Correlation with digital dermoscopic images can help dermatopathologists to diagnose equivocal skin tumours

• Published in: British journal of dermatology (1951) Vol. 155; no. 3; pp. 546 - 551
• Main Authors: Bauer, J., Leinweber, B., Metzler, G., Blum, A., Hofmann-Wellenhof, R., Leitz, N., Dietz, K., Soyer, H.P., Garbe, C.
• Format: Journal Article
• Language: English
• Published: Oxford, UK Blackwell Publishing Ltd 01-09-2006; Blackwell
• Subjects: Adolescent; Adult; Aged; Aged, 80 and over; ambiguous tumours; Biological and medical sciences; Child; Child, Preschool; Dermatology; dermatopathology; dermatoscopy; dermoscopy; Dermoscopy - methods; Diagnosis, Differential; Female; Humans; Infant; Male; Medical sciences; Melanoma - pathology; Middle Aged; Nevus, Pigmented - pathology; Reproducibility of Results; Retrospective Studies; Skin Neoplasms - pathology; teledermatology; Teleradiology; Human; dermatopathology; Correlation; Skin disease; Dermatology; Digital image; Anatomic pathology; Telemedicine; Tumor; Skin; Diagnosis; teledermatology; Dermatoscopy; ambiguous tumours; dermoscopy
• ISSN: 0007-0963; 1365-2133
• DOI: 10.1111/j.1365-2133.2006.07342.x

Summary
Background  A variety of pigmented skin tumours can lead to diagnostic difficulties in dermatopathology.
Objectives  To investigate whether the interobserver agreement between histopathological diagnoses of equivocal pigmented tumours made by two referral centres can be improved by additional use of dermoscopic images.
Material and methods  Retrospective study using 160 tumours excised in the pigmented skin lesions clinic in Graz and 141 from Tübingen. Tumours were diagnosed in the referring centres using clinical data, histopathology and, if required, immunohistochemistry. The tumours were initially diagnosed as 74 melanomas, 218 melanocytic naevi and nine nonmelanocytic tumours. Haematoxylin and eosin sections, patients’ age and sex, tumour localization and digital dermoscopic images were then exchanged between the participating centres. Then, diagnoses were made initially based solely on dermatopathology and clinical information. After a washout phase, the same sections were reevaluated with the additional use of dermoscopic images. The main outcome measures were the Cohen's κ‐coefficients of the initial diagnoses of the centre submitting the cases and the diagnoses of the other centre without and with dermoscopy.
Results  The κ‐coefficient between the initial diagnoses with those made by the second centre without dermoscopy was 0·90 in Graz, 0·73 in Tübingen, and 0·81 overall. With the additional use of dermoscopy the κ‐value was invariably high with 0·89 in Graz, and improved to 0·87 in Tübingen, and to 0·88 overall.
Conclusions  The additional use of digital dermoscopic images further improved the overall very good agreement of histopathological diagnoses between two referral centres.