In an accompanying editorial,¹ Karla Kerlikowske, MD, of the University of California, San Francisco, and colleagues point out that while the finding of Killelea et al that digital mammography screening was not associated with downstaging of disease may not be surprising, given the similar accuracy of digital and film screening for cancer detection in older woman that has been observed elsewhere, the data are reassuring insofar as they suggest little evidence of increased harms with the shift to digital mammography.

They note that the study showed a small increase in diagnostic mammography and a small decrease in biopsy rates, suggesting that overdiagnosis has not been increased. However, they also emphasize that the increased costs associated with the use of digital mammography and computer-aided detection are troubling and require addressing if principles of high-quality care are to be upheld.

Reduced Cost Efficiency

The commentators point out that the Killelea et al study adds to the literature indicating higher costs of digital mammography without clear added benefit in women aged ≥ 65 years. Although some data suggest a small health gain at increased cost among women aged 50 to 79 years with biennial digital screening, other data indicate no overall benefit compared with film screening, and there is no convincing evidence to support use of computer-aided detection (which increased from 3.2% to 33.1% in the Killelea et al study).

As the commentators stated, “Thus, the transition to digital breast cancer screening in the US has increased screening mammography costs for possibly small or no health gains, resulting in screening mammography being less cost efficient than in the past.” This reduced cost efficiency has to be viewed in the context of the huge cost of mammography screening in the United States (estimated at $7.8 billion in 2010) and the fact that Medicare reimburses more for digital than film screening.

Risk-Based Screening

Kerlikowske and colleagues suggest approaches to address this situation based on risk-based screening. First, the frequency of screening could be reduced using a risk-based approach in which low-risk women could stop screening or be screened at greater intervals, thus reducing the overall numbers of screenings, false-positives, and biopsies, as well as cases of overdiagnosis. The commentators state that breast imaging registries and organized screening programs are suited for use in evaluating risk-based strategies, given the risk factor data collected by these programs and the ability to link such data with outcomes.

Second, risk-based screening could also include strategies to discontinue screening in older women who are unlikely to benefit from screening. The commentators note that 55% of the Medicare patients in the Killelea et al study were aged ≥ 75 years and that > 40% had at least one comorbidity. However, there is no randomized trial evidence indicating a benefit of screening in women aged ≥70 years.

In this regard, the commentators cite findings of the Swedish Two-Country trial, which showed no statistically significant reduction in breast cancer mortality among women aged 70 to 74 years screened every 24 to 33 months, a U.S. modeling study showing approximately two additional breast cancer deaths prevented per 1,000 women undergoing biennial screening from ages 70 to 74 years, and a recent study indicating that biennial vs annual screening did not increase risk of advanced-stage tumor incidence and reduced false-positive rates among women aged 66 to 89 years. They further point out that the benefit of screening is likely reduced among older women with moderate or severe comorbidities and limited life expectancy, whereas the harms likely increase, since risk of overdiagnosis increases with age.

Thus, Dr. Kerlikowske and coauthors propose: “Risk-based screening that incorporates defined stopping ages based on breast cancer risk in combination with comorbidities and life expectancy may further improve the balance of benefits vs harms for older women, because screening would not be offered to elderly women with limited life expectancy who are unlikely to benefit.”

Conclusions

The commentators concluded: “To be responsible advocates for high-quality medical care, our enthusiasm for new technologies should not replace strong, consistent evidence that the benefits of the new technology outweigh the harms in a clinically important way. Once manufacturers demonstrate at least comparable performance, timely evaluation of emerging technologies using breast imaging registry risk factor and performance data can provide important comparative effectiveness evidence for guiding clinical applications.”

They note that efforts in this regard will be particularly important in evaluating whether tomosynthesis—“the newest breast cancer screening technology rapidly diffusing into community practice with minimal comparative effectiveness evidence”—provides any clinically important benefits over digital mammography. ■

Disclosure: The work was supported by National Cancer Institute-funded Breast Cancer Surveillance Consortium grants. The study authors reported no potential conflicts of interest.

Reference

1. Kerlikowske K, Hubbard R, Tosteson ANA: Higher mammography screening costs without appreciable clinical benefit: the case of digital mammography. J Natl Cancer Inst 2014; doi:10.1093/jnci/dju191.