Failure of gadopentetate dimeglumine-enhanced, high-resolution magnetic resonance imaging to differentiate among melanin-containing skin tumors.

Abstract

We evaluated the diagnostic potential of gadopentetate dimeglumine-enhanced, high-resolution magnetic resonance (MR) imaging to differentiate benign from malignant melanin-containing skin tumors. Forty-five patients were prospectively examined using high-resolution MR imaging at 1.5 T using a 2.5-cm surface coil. For tumor assessment, T1-weighted and T2-weighted transverse spin-echo sequences were acquired. After intravenous administration of gadopentetate dimeglumine (0.1 mmol/kg), the T1-weighted transverse sequence was repeated. Contrast enhancement was quantitatively determined as the percentage increase of signal intensity. Histologic findings were correlated using the Wilcoxon signed-ranks test. The quality of contrast enhancement was assessed by three independent investigators who were unaware of the patients' history and histologic data. The signal-to-noise ratio (SNR) was calculated in the T2-weighted sequence. Significance was tested using the Wilcoxon signed-ranks test. In all tumors, contrast enhancement was visually discernible. Half of the cases were enhanced inhomogeneously. The percentage of contrast enhancement did not correlate with histologic findings. Malignant melanomas could not be differentiated from benign melanocytic nevi with the use of gadopentetate dimeglumine. Determination of the SNR in T2-weighted sequences revealed no significant difference for histologic subgroups or tumor type. Gadopentetate dimeglumine-enhanced MR imaging does not differentiate malignant melanomas from benign melanocytic nevi. Determination of the SNR in the T2-weighted sequences revealed no significant difference for histologic subgroups.