUROC 0.68, 95%-CI 0.55-0.80).

Conclusion: No protein biomarkers with additive value to clinical predictors for predicting IR to ‘MTX+’ in new-onset RA could be externally validated. The relatively easily obtainable clinical predictors still remain the most accurate factors to predict the need for adding a biologic DMARD to the initial MTX-based treatment strategy.

References:

1. Teitsma XM, et al. Ann Rheum Dis 2018. May 14. [Epub ahead of print]

2. Teitsma XM, et al. Ann Rheum Dis 2016;75 (Suppl 2):440.

3. Bijlsma JWJ, et al. Lancet 2016 23;388(10042):343-355.

4. de Jong PH, et al. Ann Rheum Dis 2014;73(7):1331-1339.

Disclosure: X. M. Teitsma, None; J. W. G. Jacobs, None; P. H. de Jong, None; J. Hazes, None; A. E. Weel, None; P. M. Welsing, None; A. Pethö-Schramm, F Hoffmann-La Roche, 3; M. E. Borm, Roche Nederland BV, 3; J. van Laar, Arthrogen, MSD, Pfizer, Eli Lelly, BMS, Astra Zeneca, Roche-Genentech, 2, 5; J. W. J. Bijlsma, Roche, AbbVie, Bristol-Myers, Squibb, Merck Sharp and Dohme, Pfizer, UCB, 2, 5; F. P. Lafeber, None.

To cite this abstract in AMA style:

Teitsma XM, Jacobs JWG, de Jong PH, Hazes J, Weel AE, Welsing PM, Pethö-Schramm A, Borm ME, van Laar J, Bijlsma JWJ, Lafeber FP. Predicting Response to Methotrexate Therapy in New-Onset RA: No Value of Adding Protein Biomarkers to Clinical Predictors [abstract]. Arthritis Rheumatol. 2018; 70 (suppl 9). https://acrabstracts.org/abstract/predicting-response-to-methotrexate-therapy-in-new-onset-ra-no-value-of-adding-protein-biomarkers-to-clinical-predictors/. Accessed .

ACR Meeting Abstracts - https://acrabstracts.org/abstract/predicting-response-to-methotrexate-therapy-in-new-onset-ra-no-value-of-adding-protein-biomarkers-to-clinical-predictors/