Blood-Based Assays Are Addition, But Not Replacement for Lung Tissue Biopsies

Leora Horn, MD, MSc

The boom of blood-based biomarkers has led to a turning point in clinical practice for physicians treating patients with non—small cell lung cancer (NSCLC). While tissue biopsies remain the standard approach, plasma assays—if positive—can direct patients to a first-line targeted treatment quicker.

“Blood-based testing does have a role in patients with NSCLC,” said Leora Horn, MD, MSc. “The blood can be potentially used as a surrogate for markers for directing for therapy. “But if blood testing is negative, it is not enough to say that a patient is not positive. Those patients do need to go on to get a biopsy.”

OncLive: What did you discuss in your lecture at the meeting?

In an interview during the 2017 OncLive^® State of the Science Summit™ on Advanced Non—Small Cell Lung Cancer, Horn, an associate professor of medicine (hematology and oncology), assistant director of the Educator Development Program, clinical director of the Thoracic Oncology Program, and medical oncologist at Vanderbilt-Ingram Cancer Center, discussed the utility of blood-based biomarkers, how they have transformed treatment approaches for patients with NSCLC, and the ongoing work currently being done with them.Horn: The talk was about biomarkers and, specifically, the use of blood-based biomarkers in lung cancer. As you know, last year, the FDA approved EGFR testing via blood for patients with stage IV NSCLC that was adenocarcinoma. We talked about the role of a biomarker, both [as] a predictive [and] a prognostic factor. Predictive is the more important one, which can tell you if a patient will likely or not likely benefit from a specific therapy.

What impact has this type of assay had on clinical practice?

Also, for blood-based biomarkers, they are not perfect. If a patient has stage IV lung cancer and you send for a blood-based test and it comes back as negative, that is not enough to say we shouldn’t treat a patient with a certain therapy. That patient should go on to get a tissue biopsy, which would allow you to get better results. Blood-based biomarkers are only about 80% sensitive; that means that, in 20% of patients, we are missing a correct result. We also talked about the fact that PD-L1 testing has become standard of care for first-line therapy for patients with lung cancer, but it cannot be done by a blood [biomarker] yet. That is likely to be something that emerges in the near future.It is a nice option for patients who have had maybe 1 or 2 biopsies that have been unsuccessful. I’ll discuss it with patients and tell them, “If the blood is positive, it gives you an answer. If it’s negative, it does not give you an answer.” For those patients who maybe had a couple of unsuccessful biopsies where we didn’t get enough tissue for molecular testing, a blood-based test can hopefully provide that information for us in helping select therapy.

They are also useful for patients who are on therapy and look like they are developing resistance. You can send the blood-based testing and, based on that test, you can get a result. The best result is with EGFR inhibitors. If patients on a first- or second-generation EGFR inhibitor look like they are having progression of disease on a scan, you can send for a blood-based test. If it is T790M-positive, we can use that result to treat them with the third-generation drug approved, osimertinib (Tagrisso).

What trials related to blood-based biomarkers have the potential to change practice?

If they are T790M-negative via blood, those patients need to go on to get a tissue biopsy because sometimes blood will miss a correct result. For blood-based testing as well, you need a critical volume of disease. Therefore, for patients who just have disease confined to the chest, they are less likely to be able to pick up the result by blood. However, if they have disease in the chest and cancer that has spread to the bone, liver, or brain, they are more likely to be able to pick up the accurate results via blood. A lot of the trials that are being done at the moment are collecting blood, but not necessarily as a primary endpoint of the study. You often hear the primary endpoint of the study as survival or response. There are secondary endpoint correlatives, which are often collecting blood, and, especially in the TKI treatment setting, there are studies that are starting to emerge. For example, what if you detect that a patient is developing resistance via blood, but their computed tomography (CT) scan looks fine. Should you switch or continue therapy? There are studies starting to emerge that are looking at that question.

What challenges are related to these blood-based biomarkers?

We sometimes see some interesting blood-based testing going on in small cell lung cancer (SCLC). Dr Christine Lovly has been leading that effort at Vanderbilt-Ingram Cancer Center, where she is looking at blood in different mutations of patients with SCLC, and you can actually detect early resistance or recurrence. There are also studies ongoing in the adjuvant setting who are treated with curative intent. As part of their monitoring, they are not only doing CT scans but looking at blood testing. Can you potentially detect recurrence in blood before you see something on a scan? Those are all in the context of clinical trials; it’s not ready for primetime use yet. There are multiple different platforms out there. There is fairly good concordance between blood and tissue, but it’s not perfect. With the PD-1/PD-L1 inhibitors, as a reality for the treatment of patients with lung cancer, it would be nice to be able to test PD-L1 in the blood as well. This is because even if patients are molecularly negative, you still need to know their PD-L1 status prior to selecting therapy. It would be nice if the blood-based testing could not only predict who is going to respond, but also predict who is not going to respond—especially in the realm of immunotherapy. There is work being done in that area.

The other interesting part will come with some of the ALK inhibitors, where the mechanism of resistance is not as clean as with EGFR inhibitors and the different second- and third-generation ALK inhibitors target specific resistant mutations. If we could detect them via blood, we can potentially prevent patients from having 3 or 4 biopsies. There are multiple ALK inhibitors that are either in clinical trials or already FDA approved, and they can help us sequence therapies for these patients.

Along that line for patients with ALK-positive lung cancer, we know that about 50% of patients at the time of resistance will have central nervous system metastases. If we could potentially detect resistance earlier on and we switch therapy, could we prevent those patients from developing brain metastases?