An Evaluation of the Association Between pCR and Survival Outcomes in the I-SPY2 Study

A follow-up analysis of results of the I-SPY2 trial of neoadjvuant therapy in high-risk, early-stage breast cancer provided support for the use of pathologic complete response (pCR) as a prognostic biomarker in this setting. These findings from the I-SPY2 Trial Consortium were published in JAMA Oncology.

The I-SPY2 trial (ClinicalTrial.gov Identifier: NCT01042379) is an ongoing, open-label,  adaptive, randomized, phase 2 study with parallel arms that is evaluating multiple investigational agents in combination with standard neoadjuvant chemotherapy compared with standard neoadjuvant chemotherapy alone in patients with high-risk, early-stage breast cancer.

Adaptive randomization in the I-SPY2 study is performed across 8 breast cancer subtypes defined by hormone receptor and ERBB2 status, as well as the risk of disease recurrence according to a 70-gene assay.

Furthermore, the I-SPY2 trial has a platform design in which arms are being opened and closed over the course of the study.

The primary endpoint of the I-SPY2 study is pCR, defined as the absence of invasive disease in breast and axillary nodes. Based on data from a meta-analysis of results of clinical trials of neoadjuvant therapy in breast cancer, pCR was recognized by the US Food and Drug Administration (FDA) as a primary endpoint to support accelerated approval of drugs and biologics in the setting of high-risk, early-stage breast cancer.^2,3

While the authors of this analysis noted that “the goal of the I-SPY2 trial was to rapidly identify active agents and combinations to increase the chance of achieving pCR,” they clarified that the goal of this analysis was to investigate whether there is “an association between pCR and survival end points in neoadjuvant treatment of early breast cancer with various novel therapeutics.”

Toward that end, the relationships between pCR status and 3-year EFS and distant recurrence-free survival (DRFS) were investigated for the first 950 patients randomly assigned to receive 10 different therapies — including standard chemotherapy alone — in the I-SPY2 trial.

A key study finding was that at a median follow-up of 3.8 years, 3-year event-free survival (EFS) was observed for 95% and 78% of those achieving pCR vs not, respectively (hazard ratio [HR], 0.19, 95% CI, 0.12-0.31; P <.001). Similar findings were observed when 3-year DRFS was compared for those with pCR (95%) vs non-pCR (81%; HR, 0.21; 95% CI, 0.13-0.34; P <.001).

These results were seen “regardless of high-risk subtype and type of treatment,” the study authors emphasized.

The findings of this study “suggest that pCR may be a robust prognostic biomarker for excellent long-term outcomes at the individual patient level for patients with early, high-risk breast cancer,” the authors wrote.

In an accompanying editorial, authors Yu Shyr, PhD, and Derek Shyr, MS, wrote that the I-SPY2 investigators “did not provide the evidence needed to validate pCR as a surrogate outcome for EFS and DRFS.”⁴

In that editorial, Drs Shyr and Shyr argued that although the I-SPY2 Trial Consortium results showed an association between pCR and the true survival endpoints of EFS and DRFS, validation of pCR as a true surrogate endpoint would require demonstration of an “association between the effect of treatment on the surrogate end point (eg, odds ratio for pCR), and the effect of treatment on the true end point (eg, hazard ratio for EFS or DRFS).”⁴

“The I-SPY2 trial is not powered or intended to evaluate whether an increase in pCR rate translates to an improvement in EFS for individual experimental therapies relative to control,” the study authors noted in their concluding remarks. “Instead, I-SPY2 is designed to identify large pCR improvements and provide the rationale for larger confirmatory trials in appropriate subsets of patients.”

Disclosure: Some of the authors of the I-SPY2 Trial Consortium reported financial relationships with the pharmaceutical industry and/or the medical device industry. For a full list of disclosures, please refer to the original study.

References