Nathan Pennell, MD, PhD, discusses the emerging widespread role of liquid biopsies in lung cancer and how to best utilize the approach in clinical practice.
Nathan Pennell, MD, PhD
Not all patients with advanced non—small cell lung cancer (NSCLC) are recommended for genomic testing when treated in a community setting, and are therefore not being appropriately tested for potential molecular drivers of their disease, explained Nathan Pennell, MD, PhD.
In a retrospective study of 1203 patients with stage IIIb or IV NSCLC treated since January 1, 2017, records were manually reviewed to find evidence of molecular testing and targeted treatment. Between 11% and 54% of the identified patients had any evidence of testing for at least 1 of 7 genes recommended for testing by the National Comprehensive Cancer Network at the time: EGFR, ALK, ROS1, BRAF, RET, MET, and ERBB2 (HER2). The genes EGFR, ALK, ROS1, and BRAF can be targeted with FDA-approved treatment options while the rest have investigational agents being tested in ongoing studies.
Overall, 54% of patients received testing for EGFR mutations and the rates decreased for the rest of the genes.1 Testing for all 4 genes with FDA-approved treatment options were given to 22% of patients. Furthermore, 48% of patients received testing for PD-L1 expression.
However, liquid biopsies have the potential to increase the number of patients who receive molecular testing with a quicker turnaround time and less-invasive approach than standard tissue biopsies, according to Pennell.
“The benefit is you can test a larger number of people, you can get all the testing that's necessary, and it can be done fast,” said Pennell. “You also don't need to do repeat tissue biopsies, making it very low risk.”
Additionally, results of the Noninvasive versus Invasive Lung Evaluation (NILE) study showed that a liquid biopsy test detected all of the guideline-recommended biomarkers in newly diagnosed patients with metastatic NSCLC at a similar rate but faster turnaround time compared with tissue genotyping. The 73-gene next-generation sequencing panel Guardant360 increased biomarker detection rate by 48%, from 60 patients identified with ≥1 guideline-recommended biomarker using tissue-based tests alone to 77 patients with liquid biopsy (21.3% vs 27.3%; P <.0001). Moreover, the time it took to receive results after ordering was 9 days for liquid biopsy versus 15 days for tissue biopsy.2
In an interview with OncLive®, Pennell, director of the Lung Cancer Medical Oncology Program at the Cleveland Clinic Taussig Cancer Institute, discussed the emerging widespread role of liquid biopsies in lung cancer and how to best utilize the approach in clinical practice.
OncLive: Why is molecular testing underutilized in practice?
Pennell: It's a number of things. There is certainly an aspect of education and need for people to understand that testing is necessary, but most biomarker testing for cancer is not done at the oncologist level; rather, it's done at the pathologist level. For example, if someone does a biopsy and is diagnosed with breast adenocarcinoma, then the pathologists automatically order a panel of important biomarker tests, such as estrogen receptor, progesterone receptor, and HER2 expression. With lung cancer and a lot of cancers, on the other hand, [biomarker testing] is only done when requested by an oncologist. However, even when oncologist makes a request, the kind of testing done is determined by the hospital's pathology department, whether they do their own internal testing or whether they have contracted an outside vendor to get it done.
There is a wide variety of how [biomarker testing] is done. There are some places, such as Cleveland Clinic, where it's done reflexively at the time of diagnosis. However, for most places, there has to be a request and then it will be sent out and done from a commercial company that does broad testing or a series of hotspot tests—where they do single gene tests in a combination. That same process could also be done internally, either with a broader, next-generation sequencing test that covers all targets of interest, or a series of single-gene tests. It's a big mess of how testing is done, which is not a good way to systematically design something that's important.
If you look at the most common genetic tests for EGFR mutations, we do a respectable job at testing for that. Probably between 80% and 90% of eligible patients get tested for EGFR mutations in the United States, depending on what study you look at. It's been a decade since that has been recommended, making it in everyone's mind. Additionally, for more than 8 years, we have had treatments for ALK-positive lung cancer, making its prevalence close behind at 70% to 80% of patients getting tested for ALK.
Beyond those 2 biomarker tests, there is PD-L1 testing. Other genetic markers, such as ROS1, BRAF, MET, RET, KRAS, HER2, and NTRK, are done at a much lower rate, either because they're not ordered or because the [tissue] is used up. After the first 2 single-gene tests, there is a major drop off in the ability to do any further testing because the [there is no more available tissue]. For patients with metastatic lung cancer, many weeks, or a month, have gone by when they get the results of this test and a lot of them may have proceeded to other treatments, such as immunotherapy, because they may not feel they can wait for the results.
Additionally, they may not have tissue available, and they may not be able to be biopsied in an area that's easily accessible. Even if they have [tissue], there may not be enough to do all the testing and the patient might require another biopsy. The patient may not want to go through that or the doctor may not feel it's safe or there is enough time. There are lots of different process-related reasons for why testing is inadequate.
How can liquid biopsies improve these testing rates?
The major advantage of plasma testing or liquid biopsy, which is a blood test, is that you can get blood from every single patient. You don't have to worry about where the biopsy was done, if there's enough material from the biopsy, or if it was done properly. The major commercial vendors for plasma testing do a comprehensive multiplexed test that covers all of the targets of interest. You can get testing on every patient and when you order the testing, you can get all of the targets of interest upfront. That immediately takes away a lot of the issues with it. That's something that you can automate: to make it standard to test everyone [upon diagnosis].
The other big advantage with liquid biopsy is the timing because, with preparation of material in order to perform the test, [results take a lot] longer [to come back] with tissue. First, you have to locate the tissue and cut sections. Then, you either have to process them to extract the DNA and RNA or send it to another company. If you send it to another company, you have to consider shipping and logistical processing times before they can extract the material and you can get those results.
