Perspectives From the Frontline: What Is Happening on the Ground With NGS Testing in NSCLC?

Benjamin P. Levy, MD

As molecular drivers take a more prominent role in dictating courses of care, collaborations has become essential for ensuring that patients with non–small cell lung cancer (NSCLC) receive optimal treatment beginning at diagnosis. The rise of multidisciplinary care in this arena calls together thoracic oncologists, pathologists, interventional radiologists and pulmonologists, as well as a plethora of supportive staff to ensure that tissue samples and testing are executed within the parameters of a new standard—one that calls for larger sample sizes, next-generation sequencing (NGS) and other biomarker assays, and division of responsibilities across laboratories and institutions.

In an OncLive Insights^® program, Benjamin P. Levy, MD, discussed the challenges with tissue testing facing care teams. For a community perspective, he spoke with Bhuvana Ramkumar, MD; and Neil Morganstein, MD, who walked through the multilevel hurdles in procurement, analysis of results, and insurance.

Challenges With and Strategies for Tissue Procurement in NSCLC

Levy: In my mind, lung cancer is the poster child for precision medicine. EGFR and ALK are chapters 1 and 2 of the targeted therapy story that is still being written. We can start off there, but things have gotten incredibly complex. My running joke is that I went into lung cancer because it was incredibly easy 15 years ago, and this has certainly changed. The complexity [of the treatment landscape] has generated a tremendous amount of questions for patients, but also enthusiasm on behalf of both the oncologist and the patient. That has translated, fortunately, to include [improved outcomes].

We have now 10 biomarkers [with drug] approvals: EGFR, ALK, BRAF V600E, MET exon 14 skipping mutations, EGFR exon 20, ROS rearrangements, ALK rearrangements, NTRK rearrangements, HER2 mutations, and RET fusions. It’s getting exciting because can let these targets to specific therapies that can improve outcomes. But you can’t give the drug unless you identify the target.

How is tissue procured for patients with NSCLC within your practice?

Morganstein: It depends on where the patient is coming from, what their status is. Most patients are already coming from [having received] a diagnosis of some sort. They are coming from their interventional radiologist, their plastic surgeon, [or] interventional [pulmonologist]. Sometimes when we see a patient we’re a little bit behind [knowing] whether the patients have had NGS sent off from pathology. Typically, when we see individuals, particularly those with metastatic disease, we’ll send for a liquid biopsy, and then try to procure the appropriate specimen from pathology and send it off to one of the companies of our choice [for NGS].

We are working on getting reflex testing, but it’s challenging, and it hasn’t worked perfectly. Some tests are being done in-house such as for ALK and ROS mutations, they are done easily with FISH [fluorescence in situ hybridization]. But sometimes there is a delay with tissue procurement and obtaining information.

Levy: That is not just in the community practice. Even in the academic centers like mine, we’re having challenges of procurement and delays. Dr Ramkumar, what is your experience with tissue procurement and challenges in your practice?

Ramkumar: To add to some of the points Dr Morganstein spoke to, I see a lot of patients in the hospital, too. They are admitted and desperately sick. [These are patients who] are not screened and then they come with stage IV [disease] and are admitted quick. We try to get a biopsy when they are in the hospital and that gets technically challenging to move the patient or to get insurance reimbursements.

If the patient comes back and I see them in the clinic, in the community I [send the patient to an interventional] radiologist who does [tissue procurement] a lot. But if their disease [involves] the bronchi and needs tissue, I go with an interventional pulmonologist at the cancer center. [In this case] I have to arrange and [make sure they] know to get more tissue because [even though we have been] telling them that for a while, they do what they can do, and they try to get what they can and do their best. But then we are stuck when we get the results and realize it is not enough [tissue] and then we must do the entire thing again. That gets the patient frustrated. Now that we need more testing, we need more tissue.

Levy: Dr Ramkumar, anecdotally, how often does that happen that, even in the best circumstances, you’re getting insufficient results?

Ramkumar: Right now, it is much better than it was 5 years ago because I know the radiologist I work with so I know they are going to get what I need. But approximately 10% to 15% of the time, I might be planning to do more [tests] on the tissue, that’s where it gets harder. We now have the liquid biopsy backup, and we know that once we can use what we can from the tissue, and we are all doing liquid [testing as well]. I am doing liquid biopsy for everyone I am seeing so I know I have that backup.

Enhancing Communication Among the Multidisciplinary Care Team

Levy: Communication is key before the biopsy. Dr Morganstein, what is your experience with insufficient tissue and is there a way to overcome this challenge? Are you talking to the proceduralist as well as the pathologist? Or does that happen after the fact?

Morganstein: Both and [the communication] is ongoing. This role of interventional pulmonary is new and we just started using this term in the past couple of years. The interventional pulmonologist, many of them are much younger [in age] and they’ve grown up with NGS and know they need [sufficient] tissue. If you read the reports you know they need 5 passes or 6 passes, they know they need tissue. With rapid on-site evaluation it’s become much less of an issue. They know getting 1 skinny needle biopsy that has a few cells isn’t adequate.

