Background/Purpose: Nailfold Capillaroscopy (NFC) is a non-invasive assessment of the nailfold area of the fingers which aids in the diagnosis of Juvenile Dermatomyositis (JDM) by providing images where the number of capillary end row loops/ mm (ERL) can be both visualized and counted.  Previous studies documented that ERL numbers are usually decreased at diagnosis and often change over time.  Although   many NFC studies employ non-digital methods of photo capture and analysis, new digital tools are well adapted for NFC photography, but their reliability   is unknown. The objective of this study was to explore the utility and reliability of digital methods of NFC in children with JDM.

Methods: First, the relationship between the Disease Activity Score (DAS) and end row capillary loops (ERL) was explored.  109 untreated children with JDM (78% females, 92% White or Hispanic,  mean age  6.7 ± 3.4 years,  mean duration of untreated disease 11.2 ± 16.9 months) had DAS and NFC obtained at diagnosis (IRB# 2012-14858), using the non-digital method.   To assess the standardization of the digital method, 12 randomly selected treated patients with JDM (42% Female, mean age 14.8± 6.4 years, White or Hispanic) were enrolled (IRB# 2011-14651).  NFC was performed on 8 fingers on each of the 12 patients, using three methods, by the same experienced investigator; there were 2 readers.  The first method, freeze frame video microscopy (FFVM), produced photos that were printed and not digitally stored.  For each method, the NFC were analyzed by counting the ERL per mm, which were then averaged. DermLite II ProHR (DL) (portable) and Leica Z6APO stereomicroscope (LS) (stationary) served as the two digital methods for image acquisition.  These photos were analyzed with Adobe Photoshop CS5 using the same ERL formula.  All 288 images were evaluated by each of the two investigators, one very experienced in NFC analysis (R1), and one a novice (R2).

Results:   Untreated children with JDM had a significant negative correlation between skin DAS and ERL (p=.0027), which was not found for muscle DAS and ERL (p= 0.43).    Both digital methods, DL and LS, were quite reliable in terms of quantification of the number of ERL, compared with the FFVM,  when read by R1 (r_c= 0.833, r_c=.846, Concordance Correlation Coefficient).  R2’s reliability was low using both digital methods (DL r= 0.436, LS r= 0.543).  This lack of concordance between the readers was further documented when DL was compared to LS (R1 r_c= 0.832, R2 r_c= 0.41).  The data from both the experienced and the novice investigators were concordant using the FFVM images (r_c= 0.834); however, the inter-rater reliability was poor when R1 and R2 were compared for the digital methods (DL r_c= 0.676, LS= 0.577).   

Conclusion: NFC can be used in the clinical setting to assess skin associated vascular changes  in JDM, providing  the  ERL number at diagnosis and over time to assess the vasculopathy.  FFVM is more easily interpreted by a novice, but digital methods pose a challenge to the novice. However, when read by an experienced investigator, the results from both methods are highly concordant. We speculate that this new, easily accessible technology may allow NFC to become a standard of care for children with JDM.

---

« Back to 2013 ACR/ARHP Annual Meeting

ACR Meeting Abstracts - https://acrabstracts.org/abstract/utility-and-reliability-of-digital-nailfold-capillaroscopy-in-children-with-juvenile-dermatomyositis-three-methods/