Background
Phase III RCT in node-positive (T0-3, N1/2, M0) breast adenocarcinoma testing dose density (q2wk vs q3wk) and sequence (sequential vs concurrent) of adjuvant doxorubicin, paclitaxel, and cyclophosphamide in a 2x2 design.
Interventions and follow up
Arm A: Sequential A x4 → T x4 → C x4, q3wk
Arm B: Sequential A x4 → T x4 → C x4, q2wk + filgrastim
Arm C: Concurrent AC x4 → T x4, q3wk
Arm D: Concurrent AC x4 → T x4, q2wk + filgrastim
Primary endpoint: DFS and OS
mFollow up: 36mo
Arm B: Sequential A x4 → T x4 → C x4, q2wk + filgrastim
Arm C: Concurrent AC x4 → T x4, q3wk
Arm D: Concurrent AC x4 → T x4, q2wk + filgrastim
Primary endpoint: DFS and OS
mFollow up: 36mo
Results
1-yr DFS (dose-dense vs conventional): 97% vs 95%
2-yr DFS: 91% vs 87%
3-yr DFS: 85% vs 81%
4-yr DFS: 82% vs 75%
OS: Significantly prolonged with dose-dense; RR 0.69, P=.013
Sequence: No DFS/OS difference between concurrent and sequential; no density-sequence interaction
2-yr DFS: 91% vs 87%
3-yr DFS: 85% vs 81%
4-yr DFS: 82% vs 75%
OS: Significantly prolonged with dose-dense; RR 0.69, P=.013
Sequence: No DFS/OS difference between concurrent and sequential; no density-sequence interaction
Adverse events
Hematologic: Severe (grade 4) neutropenia less frequent in dose-dense arms due to filgrastim support.
Overall: Dose-dense scheduling was well tolerated with growth-factor support.
Overall: Dose-dense scheduling was well tolerated with growth-factor support.
Conclusions
Dose-dense (q2wk) adjuvant chemotherapy with filgrastim reduces recurrence (26% relative) and death (31% relative) versus conventional q3wk dosing; sequence does not matter.
Key Limitations
Short 36-month follow-up at primary report. No HER2 stratification. TNBC subgroup recurrence reduction (32% vs 19%) not statistically significant. Mandatory filgrastim only in dose-dense arms confounds the schedule comparison.
Clinical Context
Established dose-dense AC→T (with G-CSF) as a preferred adjuvant regimen for higher-risk early breast cancer per ASCO/ESMO, later supported by the EBCTCG dose-dense meta-analysis. Built on the AC→T paradigm from CALGB 9344.