Results from the Investigation of Serial Studies to Predict Your Therapeutic Response With Imaging and Molecular Analysis 2 (I-SPY2) trial showed that pathological complete response (pCR) predicted 3-year event-free survival (EFS) and distant recurrence-free survival (DRFS) (Abstract GS3-08). These findings support using pCR as a surrogate endpoint for neoadjuvant trials to more quickly evaluate new drugs in combination with standard chemotherapy for breast cancer patients. The study results were presented at the 2017 San Antonio Breast Cancer Symposium.
pCR as a Surrogate Endpoint
The FDA can grant new neoadjuvant drugs accelerated approval for use in breast cancer patients if they demonstrate a pCR that predicts long-term clinical benefit (Lancet 2014;384(9938):164-172). Accelerated approval allows patients access to the drug, but a larger clinical trial and reporting of long-term clinical outcomes are still required to gain full FDA approval. For example, pertuzumab was granted accelerated approval in 2013 for use in combination with trastuzumab and chemotherapy in the neoadjuvant setting. It was recently granted full approval in December 2017 when the long-term clinical outcomes were reported in a trial.
“This is a way to get drugs to patients sooner by looking at pathologic complete response rate as a sort of surrogate that is not definitive but indicates a higher chance [of long-term benefit],” said Debasish Tripathy, MD, Chairman in Breast Medical Oncology at the University of Texas MD Anderson Cancer Center, Houston. He was not involved in this particular I-SPY2 study, but was an investigator on the I-SPY2 trial in the past.
He explained that when trial investigators are following patients for long-term outcomes, like disease-free survival, they have to follow them for 5 years or longer, whereas it only takes about 6 months to obtain pCR rates; pCR gives a much quicker answer.
He noted that pCR as a surrogate endpoint is a controversial topic in the field. “The pathologic complete response rate alone doesn't guarantee that you're going to have a longer survival, but the data suggest that it's likely.”
The I-SPY2 trial is an open-label, randomized, phase II clinical trial designed to evaluate whether adding investigational drugs to standard neoadjuvant chemotherapy is better than chemotherapy alone for patients with advanced breast cancer. Investigational drugs that pass a certain statistical threshold graduate to the next phase of study.
The I-SPY2 investigators enrolled women with stage II or III breast cancer with a high risk of recurrence and tumors that were greater than 2.5 cm. Breast cancer subgroups were triple-negative, hormone receptor-positive/HER2-negative, hormone receptor-negative/HER2-positive, and triple-positive.
After tissue biopsy and MRI, participants were adaptively randomized to one of several treatment arms. The control arm was paclitaxel for 12 weekly cycles followed by anthracycline. Those in the remaining arms were given the investigational agent for 12 weekly cycles followed by anthracycline. For this analysis, 11 investigational agents or combinations, including the control, had been given to participants.
The primary endpoint was pCR, which was defined as no residual invasive cancer in breast or lymph nodes. Participants were labeled non-pCR if they did not complete the assigned therapy. Secondary endpoints were residual cancer burden and EFS.
“Importantly, we need to note that hormone receptor-positive/HER2-negative patients with low-risk MammaPrint scores are not enrolled in I-SPY2,” said study presenter Douglas Yee, MD, Professor of Medicine and Pharmacology in the Division of Hematology, Oncology and Transplantation at the Masonic Cancer Center, University of Minnesota, Minneapolis. “These are genomically low-risk patients we felt should not be exposed to chemotherapy.”
The evaluable patient population was 746 patients; 259 patients (35%) achieved pCR and 487 patients (65%) did not. The median follow-up was 2.7 years; 126 EFS events and 109 DRFS events occurred. A total of 12 patients did not undergo surgery and were considered non-pCR.
When interpreting the data, Yee stressed, “We are not comparing one arm versus another. We are comparing all patients who achieved a path CR.”
For all participants who achieved pCR, 3-year EFS was 94 percent and 3-year DRFS was 95 percent, both with hazard ratios of 0.20. Yee described this result as “exceptional.”
The results for the breast cancer subgroups were similar. For the triple-negative breast cancer group, the EFS was 92 percent with a hazard ratio of 0.17 and DRFS was 94 percent with a hazard ratio of 0.16. For the hormone receptor-positive/HER2-negative breast cancer group, EFS was 94 percent with a hazard ratio of 0.21 and DRFS was 94 percent with a hazard ratio of 0.22. For the hormone receptor-negative/HER2-positive breast cancer group, EFS was 93 percent with a hazard ratio of 0.10 and DRFS was 93 percent with a hazard ratio of 0.14. For the triple-positive subgroup, the EFS was 96 percent with a hazard ratio of 0.26 and DRFS was 97 percent with a hazard ratio of 0.19.
“When you plot these event-free survival rates on a single graph that's stratified by path CR or non-path CR, you see that all the subgroups who achieved a path CR enjoy a very similar event-free survival rate,” said Yee. “While there is some splay in the non-pCR rates, it is very clear that patients who achieved a path CR have a better event-free survival than any of the subgroups of patients that did not.”
“We believe path CR is equally predictive across all tumor subsets, and beyond that, it is also across multiple agents to achieve path CR,” noted Yee. “Path CR as an endpoint enables rapid evaluation of novel therapeutic combinations. It can accelerate the identification of effective and potentially less toxic regimens, so we believe that achieving path CR through any therapy for any subtype is a sufficient endpoint.”
Tripathy agreed, saying the I-SPY2 trial findings provide “greater evidence” for using pCR as a surrogate endpoint. He noted that the findings do not actually change practice.
“It's not like we can take these data and start giving all these biological drugs to our patients without FDA approval. It further supports the whole notion that you can test a drug in the neoadjuvant setting and get a readout much more quickly than you would looking at disease-free survival,” said Tripathy.
He explained that a larger confirmatory trial will still be needed; the I-SPY2 trial only showed a tight correlation between pCR and outcomes.
“The idea that we can accelerate drug testing using neoadjuvant therapy now has more support, and there's more certainty around it, but not to the point that it's definitive, not to the point that we can fully approve a drug just on the basis of a neoadjuvant trial,” stated Tripathy about the study findings.
Call for Action
“We need to develop minimally invasive techniques, such as MRI and biopsy, to identify path CR prior to definitive surgery,” Yee explained. “We need to validate robust MRI and tissue predictors of path CR, and if we can do that, we can then begin to de-escalate toxic therapy. For example, we could begin to exclude AC [anthracycline] if we have confidence that path CR is obtained early.
“The contrary is also true,” he added. “We need to reassign patients to new therapies if path CR is not predicted.”
Christina Bennett is a contributing writer.