Hope S. Rugo, MD, University of California San Francisco Comprehensive Cancer Center, discusses the clinical significance of the phase 3 SOPHIA trial, the findings of which were presented at the 2019 ASCO Annual Meeting.

Transcript

I'm Hope Rugo, and I'm a Breast Medical Oncologist and Professor of Medicine and Director of Breast Oncology and Clinical Trials Education at the University of California at San Francisco's Comprehensive Cancer Center, here in San Francisco, California.

I am talking to you today about results, the first result, of the phase 3 SOPHIA trial, a trial that compared a novel antibody to HER2 called margetuximab to trastuzumab, combined with a menu of chemotherapy agents in patients with metastatic HER2-positive breast cancer that have progressed on prior therapy.

This is the first progression-free survival data from this phase 3 trial. We chose the population that we studied because treatment after established lines of therapy for HER2-positive breast cancer is still an unmet need.

The first-line therapy for HER2-positive breast cancer is trastuzumab–pertuzumab in combination with a taxane. Second line is PDM-1. Both of these treatments were shown to improve survival, compared to the standard of care.

Subsequent lines of therapy usually include continued to HER2-targeted therapy, depending on where in the world you are with chemotherapy. That could include an antibody like trastuzumab or an oral tyrosine kinase inhibitor, like lapatinib, or now, potentially, in the not-too-distant future, sorafenib, or other oral tyrosine kinase inhibitors. We really don't know what treatment to use next, and we don't have survival benefit data from any therapeutic intervention after second-line.

Margetuximab is a novel Fc-engineered HER2-targeted antibody that was designed to activate immune responses.

What we know is that there are polymorphisms of Fc gamma receptors that influence immune responses. It's quite interesting, just like there are polymorphisms for many of the enzymes that metabolize drugs.

Margetuximab is an antibody, and the Fc portion of that antibody was engineered to enhance the affinity for both low-affinity and high-affinity activating Fc gamma receptor. The Fc gamma receptor is called CD16a. The low-affinity allele is called F, and the high-affinity allele is called V.

Margetuximab engineering also was designed to reduce binding or activation of an inhibitory Fc gamma receptor called CD32b. We actually were interested in what information was available about polymorphisms and responsiveness to trastuzumab.

There actually were 2 retrospective studies that showed that patients who carry the low affinity CD16a F allele had a lower response to trastuzumab, both in the metastatic and adjuvant setting. However, controversy exists, because 2 other studies that looked at patients in the adjuvant setting showed no impact of the CD16a low-affinity F allele.

Our hypothesis was that in pretreated patients with HER2-positive metastatic disease that we would see greater benefit from margetuximab in patients who carry the F allele compared to trastuzumab, because trastuzumab has no particular advantage in that group, because of this engineering, which generates higher affinity.

SOPHIA actually is the first prospective trial to assess the impact of Fc gamma receptor alleles on the activity of trastuzumab and other HER2-targeted agents.

In vitro studies have shown that margetuximab enhances innate immunity, looking at NK cells with different Fc alleles, both the F allele and the V allele, and clearly, the impact of the margetuximab on these NK cells in terms of cell cytotoxic activity was greater in these NK cells that had the F allele, compared to a surrogate for trastuzumab. There was still a benefit in the V-V homozygous—the homozygous NK cells—but it was less of a difference between margetuximab and trastuzumab.

There actually was a previous phase 1 trial that tested margetuximab in 24 patients with pre-treated HER2-positive metastatic breast cancer, and showed a response rate of 17%. Really, I think quite remarkably, 3 patients are still on treatment, who are now 4-to-6 years out, as of May 2019.

Peripheral blood mononuclear cells were looked at before exposure and after exposure to margetuximab, and demonstrated that 94% of patients had activated and had HER2-specific T cell and antibody responses in response to a monotherapy with margetuximab. This is very exciting and led to the phase 3 trial that I explained earlier.

We randomized 536 patients with pretreated HER2-positive breast cancer to receive either margetuximab or trastuzumab in combination with a menu of chemotherapy agents, and we had eligibility, which included at least 2 prior anti-HER2 therapies that included pertuzumab and 1-to-3 lines of therapy in the metastatic setting.

We actually had sequential primary end points, including progression-free survival and overall survival. Then one of the secondary end points included an exploratory, preplanned analysis of the efficacy of margetuximab, based on the Fc gamma receptor alleles that patients had.

The demographics and disease characteristics were overall well-balanced in this trial. In terms of prior cancer therapy, by study design, 100% of patients had received prior trastuzumab and pertuzumab, and nicely for the study results, over 90% of patients had received TDM1, so very encouraging.

More patients receiving margetuximab than trastuzumab had received prior adjuvant or neoadjuvant therapy. In the treatment itself, more than 1/3 of patients had received more than 2 lines of therapy in the metastatic setting.

