Fax: (713) 794-4385
Summary of aromatase inhibitor clinical trials in postmenopausal women with early breast cancer†
Article first published online: 10 DEC 2007
Copyright © 2007 American Cancer Society
Volume 112, Issue S3, pages 700–709, 1 February 2008
How to Cite
Buzdar, A. U., Coombes, R. C., Goss, P. E. and Winer, E. P. (2008), Summary of aromatase inhibitor clinical trials in postmenopausal women with early breast cancer. Cancer, 112: 700–709. doi: 10.1002/cncr.23193
Presented at Endocrine and Targeted Manipulation of Breast Cancer: Proceedings of the Sixth Cambridge Conference, Cambridge, Massachusetts, April 30-May 1, 2007.
- Issue published online: 18 JAN 2008
- Article first published online: 10 DEC 2007
- Manuscript Accepted: 11 OCT 2007
- Manuscript Revised: 2 OCT 2007
- Manuscript Received: 30 JUL 2007
- Bristol-Myers Squibb
- Eli Lilly
- aromatase inhibitor;
- breast cancer;
- endocrine therapy;
- clinical trial
Five years of adjuvant therapy with tamoxifen was considered the gold-standard treatment for postmenopausal women with estrogen receptor-positive breast cancer for many years. Data from a core group of clinical trials investigating the safety and efficacy of aromatase inhibitors (AIs) have challenged this perception. These studies were designed to evaluate the safety and efficacy of AIs in the following clinical settings: 1) as initial adjuvant therapy (the Arimidex, Tamoxifen, Alone or in Combination trial, Breast International Group Trial 1-98), 2) in a “switched setting” after 2 to 3 years of treatment with tamoxifen (Arimidex-Nolvadex 95, the Austrian Breast and Colorectal Cancer Study Group 8 [ABCSG 8] trial, the Italian Tamoxifen Anastrozole study, the Intergroup Exemestane Study), and 3) in extended settings (National Cancer Institute of Canada Trial MA.17, ABCSG 6a, National Surgical Adjuvant Breast and Bowel Project 33). The efficacy data from these studies suggested that AIs have added substantial benefit in terms of disease outcome. AIs were tolerated well, and patients who received them experienced fewer thrombolic events and less endometrial cancer, hot flashes, night sweats, and vaginal bleeding compared with patients who receive tamoxifen. However, patients who received tamoxifen had less skeletal events and accelerated bone resorption compared with women who received AIs. AIs should be considered when planning a patient's endocrine therapy, taking into account the differences in tolerability and end-organ effects of the classes of endocrine therapy. Outstanding issues to optimize AI therapy include identifying the optimal duration, agent, and patients for these therapies. Cancer 2008. © 2007 American Cancer Society.
Five years of adjuvant therapy with tamoxifen was considered the gold-standard treatment for postmenopausal women with estrogen receptor (ER)-positive breast cancer for many years. However, data from a core group of clinical trials investigating the safety and efficacy of aromatase inhibitors (AIs) have challenged this perception. These studies were designed to evaluate AIs in the following clinical settings: 1) as initial adjuvant therapy (the Arimidex, Tamoxifen, Alone or in Combination [ATAC] study, Breast International Group [BIG] study 1-98), 2) in a “switched setting” after 2 to 3 years of treatment with tamoxifen (Arimidex-Nolvadex [ARNO] trial 95, the Austrian Breast and Colorectal Cancer Study Group 8 trial [ABCSG 8], the Italian Tamoxifen Anastrozole (ITA) trial, the Intergroup Exemestane Study [IES]), and 3) in extended settings (National Cancer Institute of Canada Trial MA.17, ABCSG 6a, and National Surgical Adjuvant Breast and Bowel Project 33 [NSAPB-33]). A summary of these studies, including study designs, results to date, interpretations of results, and the impact on patients, are provided below. In addition, questions that were left unanswered by these trials are discussed.
