NN: Welcome to this ecancer discussion on tumour agnostic biomarkers and, more generally, precision medicine in gastrointestinal cancer. I’m Nicola Normanno, Director of the Department of Translational Research at the National Cancer Institute of Naples in Italy and today with me is Katja Schmitz. Katja, do you want to introduce yourself, please?
KS: My name is Katja Schmitz. Thanks for giving me the opportunity to take part in this discussion.
I’m Head of the Molecular Pathology Lab at the University Hospital in Innsbruck and I’m a board certified pathologist.
NN: Well, Katja, today the first topic that we need to discuss is about tumour agnostic biomarkers. Actually, there are a few tumour agnostic biomarkers that have been approved up until now with basically, NTRK fusions and TMB. The NTRK fusions are really rare and there are some doubts on using TMB as an agnostic biomarker for all tumour types. Let’s start with NTRK, these mutations, these genetic alterations are quite rare. What is your approach in your daily practice to find NTRK [?? 1:18], of course, focussing on GI cancers?
KS: Depending on which cancer type we are analysing, we are performing an RNA-based NGS sequencing. So we have different panels which are covering NTRK 1-3. So if I’m performing an NGS fusion panel that’s enough and, for example, for colon cancers where I only would do a DNA-based NGS we are applying first NTRK immunohistochemistry with a pan-NTRK antibody.
We use the NTRK immunohistochemistry for screening and if we have any positive staining, so any intensity or any percentage of tumour cells being positive, then we are performing an RNA-based NGS analysis.
NN: This is interesting. Are you screening all your colorectal cancer patients for NTRK immunohistochemistry because most centres don’t have this approach?
KS: Yes, it’s a very small percentage, obviously, and we have some false negative cases in the immunohistochemistry. But our clinicians want us to do that so for our patients in our hospital we are doing that.
NN: Another approach would be to focus on patients who have either microsatellite instability because that cohort usually is enriched for fusions. So this is a suggestion that we could also give to our colleagues. Another point is that, for example, in cholangiocarcinoma since we are already screening for FGFR fusions, all the panels that screen for FGFR as a matter of fact you could also do NTRK. So by using this approach, by using NGS for our cholangiocarcinoma patients we can also identify the NTRK fusion positives, right?
KS: Yes, correct.
NN: Very good. And what about TMB? Do you have requests and tests for TMB in patients with GI cancer?
KS: No, we don’t have any requests for TMB. I think we had one case in the last year and we are, right now, not performing it in our laboratory but we are collaborating with another lab in the north of Tyrol. So we are working with them together.
NN: Actually, as a matter of fact, TMB, tumour mutational burden, has been approved as an agnostic biomarker for the use of pembrolizumab. But there are data suggesting that in many cancers that patients [?? 4:01] don’t really respond to treatment. For example, at the last ASCO GI, there was presented an abstract with the combination of nivolumab ipilimumab in patients with microsatellite stable colorectal cancer and high TMB with a cut-off of 9 and actually very little activity was observed. Therefore maybe we still need to start additional investigations in this.
KS: Yes, and right now we are building up our laboratory, so we are buying a new sequencer so we will be able to do TMB ourselves.
NN: Thank you. In this respect to me it’s clear that one of the biggest innovations in the past few months has been the introduction of immune therapy for patients with microsatellite instability.
All the data that we have are confirming that actually immunotype is very active in this subset of patients. Nevertheless, there have still been only a fraction of patients here responding to immune therapy, about 40-45%. So it may be that we need additional biomarkers also in this field. Are you working in this area? Do you have any insight to this?
KS: No, not really, we’re just doing the MSI analysis for all our patients with colorectal, gastric and cholangiocarcinoma.
NN: Actually there are some intriguing data suggesting that if you combine TMB with MSI you can identify the patients with the highest level of response, but these are really preliminary data. On the other hand, we spoke about fusions and of course one of the hot topics is FGFR fusions in cholangiocarcinoma. Actually at the ASCO GI meeting we presented one of our abstracts on the activity of drugs, also immune drugs, in patients who have FGFR fusions. Should we look only for fusions or also other genomic alterations? What is your feeling about it?
KS: I would suggest you look not only for fusions because there are data available showing that patients with FGFR mutations are also responding to the drugs. Our panels are constructed that we can only find the fusion if we are doing the NGS RNA-based. We can also find the mutations and I am including them in my report. Of course, if I would only do a FISH analysis for break-aparts I would miss all the mutations in FGFR.
NN: I fully agree that we need NGS. With NGS up until now only the fusions have been approved as a biomarker but to investigate mutations is important even because some of these mutations, in particular in patients who are treated with FGFR kinase inhibitors, can be associated with resistance. Maybe these mutations can be tracked and identified by using liquid biopsy; this is a possibility, of course.
KS: Yes, I totally agree.
NN: And, actually, the availability of new FGFR kinase inhibitors will maybe make possible what we are already doing in lung cancer with EGFR or ALK in which we alternate different inhibitors that are able to block specific mutations that are associated with the disease. So this is a new field and some new data have been also presented at the GI meeting. It’s very interesting to explore.
