Network meta-analysis of novel diagnostic biomarkers for rheumatoid arthritis: comparative performance of anti-CarP, anti-MCV, and emerging markers.

OBJECTIVE

Rheumatoid arthritis (RA) diagnosis remains challenging due to the limitations of conventional biomarkers. This network meta-analysis (NMA) aimed to compare the diagnostic performance of seven novel biomarkers for RA.

METHODS

We systematically searched EMBASE, PubMed, and Cochrane databases, identifying 1,713 records. Twelve studies met inclusion criteria. Diagnostic accuracy metrics (sensitivity, specificity, log diagnostic odds ratio [lnDOR]) were extracted. A frequentist NMA was performed using the netmeta package in R, with biomarkers ranked via P-scores against the ACR/EULAR RA criteria. Heterogeneity, transitivity, and publication bias were assessed.

RESULTS

Anti-carbamylated protein (Anti-CarP) antibody showed the highest diagnostic efficacy (P-score = 0.80; sensitivity 62 - 64%, specificity 89%), followed by anti-mutated citrullinated vimentin (Anti-MCV; P-score = 0.69). miR- 146a and power Doppler ultrasound (PDUS) demonstrated moderate performance (P-scores 0.46 and 0.45), while 14-3-3 η protein (P-score = 0.26) and its combination with ACPA (P-score = 0.07) ranked lowest. The network exhibited a star-shaped geometry with no significant inconsistency or publication bias (Egger 's test p = 0.52).

CONCLUSION

Anti-CarP and Anti-MCV are the most promising diagnostic biomarkers for RA, Anti-CarP and Anti-MCV are the most promising diagnostic biomarkers for RA. Their potential utility in seronegative or early-stage RA warrants further investigation. The limited performance of 14-3-3 η challenges prior assumptions about its utility. Standardized thresholds and direct comparisons are needed for validation.