Bart V. J. Cuppen

Personalized biological treatment for rheumatoid arthritis: a systematic review with a focus on clinical applicability

OBJECTIVES To review studies that address prediction of response to biologic treatment in RA and to explore the clinical utility of the studied (bio)markers. METHODS A search for relevant articles was performed in PubMed, Embase and Cochrane databases. Studies that presented predictive values or in which these could be calculated were selected. The added value was determined by the added value on prior probability for each (bio)marker. Only an increase/decrease in chance of response ⩾15% was considered clinically relevant, whereas in oncology values >25% are common. RESULTS Of the 57 eligible studies, 14 (bio)markers were studied in more than one cohort and an overview of the added predictive value of each marker is presented. Of the replicated predictors, none consistently showed an increase/decrease in probability of response ⩾15%. However, positivity of RF and ACPA in case of rituximab and the presence of the TNF-α promoter 308 GG genotype for TNF inhibitor therapy were consistently predictive, yet low in added predictive value. Besides these, 65 (bio)markers studied once showed remarkably high (but not validated) predictive values. CONCLUSION We were unable to address clinically useful baseline (bio)markers for use in individually tailored treatment. Some predictors are consistently predictive, yet low in added predictive value, while several others are promising but await replication. The challenge now is to design studies to validate all explored and promising findings individually and in combination to make these (bio)markers relevant to clinical practice.

Metabolomics profiling of the free and total oxidised lipids in urine by LC-MS/MS: application in patients with rheumatoid arthritis

Oxidised lipids, covering enzymatic and auto-oxidation-synthesised mediators, are important signalling metabolites in inflammation while also providing a readout for oxidative stress, both of which are prominent physiological processes in a plethora of diseases. Excretion of these metabolites via urine is enhanced through the phase-II conjugation with glucuronic acid, resulting in increased hydrophilicity of these lipid mediators. Here, we developed a bovine liver-β-glucuronidase hydrolysing sample preparation method, using liquid chromatography coupled to tandem mass spectrometry to analyse the total urinary oxidised lipid profile including the prostaglandins, isoprostanes, dihydroxy-fatty acids, hydroxy-fatty acids and the nitro-fatty acids. Our method detected more than 70 oxidised lipids biosynthesised from two non-enzymatic and three enzymatic pathways in urine samples. The total oxidised lipid profiling method was developed and validated for human urine and was demonstrated for urine samples from patients with rheumatoid arthritis. Pro-inflammatory mediators PGF2α and PGF3α and oxidative stress markers iPF2α- IV, 11-HETE and 14-HDoHE were positively associated with improvement of disease activity score. Furthermore, the anti-inflammatory nitro-fatty acids were negatively associated with baseline disease activity. In conclusion, the developed methodology expands the current metabolic profiling of oxidised lipids in urine, and its application will enhance our understanding of the role these bioactive metabolites play in health and disease.

Exploring the Inflammatory Metabolomic Profile to Predict Response to TNF-α Inhibitors in Rheumatoid Arthritis

In clinical practice, approximately one-third of patients with rheumatoid arthritis (RA) respond insufficiently to TNF-α inhibitors (TNFis). The aim of the study was to explore the use of a metabolomics to identify predictors for the outcome of TNFi therapy, and study the metabolomic fingerprint in active RA irrespective of patients’ response. In the metabolomic profiling, lipids, oxylipins, and amines were measured in serum samples of RA patients from the observational BiOCURA cohort, before start of biological treatment. Multivariable logistic regression models were established to identify predictors for good- and non-response in patients receiving TNFi (n = 124). The added value of metabolites over prediction using clinical parameters only was determined by comparing the area under receiver operating characteristic curve (AUC-ROC), sensitivity, specificity, positive- and negative predictive value and by the net reclassification index (NRI). The models were further validated by 10-fold cross validation and tested on the complete TNFi treatment cohort including moderate responders. Additionally, metabolites were identified that cross-sectionally associated with the RA disease activity score based on a 28-joint count (DAS28), erythrocyte sedimentation rate (ESR) or C-reactive protein (CRP). Out of 139 metabolites, the best-performing predictors were sn1-LPC(18:3-ω3/ω6), sn1-LPC(15:0), ethanolamine, and lysine. The model that combined the selected metabolites with clinical parameters showed a significant larger AUC-ROC than that of the model containing only clinical parameters (p = 0.01). The combined model was able to discriminate good- and non-responders with good accuracy and to reclassify non-responders with an improvement of 30% (total NRI = 0.23) and showed a prediction error of 0.27. For the complete TNFi cohort, the NRI was 0.22. In addition, 88 metabolites were associated with DAS28, ESR or CRP (p<0.05). Our study established an accurate prediction model for response to TNFi therapy, containing metabolites and clinical parameters. Associations between metabolites and disease activity may help elucidate additional pathologic mechanisms behind RA.

