By: Charles Bankhead, Senior Editor
80% lower risk of recurrence with surrogate endpoint, but link to survival unproven
Neoadjuvant therapy that resulted in pathologic complete response (pCR) reduced the risk of breast cancer recurrence by 80%, pooled data from a phase II clinical trial platform showed.
The results held up across all breast cancer subtypes and treatment regimens, which included nine novel combinations. About a third of patients achieved pCR with neoadjuvant therapy, and those patients had 3-year event-free survival (EFS) and distant recurrence-free survival (DRFS) rates of about 95%.
“The strength of the I-SPY2 study is that it shows that pCR, regardless of high-risk subtype and type of treatment and across nine investigational targeted biologics, is associated with a much better outcome for individuals who achieve pCR,” Laura J. Esserman, MD, of the University of California San Francisco, and co-authors wrote of their findings. “Larger confirmatory neoadjuvant trials will be able to generate sufficient data to establish whether pCR can be considered a validated surrogate of EFS.
“Perhaps more importantly, these data should drive and inspire us to think about how to maximize the chance that each individual can achieve pCR. These data provide a clear experimental rationale for serial adjustments to systemic therapy prior to surgery with therapies directed at specific subtypes,” the researchers said.
The analysis demonstrated a “strong individual-level association between pCR and the survival endpoints,” according to the authors of an accompanying editorial, Yu Shyr, PhD, of Vanderbilt University School of Medicine in Nashville (and associate editor of statistics for JAMA Oncology), and Derek Shyr, MS, of the Harvard T.H. Chan School of Public Health in Boston. However, they added, the data remained insufficient to serve as validation of pCR as a surrogate outcome for EFS or DRFS.
To prove that a surrogate endpoint (pCR, in this case) is equivalent to a true endpoint, studies of the endpoint must demonstrate two types of associations: the I-association (between the surrogate and true endpoint) and the T-association (effect of treatment on the surrogate and on the true endpoint). The T-association is particularly challenging, requiring trial-level data with a sufficient number of patients and number of events per treatment type, the editorialists explained.
“Unfortunately, the I-SPY2 trial has a small number of patients and events per treatment type, making it difficult to evaluate T-association,” Shyr and Shyr continued. “In addition, the small number of events would have translated to large confidence limits around the summary measures of trial-level association owing to the large sampling error for each data point in a trial-level analysis.”
Offering a potential example of a surrogate endpoint gone awry, the editorialists noted that the Prostate Cancer Prevention Trial with finasteride used the surrogate endpoint of biopsy-proven prostate cancer at 7 years, instead of a “true” endpoint, such as elimination of symptomatic disease or reduced mortality. Use of the surrogate reduced the required sample size by 50,000 and showed a significant decrease in the incidence of biopsy-proven prostate cancer.
However, further research showed that finasteride reduced the incidence of biopsy-proven prostate cancer because of its effect on prostate-specific antigen levels, which altered the biopsy sampling pattern and affected the false-positive rate.
Barbara Jacoby is an award winning blogger that has contributed her writings to multiple online publications that have touched readers worldwide.