How much should mammographers fret about overdiagnosis?

Not at all, according to a commentary published in the August issue of Radiology. The overdiagnosis myth overshadows the real problems—suboptimal patient management and overtreatment. The authors urged imagers to remain vigilant in detection and suggested they play an active role in management decisions.

Screening mammography continues to trigger both vocal opposition and ardent support with some critics asserting that aggressive screening in the U.S. has led to an epidemic of overdiagnosis.

“The term overdiagnosis is often used in population–based studies to describe the difference between cancer detection and subsequent treatment of abnormal findings and the actual corresponding population-based effect on mortality,” wrote David Gur, ScD, and Jules H. Sumkin, DO, both from the department of radiology at University of Pittsburgh.

Overdiagnosed cancers represent disease that is detected and treated which would not have affected mortality if it had been left alone.

A study published Nov. 22, 2012, in the New England Journal of Medicine, estimated that 31 percent of diagnosed breast cancers likely represent overdiagnosis. The researchers attributed the majority of the progress in reducing breast cancer mortality to improvements in treatment rather than improvements in detection.

This leads to the assumption, according to Gur and Sumkin, that radiologists could solve overdiagnosis by not searching for preclinical abnormalities. However, the authors contend good and correct imaging data are always good.

“What is not always good is the way this information is used when making optimal management decisions… Once a decision to screen is made, the role of radiology in screening for early detection of disease has been and needs to remain diligence in detecting and correctly diagnosing all depicted abnormalities at the earliest stage possible.”

Correctly diagnosed abnormal findings can be optimally managed, suboptimally managed, mismanaged and possibly overtreated, continued Gur and Sumkin.

The decision then becomes how to effectively manage a finding that is not considered indicative of clinically important disease. As society makes these decisions, the role of imaging may be to follow findings until they progress to a point where a change is management is required, e.g. grade 1 ductal carcinoma in situ.

This approach follows other established protocols in imaging; for example, some findings in the bones and chest are recognized as “leave-me-alone” lesions.

The hitch in breast imaging is the dearth of data on how to manage some lesions that may reflect overdiagnosis. Gur and Sumkin cited treatment of lobular neoplasia as one example.

They encouraged radiologists to focus on detection, correct diagnosis and appropriate treatment of early clinical disease and remain actively involved in management decisions, “as we cannot fully blame our colleagues for suboptimal use of information we provide if we are not actively trying to participate in the process.”

Finally, the authors emphasized that medicine and society needs to address the real problem—overtreatment. “Until there are validated and accepted alternatives to imaging-based screening, as imagers finding these lesions, we are doing exactly what we are supposed to do and exactly what women expect us to do.”