Diagnostic and Clinical Utility of Novel Anti-PAD4 Antibodies in Anti-CCP2-Negative and Anti-CCP2-Positive Rheumatoid Arthritis
Tyler O’Malley, Brittany Partain, Cynthia Alvarez, Clayton Nagle, Michael MahlerBackground/Objectives: Diagnosis of rheumatoid arthritis (RA) remains challenging, particularly in anti-CCP2-negative disease. Novel biomarkers, including anti-peptidyl arginine deiminase 4 (PAD4) autoantibodies, have emerged as promising tools for improving diagnostic accuracy in RA patients. The aim of this study was to analyze the performance of the first clinically validated anti-PAD4 assay. Methods: A total of 192 well-characterized RA patients and 383 controls were included in this study and tested for anti-CCP2, RF IgA, and RF IgM (all Phadia, Thermo Fisher) as well as anti-PAD4 IgA and anti-PAD4 IgG by ELISA (laboratory developed test, Exagen, Inc., Vista, CA, USA). Results for the anti-PAD4 IgG ELISA for 143 patients were used for method comparison with a particle-based multi-analyte technology (PMAT, Inova Diagnostics, San Diego, CA, USA). Results: ELISA and PMAT showed moderate-to-good agreement for anti-PAD4 IgG (rho = 0.66; Passing–Bablok slope 0.83, 95% CI 0.71–0.92; kappa = 0.67). Anti-PAD4 IgG was significantly correlated with anti-PAD4 IgA (rho = 0.47) and weakly correlated with RF IgM (rho = 0.29), RF IgA (rho = 0.28), and anti-CCP2 (rho = 0.22) (all p < 0.001). The PAD4 IgG and IgA ELISAs were validated in the RA population, yielding sensitivities/specificities of 31.8%/92.7% for anti-PAD4 IgG, 18.8%/93.7% for anti-PAD4 IgA, 43.2%/91.1% for RF IgM, 23.4%/96.1% for RF IgA, and 42.2%/99.2% for anti-CCP2, respectively. Anti-CCP2 IgG achieved the highest discriminatory performance (AUC = 0.86), followed by RF IgA (AUC = 0.69) and RF IgM (AUC = 0.67). Anti-PAD4 IgG and anti-PAD4 IgA showed modest discrimination (AUC = 0.55 and 0.52, respectively). Among the biomarkers evaluated, anti-CCP2 demonstrated the strongest association with RA (OR = 110.27), followed by RF IgM (OR = 7.40), RF IgA (OR = 7.14), anti-PAD4 IgG (OR = 5.92), and anti-PAD4 IgA (OR = 3.48) (all q < 0.001). In the anti-CCP2-negative population, the strongest association was found for anti-PAD4 IgG (OR = 2.87, q = 0.005), followed by RF IgM (OR = 2.20, p = 0.024). After adjustment for anti-CCP2, RF, age, and sex, anti-PAD4 IgG remained independently associated with RA (OR = 2.69, p = 0.003), with a small gain in discrimination (ΔAUC = 0.02–0.04). Conclusions: Our study established comparability between methods and shows that anti-PAD4 provides modest but statistically significant diagnostic information in the anti-CCP2-negative RA population, independent of anti-CCP2 and RF, in the anti-CCP2-negative RA population, where conventional markers are uninformative.