A couple of weeks ago, the American Society of Clinical Oncology (ASCO) released a set of revised guidelines for stage IV NSCLC. While grounded in good evidence, they are striking for the contrast they offer with the recommended strategy from the National Comprehensive Cancer Network (NCCN) guidelines for this population. Most notably, the difference is related to molecular testing and subsequent individualized treatment based on this. As many of you who follow the discussions here have observed, the question of who to test and what to test for is currently among the most timely, evolving, and still somewhat controversial questions in the practice of lung cancer today. This is especially true in the setting of the approval of the ALK inhibitor XALKORI (crizotinib) by the FDA just a few weeks ago, since this agent is only going to be made commercially available to the 4-5% of patients with NSCLC who test positive for an ALK rearrangement at the one designated central lab approved along by the FDA along with the drug. So here, the approval of the drug is actually predicated on molecular testing for patients who would hope to receive this agent.

The NCCN guidelines, released earlier this year, recommend testing for the EGFR mutation all patients with advanced/metastatic non-squamous NSCLC, with the subsequent recommendation of the EGFR tyrosine kinase inhibitor (TKI) Tarceva (erlotinib) as first line treatment for patients with a demonstrated EGFR mutation. While ALK testing and crizotinib are not mentioned, this is not surprising, since neither was commercially available when the last version of their guidelines was released. Frankly, I found these guidelines to be just a little "forward-thinking" in calling for anyone with non-squamous NSCLC to be tested, since people with a very significant smoking history would have a quite low, even if non-zero, chance of having an EGFR mutation. Moreover, while many studies have now shown that people with an EGFR mutation have a consistently longer progression-free survival and response rate with first line EGFR TKI compared with first line chemo, there isn't clear evidence that survival is worse if people with an EGFR mutation receive an EGFR TKI after first line (in other words, A + B = B + A). Since EGFR inhibitors are readily available regardless of smoking status or NSCLC subtype (in the US, at least), it's not quite clear that it's critical to test everyone immediately after diagnosis. They key may just be that everyone who could possibly have an EGFR mutation gets their opportunity with an EGFR TKI, sooner or later. That said, the ALK story adds a level of complexity now, since XALKORI isn't an option if you aren't tested. But that's not yet addressed in any formal guidelines.

ASCO guidelines offer a very different interpretation. Specifically, they note that collecting tumor tissue can be a practical challenge and that molecular testing is currently considered as investigational rahter than the standard of care. Moreover, even patients who are found to have an activating EGFR mutation, an EGFR TKI is considered an acceptable alternative to standard chemotherapy but not a definite superior choice, since at this point there hasn't been a significant survival benefit to first line EGFR TKI demonstrated in such patients. (And again, not surprisingly, no mention of ALK testing or XALKORI.)

My interpretation of these two sets of guidelines, with their very different, rather polarized views, is that most oncologists will (and probably should) fall within the spectrum between these extremes. Personally I believe that the NCCN decree for testing in all non-squamous patients, without evidence that an EGFR mutation is present in more than a vanishingly small proportion of longtime smokers with large cell NSCLC, for instance, and without evidence that testing delivers a survival benefit, is rather heavy handed. I see the guidelines as a reflection of the bias by many academic oncologists who don't have much of a concept of the practical limitations of obtaining tissue in clinics outside of academic centers, or the frustrations and delays of ordering mutation testing in labs in another state rather than in the same building where you work.

On the other hand, I would definitely say that the data supporting treating patients with an EGFR mutation with an EGFR TKI relatively early, including in the first line setting most of the time, supports the value of testing patients with a reasonable probability of having an EGFR mutation (many patients with an adenocarcinoma, many patients who are ex-smokers or never-smokers, regardless of tumor histology), and of favoring an EGFR TKI over chemo in EGFR mutation patients. Even if the data don't support a declaration that EGFR mutation or other molecular testing is a mandate, calling all such testing "investigational" is, in my mind, an oversimplification in the other direction. As with the NCCN guidelines, the ASCO guidelines are a reflection of the group that developed them: ASCO represents the broad range of rank and file oncologists, who see more patients in the community setting who may have little tissue available from an initial biopsy and who may be more resistant to another invasive procedure, especially if molecular testing entails a delay of several weeks before results are available and treatment can be initiated.

The ground has been shifting under our feet, as new information comes in, pretty much in real time, with every new cancer meeting and even with new issues of our monthly journals. This debate will continue, a product of differences in how much evidence different oncologists (and patients) need to change their practice.

Today, molecular testing for a group of patients with a high pre-test probability can provide the only way to receive crizotinib for appropriate patients, it can direct us to give EGFR TKI therapy as early as possible to patients with an EGFR mutation (who might develop a complication along the way that precludes them from getting it later), and I do honestly believe it will come to revolutionize how we see, categorize, and treat lung cancer (and other cancers) over the next decade. It's our way into the future, but I don't think the data quite support mandating it for nearly every patient in the present.