Histopathologic abundance of pigmentation correlates with disease-specific survival in malignant melanoma but is not independent of current AJCC pT stage.

The increasing number of melanoma patients makes it necessary to develop best possible strategies for prognosis assessment in order to recommend appropriate therapy and follow-up. The prognostic significance of tumor cell pigmentation has not been fully elucidated. Hematoxylin and eosin (H&E)-stained sections of 775 melanomas diagnosed between 2012 and 2015 were independently assessed for melanin pigment abundance by two investigators, and the impact on melanoma-specific survival was calculated. Unpigmented melanomas (n = 99) had a melanoma-specific survival of 67.7%, melanomas with moderate pigmentation (n = 384) had a melanoma-specific survival of 85.9%, and strongly pigmented melanomas (n = 292) had a melanoma-specific survival of 91.4% (p < .001). In an analysis of melanoma-specific survival adjusted for pT stage and pigmentation, we found a nonsignificant impact of pigmentation abundance with a hazard ratio of 1.277 (p = .74). The study presented here provides evidence in a German cohort that patients with pigmented melanomas have a more favorable prognosis than those diagnosed with nonpigmented melanomas. Moreover, the abundance of pigmentation already seems to provide a first prognostic estimate. However, it does not appear to provide significant additional value for prognostic assessment according to the AJCC 2017 pT classification.

Development of an Image Analysis-Based Prognosis Score Using Google's Teachable Machine in Melanoma.

BACKGROUND: The increasing number of melanoma patients makes it necessary to establish new strategies for prognosis assessment to ensure follow-up care. Deep-learning-based image analysis of primary melanoma could be a future component of risk stratification. OBJECTIVES: To develop a risk score for overall survival based on image analysis through artificial intelligence (AI) and validate it in a test cohort. METHODS: Hematoxylin and eosin (H&E) stained sections of 831 melanomas, diagnosed from 2012-2015 were photographed and used to perform deep-learning-based group classification. For this purpose, the freely available software of Google's teachable machine was used. Five hundred patient sections were used as the training cohort, and 331 sections served as the test cohort. RESULTS: Using Google's Teachable Machine, a prognosis score for overall survival could be developed that achieved a statistically significant prognosis estimate with an AUC of 0.694 in a ROC analysis based solely on image sections of approximately 250 × 250 µm. The prognosis group "low-risk" (n = 230) showed an overall survival rate of 93%, whereas the prognosis group "high-risk" (n = 101) showed an overall survival rate of 77.2%. CONCLUSIONS: The study supports the possibility of using deep learning-based classification systems for risk stratification in melanoma. The AI assessment used in this study provides a significant risk estimate in melanoma, but it does not considerably improve the existing risk classification based on the TNM classification.