The FLEX trial raises a number of additional points as we struggle to determine how to integrate Erbitux (cetuximab) into the current standards of care. One question is whether we can refine how well we do with Erbitux by using clinical or molecular variables to select better or worse candidates for it. I already mentioned in my prior post that Asian patients (among whom 52% were never-smokers) had a far better survival than Caucasian patients overall, I believe because they often received subsequent oral EGFR tyrosine kinase inhibitors like tarceva (erlotinib) or iressa (gefitinib). To me, it seems likely that whatever modestly beneficial effects may or may not have been conferred from erbitux was overwhelmed by the effects of subsequent treatment. It’s remotely possible that Asian patients could just be otherwise much healthier than Caucasian patients, but I doubt that this is a real explanation.

There was some evidence that patients with adenocarcinomas received a little more benefit with erbitux than the patients with squamous cancers, as shown in the table below (which also highlights the race-based difference)

(click on image to enlarge)

However, many patients with adenocarcinomas are also eligible for avastin, which has been associated with a two month advantage in median overall survival and is FDA approved already in this same setting. I think that for patients in whom you might be particularly worried about bleeding risk but are still technically avastin-eligible, such as the patients with central cavitary tumors or a history of coughing up small amounts of blood, erbitux may be an attractive alternative. There are also some patients who develop problematic side effects on avastin, whether with nosebleeds or kidney problems or high blood pressure, who might switch from avastin to erbitux.

But I think that the easiest transition of use of erbitux will be in patients with squamous tumors, despite the rather modest median overall survival advantage of 1.3 months. There are several real challenges, including the problematic rash, the weekly infusion schedule, and other potential side effects, including higher risk of fevers with low white blood cell counts (known as febrile neutropenia) that we usually hospitalize patients for. There is also a cost that will average tens of thousands of dollars per patient (which for many patients amounts to a very significant co-payment with each treatment), which may be considered to be too expensive to gain what many will perceive as relatively little. I think it will be worth a discussion with patients, but many doctors may recommend that it doesn’t offer enough benefit to a general population to justify adding it, that the juice may not be worth the squeeze.

The equation could change, though, if we could better identify a target population. One factor that I think could emerge as extremely important is selection by EGFR gene amplification, detected by a lab test called fluorescence in situ hybridization, or FISH. I described this very provocative issue in a prior post, in which it appeared that within the limited subset of patients who had received carbo/taxol chemotherapy combined in some way with erbitux in whom tumor tissue was available for EGFR FISH testing, both the progression-free and overall survival were markedly greater (doubled, in fact) in patients who were EGFR FISH positive vs. EGFR FISH negative (pretty evenly divided between two categories). Those are small numbers and not enough to change practice at this point, most experts would agree, but additional tumor tissue from ongoing trials is being tested in the coming months. If that work supports the same conclusion, it would support adding this variable into future trials and arguably using this factor as a clinical decision-making tool.

Another concept that may be relevant here is the presence or absence of a K-Ras mutation (see prior post), one of the important cancer-inducing factors, in the tumor. One of the biggest stories at ASCO this year was work that clearly demonstrated that patients with colon cancer who have a K-Ras mutation in their tumor are very unlikely to benefit from erbitux. There has been work in lung cancer that has shown that patients with K-Ras mutations (about 20-30% of lung cancer patients overall, and usually correlated with smoking more) have almost no chance of responding to tarceva, and they also do very poorly with chemo – so it’s not clear what would be a useful treatment for these patients. I don’t think we’ve seen any work yet on the topic of K-Ras mutations correlating with benefit on erbitux in the setting of NSCLC, but that’s a subject we need to address.

I’ll just close by saying that this work, whether talking about differences between patients with adenocarcinomas or squamous tumors, or considering EGFR FISH and K-Ras testing, all require a reasonable amount of tumor tissue. After years of settling on the smallest amount of tumor tissue possible, often with a fine needle aspirate that collects just a few cells, we need to use larger bore needles to perform core biopsies or even remove entire lymph nodes, so that we can collect enough tissue to more reliably determine the histology (type of lung cancer) and molecular features of a cancer. Accounting for these variables, we may well be able to focus on a narrower population of patients who can gain a more clinically significant benefit from drugs like erbitux, and potentially others.

Next, I’ll turn to some of what we learned about alimta in the previously described trial about maintenance (or early second line) chemotherapy after four cycles of first line chemo. This work could change clinical practice and also provided further information about alimta working in some NSCLC subgroups but not others.