Can baseline serum microRNAs predict response to TNF-alpha inhibitors in rheumatoid arthritis?

BackgroundIn rheumatoid arthritis, prediction of response to TNF-alpha inhibitor (TNFi) treatment would be of clinical value. This study aims to discover miRNAs that predict response and aims to replicate results of two previous studies addressing this topic.MethodsFrom the observational BiOCURA cohort, 40 adalimumab- (ADA) and 40 etanercept- (ETN) treated patients were selected to enter the discovery cohort and baseline serum profiling on 758 miRNAs was performed. The added value of univariately selected miRNAs (p < 0.05) over clinical parameters in prediction of response was determined by means of the area under the receiver operating characteristic curve (AUC-ROC). Validation was performed by TaqMan single qPCR assays in 40 new patients.ResultsExpression of miR-99a and miR-143 predicted response to ADA, and miR-23a and miR-197 predicted response to ETN. The addition of miRNAs increased the AUC-ROC of a model containing only clinical parameters for ADA (0.75 to 0.97) and ETN (0.68 to 0.78). In validation, none of the selected miRNAs significantly predicted response. miR-23a was the only overlapping miRNA compared to the two previous studies, however inversely related with response in one of these studies. The reasons for the inability to replicate previously proposed miRNAs predicting response to TNFi and replicate those from the discovery cohort were investigated and discussed.ConclusionsTo date, no miRNA consistently predicting response to TNFi therapy in RA has been identified. Future studies on this topic should meet a minimum of standards in design that are addressed in this study, in order to increase the reproducibility.

Contribution of the Subjective Components of the Disease Activity Score to the Response to Biologic Treatment in Rheumatoid Arthritis

A significant proportion of patients with rheumatoid arthritis do not respond adequately to biologic treatment. We hypothesized that lack of response to (biologic) disease‐modifying antirheumatic drugs (DMARDs) is high in patients in whom the subjective, patient‐reported component of the Disease Activity Score 28 (DAS28) is high at baseline. The primary aim of our present study was to investigate the contribution of the more subjective versus the objective components of the DAS28 to response to biologic agents in RA patients, as well as the changes in this contribution over time. The secondary aim was to examine whether the value of this subjective contribution at baseline affects the response to treatment.

THU0081 Towards Individualized Risk Determination in RA: A Prediction Model for TNFI Discontinuation within the First Year After Start

Background TNF-alpha inhibitor (TNFi) treatment has dramatically improved the outcome of RA patients. A substantial number of patients, however, fails to clinically respond to this therapy or experiences adverse-effects that necessitate discontinuation of therapy. In that sense, treatment success could be envisioned as a matter of balance between drug efficacy and tolerability. Prediction of discontinuation gives insight in patients at risk for suboptimal treatment success, and possibly factors that interacts the decision making of physician and patient. Objectives To predict TNFi discontinuation within the first year of use in an observational cohort and gain more insight in parameters predictive of treatment success. Methods Data was used from the Biologicals and Outcome Compared and predicted Utrecht region in Rheumatoid Arthritis (BiOCURA) cohort. Eight hospitals of the Society for Rheumatology Research Utrecht included patients eligible for biological treatment, which were followed up to one year after start of this treatment. A univariate preselection (p<0.2) was performed to enter variables in the multivariable cox-regression model with backward selection (p<0.1). To develop a quick tool for clinical practice, the linear predictor was simplified by adjusting the coefficients to usable scores. Results After one year of follow-up of 192 TNFi patients, 75 (39%) discontinued treatment, because of inefficacy (64%) or side-effects (33%). Discontinuation was predicted by a combination of female gender (HR=2.1, p=0.02), currently smoking (HR=1.8, p=0.03), RF positivity (HR=0.67, p=0.10), high VAS-GH score (HR=1.02/mm, p<0.01) and higher number of previous biological DMARDs (HR=1.5/biological, p<0.01). A simplified score for use in clinical practice was made (see table). For each of the five variables a patient scores points. The sum of these scores can determine if there is an absolute chance of discontinuation within the first year of 67% (score >7.00) or 83% (score >7.75). Analyses of the high versus low risk patients revealed no different reasons for discontinuation (p=0.27 and 0.90 for different cut-offs respectively). Interestingly, the course over time of DAS28 and DAS28-inflammation and pain component in high risk (score >7.00) versus low risk patients, revealed differences that put discontinuing in a more subjective context: Although absolute DAS28 scores were higher over time for the high risk patients (0.27-0.75), these differences could be explained by a higher relative pain component of the DAS28, which was increased up to 1.3 fold three months after start (p<0.01).Table 1 Points Female gender, yes 3.0 Currently smoking, yes 2.0 RF negativity, yes 1.5 No. of prev. biologicals, per biological 1.5 VAS-GH, per 10mm 0.2 Total … Conclusions TNFi discontinuation within the first year of use in an observational cohort can be predicted by a simple prediction score. The reported pain by patients is probably an underestimated factor in the clinical decision of discontinuation. To investigate if these findings are reproducible, validation will be performed at short notice when 100 subsequent TNFi treated patients are included in the BiOCURA cohort (currently n=75). Disclosure of Interest None declared