Now, liquid biopsies are interesting to monitor the response but now there are also initial applications in [?? 7:39]. There has been an interesting abstract on the use of liquid biopsies to help the diagnosis of patients with suspected pancreatic cancer or biliary tract cancer. I don’t think we can really replace tissue biopsy with liquid biopsy but in a patient in which there is a high suspect of a tumour if we ran a liquid biopsy and we found a genomic alteration that is specific for that tumour, do you think that this can help for the diagnosis that we use for our patients?
KS: Yes, I think that can help, definitely. Because especially those fine needle aspirates from pancreatic cancers where you have a very little amount of tumour cells, I think that liquid biopsy is always worth a try. So if you get a non-conclusive result then it doesn’t help, but if you really find a mutation or even a fusion then it can definitely help.
NN: I fully agree with you. At least liquid biopsy can provide complementary information to tissue biopsy so this is a tool that we have available and we need to start to work also with this tool. [?? 8:50] there is increasing evidence also that liquid biopsy could be very useful in patients with early colorectal cancer to identify minimal residual disease after surgery. What do you think? Is this going to be soon applied in the clinics? Are we close to using liquid biopsy in the clinic for these patients?
KS: Yes, I think that will come very quickly that we will have to diagnose minimal residual disease, just to give the clinicians a follow-up. So it’s very important to have the liquid biopsy initially at the date of the diagnosis and take it at the time when the tissue biopsy is taken. Then we can always correlate our results. So I wouldn’t only rely on the liquid biopsy but I would always want. to correlate it with a tissue biopsy on the initial diagnosis. Then it’s much easier to follow-up the patient, to look for minimal residual disease to see how he responds to the therapy. Then we have the opportunity to find the resistance mutations if we still see the initial genetic aberration.
NN: Yes, very nice. Actually, the point is I keep looking at all the genetic data we have. It’s clear that if a patient has a positive liquid biopsy after surgery or after giving therapy the risk of relapse is very high, with no doubt. The point is that, unfortunately, some of the patients who are liquid biopsy negative will have a relapse. It’s a small fraction but still has been found consistently in all the studies. So it’s clear that in the patients that are positive maybe there is room for an intensification of the treatment, however, I’m not sure whether in the patients that are negative we should not give a different chemotherapy because actually we don’t manage to identify all the patients at risk. We don’t know whether this is due to the limit of sensitivity of the techniques or the biology of the tumour because maybe some mets could be dormant for a while before they provide any kind of significant release of circulating tumour DNA.
Katja, still only a fraction of patients who carry FGFR fusion respond to treatment with the FGFR inhibitors. So there are, most likely, some additional mechanisms that might limit the sensitivity to these drugs. Is there any evidence from the literature or possible additional biomarkers that can help us to better define some subpopulations of FGFR positive cholangiocarcinoma patients who will really respond to treatment?
KS: Generally mutations or fusions in cholangiocellular carcinomas are quite rare but we can find additional mutations like IDH mutations, BRAF or KRAS, NRAS mutations, which generally show that the patients have less response to TKIs.
NN: Actually the presence of other drivers, of course, will be important. There are some preliminary
data suggesting that the presence of genetic alteration tumour suppressor genes might be associated with a resistance response but I’m not sure whether this is really a true activity of these tumour suppressor genes or the presence of these mutation tumour suppressor genes suggests that maybe there is a tumour heterogeneity that leads to resistance. So this is something that we will need to address more in the next studies. Of course we need data and [?? 12:51] clinical trials to investigate these questions.
Now, there are also other emerging biomarkers in GI cancer, for example, BRCA in pancreatic cancer. What do you think? Do you already use this in your clinical practice? Do you test your pancreatic cancer patients for BRCA mutations?
KS: Yes we do test them but because of the law on human genetics we are only testing the patients when we talk to the clinicians before. But since we are only testing the tumour tissue and we are not looking for germline mutations it’s actually safe as a pathologist to do the BRCA1 and 2 analysis in cancers in general.
NN: I agree. And actually there are also some interesting data on BRAF mutant cancers. It is confirmed it is found also in biliary tract cancer. Again, is BRAF included in your panel when it’s in biliary tract cancer? Do you have access to drugs or to clinical trials in your centre?
KS: Yes. BRAF is included in all of our panels in DNA- and RNA-based NGS panels and our clinicians do have access to BRAF inhibitors or MEK inhibitors.
NN: The response rates are usually not that high but this is true also for colorectal cancer. So it’s possible that, again, even for BRAF mutant cancers we need to go more and deep in genomic profiling of these patients and identify additional markers that can predict or not response. My last point actually is also on this story of extended RAS. There are different opinions, there are labs that are still are using only real-time PCR to test for RAS mutations or even [?? 14:49]. Are you using NGS and up front? Do you test only for KRAS and NRAS and BRAF or do you also test for additional genomic alterations, for example, HER2 amplification?
KS: Right now we are looking for the mutations and fusions only and not for the FGFR amplifications. We used to do that for lung cancer, for FGFR1 but right now we don’t do that anymore.
NN: Yes, I was talking about HER2 amplification and ERBB2 amplification in colorectal cancer.
KS: Yes, sorry. Yes, we are looking for ERBB2 amplifications. We are screening with immunohistochemistry and then performing the FISH.
NN: Okay, so, as you see even in colorectal cancer we are always expanding the number of biomarkers that we are testing more generally. There are more and more biomarkers even for our GI cancer patients So, many thanks for watching this video. I hope that we could provide some valuable information for your clinical practice. Thanks again.