Scientists at the National Cancer Institute recognized the limitations of the mammography screening trial conducted during 1963-1969 by the Health Insurance Plan of Greater New York (HIP-GNY). Those limits prevented the trial from supporting a widespread screening protocol for breast cancer among asymptomatic women of average risk. (See previous post, “Mammography’s Shadows, II: Numbers and the First Large X-Ray Screening Trial.) But their concerns were overridden by media inflated fears of breast cancer, and the organizations whose members stood to gain from widespread breast cancer screening.
For the careers and institutions committed to the understanding, diagnosis, and treatment of cancer, 1971 was an annus mirabilis. That year the Journal of the American Medical Association published a report of the HIP-GNY breast cancer screening study that purportedly proved that early detection with mammography screening “saves lives.”
It was also the year that Congress passed, and President Richard Nixon signed, the National Cancer Act of 1971, which had strong bi-partisan support. Provisions in the legislation ensured that the NCI, established in 1937, would become a virtually autonomous unit of the National Institutes of Health (NIH), First, the NCI director would be appointed by the president. Then the institute would be able to submit its own budget to the White House, and have its own independent procurement authority, which meant that it could award grants and contracts without NIH review. These provisions signaled the success of a masterful maneuver for those who lobbied aggressively for the legislation, among them, the American Cancer Society (ACS).
The influence the American Cancer Society could wield over the National Cancer Institute was foretold by President Nixon himself, who nominally originated the legislation with a message to congress earlier in 1971. During the bill’s White House signing ceremony Nixon invited to the dais only two (non-congressional) dignitaries: chair of the President’s Cancer Panel Benno Schmidt, a lawyer and wealthy partner in the New York investment firm J.H. Whitney & Co., and Dr. Alva H. Letton, physician and president of the American Cancer Society. Each man received one of the two pens Nixon used to sign the act. Nixon invited Letton to speak, which he did, briefly, commending the new War on Cancer as “the greatest thing ever done by the United States.”[1]
Shortly thereafter the American Cancer Society planned its own campaign in the War on Cancer. The campaign had two leading salients–a national membership campaign, and a marketing campaign for routine screening mammography. The mammography
screening campaign would be named the national Breast Cancer Detection Demonstration Project (BCDDP), thereby implying that the detection method being used had been proven to be both safe and effective at defeating breast cancer.
Since its lobbying had resulted in a substantial increase in the NCI’s budget, the American Cancer Society turned to the NCI for funding for its project to recruit over a quarter-million women for annual screening mammography. The society’s objective was to “search out people who have undetected cancer” by convincing the public that “early detection saves lives.”[2] Like all large screening programs, the campaign would increase revenue streams not only for physicians and clinics, but for x-ray equipment manufacturers and pharmaceutical companies. [3]
The National Cancer Institute’s own civil service scientists undoubtedly knew at the time that the carcinogenic effects of x-rays had been demonstrated before World War II, and confirmed thereafter by the National Academy of Sciences’ panels on the biological effects of exposure to ionizing radiation. Moreover, the probability that physical compression of, and piercing or cutting into, breast tumor cells might cause metastasis, could have been readily suspected. In any event, physicians had been warned of that possibility as early as 1958.[4] Questioning the validity of the HIP-GNY study, which was cited as having proven the value of screening mammography, scientists at the National Cancer Institute opposed NCI support of the American Cancer Society’s screening mammography campaign.
