True extent of calcifying DCIS components of invasive breast cancer: radiological-pathological correlation in slice radiograms of breast conserving therapy specimens

Aim:

Tumor extent in breast conserving therapy (BCT) specimens is a major risk factor for local recurrence in invasive breast cancer with extratumoral ductal in situ carcinoma (DCIS) components. Surgical strategies mainly depend on extent of tumor and microcalcifications in mammograms, but since not all ducts containing in situ carcinoma bear microcalcifications, final extent of DCIS may differ from the radiologically suspected volume. Our aim was to compare radiological and pathological extent of DCIS associated with invasive cancer in BCT specimens.

Methods:

The extent of microcalcifications in 265 slice radiograms of 20 BCT specimens was measured by two radiologists and correlated with histological DCIS size in mega- or jumbo-slides of whole slices.

Results:

151 slices contained microcalcifications, 108 slices contained DCIS. Positive predictive value for presence of DCIS in slice radiograms containing suspicious microcalcifications was 0.63, negative predictive value for absence of DCIS in slice radiogram without microcalcifications was 0.88. There was a low positive correlation between radiological extent of microcalcification and final pathological extent of DCIS (R = 0.18, p = 0.057).

Slice radiography overestimated DCIS extent in 55% (60/108) and underestimated DCIS extent in 29% (31/108) of slices containing microcalcifications. Deviations of ≤2 mm between DCIS and microcalcification extent were found in only 16% (17/108 slices).

Conclusions:

Presence or absence of microcalcifications in slice radiograms of BCT specimens did not reliably predict existence or absence of DCIS. Extratumoral microcalcifications in invasive breast cancer should prompt thorough specimen sampling in and beyond calcified areas including radiological-pathological correlation to ensure accurate disease extent measurement.