Three-Year Follow-up of Neoadjuvant Chemotherapy With or Without Anthracyclines in the Presence of Dual ERBB2 Blockade in Patients With ERBB2-Positive Breast Cancer: A Secondary Analysis of the TRAIN-2 Randomized, Phase 3 Trial

Abstract

Importance: Primary analysis of the TRAIN-2 study showed high pathologic complete response rates after neoadjuvant chemotherapy with or without anthracyclines plus dual ERBB2 (formerly HER2) blockade.

Objective: To evaluate 3-year event-free survival (EFS) and overall survival (OS) of an anthracycline-free and anthracycline-containing regimen with dual ERBB2 blockade in patients with stage II and III ERBB2-positive breast cancer.

Design, setting, and participants: A total of 438 patients with stage II and III ERBB2-positive breast cancer were enrolled in this randomized, clinical, open-label phase 3 trial across 37 hospitals in the Netherlands from December 9, 2013, until January 14, 2016. Follow-up analyses were performed after a median follow-up of 48.8 months (interquartile range, 44.1-55.2 months). Analysis was performed on an intention-to-treat basis.

Interventions: Participants were randomly assigned on a 1:1 basis, stratified by age, tumor stage, nodal stage, and estrogen receptor status, to receive 3 cycles of fluorouracil (500 mg/m2), epirubicin (90 mg/m2), and cyclophosphamide (500 mg/m2), followed by 6 cycles of paclitaxel and carboplatin or 9 cycles of paclitaxel (80 mg/m2 days 1 and 8) and carboplatin (area under the concentration-time curve, 6 mg/mL/min). Both groups received trastuzumab (6 mg/kg; loading dose 8 mg/kg) and pertuzumab (420 mg intravenously; loading dose 840 mg) every 3 weeks.

Main outcomes and measures: Three-year EFS, OS, and safety.

Results: A total of 438 women were randomized, with 219 per group (anthracycline group, median age, 49 years [interquartile range, 43-55 years]; and nonanthracycline group, median age, 48 years [interquartile range, 43-56 years]). A total of 23 EFS events (10.5%) occurred in the anthracycline group and 21 EFS events (9.6%) occurred in the nonanthracycline group (hazard ratio, 0.90; 95% CI, 0.50-1.63; favoring nonanthracyclines). Three-year EFS estimates were 92.7% (95% CI, 89.3%-96.2%) in the anthracycline group and 93.6% (95% CI, 90.4%-96.9%) in the nonanthracycline group and 3-year OS estimates were 97.7% (95% CI, 95.7%-99.7%) in the anthracycline group and 98.2% (95% CI, 96.4%-100%) in the nonanthracycline group. The results were irrespective of hormone receptor and nodal status. A decline in left ventricular ejection fraction of 10% or more from baseline to less than 50% was more common in patients who received anthracyclines than those who did not (17 of 220 [7.7%] vs 7 of 218 [3.2%]; P = .04). Two patients treated with anthracyclines developed acute leukemia.

Conclusions and relevance: This follow-up analysis of the TRAIN-2 study shows similar 3-year EFS and OS estimates with or without anthracyclines in patients with stage II and III ERBB2-positive breast cancer. Anthracycline use is associated with increased risk of febrile neutropenia, cardiotoxic effects, and secondary malignant neoplasms.

Trial registration: ClinicalTrials.gov Identifier: NCT01996267.

Conflict of interest statement

Conflict of Interest Disclosures: Dr Wesseling reported receiving grants from KWF Dutch Cancer Society, Cancer Research UK, and ZonMW Health Research Netherlands outside the submitted work. Dr Linn reported receiving institutional research grants from Roche during the conduct of the study; institutional research grants from AstraZeneca, Eurocept Pharmaceuticals, Genentech/Roche, Novartis, Pfizer, Tesaro (now owned by GlaxoSmithKline), Immunomedics, and Agendia; unrestricted educational grants from Bayer and the Daiichi-Sankyo Faculty Endeavour Program; nonfinancial support from AstraZeneca, Genentech/Roche, Tesaro (now owned by GlaxoSmithKline), Novartis, and Immunomedics; serving on the scientific advisory board for Cergentis; and serving as an IBM advisory board member outside the submitted work. Dr Sonke reported receiving institutional research grants from Roche during the conduct of this study and institutional research grants from AstraZeneca, Merck, and Novartis outside the submitted work. No other disclosures were reported.

Figures

Figure 1.. Event-Free Survival and Overall Survival According to Treatment Group

A, Kaplan-Meier estimates of event-free survival in the intention-to-treat population, 4 years after randomization of the last patient. B, Kaplan-Meier estimates of overall survival in the intention-to-treat population, 4 years after randomization of the last patient. HR indicates hazard ratio.

Figure 2.. Subgroup Analysis Event-Free Survival According to Treatment Arm

Hazard ratios are for event-free survival in Cox proportional hazards regression models (including subgroup, treatment, and interaction) including 95% CI (P value for interaction not given because the analysis is descriptive). The sizes of the data markers are proportional to the number of patients. aHazard ratio less than 1 favors nonanthracyclines. bDisease stage II includes 1 patient with stage I disease (nonanthracycline group) and disease stage III includes 1 patient with stage IV disease (anthracycline group). cTumor grade was missing for 11 patients in both groups.

Figure 3.. Disease-Free Survival According to Response

Kaplan-Meier estimates of disease-free survival, 4 years after randomization of the last patient by pathologic complete response (pCR). HR indicates hazard ratio.