Article Text
Abstract
Background: Psoriatic Arthritis (PsA) is a chronic, inflammatory disease that severely impacts the quality of life. The disease develops in 20-30% of people with psoriasis (PsO). Known clinical and genetic risk factors lack sensitivity and specificity in ability to predict progression from PsO to PsA. The identification of molecular biomarkers of PsA in those with PsO could support earlier diagnosis of PsA and the initiation of management and treatment strategies to prevent or ameliorate disease impact. In the IHI project HIPPOCRATES, multi-omics approaches are being used to initially identify molecular biomarkers that may predict or measure progression from PsO to PsA.
Objectives: To develop a multi-iomic molecular signature for early diagnosis of psoriatic arthritis in those with psoriasis.
Methods: Serum samples from 90 individuals participating in a prospective PsO study (BioCOM) at St. Vincent’s University Hospital (Dublin, Ireland), were analyzed by HIPPOCRATES partners with expertise in LC-MS/MS-based proteomics, targeted affinity proteomics, metabolomics, lipidomics, and genomics. Of the 90 participants, 30 had established PsA, 30 had PsO with no clinical features to suggest musculoskeletal disease (MSK), and 30 had PsO and clinical symptoms suggestive of early MSK involvement.
Results: Using LC-MS/MS, targeted proteomics (Multiple Reaction Monitoring) assays, and random forest algorithms, a panel of relevant peptides was identified, which distinguished PsO from PsA with an ROC AUC = 0.8. Targeted proteomics using Olink PEA assays, revealed quantitative differences in LASSO-selected inflammatory proteins between PsA and PsO groups. Furthermore, lipidomics and metabolomics data for the samples were obtained using a combined approach for targeted and non-targeted analysis and indicated differences between the two groups for different lipid species. Once uploaded to the Secure HIPPOCRATES Data Management Platform the data will be analysed using ML and AI methods.
Conclusion: We have identified candidate molecular markers that may differentiate PsO from PsA. In ongoing studies, the multi-omics data is being combined and subjected to integrated Machine Learning analysis. These initial findings will require evaluation in additional PsO and PsA cohorts - ideally longitudinal cohorts.
REFERENCES: [1] FitzGerald O et al. Ther Adv Musculoskelet Dis. 2023;15:1759720X231192315.
[2] Gurke R, et al. Biomedicines. 2022; 10(10):2387.
[3] FitzGerald O, Pennington SR. Nat Rev Rheumatol. 2022;18:123-124.
[4] Psoriatic arthritis.
[5] FitzGerald O, et al. Nat Rev Dis Primers. 2021;7(1):59.
[6] Jadon DR, Stober C, Pennington SR, FitzGerald O. Nat Rev Rheumatol. 2020 Nov;16(11):609-627.
Acknowledgements: This project has received funding from the Innovative Medicines Initiative 2 Joint Undertaking (JU) under grant agreement No 101007757 (HIPPOCRATES). The JU receives support from the European Union’s Horizon 2020 research and innovation programme and EFPIA.
Disclosure of Interests: None declared.
- Prognostic factors
- Diagnostic test
- Biomarkers
- ‘-omics
- Best practices