Transcript:John R. Gosney, MD, FRCPath: Stage III non—small cell lung cancer is in the middle of the spectrum. In stage I and II disease, we hope that treatment using surgery is the ideal scenario, which of course has curative intent. With stage IV disease, which is essentially disseminated, the approach is different. And here we’re really talking about control rather than cure. Stage III is in the middle. And in the third of patients who will fall into that category, really the intent is curative, but it depends rather more on radiotherapy and/or chemotherapy.

When the first targeted therapies appeared about a decade ago, essentially TKIs [tyrosine kinase inhibitors] against tumors with EGFR mutations, they were concentrated on disseminated disease, and that tends to have been the case over the past decade. The extension of targeted therapy using durvalumab into stage III disease therefore has brought something of a problem. This is slightly new. And because this is new, there’s a fear out there, which is justified that patients in the middle of the spectrum with stage III disease are not being routinely tested for the appropriate genomic drivers and for PD-L1 [programmed death-ligand 1] expression. Hence, the fear that they might essentially slip through the net.

And it’s very important to test them because with the advent of targeted therapies—essentially immune modulators—to patients with intermediate stages of disease, it’s very important that this testing is done. My practice has always been to test all patients stage II and beyond at diagnosis for genomic drivers and for PD-L1 expression. So automatically patients with stage III disease will be included in that cohort. But in many centers—especially where the decision to test is not made until the tumor board, the multidisciplinary team, meets—that is not the case, and that’s where the problem lies. But absolutely, I think all patients with stage III disease should indeed be tested for the usual drivers and for PD-L1 expression.

The profiling of non—small cell lung cancer started really a decade ago with the advent of tyrosine kinase inhibitors active against tumors with EGFR mutations. And then we moved on to looking for ONC gene rearrangements, and that was maybe 3 or 4 years after that. And the most recent development, which has a major one, has been in the need to test for PD-L1 expression to guide the use of the new generation of immune-modulating drugs.

And the testing for EGFR mutations and ALK rearrangements, of course, is essentially genomic testing; whereas the testing for PD-L1 expression to guard the use of immune modulators depends on immunochemistry and is rather more subjective. So the nature of the test has changed over the passage of time. And most recently we’ve added the need to test for ROS1 gene rearrangement on top of that. So there’s really been a development in terms of the number of tests that we need to do and their nature over the past 10 years. And there’s no sign that this is ending, so the momentum continues.

For our profiling of non—small cell lung cancer, we prefer single-gene testing. And that’s essentially because of its flexibility. We’re driven practically by the availability of drugs and their reimbursement, and we use that as a guide to what we will test routinely, which essentially at the moment is EGFR mutations, ALK and ROS1 rearrangements, and PD-L1 expression.

And we do the genomic test using single-gene testing and immunochemistry for PD-L1 expression. And this means that, because of the flexibility and the speed with which we can operate, we can turn around the full panel of profiling within 6 days, and we do this routinely. Now we will be moving to NGS [next-generation sequencing] testing because that carries advantages, of course. You get a lot more information, and you get information on different genomic drivers. But of course, it takes more time. There’s a need for batching, and it requires more tissue. So it does carry disadvantages.

Ultimately, of course, because single-gene testing with its flexibility can produce a complete profile, usually within 5 or 6 days, which is our usual turnaround, that’s a huge advantage with treatment planning because it means that all the necessary information is available to the oncologist very early on in the course of the patient’s management, and that’s very important.

Transcript Edited for Clarity