Adjuvant Systemic Therapy
Stage and molecular features determine the need for adjuvant systemic therapy and the choice of modalities used. For example, estrogen and/or progesterone receptor–positive patients will receive hormone therapy. HER2overexpression is an indication for using adjuvant trastuzumab, usually in combination with chemotherapy. When neither HER2 overexpression (e.g., triple negative, which is common in the basal-like tumors) nor hormone receptors are present, adjuvant therapy relies on chemotherapeutic regimens, which are often combined with experimental targeted approaches.
Hormone therapy
If ER status is used to select adjuvant treatment, the study should be performed in a well-established, skilled laboratory. Immunohistochemical assays appear to be at least as reliable as standard ligand-binding assays in predicting response to adjuvant endocrine therapy.[84]
Tamoxifen
The EBCTCG performed a meta-analysis of systemic treatment of early breast cancer by hormone, cytotoxic, or biologic therapy methods in randomized trials involving 144,939 women with stage I or stage II breast cancer. The most recent analysis, which included information on 80,273 women in 71 trials of adjuvant tamoxifen, was published in 2005.[85] In this analysis, the benefit of tamoxifen was found to be restricted to women with ER-positive or ER-unknown breast tumors. In these women, the 15-year absolute reductions in recurrence and mortality associated with 5 years of use were 12% and 9%, respectively.[85][Level of evidence: 1iiA]
Allocation to approximately 5 years of adjuvant tamoxifen reduces the annual breast cancer death rate by 31%, largely irrespective of the use of chemotherapy and of age (<50 years, 50–69 years, =70 years), PR status, or other tumor characteristics.[85] This EBCTCG meta-analysis also confirmed the benefit of adjuvant tamoxifen in ER-positive premenopausal women.[85] Women younger than 50 years obtained a degree of benefit from 5 years of tamoxifen similar to that obtained by older women. In addition, the proportional reductions in both recurrence and mortality associated with tamoxifen use were similar in women with either node-negative or node-positive breast cancer, but the absolute improvement in survival at 10 years was greater in the latter group (5.3% vs. 12.5% with 5 years of use).[85][Level of evidence: 1iiA] Similar results were found in the IBCSG-13-93 trial.[86] Of 1,246 women with stage II disease, only the women with ER-positive disease benefited from tamoxifen.
The optimal duration of tamoxifen use has been addressed by the EBCTCG meta-analysis and by several large randomized trials.[85,87-89] Results from the EBCTCG meta-analysis show a highly significant advantage of 5 years versus 1 to 2 years of tamoxifen with respect to the risk of recurrence (proportionate reduction 15.2%; P<.001) and a less significant advantage with respect to mortality (proportionate reduction 7.9%; P = .01).[85] Results from the NSABP-B-14 study, which compared 5 years of adjuvant tamoxifen to 10 years of adjuvant tamoxifen for women with early-stage breast cancer, indicate no advantage for continuation of tamoxifen beyond 5 years in women with node-negative, ER-positive breast cancer.[87][Level of evidence: 1iA]
Another trial that included both node-positive and node-negative women also demonstrated the equivalence of 5 years and 10 years of therapy.[88][Level of evidence: 1iiDii] In both trials, there was a trend toward a worse outcome associated with a longer duration of treatment. In one trial EST-5181, node-positive women who had already received 5 years of tamoxifen following chemotherapy were randomly assigned to continue therapy or observation.[89] In the ER-positive subgroup, a longer time to relapse was associated with continued tamoxifen use, but no improvement in OS was observed. The current recommendation is that adjuvant tamoxifen be discontinued after 5 years in all patients as current standard therapy.[90] Clinical trials such as the Adjuvant Tamoxifen Longer Against Shorter (ATLAS) trial and the Adjuvant Tamoxifen Treatment–Offer More? (CRC-TU-ATTOM) trial have addressed different durations of adjuvant tamoxifen, and results are pending.
(Refer to the Letrozole section in the Aromatase inhibitors section of this summary for more information.)
Tamoxifen and chemotherapy
That chemotherapy should add to the effect of tamoxifen in postmenopausal women has been postulated.[91,92] In a trial (NSABP-B-16) of node-positive women older than 50 years with hormone receptor–positive tumors, 3-year DFS and OS rates were better in those who received doxorubicin, cyclophosphamide, and tamoxifen versus tamoxifen alone (DFS was 84% vs. 67%; P = .004; OS was 93% vs. 85%; P = .04).[93][Level of evidence: 1iiA] The NSABP-B-20 study compared tamoxifen alone with tamoxifen plus chemotherapy (cyclophosphamide, methotrexate, and fluorouracil [5-FU] [CMF] or sequential methotrexate and 5-FU) in women with node-negative, ER-positive breast cancer.[94] After 12 years of follow-up, the chemotherapy plus tamoxifen regimen resulted in 89% DFS and 87% OS compared with a 79% DFS and 83% OS with tamoxifen alone.[94][Level of evidence: 1iiA]
In another study of postmenopausal women with node-positive disease, tamoxifen alone was compared with tamoxifen plus three different schedules of CMF. A small, DFS advantage was conferred by the addition of early CMF to tamoxifen in women with ER-positive disease.[95][Level of evidence: 1iiDii] However, another study in a similar patient population, in which women were randomly assigned to receive adjuvant tamoxifen with or without CMF, showed no benefit in the chemotherapy arm; in this study, intravenous (day 1 every 3 weeks) rather than oral cyclophosphamide was used.[96][Level of evidence: 1iiA] The overall results of the available evidence suggest that the addition of chemotherapy to tamoxifen in postmenopausal women with ER-positive disease results in a significant, but small, survival advantage.