Initial Adjuvant Therapy
The ATAC trial assessed the efficacy and tolerability of the AI anastrozole versus tamoxifen as initial adjuvant treatment for postmenopausal women with hormone receptor-positive disease (Table 1). At a median follow-up of 68 months, anastrozole significantly prolonged disease-free survival (DFS) and the time to recurrence compared with tamoxifen (DFS: hazards ratio [HR], 0.87; 95% confidence interval [95% CI], 0.78-0.97 [P = .01]; time to recurrence: HR, 0.79; 95% CI, 0.70-0.90 [P = .0005]).1 Distant recurrences and contralateral breast cancers also were reduced (distant recurrences: HR, 0.86; 95% CI, 0.74-0.99 [P = .04]; contralateral breast cancers: 42% reduction, P = .01). These differences favoring anastrozole were observed irrespective of baseline prognostic factors. An absolute difference in DFS was evident within 1 year and continued to increase beyond completion of treatment. The safety and risk benefit analysis of ATAC data showed better tolerability and risk benefit profiles for anastrozole over tamoxifen.2 Treatment-related adverse events (AEs) occurred significantly less often with anastrozole (61%) compared with tamoxifen (68%; P < .0001); serious AEs and withdrawals because of AEs also were significantly fewer with anastrozole. The risk of life-threatening AEs, such as venous thromboembolic events, stroke, and endometrial cancer, all were significantly lower with anastrozole.1–3 Bone fractures were more common with anastrozole than with tamoxifen (11% vs 8%; P < .0001): Hip fractures occurred at a similarly low incidence in both treatment groups.2 Data from a substudy of this trial suggest that, although anastrozole was associated with significant decreases in bone mineral density (BMD) in the lumbar spine and total hip compared with tamoxifen,4 no patients with normal BMD at baseline became osteoporotic at 5 years. The ATAC trial endometrial subprotocol examined the effect of these agents on endometrial pathology. At 6 years of follow-up, patients who received anastrozole had fewer endometrial abnormalities than patients who received tamoxifen (27.3% vs 43.9%; P = .017).5 Furthermore, in the main ATAC trial, the differences in gynecologic AEs resulted in a 4-fold increase in the number of hysterectomies in patients who received tamoxifen compared with patients who received anastrozole (5.1% and 1.3%, respectively; P < .0001).2 The impact of both agents on health-related quality of life (QoL) was similar.6
|Setting/trial*||No. of randomized patients||Treatment arms||Primary/secondary endpoints||Main findings|
|Initial adjuvant therapy|
|ATAC1–6||9366||Anastrozole (n=3116), tamoxifen (n=3135), combination (n=3125)||Primary: DFS, safety/tolerability; secondary: incidence of contralateral cancer, time to distant recurrence, OS||At 33-mo follow-up, combination arm showed no benefits over monotherapy and was discontinued; at 68-mo follow-up, there were significant advantages for anastrozole over tamoxifen in DFS, time to recurrence, and contralateral breast cancer|
|Initial adjuvant and sequential|
|BIG 1-987, 8||8028||Tamoxifen (n=2459), letrozole (n=2463), tamoxifenletrozole (n=1550), letrozoletamoxifen (n=1545)||Primary: DFS; secondary: OS, systemic DFS, time to distant recurrence||Initial adjuvant analysis (monotherapy tamoxifen vs. letrozole): 1) letrozole significantly increased DFS, reduced distant recurrences, prolonged time to distant metastasis; 2) tamoxifen group experienced more thromboembolism, endometrial cancer, hot flashes, night sweats, and vaginal bleeding; 3) patients on letrozole experienced more skeletal and cardiac events and hypercholesterolemia. Analysis of sequential treatment arms: data not yet available|
|ARNO 959||1059||Randomized after 2 y of tamoxifen: Anastrozole (n=489) or tamoxifen (n=490)||Primary: DFS, local or distant recurrence, contralateral breast cancer, death; secondary: OS, safety||1) Anastrozole treatment significantly reduced the risk of recurrence and improved OS; 2) the incidence of serious AEs was lower in the anastrozole group; 3) no new safety issues identified|
|ABCSG 810||2926||Randomized after 2 y of tamoxifen: Anastrozole (n=1472) or tamoxifen (n=1454)||Primary: EFS; secondary: distant recurrence-free survival, OS, tolerability||1) Switching to anastrozole showed a significantly reduced risk of recurrence, longer duration of DFS and EFS, and improved OS; 2) thromboses, embolism, and endometrial cancer were more common in patients continuing on tamoxifen, whereas serious fractures were more likely in those switching to anastrozole|
|ITA11||448||Randomized after 2-3 y of tamoxifen: Anastrozole (n=223) or tamoxifen (n=225)||Primary: DFS; secondary: EFS, OS, safety||1) DFS and EFS were significantly longer in the anastrozole group; 2) more patients in the anastrozole group experienced at least 1 AE, but numbers of patients experiencing serious AEs were comparable|