Most vocal among them was NCI’s John C. Bailar, III, a noted epidemiologist who had also trained in medicine at the Yale University of Medicine before turning to statistics. As NCI’s deputy associate director for cancer control he reviewed the ACS’s proposed program. Bailar was soon troubled not only by the seemingly indiscriminate exposure to x-radiation of thousands of asymptomatic women, but by the fact that there were no controls to ensure consistent compliance with standard mammography procedures or dosages among the 27 screening centers involved.[5]
His concerns were dismissed at NCI by the explanation that the project’s “objectives have been pretty well dictated by the American Cancer Society. In summary they are ‘Can American Cancer Society volunteers mobilize a large number (100,000 per year) of volunteers for thermography, mammography, and a physical exam?’”[6]
Civil-service scientists do not typically question work sponsored by their own organizations, so one must suspect that Bailar had considerable support from his peers when he submitted a manuscript to the Annals of Internal Medicine in June of 1975 in which he “regretfully” concluded “that there seems to be a possibility that the routine use of mammography in screening asymptomatic women may eventually take almost as many lives as it saves.” Using ionizing radiation risk estimates developed by the National Academy of Sciences, he cautioned that mammography doses could substantially increase not only the incidence of breast cancer, but “lesser hazards of radiation induction of leukemia, lung cancer, or other conditions.”[7]
Bailar noted that the methodological deficiencies in the HIP-GNY study (e.g., self-selection by participants and lack of clarity about the clinical breast exam techniques used for both study and control groups) rendered useful interpretation of its numbers difficult. Perhaps the greatest pitfall of attempting to quantify in years survival as a result of mammography, Bailar pointed out, was reflected in two biases evident in the HIP study: “lead time” and “length time.” The lead time bias invites the mistaken view that a cancer detected through mammography three years before it would have been detected clinically or through breast self-exam results in the addition of three years’ life—“survival”–to the patient’s prognosis.
The length-time bias results when an asymptomatic and indolent neoplasm, which would likely result in a good prognosis, is detected through mammography and survival is attributed to mammography.[8] In short, given the complexities of cancer such as wide variations in latency and aggressiveness, attempting to quantify the relationship of early detection to mortality would produce debatable conclusions at best.
Bailar was not alone in his misgivings about a project that would be largely funded by the NCI.[9] Scientists at the Institute voiced concerns not only about the lack of standardization in x-ray equipment, but “the possible hazards of ionizing radiation and the ability of an uncontrolled study to generate any meaningful conclusions about the value of screening mammography.”[10] Other statisticians agreed, also noting “the absence of a control group” which would prevent one from comparing results from the BCDDP with “data derivable from other sources.”
One leading radiation scientist questioned whether “mass screenings involving mammograhy is consistent with good radiation protection theory.”[11] Roswell Park Memorial Institute (cancer research center) biostatistician Irwin D. J. Bross warned, “many more cancers are produced than cured by x-rays …. This exposure to diagnostic x-ray will probably result in the worst iatrogenic epidemic of breast cancer in history.”[12]
In spite of the NCI scientists’ concerns, director Frank J. Rauscher, Jr. approved the joint project. Rauscher, who had been a scientist at the NCI since 1959, was appointed to the post by the ACS’s friend Richard Nixon, undoubtedly with the organization’s approval, if not at its suggestion. He was rewarded in 1976 for his support for the ACS project with a new position as Vice-President for Research at the ACS, at nearly double the salary he had been receiving at the NCI.[13] His new boss at the ACS, R. Lee Clark (Byrd’s successor), had previously served on the President’s Cancer Panel and president of the Cancer Center at the M.D. Anderson Hospital and Tumor Institute,[14] where some of the earliest uses of mammography had occurred. That institution would be one of the recipients of NCI funds to participate in the BCDDP.
By October 1975 the NCI was under assault for supporting the project. To respond to its critics the NCI created three expert panels to review the efficacy and safety of screening mammography. Having reviewed the HIP-GNY study, which the ACS asserted validated the “benefit” of mammography screening through a random controlled trial, the NCI panels’ report concluded that half of the cancers claimed to have been found only by mammography in the HIP study were, in fact, “palpable, in no sense clinically occult.” Proper review of any uncertain physical exam results might have prevented the resort to mammography at all in those cases. At the same time, “mammography was negative” for over 40% of the cancers detected through clinical breast exam. [15]
The institute’s review panels’ report, issued in March of 1976, should have brought the project to a halt. The report concluded that it was “difficult to assess the independent contribution of mammography” to any benefit of screening for women over 50 [italics added]. But rather than suggest that screening be stopped, the report temporized by recommending that “x-ray dosages should be kept as low as possible.” As for mammography for younger women, NCI support for such screening should “be concentrated on validating its use through appropriately designed randomized trials.”[16]
The Breast Cancer Detection Demonstration Project was completed in 1981. During that period controversy over the program continued unabated, so the National Institutes of Health, parent department to the NCI, did what many institutions do when faced with serious controversy: It sought to achieve a negotiated resolution of the issues—relying on process to obtain an outcome that a majority would accept. In 1977 the NIH convened the first of several “consensus conferences” held during the next two decades on the merits of mammography. Conference topics in 1977 included reviews of pathology specimens from biopsies conducted as a result of BCDDP screening.
