Dose-Dense Chemo Ups Survival in Node+ Breast Cancer

SAN ANTONIO—A condensed, 14-day schedule of doxorubicin (Adria-mycin), cyclophosphamide, and pacli-taxel (Taxol) proved superior to conventional scheduling of the three drugs in a Cancer and Leukemia Group B (CALGB) trial reported at the 25th Annual San Antonio Breast Cancer Symposium (abstract 15). According to study investigator Marc L. Citron, MD, the dose-dense scheduling resulted in significant improvements in both disease-free and overall survival—and, with filgrastim (Neupogen) support, did not increase drug toxicity.

Beginning in 1997, the phase III trial (CALBG 9741) randomized 2,005 women with node-positive stage II/IIIa breast cancer to one of four regimens of Adria-mycin (A), 60 mg/m²; cyclophosphamide (C), 600 mg/m2; and Taxol (T) 175 mg/m². The regimens were designed to compare concurrent therapy (AC, followed by T, or AC-T) to sequential therapy (A-T-C) in both a conventional 21-day schedule and a dose-dense 14-day schedule (see Table). Cumulative and individual treatment doses were identical in each arm.

All patients on the two dose-dense regimens also received filgrastim (5 µg/kg, starting on day 3 for a total of seven doses). Upon completion of chemotherapy, all premenopausal, estrogen-receptor (ER)-positive patients and all postmenopausal patients were advised, but not required, to begin tamoxifen (Nolvadex) (20 mg/d) within 12 weeks.

Dr. Citron, clinical professor of medicine, Albert Einstein College of Medicine, Bronx, New York, noted that patient eligibility requirements were kept as broad as possible in order to include the types of patients seen in everyday practice. The study population included T0-T3, N1/N2 patients. Participants had undergone primary surgery with lump-ectomy and axillary lymph node dissection or modified radical mastectomy, with clear margins. Chest x-ray was the only specific staging criterion.

The researchers stipulated a baseline granulocyte count of 1,000 µL for entry into the trial for the first and every dose of chemotherapy in order to keep dose reductions and delays to a minimum.

The median patient age was 50, approximately 50% were premenopausal, and the median number of involved lymph nodes was 3. Approximately one third of the patients had ER-negative tumors, and about one third had undergone lumpectomy. Equal numbers of patients in all four arms received tamoxifen. Of the 2,005 patients enrolled, 1,973 received therapy and were evaluable.

At this first protocol-specified analysis at a median follow-up of 3 years, 315 patients had experienced recurrences, compared with the approximately 515 recurrences that had been expected.

At 4 years, the disease-free survival of patients on the two dose-dense schedules was 82%, compared with 75% for those on the two conventional 21-day schedules—for a 26% proportional reduction in the risk of breast cancer recurrence (P = .01). Study results showed no differences in efficacy between sequential and concurrent therapy.

Overall survival also was higher among patients on the dose-dense regimens, with 92% alive at 3 years vs 90% of those on a conventional schedule—a 31% overall reduction in mortality (P = .013).

No treatment-related deaths occurred. "The major difference in toxicity was granulocytopenia, which was much more significant in the every-3-week scheduling," Dr. Citron said. "Less than 1% of all cycles were accompanied by hospitalization for febrile neutropenia." Red blood cell transfusions were required in 13% of patients on the dose-dense concurrent regimen, compared with less than 4% in the other three treatment groups.

The investigators noted a slight increase in postchemotherapy neurologic toxicity among patients in the concurrent treatment arms, compared with sequential treatment.

The incidence of secondary acute myelogenous leukemia (AML) was 0.18%, similar to what was observed in CALBG 9344, the trial that made AC+T the standard adjuvant therapy, Dr. Citron said, and the incidence did not appear to be influenced by filgrastim. In fact, the incidence of AML was slightly higher in the conventional-scheduling arms.

Dose reductions were uncommon in all arms, although there were slightly more reductions in paclitaxel dose in the dose-dense group. Overall, fewer than 8% of cycles in all arms were delayed.

Conclusions and Caveats

Dose-dense treatment improved disease-free survival and overall survival by 26% and 31%, respectively, compared with conventional dose scheduling, Dr. Citron said. Moreover, these advantages were not accompanied by increased toxicity. "There are caveats," he said. "Additional follow-up is needed to confirm the survival benefits. Our maximum follow-up is only 5 years, and treatment-related patterns of toxicity and recurrence may emerge. And certainly filgrastim adds to the cost of the regimen."

He described the results as being "consistent with Dr. [Larry] Norton’s math-ematical models of tumor kinetics [see box], and we believe that these math-ematical concepts may be generalizable to other drugs and other diseases."

Dr. Norton, head of the Division of Solid Tumor Oncology, Memorial Sloan-Kettering, and senior author of the study, developed the mathematical model, known as the Norton-Simon hypothesis, along with Dr. Richard Simon of the National Cancer Institute.