Background/Purpose: Our objective is to examine the additional contribution of 14-3-3η levels to known predictors of radiographic progression in inflammatory polyarthritis, both over several years and over the following year.

Methods: All variables were measured at initial and annual visits up to 4 years in patients with recent onset polyarthritis treated to remission. HIGH 14-3-3η was ≥0.50 ng/ml; elevated CRP ≥8 mg/L; rapid radiographic progression (RRP) was ≥5 Units per year in Sharp van der Heijde (Sharp) Erosion score. General linear models (Glimmix) combining baseline predictors and treatments over time were performed with random effect for repetitions over time to assess the independent effect of 14-3-3η on radiographic progression; Model 1: Positive anti-CCP2; Model 2: Positive Rheumatoid Factor (RF); Model 3: Positive anti-CCP2 and RF plus elevated CRP; Model 4: Age, Gender, Symptom Duration, Smoking, DAS28-CRP, Sharp, HAQ and Treatments, Model 5: All variables in Models 3 and 4. Generalized Estimated Equations (GEE) assessed the association with RRP of anti-CCP2, elevated CRP, and HIGH 14-3-3η, alone and in combination.

Results: Mean age was 58.5 years, 60.2% women; median duration at inclusion 3.4 months. Out of 1529 complete evaluations (including baseline) in 533 patients, 511 (33.4%) were HIGH 14-3-3η, 590 (38.6%) positive anti-CCP2, 554 (36.2%) positive RF, 476 (31.1%) elevated CRP; 157 (10.3%) RRP episodes.

In univariate analyses, baseline 14-3-3η levels (as did positive anti-CCP2, positive RF and elevated CRP) predicted erosion scores at 42 months (median (IQR): 5.0 (1.0-10.0) vs 2.0 (0.0-7.0), p<0.001) and radiographic progression (mean ± SD: 5.7 ± 10.2 vs 3.3 ± 8.5, p<0.001). HIGH 14-3-3η, positive anti-CCP2 and elevated CRP were each associated with increased Relative Risks (RR) of RRP: 1.82 (1.36-2.43) p<0.001; 2.33 (1.65-3.30) p<0.001; 2.21 (1.65-2.95), p<0.001, respectively.

In multivariate Glimmix models, continuous 14-3-3η remained a highly significant independent predictor of Erosion progression in all Models, including the 5 described above (all p <0.001).

In GEE, relative to being negative for all three, being positive for all of HIGH 14-3-3η, anti-CCP2, and CRP was associated with a RR of 6.19 (3.82-10.05), p<0.001; RRP occurred following 34.1% of visits when all three were positive. Being positive for only 2 variables identified subsets of patients at intermediate RR (1.84 to 4.42) of RRP. The contribution of HIGH 14-3-3η to predict RRP was highest in patients with positive anti-CCP2 and/or elevated CRP; sensitivity was then 46.2%, specificity 84.0% and negative predictive value 93.3%.

Conclusion: In patients treated to remission, baseline 14-3-3η levels were predictive of erosion scores and erosion progression, but not of narrowing, over the following 4 years, even when combined with multiple baseline predictors of erosive progression and taking into account treatments received from baseline. The presence of HIGH 14-3-3η amplified the risk (up to 34%) for RRP conferred by anti-CCP2 and/or CRP. Adding 14-3-3η measurement to anti-CCP2 and clinical measures over the course of RA may inform therapeutic strategies tailored to halt rapid joint damage progression in the most susceptible patients.

« Back to 2018 ACR/ARHP Annual Meeting

ACR Meeting Abstracts - https://acrabstracts.org/abstract/levels-of-14-3-3eta-are-an-independent-predictor-of-long-term-radiographic-erosive-progression-and-of-short-term-rapid-radiographic-erosive-progression-in-patients-with-inflammatory-polyarthritis/