We do battle insurance issues, and it’s on ongoing battle to get tissue sent out immediately, there are [staffing issues], even when we request tissue it doesn’t get sent out. There are all kinds of hang ups along the way, physicians know what to do. There isn’t an educational [issue], it’s more of a logistical aspect, and a structural aspect to get the appropriate test run.

[We are in] communication and [we have] what we need [to do it], it just often does not happen.

Levy: In small systems or large systems we’ve tried very hard to have this communication [with the] interventionalists and pathologists prior [to biopsy] but even under the best circumstances we are also struggling with reflex testing. Is reflex testing being done at your centers and then once that happens, what type of communication is going on between you, the pathologist, and the care team?

Ramkumar: I work in the community [clinic] and the community hospital pathologist are [hesitant to do] reflex testing right now because then it falls on them to do the authorization. In the cancer center, it’s not a problem, PD-L1 testing gets done. But for the other [tests], we have to do the paperwork for the prior authorization, or pull the paperwork for [NGS testing], and pathologists are not keen about sending [the samples].

If the patient is admitted to the hospital, and they have a chest tube [or other procedures], we can get the tissue, but we have to wait for [the patient] to be discharged for 24 hours to do the NGS. In the community setting it’s very hard, at the center we can still do the EGFR and ALK [testing] and they get paid, but in the community setting it’s harder.

Levy: You mentioned PD-L1 testing, and I’m curious from the community perspective, is the message out there that if you have a PD-L1 result you still need to wait for the testing or do you find that a lot of your colleagues are using [those results] and are not following up on the molecular report? We have a lot of data that suggest that if you have a PD-L1 that is high, then if they have an EGFR mutation, single-agent immunotherapy doesn’t work that well nor does chemoimmunotherapy. Is that permeated in the community?

Ramkumar: As far as community oncologists are concerned, I think everybody knows to wait for the molecular results before acting on PD-L1 unless a patient needs to be treated right away. We do just chemotherapy to start with as we are waiting on the molecular or PD-L1 [results]. But I don’t think any community oncologist will just go by PD-L1 and get started.

Morganstein: I think it is pretty well-documented, the lack of efficacy if their disease is EGFR-positive and PD-L1 and I think that has trickled down both to the pathologist and the community oncologist. They are well aware that you need the entire panel and it’s been repeated over and over again that you really need the full picture and panel before you can really treat [an individual].

Of course, if a patient needs urgent treatment, we have standard chemotherapy which is fine. But it’s harder to give the appropriate treatment. I always bring this back to the breast cancer oncologist who did much better than we have—they won’t treat somebody unless they have ER [estrogen receptor] or HER2 [results]. We haven’t gotten that good yet. We still see patients all the time without all the information. If you try to see a breast cancer clinician without ER, PR [progesterone receptor] or HER2, see what happens to the pathologist. We’re just not there, we’re getting there but we’re not close.

Best Practices for Accurate Tissue Genotyping in NSCLC

Levy: I frequently say that we are learning a lot from the breast cancer world in terms of diagnostics [and] therapeutically [such as with] antibody-drug conjugates being used in breast cancer that we are just starting to learn about in lung cancer. I am leaning on my breast cancer colleagues quite frequently and, again, despite our best efforts, the nihilism of lung is in some ways preventing [reflex testing] from happening at a lot of institutions. This should be done ideally before we see the patient.

I’ll ask the unfair question, there is no right or wrong here, but what should be best practice if you look at the tissue journey from diagnosis to treatment?

Ramkumar: Once we do the biopsy and the pathologist knows it is NSCLC, [we have to have a preferred] panel which they can then reflex and send [the tissue] to. It’s a little complicated now because we have a few [panels available], but if we have a standard they can work with and if they are able to get that started, by the time [a patient] comes to see us, if the NGS is already started and we know it’s in the process we might have it by the time they see us. It is going to take time for that to happen especially for the community pathologist to come on board, I don’t know about the cancer centers, they may be ahead of us.

Levy: Not really.

Ramkumar: If [pathologists] don’t run into any insurance issues and they get comfortable doing it will all fall in place and time will tell.

Levy: That is good insight. Is there an opportunity to educate the proceduralist, whether that be the interventional pulmonologist or the interventional radiologist about the importance about how much tissue they get because that is a big thing? Or do you find that they are aware at this point that you need a lot of tissue?

Morganstein: Both. In some ways they know that they need a lot of tissue, but I think that [we should] always be reminding them that we need more and that it is not adequate just to make a tissue diagnosis and overly repeat. Tissue by itself is no longer adequate to treat lung cancer. A diagnosis of adenocarcinoma gives us less than 50% of our information. We know [that the individual] has cancer but it leaves us hamstrung. They know it but repeating is always useful, it’s on us as the treating oncologist to throw fits until we get it. We just got our pathologists to send immediate NGS on all lung cancers whether they get paid or not because it is inadequate care. If you just keep saying it, it is a quality issue, eventually somebody will listen.

