From Medscape Medical News
Roxanne Nelson
March 15, 2010 — New findings suggest that mammography is of little value in young women with a familial risk for breast cancer who have access to quality-assured magnetic resonance imaging (MRI).
In a study published online February 22 in the Journal of Clinical Oncology, researchers found that using mammography, an ultrasound examination, or clinical breast examination does not increase the "cancer yield," compared with what is achieved with MRI alone.
The team was headed by Christiane Kuhl, MD, professor and vice chair of radiology at the University of Bonn in Germany, who has previously reported the superiority of MRI and is an advocate for this technology.
In this study, the researchers report that the cancer yield of ultrasound (6.0 of 1000) and mammography (5.4 of 1000) was equivalent, and increased nonsignificantly if both modalities were combined (7.7 of 1000). With MRI alone, the cancer yield was 14.9 of 1000, which is significantly higher than with either mammography or ultrasound.
In addition, the positive predictive value was 39% for mammography, 36% for ultrasound, and 48% for MRI.
Dr. Kuhl and colleagues suggest that, as a result, existing screening guidelines for young women at high and moderate risk for breast cancer — which currently recommend mammography — need to be revised.
This study adds to accumulating evidence that MRI is the most important screening modality for women with familial and genetic risks for breast cancer. However, for many reasons, mammography should not be abandoned just yet, according to an accompanying editorial.
Current Guidelines Not Based on Randomized Trials
Guidelines in both Europe and the United States recommend that women who face an increased risk for breast cancer begin mammographic screening at the age of 25 to 30 years, and should continue until age 70, Dr. Kuhl explained.
"It is important to realize that mammographic screening has been established by randomized controlled clinical trials only for women aged 40 and older — and many would argue only for women older than 50," Dr. Kuhl told Medscape Oncology. "Mammographic screening of women aged 25 to 39 has never been investigated or established by data from randomized controlled clinical trials."
She added that for women in this age range, existing guidelines are based on "expert opinion" only, with the experts being mainly oncologists, not radiologists. "But now, 10 years down the road from setting up these guidelines, we have evidence to suggest that mammography, if performed in addition to MRI, does not increase the cancer yield over what is achieved by MRI alone," she said.
Because the use of mammography is indicated in women older than 40 years, irrespective of risk, and its use is supported by randomized clinical trials, "we would stick to the recommendations to use mammography in addition to MRI in these high-risk women," said Dr. Kuhl.
But recent data do not support the use of screening mammography in women younger than 40 years who undergo quality-assured breast MRI, she added. "Since there are no data from randomized clinical trials that support the use of screening mammography for this age group anyway, we suggest the discontinuation of screening mammography in this age group," she said.
Changing Standards of Care?
In their paper, Dr. Kuhl and her colleagues provide a meta-analysis of the existing evidence, in the entire medical literature, on diagnosing breast cancer in young women. "The results clearly support our conclusion," she said.
But even though the data from the current trial, along with the meta-analysis, are clear and statistically highly significant, Dr. Kuhl noted that she still has "the following reservation when I am asked whether we should now change the existing standards of care."
She explained that, unlike studies that relate to treatment with a new drug, it is much more difficult to predict whether results of diagnostic trials will be completely reproducible in all global settings. This has an affect on the generalizability of results.
In a randomized trial comparing 2 different drugs, the difference between the success of treatment A and that of treatment B will most likely be transferable to all parts in the world, because drug A and drug B will be exactly the same, Dr. Kuhl explained. "This is different for diagnostic studies because there is a 'human factor' involved, and, at least for MRI, there are many different ways to image the breast."
"Therefore, to account for this variability of test performance that is true for all diagnostic studies, we chose the somewhat clumsy way to summarize our results as [being applicable to] 'women who have access to quality-assured breast MRI'," said Dr. Kuhl.
Higher Cancer Detection With MRI
In the current study, Dr. Kuhl and colleagues investigated the effectiveness of clinical breast examination, mammography, ultrasound, and quality-assured breast MRI, used alone or in combination, for screening women at elevated risk for breast cancer. The Evaluation of Imaging Methods for Secondary Prevention of Familial Breast Cancer (EVA) trial was conducted at 4 German academic breast centers and consisted of 687 women with a lifetime risk of developing breast cancer of 20% or greater.
The participants underwent 1679 screening rounds — which included annual MRI, mammography, ultrasound, and clinical breast examination — that were read independently and in different combinations. A subgroup of 371 women underwent 6-month screening ultrasound examinations and clinical breast exams. The mean follow-up was 29.18 months and the median follow-up was 29.09 months.
During the study period, 27 women (3.9%) were diagnosed with breast cancer. Of these, 16 (59%) were invasive cancers and 11 (41%) were ductal carcinoma in situ (DCIS). All of the cancers were detected during the annual screenings, and none were interval cancers. The mean age at diagnosis was 43.1 years.
The majority of women (n = 21; 77%) were diagnosed with minimal cancers, and 9 of the cancers occurred in women with a history of the disease (7 were contralateral or second primary cancers and 2 were local recurrences).
The researchers found that the sensitivity achieved by ultrasound alone (37%) and by mammography alone (33%) was comparable (P = .72). Combining mammography and ultrasound yielded a slightly higher sensitivity, but it was not significant (48%; P < .12).
However, MRI used alone was significantly more sensitive (93%) than either mammography or ultrasound alone or in combination (P < .005). The addition of mammography to MRI did not result in a statistically significant increase in sensitivity (P = .5).
Overall, 2 cancers were diagnosed with mammography alone (7%) and none were diagnosed with ultrasound alone, but 14 were diagnosed by MRI alone. Clinical breast examination was positive in 110 screening rounds, but only 1 palpable mass corresponded to a cancer diagnosis. The remaining cancers diagnosed during the study period were all clinically occult.
Contribution to Accumulating Evidence
The results of the EVA trial contribute to "the accumulating evidence that MRI is not only the most important screening modality in gene-mutation carriers, but also in women with a familial risk without a documented BRCA1/2 mutation," notes Jan G.M. Klijn, MD, in an accompanying editorial.
"But for a number of reasons, many radiologists and oncologists think that it is still too early to abandon mammography," writes Dr. Klijn, who is from Erasmus University in Rotterdam, the Netherlands. One reason is that in other large studies, combined mammography and MRI were significantly superior to either mode of screening alone.
Another reason is that "the experience of MRI is variable among centers participating in trials and certainly outside trials mandating a learning curve," he writes, adding that future studies will be needed to confirm the very high MRI sensitivities and specificities for both the invasive cancers and DCISs that were reported in this paper, using comparable technology and quality-controlled protocols.
The study was supported by a grant from the German Cancer Aid Society. None of the authors nor the editorialist have disclosed any relevant financial relationships.
J Clin Oncol. Published online February 22, 2010. Abstract, Abstract
No comments:
Post a Comment