Progression-free survival was analyzed by central-blinded analysis, and was improved in patients receiving margetuximab compared to trastuzumab with a hazard ratio of .76 and a P value of .033. The absolutely difference was relatively small at about 1 month in progression-free survival, going from 4.9 to 5.8 months, with continued separation of the curves over time. We also looked at investigator-assessed progression-free survival, and that was also longer in patients receiving margetuximab.

A subgroup analysis was performed and showed that, although progression-free survival was improved in the majority of patient subgroups receiving margetuximab, it was a relatively enhanced impact in patients who had IOT3-positive disease, more than metastatic sites, hormone receptor-negative disease, and those who had received prior adjuvant therapy for breast cancer.

We looked at, as I mentioned earlier, planned exploratory progression-free survival analysis by Fc gamma receptor alleles.

What we found was that progression-free survival appeared to be enhanced in patients receiving margetuximab, based on their CD16a allele, so that those patients who carried the F allele appeared to have a relatively greater benefit from margetuximab versus trastuzumab, compared to the degree of benefit seen in the intent to treat population.

What we found is that when we looked at patients who carried at least 1 F allele—so F-F or F-V—that the progression-free survival was longer with patients receiving margetuximab versus trastuzumab, with a hazard ratio of .68 and a P value of .005.

The absolutely benefit actually was 1.8 months. If we looked at patients who had a V-V genotype—that's a higher affinity Fc gamma CD16a allele—those patients appeared to have a relatively similar benefit from margetuximab and trastuzumab.

It's important to note that 86% of patients carry at least 1 F allele, and only 14% of patients were homozygous for the V allele. The ability to look at the V allele is relatively limited.

We also wanted to know whether it made a difference whether you were homozygous or heterozygous for the F allele, so we looked at patients who had F-F versus F-V (relatively equal distribution between that group, a little more patients with F-V).

What we found was an absolute difference between the margetuximab-treated patients and trastuzumab patients that appeared even greater with margetuximab in this population at 2.6 months absolute difference in PFS.

The hazard ratios between F-F and F-V patients were relatively similar, at .68 in patients who had the homozygosity for the F allele, and .71 for patients who had a heterozygous F allele (so F-V).

That’s very encouraging, and suggests that there might be a real difference in outcome based on the CD16a Fc gamma receptor allele that you inherit and have, based on the affinity to the antibody that you're receiving in combination with chemotherapy.

We had an interim overall survival analysis planned in the study. The data cutoff for that interim overall survival was October of 2018, and there are 2 more analyses planned—a second interim analysis and a final analysis.

At the first interim overall survival analysis, we only had 41% of the events needed for the final analysis, so less than 50% of the events—this is very preliminary. For the intent-to-treat population, we saw a numeric improvement in overall survival that was not statistically significant; that difference was 1.7 months.

However, we also had a pre-planned analysis in the patients based on Fc gamma receptor genotype, What we found was that patients who carried at least 1 F allele—so the F-F and F-V patients, remember, that was 86% of the total trial population—had an absolute difference in overall survival favoring margetuximab of 6.7 months.

Very exciting and early look at this data, but still interesting, because we have already seen this difference in patients in terms of progression-free survival. In fact, we have done an interaction test on the progression-free survival, based on the F allele presence. That was significant also, with a P value of .012.

Taken together, along with the data that suggest that overall survival may have a greater impact on progression-free survival, when we use antibodies that impact the immune system, this is exciting.

We have a second planned interim analysis that should be very informative, planned after 270 deaths. That's a much greater number of deaths—way more than 50%—and that should happen later in 2019.

Of course, we looked at other secondary end points, including overall response and safety. Overall response was numerically longer with patients receiving margetuximab than trastuzumab, with a P value of .06. The clinical benefit rate was significantly longer, significantly higher, 38.6% versus 24.8%, with a P value of .003. That was encouraging as well. In terms of adverse events, margetuximab was well-tolerated.

There was no difference in serious adverse events, grade ≥3 adverse events, or treatment discontinuation versus adverse events, or no deaths attributed to antibody treatment. When we looked overall at all adverse events, regardless of causality, we saw one difference that stood out—that was that patients receiving margetuximab had a higher incidence of infusion-related reaction. Most of them occurred in the first cycle and were low grade, but 3 patients discontinued treatment due to infusion-related reactions.

We found that these reactions could be well-controlled with subsequent premedication before the infusion, so that was encouraging.

We've actually demonstrated that there's a novel Fc-engineered HER2-targeted antibody that stimulates mechanisms of both innate and adaptive immunity called margetuximab, that in a randomized phase three trial, we saw a prolongation of progression-free survival that was statistically significant.

In this first prospective analysis of the impact of Fc gamma receptor genotype on efficacy with HER2-targeted therapy, we saw the first evidence that there may be a real difference in pre-treated patients with HER2-positive metastatic breast cancer that favors an engineered, higher affinity binding antibody in patients who carry the lower affinity F allele for CD16a.

Safety was acceptable, as I mentioned, with an increase in infusion-related reactions that could be managed fairly easily.

Our next milestone is the second interim overall survival analysis that will occur later in 2019. We hope to have additional data soon. Thank you.