FDA Panel Rejects Breast Cancer Screening Device


GAITHERSBURG, MD. — A device intended for use in the annual breast cancer screening of women aged 30–39 years who have no family history of the disease and a negative clinical breast examination was found not effective in a unanimous vote by the Food and Drug Administration Obstetrics and Gynecology Devices Panel in August.

The T-Scan 2000ED is manufactured by Mirabel Medical Systems and uses a 1-volt electrical current to collect 306 measures of impedance for each breast. Mirabel believes differences in impedance can indicate the presence of cancer; women with positive results would then be referred for mammography or ultrasound. A positive T-Scan result is shown on the operator's monitor as a red line, and a negative result appears as a green line; no tissue imaging is produced.

A principal objection of the panel was that the sensitivity arm of the study included women aged 30–45 years, rather than just the 30–39 target age group included in the specificity arm. Moreover, both arms included clinical sites in Israel, where the population is dissimilar to that of the United States in ethnic makeup, body mass index, and possibly brassiere cup size and other factors, panel members noted.

The racial composition of both study arms also raised concerns; minority patients were underrepresented in comparison with the U.S. population, several panel members noted. In the specificity arm, 1.5% of the patients were of Asian descent, 2.9% were black, and 2.7% were Hispanic. In the sensitivity arm, 2.3% were Asian, 8% were black, and 4.6% were Hispanic.

The prospect of needlessly alarming patients with positive T-Scan results later belied by negative mammograms was raised by the panel. The FDA deduced from Mirabel's data that using a prevalence of 0.0015 in a population of 10,000 women, 530 would receive false positives. At the same time, T-Scan would identify 4 of 15 expected cancers in the group and miss roughly 11 cancers. Some panel members worried that false positives would lead to increased patient anxiety that might not be assuaged by negative mammogram testing. Specificity was calculated at 94.7.

Dr. Mark Akin, a private-practice gynecologist in Austin, Tex., and one of several physicians to use the device in clinical trials, believes any extra anxiety is minimal and worth the potential benefit. “Whatever level of anxiety [my patients] may have had, I'm certain that it pales in comparison with the devastation a woman in her 30s experiences when she is told she has advanced breast cancer and will never see her children grow up. What I saw in my study were young women who were grateful that I cared enough to test a new device in order to potentially find breast cancer when it was early enough to have a chance at cure,” he said in an interview. “Although the pooled data had a 95% specificity, out of my patients, I only had six positive findings. Statistically, [that] should have been 18. These T-Scan-positive patients were told they had a 99% chance of being normal. All six subsequently had a screening mammogram that was normal. I did not observe any more anxiety in these patients than when a patient has an abnormal Pap smear and goes on to have a normal colposcopy.

Also of concern to some panel members was the predetermined requirement of 2 for relative probability, meaning that any woman in the target age range with a positive T-Scan finding would need to be twice as likely to have breast cancer as the average woman in the 30- to 39-year-old population. Relative probability was ultimately determined to be 3.6. However, at least one panel member regarded the threshold of 2 as inappropriate. “I think the 2 hurdle is much too low. … It absolutely must be associated with age. A 30-year-old is incredibly different from a 39-year-old in terms of risk for breast cancer,” said Donald Berry, Ph.D., chairman of biostatistics and applied mathematics and director of cancer research at the University of Texas M.D. Anderson Cancer Center, Houston.

According to the FDA's analysis of the sensitivity trial, there were four confirmed cancers in women aged 30–39 who had negative clinical breast examination findings and negative family history of cancer. Two cases were in the United States, and two were in Israel, where one was detected by the T-Scan. An additional 11 cancers were confirmed in the 40–45 age group, of which 4 were detected by T-Scan. Thus the device detected 5 of 15 overall cancers and 1 of 4 cancers in women younger than 40, for a sensitivity rate of 25%, the FDA concluded.


Next Article: