Serious Scientific Flaws in the Canadian National Breast Screening Study Trials Revealed in a Report from Breast Cancer Screening Experts

Take-Away Points

• ■ Major Focus: To emphasize shortcomings of the two Canadian National Breast Screening Study (CNBSS) trials, which showed no benefit of screening mammography for reducing breast cancer mortality.

• ■ Key Result: Multiple serious flaws in the CNBSS randomized controlled trials (RCTs) included insufficient statistical power to detect differences in breast cancer mortality, substandard mammogram quality, and inclusion of patients with breast cancer symptoms.

• ■ Impact: Scientific flaws in CNBSS trials make them inappropriate for guiding global screening policies.

The two CNBSS trials intended to study the efficacy of screening mammography in 40- to 49-year-old women (CNBSS-1) and 50- to 59-year-old women (CNBSS-2). While eight international RCTs demonstrated the benefit of screening mammography to reduce breast cancer mortality, only the two CNBSS trials failed to show this benefit. Decades later, new evidence reveals serious scientific flaws in the CNBSS trials.

Seely, Eby, and Yaffe analyzed the CNBSS trials in the context of fundamental features of high-quality RCTs. The first major flaw is insufficient statistical power to detect breast cancer mortality differences between study groups. Contributing factors included insufficient recruitment due to fears of radiation during the 1960s–1980s trial period, high contamination and noncompliance rates, and patient loss to follow-up due to research staff error. Second, the CNBSS trials employed extremely poor quality mammography according to evaluations by external expert radiologist reviewers. Issues included the use of old or outdated mammography equipment, lack of photo timers for automatic exposure control, and use of mediolateral (captures less breast tissue) rather than mediolateral oblique (captures more breast tissue) views for the first 5 years of the 8-year study. Poor mammogram quality combined with lack of training for radiologists and technologists resulted in limited sensitivity (69%) of mammography for helping detect breast cancer compared with other RCTs (eg, UK Edinburgh trial, 94.5% and Swedish Two County trial, 95%).

Third, the CNBSS trials did not explicitly exclude women with breast cancer symptoms and thus inappropriately enrolled diagnostic mammography participants in a screening trial. For example, the study recruited women with breast lumps and used a breast surgical clinic as a recruitment site. Additionally, eyewitnesses reported violations of randomization (assignment of women with physical signs of breast cancer to the screening arm in CNBSS), and an excess of advanced breast cancers in the screening arm of CNBSS-1 provides additional indirect evidence of nonrandom group assignments.

Multiple major flaws in study design and execution make the CNBSS trials scientifically unreliable for evaluating the efficacy of screening mammography, and they should not be used to guide public policy decisions regarding breast cancer screening.

Highlighted Article

• Seely JM, Eby PR, Yaffe MJ. The fundamental flaws of the CNBSS trials: a scientific review. J Breast Imaging 2022;4(2):108–112. doi: https://doi.org/10.1093/jbi/wbab099