Home > Rheumatology > WPPAC 2021 > Disease Progression > Active PsA: Biomarkers distinguish radiographic progressors from non-progressors

Active PsA: Biomarkers distinguish radiographic progressors from non-progressors

WPPAC 2021
Phase 3, SPIRIT-P1

Patients with psoriatic arthritis (PsA) who progress to radiographic damage can potentially be discriminated from those who do not progress by 103 biomarker peptides, corresponding to 69 proteins. This has been identified using 2 complementary proteomic approaches and a combination of univariate and machine learning statistical analysis [1].

A delay in the diagnosis and treatment of patients with PsA leads to poor radiographic and functional outcomes [2]. So, identifying which patients might progress radiographically is essential. This unmet need has been recognised by the Group for Research and Assessment of Psoriasis and Psoriatic Arthritis (GRAPPA) [3]. Biomarkers for radiographic joint damage could assist in early-stage identification of patients likely to progress as well as identifying those who are progressing despite therapy.

The previous phase 3 SPIRIT-P1 trial (NCT01695239) demonstrated that treatment with the high-affinity IL-17A antagonist ixekizumab resulted in reduced progression of structural damage in patients with active PsA [4]. Nonetheless, 5–10% of patients who did progress may have benefitted from a more aggressive treatment if identified using biomarkers at baseline.

The aim of the current sub-analysis was to identify protein biomarkers that might distinguish at baseline those patients who progressed to joint damage from those who did not [1]. To this end, mass spectrometry-based proteomics was performed. Baseline serum samples were obtained from 83 participants of the SPIRIT-P1 trial. Radiographic progression was defined as those who showed a >0.5 change from baseline modified total Sharp score (mTSS) at week 24 or 52.

On univariate analysis, targeted proteomics identified 4 differentially expressed candidate peptides (P<0.01). With subsequent machine-learning random forest modelling, the top-15 candidate peptides of the previously used protein set were identified (ROC AUC of 0.85). The unbiased analysis found 74 peptides that were significantly differentially expressed between those who progressed and those who did not (P<0.01). Subsequent random forest modelling based on unbiased proteomics revealed a set of 15 proteins –distinct from the targeted set– that could distinguish progressors from non-progressors (ROC AUC of 0.94).

Baseline discrimination of progressors versus non-progressors was obtained in both discovery and targeted analysis. Peptides from the univariate and random forest models of the targeted and discovery analyses were combined to generate a list of 103 peptide candidate biomarkers of progression to joint damage in PsA.

Further work needs to verify whether the peptides within the 103 candidates can discriminate progression from non-progression and subsequently validate them using patient samples from a separate cohort.

  1. Coleman O. Identification of serum protein biomarkers at baseline to distinguish radiographic progressors from non-progressors in patients with active Psoriatic Arthritis (PsA). Abstract O1, 6th World Psoriasis & Psoriatic Arthritis Conference, 30 June–3 July 2021.
  2. Ocampo V, Gladman D. F1000Res. 2019;8:F1000 Faculty Rev-1665.
  3. Ritchlin CT, et al. J Rheumatol. 2010;37:462-7.
  4. Mease PJ, et al. Ann Rheum Dis. 2017;76:79-87.


Copyright ©2021 Medicom Medical Publishers

Posted on