Tamoxifen toxic effects
The use of adjuvant tamoxifen has been associated with certain toxic effects. The most important effect is the development of endometrial cancer which, in large clinical trials, has been reported to occur at a rate that is two times to seven times greater than that observed in untreated women.[97-100] Women taking tamoxifen should be evaluated by a gynecologist if they experience any abnormal uterine bleeding. Although one retrospective study raised concern that endometrial cancers in women taking tamoxifen (40 mg/day) had a worse outcome and were characterized by higher-grade lesions and a more advanced stage than endometrial cancers in women not treated with tamoxifen, other larger studies using standard tamoxifen doses (20 mg/day) have not supported this finding.[97,101,102] Similar to estrogen, tamoxifen produces endometrial hyperplasia, which can be a premalignant change. In a cohort of women without a history of breast cancer who were randomly assigned to receive tamoxifen or placebo on the British Pilot Breast Cancer Prevention Trial, 16% of those on tamoxifen developed atypical hyperplasia at varying times from the start of treatment (range, 3 months–75 months; median, 24 months), while no cases occurred on the control arm.[103] The value of endometrial biopsy, hysteroscopy, and transvaginal ultrasound as screening tools is unclear.[104,105] Of concern is an increased risk of gastrointestinal malignancy after tamoxifen therapy, but these findings are tentative, and further study is needed.[106]
Tamoxifen use is also associated with an increased incidence of deep venous thrombosis and pulmonary emboli. In several adjuvant studies, the incidence ranged from 1% to 2%.[87,93,107-109] Clotting factor changes have been observed in controlled studies of prolonged tamoxifen use at standard doses; antithrombin III, fibrinogen, and platelet counts have been reported to be minimally reduced in patients receiving tamoxifen.[110] The relationship of these changes to thromboembolic phenomena is not clear. Tamoxifen use may also be associated with an increased risk of strokes.[109,111,112] In the NSABP Breast Cancer Prevention Trial (NSABP-P-1), this increase was not statistically significant.[111]
Another potential problem is the development of benign ovarian cysts, which occurred in about 10% of women in a single study.[113] The relationship between tamoxifen and ovarian tumors requires further study.[114] Short-term toxic effects of tamoxifen use may include vasomotor symptoms and gynecologic symptoms (e.g., vaginal discharge or irritation).[115] (Refer to the PDQ summary on Sexuality and Reproductive Issues for more information.) Ophthalmologic toxic effects have also been reported in patients receiving tamoxifen; patients who complain of visual problems should be assessed carefully.[116-118] Because the teratogenic potential of tamoxifen is unknown, contraception should be discussed with patients who are premenopausal or of childbearing age and are candidates for treatment with this drug.
Tamoxifen therapy may also be associated with certain beneficial estrogenic effects, including decreased total and low-density lipoprotein levels.[119,120] A large controlled Swedish trial has shown a decreased incidence of cardiac disease in postmenopausal women taking tamoxifen. Results were better for women taking tamoxifen for 5 years than for women taking it for 2 years.[121] In another trial, the risk of fatal myocardial infarction was significantly decreased in patients receiving adjuvant tamoxifen for 5 years versus those treated with surgery alone.[120] In the NSABP-B-14 study, the annual death rate due to coronary heart disease was lower in the tamoxifen group than in the placebo group (0.62 per 1,000 vs. 0.94 per 1,000), but this difference was not statistically significant.[122] To date, three large controlled trials have shown a decrease in heart disease.[120-122]
Controlled studies have associated long-term tamoxifen use with preservation of bone mineral density of the lumbar spine in postmenopausal women.[123-125] In premenopausal women, decreased bone mineral density is a possibility.[126]
Ovarian ablation, tamoxifen, and chemotherapy
The EBCTCG meta-analysis included almost 8,000 premenopausal women who were randomly assigned to undergo ovarian ablation with surgery or radiation therapy (4,317) or ovarian suppression with luteinizing hormone-releasing hormone (LHRH) agonists (3,408). Overall, ovarian ablation or suppression reduced the absolute risk of recurrence at 15 years by 4.3% (P < .001) and the risk of death from breast cancer by 3.2% (P= .004).[85] No evidence showed that the relative benefit of suppression differed from that of ablation, but the benefit of either was less in patients who received chemotherapy.[127][Level of evidence: 1iiA]
A single study of more than 300 patients that compared cyclophosphamide, methotrexate, 5-FU, and prednisone (CMFP) with the same chemotherapy regimen plus surgical oophorectomy showed no additional survival benefit from oophorectomy.[128][Level of evidence: 1iiA] Three trials (including the International Breast Cancer Study Group’s trial [IBCSG-VIII] and the Eastern Cooperative Oncology Group’s trial [EST-5188]) involving more than 3,000 patients assessed the impact on DFS and OS from the use of an LHRH analogue (in one trial, 50% of the patients had radiation oophorectomy rather than an LHRH analogue) in addition to chemotherapy.[127,129,130][Level of evidence: 1iiA] None of the trials identified a statistically significant benefit in OS or DFS from ovarian suppression.
As an adjuvant strategy, ovarian ablation has also been compared with chemotherapy in premenopausal women. In a direct comparison of surgical or radiation ablation and CMF, DFS and OS rates were identical in 332 premenopausal women with stage II disease.[131][Level of evidence: 1iiA] A trial of 599 premenopausal node-positive patients found leuprorelin acetate to be similar to CMF with respect to DFS and OS.[132] A Danish trial compared ovarian ablation or suppression to CMF (9 cycles intravenously every 3 weeks) in premenopausal, ER-positive women and found no difference in OS or DFS in the two study groups.[133,134] The study did not have tamoxifen as an adjuvant arm and also did not use taxanes or anthracyclines. Results may have been different with these two contemporary modifications to the study. A trial of CMF versus tamoxifen plus ovarian ablation (e.g., by surgery, radiation therapy, or gonadotropin-releasing hormone [GnRH]) in premenopausal or perimenopausal women with receptor-positive tumors has been reported.[135][Level of evidence: 1iiA] In this small trial, which did not meet its target accrual, the combination of tamoxifen and ovarian ablation provided comparable DFS and OS rates. In three larger trials in which medical ovarian ablation with goserelin was used, the impact of goserelin alone on DFS was found to be comparable to CMF in the subgroup of ER+ patients,[127,136][Level of evidence: 2Dii] whereas the combination of goserelin and tamoxifen was associated with prolonged DFS compared with CMF alone.[137][Level of evidence: 1iiDii] Whether tamoxifen or aromatase inhibitors add to ovarian ablation, and the elucidation of the optimal roles for endocrine manipulation and chemotherapy in receptor-positive premenopausal women, require further evaluation.[138] These issues are the subject of several trials.
