Battling Lung Cancer in the Tobacco Belt

I direct the lung cancer program at East Carolina University, which is not only in the Bible Belt but also the Tobacco Belt. And that means we see a lot of lung cancer – on average, 8–10 new cases every week. Clearly, we need to make good treatment decisions fast. For the past two years, we’ve been using liquid biopsy – specifically, droplet digital PCR (ddPCR), which lets us zero in on the actionable mutations in lung cancer (EGFR, ALK, ROS1, BRAF, and KRAS). Because ddPCR looks only at the “hotspots,” it has a quick turnaround time – about three days. The speed dramatically affects decision-making – to the point where I would describe it as the first step toward truly personalized oncology.

We can now identify a cancer’s mutations and respond with appropriate therapy without the issues inherent in solid tissue biopsy – not only the delays, but also the intratumoral heterogeneity that is often missed when sampling only one or a few tumor sites. In fact, liquid biopsy has a particular advantage in that it captures the most aggressive clone; because it is the site of greatest apoptosis, it sheds the most cell-free DNA (cfDNA). When we began using ddPCR for cancer testing, we began to see mutations we never expected – and, in some cases, they completely changed our treatment decisions. And that’s when we began to ask, “How did we ever make decisions without liquid biopsy?”

Breaking barriers and learning lessons

Liquid biopsy surely has great advantages, but we also face integration hurdles. One huge stumbling block is that the technology is hard to understand for those in other fields. Medical oncologists think biologically and therapeutically, and not from a technology standpoint, so there’s been a lot of confusion. We have 12 medical oncology fellows at our institution, and I try to teach them that there are really two types of liquid biopsy. There are the ones that are going to give you information on actionable and/or druggable hotspot mutations – and the more limited the search you make, the quicker you get results. Whereas the other option – next generation sequencing (NGS) – can be endless: it can be 92 genes, or 582, or even the whole exome. You can get a lot of information, but you’re not going to act on 99 percent of it. I think a lot of medical oncologists have used NGS and received a 12-page report and thought, “this is never going to be helpful!” By using the right method, you can get useful information that will help guide treatment.

Liquid biopsy is already teaching us a great deal. Cancer is a lot more genetically diverse than we suspected. You’re not “supposed” to have co-occurring EGFR and KRAS mutations, but we’ve seen them. You’re not “supposed” to have an EGFR T790M mutation in squamous cell lung cancer, but we’ve seen that. We’ve learned that not all lung cancers respond to platinum – for instance, those with ALK mutations, or KRAS G12C. These are aggressive cancers that require rapid response, and understanding their genetic landscapes gives us the ability to use the best treatment right from the start. And when you consider a variety of liquid biopsies – that is, not only those from blood, but also those from urine, cerebrospinal fluid, or any other source – you capture an even broader range of genetic alterations, because not every compartment has the same mutations.

A good example of liquid biopsy’s utility is “chemoimmune” therapy (pemetrexed, carboplatin and pembrolizumab) for non-squamous cell lung cancer. It’s a promising approach – provided the tumor doesn’t have an EGFR or ALK mutation. With ddPCR, we know within three days whether or not those mutations are present in our fluid of choice, which lets us determine the best treatment option for a rapid and durable response. For me, that really crystallized the benefits of having a quicker test – and it helps in conversations with patients, too, because we can reassure them that the results of the treatment will really be worth the downsides.

Tailoring treatments

I view solid tissue and liquid biopsy as complementary techniques – each makes up for the other’s failings. When biopsying tumors, you may not always get sufficient tissue, especially in lung cancer. Liquid biopsy material is easy to access, so you can perform deeper analyses. In our program, we perform both the “hotspot” actionable mutation test and NGS (of tissue samples); the latter takes longer, but leaves us with information we can continue to use as precision medicine and immunotherapy evolve.

The United States has an additional, unique problem with tissue biopsies. If a hospitalized patient is under Medicare – as most over the age of 65 are – the material can’t be sent out for 14 days. That means you have to wait until either two weeks have passed or the patient is discharged, and then the same amount of time again even for initial results. That’s a full month of delays in providing treatment, and not every lung cancer patient has that time to spare. The problem is only exacerbated when the disease develops treatment resistance; re-biopsying a lung tumor is a difficult process at the best of times.

On the other hand, tissue provides us with a wealth of information that becomes increasingly valuable. From a pathology standpoint, I know the most important lung cancer question has always been, “Is it non-small cell or small cell?” Once upon a time, that was all the pathologist had to do. Now, with an adenocarcinoma, there’s this never-ending algorithm – CK7, CK20, TTF1, p40, p63… to the point where the tissue sample is exhausted on tests with very little impact on therapy. We need to break away from that pattern of algorithms and focus on the data that will actually affect treatment decisions.

Historically, we clinical oncologists have referred to certain patients as “exceptional responders,” because therapy has been highly effective even with advanced metastatic disease. Now, we’re finding out the truth – they’re not “exceptional responders,” they’ve just received the right treatment appropriate to their particular disease.

