Mammography’s Midlife Crisis

Posted on March 23, 2013 by Handel Reynolds

Mammography turns 40 this year. It was in 1973 that the American Cancer Society and The National Cancer Institute jointly launched the Breast Cancer Detection Demonstration Project. This was a nation-wide screening program involving 280, 000 healthy women between the ages of thirty-five and seventy-four who were invited to undergo annual breast cancer screening with mammography and physical examination. It was the first time most Americans had ever heard of mammography and it marked the beginning of our national effort to eradicate breast cancer through early detection. As the beleaguered test hits the “Big 4-0”, it finds itself in the throes of an all-out midlife crisis; its relevance questioned, its impact unclear, its future uncertain.

As the beleaguered test hits the “Big 4-0”, it finds itself in the throes of an all-out midlife crisis; its relevance questioned, its impact unclear, its future uncertain.

It was very different at the start. In 1973, mammography was surrounded by an aura of unbridled hope and optimism. After all, it was a mere twenty years earlier that the American Cancer Society had played a major role in the nation-wide implementation of another screening test. The Pap smear proved well accepted by the public and highly successful in reducing cervical cancer death rates. This experience would later inform the organization’s approach to mammography. The mammogram, it was hoped, would be breast cancer’s Pap smear. It was a dream, however, that would not be realized. Cervical cancer screening’s largely tranquil history stands in stark contrast to the controversy and acrimony that have so characterized mammography’s American journey.

It is notable that mammography has never been a particularly controversial test in other developed countries. This bifurcated history can be traced to two crucial early decisions made by mammography’s backers. The first concerns the appropriate age demographic for screening. In Europe and other developed countries with active screening programs, the test is generally accepted to be most appropriate for postmenopausal women. In such places, it is recommended for women starting at age fifty. In the United States, mammography’s primary proponents, the American College of Radiology and the American Cancer Society, adopted a position that women under fifty should be screened with mammography as far back as the 1970s. At that time, there was absolutely no scientific evidence that mammography benefitted these younger women. Rather than evidence, this decision was based on hope; hope that someday the desired evidence would be uncovered. This one question: whether women younger than fifty should be screened, has dogged mammography incessantly, residing at the heart of almost every major eruption of mammography controversy since its inception. It has been mammography’s dispute that will not die.

The second, and more consequential, decision made by mammography’s proponents, regards the manner in which the test would be presented to the public. Throughout its history, mammography has been promoted using simple, one-sided messages; a “just do it” approach that presents only the potential benefits with no mention of possible harms or limitations. This course, designed to minimize confusion and maximize compliance, has been successful in achieving high rates of mammography utilization. It has contributed nothing to true public/consumer education.

Routine screening with mammography reduces a woman’s likelihood of dying of breast cancer. This is true for women in their forties, fifties, sixties and possibly beyond. The magnitude of this benefit, however, increases with age. Thus, older women derive a greater degree of risk reduction than younger women. At the same time, mammography is associated with certain potential harms. Chief among these are false positive readings, where the mammogram shows a suspicious finding requiring further imaging or biopsy, only to be proven to be non-cancerous; false negatives, where breast cancer is missed, often due to dense breast tissue; and overdiagnosis, a scenario where a cancer that would not have otherwise come to light in the woman’s lifetime, is detected. While estimates vary, up to a third of mammography-detected cancers may fall in this category. Thus, paradoxically, while mammography may decrease a woman’s likelihood of dying of breast cancer, it increases her chances of being diagnosed with the disease. Some of mammography’s harms, for example false positive readings, are more common among women in their forties. In this way, screening involves a harm/benefit calculus; one that becomes more favorable to screening as women age.

Despite its failings, mammography is, and for the foreseeable future, will continue to be, an important component of women’s healthcare. The test has undoubtedly played a role in the documented decline in breast cancer mortality rates over the past twenty years. Furthermore, first-dollar insurance coverage is mandated under the Affordable Care Act, guaranteeing barrier-free access. However, as the test enters its fifth decade, it is time to abandon the paternalistic “just do it” approach to mammography promotion in favor of true informed decision-making; a fresh paradigm where women are given all the facts about mammography: its benefits and potential harms, and are empowered to make their own choices. Modern women deserve no less.

H. Gilbert Welch Endorses The Big Squeeze in LA Times Op-Ed

Posted on March 18, 2013 by Handel Reynolds

In case you missed it, here is a great Op-Ed by Dr. H. Gilbert Welch published recently in the Los Angeles Times. In it, he strongly endorses The Big Squeeze.

Follow the Money Trail in 3D Mammography Research

Posted on January 17, 2013 by Handel Reynolds

Last week, in one of the largest studies of 3-D mammography (also referred to as breast tomosynthesis) to date, Norwegian researchers reported on an investigation involving almost 13,000 women. These women underwent standard (2D) mammography plus tomosynthesis (3D). The studies were read independently of each other by four radiologists. They found that the 3D technique boosted the detection of invasive breast cancer by 40% and reduced false positive readings (see below) by 15% over standard mammography alone. This study adds to a growing body of research showing superiority of the combined 3D/2D mammogram over standard 2D mammography.