Aromatase inhibitors
Based on DFS advantage as described below, aromatase inhibitors have become the first-line adjuvant therapy for postmenopausal women; however, because there is no demonstrated survival advantage to aromatase inhibitors, tamoxifen remains a reasonable alternative.[139,140]
Anastrozole
A large randomized trial of 9,366 patients has compared the use of the aromatase inhibitor anastrozole and the combination of anastrozole and tamoxifen with tamoxifen alone as adjuvant therapy for postmenopausal patients with node-negative and node-positive disease.[141,142] Most (84%) of the patients in the study were hormone-receptor positive. Slightly more than 20% had received chemotherapy. With a median follow-up of 33.3 months, no benefit was observed for the combination arm relative to tamoxifen. Patients on anastrozole, however, had a significantly longer DFS (hazard ratio [HR] = 0.83) than those on tamoxifen. In an analysis conducted when all but 8% of the patients had completed protocol therapy at a follow-up of 68 months,[142] the benefit of anastrozole relative to tamoxifen with respect to DFS was slightly less (HR = 0.87; 95% CI, 0.78–0.96; P = .01). A greater benefit was seen in hormone receptor-positive patients (HR = 0.83; 95% CI, .73–0.94; P = .05). There was an improvement in time to recurrence (HR = 0.79; 95% CI, 0.70–0.90; P = .005), distant DFS (HR = 0.86; 95% CI, 0.74–0.99; P = .04) and contralateral breast cancer (42% reduction; P = .01) in patients who received anastrozole.[142][Level of evidence: 1iDii] No difference was shown in OS (HR = 0.97; 95% CI, 0.85–1.12; P = .7 ). Arthralgia and fractures were reported significantly more often in patients who received anastrozole, whereas hot flushes, vaginal bleeding and discharge, endometrial cancer, ischemic cerebrovascular events, venous thromboembolic and deep venous thromboembolic events were more common in patients who received tamoxifen.[142] An American Society of Clinical Oncology (ASCO) Technology Assessment panel has commented on the implications of these results.[143,144]
Three trials examined the effect of switching to anastrozole to complete a total of 5 years of therapy after 2 to 3 years of tamoxifen.[145-147] One study, which included 448 patients, demonstrated a statistically significant reduction in DFS (HR = 0.35; 95% CI, 0.18–0.68; P = .001) but no difference in OS.[145][Level of evidence: 1iiA] The other two trials were reported together.[146] A total of 3,224 patients were randomized after 2 years of tamoxifen to continue tamoxifen for a total of 5 years or to take anastrozole for 3 years. After a median follow-up of 78 months, an improvement in all-cause mortality (HR = 0.61; 95% CI, 0.42–0.88; P = .007) was observed.[147]
A meta-analysis of these three studies showed that patients who switched to anastrozole had significant improvements in DFS (HR = 0.59; 95% CI, 0.48–0.74; P < .001), EFS (HR = 0.55; 95% CI, 0.42–0.71; P < .001), distant DFS (HR = 0.61; 95% CI, 0.45–0.83; P= .002), and OS (HR = 0.71; 95% CI, 0.52–0.98; P = .04) compared with the patients who remained on tamoxifen.[148]
Letrozole
A large double-blinded randomized trial of 8,010 postmenopausal women with hormone receptor-positive breast cancer compared the use of letrozole versus tamoxifen given continuously for 5 years or with crossover to the alternate drug at 2 years.[149] In an updated analysis from the Breast International Group (IBCSG-1-98) including only the 4,922 women who received tamoxifen or letrozole for 5 years, at a median follow-up time of 51 months, DFS was significantly superior in patients treated with letrozole (HR = 0.82; 95% CI, 0.71–0.95; P = .007; 5-year DFS = 84.0% vs. 81.1%).[150][Level of evidence: 1iDii] OS was not significantly different (HR = 0.91; 95% CI, 0.75–1.11; P = .35). Patients on letrozole had significantly fewer thromboembolic events, endometrial pathology, hot flashes, night sweating, and less vaginal bleeding. Patients on tamoxifen had significantly fewer bone fractures, arthralgia, hypercholesterolemia, and cardiac events other than ischemic heart disease and cardiac failures.[150]
A large, double-blinded, randomized trial (CAN-NCIC-MA17 [NCT00003140]) of 5,187 patients compared the use of letrozole versus placebo in receptor-positive postmenopausal women who received tamoxifen for approximately 5 (4.5–6.0) years.[151] After the first planned interim analysis, when median follow-up for patients on study was 2.4 years, the results were unblinded because of a highly significant (P < .008) difference in DFS (HR = 0.57) favoring the letrozole arm.[151][Level of evidence: 1iDii] After 3 years of follow-up, 4.8% of the women on the letrozole arm had developed recurrent disease or new primaries versus 9.8% on the placebo arm (95% CI for the difference, 2.7%–7.3%). Women on letrozole had significantly more hot flashes, arthritis, arthralgia and myalgia, but less vaginal bleeding. New diagnoses of osteoporosis were more frequent on letrozole (5.8% vs. 4.5%), though the difference was not statistically significant (P = .07). Because of the early unblinding of the study, longer-term comparative data on the risks and benefits of letrozole in this setting will not be available.[152,153] An updated analysis including all events prior to unblinding confirmed the results of the interim analysis.[154] In addition, a statistically significant improvement in distant DFS was found for patients on letrozole (HR = 0.60; 95% CI, 0.43–0.84; P = .002). Although no significant difference was found in the total study population, the node-positive patients on letrozole also experienced a statistically significant improvement in OS (HR = 0.61; 95% CI, 0.38–0.98; P = .04), though the P value was not corrected for multiple comparisons. An ASCO Technology Assessment panel has commented on the implications of these results.[143,144]
Exemestane
A large double-blinded randomized trial (EORTC-10967 [ICCG-96OEXE031-C1396-BIG9702]) of 4,742 patients compared continuing tamoxifen with switching to exemestane for a total of 5 years of therapy in women who had received 2 to 3 years of tamoxifen.[155,156] After the second planned interim analysis, when median follow-up for patients on the study was 30.6 months, the results were released because of a highly significant (P < .005) difference in DFS (HR = 0.68) favoring the exemestane arm.[155][Level of evidence: 1iDii] After a median follow-up of 55.7 months, the HR for DFS was 0.76 (95% CI, 0.66–0.88; P = .001) in favor of exemestane.[157] At 2.5 years after randomization, 3.3% fewer patients on exemestane had developed a DFS event (95% CI, 1.6–4.9). The HR for OS was 0.85 (95% CI, 0.7–1.02; P = .08).[157][Level of evidence: 1iA] Women on exemestane had significantly more arthralgia, diarrhea, hypertension, fractures, arthritis, musculoskeletal pain, carpal tunnel syndrome, insomnia, and osteoporosis, but women on tamoxifen had significantly more gynecologic symptoms, muscle cramps, vaginal bleeding and discharge, thromboembolic disease, endometrial hyperplasia, and uterine polyps. (For information on diarrhea, refer to the PDQ summary on Gastrointestinal Complications; for information on insomnia, refer to the PDQ summary on Sleep Disorders.)
