SAN ANTONIO -- Tamoxifen has a "true preventive effect" on breast cancer in women with a strong family history of the disease -- but it may take several years before the benefit is seen.
SAN ANTONIO, Dec. 18 -- Tamoxifen has a "true preventive effect" on breast cancer in women with a strong family history of the disease -- but it may take several years of treatment before the benefit is seen.
The finding emerged in the second decade of the long-running Royal Marsden cancer prevention study, according to Trevor Powles, Ph.D., of the Royal Marsden Hospital in London.
Two decades after the randomized, placebo-controlled trial started, women in the tamoxifen arm, with a median follow-up of 13 years, have a significantly lower rate of estrogen receptor-positive breast cancer than those getting placebo, Dr. Powles told the San Antonio Breast Cancer Symposium here.
And when the researchers restricted the analysis to the so-called post-treatment period -- the years after participants stopped taking the medication -- there was a significant benefit for all types of breast cancer, Dr. Powles said.
The finding was revealed in a planned analysis, triggered when the number of women in the trial with cancer reached 200, he said. An earlier report in 1998 -- triggered at 70 events -- showed no differences between the arms regardless of cancer type.
That's because the effect only began to show up, Dr. Powles said, after the treatment period, implying that the drug is actually preventing disease, not treating small cancers that had been undetected.
The study, combined with another study presented here, the 7,000-patient IBIS-1 trial, added to the data supporting the use of tamoxifen to prevent breast cancer, according to Kent Osborne, M.D., of Baylor College of Medicine in Houston.
With a median follow-up of eight years, the IBIS-1 trial also showed a reduction in invasive breast cancers that continued to be seen after the treatment ended.
"Both studies show that with longer follow-up, the reduction in the incidence of breast cancer does not go away -- it keeps going for many years," Dr. Osborne said.
One puzzle, he said, is that in the Royal Marsden study, "there was no effect for the first five years and then you started to see it." Researchers can't yet explain the delay, he added.
Dr. Osborne added that women rarely take tamoxifen to prevent breast cancer in the U.S., largely because of fears of adverse events associated with the drug. The findings of these studies seem unlikely to change that, he said.
The Royal Marsden study enrolled 2,494 healthy women between 1986 and 1996, and randomized them to either 20 mg/day of tamoxifen for eight years or placebo. Median follow-up of the 2,471 women who were analyzed for this report was 13.2 years, he said.
The women were considered at high risk for breast cancer because they had a strong family history of the disease, Dr. Powles said. Once in the trial, they underwent a clinical review every six months and mammography every year.
The study remains blinded to its participants and their doctors, he said.
This analysis found that overall, women taking tamoxifen had a hazard ratio for breast cancer of 0.78, but the 95% confidence interval included unity and the reduction was not statistically significant.
The hazard ratio remained unchanged (and non-significant) when the researchers restricted the analysis to the treatment period, but became significant in the post-treatment period. The latter hazard ratio was 0.67, which was significant at P=0.05.
Much of the reduction was driven by decreases in estrogen receptor-positive cancer, Dr. Powles said.
Over the whole trial period, Dr. Powles said, tamoxifen is now seen to reduce the risk of estrogen receptor-positive cancer by 38%. The hazard ratio was 0.62, which was significant at P=0.05)
But again, most of the effect is after treatment ended, he said. In the treatment period, there was no significant difference in cancer incidence, as the 1998 report indicated.
Yet in the post-treatment period, he said, tamoxifen reduced the risk of estrogen receptor-positive cancer by 51%, which was statistically significant at P=0.005.