Seven-Year Follow-Up Analysis of Adjuvant Paclitaxel and Trastuzumab Trial for Node-Negative, Human Epidermal Growth Factor Receptor 2-Positive Breast Cancer

Abstract

Purpose: The Adjuvant Paclitaxel and Trastuzumab trial was designed to address treatment of patients with small human epidermal growth factor receptor 2 (HER2)-positive breast cancer. The primary analysis of the Adjuvant Paclitaxel and Trastuzumab trial demonstrated a 3-year disease-free survival (DFS) of 98.7%. In this planned secondary analysis, we report longer-term outcomes and exploratory results to characterize the biology of small HER2-positive tumors and genetic factors that may predispose to paclitaxel-induced peripheral neuropathy (TIPN).

Patients and methods: In this phase II study, patients with HER2-positive breast cancer with tumors 3 cm or smaller and negative nodes received paclitaxel (80 mg/m²) with trastuzumab for 12 weeks, followed by trastuzumab for 9 months. The primary end point was DFS. Recurrence-free interval (RFI), breast cancer-specific survival, and overall survival (OS) were also analyzed. In an exploratory analysis, intrinsic subtyping by PAM50 (Prosigna) and calculation of the risk of recurrence score were performed on the nCounter analysis system on archival tissue. Genotyping was performed to investigate TIPN.

Results: A total of 410 patients were enrolled from October 2007 to September 2010. After a median follow-up of 6.5 years, there were 23 DFS events. The 7-year DFS was 93% (95% CI, 90.4 to 96.2) with four (1.0%) distant recurrences, 7-year OS was 95% (95% CI, 92.4 to 97.7), and 7-year RFI was 97.5% (95% CI, 95.9 to 99.1). PAM50 analyses (n = 278) showed that most tumors were HER2-enriched (66%), followed by luminal B (14%), luminal A (13%), and basal-like (8%). Genotyping (n = 230) identified one single-nucleotide polymorphism, rs3012437, associated with an increased risk of TIPN in patients with grade 2 or greater TIPN (10.4%).

Conclusion: With longer follow-up, adjuvant paclitaxel and trastuzumab is associated with excellent long-term outcomes. Distribution of PAM50 intrinsic subtypes in small HER2-positive tumors is similar to that previously reported for larger tumors.

Trial registration: ClinicalTrials.gov NCT00542451.

FIG 1.

Trial enrollment and follow-up.

FIG 2.

Disease-free survival (DFS). (A) Kaplan-Meier plot of DFS in the intention-to-treat population. (B) DFS according to hormone-receptor status. Abbreviations: neg, negative; Point est, point estimate; pos, positive.

FIG A1.

Tumor sample analyses. (*) Tumors less than 4 mm were excluded (insufficient invasive tissue). (†) NanoString set a predefined threshold to maintain the integrity of the subtyping call. A 100-count Geomean for the housekeeping genes (HKs) is the minimal amount needed to generate a confident subtype call. The algorithm requires data above the background to make a reasonable call of subtype, and HK was used as a marker for the rest of the targets. Cases with borderline HK Geomean were excluded from the risk of recurrence (ROR) score evaluation. (‡) Five cases previously incorrectly marked as failed subtyping were corrected and included in the ROR scoring. Therefore, there is a total discrepancy of 14 cases between the number of cases with PAM50 (Prosigna) results and number of cases with ROR scores. QNS, quantity not sufficient.