Background and objectives The afferent lymphatic system and the draining lymph node (LN)act as complementary check-points through progressive steps of the inflammatory and resolution processes. We have previously shown in a proof-of-concept study that LN Power Doppler ultrasound (PDUS) allows detection of qualitative changes in rheumatoid arthritis (RA). In this study we developed a quantification approach for PDUS assessment of the inflammatory node and investigated its clinical and pathologic significance.
Materials and methods Forty RA patients refractory to conventional DMARDs were evaluated through complete clinical and PDUS examination in hands/wrists and axillary LN. Twenty healthy individuals were recruited as controls. Thirty-one patients starting anti-TNF treatment were followed-up prospectively at weeks 4 and 24. Axillary LN were investigated by digital image analysis and semi-quantitatively focusing on volumetric measures, vascular perfusion and micro-structure of the lymphocyte-rich cortex. Patient-related US indices for each parameter were constructed for correlative, response-to-treatment and predictive analyses.
Results LN perfusion and morpho-structural features demonstrated specific quantitative increase in RA LN leading to distinct sonotypes differentiating patients from controls. LN PDUS parameters showed however a high degree of heterogeneity with values exceeding the threshold of controls restricted to 42.5% of the cases. No differences in LN status were observed in relation to autoantibody positivity. LN indices showed instead a link with peripheral joint inflammation as inferred by the significant correlation (at systemic and ipsilateral level) between LN and peripheral joint PD (r = 0.35, p = 0.03). At week 24, 61.3% of the patients were EULAR responders. Patients with different treatment outcomes did not show any significant difference in baseline clinical features and joint US. In contrast, the LN PD index was significantly lower in non-responders (median [IQR] 0 [0–0.5] vs 2 [1–4.75], p = 0.007). Lack of changes of LN PD above the threshold of controls (PD index≤2) was the only predictor of failure to achieve response with an OR of 9.9 (p = 0.04).
Conclusions Synchronised and longitudinal PDUS monitoring of joint inflammation and LN reactivity is an achievable goal in patients with RA. Lack of draining LN involvement in patients with active disease seems to associate with a less favourable anti-TNF treatment response.