Architectural Distortion

Tomosynthesis Features of Architectural Distortion

AD is defined as distorted parenchyma, often with spicules or tethered Cooper ligaments radiating from a common point with no obvious associated mass ( Fig. 10.1 ). AD can be associated with an asymmetry and/or microcalcifications. It is important to recognize that our understanding of AD to date has largely been based on 2D mammography, where complete assessment of lesions is more limited. With tomosynthesis, AD may be found to be associated with a subtle, low-density mass. Classic clues used to identify a subtle AD include focal retraction, distortion, and/or straightening of the tissue along the fat and parenchymal borders. These findings are usually more conspicuous on tomosynthesis, thereby increasing mammographic sensitivity and improving lesion characterization ( Fig. 10.2 ).

AD is usually isodense to fibroglandular breast tissue and can be difficult to recognize on 2D mammography in both dense and nondense breast tissue. Tomosynthesis can more clearly depict the radiating spicules, allowing for improved mammographic sensitivity and specificity in all breast densities ( Fig. 10.3 ).

When analyzing the tomosynthesis mammogram, it is essential to scroll slowly through the tomosynthesis stack, focusing only on a single area of the breast, as the AD may be best seen on a single view and only be apparent on one or two image slices. Once the potential AD is identified, scrolling back and forth just through the level of the suspected AD can help confirm that a true lesion exists. By determining the location of the tomosynthesis slice best demonstrating the AD, the orthogonal view can be carefully scrutinized in order to increase the likelihood of identifying a subtle AD on both the craniocaudal (CC) and mediolateral oblique (MLO) projections ( Fig. 10.4 ). In these cases, identification of nearby landmarks can also assist in locating a subtle AD on the orthogonal view.

Pathology of Architectural Distortion

When AD is identified, one must always first correlate the finding with the patient’s surgical history, as the most common cause of AD is previous surgery, secondary to either benign or malignant disease. If no surgical history is found, AD is considered a suspicious finding and malignancy must be excluded.

Malignant etiologies include invasive ductal (IDC) or lobular cancer (ILC). Given that IDC is more common, most malignant AD will represent IDC, including NOS (not otherwise specified) or well-differentiated invasive tubular carcinoma. ILC traditionally accounts for 5% to 10% of all invasive breast cancers and can be extremely subtle or occult on 2D mammography. Early studies indicate that the rate of ILC may be higher among suspicious AD seen only on tomosynthesis, thus potentially improving the rate of detection of this invasive subtype.

Although infrequent, ductal carcinoma in situ (DCIS) may also present as AD seen only on tomosynthesis. DCIS typically presents as suspicious linear, branching, and/or pleomorphic microcalcifications, but has been reported to present as AD on 2D mammography in 2% to 10% of DCIS cases. On tomosynthesis, DCIS may present as a subtle AD with variable density, with or without associated suspicious microcalcifications ( Fig. 10.5 ).

In addition to postoperative scars and invasive and/or in situ cancers, other histologies presenting as AD include high-risk radial sclerosing lesions, atypical ductal hyperplasia (ADH), atypical lobular hyperplasia (ALH), and lobular carcinoma in situ (LCIS). Stromal fibrosis, sclerosing adenosis, and rarely granular cell tumor or breast fibromatosis may also present as an AD seen only on tomosynthesis. Radial sclerosing lesions, composed of both radial scars and complex sclerosing lesions, may also present as AD on tomosynthesis and are indistinguishable from carcinoma ( Fig. 10.6 ). These lesions are histologically similar and are unrelated to trauma or prior surgery and may be secondary to a localized inflammatory reaction and chronic ischemia. Complex sclerosing lesions are distinguished from a radial scar by being more complex histologically and generally greater than 1 cm. Radial sclerosing lesions have been reported to be associated with atypia and/or malignancy, although the data has been primarily based on 2D mammography. The understanding of the association of malignancy with radial sclerosing lesions found on tomosynthesis is still evolving.