Background Genetic factors have a substantial role in determining development of rheumatoid arthritis (RA), and are likely to account for 50–60% of disease susceptibility. Genome-wide association studies have identified non-human leucocyte antigen RA susceptibility loci which associate with RA with low-to-moderate risk. Objectives To investigate recently identified RA susceptibility markers using cohorts from six European countries, and perform a meta-analysis including previously published results. Methods 3311 DNA samples were collected from patients from six countries (UK, Germany, France, Greece, Sweden and Denmark). Genotype data or DNA samples for 3709 controls were collected from four countries (not Sweden or Denmark). Eighteen single nucleotide polymorphisms (SNPs) were genotyped using Sequenom MassArray technology. Samples with a >95% success rate and only those SNPs with a genotype success rate of >95% were included in the analysis. Scandinavian patient data were pooled and previously published Swedish control data were accessed as a comparison group. Meta-analysis was used to combine results from this study with all previously published data. Results After quality control, 3209 patients and 3692 controls were included in the study. Eight markers (ie, rs1160542 (AFF3), rs1678542 (KIF5A), rs2476601 (PTPN22), rs3087243 (CTLA4), rs4810485 (CD40), rs5029937 (6q23), rs10760130 (TRAF1/C5) and rs7574865 (STAT4)) were significantly associated with RA by meta-analysis. All 18 markers were associated with RA when previously published studies were incorporated in the analysis. Data from this study increased the significance for association with RA and nine markers. Conclusions In a large European RA cohort further evidence for the association of 18 markers with RA development has been obtained.

Objectives To evaluate (i) the association between random certolizumab drug levels, antidrug antibodies (ADAbs) and treatment response in patients with rheumatoid arthritis (RA); (ii) longitudinal factors associated with ADAbs and certolizumab drug levels. Methods This prospective cohort included 115 patients with RA treated with certolizumab. Serum samples were collected at 3, 6 and 12 months following treatment initiation. Drug levels and ADAbs were measured using ELISA and radioimmunoassay, respectively, at 3, 6 and 12 months. Disease Activity Score in 28 joints (DAS28) were measured at each visit and 12 months European League Against Rheumatism (EULAR) response was calculated. Patient self-reported adherence was collected longitudinally. Ordinal logistic regression and generalised estimating equation were used to test the association: (i) between drug levels, from serum sampled and treatment response; (ii) between ADAbs and drug levels; (iii) patient-centred factors and drug levels. Results ADAbs were detected in 37% (42/112 patients by 12 months). The presence of ADAbs were significantly associated with lower drug levels over 12 months (β=−0.037, 95% CI −0.055 to 0.018, p<0.0001) but not independently with 12 months EULAR response (β=0.0013 (95% CI −0.0032 to 0.00061), p=0.18). Drug level was associated with 12 months EULAR response (β=0.032 (95% CI 0.0011 to 0.063), p=0.042). In the multivariate model, ADAb level and adherence were significantly associated with drug concentrations. Conclusions This is the first study to demonstrate that higher certolizumab drug levels are associated with better 12 months EULAR response. ADAbs in certolizumab-treated patients with RA were detected at higher levels than previous studies and help determine the aetiology of a low drug level.

Methotrexate (MTX) is a common first-line treatment for new-onset rheumatoid arthritis (RA). However, MTX is ineffective for 30–40% of patients and there is no way to know which patients might benefit. Here, we built statistical models based on serum lipid levels measured at two time-points (pre-treatment and following 4 weeks on-drug) to investigate if MTX response (by 6 months) could be predicted. Patients about to commence MTX treatment for the first time were selected from the Rheumatoid Arthritis Medication Study (RAMS). Patients were categorised as good or non-responders following 6 months on-drug using EULAR response criteria. Serum lipids were measured using ultra‐performance liquid chromatography–mass spectrometry and supervised machine learning methods (including regularized regression, support vector machine and random forest) were used to predict EULAR response. Models including lipid levels were compared to models including clinical covariates alone. The best performing classifier including lipid levels (assessed at 4 weeks) was constructed using regularized regression (ROC AUC 0.61 ± 0.02). However, the clinical covariate based model outperformed the classifier including lipid levels when either pre- or on-treatment time-points were investigated (ROC AUC 0.68 ± 0.02). Pre- or early-treatment serum lipid profiles are unlikely to inform classification of MTX response by 6 months with performance adequate for use in RA clinical management.

Data-independent acquisition mass spectrometry allows for comprehensive peptide detection and relative quantification than standard data-dependent approaches. While less prone to missing values, these still exist. Current approaches for handling the so-called missingness have challenges. We hypothesized that non-random missingness is a useful biological measure and demonstrate the importance of analysing missingness for proteomic discovery within a longitudinal study of disease activity. The magnitude of missingness did not correlate with mean peptide concentration. The magnitude of missingness for each protein strongly correlated between collection time points (baseline, 3 months, 6 months; R = 0.95–0.97, confidence interval = 0.94–0.97) indicating little time-dependent effect. This allowed for the identification of proteins with outlier levels of missingness that differentiate between the patient groups characterized by different patterns of disease activity. The association of these proteins with disease activity was confirmed by machine learning techniques. Our novel approach complements analyses on complete observations and other missing value strategies in biomarker prediction of disease activity. Supplementary data are available at Bioinformatics online.

Adding biomarker information to real world datasets (e.g. biomarker data collected into disease/drug registries) can enhance mechanistic understanding of intra-patient differences in disease trajectories and differences in important clinical outcomes. Biomarkers can detect pathologies present early in disease potentially paving the way for preventative intervention strategies, which may help patients to avoid disability, poor treatment outcome, disease sequelae and premature mortality. However, adding biomarker data to real world datasets comes with a number of important challenges including sample collection and storage, study design and data analysis and interpretation. In this narrative review we will consider the benefits and challenges of adding biomarker data to real world datasets and discuss how biomarker data have added to our understanding of complex diseases, focusing on rheumatoid arthritis.