Objective: We recently showed that frozen section diagnosis has an almost perfect specificity for the diagnosis of malignancy in an adnexal mass, whereas the sensitivity was, though good, not perfect. The aim of the present study was to examine whether the accuracy of frozen section diagnosis is affected by the size of the adnexal masses.
Methods: We included women that underwent frozen section diagnosis for adnexal mass surgery. For each patient we recorded serum CA125 level, menopausal status, tumour size, and histologic classification both at frozen section diagnosis and at definite histological examination. We calculated sensitivity and specificity of frozen section diagnoses, both in tumours below and above 10 cm diameter.
Results: We included 257 patients, of whom 142 had a benign tumour, 28 had a borderline tumour and 87 had a malignant tumour at definitive histological assessment. In case frozen section diagnosis showed malignancy, this was always confirmed at final histological assessment. In women with a tumour <10 cm, there was only 1 false negative diagnosis in 50 women with a benign frozen section diagnosis, whereas there were 11 false negative diagnoses in 97 women in women with a tumour > or = 10 cm. The corresponding likelihood ratios of a benign diagnosis for presence of malignancy were .15 for tumours > or = 10 cm and 0.03 for tumours <10 cm, respectively.
Conclusion: The accuracy of frozen section diagnosis is dependent on tumour size. In adnexal masses > or = 10 cm, a benign result of the frozen section diagnosis is less reliable than in women with a tumour <10 in cm.