Advanced Pancreatic Cancer: The Road to Personalized Care - Episode 12

Immunotherapy for MSI-High Pancreatic Tumors


John L. Marshall, MD: Every single bus stop, and every single building in Chicago right now has an ad for an immunotherapy product. O’Hare airport is littered with them. Clearly, immunotherapy has been the big thing for a lot of cancers. Where are we with this in pancreas? I’ve never even seen an MSI-high pancreas; has anybody?

Tanios S. Bekaii-Saab, MD: Yes.

Kabir Mody, MD: Yes.

Eileen M. O’Reilly, MD: Yes.

John L. Marshall, MD: You have had some? Tell us a story.

Tanios S. Bekaii-Saab, MD: This was part of the study, but we individualized it to the level of the patient. This was a patient with Lynch syndrome, who had gone through colon cancer and uterine cancer, and then pancreas cancer. In fact, had it resected and then came back and was treated with gemcitabine/nab-paclitaxel with no response, and then were referred to a study with pembrolizumab. She then has an MSI-high tumor. This was 4 years ago.

Within a few months, the patient went into complete remission. Her CA 19-9 normalized completely. At 9 months she started having some neuropathy, which was very intriguing, and it ended up being an autoimmune reaction, and we had to stop pembrolizumab at 10 months. She remained in CR and she remains in CR now 4 years after.

John L. Marshall, MD: I think about immunotherapy toxicities and I’m very glad I went to internal medicine training. How frequent is this problem; this genetic abnormality; and how important is this for a pancreas cancer?

Eileen M. O’Reilly, MD: I mean, I think it’s important but it’s a rare event and we certainly have to keep that in mind. I would say the best data we have suggest about 1% to 2%. We looked at 800-plus people, we found 7. So, it was 0.8% incidence.

John L. Marshall, MD: Is that people living in Manhattan or no?

Eileen M. O’Reilly, MD: That’s all the people that we sequenced. We were able to look for somatic or germline mutations and MMR genes, consequent of next-generation sequencing. All had germline Lynch, and 4 of those individuals benefitted from immunotherapy. One lovely story, in a way, like Tanios’ patient, lived for 10 years plus with advanced pancreas cancer. Died actually of another Lynch-related malignancy and had tangible benefits, durably so, from immunotherapy.

Kabir Mody, MD: We’ve all been so very excited about MSI-high and the approvals founded in some of the colorectal and other GI malignancies. But the more we’ve been looking at the data of a single agent; and even combination immunotherapy, even in the MSI-high patients; it’s not like it’s a homerun. It’s not like kidney and melanoma.

Eileen M. O’Reilly, MD: Well to add to that, this gentleman developed 2 other Lynch-related malignancies while on pembrolizumab, but yet his metastatic pancreas cancer remained controlled. Fascinating, right? We try to learn as much as we can from this.

George P. Kim, MD: I would just add that you probably need to give chemotherapy first. I had the experience of giving a checkpoint inhibitor to a patient before chemotherapy first-line. I called Hopkins and they gave me approval. The patient basically had lots and lots of toxicities. So, it made me very aware of some of the trials in other GI cancers, where we’re giving chemotherapy upfront, culling the disease, reducing some of the inflammation, and then using some of the checkpoint inhibitors.

The other thing I’d say about this MSI; there is an MSI-like—the Europeans look at that in colorectal—I think we need to explore that more. Some of the tumor mutational load work may also be important, to try to broaden the number of patients that can benefit.

John L. Marshall, MD: To get to that molecular testing you need a piece of tissue, and I struggle with this part; that very often the diagnosis is made with a fine-needle aspiration and a few cells, and we can’t do really broad testing. We talked a little bit about that earlier. Is MSI enough to take a patient with metastatic disease and biopsy of them for molecular profiling, or is it part of a larger puzzle? Would you rebiopsy somebody to do genetic testing if you didn’t have adequate tissue in the bank?

Kabir Mody, MD: That’s a good question. I think that the outcomes with immunotherapy in MSI-high patients are enough to justify it. Because it’s another option for them; it’s a potentially meaningful option as Tanios has illustrated. I mean, obviously not everyone has great outcomes on these drugs, but certainly I think that the chances of that are significant enough that I would put them through it.

