ASCO 2013 Abstract #5: aTTom

CARE Breast Cancer Faculty Canadian Perspectives from ASCO 2013

Abstract 5: aTTom: Long-Term Effects of Continuing Adjuvant Tamoxifen to 10 Years Versus Stopping at 5 Years in 6,953 Women With Early Breast Cancer. R.G. Gray, D. Rea, K. Handley, et al.

Background: Five years of adjuvant tamoxifen in estrogen-receptor-positive (ER+) early breast cancer (EBC), reduces breast cancer recurrence and death rates. Due to preliminary studies, uncertainties remained however as to whether extended tamoxifen duration beyond five years would provide additional benefit. ATLAS2 was the first large, international, randomized trial presented at SABCS 2012, which demonstrated the additional benefit of extended duration tamoxifen from five to ten years. A similarly designed UK-LED study aTTom2 was the subject of much anticipation and was presented at the plenary session at ASCO 2013. 

Methods: During 1991–2005, 6,953 women with ER+ (n = 2,755), or ER untested (4,198, estimated 80% ER+ if status known) invasive EBC from 176 UK centers were randomized after five years of tamoxifen, to stop tamoxifen or to continue on to ten years.

Results: Allocation to continued tamoxifen use to ten years reduced breast cancer recurrence overall and was reported to be time dependent and only significant after year six. Longer duration treatment also reduced breast cancer specific mortality and again time dependent, with significance noted only after year nine.  Similar benefits were seen with overall mortality and no major increase was seen in non-breast cancer mortality. As expected, increased rates of endometrial cancers were observed in the ten-year duration group.

The aTTom1 authors reported that combining its similarly positive results to its international counterpart ATLAS2 enhances statistical significance of longer duration tamoxifen (ten vs. five years) with respect to recurrence (p < 0.0001), breast cancer mortality (p = 0.002) and overall survival (p = 0.005) benefits. Taken together with the reduction in breast cancer deaths seen in trials of five years of tamoxifen vs. none, these results indicate that ten years of adjuvant tamoxifen, compared to no tamoxifen, reduces breast cancer mortality by about one third in the first ten years following diagnosis and by a half subsequently thereafter.

How to interpret and implement this historical data and how to correctly identify those patients most likely to benefit from the extended tamoxifen strategy in the clinical setting today is the challenge at hand.

Although the largest cohorts in the trial were postmenopausal, the greatest uptake for adjuvant ten year tamoxifen will likely be seen in the ER+ premenopausal subgroup where no evidence based, extended duration endocrine strategy was previously available in the EBC stetting.

For the ER+ postmenopausal subgroup, having exposure to an aromatase inhibitor (AI) at some point in the adjuvant setting (upfront, switch or extended duration strategy) is of additional advantage (particularly in the higher risk, lymph node positive subset) compared to tamoxifen alone for five years. In those postmenopausal women with AI contraindications or AI intolerance, the ten year adjuvant tamoxifen strategy appears to be a new and reasonable treatment option.

Despite the question of tumor dormancy and the eagerness to extend any perceived benefit, there is no level one evidence at this time, to support the use of either (a) ten years of adjuvant AIs, (b) tamoxifen after completion of five years of an AI, or (c) indefinite adjuvant endocrine therapy in the EBC setting. The risk of endometrial cancer and other extended duration side effects also needs to be carefully considered and balanced by both the oncologist and patient alike.

Lastly, given the difficulty in patient treatment compliance due to side effects or “pill fatigue” with even five years of adjuvant endocrine therapy, efforts to further improve the therapeutic ratio of the adjuvant ten year tamoxifen strategy will be paramount. Additional studies to reliably define the subsets of patients who benefit clinically (e.g., those with lymph node positive disease3) or with additional molecular refinement of prognostic4 or predictive factors are needed.
— CARE Breast Cancer Faculty