Adjuvant Hormonal Therapy in Premenopausal Women With Operable Breast Cancer: Not-So-Peripheral Perspectives
Adjuvant Hormonal Therapy in Premenopausal Women With Operable Breast Cancer: Not-So-Peripheral Perspectives
ABSTRACT: Reviews of issues around adjuvant hormonal therapies for breast cancer in premenopausal women often focus on recent and current large clinical trials, and fail to address other subjects that are very germane to evidence-based and investigatory clinical practice. These topics include: (1) the descriptive epidemiology of breast cancer globally, (2) critical issues in tumor hormone receptor testing, (3) compelling data demonstrating that hormone receptor–positive breast cancer is a chronic disease, (4) data supportive of combined hormonal therapy with tamoxifen as the standard of care, and the limited justifications for awaiting the SOFT and TEXT trial results, (5) pharmacogenetic hypotheses with tamoxifen, (6) ethical issues in ovarian suppression vs ablative treatment, and (7) emerging data about the importance of primary tumor removal surgery itself and “surgical stress” in solid tumor management.
While optimal adjuvant hormonal therapies for premenopausal women with operable breast cancer have yet to be defined, discussions and reviews of the state of the art and “areas of confusion” often fail to consider developments that are germane to keeping evidence-based clinical practice truly up-to-date. The current communication is prompted by this perspective and a recent review and its commentaries.
The late Jonathan Mann often said that the way we frame issues dictates how we approach them. Framing the challenges of getting to more effective adjuvant therapies for premenopausal women with hormone receptor–positive tumors within the context of our most recent and ongoing larger clinical trials only—which is what is usually done—is to ignore the richness and relevance of other provocative and emerging data that are directly applicable to clinical and investigative practice now. Here, I review such data in seven areas.
Descriptive Epidemiologic Data of Breast Cancer Globally
There are several reasons why we need to move away from perspectives and data based on North America only (or high-income countries only) in discussing breast cancer management. First, the cancer treatment and reading community is global, and we are increasingly called upon to be global citizens and speak to the needs of patients everywhere. In 2010, we will move to situations in which, worldwide, the majority of new annual cases of breast cancer will develop in Asian women (~800,000 of 1.5 million) and half will be in poor premenopausal women (~740,000 vs 44,000 premenopausal cases in the United States).[3,4]
Further, the fact that the overwhelming majority of our treatment data come from studies in women of northern European genetic background—with likely very specific tumor gene profiles and certainly different metabolic gene profiles—make the available data of uncertain relevance to this new majority of affected women who live in non-Western countries, as well as those women in Western countries of different genetic/ethnic backgrounds.[5-7] Finally, a common perception is that hormone receptor–positive breast cancer is less frequent in pre- than in postmenopausal women, but data from the Philippines, Vietnam, Taiwan, and China do not support this general conclusion.[8,9, and personal communications from Zhi Ming Shao, November 7, 2009; and from Ta Van To, May 2009]
In sum, our discussions about breast cancer management need to be more broadly sensitive and considerate of the global realities.
Tumor Hormonal Receptor Testing
While for some time there have been expressions of concern regarding quality control issues surrounding tumor hormonal receptor testing, the implications for practicing clinicians (and possible remedial actions) have not been obvious. With the upcoming publication of the American Society of Clinical Oncology (ASCO)–College of American Pathologists (CAP) guidelines on hormonal receptor quality assurance, this situation should change. In the meantime, various data regarding one broad issue, which will be addressed in the guidelines, deserve all clinicians’ attention: choices of tissue specimens and their management prior to laboratory testing.
