Molecular testing is an invaluable resource in lung cancer, but the latest evidence shows that many US practitioners do not use it. Real world data in patients with stage IIIB or IV non-small cell lung cancer (NSCLC) found that only 54% were tested for epidermal growth factor receptor (EGFR) and 22% for anaplastic lymphoma kinase (ALK), ROS-1, and BRAF. Even when testing is done, it may not be enough to inform treatment decisions. A 2019 study at JAMA reported that only two thirds of patients with EGFR mutation or ALK rearrangements never receive an appropriate inhibitor during the course of their disease.

In comparison, Britain seems to have a lot to be proud of when it comes to molecular testing. Recently reported data from the United Kingdom National Lung Cancer Audit revealed that 83% of patients with advanced adenocarcinoma underwent testing EGFR, ALK, ROS-1, and programmed death ligand 1 (PD-L1). However, audit findings at the level of patients receiving appropriate targeted care and average survival indicate that the UK also has room to improve.

To find out what caused the differences in the UK and US approaches, Medscape contributor Dr. Jack West spoke with Dr. Sanjay Popat, professor of chest oncology at The Institute of Cancer Research and consultant in medical oncology at Royal Marsden Hospital in London, United Kingdom.

H. Jack West, MD: Let’s discuss what works and what hurts in both of our systems, because that is an important step to fix it.

Sanjay Popat, MD: In Britain, some things are a little better, some are similar to problems in the United States, and some are a little worse.

What we do very well is the penetration that testing has in the population. That’s clear in the 83% testing rate for biomarkers funded by the National Health Service (NHS). The data is for every patient who walks through the doors of a number of centers in 2017.

But when you look a little deeper, you see the problem. Of those identified EGFR positive mutants, only 75% received EGFR tyrosine kinase (TKI) inhibitors upfront. Even worse, only 58% ALKPositive patients receive ALK inhibitors. I feel this is very sad because this is the most effective medicine we have.

There must be a reason why this genotype was not translated into TKI recipes. Do cancer experts really take the results and find out the next step based on this limited amount of biomarkers? In general, I think they did, because to prescribe immunotherapy in the UK, you have to prove on the form that patients are the wild type to EGFR and ALK.

One possible cause is the turnaround time, which we have a problem with. Or maybe more complex.

West: Completion time is certainly a major challenge in the United States, where it routinely takes 3 weeks or more to complete network-based testing. Many patients want to be treated very quickly.

If molecular testing will take 3 to 4 weeks or more, even cancer experts who have a preference for it might feel that they might not be wrong in giving chemotherapy and immunotherapy. Or say you get the PD-L1 returned very quickly and then have to wait for molecular testing; it is very tempting to act at high PD-L1 levels and maybe just starting a single agent pembrolizumab.

Then there is a sense of competition in the United States. Patients who are not inclined to wait can leave your place of practice for an oncologist at the end of the road who will start treating them quickly. Many of my colleagues have expressed concern about having to convince patients that it is better to wait for the test results than to start treatment now, just to get things going.

Popat: We do not have the same commercial pressures that desperately need to start some form of systemic therapy, although it may not be clinically indicated. Patients have no hope of going to other practices, which may not have availability even if they do.

There are some cultural differences too. The NHS system does not allow you to prescribe pembrolizumab unless you really confirm that the patient is a wild type EGFR and ALK. That is the same thing for chemotherapy plus immunotherapy. You can prescribe it, but you have to prove that the patient is a wild type.

Another thing to consider is that in many centers, if you decide you want to provide some kind of systemic therapy, the patient may wait 2 weeks (if not longer) for the first cycle. That allows a little time buffer so that molecular results can appear.

Biopsy on Both Sides of the Atlantic

West: About 1 in 10 patients in the National Lung Cancer Audit report has insufficient tissue. In the United States, these patients can start chemotherapy plus temporary immunotherapy. It does not seem like a real possibility in the UK, because basically you are required to do whatever needs to be done to get molecular data. What is the impact of retesting?

