Mammographically indeterminate microcalcifications--can we do any better?

In the first round of the mammographic screening program of the South Australian Breast X-ray Service, 80 (0.73%) of the first 10,848 women screened over 18 months were referred after primary and second-stage assessment for definitive histology because of microcalcifications. Obvious mass lesions associated with calcification were excluded from this study, as were women whose calcification was regarded as sufficiently benign to warrant routine rescreening in two years. After classic patterns of malignant microcalcification were excluded, a large group (75%) remained, whose calcifications fell into the indeterminate grades of radiological suspicion. Of these, only 15% proved to have cancer, and in one third of these the cancer was mammographically occult. A high rate of discordant readings was noted in lesions which ultimately proved benign. Neither family history, distribution of calcification nor the presence of a faint soft tissue density proved to be unfailingly reliable predictors of benign or malignant histology in this group. Vigorous pursuit of histopathological correlation and performance statistics are urged to monitor and minimise the proportion of women who remain in this indeterminate group and to follow their natural history. Current mammographic techniques are still inadequate for the provision of definitive information on microcalcification in all cases, but a sustained commitment will reduce the number proceeding to histology for benign disease.