key: cord-0063314-txvr42cu
authors: Shipa, Muhammad R; Yeoh, Su-Ann; Embleton-Thirsk, Andrew; Mukerjee, Dev; Ehrenstein, Michael R
title: O07 The increase in erythrocyte mean corpuscular volume by methotrexate is potentiated by hydroxychloroquine and is an early indicator of clinical response in rheumatoid arthritis
date: 2021-04-26
journal: Rheumatology (Oxford)
DOI: 10.1093/rheumatology/keab246.006
sha: 696800b7e2c654d97e1c94a663c71348a608b947
doc_id: 63314
cord_uid: txvr42cu

Background/Aims Rheumatologists are facing a significant challenge in the management of early rheumatoid arthritis (RA) due to limitations placed on outpatient visits during the COVID-19 pandemic. Frequent clinical assessments of disease activity are recommended during implementation of the treat to target strategy to achieve remission. A biomarker indicating response to methotrexate during the early phase of therapy could complement clinical examination. Methotrexate increases erythrocyte mean corpuscular volume (MCV), which is measured routinely, prompting us to investigate whether changes in MCV could act as an early indicator of response. Methods Patients with early RA who were started on methotrexate therapy were included from two independent cohorts. The larger cohort (discovery cohort, n = 655) was used to build the model and the second cohort (validation cohort, n = 225) was applied to test the prediction of the model. Conventional statistical, and machine learning approaches were adopted to identify key determinants that influence the potential relationship between MCV and clinical response, defined as attainment of remission or low disease activity, at six months after starting methotrexate. Results A LASSO penalised logistic regression model was built with the discovery cohort [area under the receiver operating characteristics (AUROC) curve = 0.76], where change of MCV from three months [Odds ratio (OR) 1.53 (95% CI 1.38-1.70)], concomitant use of hydroxychloroquine [OR 2.16 (95% CI 1.52 - 3.07, p < 0.001)], and seropositivity [OR 1.83 (95% CI 1.12 - 3.03, p = 0.02)] were associated with favourable methotrexate response [accuracy 81% (95% CI 76%-86%) of the model testing against discovery model]. Different machine learning classification methods were applied. Random forest exhibited the maximum accuracy and AUROC (89%, and 86%, respectively), confirming the above three predictors as the most significant. Two latent classes (class 1: smaller MCV increase and class 2: greater MCV increase) were identified based on the MCV changes over six months. Class 1 had fewer responders and a lower number of patients on hydroxychloroquine compared to class 2. The earliest time point of significant difference of MCV between responders and non-responders was three months [mean difference 1.43 (95% CI 0.57-2.3)]. Combination hydroxychloroquine and methotrexate caused the greatest increase in MCV with a difference between responders and non-responders at 2 months. Change of MCV at three months showed AUROC of 0.75 to predict treatment response to the combination of methotrexate and hydroxychloroquine at six months with an optimal cut-point of MCV 3.5 fL (95% CI 3.5-3.6) with 71% sensitivity and 75%, specificity. Conclusion Our data provides mechanistic insight into the synergistic clinical benefit of concomitant hydroxychloroquine with methotrexate, boosting the rise in erythrocyte MCV which could serve as an early biomarker of treatment response. Disclosure M.R. Shipa: Grants/research support; Versus arthritis. S. Yeoh: Grants/research support; Royal College of Physicians, Rosetrees Trust, NIHR University College London Hospitals Biomedical Research Centre, UCLH Charities. A. Embleton-Thirsk: None. D. Mukerjee: None. M.R. Ehrenstein: Grants/research support; University College London Hospital Biomedical Research Centre.

Methotrexate (MTX) is the most common treatment for rheumatoid arthritis (RA). The prevalence of adverse events (AEs) associated with MTX treatment for RA have been studied extensively, but there are limited data on the predictors of these AEs. This study aims to summarise the prevalence rates of MTX AEs, including gastrointestinal (GI), neurological, mucocutaneous, and elevated alanine transaminase (ALT) enzyme, and to identify baseline demographic and clinical predictors of these AEs.

The Rheumatoid Arthritis Medication Study (RAMS) is a UK multicentre prospective cohort study of patients with RA starting MTX for the first time. Relevant demographic, medication, clinical and disease related data were collected at baseline. AEs were reported at six and twelve months follow-ups. The prevalence rates of AEs were calculated based on the proportions of patients who reported having had an AE within one year of follow-up. The associations between candidate baseline predictors and AEs were assessed using multivariable logistic regression.

A total of 2,089 patients were included with a mean age of 58.4 (standard deviation: 13.5) years, 1390 (66.5%) were women. 1,814 and 1,579 patients completed the 6 and 12 months follow-up visits, respectively. The prevalence rates of the AEs within one year of followup were: GI ¼ 777 (40.6%), mucocutaneous ¼ 441 (23.1%), neurological ¼ 487 (25.5%), elevated ALT (> upper limit of normal [ULN]) ¼ 286 (15.5%). Younger age and being a woman were associated with increased risk of GI AEs, (age: OR 0.97 per year increase in age, 95% CI 0.98, 1.00; male sex: OR 0.58 vs female, 95% CI 0.46, 0.74) ( Table  1) . Higher baseline Health Assessment Questionnaire (HAQ) score was an independent predictor of GI, mucocutaneous, and neurological AEs. Furthermore, having ALT >1xULN at baseline or history of diabetes was associated with increased risk of subsequent ALT elevation during the study follow-up.