|Meta-analysis of ARNO 95, ABCSG 8, and ITA12||4006||Anastrozole (n=2009) or tamoxifen (n=1997)||Primary: DFS, EFS, distant-recurrence-free survival, OS; secondary: AEs||There were significant improvements in all efficacy endpoints for anastrozole over tamoxifen|
|IES13–15||4724||Randomized after 2 to 3 y of tamoxifen: Exemestane (n=2362) or tamoxifen (n=2380)||Primary: DFS; secondary: OS, Safety, BMD, biomarkers of bone metabolism, fracture incidence, QoL tools (FACT-B, ES)||1) Exemestane treatment after 2-3 y of tamoxifen significantly improved DFS compared with 5 y of tamoxifen treatment; 2) No significant difference between the 2 treatment groups for OS, but improved OS if ER-negative patients excluded; 3) both treatment groups well tolerated with rare severe toxic effects; 4) BMD was decreased within 6 mo of switching to exemestane, no patient with normal BMD at entry became osteoporotic; 4) fracture incidence significantly greater in the exemestane group; 5) QoL was good and stable for 2 y with no clinically meaningful differences observed between the 2 groups|
|Extended adjuvant therapy|
|MA.1716||5187||Randomized after 5 y of tamoxifen: Letrozole (n=2593) or placebo (n=2594)||Primary: DFS; secondary: OS, QoL tools (SF-36, MENQOL), lipids, BMD||1) Letrozole significantly improved DFS survival after 5 y of tamoxifen; 2) AEs (primarily grade 1 and 2): hot flashes, arthritis, arthralgia, and myalgia more common in letrozole than placebo group; trend toward higher rate of newly diagnosed osteoporosis in letrozole group|
|NSABP B-3317||1598||Randomized after ∼5 y of tamoxifen: Exemestane or placebo||Primary: DFS; secondary: OS, safety, QoL||Women who received exemestane after 5 y of tamoxifen therapy were 56% less likely to have relapse of breast cancer|
|ABCSG 6a18||856||Randomized after 5 y of tamoxifen with or without AG: anastrozole (n=387) or no treatment (n=469)||Primary: recurrence-free survival; secondary: OS, safety||1) Patients receiving anastrozole experienced a significantly reduced risk of recurrence; 2) anastrozole was well tolerated, no unexpected AEs were reported|
The BIG 1-98 study is an ongoing, Phase III, international, multicenter, randomized, double-blind trial that compares the safety and efficacy of 5 years of treatment with 4 different adjuvant endocrine therapy regimens (Table 1). Patients were randomized to receive 1 of the following daily monotherapy or sequential treatments for 5 years: 1) tamoxifen, 2) letrozole, 3) tamoxifen for 2 years followed by letrozole for 3 years, or 4) letrozole for 2 years followed by tamoxifen for 3 years. The objectives of this trial were to compare adjuvant letrozole versus tamoxifen administered for 5 years and to compare these treatment regimens given sequentially versus continuously. Efficacy was measured as the proportion of patients who had DFS at yearly intervals after randomization. Secondary efficacy parameters are summarized in Table 1.
In total, 8028 women were randomized into this trial, as outlined in Table 1. To date, 2 planned analyses of the data have been reported. The BIG 1-98 collaboration group first reported on the comparison of tamoxifen with letrozole at a median follow-up of 25.8 months.7 That analysis included all data from patients who were assigned to the monotherapy arms and data from the sequential therapy arms that was censored at the time of therapy switch. In the analysis, compared with tamoxifen, letrozole significantly reduced the risk of an event (invasive recurrence, second malignancy, or death without prior cancer event) ending a period of DFS (HR, 0.81; 95% CI, 0.70-0.93 [P = .003]), particularly the risk of distant recurrence (HR, 0.73; 95% CI, 0.60-0.88 [P = .001]). With regard to safety, patients who were receiving tamoxifen experienced more thromboembolism, endometrial cancer, and vaginal bleeding. Patients who were receiving letrozole experienced more skeletal events (bone fractures, arthralgia), hypercholesterolemia, and cardiac events. A second analysis of the data was conducted with a median of 51 months of follow-up data and included only those patients who had been randomized to the monotherapy treatment arms.8 Similar to the first analysis, a greater proportion of patients in the letrozole group remained alive and disease free at 5 years (HR, 0.82; 95% CI, 0.71-0.95 [P = .007]). The safety profile was similar to that reported in the first analysis. These longer term data confirmed the safety and efficacy findings from the first analysis. Letrozole was superior to tamoxifen in the defined primary endpoint, DFS, and in the secondary efficacy endpoints, including the time to distant recurrence.
To date no statistical comparison of the sequential treatment groups with one another or with monotherapy have been conducted. However, the data to date regarding the monotherapy arms suggest that patients can benefit from receiving letrozole for a period of 5 years without suffering from increasing adverse sequelae.
