Rate of Onset of Response
It is often stated that combination chemotherapy has a more rapid onset of action than endocrine therapy and that this is an advantage for chemotherapy. The median time to response was 66 days for letrozole and 70 days for tamoxifen in the Letrozole 024 study; ie, half of the tumors designated as in partial or compete remission at 4 months had already met response criteria at 2 months.[13] This suggests that for some tumors at least, the response to aromatase inhibitor therapy is fairly rapid. Time to response is rarely quoted in the neoadjuvant chemotherapy literature, and thus, indirect comparisons are once again difficult to make. Long-Term Success of Neoadjuvant Treatment-Assisted Breast Conservation
Data on the long-term outcome of patients given neoadjuvant aromatase inhibitors are sparse at present, and this is certainly the major obstacle to widespread adoption of this treatment approach. Reassuring data regarding local recurrence rates are emerging from patients treated in Edinburgh, where breast-conserving surgery after neoadjuvant endocrine therapy produced excellent local control rates as long as the breast was treated with radiotherapy (Table 3).[12] Biomarker Research The molecular basis for the response of hormone-receptor-positive disease to endocrine treatments is poorly understood. Neoadjuvant endocrine therapy trials, therefore, provide a critical opportunity to conduct correlative science studies that address this question. Protein Biomarkers
The initial phase of these investigations has focused on protein biomarker levels estimated by immunohistochemistry, including the proliferation marker Ki67, tyrosine kinase-linked peptide growth-factor receptors, epidermal growth-factor receptors HER1 and HER2, and the estrogen-regulated proteins, trefoil factor 1 (TFF1), ER, and PR. Analysis of these biomarkers can be viewed from two perspectives: first as baseline predictors of response, and second as pharmacodynamic biomarkers of the effect of neoadjuvant endocrine therapy. One of the more interesting results concerning baseline predictors of response was the finding that differences in neoadjuvant response rates between letrozole and tamoxifen were particularly marked for tumors that were ERpositive and also positive for HER1 and/or HER2 by immunohistochemistry (88% vs 21%, respectively; P = .0004).[32] This suggests that tumors in this category are very sensitive to estrogen-deprivation therapy but somewhat resistant to tamoxifen. Theoretically, therefore, HER1 and HER2 analysis could be used to identify subgroups of patients in whom an aromatase inhibitor should be favored over tamoxifen as adjuvant therapy. An investigation is currently under way with tissue blocks obtained from patients who received treatment in the context of the Arimidex, Tamoxifen Alone or in Combination (ATAC) trial to further investigate this hypothesis (personal communication, Prof. M. Dowett, 2003). Given that the pathologic complete response rate is low with neoadjuvant aromatase inhibitor therapy, alternative clinical or biomarker surrogates for therapeutic efficacy must be identified. Levels of Ki67, a cell-cycle- regulated protein, typically fall with neoadjuvant endocrine treatment, and this compound was recently shown to be suppressed more effectively by letrozole than by tamoxifen in samples from the Letrozole 024 trial.[33] This finding correlates well with the clinical advantages of third-generation aromatase inhibitors in the neoadjuvant, adjuvant, and advanced disease settings[34] and supports a role for Ki67 as a surrogate end point in phase III neoadjuvant studies that aim to improve upon the efficacy of single-agent aromatase inhibitor treatment. Levels of the estrogen-regulated proteins PR and TFF1 also fall dramatically with letrozole in contrast to tamoxifen therapy, which, if anything, produces a trend for an increase in PR.[33] These observations underscore the profound differences in tumor gene regulation that take place when tumors are treated with estrogen deprivation as opposed to a selective estrogen-receptor modulator (SERM). Another finding from these investigations was that tamoxifen treatment caused profound downregulation of ERs in some tumors; in posttreatment samples, ER levels were lower in the tamoxifen arm than in the letrozole arm. These data suggest that the mechanism responsible for ER downregulation with SERM therapy is different from that of aromatase inhibitors, and ER levels should be interpreted with caution in posttreatment samples.[33] Gene Expression
Currently, gene expression profiling is under investigation both as a means of discovering new biomarkers for responsiveness to aromatase inhibitor therapy, and to better understand the molecular basis of response to these agents.[35] Limited microarray information from a responding case in an ongoing phase II investigation of letrozole can be presented to illustrate how estrogen deprivation downregulates expression from multiple genes involved in critical pathways that drive the neoplastic phenotype (Table 4). Genes showing the greatest decrease with treatment included members of a proliferation cluster (topoisomerase [DNA] II alpha (170 kD), ribonucleotide reductase M2 polypeptide, 5-methyltetrahydrofolatehomocysteine methyltransferase reductase, and cell division cycle 2, G1 to S and G2 to M, an invasion cluster (matrix metalloproteinase 1 [interstitial collagenase]), carboxypeptidase B1 (tissue), CD36 antigen (collagen type I receptor, thrombospondin receptor), protein regulator of cytokinesis 1, and a cell survival cluster (baculoviral IAP repeatcontaining 5 [survivin]), and nucleolar protein 3 (apoptosis repressor w/ CARD domain) This ongoing study is powered to explore baseline gene expression profiles to delineate "metagenes" that identify patients with the greatest chance of responding to neoadjuvant aromatase inhibitor therapy.