With something like plasma testing, the median time for testing results in the NILE trial was 9 days, which was significantly faster than the turnaround time for tissue. Those are the big advantages and has the potential to dramatically increase the number of people who get tested for the appropriate genetic markers for lung cancer.
How else do liquid biopsies compare with standard tissue biopsies?
People have spent a lot of focus on the disadvantages of the liquid biopsy, including the sensitivity. If you're extracting DNA from an actual tumor and you have a piece of the tumor, you're getting DNA that's coming from that cancer. If you test it, you have a high likelihood that you'll accurately be able to assess if there's a mutation or a gene fusion in that cancer because you're testing the cancer itself.
If you're looking at circulating-free DNA in the blood, then you're relying on some of the DNA from the cancer cells, which is not in the blood and is usually in the lung, bone, or wherever the tumor is, to leak into the bloodstream and become detectable amid a huge amount of DNA that is not tumor DNA; it's normal DNA from the person. The technology has gotten phenomenally better at being able to detect extremely low amounts of circulating tumor DNA and being able to filter it out to detect the difference between the tumor and normal DNA from a person.
However, not every cancer is going to be detectable in that way. Some cancers don't shed DNA as much as others. Some patients who have a relatively low amount of cancer and a low tumor burden may not have as much detectable DNA. The sensitivity is always going to be a relative disadvantage of doing liquid biopsies compared with an actual tumor biopsy. If you look at comparisons between tumor biopsies and plasma biopsies, for example, in the NILE study, you probably get somewhere in the range of about 80% sensitivity of mutations and other genetic alterations that are known to be in the tumor from a tumor biopsy that are detectable in the blood, meaning you may miss 20% of actual findings that are there by solely doing a blood test.
What targets can be identified via liquid biopsy?
In 2019, there are 4 must-have biomarker tests for genetic testing in lung cancer. You have to test for EGFR, ALK, ROS1, and BRAF V600E in every patient with lung adenocarcinoma and nonsquamous non—small cell lung cancer. Then, there are a number of tests that should be done on everyone but we don’t yet have FDA-approved treatments for—including RET and MET exon 14 skipping mutations. Therefore, you could argue that they are not necessary, but getting very close [to being necessary].
Similarly, the FDA is reviewing data from drugs for approval in this space. HER2 activating mutations are still investigational, but there are promising drugs and trials. KRAS mutations, historically, have not been targetable, but that's changing with multiple drugs now in early-phase trials. KRAS is also very valuable because it’s very common, and these drivers are mutually exclusive. If you find it, there likely are not any other targetable oncogenes present.
There are also NTRK fusions, which are extremely rare but have extremely effective treatments if you identify them. When you're doing a broad multiplex test, you can test for all of these at around the same cost as testing for any of them [individually].
The NILE trial looked into the impact of liquid biopsies in lung cancer. Could you highlight the findings from this study?
This is a very simple test. They took 282 patients with lung cancer who were going to undergo standard tissue testing. All of them had their blood drawn and sent for plasma testing at the same time. In the end, they described the success rate for identifying guideline-recommended biomarkers with the plasma testing compared with the tissue testing. This was a nice, real-world study because they were using a commercially available, FDA-approved plasma test and whatever was done by local testing centers for tissue testing, which is a very realistic real-world way of doing it.
In one aspect, this is a glaring way of illustrating that with many biomarkers to test, using a combination of single-gene tests is completely ineffective. There's no practical way to test for 8 different biomarkers using single-gene tests. If you just looked at EGFR and ALK, for example, more than 80% of patients were tested, which is pretty close to what literature has described regarding patients who were tested for EGFR and ALK nationwide. However, even 80% for ALK 83% for EGFR, compared with 95% of patients getting tested with plasma testing, shows that liquid biopsy captures a much larger group of eligible patients and ensures they have at least some testing done.
The disadvantage [of liquid biopsy] is the sensitivity. Although a relatively small number of patients had all of the tissue testing done, the plasma testing results in patients who had tissue testing showed they had about an 80% sensitivity. In other words, 80% of the genetic alterations that were found in tissue were also found in the plasma. The flipside is if you had done plasma and not tissue testing, you probably would have missed about 20% of those.
However, since only 18% of people got all the tissue testing done, you probably picked up a lot more actionable alterations by doing the plasma testing than you would have if you had solely done tissue testing. In a way, they are complementary. The perfect solution would be to do tissue testing and get it done correctly on 100% of patients. However, since that's impractical currently, plasma testing has a huge role to play in improving the number of people getting testing.
Is there any pushback regarding the payment aspect liquid biopsies?
A major outstanding question when it comes to plasma testing is reimbursement for paying for these tests. There are FDA-approved plasma tests. Hopefully, payers understand the importance of the role of testing and are covering this. We as a field have to be careful about routinely recommending double testing using plasma and tissue for every patient. We have to set up an algorithm that optimally try to identify [optimal testing] for patients without doing wasteful testing. I'm encouraged by what I've seen, at least in the last couple of years, with payer coverage for plasma testing when it's appropriate.
What steps can we take as a community to make liquid biopsies more widespread?
[We must work on] education and letting people know about plasma testing, including the advantages and disadvantages. I do not want to put out the message that, “Plasma testing should replace tissue testing universally, that it's every bit as good, and that the result of plasma testing is adequate if it shows you don't have an actionable result.” If you have good tissue testing and you're going to get the [results from plasma too], then that's adequate. If you have any questions about your tissue testing, then doing plasma testing upfront makes perfect sense. However, if you get a negative result and there are no actionable results from your plasma test, you have to follow up with tissue testing because you may have missed some [alterations].