Nov. 30, 1999 (Chicago) -- Radiologists often say that it is difficult to find very small lesions in dense breasts --which the breasts of younger women often are -- with mammograms, but a study from Stanford University radiologists suggests that about half of the "missed" cancers are found in non-dense, or fatty, breasts.
Robyn L. Birdwell, MD, assistant professor at the Stanford University School of Medicine, reported the finding at the 85th Scientific Assembly and Annual Meeting of the Radiological Society of North America.
Birdwell, who is also acting director of the radiology residency program at Stanford, tells WebMD "that was one of the most striking findings of our study -- half of the tumors were in non-dense breasts." Birdwell says that the study debunked most of the mammography lore that she believed. "I thought that missed lesions look different, were usually calcifications or occurred in unusual locations. In our study none of those assumptions turned out to be true."
She and her colleagues amassed records from more than 1,000 consecutive cases of breast cancer detected through mammogram screenings. In about half of the cases, prior mammograms were available for comparison. A radiologist then examined the paired mammograms and determined that in almost 300 cases, evidence of cancer was visible on the first mammogram but had not been identified. An expert panel of five radiologists, who were blinded to the results of the second mammogram, were asked to review the initial films. "A majority of the five identified 115 cancers in 110 women and recommended recall imaging," Birdwell says. She then studied the characteristics of these "missed" cancers.
"Most of them weren't calcifications. They were masses," she says. Eighty-two of the 115 were invasive cancers and almost half were in the upper outer quadrant of the breast, a "very common location," she says.
Birdwell offers several theories to explain the missed diagnoses, including a very simple one: the mammographers were too distracted by other tasks to concentrate on the films under review. She says that at most busy breast centers, films are reviewed in rooms that aren't dark enough and that are too accessible to distractions from co-workers. "I've only ever seen one ideal screening room. ... It was down a hall, away from the action, and completely dark. I don't even think there was a phone. I was envious."
She says, too, the fact that about half the lesions were found in fatty breasts suggests that sometimes mammographers may "let down our guard a little when we have a fatty breast, so that's a teaching point." Traditionally, fatty breasts are easier to detect breast cancers in than are the denser breast tissues.
A remedy for the situation might be having two people review mammograms, and Birdwell says that several institutions do attempt this "second read" approach. "But this has pitfalls too, not the least of which is legal. Who is responsible for dictating the notes? Who signs off on the case? If the first reader is [deemed] responsible, will the second reader be attentive enough?" she says.
While Birdwell reported on cancers that should have been spotted on the first mammogram, her colleague Debra M. Ikeda, associate professor of radiology at Stanford, reported on another subset of this study: 174 mammograms that had "subtle findings but that would not have been selected for recall by most members of the expert panel."
Ikeda says that "70% of these cancers looked normal." Birdwell says that it is unclear if any of these cancers could have been identified using any other technology. She says, for example, that most persons who advocate the use of ultrasound say that the ultrasound approach should be used in women with dense breasts. In the cases she and Ikeda presented, half of the women would not have been selected for ultrasound screening because they don't have dense breasts.
While Birdwell says the findings suggest that mammography is far from foolproof, she adds that the good news is that all of these women did eventually have their cancers detected by screening mammography and even invasive cancers were at a very early stage and very treatable.
- Although it is difficult for radiologists to find cancerous lesions in dense breast tissue, half of all missed lesions occurred in non-dense, or fatty, breast tissue.
- There was nothing unusual about the types of lesions that were missed, leading one researcher to suggest that mammographers were just distracted in the screening room.
- Some institutions have two reviewers read each mammogram, but this introduces legal issues, such as which person will have the ultimate responsibility in the reading of the film.