Tapping a Rich Vein

At the moment, it’s difficult to catch pancreatic cancer early. There are often few or no symptoms in the early stages of the disease, and there are no screening tests recommended by any professional gastrointestinal society. Even in high-risk populations where regular screening could catch disease early and make treatment more effective, the recommendations are controversial. It’s unclear which patients need screening and how best to do it – whether by cross-sectional imaging (like MRI or CT scans) or by endoscopic ultrasound.

Even after pancreatic cancer is diagnosed, there are still problems. The radiologic imaging we use to determine a patient’s eligibility for curative surgery has limited ability to determine postoperative risk of recurrence. Not only that, but we can’t use it to identify whether or not patients would benefit from invasive surgery or harsh neo-adjuvant chemotherapy – and that’s vital, because it helps us determine where the benefits of aggressive treatment may outweigh the risks.

Our goal at the University of Chicago Center for Endoscopic Research and Therapeutics is to pioneer innovative approaches to diagnose and treat gastrointestinal disease. We consider pancreatic cancer a primary focus because it’s a devastating disease that’s difficult to cure – often due to local advancement or distant metastasis. Recently, it has become increasingly clear that not all pancreatic cancers are equal. This offers the possibility of personalized treatment, but it’s still a challenge to adequately assess tumor molecular heterogeneity and perioperative risk of recurrence. That’s where our new test may be able to help.

The power of CTCs

Circulating tumor cells (CTCs) have been explored as a minimally invasive tool for assessing solid tumor burden. But unless the tumor burden is very high, CTCs in peripheral blood are extremely rare in patients with pancreatic cancer – possibly due to biophysical factors like platelet adhesion or cluster size and shape, which might trap the CTCs in smaller vessels. So we hypothesized that collecting blood from a different site – the portal vein – might yield a differential CTC count and provide us with the necessary biomarkers to personalize treatment.

To do that, we devised a technique that makes use of endoscopic ultrasound and a small needle to sample the portal venous blood. Our technique involves inserting a linear echoendoscope into the stomach or proximal duodenum via the mouth. Under ultrasound guidance, we identify the left and right portal veins and verify flow signal. Then, we advance a 19-gauge EUS-FNA needle transhepatically into the portal vein and take two to four 8.5 mL aliquots of blood (see Figure 1). Once aspirated, we can count and characterize the CTCs in the blood, helping us to define the patient’s diagnosis and make better treatment decisions.

What makes our method better than the existing ones? It gives us access to the portal venous blood, which contains CTCs even in gastrointestinal tumors where the peripheral blood does not. Because we can isolate sufficient numbers of those cells, we’re able to perform genomic and proteomic tumor profiling to personalize pancreatic cancer treatment. Not only that, but the technique has a good safety profile – minimally invasive, incisionless, and with no complications seen in the patients who have undergone this type of testing.

So far, we’ve used portal venous sampling to evaluate 18 patients with suspected pancreatobiliary cancers (1). We were able to detect CTCs in the portal vein blood of all 18 – but when we sampled peripheral blood, only four of the patients had detectable CTCs. On average, portal vein blood contained 118.4±36.8 CTCs per 7.5 mL, whereas peripheral blood contained only 0.8±0.4 CTCs in the same volume. For patients with less invasive disease, the technique is particularly effective; peripheral blood revealed a mean of 0.4 CTCs and a median of 0, but our portal vein samples had a mean of 83.2 CTCs and a median of 62. It seems clear that it’s much easier to locate CTCs in portal venous samples – and that portal vein blood contains enough cells to enable further analysis.

A promising prospect

Of course, CTC analysis is still in its infancy. The results we recently published came from a pilot and feasibility study; we wanted to ensure that portal venous CTC sampling could be done without excessive risk to the patient. Our next step is to conduct a prospective study, which will let us determine what role portal vein CTC isolation, enumeration and molecular profiling can have on prognostic significance, including postoperative recurrence, length of survival, or response to neo-adjuvant chemotherapy. So the test is not yet ready for clinical use – but because CTC enumeration has already been validated and is US Food and Drug Administration-approved for other solid tumors, we’re optimistic about the prospects of this new method.

The most immediate potential of portal venous CTC sampling is to provide clinical utility in stratifying risk of postoperative recurrence, as well as helping to identify patients who may benefit from more aggressive treatment. Ultimately, I hope it will supplement imaging to help doctors determine personalized cancer treatment plans for their patients. I have an additional hope for the test, too – I think it has a future role in screening high-risk patients whose cancer may not yet be visible on imaging studies.

And pancreatic cancer isn’t the end of the story. My colleagues and I believe that portal venous sampling can be applied to many different gastrointestinal cancers – for instance, colon cancer, where intraoperative surgical access has already confirmed the presence of CTCs in the portal vein. Any cancer with a high rate of metastasis to the liver is likely to have CTCs arriving via the portal vein, which also broadens the applicability of our test. In the future, we’ll need to clarify what types of CTCs harbor the most risk for seeding metastases and determine what kind of molecular profiling we can use to characterize them. In the meantime, we’ve shown the feasibility of portal vein access for effective liquid biopsy – and hopefully, this will soon translate into better research, diagnosis and treatment of pancreatic cancers.

Christopher Chapman is a gastroenterology fellow at the University of Chicago Hospitals, Chicago, USA.

Mission: Impossible by Michael Schubert

New Ab-ilities in Imaging by Michael Schubert

Surprising Subtypes by Michael Schubert

Friend or Foe? by Anne Marie Lennon

(Chemo)resistance Is Futile by Michael Schubert

Breaching Cancer’s Defenses by Ingunn Stromnes

An Epigenetic Epiphany by Michael Schubert

Building a Better Mousetrap by Michael Schubert