OP0130 A Proteomics Approach to Predict the TNF-Alpha Inhibitor Response in RA: The Added Clinical Value of a Protein Score:

Background In rheumatoid arthritis (RA) it is of major importance to distinguish non-responders to TNF-alpha inhibitor (TNFi) treatment before start to prevent a delay in effective treatment, potential side-effects and unnecessary healthcare costs. We investigated the ability of al large set of inflammatory proteins to predict (absence of) response to biological treatment. Objectives To develop a protein score predictive for response to TNFi treatment in RA and investigate its added predictive value over clinical parameters alone. Methods In consecutive RA patients eligible for TNFi treatment as included in the BiOCURA registry, serum was collected before start of treatment and analyzed on 57 inflammatory proteins using xMAP technology. EULAR response was determined after three months. A supervised cluster analysis method, partial least squares (PLS) was used to select the best combination of proteins and cross-validation to gain a reproducible protein score. Multiple imputation was used to account for missing data of baseline clinical parameters. Relevant clinical parameters for EULAR good response were selected by performing a univariate (p<0.2) and multivariable backward selection (p<0.1). The predictive ability of the final model with and without the protein score was assessed using the area under the receiving operater curve (AUC-ROC), negative predictive values (NPV) and the reclassification index (NRI). Results Response was determined for 171 of the 192 cases starting treatment. On top of CRP and DAS28 at baseline, PLS revealed 9 important proteins: sCD14, IFNγ, MCP1, MIP1b, MIP3b, TARC, sTNFRI, sTNFRII and TSLP. These markers were able to explain 31.5% of the variance in DAS28 at 3 months. Final models for prediction of TNFi response included baseline DAS28, naivety for bDMARDs, HAQ, RF positivity, concomitant MTX and GC use. The protein score did not improve the AUC-ROC of 0.80 (0.73-0.87), however, when the predefined cut-off for a NPV≥0.9 was set, the addition of the protein score resulted in the classification of 30 extra patients in the low probability category (table). An improved classification was observed of 24.2% and -4.8% for patients with and without a response respectively (NRI=19.42%). Model Category for probability of response Size (n) Probability of EULAR good response (% in cat.) Clinical variables alone Low (0–10%) 58 10.4 Normal (10–100%) 113 42.3 Clinical variables + protein score Low (0–10%) 88 9.8 Normal (10–100%) 83 54.5 Conclusions We showed that a combination of a protein score and clinical variables is able to predict absence of EULAR good response to TNF inhibiting treatment and can classify more patients at baseline in the appropriate risk category than clinical variable alone. This protein score may therefore contribute to a more patients tailored treatment, leading to a better usage of the available resources. In the near future these findings will be validated externally. Acknowledgements The Society for Rheumatology Research Utrecht (SRU) consists of University Medical Center Utrecht, Antonius Hospital Nieuwegein and Utrecht, Diakonessen Hospital Utrecht, Meander Medical Center Amersfoort, Sint Maartenskliniek Woerden, Hospital St. Jansdal Harderwijk, Tergooi Hospital Hilversum, Flevo Hospital Almere. Disclosure of Interest None declared

