Winning combination: Mammo + tomo cuts false positives and missed cancers

Twitter icon
Facebook icon
LinkedIn icon
e-mail icon
Google icon
 - breast cancer

The addition of screening breast tomosynthesis to digital mammography increased diagnostic accuracy and reduced recall rates, according to a study published online Nov. 21 in Radiology.

Standard mammography is a not an ideal screening modality, as exams may miss 20 to 30 percent of breast cancers. Digital tomosynthesis may help address the challenge by enabling analysis of individual planes of the breast while also reducing the impact of overlapping tissue. However, mammography allows comparisons with prior exams.

Combining both exams preserves the relative strengths of each, according to Elizabeth A. Rafferty, MD, from the department of radiology at Massachusetts General Hospital in Boston, and colleagues.

Despite the early promise of tomosynthesis, previous studies have not provided clear answers. Two small studies suggested reduced recall rates when combining tomosynthesis with digital mammography. Another reported no significant difference in screening accuracy between tomosynthesis and mammography.

Rafferty and colleagues designed a multi-institutional trial to compare radiologists’ accuracy and recall rates for breast tomosynthesis combined with digital mammography versus digital mammography alone.

The researchers enrolled 1,192 women between July 2006 and May 2007, and performed two enriched reader studies. Study 1 included 312 cases, including 48 cancers, and images were read by 12 radiologists. Study 2 comprised 312 cases, with 51 cancers, and images were read by 15 radiologists.

Radiologists underwent training that included 150 cases illustrating the appearance of normal tissue patterns, summation artifact and benign and malignant lesions.

During study 1, readers scored the digital mammogram and were then provided with the tomosynthesis images to record a combined score. In this study, radiologists recorded only that an abnormality requiring recall was present. An analysis suggested that radiologists tended to inappropriately dismiss lobulated lesions as benign, so supplementary training prior to study 2 addressed such cases.

In study 2, radiologists also read the mammogram first and then the combined images. In this study, they recorded lesion type and location.   

“In both studies, digital mammography plus tomosynthesis demonstrated superior diagnostic accuracy compared with digital mammography alone,” wrote Rafferty and colleagues. Average gains in the area under the curve were 7.2 percent and 6.8 percent for the two studies, a consistent and significant result, according to Rafferty and colleagues.

Average sensitivity increased by 10.7 percent and 16 percent, respectively, in the two studies. “Almost all of the gains in diagnostic sensitivity with the combined modality were attributable to the improved detection and characterization of invasive cancers, which are the cancers we are most concerned about because of their potential to metastasize,” Rafferty said in a press release.

The researchers also observed significant reductions in the recall rates for all readers.  Absolute recall rate reductions of 38.6 percent and 17.1 percent were seen in studies 1 and 2, respectively.  “Clinically, such a reduction in the recall rate can be expected to translate to a substantial number of diagnostic tests being avoided,” wrote Rafferty et al.

The researchers pointed out key limitations that could have obscured results of the combined model. The study excluded cancers that could be detected only by tomosynthesis, which would likely underestimate sensitivity gains in clinical practice. Also, prior images were not available to assist readers in their assessment, and the study mix was tilted toward biopsy lesions. Thus, the tendency to recall may have been higher than in clinical practice, so the magnitude of the reduction in the recall rate might be lower in practice.

Despite these issues, the dual benefit of improved diagnostic accuracy and a reduced false-positive recall rate could help women avoid unnecessary additional testing and decrease anxiety, convenience and cost, according to Rafferty and colleagues. “The beauty of tomosynthesis is that it addresses two major concerns with screening mammography: missed cancers and false positive rates,” Rafferty concluded.