Optical spectral transmission to assess inflammation in hand and wrist joints of rheumatoid arthritis patients

Abstract

Objective: To develop an optical spectral transmission (OST) model to measure joint inflammation, and thus disease activity, as well as to evaluate (patho-)physiological findings that could lead to misclassification of inflammation. Methods: Forty-six RA patients were included in this cross-sectional study, where US scores, duplicate OST measurements and 28-joint DAS (DAS28) were acquired. With US as a reference standard, the diagnostic performance of OST in detecting inflammation at the joint level was evaluated using receiver operating characteristic (ROC) curve analyses. At the patient level, correlations with US were analysed for DAS28 and OST, and at joint level for OST and tender and swollen joint counts (TJC and SJC, respectively). Joint pathology potentially influencing misclassification by OST [erosions, osteophytes, tendon (sheath) inflammation (ab)normal vasculature and chondrocalcinosis] was evaluated for significance in a multivariate nominal logistic regression model. Results: Diagnostic performance of OST was good for MCP [area under the ROC curve (AUC-ROC) 0.88], PIP (AUC-ROC 0.83) and wrist (AUC-ROC 0.74) joints and for all joints together (AUC-ROC 0.85). At the patient level, DAS28 correlated very poorly (ρ = 0.06) and OST moderately (ρ = 0.54) with US. At the joint level, US correlation with OST was strong (ρ = 0.64), with SJC it was weak (ρ = 0.30) and with TJC it was very weak (ρ = -0.02). Misclassification of inflammation by OST was relatively rare (17%). Dorsal erosions [odds ratio (OR) 4.0], osteophytes (OR 2.1) and extensor tendinitis (OR 4.6) increased the risk of underestimating inflammation of MCP and PIP joints and osteophytes (OR 3.0) also increased the risk of overestimating inflammation. Conclusion: OST is a sensitive, specific and objective technique to assess joints inflammation of the hands and wrists of RA patients, even though bone and tendon pathology increases the risk of misclassification.