A large randomized trial of 9,779 patients compared DFS of postmenopausal women with hormone receptor–positive breast cancer between initial treatment with sequential tamoxifen for 2.5 to 3 years followed by exemestane for a total of 5 years versus exemestane alone for 5 years.[158] The primary endpoints were DFS at 2.75 years and 5.0 years. Five-year DFS was 85% in the sequential group and 86% in the exemestane-alone group (HR = 0.97; 95% CI, 0.88–1.08; P = .60).[158][Level of evidence: 1iDii] The results of this trial support the use of exemestane, either sequentially after tamoxifen or as initial treatment for early-stage hormone receptor–positive breast cancer in postmenopausal women.
Chemotherapy
Overview of chemotherapy
Some of the most important data on the benefit of adjuvant chemotherapy came from the EBCTCG, which meets every 5 years to review data from global breast cancer trials. The year 2000 overview analysis (published in 2005) summarized the results of randomized adjuvant trials initiated by 1995.[85] The analyses of adjuvant chemotherapy involved 28,764 women participating in 60 trials of combination chemotherapy (polychemotherapy) versus no chemotherapy, 14,470 women in 17 trials of anthracycline-containing versus CMF-type chemotherapy, and 6,125 women in 11 trials of longer versus shorter chemotherapy duration.
For women younger than 50 years, polychemotherapy reduced the annual risk of disease relapse and death from breast cancer by 37% and 30%, respectively. This translated into a 10% absolute improvement in 15-year survival (HR = 42% vs. 32%). For women aged 50 to 69 years, the annual risk of relapse or death from breast cancer was decreased by 19% and 12%, respectively. This translated into a 3% absolute gain in 15-year survival (HR = 50% vs. 47%). The absolute gain in survival for polychemotherapy versus no adjuvant therapy in women younger than 50 was twice as great at 15 years as it was at 5 years (10% vs. 4.7%), while the main effect on disease recurrence was seen in the first 5 years.[85] The 15-year cumulative reduction in mortality from 6 months of an anthracycline-based regimen (e.g., fluorouracil, doxorubicin, cyclophosphamide [FAC] or fluorouracil, epirubicin, cyclophosphamide [FEC]) was 38% in women younger than 50 years, and 20% in those aged 50 to 60 years. The meta-analysis also showed that the reduction in risk of recurrence was similar in the presence or absence of tamoxifen, irrespective of age (<50 years vs. 50 to 69 years), though the result did not attain statistical significance in those randomly assigned women younger than 50 years. This finding, however, is most likely the result of the relatively small number of younger women in trials of combined chemoendocrine therapy. Few women older than 70 years had been studied, and specific conclusions could not be reached in this group. Importantly, these data were derived from clinical trials in which patients were not selected for adjuvant therapy according to ER status, and they were initiated before the advent of taxane-containing, dose-dense, or trastuzumab-based therapy.[85] As a result, they may not reflect treatment outcomes based on evolving treatment patterns.
Results of individual trials are generally in agreement with the conclusions of the meta-analysis. The NSABP-B-13 study demonstrated a benefit for chemotherapy with sequential methotrexate and 5-FU versus surgery alone in patients with node-negative, ER-negative tumors.[93,94,159,160][Level of evidence: 1iiA]
Duration of CMF-based chemotherapy
The EBCTCG meta-analysis assessed data from five trials comparing durations of at least 6 months with longer durations of 9 to 24 months. No survival benefit was demonstrated for durations greater than 6 months.[161][Level of evidence: 1iiA]
Anthracycline-based regimens
The EBCTCG meta-analysis analyzed 11 trials that began in 1976 through 1989 in which women were randomized to receive regimens containing anthracyclines (e.g., doxorubicin or epirubicin) versus CMF alone. The EBCTCG overview analysis directly compared anthracycline-containing regimens (mostly 6 months of FEC or FAC) with CMF (either oral or IV) in approximately 14,000 women, 64% of whom were under 50 years.[85] Compared to CMF, anthracycline-based regimens were associated with a modest but statistically significant 11% proportional reduction in the annual risk of disease recurrence, and a 16% reduction in the annual risk of death. In each case, the absolute difference in outcomes between anthracycline-based and CMF-type chemotherapy was about 3% at 5 years and 4% at 10 years.[161][Level of evidence: 1iiA]
The largest direct comparison of cyclophosphamide, doxorubicin, and 5-fluorouracil (CAF) (six cycles) versus CMF (six cycles) occurred in a U.S. Intergroup study (SWOG-8897), which was not included in the meta-analysis.[162] In this study, 2,691 patients were randomized to receive CAF or CMF with a second randomization to 5 years of tamoxifen versus no tamoxifen. Ten-year follow-up estimates indicated that CAF was not significantly better than CMF (P = .13) for the primary outcome of DFS (77% vs. 75%; HR = 1.09; 95% CI, 0.94–1.27). CAF had slightly better OS than CMF (85% vs. 82%, HR = 1.19 for CMF vs. CAF; 95% CI, 0.99–1.43), though values were statistically significant in the planned one-sided test (P = .03). Toxicity was greater with CAF and did not increase with tamoxifen. Overall, tamoxifen had no benefit (DFS, P = .16; OS, P = .37), but the tamoxifen effect differed by high-risk groups. For high-risk node-positive patients, tamoxifen was beneficial (DFS: HR = 1.32 for no tamoxifen vs. tamoxifen; 95% CI, 1.09–1.61; P = .003; OS: HR = 1.26; 95% CI, 0.99–1.61; P = .03) but not for high-risk node-negative patients (DFS: HR = 0.81 for no tamoxifen vs. tamoxifen; 95% CI, 0.64–1.03; OS: HR = 0.79; 95% CI, 0.60–1.05). The conclusion of this trial was that CAF did not improve DFS compared with CMF; and, there was a slight effect on OS. Given greater toxicity, CAF cannot be concluded to be superior to CMF. Tamoxifen was effective in high-risk node-positive disease but not in high-risk node-negative disease.[162][Level of evidence: 1iiA]
Several investigators have attempted to improve outcomes by combining CMF and anthracycline-containing regimens. Two Italian studies have evaluated these regimens.[163,164] In one study, 490 premenopausal and postmenopausal women with one to three axillary lymph nodes were randomized to receive CMF (12 cycles) or CMF (eight cycles) followed by doxorubicin (four cycles).[163] After a median observation of 17.5 years, no statistically significant difference was documented in the first study (relapse-free survival [RFS], HR = 1.06; total survival, HR = 1.03). In contrast, the delivery of doxorubicin first, followed by CMF significantly reduced the risk of disease relapse (HR = 0.68; 95% CI, 0.54–0.87; P =.0017) and death (HR = 0.74; 95% CI, 0.57–0.95; P = .018) compared with the alternating regimen. In the other study, 403 premenopausal and postmenopausal women with four or more positive axillary lymph nodes were randomized to receive doxorubicin (four cycles) followed by CMF (eight cycles) or CMF (two cycles) alternating with doxorubicin (one cycle) for a total of 12 cycles.[164] Women who received doxorubicin followed by CMF had better RFS (42% vs. 28%; P = .002) and OS (58% vs. 44%; P = .002).[164][Level of evidence: 1iiA]
The NSABP-B-15 trial randomized 2,194 patients with axillary node-positive breast cancer and tumors determined nonresponsive to tamoxifen to doxorubicin and cyclophosphamide (AC) (four cycles), CMF (six cycles), or AC (four cycles) followed after a 6-month delay by CMF (three cycles).[165] No differences were seen in DFS or OS among the three groups.[165][Level of evidence: 1iiA] This study has also shown no difference in survival rates between four cycles of AC and six cycles of CMF.