One such review found that 66 of 506 pathological specimens taken thus far in the program “contained neither cancer nor carcinoma in situ.” As a result, 58 of the women from whom those specimens had been taken suffered unnecessary mastectomies.[17] Rather than regard these grievous errors as a serious risk of screening mammography, NCI declined to halt the program and referred the matter to those who made the diagnoses and performed the surgeries. As for whether to inform the misdiagnosed women of the mistaken results, the NCI side-stepped that moral obligation by leaving notification up to the women’s own physicians.[18]
When women recruited for the BCDDP were asked the source of their information about the campaign, their answers must have pleased the ACS, for which the project had always been essentially a marketing campaign. Slightly over 90% had heard about it through social contacts, the media, or ACS volunteers.[19] But what that number also revealed was that less than 10% had been recruited by their physicians on the basis of individual medical assessment. Thus over 90% of the BCDDP’s research subjects were self-selected for what most likely was a combination of personal reasons, social persuasion, and American Cancer Society assurances that they would benefit.
Slightly over a quarter of a million women participated in the Breast Cancer Detection Demonstration Project, or only .67% of the entire population in the country of women between the ages of 35 and 74. (In 1970 there were 40.4 million women resident in the United States in that age group.) Since there is no way we could consider the participants in the BCDDP as representative of that larger population, using data from the project as a basis for establishing health policy or medical screening protocols was scientifically unsupportable. Given the risks these women were subjected to, risks of which they knew little if anything, it was also ethically reprehensible.
Never adequately examined have been the consequences for these women of the radiation exposures they received during their five annual screenings. In the fall of 1976, not long after the NCI’s review panels issued their report, the Health Research Group of Public Citizen called for stronger informed consent requirements for “women undergoing mammographies at … federally sponsored screening centers across the country.”
In its letter to Dr. Theodore Cooper, assistant secretary at the department of Health, Education and Welfare which oversees the National Institutes of Health, the group cited “mammography machines which are emitting enough excessive radiation so that the absorbed dose is more than one rad per exam, a danger even to women over 50.” [20] (The NCI’s own BCDDP review panels had found “substantially higher [radiation] exposure levels in unmonitored screening clinics.”[21])
While supporters of mammography screening relied exclusively on unreliable quantitative data in an effort to prove its benefits, quantitative data became magically unimportant when considering its risks. Responding to the Health Research Group’s warnings about excessive radiation exposure, a National Cancer Institute official declared “’that’s uncertain. There just isn’t enough information to quantify the risk that much.”[22]
Women who entered the program at ages 35-44 had nearly twice as many instances of new breast cancers detected between the first and fifth year of screening as did women who began five years of screening at age 45 and older (which does not support claims that older women are at greater risk of breast cancer than younger women, and thus can benefit from mammography screening). At the same time, women between the age of 40 and 45 participating in the BCDDP experienced double the number of new cancers detected in the fifth year as were detected in their first year.[23]
Thus the longer women remained in the program, the more likely they were to have new breast cancer detected in the fifth year of their participation.[24] The possibility that a higher detection rate in the fifth year might be attributable to cumulative x-ray exposure during the previous four years was not mentioned in the BCDDP five-year summary report.
As with virtually all attempts to justify screening mammography by linking it to breast cancer mortality data, BCDDP-reliant articles in medical literature sampled from the 1980s and 1990s overlooked the important role of breast cancer treatment on survival rates. Disregarding treatment as a critical variable in cancer mortality assumes that the same treatment was provided to all of the women diagnosed with cancer at the same age and stage in their cancers, and that all women responded to that same treatment uniformly. Such assumptions are clearly untenable.
In any event, in 1997 the National Institutes of Health convened a 12-member panel for a ‘Consensus Development Conference’ with an audience of 1,100 in hopes of resolving the question of whether women in their 40’s should undergo routine screening mammography. The panel concluded that “the data currently available do not warrant a universal recommendation for mammography for all women in their forties.” That decision belonged to individual women, deciding for themselves, with “access to the best possible relevant information regarding both benefits and risks.”
