'Flawed' Mammogram Study Adds to Controversy

Dr. Patrick I. Borgen

Patrick I. Borgen, MD

The large Canadian study that has caused a stir by indicating that mammograms are of no use in women aged 40 to 59 years, and in fact can lead to over-diagnosis of breast cancer, is flawed and misleading, according to the program chairman of the Miami Breast Cancer Conference.

The Canadian National Breast Screening Study¹—the results of which were published in mid-February in the British Medical Journal and which were reported widely in newspapers and on television—will be the subject of discussion during the conference, the chairman, Patrick I. Borgen, MD, said.

How Canadian Study Unfolded

“The Canadian study is probably the single most flawed, confounded trial in the history of trials—certainly in trials of mammography,” Borgen said. “This is a trial from 34 years ago, using ancient technology and a very flawed randomization technique, which put women with breast lumps into the mammography arm and normal-exam women into the control arm—and, of course, after 30 years, it showed no difference. Applying that to today is simply outrageous.” Borgen will address the topic in a talk titled “The Great Mammography Debate” at 8:45 a.m. on Friday, March 7. “We’ll be spending a lot of time at the Miami Breast Cancer Conference dissecting this trial, talking about it, and, frankly, developing a group strategy for how to overcome it,” he said. “Mammograms save lives, and we’ll be getting that message out loud and clear at Miami.”In the study, Anthony B. Miller, MD, professor emeritus at the University of Toronto, and colleagues compared 25-year cancer outcomes in 89,835 women aged 40 to 59 years who were randomly assigned to either receive or not receive annual mammography screening. The trial was launched in 1980, and the recruitment period continued through 1985. In women who received mammography, screening took place over 5 years at 15 centers in six Canadian provinces. Two-view film screen mammography was used, with craniocaudal and mediolateral views taken until 1985; after that, in accordance with updated guidelines, craniocaudal and mediolateral oblique views were taken.

Women aged 40 to 49 years in the mammography arm and women aged 50 to 59 years in both arms received annual physical breast examinations, while women aged 40 to 49 in the control arm received a single physical breast examination followed by usual care in the community, the authors wrote. All women in the study were taught breast self-examination by trained nurses prior to randomization. During the 5-year screening period, 666 invasive breast cancers were diagnosed in the mammography arm (nearly three-fourths of them screen-detected), with 180 of those cancers becoming fatal within 25 years. In the control arm, 524 invasive breast cancers were diagnosed, 171 of them fatal within 25 years (Table).

The overall hazard ratio (HR) for death from breast cancers diagnosed during the screening period by mammography was 1.05 (95% confidence interval [CI], 0.85-1.30). Age made virtually no difference in the survival outcomes of these women, according to the authors.

Over the entire study period, 3250 women in the mammography arm and 3133 in the control arm were diagnosed with breast cancer, and 500 and 505, respectively, died of the disease. The cumulative mortality from breast cancer did not significantly differ between women in the mammography arm and in the control arm (HR 0.99; 95% CI 0.88-1.12), Miller et al wrote.

Table. Detected Invasive Breast Cancers and Related Deaths, Canadian National Breast Screening Study

Detection Outcomes (n = 89,835)

Mammogram

(n = 44,925)

Physical Exam

(n = 44,910)

Hazard Ratio

Cancers detected at 5-y screening

666

524

n/a

Cancers detected at 5-y screening that became fatal over 25 y

180

171

1.05

Cancers detected over 25 y

3250

3133

n/a

Cancers detected that became fatal over 25 y

500

505

.99

Of the cancers found during the screening period, all those in the control arm and 68.2% in the mammography arm were palpable, but rates of node-positive cancer in the groups were similar (32.4% in the control arm and 30.6% in the mammography arm [P= .53 for difference]), the investigators found. Palpable cancers were generally larger than cancers detected only by mammography (2.1 cm vs 1.4 cm;P<.001), and were more likely to be node positive (34.7% vs 16.5%;P<.001), they wrote.

While the cancers detected strictly by mammography were associated with better survival outcomes than those that were palpable upon diagnosis, that result was not meaningful, the authors argued. &ldquo;Although the difference in survival after a diagnosis of breast cancer was significant between those cancers diagnosed by mammography alone and those diagnosed by physical examination screening, this is due to lead time, length-time bias, and over-diagnosis,&rdquo; they wrote.

