[Cytological diagnosis of thyroid nodules. Comparison of results obtained with guided echography with those from of a "blind" biopsy].

The goal of the study was to verify if echographic fine needle biopsy (US-FNB) improved the diagnostic accuracy of thyroid nodules with respect to a "blind" technique (B-FNB). The study group involved 6,693 patients who underwent aspiration biopsies (989 males and 5,704 females; mean age 46.3 +/- 13.7; range 5-88 years). Total biopsies were 7,952 as some patients underwent two or more biopsies in different nodules. 7,324 nodules with B-FNB and 628 nodules with US-FNB were studied; in this latter cases 136 nodules were not palpable. The diagnosis of malignancy was higher in US- FNB than B-FNB. However, non-diagnostic samples were more prevalent in US-FNB with respect to B-FNB. A histological diagnosis was available in 27 patients with US-FNB and 379 with B-FNB. In US-FNB, accuracy was 81.8%, sensitivity 86.6% and specificity 77.7%. In B-FNB, accuracy was 80.6%, sensitivity 85.1% and specificity 78.1%. These data show that US-FNB diagnostic accuracy was not better than B-FNB. The higher incidence of malignancy found in US-FNB was not dependent on better accuracy, sensitivity or specificity of the US-FNB technique with respect to the B-FNB examination. It is possible that the higher number of malignancies in patients studied with US-FNB was due to the fact that the US-FNB technique aspirated smaller nodules and particular areas of echographic suspicion, thus increasing the possibility of finding a malignancy. Based on our data, we feel that US-FNB results in a greater number of diagnostically inadequate samples. This may depend in part on the smaller size of the biopsied nodules (which are relatively more vascular than larger nodules) and in part on the greater difficulty in carrying-out this technique which directed towards particular echographic areas of suspicion resulting in greater biopsy blood contamination. The authors feel that US-FNB should be carried-out in all cases of non-palpable nodes found incidentally so as to rule out malignancy.