NEW YORK (GenomeWeb) – New data from a clinical study funded by the breast cancer charity Side-Out Foundation indicates the ability of multi-omic analyses to improve the effectiveness of therapy in patients with metastatic disease.
According to Emanuel Petricoin, co-director of the Center for Applied Proteomics and Molecular Medicine at George Mason University and one of the leaders of the study, initial results from the second Side-Out study suggest that using molecular data to guide patient treatment extends progression-free survival in metastatic breast cancer patients.
This is in line with results from the first Side-Out study, which found that in 13 of 25 subjects therapies selected using genomic and proteomic analyses of patient tumors extended progression-free survival by more than 30 percent compared to the patient's last treatment regimen.
That finding was notable, Petricoin said, in that typical progression-free survival decreases as patients move through treatment options.
"What we reported in the first trial is actually more than half the patients had… an increase in progression-free survival," he said, adding that while he and his colleagues are still completing analysis of data from the second Side-Out study, the results appear to be "in keeping with" those from the first study.
Assuming the new results do, in fact, show similarly significant improvements in progression-free survival, they will bolster the case not just for such multi-omic analyses generally, but for phosphoproteomic data specifically, as this was the primary type of data used for guiding therapy in the second study.
Both studies used reverse-phase protein arrays to measure phosphorylation levels of proteins in cancer signaling pathways, allowing them to assess whether or not those pathways were activated with the aim of determining whether or not they might be targets for therapy. However, while the first study also used extensive genomic data including clinical exome sequencing and RNA-level data generated by Caris Life Sciences, the second study relied mainly on the RPPA data.
The idea underlying this decision, Petricoin said, was that phosphorylation provides a more direct read-out of pathway activation, and so should be particularly informative in terms of identifying candidate proteins and networks for targeting with drugs.
"It was felt that the RPPA was well suited for this because of its content, so we wanted to test that concept," he said.
In the first study, however, the tumor boards tended to use the RPPA data more for eliminating particular treatment options than selecting new ones, Petricoin said.
"You could have a patient where, say, the next-gen sequencing data said, 'Aha, you should give a HER2 inhibitor,' because they are HER2 amplified, but on the RPPA we saw no HER2 activiation," he said. "And so then the doctors would have said, 'No, we aren't going to give a HER2 inhibitor because there is no activation of HER2,' and they might have selected another NGS-based target instead."
This is in line with trends in proteogenomic research more broadly, where an emerging use of the approach is looking at protein-level data to determine which genomic alterations are biologically or clinically relevant.
"I think it is that negative [selection] that physicians seem to be most interested in right now rather than, "Hey, NGS says this is not mutated, but RPPA says it is activated," Petricoin said.
He added, though, that in other work, including through the companies Perthera and Theranostics Health (now part of Avant Diagnostics), which he co-founded, he has found physicians more willing to select, as opposed to rule out, therapies based on RPPA phosphoprotein measurements.
He cited as an example of RPPA's potential in this respect research he and others have done through the I-SPY 2 trial indicating that even in triple-negative breast cancer patients, phosphorylated HER2 and EGFR helped predict which patients would benefit from the HER2 inhibitor neratinib.
Petricoin said he hopes to further establish the effectiveness of RPPA data for guiding therapy in the third Side-Out study, which is currently enrolling. That study, he said, will seek to determine if RPPA data can identify non-responders among metastatic breast cancer patients to the CDK4/6 kinase inhibitors palbociclib and ribociclib (marketed, respectively, by Pfizer as Ibrance and by Novartis as Kisqali), which are approved for treatment of ER-positive and HER2-negative breast cancer.
Some 20 to 40 percent of patients don't receive benefit from these drugs, Petricoin noted, but there are no well validated biomarkers that can identify those patients.
"We believe that phosphoproteins hold the key," he said. "And we are going to try to prove it by testing the hypothesis that looking at CDK4/6 substrates by RPPA will be predictive [of patient response]."
Beyond reducing overtreatment with CDK4/6 inhibitors, good response markers could also help doctors move more quickly to alternative treatments like the MTOR inhibitor everolimus (marketed by Novartis as Afinitor), which Petricoin said is indicated for the same patient population as the CDK4/6 inhibitors.
Affinitor is more toxic than the CDK4/6 inhibitors, he said, "but if we can identify patients that received no benefit from [a CDK4/6 inhibitor], their doctor can start them on Afinitor."
In the trial, which is currently recruiting at four sites around the country and is slated to last two years, Petricoin and colleagues will use RPPA to assess the phosphorylation status of a set of CDK4/6 substrates they have hypothesized could help distinguish between responders and non-responders. Patients will undergo another biopsy if and when they are determined to have progressed on the CDK4/6 inhibitors and that biopsy will undergo testing by the full battery of genomic, proteomic, and phosphoproteomics approaches used in the first two Side-Out studies.