The relevant data include time to penetration and fixation of subsequently tested tumor tissues (optimally < 30 minutes), pH of fixative (optimally neutral, not acidic), and duration of fixation (optimally > 8–10 hours; less critically < 48 hours). Each of these factors influence determination of the presence and levels of hormonal receptor proteins.[10-12] Inattention to these parameters leads to more frequent findings of hormone receptor–negative tumors and lower levels of hormonal receptor proteins. When hormonal receptor determinations are done on core biopsy specimens, approximately 10% more tumors are found to be hormone receptor–positive, compared with when tests are done on subsequent mastectomy (and lumpectomy?) specimens.[13,14] The implications of these findings are clear:
• False-negative findings of hormonal receptor protein lead to depriving patients of important, recurrence-preventing hormonal therapies and in many circumstances choices of usually more toxic chemotherapy treatments.
• Clinicians need to be involved in the complete management sequence of tissue specimens obtained when hormonal receptor testing is part of the diagnostic panel.
• The diagnostic sequence for breast masses may, in some circumstances, need reconsideration. When the sequence includes a fine-needle aspiration biopsy, and hormonal receptor determinations are then done on subsequent mastectomy (as practiced in most of the world) or lumpectomy specimens, even greater attention to tissue management procedures is warranted. The case for core biopsy as the first diagnostic procedure, with this specimen used for hormonal receptor testing and with associated attention to optimal tissue management, deserves renewed consideration.
Hormone Receptor–Positive Breast Cancer as a Chronic Disease
Data from multiple sources are reinforcing what clinicians have been aware of for some time, but this awareness has not yet completely translated into thoughtful clinical practice and investigative medicine. For example:
• Measurable and steady rates of recurrence characterize meta-analysis populations of patients with hormone receptor–positive breast cancer (more so than those with hormone receptor–negative tumors) through 15 years after diagnosis.[15,16] There is a lag in return to higher rates of recurrence in the immediate years after hormonal therapies are stopped (eg, ~5 years).[15,17-19]
• In cases where hormonal therapies are given for longer than 5 years or started after 5 years from diagnosis, lower rates of recurrence occur in the 5- to 10-year postdiagnosis window.[17,20-22]
Current National Comprehensive Cancer Network (NCCN) guidelines suggest the use of adjuvant hormonal therapy with an aromatase inhibitor after 5 years (ie, for postmenopausal women, a status that all premenopausal women can achieve with chemotherapies, ovarian ablation, or continuing suppression therapies). However, there does not appear to be a consensus on this recommendation, at least as manifested in recommendations for therapies after 5 years in ongoing adjuvant studies. Clearly, there are many uncertainties about risks and benefits for subsets of patients, but the clear conclusion that hormone receptor–positive breast cancer is, for many (and perhaps the majority of patients), a chronic disease, must command more of our collective attention. Part of the reticence to more frequently and forthrightly consider this issue comes from the fact that patients find this to be an upsetting perspective, particularly because this is not how the disease has been framed in the past.
Combined Ovarian Suppression/Ablation and Tamoxifen Therapy as the Standard of Care
In 2003–2004, it may have been reasonable to assert that tamoxifen alone was the hormonal therapy standard of care for premenopausal women with hormone receptor–positive tumors. This was the conclusion of the Suppression of Ovarian Function Trial (SOFT) investigators, who assessed tamoxifen vs ovarian function suppression or ablation plus tamoxifen vs ovarian function suppression or ablation plus an aromatase inhibitor (exemestane [Aromasin]), giving each therapy for 5 years. Many in the research community have continued to maintain that tamoxifen alone is the standard of care in this setting.
Austrian investigators reporting on the issue of ovarian suppression plus tamoxifen or an aromatase inhibitor (trial discussed below), whose study began accrual in 1999, apparently did not consider tamoxifen alone to be the standard of care a decade ago. In 2009–2010, however, it is neither reasonable nor appropriate to assert (1) such equivalence of tamoxifen alone and combined therapy, and (2) to call for continued accrual to SOFT and the Tamoxifen/Exemestane Trial (TEXT), withholding judgment on the role of aromatase inhibitors in combined therapy until these trials report their results. The following findings support these contentions:
• In metastatic hormone receptor–positive disease, four individual trials and a meta-analysis have demonstrated improved outcomes with combined ovarian suppression plus tamoxifen therapy over either therapy alone.[24,25]
• In a large Intergroup trial, luteinizing hormone-releasing hormone (LHRH) alone after CAF chemotherapy (cyclophosphamide, doxorubicin [Adriamycin], fluorouracil [5-FU]), was inferior to LHRH plus tamoxifen (disease-free survival difference at 9 years = 8%; overall survival difference at 9 years = 3%). In the metastatic setting, LHRH alone and tamoxifen alone appear to be equivalent therapies.
