Nonmammographic Screening for Breast Cancer

To the Editor: The Editorial by Dr Kuhl¹ advocated breast MRI and other additional screening for breast cancer. However, for younger women there is an important issue of whether the average absolute benefit of any breast cancer screening is worth the harms and the opportunity costs.

Kuhl dismissed the first major harm inherent in screening healthy women: false-positive evaluation and biopsy-related anxiety. Screening advocates should help quantify this psychological distress and the opportunity costs for use in cost-effectiveness analyses.

Moreover, Kuhl did not discuss the second major potential screening harm: ductal carcinoma in situ overdiagnosis. This extreme form of length bias (in which screening preferentially identifies slower growing tumors without survival benefit) is one reason why the use of surrogate end points to replace randomized trials is not valid.²

Screening prevents some breast cancer deaths but not breast cancer. There is a 14.4% lifetime development risk for in situ and malignant breast cancer, along with a lifetime death risk of 2.9%.³ The complement means that 97% of women will not die from breast cancer. Over a decade, the absolute benefit or increase in survival percentage from screening mammography is only 0.05%.⁴ Ten times more women will receive overdiagnosis and overtreatment, including delayed radiation-induced coronary artery disease.⁴ This harm is why measuring the overall mortality effect in any randomized screening trial is optimal. Furthermore, what fraction of the women with additional cancers detected by additional imaging and biopsy (assuming any incremental risk reduction) and treated earlier than those detected by mammography would be “protected” or have their lives saved?

Randomized trials attempting to demonstrate a survival benefit from using breast MRI or ultrasound that is greater than mammography alone may give the wrong answer from the perspective of screening advocates. In the most recent randomized trial, there has been no statistically significant mortality benefit by starting mammography screening at age 40 years.⁵ The expense of a randomized trial is a bargain compared with the billions of dollars that might be wastefully spent annually on additional population screening instead of finding a cure.