Breast Cancer: Tamoxifen Benefit Durable
Published: Dec 12, 2014 | Updated: Dec 12, 2014
SAN ANTONIO -- Tamoxifen's primary prevention benefits continue to accrue long-term, reaching a number needed to treat of just 22 at 2 decades of follow-up in the IBIS-I trial.
Incidence of any breast cancer was 7.8% with a 5-year course of the anti-estrogen treatment compared with 12.3% in the placebo group of the trial (P<0.0001), Jack Cuzick, PhD, of Queen Mary University London, and colleagues found in the IBIS-I trial at 20 years.
The curves continued to separate after 10 years of follow-up for that endpoint and for invasive estrogen receptor (ER)-positive cancers, they reported here at theSan Antonio Breast Cancer Symposium and simultaneously online in the Lancet Oncology.
"With an additional 10 years of follow-up we have substantially larger differences," Cuzick said, "... which bodes well for the prospects that the prevention may even be longer than 20 years -- potentially a lifetime."
That number needed to treat of 22 "is very favorable compared to any preventive treatment for heart disease or anything else in a general population," Cuzick said at a press conference.
The fundamental message is that chemoprevention with tamoxifen works, William Gradishar, MD, of Northwestern University told MedPage Today.
And whereas in the adjuvant breast cancer treatment setting, oncologists are thinking about extending tamoxifen beyond 5 years, that doesn't appear to be needed in the prevention setting, he noted. "The data don't look like there would be an added benefit."
Rowan T. Chlebowski, MD, of the Los Angeles BioMedical Research Institute at Harbor-UCLA Medical Center, agreed in an accompanying editorial that the number needed to treat was "very favorable." But he questioned the implication of mortality findings in the trial.
Mortality Discrepancy
Despite the 29% reduced relative risk of developing breast cancer, there were five more deaths from breast cancer -- 31 versus 26 -- in the tamoxifen group than in the placebo group, albeit not a significant difference (odds ratio 1.19, P=0.8).
That discordance had been even more pronounced at the 10-year follow-up (18 versus nine deaths, OR 2.00), although still not significant at P=0.08.
"To resolve such differences is not easy, since the authors state that the survival results could not be attributed to differing menopausal hormone therapy use or estrogen receptor-negative disease," Chlebowski noted.
The trial wasn't powered for mortality, Cuzick noted, pointing out that there was no overall mortality effect after 10 years (95 versus 96 cases).
"Less than 10% of the women who had breast cancer have died from it," he told reporters. "It's actually probably too early to make any pronouncements, favorable or unfavorable, about deaths from breast cancer. That's actually going to take another 10 years of follow-up."
Nevertheless, "additional events would be unlikely to appreciably change the results since the trend is in the opposite direction and survival curves that cross are unlikely to become positive," Chlebowski argued.
"The discordance between tamoxifen's effects on breast cancer incidence and outcome noted in the IBIS-I update could merely represent the effects of chance alone," he concluded, "or alternatively might indicate that tamoxifen mainly decreases the incidence of cancers with a very favourable prognosis, increases cancers with unfavourable outcomes, or both.
How these alternative ideas are viewed will determine the effect of the IBIS-I update on breast cancer chemoprevention practice in the clinic."
Subgroups
That explanation based on incidence of a differential impact on favorable-prognosis cancers found some support in the data.
The International Breast cancer Intervention Study (IBIS-I) included 7,154 higher-risk women recruited from genetics clinics and breast care clinics in eight countries.
It showed the most benefit in preventing ER-positive cancer (HR 0.66 at the median 16 years of follow-up,P<0.0001) and a similar albeit not statistically significant impact in preventing ductal carcinoma in situ (HR 0.65, P=0.05).
But invasive ER-negative breast cancer numerically went the opposite direction (HR 1.05, P=0.8).
"I don't think any of them are actually caused by tamoxifen," Cuzick said. "My explanation, and this is hypothetical still, is that there are cancers that would appear as ER-positive if untreated. Tamoxifen will hold them in bay for some years, but when they escape tamoxifen control they do so as ER-negative tumors.
"That has been seen in the adjuvant treatment setting, and it's likely to be the case here."
The benefit was also reduced among women who got menopausal hormone therapy during the time of active treatment with tamoxifen (P=0.04).
That was particularly true in the case invasive estrogen receptor-positive cancers. Women who got menopausal hormone therapy at the same time as their anti-estrogen therapy had a nonsignificant hazard ratio of 0.87 with tamoxifen versus none while other women had a hazard ratio of 0.55 (P=0.03).
Age was not a significant factor overall.
The trial also showed the expected increase in endometrial cancers with tamoxifen (29 versus 20 cases, OR 1.45, 95% CI 0.79-2.71).
While expected, Cuzick noted how important this risk is, pointing to the five deaths from endometrial cancer in the tamoxifen group versus none in the placebo group over the long-term follow-up.
The study remains blinded, so there hasn't been much crossover, he noted. "The results are pretty clean. There are no surprises with regard to side effects."
Không có nhận xét nào:
Đăng nhận xét