Caris Life Sciences presented two posters at this week's American Society of Clinical Oncology Breast Cancer Symposium, revealing some of the ways the company has begun to search for new insights in data collected from 25,000 patients for its Target Now molecular profiling service.
One poster presented data on the prevalence of SPARC (secreted protein acid rich in cysteine) — a protein biomarker known to predict response to the drug nab-paclitaxel in certain cancer types, including breast — in a large cohort of breast cancer samples. The researchers showed that SPARC was about equally prevalent in the different breast cancer subtypes and could provide a potentially exciting target in the particularly hard-to-treat triple negative subtype.
The second presentation demonstrated the feasibility of performing biomarker profiling on body cavity fluid samples from patients with breast cancer metastasis limited to the body cavities. According to Raheela Ashfaq, medical director of oncologic pathology at Caris and an author of both posters, the company hopes to eventually incorporate body fluid profiling into the Target Now service package.
Ashfaq told PGx Reporter this week that these two presentations are representative of a larger thrust at Caris to mine its repository of samples for information that may enhance the "knowledge out in the community" on the implications of various biomarkers for cancer treatment.
"We feel that we have a lot to contribute," she said.
For the two studies presented at the ASCO meeting, Ashfaq said what the group did was "basically just retrospective analysis on the cohort of cases that we routinely get to profile."
Both posters represent research done on a database of information gleaned from samples sent in for the company's Target Now service, a "platform agnostic" system that uses several technologies to profile a patient's tumor for a variety of biomarkers. The service provides oncologists with a report of the biomarker findings and their associations with response to particular chemotherapeutic agents.
"When we issue the results — be it immunohistochemistry, or microarrays, or gene sequencing analysis, or FISH — all the results are in a back-end database, and … so far we have actually profiled 25,000 patients," Ashfaq said.
In one of the two studies presented at the meeting, the researchers evaluated Target Now's various profiling methods on body cavity fluid. The researchers took "all the fluid samples [they] had received from breast cancer patients," Ashfaq said, and looked at the feasibility of doing all the technology platforms used in the Target Now service, but on a fluid sample instead of a paraffin-embedded tissue sample.
This all started, she said, because oncologists had begun asking if Caris could evaluate fluid. "They were contacting us and saying, 'Look guys, we do not have tissue samples. This patient has only tumors in the body cavity,'" Ashfaq said. "And if there is a patient with advanced breast cancer and there is no mass in an organ system where you can approach it with a biopsy … then pretty much, you do not know what the molecular characteristics of that tumor are." In this scenario, oncologists "are trying to guesstimate what treatment regimen would work in that particular patient."
According to Ashfaq, fresh fluid that came in to the company was immediately spun down and the cells embedded in paraffin, so that the usual roster of platforms could be used. "We have essentially taken the upfront sample with the cells in a dispersed state, and we've blocked it to resemble a tissue block," she said.
"And from then we go and do all of the studies."
Ashfaq said the group was successful in profiling the fluid, though she said there was some variation because of the variability in the number of cells in each sample. "Some may contain millions, some may contain very few," she said.
Of 28 breast cancer fluid samples, 10 IHC biomarkers could be profiled in 20 samples, while seven samples were deemed "insufficient," and one yielded between one and nine markers, the researchers reported in their poster. DNA microarray analysis was done in 10 and combined results of predictive markers provided information on therapeutic guidance in 21 of 28 cases, they wrote.
"We didn't perform 18 biomarkers on every sample. But we were able to do a good number by immunohistochemistry, and in some we were able to do microarray, and we were able to routinely do FISH, though with sequencing we didn't have any luck," Ashfaq said. The results were successful enough that the researchers concluded molecular profiling of body cavity fluid can "provide insight into the molecular characteristics of malignant cells," and inform targeted therapy decisions. According to Ashfaq, the plan is to eventually make body cavity fluid a sample type option for Target Now.
In the company's other poster, the study of SPARC expression, the researchers evaluated 885 patient samples from the database, a number Ashfaq called "large" in the context of the existing data on this protein. Samples were divided into the five established subtypes based on ER and HER2 status.
According to the group's poster, the study was designed to evaluate the frequency distribution of
SPARC with a special emphasis on triple-negative patients. Though the results showed about the same frequency of SPARC in all five subtypes — between 42 and 47 percent — Ashfaq said that for triple-negative cancers the implication for treatment is greater because there are so few effective treatment options.
"We know that there are very few options when you are dealing with triple negative. So, if we know upfront that this particular triple negative tumor is expressing SPARC we now know we have a target."
This is not to diminish SPARC's relevance in other cancer types, though. For example, while Genentech's Herceptin is indicated for HER2-overexpressing patients, Ashfaq noted that only a subset of HER2-positive patients respond to the drug. "With this study we are outlining additional targets that could be used in treatment decisions," she said.
Though the study presents a valuable picture of SPARC distribution, Ashfaq noted that one weakness of the project is that the company does not have outcome data from patients given nab-paclitaxel in the context of their SPARC status.
The company is hoping to gather this data, however. Caris last year launched a cancer registry hoping to inform payors and healthcare providers about how oncology treatments administered based on patients' molecular profiles impact their outcomes (PGx Reporter 2/17/2010).
Ashfaq said that oncologists will provide the company with outcomes after recommending drugs based on Target Now molecular profiling. She said Caris plans to start gathering some data and presenting studies based on these clinical outcomes, but did not give a timeline for when the firm plans to release this data.
She also said the company will continue to present data extracted from its sample database, which she expects to be useful for the field especially as long as the large majority of clinical trials continue to follow a "one-biomarker/one-drug" model.
Ashfaq acknowledged that this model is changing, citing as an example the Biomarker-integrated Approaches of Targeted Therapy for Lung Cancer Elimination, or BATTLE, trial for lung cancer, which is classifying patients into four biomarker profile groups to determine which of four treatment strategies they should follow (PGx Reporter 4/16/2011). However, she noted that in the meantime, "we have such rich data — not just in breast cancer but in all different cancers we receive for profiling — that we feel … we have a lot to contribute to the biomarker literature."
Ashfaq plans to present more comprehensive data at the American Society of Cytopathology annual meeting in November that expands the company's research on profiling body cavity fluid in a greater variety of metastatic cancers
Have topics you'd like to see covered in Pharmacogenomics Reporter? Contact the editor at mashford [at] genomeweb [.] com.