乌鸦传媒

Five days of prophylactic granulocyte colony-stimulating factor (G-CSF) proved as effective as 7 to 10 days for preventing chemotherapy-induced febrile neutropenia in patients with early breast cancer, a randomized trial showed.

Patients assigned to the 5-day regimen of filgrastim (Neupogen) had a 2.09% incidence of febrile neutropenia or treatment-related hospitalization per cycle of chemotherapy versus 3.60% with 7 to 10 days of G-CSF. The difference fell within the prespecified threshold of 3% for noninferiority.

Analysis of the two components of the composite endpoint showed a 0.11% difference per cycle in febrile neutropenia and -1.68% for treatment-related hospitalizations between the 5-day and 7- to 10-day regimens, as reported online in .

"Despite its considerable cost and widespread use for two decades, the most effective duration of filgrastim as primary febrile neutropenia prophylaxis for early breast cancer patients has remained unknown," Mark Clemons, MD, of the Ottawa Hospital Cancer Centre in Ontario, Canada, and co-authors wrote in their discussion of the findings. "This may in part reflect the fact that many studies in the literature are retrospective in nature with resulting methodological challenges. To our knowledge, this is the only randomized trial to prospectively answer this important clinical question."

The group added that "given the recognized toxicity and cost of this agent, as well as the impact on health economics and patient morbidity, the shorter duration should be considered standard of care."

The findings will likely have limited impact on clinical practice in the U.S., where the standard for most patients is a long-acting growth factor that requires a single administration the day after chemotherapy, said Joanne Mortimer, MD, a breast cancer specialist at City of Hope in Duarte, California.

"Most often we use the growth factors that are long acting and they get one shot the day after chemo, whether it's by a device or whether it's coming into clinic and getting a shot," Mortimer told 乌鸦传媒. "[Patients] don't have to come in for five or seven days for treatment."

Exceptions include patients whose insurance does not cover the long-acting agents and patients who receive weekly chemotherapy, which would overlap and conflict with the activity of the long-acting growth factors.

"This would probably apply to a fairly small population of patients in the U.S.," said Mortimer.

The findings came from a multicenter randomized trial involving 466 patients with early-stage breast cancer. The patients received either a 5-day course of filgrastim (n=184) or 7 to 10 days of growth factor support (n=282).

The primary endpoint was the difference in the composite of febrile neutropenia or treatment-related hospitalization during each cycle of chemotherapy. Investigators defined febrile neutropenia as an absolute neutrophil count <0.5 × 10⁹/L in association with a temperature >38.3° C (~100.9° F) or two consecutive temperature readings >38.0° C for 2 hours. Key secondary outcomes included rates of dose delay/reduction and discontinuation.

The patients had a median age of 56, and 22.5% of the study participants were 65 or older. The most commonly used chemotherapy regimens were paclitaxel-cyclophosphamide (33.9%), dose-dense doxorubicin-cyclophosphamide and paclitaxel (29%), and fluorouracil-epirubicin-cyclophosphamide plus docetaxel with filgrastim beginning in cycle 1 (6.9%) or cycle 4 (21%). The filgrastim dose was 300 µg in 97.2% of evaluable cases.

Patients in the 5-day arm received 897 cycles of chemotherapy, whereas those randomized to the 7- to 10-day schedule received 1,302 cycles.

The primary endpoint was evaluated by per-protocol analysis, but investigators also performed an intention-to-treat (ITT) analysis. Investigators excluded 14 patients from the per-protocol analysis, consisting of three patients found to have metastatic disease, two who withdrew consent, four who refused G-CSF injections, three whose physicians prescribed peg-filgrastim (Neulasta), and two patients who refused chemotherapy.

The 1.52% lower rate of the composite endpoint in the 5-day arm satisfied statistical criteria for noninferiority. The ITT analysis yielded a 0.78% lower rate of the composite endpoint in the 5-day arm, also meeting noninferiority criteria. Treatment-related hospitalization accounted for most of the difference.

Per-protocol analysis of secondary outcomes showed that 34.64% of patients in the 5-day arm had at least one chemotherapy dose reduction, delay, or discontinuation as compared with 37.5% of patients who received extended-duration G-CSF. The ITT analysis showed rates of 37.64% in the 5-day arm and 39.42% in the 7- to 10-day arm.

Comment