Teams led by researchers at George Mason University have published a pair of studies exploring protein signaling in breast and colon cancer that could offer new insights into treatments for the diseases.
Using reverse-phase protein microarrays, the scientists measured HER2 signaling pathways in samples from 415 breast cancer patients and, in a separate effort, profiled protein pathway activation in 34 matched primary colorectal cancers and synchronous hepatic metastasis.
In particular, the breast cancer study – published this week in the online edition of Clinical Cancer Research – could have important therapeutic implications, possibly expanding the population of patients eligible for treatment with anti-HER2 drugs like trastuzumab – marketed by Roche as Herceptin – said Emanuel Petricoin, director of GMU's Center for Applied Proteomics and Molecular Medicine and an author on both studies.
In their phosphoproteomic study of 415 breast cancer patients, the researchers identified a subset of roughly 40 patients who, while negative for HER2 according to conventional immunohistochemistry and fluorescence in situ hybridization tests, showed phosphorylated HER2 and activation of HER2 signaling pathways.
Trastuzumab is typically given only to breast cancer patients who test positive for heightened HER2 expression via either IHC or FISH. However, Petricoin told ProteoMonitor, the identification of a population that tests negative for increased HER2 via IHC and FISH but tests positive for HER2 phosphorylation and pathway activation suggests these conventional tests could be missing some patients that could benefit from trastuzumab treatment.
Previous studies have hinted at such a population said Julia Wulfkuhle, a GMU researcher and first author on the paper.
"There is data [from past studies] indicating that there are certain patients that are not HER2 positive that are responding to anti-HER2 therapy," she told ProteoMonitor. "There have been observations from several trials… that [such] a population of patients might exist."
However, scientists haven't known what underpinned this response on a molecular level, she said. "I would say this is the first characterization of a potential population that might be representative of those previous observations."
"It wasn't just that we saw this phosphoHER2 positive group [among the HER2 negative patients]," Petricoin said. "We [also] saw evidence that the phosphorylation event that was occurring was biologically productive – meaning that we saw evidence of receptor heterodimerization with other HER family members [and] increased phosphorylation of downstream biochemically linked substrates that were identical in level to those patients [who] really were HER2 positive."
The next step, he said, is demonstrating that the finding is, in fact, clinically relevant. To that end the researchers have a collaboration planned with Soonmyung Paik, director of the Pathology Division at the National Surgical Adjuvant Breast and Bowel Project, to measure HER2 phosphorylation and signaling in sample sets like those from the project's B-31 clinical trial that found evidence of trastuzumab response in HER2-negative patients.
They also plan to look at samples from the Biomarkers Consortium's I-SPY breast cancer study to see if that population has an incidence of HER2-negative, phospho-HER2 positive patients comparable to the Clinical Cancer Research study.
"If we see that the incidence rate is there – that we actually see this cohort [in the I-SPY samples], then the plan would potentially be to bring in phosphoHER2 measurement as a stratification [tool] in the I-SPY2 trial," Petricoin said. "So we could look at patients in a neoadjuvant setting who are phosphoHER positive but IHC or FISH negative to see if they receive clinical benefit [from trastuzumab]."
The researchers are also "discussing this new phenotype with HER2 thought leaders about potential prospective clinical trials where this could be baked in," he added.
The goal, Petricoin noted, isn't to replace conventional IHC and FISH testing. "If a patient is IHC or FISH positive they would be given anti-HER2 therapy just as the standard of care," he said. "What this type of assay could potentially be used for is identifying a brand new cohort for therapeutic stratification that otherwise would likely have been placed into a bucket for triple negative disease and get an entirely different [treatment] regimen."
The GMU study is just the most recent to show the potential of phosphoproteomic analysis – and RPPA technology specifically – for identifying new patient subtypes within cohorts once thought to be homogenous. Last month, a study by the National Cancer Institute's Cancer Genome Atlas consortium used RPPAs to identify two apparently distinct phosphoproteomic-based subtypes within the larger gene expression-based HER2 subtype – one exhibiting high HER2 and HER1 signaling activity and the other exhibiting lower levels of such activity (PM 9/28/2012).
In a separate study published last week in Clinical and Experimental Metastasis, Petricoin's team built phosphoproteomic profiles of 34 matched primary colorectal cancers and synchronous hepatic metastasis, demonstrating that the primary tumors and metastases differed significantly in their protein signaling architecture.
This extent of these differences is a key question for molecularly targeted treatment of metastatic cancer, Petricoin noted.
"What often happens now is [clinicians diagnose] a patient … with metastatic cancer; they resect the primary tumor; and then they look at that primary tumor with immunohistochemistry and molecular profiling," he said. "But [the molecular profile of the primary tumor] has nothing to do with the metastasis that is going to kill the patient. So [clinicians] give drugs based on [the primary tumor] and they don't work because they didn't look at the cancer that is going to kill the patient."
Using RPPAs to compare protein expression and phosphorylation in the primary and metastatic lesions across 86 key signaling proteins, the researchers found "systemic pathway changes [between the primary and metastatic lesions] that extended across all 34 patients," Petricoin said.
Although he cautioned that it was difficult to draw a conclusion based on just 34 patients, Petricoin noted he was surprised by the homogeneity of the metastases.
"I felt that there was going to be much more heterogeneity in the metastases from patient to patient," he said.
He also noted that the study's findings lend additional support to the multi-kinase inhibitor regorafenib – marketed by Bayer as Stivarga – which the US Food and Drug Administration approved last month for treatment of metastatic colorectal cancer.
Regorafenib, Petricoin said, "targets specific proteins that we found to be elevated in the metastases versus the primary tumor."
The results of the Clinical and Experimental Metastasis study would appear to contradict results published in Molecular Systems Biology last month by the lab of Max Planck Institute of Biochemistry researcher Matthias Mann that found few changes between primary tumors and metastases in colorectal cancer (PM 9/21/2012).
However, Petricoin noted, the MSB paper looked at nodal metastases, which occur in the lymph nodes, rather than metastatic tumors that have invaded an organ.
"In some ways the [MSB] paper was about a comparison of a completely different biological process," he said. "It may suggest that the molecular difference between the primary tumor and the lymph [tumor] didn't appear to be much… but our paper shows tremendous changes when you actually look at the metastatic lesion."
Getting hold of the samples needed to do such a comparison is no easy task, said GMU researcher Mariaelena Pierobon, a co-author on both the Clinical Cancer Research and Clinical and Experimental Metastasis studies. The matched primary and metastatic samples came from the University of Padova, where Pierobon was previously a surgical resident.
"It's a pretty unique [sample set]," she told ProteoMonitor. "They were collected in the same patient at the same time, both the primary tumor and metastasis. And that is something that is pretty hard to get, usually."
Petricoin said he hoped studies like the GMU effort would increase the availability of such samples by providing evidence of their utility.
"There is still – I would call it almost folklore – that you can look at the primary tumor and get what you need out of that from a molecular standpoint," he said. "But I think there is a growing sense that we're likely missing the boat, and we're really going to have to collect metastases to be able to target the therapies."
"Biopsying metastatic lesions is an invasive procedure, so you don't want to biopsy a patient for no reason," he added. "We hope that some of the publications are showing the need."
Pierobon is also involved in a collaboration with an Italian group using rat models to study metastases. "We are trying to understand how well the rat model resembles the pathway activation we see in humans," she said, noting that the aim is to use information from the models to identify treatments for preventing or treating metastases.