Eight-Year NSABP Data Confirm Radiation Benefits in DCIS

PARIS--Detection rates of ductal carcinoma in situ (DCIS) have soaredthanks to mammography but, to date, only one randomized trial has attemptedto quell the resulting therapeutic turmoil. Now, eight-year follow-up resultsfrom that NSABP trial indicate that the advantages of breast irradiationfollowing local excision continue to be maintained over the long term."All patients benefit from radiation," Bernard Fisher, MD, scientificdirector of the NSABP, said at the Seventh International Congress on Anti-CancerTreatment (ICACT). "The most optimal prevention of second ipsilateralbreast tumors is free specimen margins and post-lumpectomy radiation."

Dr. Fisher pointed out that second tumors in the ipsilateral breastdeveloped in 27% of women with DCIS who had been treated with lumpectomyalone and that nearly half of these tumors were invasive. In contrast,only 12% of women treated with postsurgical breast irradiation developedsecond tumors, roughly a third of which were invasive.

"The important question is whether you get an invasive cancer fromDCIS, and these data show how well radiation works in controlling the developmentof ipsilateral invasive breast tumors," he said. The eight-year ratesof local-regional disease, distant disease, and deaths were equally smallin both NSABP treatment arms, however.

In an effort to pinpoint those subgroups of patients that would mostbenefit from radiation, the NSABP investigators examined the value of pathologiccharacteristics in predicting the development of a second ipsilateral breasttumor. Multivariate analysis revealed that comedo necrosis and margin involvementwere the only independent prognostic factors, Dr. Fisher said.

Nevertheless, he emphasized, radiation was effective irrespective ofnuclear grade, histologic type, tumor size, the presence or absence ofcomedo necrosis, and margin involvement.

Best and Worst Scenarios

"Even in the worst scenario, with involved margins and marked comedonecrosis, radiation reduces the risk of ipsilateral breast tumors to avery low level," he said. "And if you take the best scenario,with free margins and absent-to-slight comedo necrosis, radiation stillhas some effect in improving the situation."

Thus, Dr. Fisher maintains that "despite its heterogeneity, DCISwarrants management by local excision and radiation regardless of the presenceor absence of favorable tumor characteristics."

He warned against the use of prognostic models in clinical decisionmaking, reminding the audience that models to predict which DCIS patientscan safely forego radiation, such as the Van Nuys index, have never beenvalidated with prospective clinical trials.

In a newly completed NSABP trial, Dr. Fisher said, women with more extensiveDCIS have been treated with lumpectomy and breast irradiation, stratifiedby age, and then randomized to tamoxifen (Nolvadex) or placebo.