Switching Endocrine Therapy
Other trials, such as ARNO 95, ABCSG 8, and ITA, have evaluated the efficacy benefits of switching to anastrozole in patients who have received adjuvant treatment with tamoxifen for 2 to 3 years (Table 1).9–11 These trials evaluated the benefits of switching to anastrozole compared with continuing tamoxifen for the remainder of the 5-year treatment period. Switching to anastrozole produced a significantly reduced risk of recurrence,9 longer duration of DFS and event-free survival,10, 11 and improved overall survival.9 A recent meta-analysis of these 3 clinical trials demonstrated that switching from tamoxifen to anastrozole provided significant improvements in both DFS (HR, 0.59; 95% CI, 0.48-0.74 [P < .0001]) and overall survival (HR, 0.71; 95% CI, 0.52-0.98 [P = .04]).12 In these switching trials, it is important to distinguish between the switching and sequencing strategies, because the resulting data relate to different patient populations, and switching study populations are enriched with patients who respond well to endocrine therapy by excluding patients who have had an early recurrence despite tamoxifen treatment.
The IES trial was designed as an international, multicenter, randomized, double-blind trial to compare the efficacy and safety of continued adjuvant tamoxifen therapy with exemestane therapy in postmenopausal women with primary breast cancer who remained free of disease after receiving adjuvant tamoxifen for 2 to 3 years. Women were randomized to receive either exemestane or tamoxifen daily to complete 5 total years of treatment. The objectives of the study were 1) to determine whether switching to exemestane, a steroidal AI, after 2 to 3 years of tamoxifen could improve DFS compared with continuing tamoxifen for the remainder of the 5 year treatment period; 2) to determine whether using both treatments in sequence would produce improved efficacy and fewer side effects than either treatment alone; 3) to determine whether switching to an AI would reduce the tamoxifen-associated incidence of thromboembolism and endometrial cancer; 4) to assess the effects of either treatment on BMD, as measured by dual-energy x-ray absorptiometry, biomarkers of bone metabolism (substudy), and fracture incidence; and 5) to determine whether switching to an AI would have an impact on the patient's QoL.13
In total, 4724 postmenopausal patients with unilateral, invasive breast cancer were assigned randomly to switch to exemestane or to continue tamoxifen treatment for the remainder of a 5-year endocrine treatment period. The primary endpoint was DFS (Table 1).
After a median follow-up of 56 months, 809 DFS events were reported (354 exemestane-related events, 455 tamoxifen-related events). The unadjusted HR for DFS was 0.76 (95% CI, 0.66-0.88; P = .0001), favoring exemestane with an absolute benefit of 3.3% (95% CI, 1.6-4.9%) by the end of treatment (ie, 2.5 years after randomization). A reduction in the risk of death was observed after a switch to exemestane (222 deaths) compared with continuation of tamoxifen (261 deaths), corresponding to an HR of 0.85 (95% CI, 0.71-1.02; P = .08) and 0.83 (95% CI, 0.69-1.00; P = .05) when patients with ER-negative disease were excluded. Exemestane and tamoxifen were tolerated well, and severe toxic effects were rare.14
In the first 6 months after switching to exemestane, BMD was lowered by 0.051 g/cm3 (2.69%; 95% CI, 2.0-3.4% [P < .0001]) at the lumbar spine and by 0.025 g/cm3 (1.35%; 95% CI, 0.8-1.9% [P < .0001]) at the hip compared with baseline BMD. The decline in BMD progressively slowed thereafter. Reductions in BMD were 1.04% (95% CI, 0.39-1.70%; P = .002) and 0.84% (95% CI, 0.29-1.39%; P = .003) in Year 2 at the lumbar spine and hip, respectively. No patient with normal BMD at trial entry developed osteoporosis. Bone resorption and formation markers increased at all time points in the women who received exemestane (P = .001). With a median follow-up of 58 months for all IES participants (n = 4274 women), 162 patients (7%) in the exemestane group and 115 patients (5%) in the tamoxifen group had fractures (odds ratio, 1.45; 95% CI, 1.13-1.87 [P = .003]). These results indicate that the increase in survival reported previously with the IES switch strategy was achieved at the expense of some detriment to skeletal health; thus, the risk-benefit ratio for women needs to be assessed individually.15
A subset of IES centers participated in a QoL substudy. The Functional Assessment of Cancer Therapy-Breast (FACT-B) and the Endocrine Subscale (ES) were administered before randomization and at 3 months, 6 months, 9 months, 12 months, 18 months, and 24 months. The primary endpoint was the FACT-B composite Trial Outcome Index (TOI). Secondary endpoints included the total FACT-B plus ES score, the total ES score, and the severity of individual endocrine symptoms. Five hundred eighty-two patients from 8 countries were enrolled into the substudy. Eighty-five percent of the questionnaires were completed and returned for analysis. QoL generally was good and stable for >2 years, and no clinically meaningful differences were observed between the groups in scores on the TOI, FACT-B, or ES at any time point. No significant differences between groups were observed for any endocrine symptoms apart from vaginal discharge, which was more pronounced in women who were receiving tamoxifen (P < .002). The switch from tamoxifen to exemestane neither increased nor decreased endocrine symptoms that were present after 2 to 3 years of tamoxifen; the switch also did not initiate significant reports of new symptoms. These results indicate that the clinical benefits of exemestane over tamoxifen are achieved without significant detrimental effect on QoL.16 We conclude that the previously reported early DFS improvements observed after a switch to exemestane after 2 to 3 years of tamoxifen persist posttreatment and translate into a modest, but real, improvement in overall survival.