• Meta-analysis of adjuvant data suggests that the combination of LHRH plus tamoxifen is better than tamoxifen alone.
• In the Intergroup adjuvant trial, oophorectomy plus tamoxifen produced a disease-free survival rate of 90.3%, compared to 87.8% for tamoxifen alone.
• In two European adjuvant trials, LHRH plus tamoxifen was superior to six cycles of IV CMF chemotherapy (cyclophosphamide, methotrexate, 5-FU), while LHRH alone was equivalent to six cycles of Bonadonna CMF chemotherapy.[30,31]
• In the author’s adjuvant trial of oophorectomy plus tamoxifen, this strategy resulted in a risk reduction of 0.58; in the meta-analysis of trials assessing adjuvant tamoxifen alone in premenopausal women, the risk reduction was 0.42.[15,19]
The consistency of the evidence—although mostly indirect—and the logic that two mechanisms of action are functioning with combined therapy, make the superiority of combined ovarian suppression plus tamoxifen therapy difficult to ignore.
The repeated counterargument is that direct evidence is needed, and the SOFT and TEXT trials will provide this. For the two hypotheses under investigation in those trials, this argument deserves careful scrutiny. However, compelling direct evidence is already available. Gnant et al presented survival data from the Austrian Breast and Colorectal Cancer Study Group (ABCSG)-12 trial comparing LHRH plus anastrozole (Arimidex) and LHRH plus tamoxifen with or without zoledronic acid (Zometa) in premenopausal women with hormone receptor–positive breast cancer. A 2×2 factorial design was used (1:1:1:1), with patients randomized to LHRH plus either anastrozole or tamoxifen, with or without zoledronic acid. The study enrolled 1,803 patients and was designed to test two primary hypotheses for the outcome of disease-free survival: (1) anastrozole against tamoxifen, and (2) zoledronic acid against no zoledronic acid.
While the addition of zoledronic acid exhibited a significant benefit in terms of disease-free survival, no difference in disease-free survival was found between the anastrozole and tamoxifen groups (P = .59, hazard ratio [HR] = 1.10, 95% confidence interval [CI] = 0.78–1.53). A similar pattern was seen for recurrence-free survival (P = .53, HR = 1.11, 95% CI = 0.80–1.56), and a trend for overall survival was found in favor of tamoxifen (P = .07, HR=1.80, 95% CI: 0.95–3.38). While a failure to reject the null hypothesis in a trial designed to test for superiority does not allow us to conclude the treatments are equivalent, there is no evidence from these data to support the superiority of anastrozole over tamoxifen in this population. However, given the size of this trial, the maturity of the data (median follow up of 48 months), and the pattern and trend of results, there is little justification for the position that the superiority of anastrozole compared with tamoxifen after ovarian ablation or with ovarian suppression alone remains an open question.
Thousands of further patients and years of follow-up will be required to demonstrate even a small effect, and these Gnant results would have to be considered in reaching conclusions about the “true” comparison results. Both the SOFT and TEXT trials also allow chemotherapy treatment. Looking to these trials for a different answer to the question of aromatase inhibitor or tamoxifen superiority in premenopausal women whose ovarian function is stopped, is neither appropriate nor realistic. While a definitive conclusion cannot be drawn from the Gnant study in terms of the survival benefits of anastrozole over tamoxifen in the premenopausal setting, a reasonable decision is to proceed as though the treatment results are similar.