After 15 years of follow-up, they added, 106 cancers detected in the mammography arm were attributed to overdiagnosis, meaning they might not have otherwise become clinically apparent within the patients&rsquo; lifetimes.

&ldquo;Annual mammography in women aged 40 to 59 does not reduce mortality from breast cancer beyond that of physical examination or usual care when adjuvant therapy for breast cancer is freely available,&rdquo; the authors concluded. &ldquo;Overall, 22% (106/484) of screen-detected invasive breast cancers were over-diagnosed, representing one over-diagnosed breast cancer for every 424 women who received mammography screening in the trial.&rdquo;

The investigators added that, &ldquo;in technically advanced countries, our results support the views of some commentators that the rationale for screening by mammography should be urgently reassessed by policy makers.&rdquo;

In discussing limitations of their findings, the authors contended that &ldquo;the lack of an impact of mammography screening on mortality from breast cancer in this study cannot be explained by design issues, lack of statistical power, or poor quality mammography. It has been suggested that women with a positive physical examination before randomization were preferentially assigned to the mammography arm. If this were so, the bias would only impact on the results from breast cancers diagnosed during the first round of screening (women retained their group assignment throughout the study). However, after excluding the prevalent breast cancers from the mortality analysis, the data do not support a benefit for mammography screening (HR 0.90; 95% CI 0.69-1.16).&rdquo;

SABCS Findings Differ

In recent findings, contrary to those of the Canadian National Breast Screening Study, a different team of investigators reported that four large studies of mammography— previously thought to have divergent results&mdash;are actually fairly consistent in demonstrating that mammograms are associated with reduced breast cancer mortality. Those results²were presented in December during the 2013 San Antonio Breast Cancer Conference (SABCS) by Robert A. Smith, PhD, senior director of cancer screening at the American Cancer Society in Atlanta.

Considered in that study were the Nordic Cochrane review, the UK Independent Breast Screening Review, the US Preventive Services Task Force review, and the European Screening Network review. Each review had generated different findings regarding how many women must be screened with mammography in order to prevent 1 breast cancer death, with estimates ranging from 90 to 2000, about a 20-fold difference, Smith said.

&ldquo;We wanted to understand why these estimates differ so much,&rdquo; he added in a press release issued by SABCS. &ldquo;What we found was that the estimates are all based on different situations, with different age groups being screened, different screening and follow-up periods, and differences in whether they refer to the number of women invited for screening or the number of women actually screened. When we standardized all the estimates to a common scenario—ie, the same exposure to screening, and a similar target population, period of screening, and duration of follow-up&mdash;the magnitude of the difference between studies dropped.&rdquo; After the adjustment, he said, there remained about a 2.5-fold difference between the estimates reported in the studies.

Otis W. Brawley, MD, chief medical and scientific officer of the American Cancer Society, has suggested that the medical community take a measured approach to the findings of the Canadian National Breast Screening Study.

&ldquo;This study adds to the body of evidence, but that body of evidence still suggests that there is some benefit to mammography screening among women aged 40 and above,&rdquo; Brawley wrote in an article for the February 14 edition of The Cancer Letter. &ldquo;The US Preventive Services Task Force assessed the total body of literature, including earlier reports from the Canadian trial, and said screening of women 40 to 59 likely produces a 15 percent relative reduction in the risk of death.

&ldquo;While it is my concern that the benefits of mammography screening have been exaggerated, this does not mean that it does not save lives, or that women should not get it,&rdquo; he continued. &ldquo;It means we need to use it with caution, explain its limitations, and realize we need to develop a better test.&rdquo;

References

1. Miller AB, Wall C, Baines CJ, et al. Twenty five year follow-up for breast cancer incidence and mortality of the Canadian National Breast Screening Study: randomised screening trial.BMJ. 2014;348:g366.
2. Smith RA, Duffy S, Chen TH-H, et al. Disparities in the estimates of benefits and harms from mammography: are the numbers really different? Presented at: 2013 San Antonio Breast Cancer Symposium; December 10-14, 2013; San Antonio, TX. Abstract S1-10.