Differential performance of nailfold video capillaroscopic parameters in the diagnosis and prognosis of systemic sclerosis

Objective. The contribution of nailfold video capillaroscopy (NVC) in identifying patients with Raynaud's phenomenon (RP) at risk for systemic sclerosis (SSc) is well established. Herein we comparatively assess the performance of different capillaroscopic parameters in diagnosing SSc among patients with RP and evaluate the prognostic capacity of NVC in SSc. Methods. At baseline we clinically and capillaroscopically evaluated 242 consecutive patients referred to our department for NVC (138 with SSc); 175 were reevaluated after 3.38 +/- 1.47 years. Sixty-two healthy volunteers served as controls. Capillaroscopy pattern (normal/early/active/late) was qualitatively defined. Capillary loss, dilated, giant or ramified capillaries and micro-haemorrhages were scored semi-quantitatively. Results. Capillary loss score had the highest diagnostic accuracy at discriminating patients with an SSc-spectrum disorder from patients with RP of different etiology and controls, as defined by ROC curve analysis [AUC (95% CI)=0.905 (0.869-0.942)], followed by dilatation score [0.863 (0.818-0.907)] and giant score [0.835 (0.787-0.884)]. By contrast, micro-haemorrhages [0.720 (0.662-0.779)] and ramifications scores [0.604 (0.539-0.670)] performed worse. Multivariate analysis in 94 SSc patients indicated that active (OR=3.305, p=0.043) and late (OR=6.900, p=0.023) baseline capillaroscopy pattern predicted occurrence of a combined adverse disease outcome [forced vital capacity (FVC) deterioration>10% and/or DLCO deterioration>15% and/or mRSS deterioration>3.5 and/or first occurrence of digital ulcers and/or death)] at 3 year follow-up. Conclusion. Dilatation score performs best of all semi-quantitative NVC parameters in diagnosing SSc. In addition, our study confirms earlier reports that worse capillaroscopy pattern at baseline correlates with higher likelihood for adverse prognosis.