SAN ANTONIO – Exemestane was as effective as anastrozole – but no better – in the first large trial to compare nonsteroidal and steroidal aromatase inhibitors as adjuvant therapy in postmenopausal women with hormone receptor–positive early breast cancer.
At a median follow-up of 4 years, event-free survival was not significantly different between women receiving adjuvant exemestane (Aromasin) and those given anastrozole (Arimedex) (hazard ratio, 1.02; P = .85) in the randomized, phase III study.
Event rates were low in both arms, with 91% of patients in each arm disease free. Investigators reported no significant differences in overall survival (HR, 0.93; P = .64), distant disease-free survival (HR, 0.95; P =.46), or disease-specific survival (HR, 0.93; P = .62), either. Subgroup analyses by lymph-node status and prior chemotherapy did not alter the event-free survival results.
"We believe that exemestane is comparable to anastrozole, and provides a new option for 5 years of up-front adjuvant endocrine therapy," Dr. Paul E. Goss said, presenting results at the annual San Antonio Breast Cancer Symposium.
Researchers had hoped that a steroidal aromatase inhibitor (AI), such as exemestane, would exhibit additional antitumor effect beyond those seen with nonsteroidal AIs, such as anastrazole.
Anastrozole, a nonsteroidal reversible competitive AI, can "profoundly and specifically suppress circulating and intra- and peritumoral estrogens," noted Dr. Goss, director of breast cancer research at the Massachusetts General Hospital in Boston. It is approved for use as frontline adjuvant therapy in postmenopausal patients with hormone receptor–positive early breast cancer.
Exemestane, a steroidal irreversible suicide AI, is not approved for up-front adjuvant endocrine therapy, but rather for use after 2-3 years of initial tamoxifen. The study rationale was that exemestane may exert more potent estrogen suppression than would anastrozole. Exemestane has mild androgenic activity that suppresses sex hormone–binding globulin, and therefore could induce a secondary antitumor effect, the investigators reasoned. The androgenic/anabolic effects of exemestane might counteract bone loss, reduce cholesterol and triglyceride levels, and cause fewer menopausal symptoms.
The open-label, randomized, phase III NCIC-CTG (National Cancer Institute of Canada–Clinical Trials Group) MA 27 trial enrolled postmenopausal women with estrogen receptor–positive early primary breast cancer. Women could also be eligible if they were 55 years of age or younger without menses in 12 months but with postmenopausal FSH, or if they had undergone bilateral oophorectomy.
Patients were excluded if they had local or metastatic breast cancer, previous AI therapy, concurrent hormone therapy, concurrent selective estrogen receptor modulator therapy, or raloxifene treatment within 3 weeks of randomization.
In all, 7,576 women were recruited in 2003-2006 and randomized to either anastrozole 1 mg/day for 5 years or exemestane 25 mg/day for 5 years. Patients were stratified by lymph-node status, adjuvant chemotherapy use, and trastuzumab use.
In the original 2x2 design, patients in each treatment arm were randomly assigned to receive 400-mg celecoxib twice daily or placebo for 3 years, but this was abandoned after concerns about cardiovascular toxicities were raised. By then, 1,622 women had been randomized to celecoxib.
"In a meta-analysis of NCI celecoxib-containing trials, it was found that about 58% randomized to date in MA27 ... had moderate- to high-risk cardiac risk factors. Hence, the celecoxib/placebo randomization was discontinued," Dr. Goss said. "For the purposes of these results, the question of dual COX-2/aromatase inhibition remains unanswered."
He reported that roughly 70% of adverse events in the trial were grade 1-2. There was no significant difference between anastrazole and exemestane in menopausal-type adverse events (hot flashes, arthritis/arthralgia, and muscle pain), although vaginal bleeding was slightly less common with exemestane, said Dr. Goss. Abnormal results on liver function tests were more common in the exemestane arm.
"Importantly, cardiovascular effects were nonsignificant; there was no difference in the two arms," Dr. Goss noted. However, hypertriglyceridemia and hypercholesterolemia were less common with exemestane. Self-reported new diagnoses of osteoporosis were slightly more frequent with anastrozole. Clinical fractures (all fractures and fragility fractures) were the same for both groups.
By 3 years, about a quarter of patients in each group had discontinued treatment.
The trial was supported in part by Pfizer Inc., which makes Aromasin. Dr. Goss reported that he has received honoraria from Novartis, Pfizer, and AstraZeneca.