Extended Adjuvant Therapy
Recurrences of breast cancer continue beyond 5 yearsof standard adjuvant tamoxifen therapy, and approximately half are distant recurrences with known high clinical and financial impact. Two-thirds of breast cancer deaths occur after the completion of adjuvant tamoxifen. Extending tamoxifen beyond 5 years adds little, if any, benefit; therefore, an unmet medical need exists for extended adjuvant endocrine therapy.
Data from 3 major trials evaluating the potential benefits of extending adjuvant therapy beyond 5 years of tamoxifen are available from the MA.17, ABCSG 6a, and NSABP B-33 trials. The results of the large, double-blind, placebo-controlled MA.17 clinical trial provided evidence that extending therapy with letrozole in the extended adjuvant therapy setting provides significant additional benefit. Extended adjuvant letrozole significantly reduced the risk of recurrence by 42% (P = .00,004) and the risk of distant recurrence by 40% (P < .0001), thus corroborating the evidence for an ongoing risk of recurrence. A prospective, planned analysis revealed a 39% reduction in the risk of death for lymph node-positive patients who received letrozole (P = .04). A retrospective cohort analysis further revealed that the advantage of letrozole over placebo increased with duration of treatment. In the overall population, the HRs for DFS and distant DFS decreased significantly between 6 months and 48 months of treatment with letrozole (HR for DFS, 0.52-0.19; P < .0001; HR for distant DFS, 0.66-0.24; P = .0004). In lymph node-positive patients, the significant decrease in the risk of distant metastases also translated into a significantly decreasing HR for the risk of death (HR for overall survival, 0.72-0.43). Furthermore, the efficacy of late extended adjuvant letrozole was assessed in patients who switched from placebo to letrozole after unblinding of MA.17. This postunblinding analysis indicated that there were significant reductions in the risk of recurrence (−69%; P < .0001) and distant recurrences (−72%; P = .002) in patients who had been off prior tamoxifen for periods between 1 year and 5 years. No interaction between age and benefit from letrozole was noted in an analysis of outcome in the elderly.17 In summary, extended adjuvant letrozole provides substantial benefit in women who have completed prior tamoxifen and should be considered for such women regardless of lymph node status, prior adjuvant chemotherapy, or age.
The NSABP B-33 study was a randomized, placebo-controlled, double-blind trial that evaluated the safety and efficacy of exemestane given for 5 years in cancer-free patients, hormone receptor-positive breast cancer survivors after approximately 5 years of tamoxifen therapy. The accrual of this trial was stopped in October 2003 after the results from the MA.17 trial were disclosed. At that time, approximately 50% of the patients had been enrolled (1598), and the study was unblinded. The data reported at the San Antonio Breast Cancer Symposium in December 2006 on 30 months of median follow-up from the study suggested that women who received exemestane were 56% less likely than those who received placebo to have a recurrence of breast cancer (P = .004). However, no overall survival benefit was observed (HR, 1.2; P = .63). The most common grade 3 and 4 toxicities observed in the exemestane group were arthralgia (1%), fatigue (0.9%), and bone pain (0.5%).18
In study ABCSG 6a, patients who were recurrence-free at 5 years after tamoxifen therapy with or without aminoglutethimide were rerandomized to receive either 3 years of anastrozole or no further treatment (Table 1).19 At 62.3 months of follow-up, patients who received anastrozole experienced a significantly reduced risk of recurrence (HR, 0.62; 95% CI, 0.40-0.96 [P = .031]) compared with patients who received no further treatment.
AIs have added benefit to the outcome of women with hormone receptor-positive breast cancer over and above that conferred by tamoxifen alone. Until further data are available from long-term, randomized, controlled clinical trials, newly diagnosed patients should consider either an upfront AI or a strategy that includes an initial course of tamoxifen followed by an AI. Two important questions in this area currently are being addressed in large, randomized, controlled trials. The first is whether there is an identifiable subgroup of patients for which tamoxifen should be considered more strongly for initial therapy. The second question is whether initial AI treatment is superior to a switched or sequential regimen with both an AI and tamoxifen. It is hoped that the BIG 1-98 trial will provide data to further our understanding in this area. (Note that one of the authors [A.U.B.] considers upfront AI as the preferred treatment for all newly diagnosed patients. The other authors believe that both options should be considered.)