The results of these various studies comparing and combining CMF and anthracycline-containing regimens suggest a slight advantage for the anthracycline regimens in both premenopausal and postmenopausal patients. Uncertainty remains, however, about whether there is an advantage to combining both regimens.
Evidence suggests that particular tumor characteristics may predict anthracycline-responsiveness. Data from retrospective analyses of randomized clinical trials suggest that, in patients with node-positive breast cancer, the benefit from standard-dose versus lower-dose adjuvant CAF,[2] or the addition of doxorubicin to the adjuvant regimen,[3] is restricted to those patients whose tumors overexpress HER2/neu.[Level of evidence: 1iiA] A retrospective analysis of the HER2/neu status of 710 premenopausal, node-positive women was undertaken to see the effects of adjuvant chemotherapy with CMF or cyclophosphamide, epirubicin, and fluorouracil (CEF).[166][Level of evidence: 2A] HER2/neu was measured using fluorescence in situ hybridization, polymerase chain reaction, and immunohistochemical methods. The study confirmed previous data indicating that the amplification of HER2/neu was associated with a decrease in RFS and OS. In patients with HER2/neu amplification, the RFS and OS were increased by CEF. In the absence of HER2/neu amplification, CEF and CMF were similar to RFS (HR for relapse = 0.91; 95% CI, 0.71–1.18; P = .049) and OS (HR for death = 1.06; 95% CI, 0.83–1.44; P = .68). Similar results were seen in a meta-analysis that included 5,354 patients in whom HER2 status was known from eight randomized trials (including the one just described) comparing anthracycline-containing regimens with non–anthracycline-containing regimens.[167]
The role of adding taxanes to adjuvant therapy
A number of trials have addressed the benefit of adding a taxane (paclitaxel or docetaxel) to an anthracycline-based adjuvant chemotherapy regimen. A literature-based meta-analysis of 13 such studies demonstrated that the inclusion of a taxane improved both DFS and OS (DFS: HR = 0.83; 95% CI, 0.79–0.87; P < .0001; OS: HR = 0.85; 95% CI, 0.79–0.91; P < .0001).[168][Level of evidence: 1iiA] Five-year absolute survival differences were 5% for DFS and 3% for OS in favor of taxane-containing regimens. There were no differences in benefit observed in patient subsets defined by nodal status, hormone receptor status, or age/menopausal status. There was also no apparent difference in efficacy between the two agents. However, none of the studies reviewed involved a direct comparison between paclitaxel and docetaxel.
A European Cooperative Oncology Group–led intergroup trial (E1199 [NCT00004125]) involving 4,950 patients compared, in a factorial design, two schedules (weekly and every 3 weeks) of the two drugs (docetaxel vs. paclitaxel) following standard-dose AC chemotherapy given every 3 weeks.[169][Level of evidence: 1iiA] There was no difference observed in the overall comparison with regard to DFS of docetaxel to paclitaxel (odds ratio [OR] = 1.03; 95% CI, 0.91–1.16; P = .61) or between the 1-week and 3-week schedules (OR = 1.06; 95% CI, 0.94–1.20; P = .33). However, there was a significant interaction between the drug administered and schedule for both DFS (0.003) and OS (0.01). Thus, compared with paclitaxel given every 3 weeks, paclitaxel given weekly improved both DFS (OR = 1.27; 95% CI, 1.01–1.57; P = .006) and OS (OR = 1.32; 95% CI, 1.02–1.72; P = .01). Docetaxel given every 3 weeks was also superior in DFS to paclitaxel given every 3 weeks (OR = 1.23; 95% CI, 1.00–1.52; P = .02), but the difference was not statistically significant for OS (OR = 1.13; 95% CI, 0.88–1.46; P= .25). Docetaxel given weekly was not superior to paclitaxel given every 3 weeks. There was no stated a prioribasis for expecting that varying the schedule of administration would have opposite effects for the two drugs. Thus, these results are hypothesis generating and should be confirmed.
Dose-intensity, dose-density, and high-dose chemotherapy
Retrospective and some prospective data support the view that physicians should avoid arbitrary reductions in dose intensity.[170,171] The data for the benefit of dose escalation in breast cancer, however, are more controversial. The CLB-8541 trial compared three dose intensities of CAF in 1,550 patients with node-positive breast cancer. Patients received either CAF (300/30/300 mg/m2 every 4 weeks for four cycles; low-dose arm), CAF (400/40/400 mg/m2 every 4 weeks for six cycles; moderate-dose arm), or CAF (600/60/600 mg/m2 every 4 weeks for four cycles; high-dose arm). The high-dose arm had twice the dose intensity and twice the drug dose as the low-dose arm. The moderate-dose arm had 66% of the dose intensity as the high-dose arm but the same total drug dose. At a median follow-up of 9 years, DFS and OS on the high-dose and intermediate-dose arms were superior to the corresponding survival measures on the low-dose arm (P = .001) with no difference in these measures between the high-dose and intermediate-dose arms.[170][Level of evidence: 1iiA] The higher dose levels used in this trial are currently considered standard, so it is unclear whether this trial is supportive of the value of dose intensity or, rather, supportive of the concept of a threshold level below which treatment becomes ineffective.
