Given the similar efficacy of tamoxifen and anastrozole for women older than age 60 years, decisions about treatment should be informed by the risk for serious adverse health effects and the symptoms associated with each drug.
—Patricia A. Ganz, MD, and colleagues
No clear efficacy differences were seen between the two treatments [anastrozole vs tamoxifen].
—Jack Cuzick, PhD, and colleagues
Two recently reported phase III trials compared adjuvant anastrozole vs tamoxifen in postmenopausal women with hormone receptor–positive ductal carcinoma in situ. As reported in The Lancet by Margolese et al,1 the National Surgical Adjuvant Breast and Bowel Project (NSABP) B-35 trial, performed in women who had received lumpectomy plus whole-breast irradiation, anastrozole significantly improved the breast cancer–free interval, although the benefit was limited to patients aged < 60 years. Toxicity differed between treatments, and a study in the NASBP B-35 population, reported in The Lancet by Ganz et al,2 showed differences in patient-reported outcomes. As reported in The Lancet by Forbes et al,3 the IBIS-II DCIS trial, performed in women who had undergone local excision with or without radiotherapy, anastrozole was noninferior to tamoxifen in preventing overall disease recurrence.
Richard G. Margolese, MD, of Jewish General Hospital, McGill University, is the corresponding author of the NSABP B-35 article in The Lancet. Patricia A. Ganz, MD, of Jonsson Comprehensive Cancer Center, University of California, Los Angeles, is the corresponding author of the NSABP B-35 patient-reported outcomes article in The Lancet. Jack Cuzick, PhD, of the Centre for Cancer Prevention, Wolfson Institute of Preventive Medicine, Queen Mary University of London, is the corresponding author of the IBIS-II DCIS article in The Lancet. (For more coverage on these two trials from the 2015 San Antonio Breast Cancer Symposium, see pages 24 and 25.)
NSABP B-35 Trial
In this double-blind trial, 3,104 patients from 333 sites in the United States, Canada, and Mexico were randomized between January 2003 and June 2006 to receive daily anastrozole at 1 mg (n = 1,552) or tamoxifen at 20 mg (n = 1,552) for 5 years.1 Randomization was stratified by age < 60 vs ≥ 60 years. The primary outcome was breast cancer–free interval defined as time from randomization to any breast cancer event, consisting of local, regional, or distant recurrence, or contralateral breast cancer, invasive disease, or ductal carcinoma in situ, on intention-to-treat analysis.
The anastrozole and tamoxifen groups were generally balanced for age (47% in both < 60 years), race/ethnicity (88% and 87% white), tumor evident on mammogram (98% and 96%), comedo necrosis (43% and 47%), palpable tumor (8% and 9%), tumor size (≥ 1 cm in 25% and 24%, unknown for 41% and 40%), and body mass index (< 25 kg/m2 for 24% and 26%).
Reduced Recurrence With Anastrozole
Median follow-up was 9.0 years, with a total of 212 breast cancer–free interval events observed. Events occurred in 90 patients in the anastrozole group vs 122 in the tamoxifen group (hazard ratio [HR] = 0.73, P = .0234). A significant time-by-treatment interaction was observed (P = .0410), with event curves diverging after the first 60 months.
A significant interaction was also observed between treatment and age group (P = .0379). Among women aged < 60 years, events occurred in 34 anastrozole vs 63 tamoxifen patients (HR = 0.53, P = .0026); among those aged ≥ 60 years, events occurred in 56 vs 59 (HR = 0.95, P = .78). Five-year breast cancer–free interval rates were 96.3% in both groups, and estimated 10-year rates were 93.1% vs 89.1%.
Among all events, invasive disease occurred in 43 vs 69 patients (HR = 0.62, P = .0123), and ductal carcinoma in situ occurred in 47 vs 53 (HR = 0.88, P = .52). Ipsilateral recurrence was observed in 46 vs 55 patients (HR = 0.83, P = .34), including invasive disease in 17 vs 22 (HR = 0.39, P = .39) and ductal carcinoma in situ in 29 vs 33 (HR = 0.87, P = .59). Contralateral disease was observed in 39 vs 60 patients (HR = 0.64, P = .0322), including invasive disease in 21 vs 40 (HR = 0.52, P = .0148) and ductal carcinoma in situ in 18 vs 20 (HR = 0.90, P = .73). Disease at a distant site was observed in four vs seven patients (HR = 0.57, P = .37).
