Increased predictive value of optical spectral transmission in early rheumatoid arthritis through use of patient-adjusted cut-off scores

Abstract

The aims of this study were to suggest patient-adjusted optical spectral transmission (OST) cut-off values for the first time and to develop clinical models that predict the probability of an early rheumatoid arthritis (RA) diagnosis based on OST findings and the 2010 American College of Rheumatology/European League Against Rheumatism (ACR/EULAR) classification criteria as a reference standard. OST examinations were performed in newly diagnosed RA patients and healthy controls by the HandScan device. Moreover, RA patients underwent a full clinical [tender/swollen joint counts (TJC/SJC), disease activity score-28 (DAS28)] and laboratory evaluation. OST confounding factors were examined via logistic multivariate regression analyses and patient-adjusted OST-cut-offs were subsequently determined. Furthermore, statistical models to calculate the probability of an RA diagnosis, based on the measured OST values and the presence of OST influencing factors, were developed. Finally, correlations of OST with RA activity parameters were assessed. 1.584 joints of 72 early RA patients were examined via OST and compared to 2.200 joints of 100 healthy controls and 1.166 joints of 53 patients with non-inflammatory arthralgia (NIA), respectively. Overall OST diagnostic performance was excellent in the whole cohort between RA- and healthy control-group [Area-Under-the-Curve (AUC): 0.810 (95%CI: 0.746–0.873); p < 0.0001], and further improved in RA-patients with ≥ 1 swollen wrist/finger joint(s) [AUC: 0.841 (95%CI: 0.773–0.908); p < 0.0001]. Comparison between RA patients and patients with non-inflammatory arthralgia showed similar results by an AUC of 0.788 (95%-CI: 0.709–0.867; p < 0.0001), and further improved in RA patients with ≥ 1 swollen wrist/finger joint(s) [AUC: 0.822 (95%CI: 0.74–0.90); p < 0.0001]. For the assessment of an adjusted RA diagnosis probability, two gender-specific statistical models were developed, based on OST values and patient age. OST cut-off values of 11.2 and 18.21 were calculated for female and male patients with active disease (sensitivity 93% and 67%; specificity 71.2% and 90%), respectively. Among RA patients, OST was associated moderately/significantly with DAS28 (r = 0.42,p < 0.001) and swollen joint count (rho = 0.355,p = 0.002). The development of patient-adjusted OST cut-off values and the suggested statistical models significantly enhance OST’s diagnostic performance, supporting its utility in differentiating between RA and non-inflammatory conditions. Future research should include a broader spectrum of arthritis types to validate OST’s comprehensive diagnostic utility also across various inflammatory arthritides. DRKS00016752 (German Registry of Clinical Trials)