A multicentre pilot study investigating high-definition optical coherence tomography in the differentiation of cutaneous melanoma and melanocytic naevi.

BACKGROUND: High-definition optical coherence tomography scanners have recently been developed. OBJECTIVES: To assess the diagnostic performance of HD-OCT in the differentiation of benign melanocytic skin lesions (MSL) and cutaneous melanoma (CM). METHODS: Patients with MSL were assessed by HD-OCT. All diagnoses were histopathologically confirmed. One blinded observer evaluated both slice and en-face HD-OCT images and diagnosed MLS on the basis of an algorithm adopted from reflectance confocal microscopy, recent HD-OCT reports, and conventional OCT. RESULTS: A total of 93 MSL were studied comprising 66 benign MSL and 27 CM. The sensitivity of HD-OCT was 74.1% [95% confidence interval (CI) 53.7-88.8%)], specificity was 92.4% (95% CI 83.2-97.5%). The positive predictive value was 80%, the negative predictive value 89.7%. The performance of HD-OCT depended on tumour thickness and the presence of borderline lesions indicated by high false negative rates in very thin melanomas and high false positive rates in dysplastic naevi. CONCLUSIONS: In the distinction of MSL, HD-OCT applied in an investigator blinded fashion has a moderate diagnostic performance. The diagnostic performance of HD-OCT of MSL should be reassessed in other clinical settings.

Histopathological correlates of basal cell carcinoma in the slice and en face imaging modes of high-definition optical coherence tomography.

BACKGROUND: High-definition optical coherence tomography (HD-OCT) scanners have recently been developed, providing significantly higher resolution than conventional OCT. OBJECTIVES: To assess the relationship between recently defined histopathological HD-OCT correlates of basal cell carcinomas (BCC) and possible predictors for the most common tumour subtypes. METHODS: For HD-OCT imaging, we used the Skintell(®) device. Twenty-five BCCs were histopathologically confirmed (including both vertical and horizontal haematoxylin and eosin and Alcian blue sectioning) and correlated with HD-OCT images. RESULTS: In the en face mode, lobulated nodules were seen in 21/25 BCCs (84%), peripheral rimming in 18/25 (72%), epidermal disarray in 18/25 (72%) and variably refractile stroma in 22/25 (88%). In the slice imaging mode, we observed destruction of layering in 19/25 (76%) BCCs. In both the slice and en face modes a significant correlation was observed between peritumoral rimming and grey/dark oval structures and lobulated nodules. Alcian blue stains showed peritumoral mucin deposits correlating with peripheral rimming around the tumour nodules. In a logistic regression model, we did not observe significant independent micromorphological HD-OCT predictors for either the solid or superficial BCC subtypes. CONCLUSIONS: In agreement with recent studies we have demonstrated that HD-OCT using the slice and en face imaging modes can visualize histopathological correlates of BCC, and potentially aid noninvasive diagnostics. However, using HD-OCT correlation it was not possible to predict the superficial or solid BCC subtypes. For the first time we have shown that peripheral rimming in HD-OCT correlates with peritumoral mucin deposition.