It is, of course, possible to use a liquid sample for NGS – but, at the moment, tissue sequencing remains superior. Ultimately, I think liquid biopsy with cfDNA will replace other monitoring methods. For instance, patients won’t need regular imaging after surgery. If the cfDNA no longer contains the cancer-driving mutation, I’d say there’s at least a 90 percent likelihood that the cancer will not return. However, if you see the mutation beginning to recirculate, you know something is brewing – so you can alter the treatment accordingly.

Enhancing and extending lives

As a lung cancer specialist, I’ve also seen the impact that additional information can have on the experience of patients. Many people have seen family members and friends go through chemotherapy, and experience some really difficult side effects – and then die anyway. Being able to explain to patients that this is a targeted therapy with a high chance of extending their progression-free survival by several years is a radically different experience.

It’s also helping to combat some of the stigma surrounding lung cancer – “You did this to yourself, because you smoked, and there’s nothing much we can do.” I think we still see this attitude even with some pulmonologists and oncologists. But people are living longer, and quantity and quality of life are being extended, which I think is helping to lessen the blame game and improve the outlook for our patients.

I’ve already seen some great results. One individual, a woman who was 70 but physiologically very youthful, had received radiation therapy for Hodgkin’s disease, then later developed a widely metastatic lung cancer – it had spread to her pleura, adrenal glands, and bones. And her diagnostic tissue was just cytology from the pleural fluid, so there was insufficient tissue to get a mutational analysis. But her liquid biopsy came back with a KRAS G12C mutation, which is not thought to be very platinum-sensitive. And so, although the standard therapy would have been a platinum doublet plus or minus bevacizumab or anti-VEGF-A, we just gave her a non-platinum doublet. The response was excellent. We repeated the liquid biopsy and found that the KRAS mutation was persisting, but also that there was an EGFR T790M mutation, which are thought to be mutually exclusive. So then we went on to immune checkpoint blockade, and transformed quality of life for the patient – she went from being breathless and in pain to enjoying a two-week African safari with her husband. Her cancer is under control, and she is still living a very full life.

Another woman, a mother to two young girls, had extensive brain metastasis – it was an EGFR exon 19 mutation, and she also had a T790M mutation, which is immunogenic (the other EGFRs are not). She went on to have immune therapy and, as of her last imaging, she is completely cancer free and able to live her life. Throughout her treatment, liquid biopsy told us exactly what was going on, and helped us to choose the treatment with the best chance of success.

As a medical oncologist, every patient visit is truly a life and death situation. It’s so important to be able to look your patient in the eye and tell them what the best treatment decision is, based on all the information you have, and to know you did everything possible for them. Liquid biopsy is helping us to do that.

A fluid future

If you are looking to implement liquid biopsy in your own practice, I have two pieces of advice. First, you need to make sure your colleagues understand what this information can do for them. In our own multidisciplinary team, we have four thoracic surgeons, two radiation oncologists, and a dedicated pulmonologist; I make sure everyone sees the results we’re getting, so that they understand we’re ordering these tests for a reason – and they are impacting therapy. Your team members must understand why they need this information.

Second, there is the question of in-house versus partnering with another organization to do the testing. Some pathologists would rather do these tests in-house, but I would argue that partnering with industry is also a good solution. After all, the technology moves so fast and becomes outdated so quickly. In our practice, we chose to play to our strengths – we are innovative and aggressive with our clinical treatment, but we didn’t want to duplicate the tools we can access by partnering with a company who can provide the information for us. I think a lot of pathologists resist the idea of giving up their tissue for testing elsewhere – but liquid biopsy solves that issue.

Liquid biopsy will lead to big changes in how we monitor cancer status in patients. We’re already seeing this in chronic myeloid leukemia using BCR-ABL quantitative PCR, and there is more and more data emerging showing that after colorectal cancer surgery, cfDNA can tell us which patients are going to see cancer recurrence (1)(2). Such information can guide the decision to treat with adjuvant chemotherapy and give us an immediate answer to that crucial question – “did the surgeon get it all?”

For patients who don’t have a diagnosed cancer, liquid biopsy is also poised to change how we monitor health. I always tell our fellows that they ought to go through an MRI at least once – no one enjoys spending 45 minutes in a tube! If we can screen patients using a blood draw, we can save them from X-rays and MRIs.

I suspect things are going to move much more quickly than expected – last year, liquid biopsies were validated at ASCO, and this year they’ve extended to diagnostic, prognostic, surveillance, resistance pathways and recurrence across many tumor types (3)(4)(5)(6). I think this area is going to expand rapidly, and it won’t be long before many of our questions are being answered with a simple blood test. Looking a couple of steps ahead, we know that cell-free DNA also appears in urine, so one day you may be able to simply provide a urine sample to find out if you’re cancer-free or not. The technology is already here – and the possibilities for the future are amazing.