Other trials have clearly escalated doses beyond the standard range. The NSABP-B-22 and NSABP-B-25 trials, for example, escalated the dose of cyclophosphamide to 1,200 mg/m2 (without granulocyte-colony stimulating factor [G-CSF]) and 2,400 mg/m2 (with G-CSF), respectively, with no significant advantage observed in DFS or OS compared with the standard dose of 600 mg/m2.[172,173][Level of evidence: 1iiA]
A U.S. Intergroup study (CLB-9344) randomly assigned women with node-positive tumors to three dose levels of doxorubicin (60, 75, and 90 mg/m2). Following treatment with doxorubicin, a second randomization occurred to paclitaxel or to no further therapy. After chemotherapy, patients with ER-positive tumors were offered a planned course of tamoxifen for 5 years. No difference in DFS related to the dose of doxorubicin was found.[174] In contrast, a Canadian trial (CAN-NCIC-MA5) in which cyclophosphamide, epirubicin, and 5-FU (CEF) were given to a total dose of 720 mg/m2 for a period of six 4-week cycles demonstrated at a median follow-up of 10 years for live patients, a 10-year RFS of 52% for patients who received CEF compared with 45% for CMF patients (HR for CMF vs. CEF = 1.31; stratified log-rank, P = .007).[175] The 10-year OS for patients who received CEF and CMF was 62% and 58%, respectively (HR for CMF vs. CEF = 1.18; stratified log-rank, P = .085). The rates of acute leukemia have not changed since the original report, whereas the rates of congestive heart failure were slightly higher (four patients [1.1%] in the CEF group vs. one patient [0.3%] in the CMF group).[175][Level of evidence: 1iiA] The design of the trial does not allow a determination of whether anthracycline or dose-intensity or both is responsible for the improved outcome. A French trial showed that higher doses of epirubicin led to a high survival rate in women with poor-prognosis disease.[176] A randomized trial that increased duration of epirubicin did not lead to increased survival at 10 years in node-positive premenopausal women.[177]
A U.S. Intergroup trial (CLB-9741) compared, in a 2 × 2 factorial design, the use of adriamycin, cyclophosphamide, and paclitaxel concurrently (adriamycin and cyclophosphamide followed by paclitaxel) versus sequentially (adriamycin followed by paclitaxel followed by cyclophosphamide), given every 3 weeks or every 2 weeks with filgrastim, in 2,005 node-positive premenopausal and postmenopausal patients.[178] At a median follow-up of 68 months, dose-dense treatment improved the primary end point, DFS in all patient population (HR = 0.80; P =.018) but not OS (HR = 0.85; P =.12). There was no interaction between density and sequence. Severe neutropenia was less frequent in patients who received the dose-dense regimens.[178,179] Grade 2 anemia (hemoglobin <10g/dL) was more frequent in the adriamycin and cyclophosphamide followed by paclitaxel every 2 weeks’ arm (P < .001). At cycle five, this same arm had the lowest nadir hemoglobin of 10.7 g/dL, 0.9 g/dL lower than the other arms. Also, epoetin alpha use was highest in this arm compared with the three other arms (P = .013). In conclusion, dose-dense adriamycin and cyclophosphamide followed by paclitaxel every 14 days in C2 was associated with a greater incidence of moderate anemia, higher use of epoetin alpha, and more red cell transfusions than the other arms.[180][Level of evidence: 1iiA]
Several clinical trials (including EST-2190) have tested high-dose chemotherapy with bone marrow transplant (BMT) or stem cell support in women with more than ten positive lymph nodes and in women with four to nine positive lymph nodes.[181-188] A prospective randomized trial of 403 patients testing the use of two tandem high-dose chemotherapy courses demonstrated a statistically significant (P = .02) difference in 5-year survival (75% vs. 70%) with a 49-month median follow-up.[187][Level of evidence: 1iiA] The remaining trials comparing conventional chemotherapy to high-dose chemotherapy with BMT or stem cell support in high-risk patients in the adjuvant setting indicated no OS or EFS benefit from the high-dose chemotherapy with BMT or stem cell support.[181-186,188-190][Level of evidence: 1iiA] The information to date does not support the use of high-dose chemotherapy outside the context of a randomized clinical trial.
Also, a systematic review of nine randomized controlled trials comparing the effectiveness of high-dose chemotherapy and autograft with conventional chemotherapy for women with early poor prognosis breast cancer was performed.[188] In total 1,758 women were randomly assigned to receive high-dose chemotherapy with autograft, and 1,767 women were randomly assigned to receive conventional chemotherapy. There were 48 noncancer-related deaths on the high dose arm and four on the conventional dose arm (RR = 7.74; 95% CI, 3.43–17.50). There was no statistically significant difference in OS between women who received high-dose chemotherapy with autograft and women who received conventional chemotherapy, either at 3 years (RR = 1.02; 95% CI, 0.98–1.06), or at 5 years (RR = 0.98, 95% CI, 0.93–1.05). There was a statistically significant benefit in EFS at 3 years for the group who received high dose chemotherapy (RR = 1.11; 95% CI, 1.05–1.18). However, this significance was lost at 5 years (RR = 1.00; 95% CI, 0.92–1.08).[188]
Other chemotherapy regimens
The NSABP-B-19 trial compared CMF to sequential methotrexate followed by 5-FU in 1,095 women with node-negative, ER-negative tumors. After 13 years of follow-up, an overall benefit was seen for CMF relative to methotrexate plus 5-FU (MF) (RFS: HR = 0.59; 95% CI, 0.45–0.77, P < .001; OS: HR = 0.71; 95% CI, 0.55–0.92; P = .01). All age and menopausal groups demonstrated an RFS benefit, and most demonstrated an OS benefit.[159][Level of evidence: 1iiA] Serious toxicity (=grade 3), especially febrile neutropenia, was more frequent among CMF-treated patients. With no outcome advantage in older women and more toxic effects from the CMF regimen, the results of this study suggested that methotrexate followed by 5-FU was a reasonable substitute for CMF for older women.
