If used with appropriate diagnostic criteria, MRI is much more sensitive than mammography for detecting breast cancers before they have developed to an invasive stage, and particularly good at identifying those lesions which are more likely to progress to dangerous forms of cancer, according to a study published in The Lancet. This finding is in stark contrast to previous studies comparing the two techniques that concluded MRI can not detect early cancers as effectively as mammography.
Debates over which screening modality is best for detecting ductal carcinoma in situ (DCIS)—an early type of breast cancer in which cancerous cells are present but are confined to the milk ducts within the breasts—have in the past focused on identification of microcalcifications, small areas where cells have died and subsequently turned into calcium deposits, which are used as a hallmark of cancerous growth. Mammography can visualise these features in a breast scan whereas MRI cannot.
However, scientists have recently found that MRI can detect another feature of DCIS—the growth of new blood vessels around the cancerous cells—by virtue of the fact that the contrast agent (an injected chemical used to enhance differences between different areas or tissues under inspection) will be seen outside its usual domain of the intravascular space in places where the vessel wall is corrupted by new vessel growth.
What is more, MRI seems to be able to distinguish between subsets of DCIS which are high-grade and low-grade—classifications that refer to the likelihood of these lesions progressing to invasive breast cancer, in which the cancerous cells break out of the ducts and invade surrounding breast tissue. In high-grade DCIS, the density of small new blood vessels is high, showing up as a higher level of enhancement on MRI images, more so than in low grade DCIS, making the former easier to spot.
Using this new observation of MRI’s DCIS identification potential, Christiane Kuhl and colleagues from the University of Bonn’s radiology department, which is associated with an academic national breast centre, compared the technique with mammography and assessed the two screening modalities’ effectiveness at diagnosing DCIS, which is treated surgically to remove the risk that it will develop into invasive cancer. "We investigated the sensitivity of each method of detection and compared the biological profiles of mammography-diagnosed DCIS versus DCIS detected by MRI alone," explain the authors.
The researchers identified their study sample from the 7319 women who were referred to their breast centre over a 5 year period and received MRI in addition to mammography for diagnostic assessment and screening. The final study cohort consisted of the 167 patients who had undergone both mammography and MRI before biopsy and who had received the final surgical pathology diagnosis of pure DCIS.
Reviewing the method through which the final diagnosis had been made, the researchers found that 93 (56%) cases of DCIS were diagnosed by mammography and 153 (92%) by MRI. Of the 89 cases of high-grade DCIS, 43 (48%) were missed by mammography, but diagnosed by MRI alone and all 43 of the cases missed by mammography were detected by MRI. Overall, MRI detected significantly more cases of any grade of DCIS than did mammography and produced a lower proportion of false negatives: mammography was falsely negative in 74 (44%) cases, whereas MRI was falsely negative in 14 (8%) cases.
"Our study suggests that the sensitivity of film screen or digital mammography for diagnosing DCIS is limited," write the authors. "Mammography tends to identify breast cancers with comparatively benign biological profiles and although more low-grade DCIS will be diagnosed if MRI is used in addition to mammography, and thus more women could be overdiagnosed with a possibly prognostically irrelevant disease, one should note that 60% of the cases of DCIS diagnosed by MRI alone were high grade." There is, therefore, the researchers conclude, reason to assume that MRI helps anticipate the diagnosis of lesions that, if left undetected, would progress to invasive breast cancer.
However, the researchers caution that since breast MRI is currently used only rarely in clinical practice, their results are not really applicable to a general screening setting because few radiologists can offer a level of expertise for MRI that comes close to that required for diagnostic mammography. Nevertheless, they firmly conclude that current recommendations regarding the use of MRI for screening for DCIS are not appropriate.
In an accompanying editorial, Carla Boetes and Ritse Mann from Radboud University Nijmegen Medical Centre, note that the observation that "MRI detects many DCIS lesions that go unnoticed on mammography implies that some invasive carcinomas can be prevented by timely intervention on the basis of MRI findings. As such, MRI has the potential to increase survival when used to detect breast cancer."
MRI for diagnosis of pure ductal carcinoma in situ: a prospective observational study. Kuhl CK, Schrading S, Bieling HB, Wardelmann E, Leutner CC, Koenig R, Kuhn W, Schild HH. Lancet 2007; 370: 485-92