Deep learning model shows pathologist-level detection of sentinel node metastasis of melanoma and intra-nodal nevi on whole slide images.

Introduction: Nodal metastasis (NM) in sentinel node biopsies (SNB) is crucial for melanoma staging. However, an intra-nodal nevus (INN) may often be misclassified as NM, leading to potential misdiagnosis and incorrect staging. There is high discordance among pathologists in assessing SNB positivity, which may lead to false staging. Digital whole slide imaging offers the potential for implementing artificial intelligence (AI) in digital pathology. In this study, we assessed the capability of AI to detect NM and INN in SNBs. Methods: A total of 485 hematoxylin and eosin whole slide images (WSIs), including NM and INN from 196 SNBs, were collected and divided into training (279 WSIs), validation (89 WSIs), and test sets (117 WSIs). A deep learning model was trained with 5,956 manual pixel-wise annotations. The AI and three blinded dermatopathologists assessed the test set, with immunohistochemistry serving as the reference standard. Results: The AI model showed excellent performance with an area under the curve receiver operating characteristic (AUC) of 0.965 for detecting NM. In comparison, the AUC for NM detection among dermatopathologists ranged between 0.94 and 0.98. For the detection of INN, the AUC was lower for both AI (0.781) and dermatopathologists (range of 0.63-0.79). Discussion: In conclusion, the deep learning AI model showed excellent accuracy in detecting NM, achieving dermatopathologist-level performance in detecting both NM and INN. Importantly, the AI model showed the potential to differentiate between these two entities. However, further validation is warranted.

FLI-1/Melan-A dual stain is an alternative to PRAME in differentiating metastatic melanoma from nodal nevus: A monocentric retrospective study.

Melanocytic nevi existing in lymph nodes create a diagnostic challenge by mimicking metastases. PReferentially expressed Antigen in MElanoma (PRAME) immunohistochemical (IHC) stain can differentiate one from another. FLI-1 IHC expression has been shown in malignant melanoma with variable sensitivity while melanocytic nevi were reported to be negative. We hypothesized that FLI-1/Melan-A dual IHC staining may be used in the distinction of metastatic melanoma from nodal nevi and can be an alternative and/or complementary to PRAME. In this study, we examined 13 lymph nodes with metastatic melanoma and 13 lymph nodes with benign deposits. We stained all of the lymph nodes with FLI-1, FLI-1/Melan-A dual, and PRAME IHC stains. In addition, we stained paired skin samples of the metastatic lymph nodes with FLI-1 and PRAME. In primary cutaneous melanomas, 11 of 13 were positive for FLI-1 and PRAME expression (85%). Malignant cells in 12 and 13 lymph nodes showed positive expression of PRAME and FLI-1, respectively. Only one case with a nevic cell deposit was weakly positive for FLI-1 and the remaining benign cases were negative for both FLI-1 and PRAME. Our results show that FLI-1/Melan-A dual stain is as sensitive and specific as PRAME in distinguishing lymph nodes with metastatic melanoma from nodal nevi. Further studies with larger case numbers are needed to support our significant results.

Melanocytic nevi in sentinel lymph nodes: association with cutaneous nevi and clinical relevance in patients with cutaneous melanomas.

PURPOSE: Melanocytic nevi in lymph nodes (NNs) are an important histological differential diagnosis of initial sentinel lymph node (SN) metastasis in melanoma. Our aim was to associate NN in SNs with clinicopathologic features and survival rates in 1, 250 patients with SN biopsy for melanoma. METHODS: To compare patients with present and absent NN, we used Fisher's exact test, Mann-Whitney U test, and multivariate logistic regression models in this retrospective observational study based on a prospectively maintained institutional database. RESULTS: NN prevalence in axillary, cervical, and groin SNs was 16.5%, 19.4%, and 9.8%, respectively. NN were observed in combination with all growth patterns of melanoma, but more frequently when the primary was histologically associated with a cutaneous nevus. We observed a decreasing NN prevalence with increasing SN metastasis diameter. Multiple logistic regression determined a significantly increased NN probability for SNs of the neck or axilla, for individuals with ≥ 50 cutaneous nevi, midline primary melanomas, and for individuals who reported non-cutaneous malignancies in their parents. Cancer in parents was also significantly more frequently reported by melanoma patients who had more than 50 cutaneous nevi. In SN-negative patients, NN indicated a tendency for slightly lower melanoma-specific survival. CONCLUSIONS: We found a highly significant association between NN diagnosis and multiple cutaneous nevi and provided circumstantial evidence that cutaneous nevi in the drainage area of lymph nodes are particularly important. The trend toward lower melanoma-specific survival in SN-negative patients with NN suggests that careful differentiation of SN metastases is important.

High discordance rate in assessing sentinel node positivity in cutaneous melanoma: Expert review may reduce unjustified adjuvant treatment.

INTRODUCTION: Identification of sentinel node (SN) metastases can set the adjuvant systemic therapy indication for patients with stage III melanoma. Studies re-evaluating the diagnosis of initially positive SN biopsies are scarce. MATERIALS AND METHODS: Dutch patients with melanoma who underwent SN biopsy between 2003 and 2014 were selected from PALGA, the Dutch Pathology Registry. Histopathological slides of SN-positive patients were retrieved for review. A random sample was reassessed by an expert melanoma pathologist. Recurrence-free survival (RFS) of patients who were misclassified (false-positive) was compared with those with a true positive SN status. For comparison, a group of SN-negative patients was included. Multivariable logistic analysis was performed to assess clinicopathological characteristics associated with misclassification of SN status. RESULTS: Diagnosis was downgraded from melanoma metastasis to nodal nevus in 38 of the 322 reviewed patients (11.8%). Considering the inclusion criteria of phase III adjuvant trials, at least 4.3% of patients would have falsely qualified for adjuvant therapy. In multivariable analysis, patients with a low SN tumour burden and subcapsular SN tumour location had a significantly higher chance of being misclassified. The five-year RFS of the 38 downgraded patients was 86.7% (95% confidence interval [CI] = 72.6-96.6), similar to the 85.9% (95% CI = 84.9-86.8, p = 0.18) for 6413 SN-negative patients and better than the 53.2% (95% CI = 47.2-59.9, p = 0.009) of 284 patients who were truly SN positive upon review. CONCLUSION: More than 10% of originally positive SN biopsies of patients with melanoma concern misclassified nodal nevi. We advocate that when adjuvant treatment is considered in patients with stage III melanoma, SN biopsies should be reassessed by an expert melanoma pathologist.