Here are our highlights from the AACR-CTRC San Antonio Breast Cancer Symposium that continue to resonate with us from December:
ATLAS (Adjuvant Tamoxifen, Longer Against Shorter)
The one presentation at the symposium that will change the standard of care was the report of the updated results of the ATLAS trial. In addition to changing the standard of care, it provides a strong reminder of:
- The risks of relying on results from underpowered trials; and
- The very long time it takes to draw conclusions about the efficacy of treatments for estrogen receptor positive breast cancer.
It is well recognized that the risk of recurrence for ER-positive breast cancer persists well beyond five years. However, the recent standard of care has been just five years of tamoxifen, as it had not previously been demonstrated that longer therapy offered any additional benefit. Indeed, in 1995, the NCI issued a clinical announcement that more than five years of tamoxifen treatment was not warranted in routine clinical treatment of women with ER-positive, node-negative breast cancer.
Some investigators hypothesized that perhaps over time tumor cells mutated and could use tamoxifen in lieu of estrogen as a growth stimulant; after all, breast cancer cells could be made dependent on tamoxifen in vitro, and it is well known that a percentage of metastatic breast cancer shrink merely from withdrawal of tamoxifen therapy. However, the number of events in the studies on which the NCI announcement was based was known to be too small for firm statistical conclusions, and more adequately powered studies continued.
ATLAS randomized 6,846 women with ER-positive breast cancer who had received five years of adjuvant tamoxifen to continuing for an additional five years or stopping therapy (control group). Continuing tamoxifen for 10 years reduced the risk of breast cancer recurrence (p=0.002), breast cancer mortality (p=0.01), and overall mortality (p=0.01).
Interestingly, these results were not evident for years 5-9 of tamoxifen treatment; an earlier report from the ATLAS investigators had not shown any significant survival benefit. The relative risk for mortality was 0.97 for years 5-9 and 0.71 in later years.
ATLAS will change practice in women who remain premenopausal after five years of therapy and who have a sufficiently high remaining risk of recurrence to warrant further therapy (an increased risk of endometrial cancer and thromboses continues with continued therapy, although these risks tend to be lower for younger women). For those women who become postmenopausal after five years of tamoxifen therapy, switching to an aromatase inhibitor is likely to provide more benefit than continuation of tamoxifen.
CALOR (Chemotherapy as Adjuvant for Locally Recurrent Breast Cancer)
CALOR is another trial whose results have been a long time in coming. This trial randomized women with an isolated locoregional recurrence to receive chemotherapy or not, and demonstrated a significant improvement in disease-free survival for women receiving chemotherapy.
It is not likely to be practice-changing, though, as many oncologists in the U.S. already administer chemotherapy for at least some fraction of women with isolated locoregional recurrences, and the trial was not large enough for subset analyses to pin down exactly who does and does not benefit. The study accrued only 162 patients from 2002-2010 and had to be closed prematurely for poor accrual.
Women with an in-breast recurrence, a chest wall recurrence, and a lymph node recurrence were all included. Radiation, hormonal therapy, and HER2-directed therapies were given to both groups as specified in the protocol.
Administering chemotherapy after a completely resected isolated locoregional recurrence increased the five-year disease-free survival rate from 57 percent in the control group (no adjuvant chemotherapy) to 69 percent for those who received chemotherapy, and improved the five-year overall survival rate, from 76 to 88 percent.
Although subset analysis of such a small trial is not conclusive, the results confirm what seems reasonable: women with endocrine resistant disease—either ER-negative or ER-positive disease resistant to endocrine therapy—benefited from the addition of adjuvant chemotherapy, and those with endocrine-sensitive disease did not.
A larger trial in this population is not likely to be performed, and it is comforting that the results of this study, however small, make sense, and are consistent with current practice. Whether predictive multigene assays will be useful for women with an isolated locoregional recurrence is an open question.
Also in the category of comfortably confirming current practice (and showing that longer is not always better) were the updates from the HERA (Herceptin Adjuvant), PHARE, and NSABP31/NCCTG N9831 trials. All the analyses suggested that the current practice of administering one year of adjuvant trastuzumab for women with HER2-positive breast cancer remains the standard.
In the HERA trial, women who were disease-free at one year were randomized to stop trastuzumab therapy or to continue it for a second year. The hazard ratio for an event with two years vs. only one year of trastuzumab was 0.99, and the hazard ratio for overall survival was 1.05. Cardiac toxicity was increased with two years of therapy (the percentage of women with LVEF less than 50 percent and 10 percent below baseline, confirmed by repeat assessment, went from 4.1 to 7.2 percent).
So for adjuvant trastuzumab two years of administration was not better than one. However six months might not be as good as one year.
The PHARE (Protocol for Herceptin as Adjuvant therapy with Reduced Exposure) trial randomized women with early-stage HER2-positive breast cancer to receive six months or one year of trastuzumab, and failed to show noninferiority. This issue is not entirely closed; other trials of shorter duration will be reported in the coming years. However for now, one year of treatment remains standard.
Updated results (final planned joint analysis) from NSABP31 and NCCTG N9831, both of which randomized women receiving chemotherapy with doxorubicin/cyclophosphamide followed by paclitaxel to receive or not receive trastuzumab, showed that the benefits of one year of adjuvant trastuzumab are durable.
At a median of 8.4 years, 84 percent of women with operable HER2-positive breast cancer who were treated with trastuzumab were alive, compared with 75.2 percent of women who were not treated with trastuzumab. This is an absolute 8.8 percent reduction and a 37 percent relative reduction in the risk of death (p<0.0001). At four years the absolute difference in overall survival was 2.9 percent; at six years, it was 5.5 percent; at eight years, it was 7.6 percent; and at 10 years, it was 8.8 percent.
Whether treatment with combination anti-HER2 therapy will eventually replace treatment with single-agent trastuzumab in the adjuvant setting, and whether optimal duration of therapy will differ with different anti-HER2 agents remains to be seen.
In the “nail in the coffin” category, BEATRICE is yet another trial suggesting no benefit for the use of bevacizumab therapy in treating women with breast cancer. This was the first randomized phase III study to look specifically at early-stage triple-negative disease, and showed that a year of adjuvant bevacizumab in addition to adjuvant chemotherapy yielded no improvement in three-year disease-free survival as compared with chemotherapy alone (83.7% vs. 82.7%, p=0.18).
The initial excitement over bevacizumab for metastatic disease has waned since the E5103 study, as additional trials have failed to demonstrate as significant an improvement in DFS as E5103 did, and have failed to demonstrate any overall survival advantage. Subsequent studies have focused on the adjuvant and neoadjuvant disease settings, as it was hypothesized that bevacizumab might be more effective in the setting of low tumor burden.
The results of BEATRICE suggest that this is not likely the case. Other major adjuvant and neoadjuvant trials testing the role of bevacizumab have been completed, and their results may change the picture. But for now, there is no obvious role for bevacizumab or other angiogenesis inhibitors in any subset of breast cancer.