A total of 186 deaths were observed, including 98 in the anastrozole group vs 88 in the tamoxifen group (HR = 1.11, P = .48). Estimated overall survival was 97.9% vs 98.0% at 5 years and 92.5% vs 92.1% at 10 years. No interaction between treatment and age group was observed (P = .38). Death due to breast cancer occurred in five vs eight patients.
Second primary cancers occurred in 107 patients in the anastrozole group vs 102 in the tamoxifen group (risk ratio [RR] = 1.04, 95% confidence interval [CI] = 0.78–1.37), with a nonsignificantly greater incidence of uterine cancer in the tamoxifen group (8 vs 17 cases, RR = 0.47, 95% CI = 0.18–1.15). Fracture was nonsignificantly more common with anastrozole (69 vs 50 patients, RR = 1.38, 95% CI = 0.95–2.03).
Other adverse events did not markedly differ between the groups, except for a higher incidence of thrombosis or embolism in the tamoxifen group (0.8% vs 2.7%); there were 17 grade 4 events in tamoxifen patients and 3 grade 4 events and 1 grade 5 event in anastrozole patients. Overall, rates of arthralgia and myalgia of any grade were somewhat increased in the anastrozole group (grade ≥ 3 arthralgia in 5% vs 4%, myalgia in 2% vs 1%).
A total of 1,193 patients, consisting of 592 anastrozole patients and 601 tamoxifen patients, were included in the analysis of patient-reported outcomes.2 Patients were assessed every 6 months, with the primary outcomes being the Medical Outcomes Study (MOS) Short Form (SF)-12 physical and mental health component scale scores and vasomotor symptoms on the Breast Cancer Prevention Trial (BCPT) symptom scale. Other outcomes were assessed using the SF-36 vitality scale, Center for Epidemiologic Studies Depression Scale, and MOS sexual problems scale.
No significant differences between groups were observed for physical health scores (P = .20), mental health scores (P = .38), energy and fatigue (P = .86), symptoms of depression (P = .46), or sexual functioning (P = .56) over 5 years. Vasomotor symptoms (P = .011), difficulty with bladder control (P = .0002), and gynecologic symptoms (P < .0001) were rated as significantly more severe by patients in the tamoxifen group; the difference in vasomotor symptoms was significant at 6, 12, 30, and 36 months. Musculoskeletal pain (P = .0006) and vaginal symptoms (P = .035) were rated as significantly worse in the anastrozole group; the difference in musculoskeletal pain was significant at 6, 12, and 24 months. After adjustment for treatment, age < 60 years vs ≥ 60 years was associated with more severe vasomotor symptoms (P = .0006), vaginal symptoms (P < .0001), weight problems (P < .0001), and gynecologic symptoms (P = .014).
The investigators concluded: “Given the similar efficacy of tamoxifen and anastrozole for women older than age 60 years, decisions about treatment should be informed by the risk for serious adverse health effects and the symptoms associated with each drug. For women younger than 60 years old, treatment decisions might be driven by efficacy (favoring anastrozole); however, if the side effects of anastrozole are intolerable, then switching to tamoxifen is a good alternative.”
IBIS-II DCIS Trial
In this double-blind trial, 2,980 women with locally excised ductal carcinoma in situ from 236 sites in 14 countries (Europe, Australia, New Zealand, Chile) were randomized between March 2003 and February 2012 to receive daily anastrozole at 1 mg (n = 1,449 included in analysis) or tamoxifen at 20 mg (n = 1,489 included in analysis) for 5 years.3 The primary endpoint was all recurrence of disease, including recurrent ductal carcinoma in situ, invasive disease, and new contralateral tumors. Analyses were performed on a modified intent-to-treat basis, including all randomized women who did not withdraw consent for data to be included (22 anastrozole and 20 tamoxifen patients).
