David:

I have reviewed this issue for some evidence-based authorities, so I'll provide a concise summary of my findings, that essentially argue for a duration of no less than 10 years of total extended endocrine therapy (tamoxifen plus aromatase inhibitor (AI) therapy), and preferably out to 15 years given new findings of the chronic nature of, and late metastatic development in, hormone-positive (HR+) breast cancer, 15 years for non-indolent/higher risk subgroups (including: node-positivity, large tumors, high mitotic count or high Ki-67, or intermediate-to-high Oncotype DX RS if available).

I also argue that the most authoritative guidelines, those of ASCO (see below), are not wholly in agreement with the full body of the critically appraised data, and that in fact we have more guidance that is compelling in clinical practice than often acknowledged in the literature. Note that this entails that if tamoxifen was administered for 5 years, then AI therapy should extend to no less than another 5 years, or another ten in higher risk populations, and for example if tamoxifen exposure was shorter (2 - 3 years) under current switching strategies increasingly deployed, then these numbers would be an additional 7 - 8 years, or 12 - 13 years for the higher risk group, and this is independent of achievement of pCR, unless there is evidence of acquired endocrine resistance (in which case I advise fulvestrant14, or adding an mTOR inhibitor).

And note that as of this writing, we have several in-progress clinical trials that are investigating the optimal duration of extended aromatase inhibitor therapy: MA17.R, SALSA (Secondary Adjuvant Long-term Study with Arimidex), LEAD (Letrozole Adjuvant Therapy Duration), DATA (Different Durations of Anastrozole after Tamoxifen), NSABP B-42, and SOLE (Study of Letrozole Extension)

The ASCO Guidelines

First, this issue has been examined in the latest (2014) ASCO Adjuvant Endocrine Therapy Practice Guideline1:

"If women are postmenopausal and have received 5 years of adjuvant tamoxifen, they should be offered the choice of continuing tamoxifen or switching to an aromatase inhibitor for 10 years total adjuvant endocrine therapy",

the effect of which is to reduce the risk of breast cancer recurrence, with: Evidence Quality: High, Strength of Recommendation: Strong, based on 5 RCTs: ATLAS, aTTom, NSABP B-14, the Scottish Cancer Trials Breast Group Tamoxifen Trial, the Scottish Adjuvant Tamoxifen Trial, and the ECOG Combination E4181/E5181 Trial. In addition, others trials provide some provisional supportive data: MA.17, NSABP B-33 and ABCSG 6a. Collectively, these motivate the following options:

(1) TAM-10: tamoxifen for 10 years

(2) TAM-5 + AI-5: tamoxifen for 5 years followed by aromatase inhibitors

(3) TAM+LHRH: tamoxifen with a luteinizing hormone-releasing hormone (LHRH) agonist

(4) AI+OS (Ovarian Suppression): aromatase inhibitor with an LHRH agonist or aromatase inhibitor with oophorectomy.

However note that the ASCO Guideline concludes that for postmenopausal women who start treatment with an AI, "there are no data showing that longer durations are clinically effective" . This I have argued is not quite correct, and although no trial has yet to evaluate 10 years of AI therapy, nonetheless we data have some data to suggest benefit with extended (> 5 years total AI exposure), especially in women at high risk for relapse where it may be beneficial to extend aromatase inhibitor therapy beyond the 5-year mark2.

Chronic Recurrence Risk and Late Metastases

What's critical to remember here is that women with hormone-positive early BC (HR+/EBC) are at continuous risk of relapse for up to 15 years after diagnosis, independent of administration of 5 years of adjuvant endocrine therapy3,4, and that late metastases may occur with rates of annual recurrences in ER+ breast cancer that in fact exceed those of ER-negative tumors after 5-7 years5. A further line of evidence from the multigenomic Oncotype DX assay reinforces the fact of late recurrences in endocrine-positive tumors: the annual risk of metastasis in the low- and intermediate-score group exceeds that of the high-risk score group after 10 years6,7.

The Motivation for Extended Endocrine Therapy

So although the optimal duration of extended adjuvant endocrine therapy has not been to date settled dispositively, one exploratory analysis8 found that the hazard ratio continues to decline for both disease-free survival (DFS) and distant disease-free survival (DDFS) with longer duration, suggesting the benefit of AI therapy (in this case, letrozole (Femara) increases with longer exposure timeframes.

In addition, I would argue that the 66% post-unblinding crossover rate in MA.17 from placebo to AI (letrozole) suggests that delayed initiation of an AI may be of benefit9, since at a median follow-up of 5.3 years, a significant reduction in recurrence risk and a significant 61% improvement in DDFS was observed in patients switching to AI therapy, even despite the fact that this population had more adverse prognostic factors, establishing that therapy given more than 7 years after diagnosis can change the chronic relapse risk of HR+ BC10.

Furthermore, the data sets from three relevant trials - the ABCSG 6a Trial, the NSABP B-33 Trial, and the ATENA (Adjuvant post-Tamoxifen Exemestane versus Nothing Applied) Trial - have been analyzed in an EBCTCG meta-analysis11, showing that at a median follow-up of 2.5 years, extended adjuvant AI therapy was associated with a relative reduction of 43% (absolute = 2.9%) in BC recurrence and a relative reduction of 27% (absolute = 0.5%) in BC mortality, and as the Trialists note, the magnitude of the effects seen on DFS and OS in these analyses is likely underestimated due to post-unblinding crossover.

Beginning to Identifying Low Risk Subgroups

One final observation I will offer. Ideally, we would want some differential markers that might identify any subset of HR-positive BC patients who might obtain only slight or negligible benefit from such extended adjuvant endocrine therapy, and it appears that EndoPredict (EP), as well as the intrinsic subtype PAM50 (Pro-signa) assay, and HOXB13/IL17BR test included within the Breast Cancer Index (BCI), can provide significant prognostic information on the likelihood of recurrence post-surgery by 5 years, and they appear to clearly identify low-risk subgroups with highly favorable prognoses who derive only very minimal benefit from extended AI therapy, as demonstrated in several lines of data12 including the ABCSG PAM50 ROR (Risk-of-Recurrence) Score Study13.

In sum therefore, given the long-term chronicity of endocrine disease and the well-established phenomenon of the emergence of late metastases (out to 15 years), a duration of no less than 10 years of total extended endocrine therapy (tamoxifen plus aromatase inhibitor (AI) therapy) is supported strongly by the existing data base, with a preference of out to 15 years for other than indolent/low-risk disease.

References

Thanks Konstantine, this is a very thorough and comprehensive review on extended hormonal therapy involving AIs, covering a lot of aspects on a subject that has not clearly answered yet. In my practice, the most common problem - and difficult to answer without solid data - is what to do with those patients who started initially with an AI and now reach 5 or (more difficult) 10 years (I use to extend AIs for more than 5 years if there are no complications). CM.

I agree with Dr Christos ;

Some studies , looked at whether taking Femara for 5 years AFTER taking tamoxifen for 5 years (for a total of 10 years of hormonal therapy) could lower the risk of the cancer coming back in postmenopausal women diagnosed with hormone-receptor-positive, early-stage breast cancer. The results showed that Femara:

reduced the risk of the cancer recurrence  and  reduced the risk of metastasis .