A U.S. Intergroup study (CLB-9344) randomly assigned women with node-positive tumors to three dose levels of doxorubicin (60, 75, and 90 mg/m2) and a fixed dose of cyclophosphamide (600 mg/m2) every 3 weeks for four cycles. After AC chemotherapy, patients underwent a second randomization to paclitaxel (175 mg/m2) every 3 weeks for four cycles, and women with ER-positive tumors also received tamoxifen for 5 years. Although the dose-escalation of doxorubicin was not beneficial, the addition of paclitaxel resulted in statistically significant improvements in DFS (5%) and OS (3%).[174][Level of evidence: 1iiA] The results of a second trial, the NSABP-B-28 trial, have also been reported.[191] This trial randomized 3,060 women with node-positive breast cancer to four cycles of postoperative AC or four cycles of AC followed by four cycles of paclitaxel. All women older than 50 years, and those younger than 50 years with receptor-positive disease, received tamoxifen. In this trial, DFS was significantly improved by the addition of paclitaxel (hazard ratio [HR] = 0.83; 95% CI, 0.72–0.96; P = .006; 5-year DFS = 76% vs. 72%). The difference in OS was small (HR = 0.93), however, and not statistically significant (P = .46).[191][Level of evidence: 1iiA]
The regimen of 5-FU, adriamycin, and cyclophosphamide (FAC) compared with docetaxel plus doxorubicin and cyclophosphamide (TAC) has been studied in 1,491 women with node-positive disease in the Breast Cancer International Research Group’s (BCIRG-001) trial. Six cycles of either regimen were given as adjuvant postoperative therapy. A 75% DFS rate existed at 5 years in the TAC group compared with a 68% survival in the FAC group (P = .001). TAC was associated with a 30% overall lower risk of death (5% absolute difference) than FAC (HR = .70; 98% CI, 0.53–0.91; P < .008). Anemia, neutropenia, febrile neutropenia, and infections were more common in the TAC group. No deaths were associated with infections in either group.[192,193][Level of evidence: 1iiA] (For information on anemia, refer to the PDQ summary on Fatigue.)
The regimen of docetaxel and cyclophosphamide (TC) compared with doxorubicin plus cyclophosphamide (AC) was studied in 1,016 women with stage I or stage II invasive breast cancer. Patients were randomly assigned to receive four cycles of either TC or AC as adjuvant postoperative therapy. At 5 years, DFS was statistically significantly superior for TC compared with AC (86% vs. 80%, HR= 0.67; 95% CI, 0.50–0.94; P = .015).[194][Level of evidence: 1iiA] However, OS was not statistically significantly improved. With TC, patients had fewer cardiotoxic effects but other side effects included more myalgia, arthralgia, edema, and febrile neutropenia compared with AC.
Bisphosphonates
The role of bisphosphonates as part of adjuvant therapy for early stage breast cancer is currently under clinical study. The ABCSG-12 (NCT00295646) trial was a 2 × 2 factorial-design randomized trial that assigned 1,803 premenopausal patients with ER+ breast cancer to receive ovarian function suppression with goserelin and tamoxifen versus goserelin and anastrozole. These patients then underwent a second randomization to receive zoledronic acid (4 mg intravenously every 6 months) versus no zoledronic acid.[195][Level of Evidence: 1iiA] There was no significant difference in DFS between the anastrozole and tamoxifen groups. However, the addition of zoledronic acid to endocrine therapy, as compared with endocrine therapy without zoledronic acid, resulted in a relative reduction of 36% in the risk of disease progression (HR = 0.64; P = .01) but did not significantly reduce the risk of death.
While bisphosphonates appear to improve DFS in a population with low-to-intermediate risk breast cancer, it is not yet clear whether this benefit is generalizable to all patients with breast cancer. Results are pending from theSHEFF-AZURE trial, a randomized phase III trial of patients with stage II or III breast cancer who received chemotherapy and/or hormone therapy plus zoledronic acid versus no zoledronic acid.
Monoclonal antibodies
Five clinical trials addressing the role of the anti-HER2/neu antibody, trastuzumab, as adjuvant therapy for patients with HER2-overexpressing cancers have released the results of interim analyses.
In the HERceptin Adjuvant (HERA) (BIG-01-01) trial, which is the largest study (5,090 patients), trastuzumab was given every 3 weeks within 7 weeks of the completion of primary therapy that included an anthracycline-containing chemotherapy regimen given preoperatively or postoperatively plus or minus locoregional radiation therapy. Although the results of the comparison of 1 year versus 2 years of trastuzumab have not been released yet, there are available data for 3,387 patients (1,694 in the 1-year trastuzumab arm and 1,693 in the observation arm).[196] Of these patients, the median age was 49 years, about 33% had node-negative disease, and nearly 50% had hormone receptor (ER and PR)-negative disease. Patients who were treated with 1 year of trastuzumab experienced a 46% lower risk of a first event (hazard ratio [HR] = 0.54; 95% CI, 0.43–0.67; P < .001), corresponding to an absolute DFS benefit of 8.4% at 2 years (95% CI, 2.1–14.8). The updated results at 23.5 months’ follow-up showed an unadjusted HR for the risk of death with trastuzumab compared with observation of 0.66 (95%CI, 0.47–0.91; P = .0115), corresponding to an absolute OS benefit of 2.7%.[197] There were 218 DFS events reported with trastuzumab compared with 321 DFS events reported with observation. The unadjusted HR for the risk of an event with trastuzumab was 0.64 (0.54–0.76; P < .001), corresponding to an absolute DFS benefit of 6.3%.
In the combined analysis of the NSABP-B-31 (NCT00004067) and Intergroup NCCTG-N9831 trials, trastuzumab was given weekly, concurrently, or immediately after the paclitaxel component of the AC with paclitaxel regimen.[198] The results were confirmed in a joint analysis of the two studies, with a combined enrollment of 3,676 patients, that demonstrated a highly significant improvement in DFS (HR = 0.48; P < .001; 3-year DFS = 87% vs. 75%), as well as a significant improvement in OS (HR = 0.67; P = .015; 3-year OS = 94.3% vs. 91.7%; 4-year OS = 91.4% vs. 86.6%).[199] Patients treated with trastuzumab experienced a longer DFS with a 52% lower risk of a DFS event (HR = 0.48; 95% CI, 0.39–0.59; P < .001), corresponding to an absolute difference in DFS of 11.8% at 3 years and 18% at 4 years. The risk of distant recurrence was 53% lower (HR = 0.47; 95%CI, 0.37–0.61; P < .001) in patients treated with trastuzumab, and the risk of death was 33% lower (HR = 0.67; 95%CI, 0.48–0.93; P = .015) in these patients.
