[Comparison of the sonographic detection of experimentally induced squamous cell carcinoma compared with the...

Malignant tumours of the head and the neck are characterised by typical signs of malignancy in greyscale- and colour-coded sonography. Sometimes, such criteria cannot be verified, and in such cases it remains unclear whether typical changes do not exist or whether we just cannot detect them with our high-end ultrasound units. We therefore compared our sonographical findings with the histology obtained in experimentally induced tumours.

Experimental squamous cell carcinoma was induced subcutaneously in nude mice (n = 18), using four different cell lines. Ultrasound examination of the tumours was performed after 98 and 112 days, respectively, Central necrosis, rupture of the capsule as well as the presence of central and peripheral blood vessels were documented before the tumours were excised. They were then sliced in the same direction as sonography had been carried out before. Blood vessels were stained using CD31.

16 tumours grew subcutaneously. In ultrasound imaging, the capsule was always intact, and subcapsular vessels were detected in every tumour. Central necrosis was seen in 10/16 tumours by sonography, but in 15/16 tumours on histological examination. The failure of sonography in detecting necrosis did not correlate with the size of the tumour. Centrally located vessels were found in 12/16 tumours using colour-coded sonography. Histology, however, showed their presence in all 16 tumours. Interestingly, the failure of sonography in detecting vessels did not correlate with tumour size. It seemed that the detection of such vessels was particularly difficult in those tumours which grew very fast. This finding explains why we often find such vessels in small tumours but not in large specimens, where they would normally be expected.