In patients with RA starting MTX, GI AEs were the most commonly reported AEs during the first year of follow-up. The identified predictors of AEs may facilitate discussions between clinicians and patients prior to commencing MTX, and may lead to increased adherence and consequently improved effectiveness. 

Rheumatologists are facing a significant challenge in the management of early rheumatoid arthritis (RA) due to limitations placed on outpatient visits during the COVID-19 pandemic. Frequent clinical assessments of disease activity are recommended during implementation of the treat to target strategy to achieve remission. A biomarker indicating response to methotrexate during the early phase of therapy could complement clinical examination. Methotrexate increases erythrocyte mean corpuscular volume (MCV), which is measured routinely, prompting us to investigate whether changes in MCV could act as an early indicator of response. Methods Patients with early RA who were started on methotrexate therapy were included from two independent cohorts. The larger cohort (discovery cohort, n ¼ 655) was used to build the model and the second cohort (validation cohort, n ¼ 225) was applied to test the prediction of the model. Conventional statistical, and machine learning approaches were adopted to identify key determinants that influence the potential relationship between MCV and clinical response, defined as attainment of remission or low disease activity, at six months after starting methotrexate. . Different machine learning classification methods were applied. Random forest exhibited the maximum accuracy and AUROC (89%, and 86%, respectively), confirming the above three predictors as the most significant. Two latent classes (class 1: smaller MCV increase and class 2: greater MCV increase) were identified based on the MCV changes over six months. Class 1 had fewer responders and a lower number of patients on hydroxychloroquine compared to class 2. The earliest time point of significant difference of MCV between responders and non-responders was three months [mean difference 1.43 (95% CI 0.57-2.3)]. Combination hydroxychloroquine and methotrexate caused the greatest increase in MCV with a difference between responders and nonresponders at 2 months. Change of MCV at three months showed AUROC of 0.75 to predict treatment response to the combination of methotrexate and hydroxychloroquine at six months with an optimal cut-point of MCV 3.5 fL (95% CI 3.5-3.6) with 71% sensitivity and 75%, specificity. 

Early diagnosis and intervention improves outcomes of immune mediated rheumatic and musculoskeletal diseases (RMDs) but may be hampered by diagnostic uncertainty. The extent to which rationally selected molecular parameters add value to clinical characteristics for diagnostic prediction in undifferentiated disease states warrants investigation. B lymphocytes play an increasingly recognised role in rheumatoid arthritis (RA) pathogenesis, and cell-specific methylation patterns link environmental exposures to genetic risk. We derived and tested the practical utility of a B lymphocyte-derived DNA methylation signature for predicting RA in an early arthritis clinic cohort. Methods CD19þ B cell and peripheral blood mononuclear cell (PBMC) whole genome DNA methylation array data were available, respectively, from 109 inflammatory arthritis patients naïve to immunomodulatory drugs (Newcastle, UK; 38% confirmed to have a diagnosis of RA within 1 year) and 50 untreated undifferentiated arthritis (UA) patients (Leiden, The Netherlands; 68% classifiable RA within 1 year by 1987 ACR criteria versus alternate diagnoses). A bespoke machine learning pipeline employed a sequential model-based optimisation (SMBO) procedure for selecting, tuning and applying methods amongst ten feature-selection, six data-sampling and two classification algorithms in the Newcastle ''training cohort.'' The predictive performance of the resultant optimised molecular classifier was assessed in the independent Leiden ''test cohort'' alongside a previously described clinical prediction rule, using comparative area under receiver operating characteristic (AUROC) curves. A modification to the clinical prediction rule that incorporated a single parameter to reflect molecular classification was also assessed. The pipeline was implemented using the R machine learning package mlr.

Using the SMBO approach, 27 CpGs maximally discriminatory for RA were selected from B lymphocyte DNA methylome training data, and a molecular classifier was derived using the random forest algorithm. 

We provide a proof of principle for the application of a B lymphocytederived epigenetic signature to enhance prediction of RA in UA patients using stored PBMCs. Further refinement of our pipeline represents a plausible means to expedite the diagnosis in undifferentiated RMDs and could offer pathophysiological insight. 

IRs for deep vein thrombosis(DVT), pulmonary embolism(PE), and DVT and/or PE(DVT/PE) were also calculated for groups of patients while receiving BARI 2mg/4mg within All-BARI-RA. Major adverse cardiovascular events(MACE) were adjudicated in 5 Ph3 studies and the LTE. Results 3770 pts received BARI for 13,148 PY, with median and maximum exposure: 4.2 and 8.4 years, respectively. Overall IRs per 100 PY were: for any treatment-emergent adverse event (AE)(25.8); serious AE

HZ (3.0)

46); malignancies excluding non-melanoma skin cancer (NMSC)(0.91)

07); and gastrointestinal perforation (0.04). (IRs)[95% confidence intervals] for patients while receiving BARI 2mg (N ¼ 1077) and BARI 4mg (N ¼ 3400) were DVT

IRs for death tended to increase in later time intervals (beyond 192 weeks). No particular cause of death contributed to this increase