Skip to main content
Premium Trial:

Request an Annual Quote

Data from Side-Out Breast Cancer Study Suggests Potential of Proteomics for Personalizing Therapy


Researchers affiliated with breast cancer charity the Side-Out Foundation presented pilot study data this week at the American Society of Clinical Oncology's annual meeting suggesting proteomics' potential as a tool in personalizing therapy for breast cancer.

In the study, the researchers developed via genomic and proteomic analyses molecular profiles of tumors from 25 metastatic breast cancer sufferers, which they then used to guide treatment. In 13 of the 25 subjects, these molecularly guided therapies extended progression-free survival by more than 30 percent compared to the patient's last treatment regimen, a result, noted Nicholas Robert, study co-author and an oncologist at Virginia Cancer Specialists, that well surpassed the researchers' criteria for success.

"These are patients with a lot of prior treatment, so to see 30 percent benefit [in 30 percent of the patients] was thought to be a reasonable goal," he told ProteoMonitor.

In fact, more than 50 percent of the patients saw a 30 percent or greater benefit in progression free survival.

Bolstered by these results, the researchers are now planning a larger follow-up study scheduled to begin this month. That study will include roughly 50 breast cancer patients from nine sites around the country.

The Side-Out trial was inspired by a similar study led by Translational Genomics Research Institute researcher Daniel Van Hoff published in the Journal of Clinical Oncology in 2010, in which he and his colleagues profiled patient tumors using immunohistochemistry, fluorescent in situ hybridization, and gene expression assays to help determine personalized treatments for these subjects.

Van Hoff "presented that work at [a] meeting several years ago and showed that there was a benefit to that approach in the patients he treated," Robert said. "So we proposed to the Side-Out Foundation that we could do [the same thing] but exclusively in breast cancer, and including not only IHC, FISH, and DNA microarrays, but also proteomics."

Spanning two-and-a-half years, the study looked at metastatic breast cancer patients who had progressed off at least three previous treatment regimens. For the genomic analysis, the researchers used Caris Life Sciences' Target Now service. For the proteomics analysis, Robert and the foundation looked to George Mason University researchers Emanuel Petricoin and Lance Liotta, who have for a number of years been using their reverse phase protein array technology to profile protein signaling networks in tumors also with the aim of identifying the best therapies for a given cancer patient.

"The George Mason group was in our backyard," Robert said. "And there is an argument that phosphorylated proteins are actually the better target to identify a valid pathway" for treatment.

He noted that in the pilot study there "was a bit of a bias to use more conventional [IHC-based and genetic] targets," but, he added, the phosphoproteomic data "helped validate some of the other data that we obtained" and, in some instances, "provided an approach that we would not have otherwise considered."

In the upcoming second phase, the researchers plan to lean more heavily on the GMU team's proteomics data, Robert said. "Our plan is to deliberately look at the proteomics piece and give it a bit more weight."

This decision, he said, is partly due to "the data coming out of this study, and partly that we should test the utility of the phosphorylation data. It gives the proteomics data a chance to see if it works or not."

A long-time advocate of using proteomics data – and RPPA-based phosphoproteomic data in particular – for guiding personalized therapy, Petricoin called the study "a real example of functional protein assay tests really being involved in a trial where clinical benefit was shown."

"It provides an exciting motivation, and now we want to expand this to larger trials to really show in a larger setting that we are seeing clinical benefit," he told ProteoMonitor.

Petricoin noted that in the upcoming second study, the researchers will be adding TGEN as a partner and provider of genomic profiling including sequencing, exome analysis, and RNA-seq.

"This next phase of the trial will be far and away the most comprehensive molecular profiling of tumors done by anyone in the world," he said, adding that the larger cohort size might also allow the researchers to identify patterns of patient response that weren't visible in the smaller pilot study.

"We want... to be able to drill into questions like: Are we seeing evidence that patients with increased mTOR activation are responding to an mTOR inhibitor? Evidence that there are subgroups of patients that are receiving clinical benefit from an EGFR inhibitor? We just don't have enough of an end number in this first trial to see that pattern emerge yet."

Beyond demonstrating that molecular profiling of tumors could have clinical benefit, the pilot study was also important in that it showed it was possible to fit it into clinical workflows and on a timescale fast enough for physicians to use it in their decision making, Petricoin noted.

"An extremely important aspect was, could we do this logistically?" he said. "Could we actually get metastatic cancers biopsied at different sites around the [US], have that sent to two different laboratories, and then take those samples and perform the molecular profiling and get the data back in less than a couple of weeks?"

Such logistical challenges have proven problematic for similar studies involving genomic techniques like next-gen sequencing and exome analysis, Petricoin said. "It has been a huge issue, because patients with progressing metastatic disease cannot wait weeks and weeks to get the data. Physicians simply will not wait that long. It's unethical."

The data typically took around two weeks to generate and return, Robert said, upon which the physicians reviewed it and determined a course of treatment. In all 25 of the cases, the therapy chosen based on the molecular data was different from that which the physicians would have chosen before seeing that information.

"Timing is huge," Robert said, regarding the effort to turn around the data quickly enough for it to be useful. Another consideration, he noted, was making sure the biopsies were stored in a manner that prevented degradation that could affect the analysis – particularly on the phosphoproteomic side of things.

"We had the research team there right when the biopsy was obtained so they were able to be put in the proper solution right away," he said. "So I think that we have less artifacts because of that."

"That's a little different from when you are doing this as a procedure where you are taking out a tumor from the breast or the lung," Robert said. "I can say this as a former pathologist – tissue doesn't get prepared properly right away" in that setting.

However, he added, were such personalized medicine approaches to become more widely used, surgeons and clinicians could adjust to these assays' sample preservation requirements.

Whether such methods do, in fact, find wider clinical utility remains to be seen, Robert said. One company trying to push the sort phosphoproteomic profiling used in the study into the clinic is Theranostic Health, a start-up launched by Petricoin and Liotta to commercialize their RPPA technology. At ASCO this week, the firm introduced its first commercial assay – the TheraLink HER Family Assay, a test intended as a supplement to conventional HER2 testing for guiding therapy in breast cancer patients (PM 5/3/2013).

"I think we need to have some controlled enthusiasm for this approach," Robert said. "But frankly, [physicians need to see] a randomized trial where someone takes their assays – whatever they are, genomics or proteomics – and compares what [therapy] they choose from those assays to standard therapy. That trial needs to be done to convince us that all this expense and time is of value."