In a standard mammogram, each time the breast is compressed and an exposure is made, a single image is obtained. This image displays all of the breast tissue that is in the volume of compressed breast: from the skin surface in contact with the clear plastic compression paddle to the opposite skin surface in contact with the rigid breast support device. Thus, the resulting single image represents a composite or summation of all the structures within the breast. This leads to two of mammography’s major limitations. First, overlapping normal structures may simulate a tumor, resulting in a recommendation for a return visit to the mammography facility for additional testing. When no cancer is ultimately found, this is called a false positive (false alarm). Second, and more importantly, this overlap of normal tissue may obscure a breast cancer, resulting in a false negative (missed cancer) reading. 3-D mammography is different in that each time the breast is compressed, instead of one image, numerous images (up to 100 or more), each representing a 1-mm thick slice of breast tissue, are produced. This markedly reduces the amount of overlapping tissue and, it is hoped, the number of false positives and false negatives.

Hologic Inc., based in Bedford, Mass, became the first company to receive US Food and Drug Administration approval for breast tomosynthesis in 2011. At this writing, it remains the sole FDA-approved vendor of this technology. Currently, the added costs associated with tomosynthesis are not covered by Medicare or most private health insurance plans. Thus, women who undergo the test typically incur some out-of-pocket expenses. This, coupled with the high cost of the equipment ($700,000), has tended to restrain the adoption of the technology.

Not surprisingly, the publication of this new study provided Hologic a major boost. Its stock price immediately rose to a 52 week high. In fact, the publication of favorable tomosynthesis studies is crucial to the company’s long-term strategy for tomosynthesis. In order to petition for Medicare and private insurance coverage for the new technology, there has to be a certain critical mass of supportive studies. Without insurance coverage, hospitals will continue to balk at the high price tag of these machines. This explains why many of the major studies published recently were conducted with significant Hologic financial support. For example, in the current Norwegian study, of the thirteen listed authors of the paper, two are actual employees of the company; an additional eight received financial compensation from Hologic for the time spent working on the study and the company provided the tomosynthesis equipment free of charge. In another propitious tomosynthesis study published recently, of seven authors, one is an employee of Hologic; another is a Hologic statistician and consultant; the lead author has lectured on the benefits of the technology at many medical symposia and her academic career has been bolstered by Hologic research funding. Finally, the company provided financial grants to cover the costs of the equipment and personnel time.

While industry involvement in medical research is not new and these relationships were disclosed in the articles, the issue has rightfully been receiving greater scrutiny over the past few years. 3D mammography may indeed be superior to standard mammography but the public would be well served by a generous measure of skepticism until results from large studies, free of this degree of corporate involvement, are available.

Breast MRI in Newly Diagnosed Breast Cancer Patients: A Bridge Too Far?

Posted on December 20, 2012 by Handel Reynolds

It is becoming common practice for surgeons and radiologists to recommend preoperative breast MRI for women diagnosed with breast cancer. The trend has occurred largely without strong evidence of benefit from this additional testing. The suggested rationale is as follows. MRI provides a much more detailed picture of the tumor, including its true size and its relationship to surrounding structures, than does mammography or ultrasound. For women undergoing lumpectomy, this added information should make it more likely that the surgeon will remove the entire mass the first time, thereby limiting the number of women who need additional operations for complete clearance.

A new study in the Annals of Surgery casts serious doubt on the validity of these widely held assumptions. In it, the authors pooled the results of nine previous studies of preoperative MRI in breast cancer patients. The combined dataset included over three thousand women. They found that there was no difference in the re-operation rate for women who had undergone MRI (11.6%) versus those who had not (11.4%). In other words, both groups were equally likely to need additional surgery to achieve clear surgical margins (complete tumor removal). Even more disconcerting was the finding that women who had MRI were much more likely to undergo a mastectomy (25%) than those who did not (18%). There are at least two reasons for this difference. First, MRI finds additional sites of cancer in the diseased breast in about 16% of women. Many of these women ultimately undergo mastectomy. While at first glance this may seem quite reasonable, it should be noted that women who undergo lumpectomy typically also receive radiation therapy to the affected breast. The purpose of this radiation is to kill any other sites of undiagnosed cancer that may be present. This has been common practice for decades and it is highly effective. The chance of breast cancer recurring in the treated breast after lumpectomy and radiation therapy is about 7% at twelve years. Thus, there is already a highly reliable method for eradicating additional sites of cancer in the affected breast; one that does not require removal of the entire breast. Second, MRI generates large numbers of false positives (false alarms). Quite frequently, “bright spots” appear on MRI that, after biopsy, are found to be completely harmless. In my experience, many women are so frightened by these false alarms that they choose mastectomy out of a desperate need for peace of mind.

The authors conclude that, given the lack of proven benefit and the demonstrable harm (increased mastectomies) associated with preoperative MRI, its “routine use … in the breast cancer patient population as a whole should be abandoned.” In medicine, as in life, sometimes less truly is more.

Breast Cancer Overdiagnosis and Mammography

Posted on November 28, 2012 by Handel Reynolds

With all the recent attention being paid to the subject of mammography – driven breast cancer overdiagnosis, here is a good primer on the subject. It is an OpEd of mine, published on August 8, 2012 on Bloomberg View, the opinion platform of Bloomberg News. It is an excerpt from chapter 8 of The Big Squeeze. I welcome your questions and comments.