Average Turnaround Time for NSCLC Molecular Testing

Levy: In a deeper dive into the evolution of molecular testing, what are some of the optimal NGS techniques? We’ve come a long way, we started with Sanger sequencing which was a hot spot of 3 or 4 genes. We move to this larger hot spot panel, what we used to call NGS, which was basically fishing out 50 genes and looking at just them. Then we moved to the NGS platforms, which can test over 300 genes and I think are the gold standard.

We also have other things which may still make a difference like IHC [immunohistochemistry], which we use for PD-L1, and FISH testing, which [Johns Hopkins] was using up until 2 years ago for all the rearrangements—ALK, ROS, RET, and NTRK. When we look [at the landscape now] we have Sanger sequencing, we have the hot spot 50-gene panel, and then finally we culminate in this hybrid capture NGS, which we hear about with all the commercial assays.

[With that] we can discuss turnaround time. There are a lot of best practices for turnaround time and [Johns] Hopkins, is not emblematic of best practices, we are working on it, but we still have a lot of turnaround time issues. In your experience what is the average turnaround time from biopsy to getting the results?

Morganstein: It’s a long time. It’s weeks, 2 weeks is early and if I get something from biopsy to all the information I need in 2 weeks, that is pretty good and unusual.

Levy: Where do you see the major hold up?

Morganstein: Send out. I have to give the [NGS] companies credit, their turnaround time from when they get the block until we get the answers is [approximately] 1 week to 10 days. It’s [the time between] our request to getting it sent out to the company—this is a nebulous place now. Hopefully, the institution is going to get in-house NGS which would be great, but that is still a ways away.

Ramkumar: If we already have the biopsy and I send the NGS at that time, I say it’s 2 to 3 weeks combined with the insurance approval. But it depends on where the block is. In the community [setting], sometimes the clinician may be asking for another opinion if they are not comfortable.

If the block goes out somewhere and they have to track it through insurance approval and then track it again to get it to the place where it needs to go to have the NGS testing. I say 2 to 3 weeks, but it is usually much longer than that.

Levy: We do the same thing, asking, “Where is the tissue in its journey right now? Is it in the pathologist’s office? Is it in our FISH, cytogenetics, or NGS laboratory?” It’s hard to track it.

Ramkumar: If [by mistake] we order EGFR and ALK alone, like we used to do before because we know we can get something out of it, the sample could have gone somewhere else. [We] have to limit ourselves to PD-L1 and count on doing the NGS if it gets done at all.

Levy: Is it a major hurdle there too, just getting it out to the company?

Ramkumar: Usually it is OK, but the pathologists are a little slow on [sending it out].

Receiving Molecular Testing Results for NSCLC

Levy: Let’s talk about the actual results and this is where it gets tough for plasma and tissue results. Is there a best molecular report? Should there be a standard here?

Morganstein: Some of it is [what you’re] used to. For example, I’ve been using [Foundation CDx] for years, so my eyes know exactly where to go on the report. I’m always afraid I’m going to miss something on some of the other reports. [What] I don’t love is that some companies will give you something that is “targetable,” but not necessarily targetable in lung cancer, such as PIK3CA mutations or that you don’t know exactly what to do with the results.

Levy: I would agree that some of these reports are structured better than others. How do you convey the results to the patient so that they are informed about what they need?

Ramkumar: When we run the NGS or liquid biopsy, we tell them, we are running more tests to see if they can go on an oral agent. That’s been 1 instance where patients are OK to wait because they have already gone through so much and [if they] wait for more testing, right there is a possibility they can go on an oral agent and don’t have to get a port and IV chemotherapy…. When we get the results, we can show them that they have this mutation, here is what is FDA approved, and we go over the results.

Levy: What are your thoughts on how patients received this information? Do they know about the NGS?

Morganstein: Patients are usually overwhelmed. They have this hope for a pill, and they have a vague notion that there are some targetable mutations. But in all fairness, when I tell them that we’re going to run further testing on their tumor DNA, most individuals glaze over.

Levy: It’s exceedingly difficult for patients to understand. We said in the beginning, there are 10 different mutations and we use them in different lines of therapy. I can speak for myself, I’m struggling to keep up with all the new mutations. For patients it’s just enough to say, we need more testing, and if there’s something targetable, we’ll be able to treat you with an oral medication. Do you have molecular tumor boards?

Morganstein: We had a very successful molecular tumor board, prepandemic and postpandemic it fell apart a little bit. We present the vast majority of our cases in our thoracic tumor board and it tends to be the most molecularly valuable arena.

We would greatly benefit from a dedicated molecular board with a pathologist on site. It would be nice to have someone to bounce [these cases] off of monthly or bimonthly because sometimes you get weird mutations.

Benjamin P. Levy, MD, is the clinical director of Medical Oncology at Johns Hopkins Sidney Kimmel Cancer Center at Sibley Memorial Hospital and an associate professor of Oncology at Johns Hopkins School of Medicine in Baltimore, Maryland.

Bhuvana Ramkumar, MD, is a medical oncologist at Niagara Falls Memorial Medical Center at Roswell Park Comprehensive Cancer Center in Niagara Falls, New York.

Neil Morganstein, MD, is a medical oncologist at Overlook Medical Center at the Carol G. Simon Cancer Center in Summit, New Jersey.