Background/Purpose: PRGN is a growth factor that binds TNF receptors without triggering TNF-like activity, thereby serving as a naturally-occurring antagonist of TNF -mediated inflammatory signaling. In multiple mouse arthritis models, PRGN inhibits TNF-activated intracellular signaling and development of joint inflammation. We hypothesized that endogenous PRGN levels may affect responsiveness to TNF-antagonist therapy in RA patients. Accordingly, we conducted a pilot observational study in RA patients to assess whether serum PRGN levels would either predict and/or reflect responsiveness to TNF-antagonist therapy. TNF may play a greater role in driving disease in RA patients who bear low PRGN levels (PRGN-low) than in RA patients who bear high PRGN levels (PRGN-high).  If correct, then PRGN-low would be more responsive to TNF-antagonist therapy than PRGN-high. 

Methods: We enrolled 40 Hispanic RA patients who had never received TNF-antagonist therapy and who, in the judgment of their respective physicians, were candidates for TNF-antagonist therapy. Prior to initiation of TNF-antagonist therapy and at regularly scheduled visits for 12 months following initiation of such therapy, disease activity was assessed by DAS28-CRP, and blood samples were collected for serum PRGN levels (measured by ELISA). Assignment of level of disease activity (high, moderate, low) and degree of clinical response (no, moderate, good, remission) was based on EULAR criteria.

Results: Of the 40 RA patients enrolled. 34 were female and 6 were male: median (range) age, 53 (22-75) years;  median (range) years of symptoms, 4 (1-29); median (range) tender joints, 16 (0-28); median (range) swollen joints , 7 (0-25); median (range) ESR, 43 (7-105); median (range) CRP, 12 (1-195); median (range) DAS28-CRP, 5.59 (3.28-8.04). 31 patients had high disease activity, and 9 had moderate disease activity. 20 patients had radiographic erosive disease; 23 were taking methotrexate; 3 were taking corticosteroids; 29 were taking hydroxychloroquine; 2 were taking leflunamide; and 19 were taking sulfasalazine. Based on the distribution of baseline serum PRGN levels, 9 RA patients were classified as PRGN-high, and 31 patients were classified as PRGN-low. No statistical differences were noted in baseline parameters between PRGN-high and PRGN-low.

Of the 40 patients, 35 initiated TNF-antagonist therapy (8 PRGN-high, 27 PRGN-low). At the last follow-up visit, 19 achieved clinical remission, 4 achieved a good response, 8 achieved a moderate response, and 4 had no response. As determined by log-rank analysis, there were no differences in achievement of moderate/good responses, good responses, or clinical remission between PRGN-high and PRGN-low. Post-initiation visits were divided into 4 intervals (1-91 days; 92-182 days; 183-273 days: ≥274 days); there were no differences in DAS28-CRP between PRGN-high and PRGN-low at any interval. Serum PRGN levels declined following initiation of TNF-antagonist therapy, but changes in serum PRGN levels did not correlate with changes in DAS28-CRP.

Conclusion: Serum PRGN levels neither predict nor reflect clinical responsiveness to TNF-antagonist therapy in RA patients.

« Back to 2015 ACR/ARHP Annual Meeting

ACR Meeting Abstracts - https://acrabstracts.org/abstract/serum-progranulin-prgn-level-is-not-a-biomarker-for-responsiveness-to-tumor-necrosis-factor-tnf-antagonist-therapy-in-rheumatoid-arthritis-ra-patients/