NEW YORK – Researchers performing a translational biomarker analysis have found that patients who still had residual invasive disease after neoadjuvant HER2 targeted treatment consistently benefitted more from adjuvant use of trastuzumab emtansine (Genentech's Kadcyla) over trastuzumab (Genentech's Herceptin).
At the American Society of Clinical Oncology's virtual annual meeting, Carsten Denkert of the Philipps University in Marburg Germany presented biomarker data from the Phase III KATHERINE study, which compared the adjuvant use of trastuzumab emtansine, or T-DM1, versus trastuzumab in patients with early-stage HER2-positive breast cancer who still had residual invasive disease after receiving neoadjuvant treatment with chemotherapy and HER2-targeted treatment before surgical resection.
The data from the KATHERINE study resulted in the approval of adjuvant T-DM1 in this setting. The clinical results, which were first presented at the San Antonio Breast Cancer Symposium in late 2018 and published in the New England Journal of Medicine in February 2019, demonstrated an 11 percent improvement in invasive disease-free survival benefit with T-DM1 versus trastuzumab. After three years, more than 88 percent of patients in the trial receiving T-DM1 didn't see their cancer return versus 77 percent of those treated with trastuzumab.
"The KATHERINE data were nothing short of impressive and certainly were practice-changing," Sara Hurvitz of the University of California Los Angeles said in a discussion following Denkert's presentation. But, she noted that there was a group of patients in the study who didn't appear to benefit from T-DM1. "An interesting question to look at is … what makes them different molecularly from the patients who really needed T-DM1?"
Researchers led by Denkert embarked on a translational program to try to answer this question through DNA and RNA sequencing of patients' tumor samples. "In the translational program, we focused on those pathways that are known to be relevant for resistance to HER2 treatment," he said in his presentation. "We know that higher HER2 expression is associated with improved outcome after HER2 targeted therapy, but those data are based on pre-therapeutic samples, and we do not know if the same is still true if we look at the dual resistant post neo-adjuvant tumor after neoadjuvant therapy."
Additionally, researchers noted the PIK3CA mutations are a known poor prognosis factor in metastatic breast cancer, but the effect of these mutations in the adjuvant setting is not clear.
The researchers first performed DNA sequencing on surgical samples taken from patients following neoadjuvant treatment with chemotherapy and trastuzumab — a regimen that was given to all patients within KATHERINE. The goal was to identify whether the presence of PIK3CA mutations might inform response to T-DM1 versus trastuzumab in the adjuvant setting post-surgery.
Ultimately, this was not the case; the invasive disease-free survival benefit with T-DM1 remained consistent for patients with and without PIK3CA mutations.
Next, the researchers performed RNA sequencing on the surgical samples to assess HER2 expression, as well as the expression PD-L1, CD8, the T-cell effector signature. Among all of these subgroups, there was likewise no difference in regard to the benefit of adjuvant T-DM1. The invasive disease-free response benefit, in other words, was observed with T-DM1 regardless of the signature status.
Researchers also compared gene expression signatures using samples taken before patients received neoadjuvant treatment and surgical samples after neoadjuvant treatment. Most of the samples investigators had access to for this analysis, approximately 80 percent, came from the surgical specimens after neoadjuvant treatment, and around 20 percent of samples were from patients taken before they received neoadjuvant trastuzumab and chemo.
Using these samples, researchers looked at the association of PD-L1 expression and treatment response. For patients who had a low PD-L1 expression in their surgical samples, outcomes were worse with adjuvant trastuzumab relative to the patients who had higher PD-L1 in their surgical samples. This difference was not observed in the group treated with T-DM1.
"This may indicate that PD-L1 may be involved in some resistance mechanisms," Denkert said, pointing out that for the other gene signatures there were no differences between the two arms.
Researchers also reported HER2 gene expression patterns that may indicate signals of resistance. In most patients, HER2 expression was decreased in the post-treatment surgical samples, likely as a result of the neoadjuvant HER2-targeted therapy. But there was a subgroup of patients whose surgical samples still expressed high levels of HER2 following neoadjuvant treatment, suggesting resistance.
When researchers looked at patients' outcomes in the adjuvant treatment arms, those with high levels of HER2 expression in their surgical samples experienced worse invasive disease-free survival outcomes on trastuzumab than patients who had decreased HER2 expression in surgical samples. "What's interesting to note is how poorly the patients did if they were HER2-high expressing and were given trastuzumab," Hurvitz said in her discussion, pointing out that there were 70 invasive disease-free survival events in that group, which consisted of 200 patients.
This biomarker association was not seen for the patients who went on to receive T-DM1 as adjuvant treatment. These patients, even with the high HER2 expression, experienced the same improved benefit with T-DM1. In contrast to the trastuzumab events, only 26 patients with high HER2 expression experienced events out of 207 total.
"What we see here is a kind of resistant population in the residual disease tumors that have a high HER2 expression in this setting [and] are resistant to trastuzumab, but not to T-DM1," said Denkert, going on to suggest, "it may be an option to overcome this resistance by T-DM1 treatment."
Hurvitz, in her discussion of the KATHERINE biomarker data, pointed out that these findings are consistent with several earlier trials that assessed levels of HER2 expression — be it by mRNA, IHC, or HER2 heterogeneity by FISH — in both the neoadjuvant and metastatic settings. "These studies generally indicate that T-DM1 works best in high HER2-expressing tumors," she said.
Ultimately, the KATHERINE translational program did not reveal any biomarkers that predicted a lesser benefit from T-DM1. But as a future direction for research, Hurvitz said in her discussion, it might be illuminating to look at how patients with lower levels of HER2 expression at baseline fare with T-DM1. In the original 2018 presentation of the KATHERINE results, she pointed out, patients with an IHC staining of less than 3 for HER2 expression appeared to derive less benefit from T-DM1 than those with IHC scores above 3.