It was demonstrated in the IES trial that switching from tamoxifen to an AI has value. However, extending the AI beyond 5 years has as yet not been established and is the subject of ongoing evaluation. Extending endocrine therapy beyond 5 years adds important and substantial benefit, conferring an additional (DFS) advantage to all patients and an overall survival advantage in lymph node-positive patients. In women with primary tumors that are ER-positive and progesterone (PR)-positive, an overall survival advantage was demonstrated in the MA.17 trial. This concept of longer duration of adjuvant endocrine therapy established by the landmark MA.17 trial currently is being tested even in patients who received upfront AI therapy.
In premenopausal women, tamoxifen remains the standard of care for adjuvant endocrine treatment. However, there are 2 ongoing National Cancer Institute-Cancer Therapy Evaluation Program-sponsored trials that were designed to evaluate the efficacy of AIs and tamoxifen in this population.
The principal toxicity of aromatase inhibition is now well recognized as accelerated bone resorption. Monitoring women with BMD assessments; counseling related to diet, exercise, and vitamin D and calcium supplementation; and judicious use of bisphosphonates have been established. QoL studies generally have indicated that AIs are tolerated well and have no greater impact on QoL than tamoxifen, but they affect slightly different domains, such as body discomfort, most likely reflecting the greater musculoskeletal and joint discomfort from AIs.
In conclusion, the advent of AIs in adjuvant therapy of women with hormone receptor-positive primary tumors has added substantial benefit in terms of disease outcome. Clinicians should consider AIs when planning the patient's endocrine treatment. Outstanding issues include identifying the optimal duration, optimal agent, and optimal patients for these therapies. Prolonged endocrine therapy may become a reality for selected patients in keeping with the chronic relapsing nature of the disease. Optimizing, expanding, and extending endocrine therapy is likely to add further to patient outcome.
The questions and discussion below follow from the oral presentation given at the Sixth Cambridge Conference on Endocrine and Targeted Manipulation of Breast Cancer and do not correspond directly to the written article, which is a more general review.
Dr. Charles Coombes: What is your explanation for this very significant improvement in disease-free survival but no improvement in overall survival? Other than the finding that many patients were switched to an AI after the 5 years, might it also be the definition of disease-free survival events?
Dr. Aman Buzdar: If we look at the data in terms of survival advantage, age is a factor, because the median age of the patients, at the start, was about 3 to 4 years younger in the trials showing survival advantage. There were more deaths from other causes as these women reach the age of 70 or 75 years. These treatments are not going to make them immortal.
Dr. Eric Winer: I think the bottom line is the AIs are an advance, but at least 5 years of therapy with an AI versus 5 years with tamoxifen just isn't such a huge advance. There's a very small improvement in overall survival. My personal view is that the way we're really going to change mortality in women with ER-positive breast cancer is by finding better treatments and by extending treatment longer in those who need it longer. I think this is the first step, but it's not as big a step as we once thought that it was.
Dr. Kathleen Pritchard: It's not a quantum leap
Dr. Winer: No, but there aren't very many quantum leaps. I actually would argue that even trastuzumab in the adjuvant setting isn't a quantum leap, but it's a bigger leap than this.
Dr. Steven Come: The other thing that has concerned me is that, when you read all the published trials, I think in every single one of them, the AIs are slightly better tolerated than tamoxifen in terms ofwithdrawals and things like that; but, in clinical practice, they prove to be fairly difficult drugs for patients to take.
Dr. James Ingle: I think it's about 5-percent that get bad musculoskeletal complaints. But I find that, in most patients, low-dose celecoxib can take care of it.
Dr. Buzdar: Patients do complain of joint discomforts; but the thing is, most of the time, patients are scared, and they think every ache or pain is a recurrence of cancer. If you reassure them that this is an effect of therapy, then I think they are a little more reassured. All of us have seen patients who, despite doing everything, have persistent discomfort. I now have a couple of patients that I had to switch back to tamoxifen because their QoL was so suboptimal.
Dr. Winer: We need to know what ultimately will be the best strategy in terms of using the AIs: How long should they be used; which patients perhaps may be spared all these therapies altogether, and which patients need more than what we're currently providing? Certainly, my experience in practice is that all of these therapies are harder to give than we like to believe. When you talk to groups of women about how easy endocrine therapy is compared with treatments like chemotherapy, there invariably are a fair number of dissenters in the audience. The burden of long-term treatment, for some proportion of patients, is really substantial.
