Screening with Other Imaging Modalities: Early Investigations

Ultrasound
Several investigators are evaluating the use of ultrasound to detect breast cancer that cannot be seen with mammography. Because ultrasound cannot reliably identify microcalcifications, which are mammographically evident in approximately one-half of breast carcinomas and are the sole method of diagnosis of the majority of ductal carcinoma in situ (DCIS) lesions, ultrasound cannot replace mammography in breast cancer screening. Some have suggested, however, that it may be used in addition to mammography in some groups of women, such as those at high risk who also have dense breasts.

Gordon and Goldenberg performed 12,706 whole-breast ultrasound examinations in women who had a palpable or mammographically evident mass (the “index lesion”). Ultrasound revealed 1,575 incidental solid masses, of which 279 underwent fine-needle aspiration (FNA) biopsy. Ultrasound detected 44 carcinomas that were not evident on mammography or physical examination. Cancer was found by ultrasound in 44 (16%) of the 279 solid masses that had biopsy, in 44 (3%) of the 1,575 solid masses that were found by ultrasound and in 44 (0.35%) of 12,706 screening ultrasound examinations. These 44 cancers were in 30 women, in whom the index lesion was benign in 15 and carcinoma in 15. The mean size of these sonographically detected cancers was 1.1 cm (median, 1.0 cm; range, 0.4-2.5 cm).

Kolb et al performed a study of 11,220 consecutive patients. Screening ultrasound examinations were performed in all 3,626 women with normal mammograms, normal physical exams and dense breasts.

They found 11 surgically proven cancers. Ultrasound revealed cancers not identified on mammography or physical examination in 11 (0.3%) of 3,626 women. The rate of cancer detection by ultrasound alone was six (0.58%) per 1,043 high-risk women and five (0.19%) per 2,583 normal-risk women. The size and stage of sonographically detected cancers were not significantly different from those of 61 nonpalpable, mammographically detected cancers and were smaller than the 64 palpable cancers diagnosed in the rest of their population.

In women with dense breasts, screening ultrasound increased breast cancer detection by 17% (from 63-74 tumors), and the number of tumors revealed only by imaging increased by 37% (from 30-41 tumors).

Additional work is needed to confirm and expand upon the work of these intriguing early studies. At the current time, however, ultrasound is not part of the standard of care in breast cancer screening.

Magnetic Resonance Imaging (MRI)
Magnetic resonance imaging (MRI) has a high sensitivity in the diagnosis of breast cancer, ranging from 86-100%.  Many studies have confirmed the ability of MRI to detect breast cancer not identified with mammography or physical examination. Slanetz et al performed bilateral breast MRI in 42 patients scheduled for biopsy of a mammographically visible or palpable abnormality, of which 17 proved to represent invasive cancer. Magnetic resonance imaging identified all 17 invasive cancers in the breast of concern. In addition, eight (19%) of 42 patients had 14 contralateral lesions found on MRI, of which nine (64%) were malignant. This is one of the first reports of breast MRI in an otherwise normal breast, and the authors suggested that their data support the use of breast MRI as a screening technique. There are several limitations with breast MRI as a screening method.

First, it has poor specificity, ranging from 37-97%.  Second, it is an expensive examination and requires injection of intravenous contrast. Third, the technology for performing biopsy under the guidance of MRI is not widely available, making it difficult to sample lesions identified only with MRI.

Refinement in the technology for MRI-guided localization and biopsy is essential if MRI is to be helpful as a screening technique. Currently, breast MRI is most useful in a problem-solving setting, and is not part of the standard of care for screening.


Laura Liberman and Timothy L. Feng
Breast cancer detection demonstration project: five-year summary report. CA 2003

References

Provided by ArmMed Media