Overdiagnosis on Screening Mammography 'Unavoidable'

Pam Harrison

March 19, 2013

VIENNA, Austria — The risk for overdiagnosis on screening mammography is unavoidable, and healthcare providers need to inform patients of this risk prior to screening, investigators say.

For women screened every 2 years from the age of 50 to the age of 70, about 1% "will have a breast cancer detected and treated that otherwise would not have surfaced clinically during their lifetime," Ulrich Bick, MD, from Charité Medical University in Berlin, Germany, told Medscape Medical News.

"This is highly significant and it's something we have to talk about," Dr. Bick said here at European Congress of Radiology 2013.

When he and his colleagues tracked the 30-year trend in breast cancer incidence, they found a 20% to 25% increase in the incidence of ductal carcinoma in situ (DCIS).

If the excess risk was solely related to the overdiagnosis of DCIS, the issue could be easily resolved. "Unfortunately, it is not that simple," Dr. Bick said.

Over the same 30-year interval, the investigators found an increase in the incidence of invasive breast cancer. Some, but not all, of this was likely influenced by the widespread use of hormone replacement therapy during that timeframe, Dr. Bick explained.

This is highly significant and it's something we have to talk about.

In fact, when a team of Norwegian investigators calculated the combined incidence of overdiagnosis of DCIS and invasive breast cancer in women who participated in the Norwegian Breast Cancer Screening Programme, they found an estimated incidence of 17% to 20%.

In the same dataset, the overdiagnosis of invasive breast cancer alone was 11% to 13%, "suggesting that about two thirds of overdiagnosis actually represents invasive cancer that's detected during screening," Dr. Bick said.

Low- and even intermediate-grade DCIS lesions can have an excellent prognosis, even in the absence of treatment, but some DCIS lesions are actually high-grade in situ cancers.

"These cancers are highly relevant for mortality and they need to be diagnosed in a timely way," Dr. Bick said.

As for low-grade DCIS, data on the natural history of these breast cancers are sparse.

The numbers suggest that roughly 50% of DCIS patients will eventually develop an invasive breast cancer, but that can take several decades.

"The main problem is whether or not you consider low-grade DCIS a cancer," Dr. Bick said. Guidelines currently dictate that once detected, DCIS should be treated and, if it is treated, it needs to be excised completely.

"These lesions are often large," Dr. Bick explained, "and a large proportion of women with low-grade DCIS undergo mastectomy." In contrast, the excision of invasive breast cancer often requires that only a small amount of tissue be removed, so surgery is often minimal, he noted.

Clinical trials have not yet identified the subset of patients who can be spared radiotherapy because of the low malignant potential of their in situ carcinoma. In the absence of these data, patients could receive radiotherapy they don't need.

In fact, no clinical trial of low-grade DCIS has found an effect of any kind of adjuvant therapy on metastatic disease or long-term survival.

Radiologists can take steps to lower the likelihood of unavoidable overdiagnosis on screening mammography.

"We all know that digital mammography increases both in situ detection of lesions, independent of grade, and invasive cancer, so it's important to use digital mammography in screening to better diagnose relevant disease," Dr. Bick said.

Screening Asymptomatic Patients

Adequate pursuit of microcalcifications on breast imaging is "highly relevant" for the detection of small invasive cancers because many of the high-grade cancers have microinvasions that are important to detect, he explained.

"I believe biopsy of a microcalcification is necessary to define the nature of the underlying abnormality," Dr. Bick said. When choosing not to treat a low-grade DCIS and proceeding with a wait and see approach, "I feel much better if I have a magnetic resonance imaging [MRI] result first. Sometimes microcalcifications are only the tip of the iceberg, and you need an MRI to exclude relevant high-grade or invasive disease in other locations, distant from the area of biopsy."

These patients should be followed-up within 6 months, he noted.

"If we diagnose low-grade DCIS, it's crucial for us as radiologists to tell patients that if they do have cancer at all, it's not the invasive breast cancer they read about, that this is not a dangerous disease. And we need to encourage physicians to reduce overtreatment," Dr. Bick said.

Session cochair Harry de Koning, MD, PhD, from Erasmus Medical Center in Rotterdam, the Netherlands, agrees there will always be overdiagnosis when asymptomatic patients are screened.

"However, the amount of overdiagnosis is limited to perhaps 10% to 20% of all cancers detected on mammography screening. Given the benefits of mammography, it's a reasonable balance," Dr. de Koning told Medscape Medical News.

He explained that it is reasonable to question whether certain subtypes of breast cancer, most notably low-grade DCIS, can be treated less aggressively, or perhaps not at all.

In the United Kingdom, plans are currently underway for a randomized controlled trial to evaluate whether the prognosis of low-grade DCIS is equivalent in those who receive treatment and those who do not.

Dr. Bick reports receiving equipment support from Hologic and Toshiba, holding a patent with Hologic, and speaking on behalf of the General Electric Company and Carestream Health. Dr. de Koning has disclosed no relevant financial relationships.

European Congress of Radiology (ECR) 2013: Abstracts A63. Presented March 8, 2013.