1683: Serum APRIL Is Associated With B-Cell Activation Markers, Disease Activity, and Lymphoma Risk in Sjögren Disease (SD): Data From the prospective ASSESS Cohort

Background/Purpose: Sjögren disease (SD) is a B cell driven systemic auto-immune disease, with a need for biomarkers to better assess disease activity, risk of lymphoma and finding new therapeutic targets. While BAFF, a key B cell survival cytokine involved in precoce steps of B cell ontogeny, has been previously associated with systemic manifestations and lymphomagenesis in SD, data on a proliferation-inducing ligand (APRIL), another B cell cytokine acting in later stages of B cell ontogeny (e.g., plasmablasts and plasma cells via BCMA and TACI receptors), remain limited. This study aimed to investigate associations between APRIL, B cell biomarkers, disease activity and risk of lymphoma in SD.

Methods: Serum APRIL levels (ELISA, Invitrogen°) were assessed in 337 patients with SD from the French ASSESS cohort (17-year follow-up) and 40 healthy donors (HD). All patients met 2016 ACR/EULAR classification criteria. Patients were stratified according to the validated clustering of the ASSESS cohort: (C1) biologically active without systemic manifestations; (C2) systemic manifestations; (C3) low activity/high symptom burden; and SD-related lymphoma (incident n=9, prevalent n=16). The association between SD-related lymphoma, APRIL and lymphoma predictors at enrollment (clinESSDAI, ESSDAI, lymphocytopenia < 0.5 G/L, CD4/CD8 T cell ratio≤0.8, RF positivity, cryoglobulinemia, monoclonal component, and low C4) and other potential biomarkers (IgG levels, total gammaglobulins, immune-complexes, beta2microglobulin, kappa and lambda free light chains, quantitative [q] anti- SSA/SSB autoantibodies, digital interferon [IFN] alpha, s-FLT3, CXCL10, CCL19, CXCL13, BAFF), were analyzed in univariate and multivariate models. Multivariate analyses took into account APRIL and variables significantly associated with lymphoma in univariate analyses selected by a stepwise procedure. Estimation of incident lymphoma risk was assessed in a separate multivariate model.

Results: APRIL levels increased across subgroups: HD (0.36 ng/mL), cluster 3 (0.34), cluster 1 (0.97), cluster 2 (1.2), lymphoma (2.33), with highest levels in incident lymphoma (2.8; p< 0.0001). In multivariate analysis, including only markers significantly associated with lymphoma in univariate analyses, APRIL expression at enrollment remained significantly associated with lymphoma (OR 1.1; 95% CI [1- 1.17] p=0.045), along with BAFF levels (mean [SD]: 1085.1 pg/ml [742.4]; per 1000 BAFF units, OR 1.75; 95% CI [1.11-2.71]; p=0.0012), CD4/CD8 ratio ≤ 0.8 (OR 4.7 95% CI [1.13-18.03]; p=0.027), low C4 (OR 5.38 95% CI [1.89-16.04]; p=0.002), lymphopenia < 0.5 G/L (OR 51.02 95% CI [4.65-611.01]; p=0.001), glandular (OR 4.95 95% CI [1.49-15.97]; p=0.007), and adenopathy (OR 6.43 95% CI [1.1-31.8]; p=0.027) domains of the ESSDAI. In a separate multivariate model, APRIL (OR 1.1 95% CI [1.01-1.21]; p=0.02) and CXCL13 (OR 1.0 95% CI [1.0-1.0]; p=0.01) were found independently associated with the risk of developing incident lymphoma.

Conclusion: These results suggest the interest of APRIL, a B-cell cytokine easily measurable in patients’ sera, as a potential biomarker of SD-related lymphoma, and as a potentially relevant therapeutic target in SD.