Accuracy of Preoperative MRI with Microscopy Coil in Evaluation of Primary Tumor Thickness of Malignant Melanoma of the Skin with Histopathologic Correlation

ABSTRACT

OBJECTIVE: The purpose of this study was to evaluate the accuracy of a preoperative MRI with microscopy coils in determining the primary tumor thickness of malignant melanoma with histopathologic correlation. MATERIALS AND METHODS: Eleven patients with histopathologically confirmed malignant melanoma were included in this study. MR images of the tumors were obtained with a 47-mm microscopy coil on 1.5T MR scanners and were evaluated by two radiologists, who assessed the thickness of the primary tumor on T2-weighted images (T2WI) and gadolinium-enhanced T1-weighted images with fat suppression (Gd-T1WI) and compared the results with the histopathologic findings as the reference standard. Correlations between tumor thickness on MRI and histopathologic examination were assessed using concordance correlation coefficients (CCCs). Inter- and intraobserver variabilities of tumor measurements were also assessed by intraclass correlation coefficient (ICC). RESULTS: Among the 11 cases included in the study, 10 cases from the same number of patients were managed with surgical excision and one case was confirmed with punch biopsy. The primary tumor thickness measured on T2WI showed better correlation with histopathologic results, as compared with measurements taken on Gd-T1WI the CCC of measurements on T2WI ranged from 0.64 to 0.78, indicating a substantial agreement, whereas the CCC of measurements on Gd-T1WI ranged from 0.50 to 0.61, indicating a moderate to substantial agreement. Inter- and intraobserver agreements of readers 1 and 2 were excellent for both T2WI and Gd-T1WI, with ICC ranging from 0.86 to 0.99. CONCLUSION: MR imaging with microscopy coils may be an accurate technique in the preoperative assessment of tumor thickness in malignant melanoma, especially on T2-weighted images.