Dr. Stephen Johnston: I know that the mantra is that all patients gain or have some degree of benefit. But, again, if we're actually trying to say whether everybody needs an AI upfront, just take both the BIG 1-98 and ATAC together, is there anything really from that that starts to be strikingly different to the sort of similar subgroup analysis from the ATAC to suggest it?
Dr. Winer: My sense is that it's just that patients who have PR-negative tumors are at higher risk of recurrence, particularly early recurrence, and that it is a prognostic, rather than predictive, factor.
Dr. Johnston: So, in your practice, would that be still a subgroup that you might say, okay, we'll give you an AI rather than tamoxifen? For whatever degree of small benefit there might be.
Dr. Paul Goss: I don't, actually. I see Her-2 as another reason for high risk of early recurrence, so it is a reason for giving an AI. There is not another subset that we currently single out for AI therapy upfront.
Dr. Johnston: So you'd give them an AI?
Dr. Goss: Yes, because they're at a higher risk of relapse.
Dr. Winer: Now, of course, the issue with Her-2 is to what extent that is altered by treatment with Her-2-directed therapy. We don't know. Of course, that argument would be at it's strongest if it turns out that those groups of patients—that is, those patients with PR-negative disease, lots of positive lymph nodes, Her-2-positive disease—are at highest risk of early recurrence and then are at relatively lower risk of late recurrence. That's probably true of Her-2. It may be true of PR. I'm not sure that it's so true of nodal positivity.
Dr. Johnston: So, are there subgroups where it's reasonable to say tamoxifen?
Dr. Winer: My own view is that it's entirely reasonable to say that there are large groups of women for whom tamoxifen upfront is reasonable, given the finding that we have yet to show that there's a survival difference with any upfront strategy with an AI. At this point, I think that a reasonable position is that we simply don't know which of these strategies is better in virtually any patient.
Dr. Goss: ER-positive breast cancer patients have a very long time to be exposed to the risk of recurrence, so the worst thing that could happen to a patient is to drop out of all adjuvant therapy very early because they can't tolerate the treatment in the first 1 or 2 years. An important question is: Which women experience antiestrogen toxicities? Actually, strategies to overcome toxicity or identify reasons for toxicity, in the 10-percent of women who stop all adjuvant therapy, probably is going to contribute to a bettering of outcome more than any of these subtle differences between the drugs. There's a project that we are just initiating looking at whether drug holidays, combined with some counseling to try to bring patients back on therapy and keep them on therapy, could help the outcome of the patients with hormone-dependent breast cancer.
Dr. Winer: If it turns out that we can give 10 years of an AI safely and effectively, that would be a very important piece of information. It may turn out—and I think it likely will turn out—that, for some number of patients, extended therapy such as that will be very important. My real hope is that, at some point, we'll also identify the patients, not just who don't need an AI upfront, but who don't need endocrine therapy at all. I actually think that tests like Oncotype DX may help us a little in terms of identifying patients who, in fact, don't benefit from endocrine therapy.
Dr. Ingle: In your technology assessment panel, you stated that optimal endocrine therapy includes an AI, either initially, or after several years of tamoxifen, or after some period of tamoxifen. Do you want to comment on what the panel's conclusion on this was at St. Gallen that, first, there's a reasonable population for which 5 years of tamoxifen is appropriate and, second, if switched, most people do it after 2 or 3 years? Any comments about the tamoxifen? Is it dead as monotherapy?
Dr. Winer: I don't think tamoxifen is dead. The St. Gallen statement that 5 years of tamoxifen is reasonable is a very pragmatic statement when one looks at therapy around the world. I think the most important thing we can do for a woman with ER-positive breast cancer, based on what we know at the moment, is give her endocrine therapy. Given the finding that there are parts of the world where the cost of the AIs are relatively prohibitive, one should not dismiss tamoxifen entirely and, for that matter, for women who are intolerant of AIs or perhaps women who have quite low-risk disease and don't want to have the side effects of the AIs, tamoxifen is a reasonable treatment.
Dr. Buzdar: We should comment on longer duration of tamoxifen beyond 5 years. The question is whether there are any data to support this at the present time.
Dr. Goss: There are no data to suggest there's an additional benefit for continuing tamoxifen. The Adjuvant Tamoxifen Longer Against Shorter (ATLAS) and Adjuvant Tamoxifen Treatment Offer More (ATOM) data may show a slight difference—I'm guessing—but I doubt that the benefit of continuing tamoxifen will exceed the benefit of switching to an AI after 5 years of tamoxifen. I think to advocate 10 years of tamoxifen, at the present time, would not be optimal.