Patients in the anastrozole and tamoxifen groups were generally balanced for age (median 60 years in both), body mass index (median 26.7 kg/m2 in both), age at menarche (median 13 years in both), smoking status (34% and 33% ever), menopausal hormone therapy (47% and 44%), hysterectomy (21% and 22%), radiotherapy (71% in both), tumor size (median 13 mm in both), margins (median 5 mm in both), tumor grade (low in 20% and 19%, intermediate in 42% in both, high in 37% and 39%), and laterality (left in 51% and 53%, right in 49% and 47%, bilateral in < 1% in both).
No Difference in Recurrence Rates
Median follow-up was 7.2 years, with 144 breast cancer recurrences observed. Overall, recurrence was observed in 67 anastrozole patients (5%) vs 77 tamoxifen patients (5%), for annual rates of 0.64% vs 0.72%, and a hazard ratio of 0.89 and 95% CI of 0.64–1.23; noninferiority of anastrozole was thus established (upper 95% CI < 1.25), but superiority was not (P = .49).
On analysis adjusting for age, body mass index, menopausal hormone therapy, grade, margins, and radiotherapy, the hazard ratio was 0.83 (95% CI = 0.59–1.18, P = .31). Invasive recurrence was observed in 37 (3%) vs 47 (3%; adjusted HR = 0.72, P = .16), including ipsilateral disease in 20 (1%) vs 22 (1%; adjusted HR = 0.77, P = .44) and contralateral disease in 17 (1%) vs 25 (2%; adjusted HR = 0.68, P = .24). Ductal carcinoma in situ recurrence was observed in 29 (2%) vs 30 (2%; adjusted HR = 0.98, P = .95), including ipsilateral recurrence in 21 (1%) vs 23 (2%; adjusted HR = 1.03, P = .93) and contralateral recurrence in 8 (< 1%) vs 6 (< 1%; adjusted HR = 1.02, P = .97). Estimates of overall recurrence were 2.5% vs 3.0% at 5 years and 6.6% vs 7.3% at 10 years.
A total of 69 deaths were observed, consisting of 33 in the anastrozole group vs 36 in the tamoxifen group (HR = 0.93, P = .78), with causes of death not differing between the groups. Death from breast cancer occurred in one vs three patients.
Differences in Adverse-Event Profiles
Adverse events of any grade were reported in 91% and 93% of patients. The frequencies of fracture (9% vs 7%, P = .027), musculoskeletal events (64% vs 54%, P < .0001), hypercholesterolemia (3% vs 1%, P < .0001), and transient ischemic attacks (1% vs < 1%, P = .05), particularly cerebrovascular accident (1% vs < 1%, P = .025), were higher in the anastrozole group. The frequencies of muscle spasm (2% vs 7%, P < .0001), vasomotor or gynecologic symptoms (61% vs 69%, P < .0001), including higher rates of hot flushes, hemorrhage, discharge, candidiasis, and deep-vein thrombosis (< 1% vs 1%, P = .0011) were higher in the tamoxifen group. There was no significant difference in occurrence of other cancers between groups (61 vs 71, P = .47), but the frequencies of all gynecologic cancers (1 vs 17, P = .0002), endometrial cancer (1 vs 11, P = .0044), ovarian cancer (0 vs 5, P = .027), and nonmelanoma skin cancer (8 vs 19, P = .040) were higher in the tamoxifen group.
The investigators concluded: “No clear efficacy differences were seen between the two treatments. Anastrozole offers another treatment option for postmenopausal women with hormone receptor–positive ductal carcinoma in situ, which may be more appropriate for some women with contraindications for tamoxifen. Longer follow-up will be necessary to fully evaluate treatment differences.” ■
Disclosure: The NSABP B-35 trial was funded by the National Cancer Institute and AstraZeneca. The IBIS-II DCIS trial was funded by Cancer Research UK, National Health and Medical Research Council Australia, Breast Cancer Research Fund, AstraZeneca, and Sanofi Aventis. For full disclosures of the authors of these studies, visit www.thelancet.com.