[Accuracy of intraoperative frozen section diagnosis in non palpable breast lesions: a series of 791 cases]. / Efficacité de l'examen peropératoire des lésions infracliniques mammaires: études sur une série de 791 cas.

The aim of this study is to estimate the accuracy and reliability of intraoperative frozen section of nonpalpable breast lesions. In fact, frozen section of palpable breast lesions has proven to be valid, but its use in breast infraclinic lesions has been discussed recently, with the publication of European recommendations. Diagnosis on frozen section was routinely performed in a serie of 791 patients with nonpalpable mammographically detected abnormalities breast lesions from January 1990 through July 2000. The initial frozen section diagnoses with known mammographic pattern were compared with the diagnoses obtained on permanent paraffin sections to estimate the accuracy of frozen sections. Frozen section diagnosis was delayed until final diagnosis assessed on permanent paraffin sections in only 8 cases (1%). Frozen section diagnoses were accurate in 744 of 783 cases (95%). The diagnosis was modified on the basis of permanent sections in 39 cases; consisting of 39 false negative. No false positive diagnosis was noted. Sensitivity and specificity of frozen section diagnoses were 87,69 and 100, respectively. When the comparison between frozen and permanent section was analyzed according to the mammographic pattern, the sensitivity among patients with microcalcifications was lower (75,23) than among patients with opacities (93,86). When frozen section and permanent section diagnoses were related according to the mammographic size ( 10 mm) the sensitivity among patients with opacities measuring less than 10 mm was lower (91,75) than among patients with opacities larger than 10 mm (95,65) and the sensitivity among patients with microcalcifications larger than 10 mm was greater (77,14) than among patients with microcalcifications size less than 10 mm (74,28). This results are similar to those obtained on palpable breast lesions, and show that frozen section is a feasible and reliable procedure in nonpalpable breast lesions, particularly more relevant in mammographic opacities than in microcalcifications, whatever the lesion size.