It appears that for the average woman with newly diagnosed breast cancer, breast MRI may cause more harm than good, due to overdiagnosis leading to unnecessary ipsilateral and contralateral surgery.
Daniel F. Hayes, MD, FASCO
“New!” “Improved!” “Throw out that old [fill in the blank] and go buy a new [fill in the blank]!” Sound familiar? The key to marketing is to convince customers that they need a product without which they had previously been quite happy. All too often, this strategy is accompanied by a caveat emptor attitude—Let the buyer beware.
Introduction of new diagnostics into medicine differs remarkably from that of therapeutics. The latter is carefully controlled by the U.S. Food and Drug Administration (FDA), with a statutory mandate that a new therapeutic drug be both safe and effective. Although the definition of these two terms may be, and frequently is, debatable, introduction of a new therapeutic agent requires some form of prospective trial in which these two criteria are investigated in comparison to a control group.
In contrast, new diagnostic techniques, whether they are based on imaging or in vitro assays, are evaluated as devices by the FDA. Our colleagues at the agency do a spectacular job determining the analytic accuracy, reproducibility, and reliability of the devices under evaluation, but they are not charged with the heavier burden of determining whether use of the diagnostic device actually improves patient outcomes.
Commenting on this regulatory inconsistency is more than just a wonky academic exercise. The relatively lax requirements for introduction of a new device into practice, compared with a new therapeutic drug, have contrasting implications: (1) faster speed and lower cost of implementation and (2) less confidence that they have clinical utility for patient care. Whereas the clinical trials necessary to demonstrate that a new drug is safe and effective may take a decade or more and may cost half a billion or more dollars, introduction of a new diagnostic device fundamentally requires evidence that it does, indeed, detect what the manufacturer says. Although premarket approval studies are necessary to establish the sensitivity and specificity of the diagnosis in question, they can be performed much quicker and less expensively, since improvement in a meaningful outcome, such as overall survival or quality of life, does not need to be established, although it is often implied.
This situation may result in the rapid introduction of lifesaving new diagnostics. No one would argue that we need prospective randomized clinical trials to determine whether chest x rays or body scanning helps patient care. However, lack of such evidence may just as likely be the cause of mistreatment due to assumed, but unreal, benefits of the results of the test. The patient may be undertreated due to failure to diagnose a condition that would have otherwise been treated. More commonly, use of unproven diagnostics may lead to overdiagnosis and overtreatment due to detection of a cancer that, if left alone or treated with existing standard strategies, would never have been clinically important. In other words, these findings are biologic false-positives.
The Downside of MRI
Introduction of magnetic resonance imaging (MRI) for evaluation of women with already diagnosed breast cancer is a case study of this conundrum. Multiple prospective randomized clinical trials, with decades of follow-up, have demonstrated that more than one-half of women with newly diagnosed early-stage breast cancer, when carefully evaluated by standard mammographic and pathologic techniques, can be safely treated with breast-conserving therapy with limited surgery and adjuvant radiation—arguably one of the great advances in breast cancer treatment of the 20th century.1-5
However, with the advent of MRI technology, an alarming number of women for whom breast-conserving therapy would have been considered appropriate treatment have been told they should have more aggressive, and cosmetically disfiguring, surgery (often mastectomy). Why? Because MRI is more sensitive than mammography and detects deposits of previously unrecognized cancers in up to 15% of patients who were considered good candidates for breast-conserving therapy.6 Human nature being what it is, detection of something called “cancer” usually leads to doing something one would not have done otherwise, and therefore these women and their doctors have elected to have more surgery.
Does this increased treatment do these patients any good? It does not appear to. In the COMICE trial, patients who were considered candidates for breast-conserving therapy were randomly assigned to MRI evaluation or usual standard investigation.7,8 Not only did preoperative MRI not decrease the need for reexcision (the hypothesis was that preoperative MRI would lead to more extensive first surgery, mitigating the need for reexcision), women who were assigned to the MRI arm had more extensive surgery and mastectomies, yet local recurrence rates (and overall survival) are, so far, identical.