Such decisions would be “difficult,” because they involved not only “scientific evidence and . . . her individual medical history;” they also involved each woman’s perception of “each potential risk and benefit, the values she places on each, and how she deals with uncertainty.” Finally, “a woman’s health care provider must be equipped with sufficient information to facilitate her decision making process.” Unfortunately, information then available to women and their health care providers was inadequate to warrant annual x-ray breast cancer screening for asymptomatic women of average or low risk.
Women over the age of 50 have occupied a curious category in the ongoing effort to market screening mammography for healthy women at normal risk for breast cancer. Both proponents and a few critics of adopting the practice as a standard medical protocol opined that the women most in danger of x-ray harm were women under 50, while women over 50 had been shown to “benefit.” Since the problems with the HIP-GNY study applied to the entire study, it could not be said to support mammography screening for older women any more than younger women. Nor did the BCDDP demonstrate such a benefit.
Was it that older women were more prone to breast cancer? If so, they did not become more prone to it until after the inception of mammography screening. Cancer deaths among women over 45 as well as women over 55 declined steadily during the 1950s and 1960s, only to rise after 1970. Since a majority of breast cancers are estrogen hormone responsive, it would seem natural that post-menopausal women would be less likely to develop breast cancer. But after three decades of routine mammography screening, the cancer death rate for women over 45 and 55 (as for all women) in 2000 exceeded the rate in 1950.[25]
Absent compelling epidemiological or laboratory evidence that all women over the age of 45 or 50 are more prone to breast cancer, that age threshold begs other explanations. Was it that older women could be exposed to x-radiation with less concern about its consequences because their instrumental value was diminished, i.e., they were less likely to get pregnant, or to be nursing, or to have the care of young children? Was it that if they were, indeed, to get x-ray induced cancer, it would not appear until their late 50s or later, at which time they might be just as likely to die of other causes as well as breast cancer?
Detecting and treating breast cancer at any stage was and remains far more profitable for the medical-industrial complex[26] than is cancer prevention. It is women–not those who gain materially from their treatment–who endure debilitating chemotherapy and radiation therapy, unnatural hair loss, mutilating surgeries, and catastrophic expense for those who are inadequately insured.
A cavalier disregard for this aspect of ‘curing’ breast cancer is reflected in ACS President (in 1976) Dr. Benjamin F. Byrd, Jr.’s response, when queried about the cancer risk of x-ray screening. “Even if there is a slightly increased risk of her getting the disease in the distant future, there’s also an excellent chance that by that time science will have learned to control the disease.”[27]
In all probability the 45 or 50 year threshold for routine screening mammography appealed because it offered a compromise in a highly contested medical protocol. Screening mammography’s opponents could ‘hold the line’ for younger women, at least, who supposedly had more to lose from the ravages of breast cancer than older women. As a compromise, the 50 year threshold enabled physicians to go about their business without being stymied by the dispute. Meanwhile, millions of asymptomatic women, urged on by powerful and trusted organizations (and sometimes their own physicians) to get their yearly screening mammograms, conscientiously complied.
(Published May 26, 2015. The next post examines the radiation risk of low-dose x-ray mammography, noting that those risks were known well before the American Cancer Society’s 1970s campaign to market screening mammography.)
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
[1] “Nixon Signing the National Cancer Act of 1971,” National Cancer Institute, News and Public Affairs. Video uploaded November 8, 2009 to You Tube. See https://www.youtube.com/watch?v=E2dzEDnGqHY.
[2] Lane E. Adams, “Executive Vice President’s Report,” American Cancer Society Annual Report, 1977. http://tobaccodocuments.org/rjr/502379608-9644.html. Downloaded December 29, 2014.
[3] For example: “Guttman Institute Reopens with Expanded Screening Services,” Cancer Network (Sept.01, 1996).
[4] J. W. Hendrick, M.D., Results of Treatment of Carcinoma of the Breast, Five to 18 Years,” Annals of Surgery, Vol. 146, pp. 728-750 (November 1957); J.W. Hendrick, M.D., “Aspiration Biopsy in Breast Cancer,” Journal of the American Medical Association, Vol. 166, No. 8 (February 22, 1958), p. 892.
[5] Daniel S. Greenberg and Judith E. Randal, “The Questionable Breast X-ray Program,” Washington Post (May 1, 1977).
[6] Daniel S. Greenberg and Judith E. Randal, “The Questionable Breast X-ray Program,” Washington Post (May 1, 1977).
[7] John C. Bailar, III, “Mammography: A Contrary View,” Annals of Internal Medicine, Vol. 84 (1976), pp. 80-82.
