Pathologic complete response to neoadjuvant therapy has been proposed as a surrogate endpoint for long-term clinical benefit in breast cancer. The U.S. Food and Drug Administration (FDA) established the international Collaborative Trials in Neoadjuvant Breast Cancer (CTNeoBC) working group to perform a pooled analysis of neoadjuvant trials in breast cancer to assess correlation of different definitions of pathologic complete response with event-free survival and overall survival and to assess whether pathologic complete response could be used as a surrogate endpoint for these outcomes.

Patricia Cortazar, MD, Medical Team Leader and Scientific Liaison, Breast Oncology Group, in the Division of Oncology Products 1, Office of Hematology and Oncology Products, FDA, Silver Spring, Maryland, and colleagues reported on pathologic complete response and long-term clinical benefit in The Lancet.¹ The three most commonly used definitions of pathologic complete response—ypT0 ypN0 (absence of invasive cancer and in-situ cancer in the breast and axillary nodes), ypT0/is ypN0 (absence of invasive cancer in the breast and axillary nodes, irrespective of ductal carcinoma in situ), and ypT0/is (absence of invasive cancer in the breast irrespective of ductal carcinoma in situ or nodal involvement)—were assessed for association with event-free and overall survival. Response defined as ypT0 ypN0 or ypT0/is ypN0 was associated with improved long-term outcome, with prognostic value being greatest in aggressive tumor subtypes. However, a trial-level analysis did not support use of pathologic complete response as a surrogate endpoint for improved event-free survival and overall survival.

Study Details

The pooled analysis included 11,955 patients from 12 international neoadjuvant trials that met eligibility criteria of having ≥ 200 patients with primary breast cancer treated with preoperative chemotherapy followed by surgery, having available data for pathologic complete response, event-free survival, and overall survival, and having median follow-up ≥ 3 years.

A trial-level analysis was performed to evaluate the potential for using pathologic complete response as a surrogate endpoint for event-free survival or overall survival. The associations between treatment effects on pathologic complete response and event-free survival and overall survival were quantified using a weighted linear regression model on a logarithmic scale.

Treatment effects on event-free and overall survival were represented as hazard ratios (HRs, calculated with Cox proportional hazards models) and on pathologic complete response as odds ratios (calculated with logistic regression models) within each randomized comparison from among the eligible trials. A two-stage model (adjusting for the estimation error of treatment effect size estimates) was used to calculate the coefficient of determination (R²) to measure the correlation between pathologic complete response and event-free and overall survival by treatment effect.

Definitions and Survival

The frequency of pathologic complete response was 22% for ypT0/is, 18% for ypT0/is ypN0, and 13% for ypT0 ypN0. Overall, patients who achieved a pathologic complete response on any definition had longer event-free survival and overall survival than did those with residual invasive cancer.

Hazard ratios for event-free survival for ypT0 ypN0 (0.44, 95% confidence interval [CI] = 0.39–0.51) and for ypT0/is ypN0 (0.48, 95% CI 0.43–0.54) were better than that for ypT0/is (HR = 0.60, 95% CI = 0.55–0.66). Similarly, ypT0 ypN0 (HR = 0.36, 95% CI = 0.30–0.44) and ypT0/is ypN0 (HR = 0.36, 95% CI = 0.31–0.42) were more closely associated with overall survival than was ypT0/is (HR = 0.51, 95% CI = 0.45–0.58). The ypT0/is ypN0 definition was thus used in all other analyses.

Response and Outcome

The rate of pathologic complete response using this definition was low in patients with low-grade hormone receptor–positive tumors and more than doubled for high-grade hormone receptor–positive tumors. Rates of response were increased in HER2-positive and triple-negative tumors, and highest in patients with hormone receptor–negative tumors and those treated with trastuzu­mab (Herceptin) among the HER2-positive subgroup.

The hazard ratios for pathologic complete response were 0.49 (95% CI = 0.33–0.71) for event-free survival and 0.43 (95% CI = 0.23–0.71) for overall survival in the overall hormone receptor–positive, HER2-negative subgroup. The association between response and long-term outcome was stronger in patients with hormone receptor–positive, HER2-negative high-grade (HR = 0.27, 95% CI = 0.14–0.50, for event-free survival; HR = 0.29, 95% CI = 0.13–0.65, for overall survival) vs low-grade tumors (HR = 0.63, 95% CI = 0.38–1.04, for event-free survival; HR = 0.47, 95% CI = 0.21–1.07, for overall survival).

Among patients with HER2-positive tumors, pathologic complete response was associated with event-free survival (HR = 0.39, 95% CI = 0.31–0.50) and overall survival (HR = 0.34, 95% CI = 0.24–0.47) regardless of hormone receptor status, with the association having greater strength in the HER2-positive, hormone receptor–negative subgroup (HR = 0.25, 95% CI = 0.18–0.34, for event-free survival; HR = 0.19, 95% CI = 0.12–0.31, for overall survival) and being weaker in the hormone receptor–positive subgroup (HR = 0.58, 95% CI = 0.42–0.82, for event-free survival; HR = 0.56, 95% CI = 0.23–1.37, with trastuzumab, and 0.57, 95% CI = 0.31–1.04, without trastuzumab for overall survival).

The association between pathologic complete response and long-term outcomes was strongest in patients with triple-negative breast cancer and patients with HER2-positive, hormone receptor–negative tumors.

Surrogate for Survival?

The trial-level analysis performed to assess whether pathologic complete response could be used as a surrogate endpoint for event-free survival or overall survival included 10 randomized comparisons from among the 12 trials in the pooled analysis and excluded nonrandomized neoadjuvant groups and patients who received additional adjuvant taxane therapy. At the trial level, there was little association between increases in frequency of pathologic complete response and the effect of treatment on event-free or overall survival.

The coefficient of determination (R²) between improvement in pathologic complete response and event-free survival was 0.03 (95% CI = 0.00–0.25) and that between pathologic complete response and overall survival was 0.24 (95% CI = 0.00–0.70). Analyses of trial-level association between response and outcome by tumor subtype also showed no correlation between increased frequency of pathologic complete response and treatment effect on event-free or overall survival.

The investigators concluded, “Patients who attain pathological complete response defined as ypT0 ypN0 or ypT0/is ypN0 have improved survival. The prognostic value is greatest in aggressive tumor subtypes. Our pooled analysis could not validate pathological complete response as a surrogate endpoint for improved [event-free and overall survival].” ■

Disclosure: The study was funded by the FDA. For full disclosures of the study authors, visit www.thelancet.com.

Reference

1. Cortazar P, Zhang L, Untch M, et al: Pathological complete response and long-term clinical benefit in breast cancer. Lancet. February 14, 2014 (early release online).