Share this article on:

How I Use Tumor Genotyping in Patients with Non-Small Cell Lung Cancer

Camidge, D. Ross MD, PhD

doi: 10.1097/
Figure. D

Figure. D

For decades the only relevant distinction made in lung cancer was between small cell and “everything else”—all of which was lumped together and called non-small cell lung cancer (NSCLC). However, all of that changed with the advent of relevant molecular distinctions in NSCLC. While many molecular abnormalities are still having their significance as predictors of benefit from specific targeted therapies determined in clinical trials, markers for two such abnormalities—ALK gene rearrangements and EGFR mutations—have already entered routine clinical practice.

The following are key FAQs…

Back to Top | Article Outline

What is an ALK gene rearrangement?

The Anaplastic Lymphoma Kinase (ALK) gene is involved in various developmental processes and is then silenced in most adult tissues. In some cancers, a gene rearrangement places the promoter and a 5' dimerization motif from another gene (often, but not always EML4 in NSCLC) upstream of the kinase domain of ALK. This rearrangement then leads to re-expression and activation of ALK as an oncogenic signaling “warhead” within the resulting fusion protein.

Back to Top | Article Outline

Why does ALK testing matter?

Crizotinib, the only FDA-approved ALK inhibitor, has a molecularly specific license. Therefore one may not, and almost certainly should not, prescribe this drug unless the patient's tumor is known to be ALK positive. In patients with advanced NSCLC who are ALK positive the drug often produces rapid, dramatic, and prolonged control of their disease.

Back to Top | Article Outline

How do you get ALK testing done?

Many institutional and commercial labs now offer ALK testing. The testing is performed on standard formalin-fixed paraffin-embedded tumor tissue, and only one to two slides may be needed. As the ALK change is perceived to occur early in carcinogenesis, archival material is acceptable and fresh biopsies are usually not required, unless there is insufficient material remaining in the original sample on which to conduct the test.

Although crizotinib's license relates to patients being shown to be ALK positive using an FDA-approved test, and only one test (FISH testing using the Abbott Vysis Probe set) is currently approved, other tests, such as other FISH tests, immunohistochemistry, or RT-PCR, are used in some academic institutions and/or are commercially available. While the battles continue to rage about whose test is best, for an individual doctor only two issues are probably relevant:

First, despite the exact wording of the FDA license, there is no evidence to date of insurers being concerned about precisely which assay was used to determine ALK positivity, provided the patient is positive. Second, as each of the assays has a learning curve associated with it (partly explaining the discrepant results between assays across different institutions), an experienced, reputable provider should be used whenever possible.

Back to Top | Article Outline

What about ROS1 testing for crizotinib?

The one exception to the statement about crizotinib not being used in patients not known to be ALK positive relates to tumors driven by another target of the same drug—specifically, the ROS1 gene. This abnormality has been shown in small numbers of patients with advanced NSCLC to also predict for benefit from crizotinib.

ROS1 gene rearrangements are assessed in a manner similar to those of ALK rearrangements, and some institutional labs (the University of Colorado and Massachusetts General Hospital being two notable examples) and private providers now offer ROS1 testing commercially. Off-label use of crizotinib in ROS1 positive cases may be permissible depending on the insurer; alternatively, trials of ROS1 inhibitors that continue to accrue nationally may be considered in such patients.

Back to Top | Article Outline

What is an EGFR mutation?

EGFR mutations are short DNA sequence abnormalities that encode alterations in the kinase domain of the Epidermal Growth Factor Receptor. There are many different mutations that occur across exons 18–21 of the EGFR gene. The most common forms, L858R and exon 19 deletions, account for approximately 90 percent of detectable mutations and predict for rapid, dramatic, and durable clinical benefit from EGFR inhibitors such as erlotinib.

Back to Top | Article Outline

Why does EGFR mutation testing matter?

This can be debated. Unlike with crizotinib and ALK, EGFR mutation testing is certainly not required prior to prescribing erlotinib. Erlotinib does not have a molecularly specific license in the United States as its initial registration study demonstrated sufficient, if modest, benefit in patients who had not been molecularly preselected.

That said, it has subsequently become clear that the degree of benefit in known EGFR-mutant patients is significantly greater than in patients known to be EGFR-wildtype. However, it is also clear that in EGFR-mutant patients the benefit from EGFR inhibitors is similar regardless of the line of therapy. Consequently, a policy of ensuring all patients with advanced NSCLC get an EGFR inhibitor at some point has been suggested as a means of obviating the need for EGFR testing.

The downside of this approach is that we know there is a significant survival drop-off between lines of therapy. Therefore, EGFR testing offers a practical way to ensure that those who are going to derive maximum benefit from EGFR inhibitors are prioritized to receive them early as possible.

In contrast to erlotinib, afatinib—a second-generation EGFR inhibitor—conducted its registration study only in patients with an EGFR mutation. Consequently, afatinib is now seeking a mutation-specific license with the FDA and if that is successful, physicians will no doubt be even more actively encouraged to pursue EGFR mutation testing in the future.

Back to Top | Article Outline

How do you get EGFR testing done?

