Ranking of 20 MHz sonography of malignant melanoma and pigmented lesions in routine diagnosis.

Purpose: Eleven dermatology clinics from all over Germany took part in our multicenter prospective study with the aim of evaluating 20 MHz sonography in the preoperative diagnosis of malignant melanomas and other pigmented skin tumours. It was to be assessed how effective sonographic measurement of thickness would compare to histology and the clinical palpation of tumour thickness and also the significance of differential diagnosis in sonography of malignant melanomas. Method: The prospective multicenter study proceeced as follows. To the end of August 1997 264 patients with a primary malignant melanoma and 417 patients with benign skin tumours were examined via 20 MHz sonography. Two different examiners estimated the clinical thickness of the tumour by palpation. The tumour was then excised and examined for postoperative correlation with the histology sections. Results: The final results showed good correlation between the histological and sonographic estimation of tumour thickness (r = 0.97). Estimation of tumour thickness by palpation showed no correlation with the histology (r = 0.59). Most of the benign (44%) and malignant tumours (38.7%) were spindle shaped. There was no significant difference between the benign and malignant tumour groups in relation to the sonographic presented shapes or echo signs. No different diagnosis could be made. Conclusion: The technique of high frequency sonography in relation to preoperative diagnosis of malignant melanomas has high priority. In contrast to clinical estimation of tumour thickness, sonography provided a good correlation to histology. The effectiveness of sonography with regard to the valence of the skin tumours is limited and there is no possibility of differentiating between malignant and benign tumours from the morphological face value. Hence, there is a demand for developing a 150 MHz apparatus which will be able to supply evidence regarding the valence of skin tumours.