Liquid Biopsy in Metastatic Lung Cancer

Jared Weiss, MD:There are two strategies to integrate plasma-based testing into clinical practice. Right now, I think that the dominant way that it’s being used is for the patient who, for whatever reason, can’t get a tissue-based test. It could be that there isn’t local availability of the appropriate testing modality, the patient declines it, or the patient or doctor is afraid of side effects or complications of that biopsy to bring molecular testing to a population for whom it’s not currently available. But these tests are getting more and more validated every day, and the turnaround time is getting faster and faster every day. So, in a world where there’s a several-week turnaround time to a non—tissue-based test, I’m not willing to wait for that result to then consider tissue-based testing, because lung cancer can hurt a patient in that time span. But these tests are getting really fast. We’re just about in an era where these results can come back in a matter of days. When you’re talking about a matter of days, now the risks of biopsy might exceed the risks of waiting. That is an era that in some places we’re in and other places we’re rapidly entering.

And in that kind of a world, with good validation and very rapid turnaround, there’s a new strategy that I’m actively considering moving to. That strategy would be that any time that I want molecular characterization, whether it’s up front or at the time of acquired resistance, it would be to get the plasma-based testing. It’s a few-day turnaround to get results. If I get an actionable result, for example, in an EGFR-mutated patient with acquired resistance, I get T790M. In contrast, if I get no result back, a nonactionable result, then I’m, at that point, going to move into tissue-based testing.

There’s one caveat here from very recent data that may be worth commenting upon. There were data shown recently at the European Lung Cancer Conference looking at results of plasma and tissue-based testing for the use of third-generation TKIs. And in looking at the patients whose plasma result was negative—so they didn’t find a result in plasma—either because there is no T790M in that plasma or, in some cases, because the tumor doesn’t shed DNA, which is a noninformative result, they subtype those patients with the negative plasma into those in whom the original sensitizing mutation was found. This is a population where the cancer probably is secreting DNA and there is just no T790M versus those for whom no molecular results could be found. The original sensitizing mutation you know is present in the cancer wasn’t there. As you might expect, those whose cancer was shedding DNA, who had the original sensitizing mutation but not the T790M, had a very poor result with the third-generation TKI. Those who had an uninformative result, no DNA being shed, had better results. And so, it may be that the presence of the original classical mutation can help you in understanding if you’re just failing to find a mutation or if it’s just not there.

If I get a negative result on plasma-based testing in an up-front patient, or a patient at new diagnosis, then I’m going to get tissue on every patient. And for the location of tissue acquisition, it just has to be cancer, any spot is fine. Classically, when we’re using PCR, we avoided bone. It can be a little bit of a problem with next-generation sequencing, but much less so. So, if there’s a nonbony spot, maybe you get a nonbony spot. But mostly my decision is about safety and comfort for the patient. I’m going to go after it in the first line.

In contrast, if you have a patient with acquired resistance to a targeted therapy and you’re trying to understand the mechanism of that acquired resistance, try to get the patient a next-line targeted therapy. Then, if I get a negative plasma result, the first thing I’m looking to do is see if whether the original sensitizing mutation that I know to be present in the cancer is there or not. And this is important to me. Because if I don’t find that original classical mutation, then I know that this cancer is not secreting DNA in the way that I can find it. That makes it mandatory for me to then get tissue.

In terms of what spot to get tissue for, I’m still thinking about patient safety, I’m still thinking about patient comfort. But the additional factor that I’m adding in is that I really want to biopsy a spot that’s growing, because there can be heterogeneity inside of the patient where some spots of cancer are still sensitive to the original TKI and where others have this new resistance change. And if what you really want to know is what that resistance change is, you have to put a needle in the spot that has that resistance change.

Weiss case 2:

A 62 year-old neversmoker with stage IV adenocarcinoma

• Mutation testing showed an EGFR exon 19 mutation
• The patient was treated with erlotinib for 1 year; he developed diarrhea and rash which was successfully managed
• After 3 months, his disease showed progression on CT; he remained without symptoms.
• He subsequently developed cough and shortness of breath, correlated to progression of a centrally located lung lesion on his follow up CT scan
• Repeat biopsy and molecular testing of the central lesion showed EGFR T790M-positivity
• He was switched to osimertinib therapy and achieved a partial response
• The patient reported dramatic improvement in his well-being