BIOMARKERS FOR PANCREATIC CANCER

SAN FRANCISCO — Pancreatic cancer continues to have one of the highest mortality rates of any of the major cancers, mostly because of late diagnosis, so there is a great interest in developing improved diagnostic tools and biomarkers.

Two new studies are addressing this. In one, researchers found that circulating tumor cells (CTCs) could be used to distinguish between benign and malignant disease and could identify metastasis. In the second, patients with pancreatic cancer who had low serum levels of matrix metalloproteinase (MMP)-7 were most likely to benefit from adjuvant treatment with gemcitabine.

Although the data are very early, results from both studies were presented here at the 2014 Gastrointestinal Cancers Symposium.

CTCs Identify Cancer

The feasibility of using CTCs as an adjunct tool to improve the diagnosis of pancreatic ductal adenocarcinoma was evaluated by researchers from the University of California, Los Angeles (UCLA).

CTCs spread from the primary tumor, and can be a source of metastasis because they have the ability to colonize distant organs, explained lead author Jacob S. Ankeny MD, from the Department of Surgery at UCLA. There is also evidence that CTCs can be used when determining prognosis and/or diagnosis in several epithelial cancers.

"Additionally, circulating tumor cell numbers may correlate with responses to chemotherapy," he said. "However, most of these benefits have not been realized in pancreatic cancer, and they have not been established as a biomarker in this disease."

Currently, the diagnosis of pancreatic cancer relies on image-guided tissue sampling, which can be "expensive, inconvenient, and not without patient risk," the researchers note. The use of CTCs might improve the ability to diagnose the disease, obtain pure molecular information uncontaminated by stromal cells, and improve the ability to accurately stage patients at the time of diagnosis, they note.

Dr. Ankeny and colleagues conducted a prospective analysis (prior to treatment) of 50 consecutive patients in whom there was either suspicion or a recent diagnosis of pancreatic cancer.

A novel microfluidic NanoVelcro technology enhanced by anti-epithelial cell adhesion molecule enrichment was used to capture CTCs from serum samples. The researchers classified CTCs by size (greater than 10 µm) and immunohistochemistry staining pattern (DAPI+, CK+, and CD45–). KRASmutational status was assessed in CTCs from 3 patients to confirm that the CTCs originated from the tumor.

In the cohort, 32 patients had cancer and 18 had nonmalignant pathology on tissue biopsy. CTCs were detected in 62.5% of the patients with pancreatic cancer and in 5.5% of the patients with nonmalignant pathology.

The specificity of CTCs for the diagnosis of pancreatic ductal adenocarcinoma was 94.4%, positive predictive value was 92.5%, and negative predictive value was 58.6%.

In patients who were diagnosed with pancreatic cancer, CTC number correlated with disease stage, and the presence of 2 or more CTCs distinguished local from systemic disease.

Dr. Ankeny cautioned that this study had a number of limitations, such as a small size and lack of outcomes data, so it is "very preliminary." However, with these limitations in mind, "CTCs do have a potential role in aiding the diagnosis of pancreatic cancer," he concluded.

Intriguing But Role Unclear

Study discussant Rebecca A. Miksad, MD, MPH, from the Beth Israel Deaconess Medical Center and Harvard Medical School in Boston, agrees that CTCs could have a role to play, although that role is currently unclear.

Although CTCs have been associated with shorter survival, and data have suggested that they are more frequently found in pancreatic cancer patients with metastatic disease, "CTCs are relatively understudied," explained Dr. Miksad.

New technology — such as the NanoVelcro technology used in this study to detect CTCs in patients with pancreatic abnormalities — might improve CTC detection, she noted.

However, Dr. Miksad pointed out that CTCs detected in pancreatic cancer patients are "quite variable," especially in metastatic disease, and that CTCs are quite rare, which is one of the primary challenges in this field.

"Test cutoff levels are very close," which "raises questions about reliability and variability," she said.

Overall, CTCs are "intriguing," and the "ability of CTCs to detect the presence of cancer may differ from its use in staging," Dr. Miksad added.

MMP-7 Predictive of Gemcitabine Benefit

n the second study, a novel biomarker panel to evaluate response to adjuvant pancreatic cancer therapy was examined by Gregory Heestand, MD, assistant clinical professor of medicine at the University of California, San Diego, and colleagues.

There is a great need for improved adjuvant therapy in pancreatic cancer, either with drugs currently in development or "using our current drugs more effectively — perhaps by matching the right drug with the patient using predictive biomarkers," he explained.

However, identifying biomarkers has been a challenge in this disease, for a number of reasons, he noted.

The Radiation Therapy Oncology Group (RTOG) 9704 trial evaluated patients with resected stage I to III pancreatic cancer who were randomized to receive 5-fluorouracil (5-FU) or gemcitabine chemotherapy before and after 5-FU chemoradiation. Before receiving adjuvant therapy, patients had baseline serum samples drawn and archived for future study, Dr. Heestand explained.

The investigators used the proximity ligation assay (PLA), which has "been of great interest because it is a technique that allows us to quantify dozens of serum proteins from very small volumes of patient samples," she reported.

Using the PLA, a probe panel was built from commercially available antibodies capable of quantifying 42 key proteins from 24 µL of patient serum. The team set out to investigate whether they could find a biomarker signature that would predict benefit from adjuvant therapy.

Of the 451 eligible patients, 213 had samples available for study. There were 112 (53%) in the 5-FU group and 101 (47%) in the gemcitabine group. As the investigators expected, reduced levels of carcinoembryonic antigen (CEA) and CA19-9 were prognostic of improved overall survival in all patients.

However, they also found that decreased levels of MMP-7 were predictive of a survival benefit in the gemcitabine but not the 5-FU group.

Dr. Miksad noted that a similar finding was observed in non-small cell lung cancer, where MMP-7 was associated with platinum resistance.

However, for treatment predictors in pancreatic cancer, many questions remain, she said. For example, is the finding for MMP-7-related gemcitabine chemoresistance, or does it reflect tumor biology, she asked. Also, does the negative association hold for gemcitabine combination therapy?

Dr. Ankeny has disclosed no relevant financial relationships. The study be Dr. Heestand's team was supported by RTOG grants from the National Cancer Institute. Dr. Heestand reports serving as a consultant or in an advisory role for Bayer and Genomic Health. Some of his coauthors also report financial relationships.

2014 Gastrointestinal Cancers Symposium (GICS): Abstracts 175 and 176. Presented January 16, 2014.