This result comes as no surprise. Roland Holland published a landmark paper in the 1980s demonstrating that, when “bread-loafed” and carefully examined, mastectomy specimens contained preoperatively unrecognized cancer deposits centimeters away from the primary tumors in up to 15% of patients.9 Further, in several trials of breast-preserving surgery with or without adjuvant radiation, the risk of subsequent in-breast recurrence is close to 50% in those who do not receive radiation but far less than 10% in those who do.10 Spoiler alert: Radiation works! It appears that all MRI does is tell us what we already knew—Most women undergoing breast-conserving therapy should receive adjuvant radiation, not more surgery.
Nonetheless, MRI has been introduced into standard breast cancer management in many centers over the past 15 years. Indeed, many centers marketed that they were more up-to-date than other centers because they used MRI to evaluate patients considered to be candidates for breast-conserving therapy. This claim was made in the absence of any prospective data, or retrospective data for that matter, that even hinted patients were better off. Indeed, as I’ve argued here, it appears they were more poorly treated by being subjected to unnecessary, cosmetically disfiguring surgery!
Contralateral Breast Cancer and MRI
A secondary effect of MRI evaluation of the affected breast is inclusion of the contralateral, unaffected breast in the imaging. Several studies have demonstrated a small, but real, detection rate above and beyond that for mammography.11 Again, these observations have led to the assumption that, therefore, patients must gain benefit by virtue of finding an occult cancer early. However, again, there are no data to suggest that these findings have led to improved patient survival, and they have certainly led to more surgery, including bilateral mastectomies.
Whereas there is no question that women with one breast cancer are at slightly higher risk over the subsequent 10 to 20 years of developing another, this risk is no more than approximately 0.5% to 1.0%/year, and very few women die of contralateral breast cancer.12 The recent article by Wang and colleagues in the Journal of Clinical Oncology, reviewed in this issue of The ASCO Post, appears to support this concept.13
By examining the use of MRI in women between the ages of 67 and 94 in the SEER (Surveillance, Epidemiology, and End Results) database, they confirmed that those patients who had preoperative MRI were roughly twice as likely to have contralateral breast cancer at the time, or within the succeeding 6 months, of their first diagnosis. Not surprisingly, those who had MRI had a slightly lower rate of subsequent contralateral breast cancer over the following 5 years, although their cumulative contralateral breast cancer risk remained higher overall (synchronous plus subsequent incidence). The authors implied that many of these previously unappreciated contralateral breast cancers represented overdiagnosis.
In all fairness, the risks of subsequent metastases and breast cancer–specific death were not reported, so it is impossible to determine whether detection of these otherwise unappreciated contralateral breast cancers led to better long-term meaningful outcomes. However, the authors did report subsequent “stage II–IV” events, which over the 5-year follow-up were quite low in all patients regardless of MRI or not, suggesting it is unlikely that detection of the contralateral breast cancer with MRI was of clinical importance.
Breast MRI: More Harm Than Good?
So, should we order breast MRI evaluation in women with newly diagnosed breast cancers? There are some uncommon situations in which the answer is “yes”— for women with axillary presentation with an undetectable (by physical or mammographic examination) breast primary, and for women with either very dense breasts or a very high genetic risk.14 However, it appears that for the average woman with newly diagnosed breast cancer, breast MRI may cause more harm than good, due to overdiagnosis leading to unnecessary ipsilateral and contralateral surgery.
How could this situation arise? I maintain it is due to the inconsistent regulatory environment in the United States, in which drugs are reviewed carefully for safety and efficacy, but in which devices, both diagnostic and therapeutic, are only reviewed for accuracy and some clinical association. This environment leads to a caveat emptor atmosphere for the use of devices to care for patients with cancer. Do we really believe that caveat emptor is good medical practice? Maybe it is for purchasing a new car but not for determination of patient care.
Call for Consistent Product Review
My colleagues and I have called for revision of the Oncology Drug Advisory Committee (ODAC) of the FDA to become an “Oncology Products Advisory Committee.”15 More recently, the advocacy organization Friends of Cancer Research, led by Ellen Sigal, proposed that the FDA reorganize into an “Oncology Center of Excellence.” I totally support this recommendation. Such a move would result in application of similar criteria for review of both devices and drugs, with corporate memory among the regulators so that all oncology products are reviewed consistently.