AB0239 Personalized Biological Treatment for Rheumatoid Arthritis: A Systematic Review with a Focus on Clinical Applicability

Background Biological treatments have dramatically improved the outcome of RA patients. A substantial number of patients, however, fails to clinically respond to these therapies. Prediction of therapeutic (non) response before start of treatment could aid in clinical decision making of a personalized treatment approach. Numerous studies have previously addressed this topic. Objectives To review studies that address prediction of response to biological treatment in rheumatoid arthritis (RA) and to explore the added value in clinical applicability of these (bio)markers. Methods A search for relevant articles was performed in PUBMED, EMBASE and COCHRANE databases. Studies which presented predictive values or in which these could be calculated were selected. The added value was determined upon sensitivities, specificities and the added value on prior probability for each (bio)marker univariately. An increase/decrease in chance of response of ≥15% was considered clinically relevant enough, whereas in oncology values >25% are common. Results Out of the 52 eligible studies, 14 (bio)markers were studied multiple times and were compared for the additive predictive effect of each (bio)marker. Of the replicated predictors, none consistently showed an increase/decrease in chance of response of ≥15%. The TNF-alpha 308 polymorphism modestly predicted response to TNF-alpha inhibitors (decrease of 3.7-30.1%), and rheumatoid factor (RF) and anti-citrullinated antibodies (ACPA) predicted response to rituximab (increase of 1.9-8.9% and 1.1-7.5% resp.). Besides these, 71 (bio)markers studied once, showed remarkably high (but not validated) predictive values. Conclusions We were not able to indicate truly clinically useful baseline (bio)markers for individually tailored treatment. Few predictors are consistently predictive yet low in added predictive value, while several others are promising but await replication. The challenge now is to design studies to validate all explored and promising findings individually and in combination, to make these (bio)markers relevant to clinical practice. Disclosure of Interest None declared

FRI0024 Contribution of the Individual Components of the Disease Activity Score (DAS28) to the Total DAS28 Score among Responders and Non-Responders to Biological Therapy for Rheumatoid Arthritis

Background The commonly used Disease Activity Score based on 28 joints (DAS28) is a composite index composed of two somewhat subjective components (tender joint count and visual analogue scale general well-being) and two more objective components (swollen joint count and erythrocyte sedimentation rate). However, not all individual components of DAS28 might respond similarly to treatment. Furthermore dominance of subjective over objective components in DAS28 at start of therapy might predict less effect of therapy. This subjective contribution relative to the total DAS28 was quantified through the “contribution score” (the subjective part of the DAS28 formula divided by the total DAS28 formula). Objectives First, to investigate if the contribution score of the study population changes over time between starting a first biological (baseline) and after three months of biological use, and to establish if this change is seen and similar in the non-, moderate- and good response to treatment groups. Second, to investigate if the subjective contribution at baseline is predictive of response to treatment at three months. Methods Patients included in this study were selected from two databases of the Utrecht Arthritis Cohort study group. In the CAMERA-II trial early RA patients had been included, comparing the addition of 10mg/day of prednisone or prednisone-placebo to a two-year MTX-based tight –controlled strategy, including a final step of adding a biological. In the more recent Utrecht observational Biological cohort study, any patient with RA who started a biological could be entered. A total of 51 of the CAMERA-II trial database patients and 121 of the Biological cohort study database, all starting a biological treatment, were included. To address the first objective, ANOVAs and paired t tests were performed. To address the second objective, an ordinal logistic regression analysis was performed (correcting for age, DAS28 at the start of the biological therapy, number of prior DMARDs used, cohort, gender, RF-status and smoking), with as criterion variable EULAR response (3 levels: non-, moderate- and good response) and as predictor variable the contribution score. Results Overall, a significant decrease in contribution score was observed (p<0.001), showing -in contrast to our hypothesis- a therapy effect more pronounced in the subjective parts of the DAS28 compared to the therapy effect in the objective components. When looking into the separate response groups, this significant change was observed only in the good responders (p<0.001). The contribution score at baseline was not predictive of the different levels of response to treatment at three months (contribution score at baseline x100: proportional OR=1.001, p=0.8). Conclusions The treatment effect of this first administered biological is largest in the subjective components of DAS28, yet these subjective components of DAS28 at start of therapy do not predict treatment response levels at three months. Disclosure of Interest None declared DOI 10.1136/annrheumdis-2014-eular.4125