Popat: Retest rate in UK audits is around 10% because initial specimens are inadequate. However, that possibility is too low. I suspect that many people who start chemo, with or without immunotherapy, are in situations where we do not have enough specimens to understand their true genotype. These patients may be biopsied at some stage in the future, which is also a challenge to do after they start receiving chemotherapy.

We really want to move towards liquid biopsy. I would love to be in a position where we have the next generation of DNA sequencing (ctDNA NGS) circulation in advance. Unfortunately, Britain lags behind this, given the large costs associated with NGS ctDNA. Several commercial providers have signed schemes with each institution to allow this approach for the selection of biomarker patients for trial registration, but certainly not close to standards.

West: Liquid biopsy is available in the United States, but is not consistently used and replaced.

Another problem that we sometimes face here is that genetic testing reports can be uploaded to random media tabs from challenging and expansive electronic medical records, never seen by the human eye. How likely is it in the UK system?

Popat: In general, our system works very well in this regard. Stopping hard is basically not prescribing chemotherapy plus immunotherapy or immunotherapy alone unless you know what you are seeing. It really forces these results to be seen and verified.

Most pathology and molecular diagnostic tests occur in established tissues with oncologists, so the results get feedback from both oncologists and clinical nurse specialists. Nurse specialists have a role in instigating molecular testing, ensuring results have reached the laboratory and returned, and marking one of the positive results to the oncologist. This means there is a much smaller chance of results just drifting off to random pages in electronic medical records.

Room for Improvement in Both Systems

West: Success reported in the National Lung Cancer Audit is accompanied by quite serious results. The average survival for the wider population is only 7 months, and only 1 year for those who have EGFR mutation. That’s a lot less than I expected to see a population enriched with a targeted therapy. What do you think contributed to that?

Popat: I fully agree and, to be honest, feel very shameful. I think you must remember, however, that this is the whole population. This is not only suitable for patients you normally consider enrolling in trials, but also for those diagnosed as emergencies, sick, or unhealthy. You see the entire population, not just a crème de la crème part

Because of this, the result data is far worse than we expected. I think that can be partly explained by many things.

For example, we still provide first generation inhibitors on EGFR settings and must hunt for testing T790M before giving osimertinib. T790M testing has a problem. I am not sure that we often actually follow up on a negative T790M test with an appropriate biopsy, as the patient slowly develops.

I am also worried that the patients themselves are not interested in undergoing additional biopsies. There is always a fear of having chemotherapy, even though we know it is a very effective rescue treatment for the T790M-negative population. I want to know more about the patient’s perspective here.

There is a lot of work to be done to optimally manage oncogenic addiction. We don’t do very well oncogen oligoprogress here, because stereotactic body radiotherapy is not routinely funded. These patients must undergo trials. Radioturgic stereotactic provisions for patients with central nervous system (CNS) development vary greatly. CNS monitoring in patients with oncogenic addiction is still poor.

This all contributes to suboptimal results. We really need to realize that patients can do very well if treated aggressively, and obtaining a TKI is an absolutely essential part of this. And, clearly, acknowledging that this is a patient who is addicted to oncogenes is initially the key.

West: The message that was brought home to me was that there was clearly remarkable new progress in our care and understanding of how cancers adapted to them. It takes a lot of effort to apply that knowledge and educate the wider community to realize this promise. Even with some of the successes reported in the United Kingdom, none of us has broken this code into a great system of work.

H. Jack West, MD, associate clinical professor and executive director of employer services at the City of Hope Comprehensive Cancer Center in Duarte, California, regularly comments on lung cancer for Medscape. West serves as a web editor for JAMA ONCOLOGY, edit and write several sections about lung cancer for UpToDate, and leads a variety of continuing education programs and other educational programs, including hosting audio podcasts West wind.

Sanjay Popat, PhD, FRCP, is a professor of thoracic oncology at The Institute of Cancer Research and a consultant in medical oncology at the Royal Marsden Hospital in London, England. His research interests include identifying DNA variants that influence the development of thoracic tumors, identifying biomarkers that can predict therapeutic effects, and developing new strategies for treating thoracic tumors through clinical trials.

Follow Medscape at Facebook, Indonesia, Instagram, and Youtube