Discrepancy between SAA and CRP levels linked to the difference of SAA/CRP ratio in the patients with early rheumatoid arthritis

ABSTRACT Objectives Indeed, serum amyloid A (SAA) and C-reactive protein (CRP) reportedly seem to have moderate correlation, but discrepancies between CRP and SAA levels have often been reported in patients with early rheumatoid arthritis (ERA). This study aimed to determine the reasons for this discrepancy. Methods ERA patients (n = 206) were enrolled and treated with anti-RA drugs. Clinical features and disease activities were estimated. CRP and SAA levels were monitored, and the SAA/CRP ratio was compared. Correlations between CRP and SAA levels in individuals and between individuals and disease activity scores were examined. Results In a follow-up study, the SAA/CRP ratio remained almost constant over time in the same patients. However, SAA/CRP ratios differed widely between patients (0.233–106.3). In patients with high SAA/CRP ratios (>6.52), many (26.2%) had abnormal SAA values only. In patients with low SAA/CRP ratios (<6.52), not a few (6.8%) exhibited abnormal CRP values only. Conclusions The SAA/CRP ratio remained virtually constant in the same patients but differed dramatically between patients, which clarifies the discrepancy between CRP and SAA levels. CRP is the better marker in low-ratio patients but not in high-ratio patients; the SAA/CRP ratio is critical for its interpretation.