The predictive value of antinuclear antibodies in patients with suspected connective tissue disease for the development of connective tissue diseases.

Early confirmation of the diagnosis of connective tissue diseases (CTD) is important, as prolonged disease activity can result in irreversible organ damage. Although antinuclear antibodies (ANAs) have been shown to precede the diagnosis of SLE, this has not been investigated in large cohorts for other CTDs. In this study, we investigated whether the presence of antinuclear autoantibodies in undiagnosed patients suspected of having CTDs is predictive of development of a future CTD. We screened the Electronic Health Records of a cohort of 1030 patients, who were tested for ANAs and their specificity in 2013/2014, to evaluate whether new CTD diagnoses had been recorded by a clinician between the original blood draw date and 2020. We compared the prevalence of ANAs in patients who developed a new CTD diagnosis during follow-up with patients with similar symptoms at baseline who did not receive a subsequent CTD diagnosis and with patients with an established CTD at baseline. Sixteen out of 1030 patients had developed a new CTD in the studied time period. The mean time period between baseline blood draw and subsequent CTD diagnosis of these patients was approximately 2.3 years. Eleven out of 16 (69%) newly diagnosed patients had positive ANA screening tests, compared to 54% of patients with a CTD diagnosis at baseline (p=ns) and 30% of symptomatic undiagnosed patients (p<0.001). This resulted in a positive predictive value (PPV) of 7% and a negative predictive value (NPV) of 98% for the development of a new CTD in undiagnosed symptomatic patients. For specific ANAs associated with the suspected CTD, the PPV was 12%, with a NPV of 98%. Progression to a CTD diagnosis is rare in undiagnosed patients. Undiagnosed patients with symptoms associated with a CTD who progress to a CTD more often have ANAs than patients with similar symptoms who do not progress to a CTD. ANA testing can be used to more stringently select patients who should remain in follow-up.