Molecular Testing for NSCLC Rapidly Evolves to Include More Markers

Kristen Champion, PhD, FACMG

Molecular testing for non—small cell lung cancer (NSCLC) has rapidly evolved within the past few years to include the various biomarkers that guide physicians in creating personalized treatment approaches for patients—and it shows no signs of slowing down, explained Kristen Champion, PhD, FACMG.

“Now, we have multiple next-generation sequencing (NGS) platforms going in the lab and over 20 different panels. We have evolved a lot. It's been especially beneficial in the lung cancer realm because we keep discovering more biomarkers at such a fast pace, making it a really exciting field to be part of,” said Champion, who is senior director of the Molecular Oncology Laboratory at Quest Med Fusion.

In an interview with OncLive^® during the 2019 State of the Science Summit^TM on Precision Medicine, Champion discussed various molecular testing methods being used in NSCLC.

OncLive: How has molecular testing for NSCLC evolved over recent years?

Champion: It's really changed dramatically, even in the last few years. When I first started in the molecular oncology laboratory at Med Fusion, the menu was very small with one assay by polymerase chain reaction (PCR). When I got into the lab, we started validating our first next-generation sequencing (NGS) assay, which was exciting. We were up and running about 1.5 years later, and it's taken off ever since.

How is HER2 used as a biomarker in lung cancer?

The HER2 mutation in lung cancer is an example of how off-label use of an existing therapeutic has benefited a new population of patients. We discovered that patients with lung cancer can have these mutations in the kinase domain of the HER2 gene. Previously, the mechanism of HER2 was about amplification and overexpression. The studies in lung cancer had been quite disappointing when trying to target that biomarker, but then we discovered that [HER2] has a different mechanism.

[HER2] has activating mutations, which shed a light on how it could be targeted. We had studies come out using HER2 inhibitors in combination with chemotherapy first. About 1 year ago, we had a study, results of which showed that ado-trastuzumab emtansine (T-DM1; Kadcyla) was effective in targeting patients with these HER2 mutations in lung cancer. The study showed that patients who had these mutations, specifically in the kinase domain but also in other regions of the gene, were quite responsive. These were exciting, positive results for this therapy.

MET exon 14 has also been of interest in lung cancer. What do we know about this marker?

MET has been a target that we've known about for a while. We have gotten fatigued about it because we have tried to target it for quite a while, and the results have been disappointing. There was some pessimism about whether MET was actually a predictive biomarker in lung cancer. However, a few years back, we discovered that lung cancer, in particular, has these MET exon 14 skipping mutations. They cluster in the donor acceptor and donor splice site, flanking exon 14 in the gene. Once that mechanism was discovered, we realized that those patients were very responsive to MET inhibition.

Could you describe the assays used to test for these abnormalities?

When it comes to targeting MET exon 14 skipping mutations, the method matters. There's a wide spectrum of different mutations, from point mutations to large structural variations. Therefore, the method has to be a comprehensive method. Additionally, it's important to consider whether it's DNA- or RNA-based. DNA-based methods predict that the mutation may cause exon 14 skipping, but they don't truly know. Rather, you have to infer whether it's actually causing [exon 14 skipping] without proof. However, when you're sequencing in RNA, you can see that the mutation identified is causing a MET exon 14 skipping event.

DNA-based tests that are inferring may actually report mutations that are not fully predictive of response. You have to be cautious about what assay you're going to use to detect MET exon 14 skipping mutations. Our assay has chosen to use the anchored multiplex PCR assays for detection of MET exon 14 skipping events. That is a nice assay that can be used for the novel fusion events, as well. You can put together a comprehensive panel of fusions and exon skipping events that can all be targeted in a single assay.

How do ALK resistance mutations influence NSCLC treatment decisions?

The ALK resistance mutation story in lung cancer is reminiscent of the EGFR T790m story that we've known about for quite a while in lung cancer. There are a few important differences. For the ALK mutation, in terms of the resistance mutation, the spectrum is very wide. There are quite a few mutations that have been reported, whereas for EGFR, we had just the one mutation that could be targeted. There are a lot of mutations scattered in the gene that can pop up. The inhibitor the patient was treated with determines the mutation you're going to see and how the patient is going to respond. It's a bit more complicated in those respects.

How is lorlatinib (Lorbrena) used in ALK NSCLC?

We have seen some exciting results come out this year for lorlatinib. It's a third-generation ALK TKI, and it's currently approved for patients who progress on a first- or second-generation TKI. Importantly, it can penetrate the brain. This is good for patients with brain metastases and it's been shown to have broad spectrum inhibition on the different ALK resistance mutations that have been reported.

How do you choose what molecular assay to use on patients?

For HER2, we use an NGS method. Prior to NGS, we were using PCR fragment-sizing assay because these are small insertion events. Because the genetic diversity is small, you can use a targeted assay for HER2 mutations. However, NGS can be coupled with other targets; you can put together a comprehensive panel that includes HER2. For any of these biomarkers, you want to test with a broad panel to conserve the tissue. While you can target HER2 mutations with a single assay, it's better if it's part of a comprehensive panel.

How do you use these results overall to then make treatment decisions?

Treatment can be complicated, and it depends on the individual situation of the patient. We don't treat every patient the same. It's also important to consider the combination of biomarkers that the patient may have. For example, the patient may have a PD-L1—positive tumor, but then you also have to pay attention to their EGFR or ALK results. It's important to get a comprehensive view of the patient's molecular findings to determine the right course of treatment for them.

What type of work is your lab doing in the development of molecular testing?

In my lab, we're in the process of validating a fusion panel by NGS. In that process, we've taken fluorescence in situ hybridization (FISH)—positive cases and we put them on the fusion panel by NGS. We saw some non-correlation and took a deeper dive into that. We discovered that all cases that didn't correlate between FISH and NGS had a variant signal pattern. They had a [variant pattern] instead of a split signal pattern.

I'm not [well-versed on] FISH, so I sat down with our cytogeneticist and we looked at these pictures. It was very consistent. I looked into the literature and found that this phenomenon has actually been reported. I wanted to bring that up and point out that FISH has been the gold standard for fusion detection for a long time, but it seems to have limitations. These FISH patterns, whether it's the common split signal pattern or the variant pattern, which actually isn't that uncommon, may be important regarding whether patients will respond to therapy.