Background Pertuzumab greatly improves pathologic complete response (pCR) rates in early stage HER2-positive breast cancer. Long-term benefit of pertuzumab, however, is less well established, as follow-up of the NEOSPHERE and APHINITY trials did not show clear improvement in overall survival, although patients with high risk of recurrence (e.g. node positive) appeared to benefit. Since its long-term benefit remains uncertain, we compared outcome of patients treated with or without pertuzumab in a quasi-random experiment using real-world data.

Methods We identified all patients with stage II-III HER2-positive breast cancer in the Netherlands treated with neoadjuvant trastuzumab containing chemotherapy between November 2013 and January 2016 from the nationwide Netherlands Cancer Registry. During this period, reimbursement of pertuzumab in the Netherlands was pending and pertuzumab was only available as trial medication for patients in 37 hospitals that participated in the TRAIN-2 study. This setting created a unique opportunity to compare two quasi-random cohorts of patients treated with or without pertuzumab. We used logistic regression analysis to evaluate the association between pertuzumab use and pCR (ypT0/is, ypN0) and Kaplan-Meier estimates and Cox regression analysis for the association with overall survival (OS). Multivariate analyses included age, cT-status, cN-status, hormone receptor (HR) status and grade. Multiple imputation was used to impute missing data for multivariate analysis.

Results We identified 1,124 eligible patients of whom 453 (40%) had received pertuzumab. Baseline characteristics were comparable with and without pertuzumab: 61% of tumors were cT2, 22% cT3, 66% were node positive and 62% ER and/or PR-positive. Grade was missing for 17% in patients treated with and 46% in patients treated without pertuzumab and therefor imputed. PCR in breast and axilla could be determined in 1,091 patients. Pertuzumab use improved pCR rates (65% vs 41%, adjusted odds ratio [aOR] 3.01; 95% confidence interval [CI] 2.29-3.97; p<0.001). At a median follow-up of 59 months (IQR 53-66) 23 deaths had occurred in the pertuzumab group and 68 in the non-pertuzumab group (3-year OS 98% vs 95%; adjusted hazard ratio [aHR]: 0.61; 95% CI:0.38-1.00, p=0.048). Pertuzumab benefit appeared largest in ER/PR negative and cN+ tumors, although the number of events in each subgroup was too small for formal comparisons. Complete-case analysis showed similar aHRs, but with broader 95% confident intervals.

Conclusion This real-world quasi-experiment confirms the efficacy of pertuzumab to achieve a pCR in stage II and III HER2-positive breast cancer. In addition, these data suggest a small absolute overall survival benefit with pertuzumab, most prominently in hormone receptor negative and node positive tumors. Despite the unique setting of two quasi-random cohorts treated with or without pertuzumab and very similar baseline characteristics, residual confounding cannot be fully excluded. Breast cancer specific evaluation and translational work including central revision of tumor grade for missing cases is pending.

pCR (%)3-yr OS (%)
Ptzno PtzPtzno PtzaHR (95%CI)
Overall 65% 41% 98% 95% 0.61 (0.38-1.00) 
HR-positive 51% 32% 98% 97% 0.70 (0.36-1.33) 
HR-negative 86% 55% 97% 92% 0.54 (0.26-1.13) 
cN0 71% 46% 99% 100% 0.54 (0.14-2.00) 
cN+ 62% 38% 97% 93% 0.65 (0.39-1.09) 
pCR (%)3-yr OS (%)
Ptzno PtzPtzno PtzaHR (95%CI)
Overall 65% 41% 98% 95% 0.61 (0.38-1.00) 
HR-positive 51% 32% 98% 97% 0.70 (0.36-1.33) 
HR-negative 86% 55% 97% 92% 0.54 (0.26-1.13) 
cN0 71% 46% 99% 100% 0.54 (0.14-2.00) 
cN+ 62% 38% 97% 93% 0.65 (0.39-1.09) 

Citation Format: Anna van der Voort, Marte Liefaard, Mette van Ramshorst, Erik van Werkhoven, Astrid Scholten, Jelle Wesseling, Marie-Jeanne Vrancken Peeters, Linda de Munck, Gabe Sonke. Pathologic complete response and 3-year survival with or without pertuzumab using real-world data of stage II and III HER2-positive breast cancer [abstract]. In: Proceedings of the 2020 San Antonio Breast Cancer Virtual Symposium; 2020 Dec 8-11; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2021;81(4 Suppl):Abstract nr PS10-07.