One in ten breast cancer tests 'misleading'

A new genetic test and views on over treatment.

RB




http://www.timesonline.co.uk/article...067221,00.html

ABSTRACT


SCREENING for breast cancer detects many cancers that do not exist or pose no hazard, a study indicates.

This means that many women may be needlessly treated for cancers that will not threaten their lives. Although screening saves many lives — 1,400 a year in Britain, according to the latest estimates — women who have a wrong or needless diagnosis pay the price in worry and unnecessary treament.

A Swedish team has determined that this “overdiagnosis” amounts to about 10 per cent of women screened. Thus, in every ten women who test positive, one gains no benefit because she does not have cancer, the cancer she does have will not spread, or she will die of another cause before it can affect her. The team has been able to arrive at the figure by following up a group of women who took part in a trial of cancer screening in Malmö that began in 1976. The study included more than 42,000 women.


AND

http://professional.cancerconsultant....aspx?id=36379


ABSTRACT

The study presented at the 2005 SABCS was another validation study of the 76-gene assay that included 180 patients; 164 of whom had ER-positive breast cancer and 16 of whom had ER- negative breast cancer. Patients had received no systemic therapy, and the median follow-up was 100 months. Gene expression was measured through the Affymetrix VDX2 GeneChip.

*
Hazard ratio of distant metastasis-free survival (DMFS) comparing favorable versus unfavorable signature = 7.41 (95% CI: 2.63-20.9). P= 8.5 x 10-6
*
Hazard ratio of overall survival (OS) comparing favorable versus unfavorable signature = 5.45 (95% CI: 1.62-18.3); P = .002

The researchers concluded that the 76-gene assay appears promising in providing prognosis in patient with lymph-node negative breast cancer. However, extensive further validation is necessary to confirm these findings, as confidence intervals are large and the reproducibility issues with the Affymetrix GeneChip need resolving.