Study questions value of CAD for mammography
Results from a new study showed that women getting mammograms at facilities using CAD (computer-aided detection) software were more likely to be given false positives, boosting the likelihood of call-backs and additional biopsies. The study was published today in The New England Journal of Medicine. Overall, the study results indicate that using CAD did not clearly improve the detection of breast cancer. The study was funded by NCI, the Agency for Healthcare Research and Quality, and the American Cancer Society.

Investigators at the University of California Davis Health System, Sacramento, Calif., and colleagues in the Breast Cancer Surveillance Consortium looked at the use of screening mammography in over 222,000 women involving over 429,000 mammograms. The observation took place from 1998 through 2002 at 43 facilities in Colorado, New Hampshire, and Washington State. The study included 2,351 women who received a diagnosis of breast cancer within one year after screening and also received a mammogram that did or did not use CAD. Seven of the facilities evaluated installed the CAD technology during the study.

The investigators stated that they “calculated the specificity, sensitivity, and positive predictive value of screening mammography with and without computer-aided detection, as well as the rates of biopsy and breast-cancer detection and the overall accuracy, measured as the area under the receiver-operating-characteristic (ROC) curve.”

Through CAD use, 32 percent more women were recalled for more tests and 20 percent more women had a breast biopsy. Overall, the investigators found that use of the software had no clear impact on the early detection of the disease.

“Within three years of FDA approval, 10 percent of the mammography facilities in the country were using CAD,” said Joshua J. Fenton, M.D., UC Davis Health System, the study’s lead researcher. “There had been no large-scale community-based review of CAD efficacy despite the rapid adoption of this technology, so we did this study to see if CAD was proving to be beneficial.”

Specifically, the investigators found that diagnostic specificity decreased from 90.2 percent before implementation to 87.2 percent after implementation, the positive predictive value decreased from 4.1 percent to 3.2 percent, and the rate of biopsy increased by 19.7 percent.
The increase in sensitivity before implementation of computer-aided detection to after implementation was not significant. In analyzing the data from all 43 facilities, CAD was “associated with significantly lower overall accuracy than was nonuse,” the authors said.

“It’s unfortunate that the use of the software has proliferated so widely before we are certain of its benefits,” said Fenton. “We need studies to determine if the benefits of the software outweigh its harms and costs. There is also the potential for new studies to improve the performance of CAD software.”

However, the authors do acknowledge in their report that since “computer-aided detection is used in the screening of millions of healthy women, larger studies are needed to judge more precisely whether benefits of routine use of computer-aided detection outweigh its harms.”

The study suggests that CAD could be helpful in finding ductal carcinoma in situ. Ferris M. Hall, MD, who authored an editorial accompanying the study wrote that “Because computer-aided detection is relatively more sensitive in detecting microcalcifications than in detecting masses, it is not surprising that Fenton et al. found that this was disproportionately associated with the detection of ductal carcinoma in situ.”
Hall also pointed to a “possible flaw” in the study, “the failure to assess the time it takes to adjust to computer-aided detection…” which as been “estimated to take weeks to years,” she wrote.

NCI Director John E. Niederhuber, MD, said in a released statement that the study “points out the need for the use of other techniques to find cancer at its earliest stages. NCI is incorporating techniques for imaging at the molecular level into many of its studies and is also conducting studies to improve the use of CAD and conventional mammography.”

“In the end, technology facilitates screening. Ultimately, treatment requires radiologists working with the examining physician and the responsible surgeon to put everything together. We worry about false positives, but we certainly don’t want to miss any cancers, either,” Niederhuber added.

The study has caused some CAD vendors to fret. Bedford, Mass.-based Hologic for one sited several deficiencies with the study, stating in a release that the study involved “43 sites from just three states contributed data to the Fenton study; only seven sites contributed data on CAD. The seven sites contributing CAD cases to the Fenton study had been using CAD for a time period ranging from two months to several years with an average amount of time in use of seven months.”

Hologic also pointed out that CAD population in the study was 31,030 mammograms, which is “comparable to the sample size of many of the previously published, peer-reviewed studies that show an increase in detection rates using CAD.”

CAD technology has also evolved considerably since the data was captured for the Fenton study. Moreover, Hologic said, multiple other CAD studies have shown an “increased detection rate of breast cancers at an earlier stage.”