Many institutions and private providers offer EGFR mutation testing. While a positive test is likely to be controversial only if one of the rare (exon 18 or exon 20) mutations is uncovered (in which case specialist advice from a lung cancer expert may be in order), a negative test always raises concerns about whether it is a false negative or not.

Clearly, not all female Asian never-smokers with adenocarcinoma of the lung (the “classical” high-risk stereotype) will harbor an EGFR mutation, and most negatives, even in the face of all the known clinical risk factors, will still be true negatives. However, a low number or low percentage of tumor cells in the specimen or otherwise poor-quality tumor material all raise the possibility of a false negative, and re-biopsying the patient to get additional material to re-test may sometimes be appropriate.

Back to Top | Article Outline

Who should be tested for ALK and EGFR and when?

There are both medical and health economic answers to this question. Medically, it could be argued that everyone with advanced NSCLC should be tested for everything, unless a group with a zero percent chance of having these abnormalities can be reliably identified. Certainly, when the patient's insurance is good (or they are prepared to cover the costs themselves), then this is the policy I adopt, rationalizing that if it was my relative I would consider even a small percentage chance of positivity worthwhile to pursue.

But, of course, it's not just about medical thinking in the real world. Instead, aspects of clinical enrichment, deciding on groups to test or not test, based on such things as smoking status, histology, and the presence/absence of other known molecular abnormalities, are often considered. Such approaches—for example, conducting EGFR and ALK testing only in adenocarcinomas, or only in adenocarcinomas where patients have little or no smoking history—decreases both the cost to find each positive case and the absolute cost of the screening as fewer people are ultimately screened.

The downside of these approaches is that some patients will fall through the cracks because they do not match the clinical stereotypes for being EGFR and ALK positive. Ultimately the decision on whom to test has to be made based on the details of the risk factors of the patient and the individual payer, health insurer, or health system involved.

Perhaps the least controversial aspect of deciding whom to screen relates to stage. As EGFR and ALK inhibitors are not currently recommended for use in the adjuvant setting (although studies exploring this are underway), it is reasonable to hold off testing in patients with early stage cancer until a relapse occurs. Practically, as the majority of stage III cancers relapse fairly quickly, in addition to testing all stage IV NSCLC cases, I also test all of those with stage III disease as well.

Back to Top | Article Outline

How do you react when an EGFR or ALK test comes back positive in a patient with advanced NSCLC?

Apart from being very pleased, the first thing I do is explore where the patient is in his or her treatment journey. If they are between lines of therapy and have no evidence of active disease (e.g., a quiescent PET scan), I would consider holding off on using the relevant specific targeted therapy.

In contrast, for patients who still have obvious residual active disease, I would institute therapy immediately as a form of “maintenance” therapy. However, if they were currently on active chemotherapy, I would probably finish the full course and then either introduce or hold off on the targeted therapy, depending on the state of the disease at the time, as above.

Back to Top | Article Outline

What about KRAS?

This is a tough one. On the one hand there is no KRAS-specific licensed therapy, although several approaches are being explored in clinical trials. Therefore one could argue that in routine practice it is currently a useless test. Alternatively, as most of the different molecular markers in NSCLC function as if they are mutually exclusive (although rare exceptions do occur), discovering that a patient is KRAS positive may prompt the search for other oncogenic drivers to be called off.

Some groups have even suggested a “tiered” or sequential screening approach to minimize resource utilization, testing for KRAS first (as the most common driver abnormality in adenocarcinoma of the lung), and then only moving on to EGFR and then ALK testing if each successive result is negative.

However, just as with any enrichment policy, the cost savings of this approach need to be weighed against the human cost of both the EGFR- and ALK-positive cases that will be missed because they are exceptions to the “mutual exclusivity” rule and the clinical significance of the extra delay in getting a meaningful result back when tiered testing is used.

Back to Top | Article Outline




Add your comments in the blog version of this article——or email

Back to Top | Article Outline

What other molecular markers are out there? What will the future hold?

Many other potential predictive biomarkers for specific targeted therapies are in development in NSCLC, in both adenocarcinoma and squamous cancer. Some of them, like BRAF mutations, FGFR1 amplification, and RET gene rearrangements, represent classical genotypic abnormalities that can be detected using sequencing or FISH-based assays comparable to those described for EGFR and ALK changes.

While the variety of different assays available can become bewildering, it is important to note two things: (1) that beyond EGFR and ALK, all of these tests are relevant only in relation to clinical trials of specific agents at present; and (2) that in the next few years Next Generation Sequencing (NGS)—involving massive parallel sequencing of large parts of the cancer genome—will be able to document the presence of all of these abnormalities (and more) at costs similar to those associated with running a few individual assays.

Consequently, while we will still have to determine what to do with the data when we get it, debating what to test each individual for and in what order is going to become increasingly moot. That said, not all molecular drivers of cancer will leave footprints in the genome as obvious as mutations or gene rearrangements. Instead, even when NGS is commonplace, we may still need the help of additional, different types of assays—looking not at the DNA but at levels of key mRNAs or proteins, for example—if we hope to define the Achilles heel of every NSCLC in the future.

© 2013 Lippincott Williams & Wilkins, Inc.
Home  Clinical Resource Center
Current Issue       Search OT
Archives Get OT Enews
Blogs Email us!