A bad diagnostic is as bad as a bad drug. If we are going to use diagnostic devices, be they radiologic or in vitro tumor biomarker tests, to direct patient care, we need to have the same confidence in them that we do for our antineoplastic therapies. We owe this to our patients, who trust in our expertise and judgment to ensure that their care is truly “personalized.” ■
Disclosure: Dr. Hayes owns stock in Oncimmune and Inbiomotion; is on the advisory board for Pfizer Global; is a consultant for Lilly Oncology; received research grants from Merrimack Pharmaceuticals, Inc, Eli Lilly Company, Janssen R&D, Puma Biotechnology, Inc, Pfizer, and AstraZeneca; and receives royalties from Janssen R&D.
1. Fisher B, Anderson S, Bryant J, et al: Twenty-year follow-up of a randomized trial comparing total mastectomy, lumpectomy, and lumpectomy plus irradiation for the treatment of invasive breast cancer. N Engl J Med 347:1233-1241, 2002.
2. Veronesi U, Cascinelli N, Mariani L, et al: Twenty-year follow-up of a randomized study comparing breast-conserving surgery with radical mastectomy for early breast cancer. N Engl J Med 347:1227-1232, 2002.
3. Lichter AS, Lippman ME, Danforth DN Jr, et al: Mastectomy versus breast-conserving therapy in the treatment of stage I and II carcinoma of the breast: A randomized trial at the National Cancer Institute. J Clin Oncol 10:976-983, 1992.
4. Arriagada R, Lê MG, Rochard F, et al: Conservative treatment versus mastectomy in early breast cancer: Patterns of failure with 15 years of follow-up data. Institut Gustave-Roussy Breast Cancer Group. J Clin Oncol 14:1558-1564, 1996.
5. Blichert-Toft M, Rose C, Andersen JA, et al: Danish randomized trial comparing breast conservation therapy with mastectomy: Six years of life-table analysis. Danish Breast Cancer Coopertive Group. J Natl Cancer Inst Monogr 11:19-25, 1992.
6. Houssami N, Hayes DF: Review of preoperative magnetic resonance imaging (MRI) in breast cancer: Should MRI be performed on all women with newly diagnosed, early stage breast cancer? CA Cancer J Clin 59:290-302, 2009.
7. Turnbull L, Brown S, Harvey I, et al: Comparative effectiveness of MRI in breast cancer (COMICE) trial: A randomised controlled trial. Lancet 375:563-571, 2010.
8. Turnbull LW, Brown SR, Olivier C, et al: Multicentre randomised controlled trial examining the cost-effectiveness of contrast-enhanced high field magnetic resonance imaging in women with primary breast cancer scheduled for wide local excision (COMICE). Health Technol Assess 14:1-182, 2010.
9. Holland R, Veling SH, Mravunac M, et al: Histologic multifocality of Tis, T1-2 breast carcinomas: Implications for clinical trials of breast-conserving surgery. Cancer 56:979-990, 1985.
10. Recht A, Hayes DF, Eberlein TJ, et al: Local-regional recurrence after mastectomy or breast-conserving therapy, in Harris JR, Lippman M, Morrow M, et al (eds): Diseases of the Breast, pp 649-667. Philadelphia, PA, JB Lippincott, 1996.
11. Lehman CD, Gatsonis C, Kuhl CK, et al: MRI evaluation of the contralateral breast in women with recently diagnosed breast cancer. N Engl J Med 356:1295-1303, 2007.
12. Ji J, Hemminki K: Risk for contralateral breast cancers in a population covered by mammography: Effects of family history, age at diagnosis and histology. Breast Cancer Res Treat 105:229-236, 2007.
13. Wang SY, Long JB, Killelea BK, et al: Preoperative breast magnetic resonance imaging and contralateral breast cancer occurence among older women with breast cancer. J Clin Oncol 34:321-328, 2016.
14. Saslow D, Boetes C, Burke W, et al: American Cancer Society guidelines for breast screening with MRI as an adjunct to mammography. CA Cancer J Clin 57:75-89, 2007.
15. Hayes DF, Allen J, Compton C, et al: Breaking a vicious cycle. Sci Transl Med 5:196cm6, 2013.
In an analysis of the Surveillance, Epidemiology, and End Results (SEER)-Medicare database reported in the Journal of Clinical Oncology, Shi-Yi Wang, MD, PhD, of Yale School of Public Health, and colleagues found that the use of preoperative magnetic resonance imaging (MRI) was associated with...