Background Anti-CCP testing is a part of the clinical routine investigations for rheumatoid arthritis (RA). RA has the highest incidence and prevalence at ages 70-80 yrs.
Objectives We aimed at investigating the predictive value of anti-CCP in a large population-based cohort of high-risk age.
Methods We used a subset of the Swedish twin registry, which includes 12 590 monozygotic and dizygotic twins born 1958 or earlier. Blood samples were analyzed for Anti-CCP2 using a commercial ELISA. Cut off set by the manufacture were used to define positive (≥cut off), low positive (≥cut off - ≤3x cut off) and high positive (>3x cut off) titers. All Anti-CCP positive samples were further investigated by ELISA for the presence of antibodies against autoantigen-derived citrullinated peptides (alpha-enolase: aa5-21; collagen type II: aa359-369; fibrinogen: aa566-580 and vimentin: aa60-75).
Cases of RA in the tested population were identified by linkage to the Swedish National Patient Register.
Sensitivity and specificity as well as the positive (PPV) and negative (NPV) predictive values of Anti-CCP positivity on the risk of RA were estimated.
Odds ratios (OR) estimating the association between RA and Anti-CCP Positivity were also calculated. Age was analyzed in categories (<60; 60-65, 65-70, >70 yr) for association to Anti-CCP Positivity and RA. For these analyses of association we used a binominal logistic regression model to adjust for sex and for clustering of data.
Results Mean age in the cohort was 65 year (range 48-93) at time point for blood donation. 350 out of 12 590 tested individuals (2.8%) were positive for Anti-CCP. 1.5% (192/12590) had RA. Of these RA cases, 65% (124/192) were positive for Anti-CCP. 51% of all Anti-CCP positive individuals (81% of the Anti-CCP RA patients) also had identifiable antibodies against at least one of the tested citrullinated peptides. Antibodies against citrullinated alpha-enolase were the most commonly detected.
The prevalence of Anti-CCP positivity did not display any variation across the age groups under study, nor did the prevalence of Anti-CCP positive RA. By contrast, the prevalence of Anti-CCP negative RA increased with age (p 0.006).
Anti-CCP Positivity (OR 95, 95% CI: 68-132), high titer Anti-CCP Positivity (OR 232, 95%CI: 160-337) as well as low titer Anti-CCP Positivity (OR 11, 95%CI: 5.3-22) were associated with RA. The sensitivity and specificity for Anti-CCP testing were 65% (95%CI: 58-71%) and 98% (95%CI: 98-98%), respectively. Both Anti-CCP and high titer Anti-CCP had a good predictive value for having RA (PPV 35.4%, 95%CI 30.4-40.4% for Anti-CCP and 56.9%, 95%CI 50.1-63.8% for high Anti-CCP) and a negative predictive values close to 100%.
Conclusions Detection of anti-citrullinated proteins antibodies by the CCP-2 method has a high positive predictive value, especially high titer Anti-CCP, when used to screen a large middle-aged population. A follow up study to identify new incident cases of RA among tested individuals will allow more accurate estimates in the future.
Disclosure of Interest None Declared