Tanios S. Bekaii-Saab, MD: It’s not just MSI.

Kabir Mody, MD: Right, it’s tumor mutation.

Tanios S. Bekaii-Saab, MD: Right, it’s BRCA, it’s BRAF, it’s others. And unfortunately, again, if circulating tumor DNA was predictable enough for us. First of all, we cannot get a tumor mutational burden yet on these liquid platforms as a surrogate of MSI-high. So, we don’t have that capacity yet in pancreas cancer; I think biopsy remains king or queen.

George P. Kim, MD: I feel strongly about getting a biopsy; I think it’s justified. It’s what Tanios just said; it’s not just MSI, it is all sorts of mutations that hopefully we can target. There are also rebiopsy clinics, where after a patient progresses, you biopsy them again. We talked about this before. I think we’ve struggled with the empiric approach of treating everybody to benefit from FU. I think we need to be smarter and really understand what the tumor is doing, especially in pancreatic cancer when you have so many underlying mutations. So, I don’t have a problem with rebiopsy; it’s the insurance that has a problem.

John L. Marshall, MD: I run a nationwide precision medicine alliance, and I’m theoretically supposed to know what I’m talking about. But there’s so much I don’t understand about the way these tests are run; giving me a positive or negative. The variance of unknown significance. So, we are all learning on the fly here a bit—whether it’s determining BRCA or the interact; you need the fusion, not the mutation. I think a lot of our colleagues—I know that, because if it’s true for me, I’m assuming it’s true for others—have this report in the chart and they really don’t know how to interpret it and don’t know if it was done with the best quality.

Eileen M. O’Reilly, MD: I couldn’t agree more, John. I think we’re really in the discovery phase of this and still learning how to apply this information and try to sift out what’s important and what’s not. It’s one thing to have an actual funding, it’s another thing to have a drug, it’s another thing to combine them in pancreas cancer and say it works, right? And there’s very few observations that we’re there for. Yet we see these quotes—and I think we’re all guilty of it—saying there’s 30% or 50% actionability. We just don’t know. But nonetheless, I think it’s very important that we’re doing this so that we learn and we’ll be able to, hopefully, better apply this going forward.

John L. Marshall, MD: And I am so, in a way, grateful for the insurance industry and the FDA to look the other way. I shouldn’t say that on camera, but because this is really still research in many ways. There are some standard of care things that justify the test, but doing it in all patients is increasing; just like how we get a CAT scan on every patient. I think we’re moving to a world where we’ll have broad molecular profiling on every one of our metastatic patients.

George P. Kim, MD: All right, let me be the moderator for a second. A patient gets through all available treatments for pancreatic cancer. Are you going to give them a checkpoint inhibitor?

John L. Marshall, MD: Not going to lob the Hail Mary; no, absolutely not.

Kabir Mody, MD: I think it’s inappropriate.

Eileen M. O’Reilly, MD: Not outside of a clinical trial. I think the data—and not outside of MSI—just do not support that.

John L. Marshall, MD: The patient wants it. Well, and what’s interesting in a sort of crazy world that we’re in, the company will send it to you without much; so it’s not like they’re even charging for it. They’ll say, “It’s worth it, give it a try.” But I would say that’s not appropriate.

Tanios S. Bekaii-Saab, MD: It’s not placebo.

John L. Marshall, MD: It’s not placebo.

Tanios S. Bekaii-Saab, MD: It’s not placebo. It has its toxicities, and it has its detriment on the patient’s timelines, in terms of acceptance of what’s coming. Also, I think it doesn’t help bring closure to the patient and to the family. Now we have to think about these angles as more important than just, we’re trained as physicians that we have the duty to continue treating. That’s not true. We are always humbled in oncology to take a step back and think that death is part of life for a lot of these patients and their families. It’s important to make sure that the transition ends up being very comfortable and very comforting to the patient and to the family. So, I actually have incredible problems with the “Hail Mary” approach, at the point where I think that the patient should receive what would be appropriate; that would be hospice and comfort care.

Transcript Edited for Clarity