[8] John C. Bailar, III, “Mammography: A Contrary View,” Annals of Internal Medicine, Vol. 84 (1976), pp. 77.
[9] Daniel S. Greenberg and Judith E. Randall, “The Questionable Breast X-ray Program,” Washington Post (May 1, 1977).
[10] Barron H. Lerner, M.D., The Breast Cancer Wars: Fear, Hope, and the Pursuit of a Cure in Twentieth-Century America (Oxford University Press, 2001), Kindle edition, location 213.
[11] Barron H. Lerner, M.D., The Breast Cancer Wars: Fear, Hope, and the Pursuit of a Cure in Twentieth-Century America (Oxford University Press, 2001), Kindle edition, location 212-213.
[12] Daniel S. Greenberg and Judith E. Randal, “The Questionable Breast X-Ray Program,” and “Waging the Wrong War on Cancer,” The Washington Post (May 1, 1977).
[13] Daniel S. Greenberg, “X-Ray Mammography: Silent Treatment for a Troublesome Report,” New England Journal of Medicine, Vol. 296, No. 17 (April 28, 1977), p. 1015.
[14] Daniel S. Greenberg, “X-Ray Mammography: Silent Treatment for a Troublesome Report,” New England Journal of Medicine, Vol. 296, No. 17 (April 28, 1977), p. 1015.
[15] Daniel S. Greenberg, “X-Ray Mammography: Silent Treatment for a Troublesome Report,” New England Journal of Medicine, Vol. 296, No. 17 (April 28, 1977), p. 1016.
[16] Daniel S. Greenberg, “X-Ray Mammography: Silent Treatment for a Troublesome Report,” New England Journal of Medicine, Vol. 296, No. 17 (April 28, 1977), p. 1016.
[17] Barron H. Lerner, “To See Today With the Eyes of Tomorrow: A History of Screening Mammography.” Canadian Bulletin of Medical History, Vol. 20, No. 2 (2003), p. 311; Daniel S. Greenberg, “A Cancer Controversy: Did Doctor Know Best?” The Washington Post, April 4, 1978, p. A19.
[18] Barron H. Lerner, “To See Today With the Eyes of Tomorrow: A History of Screening Mammography.” Canadian Bulletin of Medical History, Vol. 20, No. 2 (2003), p. 311-12; “From the Editor,” New York Times, October 8, 1977.
[19] Based on 10% sample of initial patient history records. Larry H. Baker, M.D., “Breast Cancer Detection Demonstration Project: Five-Year Summary Report,” CA-Cancer Journal for Clinicians, Vol. 32, No. 4 (July/August, 1982), pp. 195-196.
[20] “Mammography Risky, Ralph Nader Warns,” The Milwaukee Sentinel, (November 3, 1976), p. 4.
[21] Daniel S. Greenberg, “X-Ray Mammography: Silent Treatment for a Troublesome Report,” New England Journal of Medicine, Vol. 296, No. 17 (April 28, 1977), p. 1016.
[22] “Mammography Risky, Ralph Nader Warns,” The Milwaukee Sentinel, (November 3, 1976), p. 4.
[23] Larry H. Baker, M.D., “Breast Cancer Detection Demonstration Project: Five-Year Summary Report,” CA-Cancer Journal for Clinicians, Vol. 32, No. 4 (July/August, 1982), p. 208.
[24] Larry H. Baker, M.D., “Breast Cancer Detection Demonstration Project: Five-Year Summary Report,” CA-Cancer Journal for Clinicians, Vol. 32, No. 4 (July/August, 1982), p. 208.
[25] Source: Centers for Disease Control, National Center for Health Statistics, Health, United States, 2013: With Special Feature on Prescription Drugs, Hyattsville, MD (2014); Centers for Disease Control, National Center for Health Statistics, Health, United States, 2010: With Special Feature on
Death and Dying, Hyattsville, MD (2011).
[26] Stevens, CW, Glatstein, E, “Beware the Medical-Industrial Complex,” Oncologist,Vol.1,No.4(1996), http://www.ncbi.nlm.nih.gov/pubmed/10388005.
[27] Daniel S. Greenberg and Judith E. Randal, “Waging the Wrong War on Cancer,” The Washington Post, (May 1, 1977).
No Comments