Analysis of Serum Proteomics Data Identifies Beta-Defensin 2 as a Predictive Biomarker of Clinical Response to IL-17 Blockade in Psoriatic Arthritis

Mathias Cardner¹, Danny Tuckwell², Anna Kostikova², Pascal Forrer³, Richard Siegel², Alain Marti³, Marc Vandemeulebroecke³ and Enrico Ferrero², ¹Novartis Global Drug Development and Novartis Institutes for BioMedical Research, Basel, Switzerland, ²Novartis Institutes for BioMedical Research, Basel, Switzerland, ³Novartis Global Drug Development, Basel, Switzerland

Meeting: ACR Convergence 2022

Keywords: Biomarkers, Interleukins, Psoriatic arthritis, TH17 Cells

Session Information

Date: Monday, November 14, 2022

Title: Spondyloarthritis Including PsA – Treatment Poster III: PsA

Session Type: Poster Session D

Session Time: 1:00PM-3:00PM

Background/Purpose: Despite several effective targeted therapies, biomarkers that predict whether a patient with psoriatic arthritis (PsA) will respond to a particular treatment are currently lacking.

Methods: We analysed proteomics data from serum samples of nearly two thousand PsA patients in placebo-controlled phase-III clinical trials of the IL-17A blocking mAb secukinumab, using the SomaScan platform. To discover predictive biomarkers of clinical response, we used statistical learning with controlled feature selection. The top candidate was validated using an ELISA and was separately assessed in a trial of almost eight hundred PsA patients treated with the anti-IL-17A mAb secukinumab or a tumour necrosis factor inhibitor (TNFi).

Results: A statistical learning model trained on the FUTURE 1, 3, 4, and 5 studies of secukinumab in PsA (n=1626) identified serum levels of the anti-microbial peptide beta-defensin 2 (BD-2) at baseline as the only serum protein predictive of ACR20 response to secukinumab but not placebo. SERPINA10 was prognostic of ACR20 but not predictive of response to sekukinumab. Baseline BD-2, but not SERPINA10, was validated in an independent test data set from the FUTURE 2 trial (Figure 1B) adding 11 percentage points to the predictive performance (ROC AUC) of a logistic regression model based on clinical predictors (Fig 1C). Although BD-2 is known to be associated with psoriasis severity, the predictivity of BD-2 was independent of baseline PASI. The association between BD-2 and ACR20 response to secukinumab was observed as early as 4 weeks and maintained up to 52 weeks (Fig 1D). Validation and quantification of BD-2 by ELISA allowed modelling a classification tree showing that baseline BD-2 is the most important factor in predicting response, and that the subset of patients with higher levels of BD-2 at baseline are highly enriched for responders to secukinumab (76%, Fig 2A). BD-2 prediction of response was enhanced in TNF-IR patients (Fig 2B), and also significantly predictive of ACR50 and 70 responses (Table 1). Analysis of samples from the EXCEED trial which compared adalimumab to sekukinumab treatment in PsA showed that BD-2 was also predictive of response to treatment with a TNFi, though this was not sustained longitudinally as it was for secukinumab. Analysis across three treatment trials showed that BD-2 levels were not significantly predictive of response to secukinumab in RA.

Conclusion: BD-2 is a novel, PsA-specific predictive biomarker of clinical arthritis response to secukinumab. Patients with high levels of BD-2 at baseline reach and sustain higher rates of clinical response after treatment with secukinumab.

Figure 1. Statistical learning based on SomaScan data identified BD_2 as a predictive biomarker of ACR20 response to secukinumab after 16 weeks of treatment. The discovery was made in the training set (panel A) and confirmed in the test set (panels B and C). A) The selection probability reflects how often a given variable was included in the underlying elastic-net regression model as a function of the regularisation step (where higher means stricter penalisation). Only variables passing stability selection are labelled. Asterisks denote secukinumab-specific variables, and the only predictive candidate was BD_2 (encoded by the gene DEFB4A). B) Box and violin plots of baseline BD_2 levels in serum in the test set, as measured by the corresponding SOMAmer. The Wilcoxon rank-sum test was used to assess differences in BD_2 distributions between ACR20 responders and non-responders, with the resulting p-values above the indicated comparison. C) Logistic regression models trained to predict ACR20 response after 16 weeks of secukinumab treatment were evaluated by ROC curves in the test set. The AUC of each model is reported under the diagonal. The clinical model incorporated sekukinumab dose level, prior exposure and response to TNF inhibitors, subject weight, and concomitant use of methotrexate D) Longitudinal response in patients with above/below median BD_2 levels at baseline. Mean rates are banded by 95% confidence intervals. Placebo arms are shown until the 16th week visit since non-responding patients were switched to secukinumab at that point.
Abbreviations: relative fluorescence unit (RFU), receiver operating characteristic (ROC), area under the curve (AUC).

Figure 2. ELISA validation of BD_2 in FUTURE 1, 2, 3, and 5, with identification of a BD_2 cut-off for ACR20 response prediction. A) A classification tree shows the optimal splits for response prediction. The colored circle nodes indicate the response predicted by the tree. Below each node the response rate in the indicated subgroup is shown in bold next to a rectangle with a corresponding height and colour. The proportion of the cohort falling into each subgroup determines the width of the corresponding rectangle and is reported below in italics. B) Response rates with 95% confidence intervals in patients with baseline BD_2 < vs ≥ 3961 pg/mL, shown per study and TNFi stratum: naïve or inadequate responder (IR).

Table 1. Odds ratios of ACR20, ACR50, and ACR70 response in patients with baseline BD_2 ≥ 3961 pg/mL, as compared with BD_2 < 3961 pg/mL. Reported in parenthesis are 95% confidence intervals along with asterisks indicating p-values below 0.05 (*), 0.01 (**) and 0.001 (***). Statistics are based on 1648 patients across FUTURE 1, 2, 3, and 5.

Disclosures: M. Cardner, Novartis; D. Tuckwell, Novartis; A. Kostikova, Novartis; P. Forrer, Novartis; R. Siegel, Novartis; A. Marti, Novartis; M. Vandemeulebroecke, Novartis; E. Ferrero, Novartis.

To cite this abstract in AMA style:

Cardner M, Tuckwell D, Kostikova A, Forrer P, Siegel R, Marti A, Vandemeulebroecke M, Ferrero E. Analysis of Serum Proteomics Data Identifies Beta-Defensin 2 as a Predictive Biomarker of Clinical Response to IL-17 Blockade in Psoriatic Arthritis [abstract]. Arthritis Rheumatol. 2022; 74 (suppl 9). https://acrabstracts.org/abstract/analysis-of-serum-proteomics-data-identifies-beta-defensin-2-as-a-predictive-biomarker-of-clinical-response-to-il-17-blockade-in-psoriatic-arthritis/. Accessed .

« Back to ACR Convergence 2022

ACR Meeting Abstracts - https://acrabstracts.org/abstract/analysis-of-serum-proteomics-data-identifies-beta-defensin-2-as-a-predictive-biomarker-of-clinical-response-to-il-17-blockade-in-psoriatic-arthritis/