[36] The long-term goal will be to determine if these same metagenes predict the risk of relapse in patients receiving adjuvant treatment with aromatase inhibitors. Future Prospects Currently, no ongoing studies are investigating neoadjuvant endocrine therapy with the same rigor applied to neoadjuvant chemotherapy. Ultimately, a practice-setting trial along the lines of NSABP B-18 will be necessary. A potential study design is outlined in Figure 3. Eligible patients would be stratified according to whether chemotherapy is considered necessary. Typically, these patients would be those with T2 or larger tumors, those with positive sentinel nodes, and those who require neoadjuvant neoadjuvant therapy anyway. For these patients, the randomization would be between neoadjuvant chemotherapy and an aromatase inhibitor. For patients randomized to receive neoadjuvant aromatase inhibitors, chemotherapy would be administered postoperatively so that exposure to chemotherapy was balanced between the two arms. For patients who do not require chemotherapy-typically those with sentinel node-negative T1 and small T2 tumors, or those in whom chemotherapy would be inappropriate- randomization would be between immediate surgery and neoadjuvant aromatase inhibitor therapy. Like NSABP B-18, the study could be powered to observe a survival advantage for neoadjuvant endocrine therapy. The hypothesis that neoadjuvant endocrine therapy improves survival in this setting was first proposed by B.J. Kennedy in a 1957 paper in which he and his colleagues described the remarkable efficacy of high-dose estrogen therapy for locally advanced breast cancer in older patients.[37] Later studies have suggested that high levels of circulating estrogen at the time of surgery may adversely affect long-term outcomes in premenopausal women with breast cancer and ablating ovarian function at the time of mastectomy could be therapeutic.[38-40] In postmenopausal women, the hypothesis that the endocrine milieu at the time of surgery affects longterm outcome could be easily addressed by randomizing patients to receive preoperative aromatase inhibitors or not. Given the remarkable effects of estrogen deprivation on ER-positive disease (illustrated by the dramatic changes in global gene expression and proliferation cited in this paper), this hypothesis seems increasingly tenable. Current trials are focusing on ways to further improve responses to neoadjuvant aromatase inhibitor therapy. This could be achieved either by using new predictive biomarkers to exclude patients with a poor chance of responding (as described above), or through the addition of a second agent that could enhance the antineoplastic effect of estrogen deprivation. Biologic therapies currently slated for testing in combination with aromatase inhibitors include cyclooxygenase-2 (COX-2) inhibitors as adjuvant therapy and HER1 and HER2 tyrosine kinase inhibitors, farnesyl transferase inhibitors, and mammalian target of rapamycin (mTOR) inhibitors in the advanced disease setting. NCI-C MA.27
At the time of this writing, a new randomized neoadjuvant endocrine therapy trial is being activated. The National Cancer Institute of Canada (NCI-C) has launched a US Breast Intergroup adjuvant trial to compare the nonsteroidal aromatase inhibitor anastrozole with the steroidal inhibitor exemestane. In a two-by-two randomization, patients will also receive the COX-2 inhibitor celecoxib (Celebrex) or placebo. Celecoxib appears to be potentially additive with exemestane in both treatment and prevention preclinical models.[41-43] Potential antitumor mechanisms include inhibition of prostaglandin E2, induction of aromatase, antiangiogenesis effects, inhibition of tumor invasion, and inhibition of tumor-induced inflammation and growth-factor production. ACOSOG Z1031 Trial
In a parallel clinical trial, the American College of Surgeons Oncology Group (ACOSOG) is considering the activation of a randomized placebo-controlled neoadjuvant trial in which postmenopausal patients with ER-positive surgical stage II/III breast cancer will receive 4 months of exemestane and placebo or exemestane in combination with celecoxib at 400 mg bid (the proposed schema is shown in Figure 4). This trial would provide an important new opportunity to determine whether small studies in the neoadjuvant setting consistently predict the outcome of large adjuvant trials. Conclusions The results of the ATAC trial are beginning to show an advantage for adjuvant aromatase inhibitor therapy over tamoxifen, but currently this improvement is rather modest, with an absolute reduction in relapse-free survival of only 2% to 3%.[44] If we are to make more dramatic improvements in treatment, we must deal with the problem of endocrine therapy resistance. The broader significance of neoadjuvant endocrine therapy, therefore, lies in the marriage of the new scientific possibilities of genomics and proteomics and a clinical context in which tumor profiling can be explored according to the response of ER-positive breast cancer to the simplest, oldest, and most successful breast cancer therapy of all-estrogen deprivation.