Abstract 5877: Evaluation of the Melanoma Institute Australia nomogram in predicting sentinel lymph node positivity in patients with cutaneous melanoma

Abstract

Abstract Background: The Melanoma Institute Australia (MIA) nomogram was developed to predict the positivity of sentinel lymph node (SLN) in patients with primary cutaneous melanoma. We sought to validate the MIA nomogram against the University of Pittsburgh SPORE database. The secondary aim was to assess whether the addition of the degree of tumor-infiltrating lymphocytes (TILs) and location of the primary melanoma increases the predictive value of the nomogram. Methods: Patients were managed in the Melanoma Center of the UPMC Hillman Cancer Center (HCC) and the Melanoma and Skin Cancer Program (MSCP) SPORE Database. The status of SLN, primary location and degree of TIL infiltration, and the parameters adopted by the MIA nomogram including age, histological type, Breslow thickness, mitotic rate, ulceration, and lymphovascular invasion (LVI) were recorded from the electronic medical record. The MIA nomogram performance was evaluated by the area under the Receiver Operating Characteristic (ROC) curve. Accuracy, sensitivity, and specificity were calculated for the most accurate cut-off value. Logistic regression was used to assess the effects of TIL and the location of melanoma. Results: In total, 68 patients were assessed, 44 (64.7%) with negative SLN, and 24 (35.3%) with positive SLN. The median age (interquartile range, IQR) was 60.5 years (49.8, 68.3). The most common histological subtype was nodular n=43 (63.2%) followed by 18 (26.5%) superficial spreading, 4 (5.8%) lentigo maligna, and 3 (4.5%) desmoplastic. Median Breslow thickness (IQR) was 3.9 mm (3.0-5.1). Ulceration was present in 47 (69.1%), and 14 (20.6%) had a LVI identified. The mitotic rate was 0, 1/mm2, 2/mm2, 3/mm2, and >=4/mm2 in 6 (8.8%), 13 (19.1%), 12 (17.6%), 8 (11.8%), and 29 (42.6%) cases, respectively. TIL was absent, non-brisk, and brisk in 16 (23.5%), 40 (58.9%), 12 (17.6%) of cases, respectively. Location for 32 (47.6) was the trunk, 29 (42.6%) the extremities, and 7 (10.3%) the head and neck area. The AUC was 72% with 95% CI (58%-85%). At the most accurate cutoff value the MIA nomogram had 54% sensitivity, 84% specificity, 65% positive predictive value, 77% negative predictive value, and 74% accuracy. Location of melanoma and TIL did not demonstrate any significant predictive value beyond the risk score (p=0.24 and p=0.92, respectively). Conclusion: The MIA nomogram demonstrated only fair predictive value. Neither location of melanoma nor TIL were significant when added to risk score. The evaluation of additional factors, such as the gene expression profile, is warranted to improve prediction of SLN status. Citation Format: Xi Yang, Lilit Karapetyan, Andrew D. Knight, Cindy Sander, William E. Gooding, John M. Kirkwood. Evaluation of the Melanoma Institute Australia nomogram in predicting sentinel lymph node positivity in patients with cutaneous melanoma [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2022; 2022 Apr 8-13. Philadelphia (PA): AACR; Cancer Res 2022;82(12_Suppl):Abstract nr 5877.