Dr. Goss: The protection of bone by tamoxifen in the overall population treated is a clear benefit
Dr. Coombes: I agree with what you said, but the fact remains if you examine the number of fractures for person years in ATAC versus IES, it's about the same. Whether it's a difference in T score or BMD loss or fractures, including traumatic, I think there's a lot of uncertainty, there is a lot of different information in the denominator.
Dr. Buzdar: The total picture of AEs from tamoxifen include thromboembolic complications, endometrial bleeding, vaginal bleeding, and endometrial cancer: When you look at the overall profile of tamoxifen, it is not as favorable as we think.
Dr. Goss: But bone toxicity needs to be looked at in context. And the tamoxifen-associated endometrial hysterectomy rate is partly a physician education problem. The hysterectomies are done by gynecologists, and we all know that 98% of these hysterectomies are unnecessary. Oncologists have pointed this out to no avail, but an educational effort is required to stop this phenomenon.
Dr. Jo Anne Zujewski: Does the risk of fractures increase as you stay on therapy, or is the rate the same over time?
Dr. Coombes: When we looked at the fracture rates, they appeared to be the same over the course of the therapy.
Dr. Buzdar: We looked at the hazard rate of fractures in ATAC. It went up on AI (anastrozole), and it remained higher, but it did not continue to go up, rise, and it remained steady over the next 2 to 3 years. Risk increases to approximately 40% with therapy, then the 2 curves run parallel to each other.
Dr. Goss: In MA.17, we exceeded the target benefit that the doctors asked us to achieve in the lymph node-negative patients and substantially exceeded it in the lymph node-positive patients. And yet, still, people around the world are reluctant to continue extended adjuvant endocrine therapy. For lymph node-positive patients coming off tamoxifen, they forego a survival advantage by not taking letrozole.
Dr. Coombes: This disappointment that you're expressing: Is this partly because people are switching patients earlier; therefore, by the time they come to 5 years, they have already had AI treatment?
Dr. Goss: I think patients should switch earlier in the majority of circumstances; but no, these are doctors in the U.K. and Europe who are going to run the 5 years of tamoxifen, finish it, and not do anything after that.
Dr. Coombes: They're making a financial decision. Some do it for lymph node-positive patients, but not lymph node-negative patients. How long after stopping tamoxifen can you switch?
Dr. Goss: Switching immediately is a good idea, the best idea; but, if you have delayed starting letrozole, there is still a big advantage. We know that 20% of the physicians in the U.S. are giving AIs for 5 years after 2 years of tamoxifen, according to research from the drug companies.
Dr. Pritchard: Is that because not everyone is stopping after 5 years?
Dr. Goss: I don't know when people stop, but doctors are telling patients, after a couple of years of tamoxifen, that they want them to switch and go to an AI for at least 5 years. The others stop tamoxifen after 5 years.
- 3A detailed analysis of the benefits of anastrozole over tamoxifen for venous thromboembolic events (VTEs) after 5 years treatment. Presented at the San Antonio Breast Cancer Symposium, San Antonio, Texas, December 14-17, 2006. Abstract 104., ; on Behalf of the ATAC Trialists' Group.
- 4Effect of anastrozole on bone mineral density: 5-year results of the ‘Arimidex’ (anastrozole), Tamoxifen, Alone or in Combination (ATAC) trial [abstract OC002]. Calcif Tissue Int. 2006; 78( suppl 1): S25., , , et al.
- 5Anastrozole is associated with a lower risk of endometrial abnormalities than tamoxifen: first report of the ATAC trial endometrial subprotocol at 6 years' follow-up [poster 4055]. Presented at the San Antonio Breast Cancer Symposium, San Antonio, Tex, December 14-17, 2006.; on behalf of the ATAC Trialists' Group.
- 9Survival benefit of switching to anastrozole after 2 years' treatment with tamoxifen versus continued tamoxifen therapy: the ARNO 95 study. ASCO Meeting Abstracts. 2006; 24(18s). Abstract 547., , , et al.
- 15Skeletal effects of exemestane on bone-mineral density, bone biomarkers, and fracture incidence in postmenopausal women with early breast cancer participating in the Intergroup Exemestane Study (IES): a randomised controlled study. Lancet Oncol. 2007; 8: 1119–1127., , , et al.
- 18Pfizer Inc. New study shows efficacy of aromasin on early breast cancer. Breast Cancer News. December 18, 2006.
- 19Extended adjuvant treatment with anastrozole: results from the Austrian Breast and Colorectal Cancer Study Group Trial 6a (ABCSG-6a). ASCO Meeting Abstracts. 2005; 23(10s). Abstract 527., , , et al.