The AVENTIS-TAX-GMA-302 study was a three-arm large trial containing two anthracycline arms (AC-D: doxorubicin, cyclophosphamide, docetaxel or AC-DH: doxorubicin, cyclophosphamide, docetaxel, and trastuzumab) and a nonanthracycline one (DCbH: docetaxel, carboplatin, trastuzumab).[200] In its second interim efficacy analysis with a median follow-up of 36 months, there were 462 DFS events and 185 deaths. For DFS, the HR was 0.61 for patients in the AC-DH arm (95%CI, 0.48–0.76; P < .001) and 0.67 for patients in the DCbH arm (95%CI, 0.54–0.83; P=.003), compared with the AC-D. This translated to absolute benefits (from years 2 to 4) of 6% and 5%, respectively with the addition of trastuzumab. Nevertheless, longer follow-up is needed in patients in the DCbH arm to warrant the omission of anthracyclines in these patients.
The Finland Herceptin (FINHER) study assessed the impact of a much shorter course of trastuzumab.[201] In this trial, 232 women younger than 67 years with node-positive or high-risk (>2 cm tumor size) node-negativeHER2-overexpressing breast cancer were given nine weekly infusions of trastuzumab concurrently with docetaxel or vinorelbine followed by FEC. At a 3-year median follow-up, the risk of recurrence and/or death was significantly reduced in patients receiving trastuzumab (HR = 0.41; P = .01; 95% CI, 0.21–0.83; 3 year DFS = 89% vs. 78%). The difference in OS (HR = 0.41) was not statistically significant (P = .07; 95% CI, 0.16–1.08).[201][Level of evidence: 1iiA]
In the Adjuvant Lapatinib and/or Trastuzumab Treatment Optimization trial (ALTTO [NCT00553358]), the role of lapatinib (in combination with, in sequence to, in comparison to, or as an alternative to trastuzumab) in the adjuvant setting is being investigated. Lapatinib is a small molecule tyrosine kinase inhibitor that is capable of dual-receptor inhibition of both EGFR and HER2 and seems to be less cardiotoxic than trastuzumab. In phase I/II studies as a single agent, lapatinib has resulted in objective responses between 4.3% and 7.8% in ER2-positive patients who had progressed on multiple trastuzumab-containing regimens with a substantial number having stable disease at 4 months (34%–41%) and 6 months (18%–21%).[202] In a phase III trial (GSK-EGF100151), lapatinib plus capecitabine was superior to capecitabine alone in women with HER2-positive advanced breast cancer that has progressed after treatment with regimens that included an anthracycline, a taxane, and trastuzumab.[203] The hazard ratio for time to progression was 0.49 (95% CI, 0.34–0.71; P < .001), with 49 events in the combination-therapy group and 72 events in the monotherapy group. The median time to progression was 8.4 months in the combination-therapy group as compared with 4.4 months in the monotherapy group.
The combination of lapatinib and trastuzumab in the ALTTO trial is further supported by a demonstration that lapatinib combined with trastuzumab confers a significantly improved progression-free survival in patients with metastatic breast cancer who experience progression on prior trastuzumab-containing treatment when compared to lapatinib alone.[204]
In September 2011, the ALTTO trial was amended to discontinue the lapatinib alone arm of the trial. The Independent Data Monitoring Committee of the trial determined that the patients assigned to the lapatinib alone arm were not likely to do as well as the patients assigned to the trastuzumab alone control arm. Final results of this trial are pending.
Cardiotoxicity with adjuvant trastuzumab
In the HERA (BIG-01-01) trial, severe congestive heart failure (CHF NYHA class III–IV) occurred in 0.6% of patients treated with trastuzumab.[196] Symptomatic CHF occurred in 1.7% and 0.06% of patients in the trastuzumab and observation arms, respectively. Fifty-one patients experienced a confirmed left ventricular ejection fraction (LVEF) decrease (defined as an EF decrease of >10 points from baseline to an LVEF <50%) with trastuzumab, which recovered or stabilized within 3 to 6 weeks of initial treatment in 86% of cases. In theNSABP B-31 (NCT00004067) trial, 31 of 850 patients in the trastuzumab arm had confirmed symptomatic cardiac events, compared with 5 of 814 patients in the control arm.[205] The 3-year cumulative incidence of cardiac events for trastuzumab-treated patients was 4.1%, compared with 0.8% of patients in the control arm (95% CI, 1.7%–4.9%). Asymptomatic decline in LVEF (defined by >10% decline or to 55%) occurred in 17% of patients in the trastuzumab arm (95% CI, 15%–20%) and 34% of patients in the control arm (95%CI, 31%–38%), with a HR = 2.1 (95%CI, 1.7–2.6; P < .001). In the NCCTG-N9831 trial, 39 cardiac events were reported in the three arms over a 3-year period. The 3-year cumulative incidence of cardiac events in arm A was 0.35% (no trastuzumab), arm B, 3.5% (trastuzumab following paclitaxel) and arm C, 2.5% (trastuzumab concomitant with paclitaxel).
In the AVENTIS-TAX-GMA-302 (BCIRG 006) trial, clinically symptomatic cardiac events were detected in 0.38% of patients in the AC-D arm, 1.87% of patients in the AC-DH arm, and 0.37% of patients in the DCbH arm.[200] There was also a statistically significant higher incidence of asymptomatic and persistent decrease in LVEF in the AC-DH arm than with either the AC-D or DCbH arms. No cardiac deaths were reported in the AVENTIS-TAX-GMA-302 trial.
In the FINHER trial, none of the patients who received trastuzumab experienced clinically significant cardiac events. In fact, LVEF was preserved in all of the women receiving trastuzumab, but the number of patients receiving adjuvant trastuzumab was very low.
Treatment options for HER2-positive early breast cancer:
- Standard treatment is 1 year of adjuvant trastuzumab therapy. Results of the 1 year versus 2 years of trastuzumab of the HERA trial are awaited, as are the results of the Protocol of Herceptin Adjuvant with Reduced Exposure (INCA-PHARE) trial, which compared 1 year versus 6 months of adjuvant trastuzumab.