Among the many challenges in clinical oncology is the fact that a very significant proportion of our patients are quite a bit more debilitated than the vast majority of patients in clinical trials that test our anti-cancer therapies. Approximately a third of the patients with advanced NSCLC have what would be considered a poor performance status (PS) of 2 or 3 (0 to 5 scale, 0 being asymptomatic, and 5 being dead), but they are extremely under-represented on our clinical trials. Because our anti-cancer therapies, whether standard chemo or targeted therapies, have challenging side effects for many patients, we struggle to balance between fighting the cancer and harming the patient. In the absence of much evidence, we need to rely on our judgment when we make treatment recommendations. Many physicians and patients have looked upon so-called targeted therapies as a potential alternative to standard chemotherapy drugs that could allow us to treat the cancer more selectively, with less collateral damage to a person’s normal tissues. The oral epidermal growth factor receptor (EGFR) inhibitors like Iressa (gefitinib) and Tarceva (erlotinib) have been the first agents that have been widely used as single agents, and they certainly have activity and can improve survival. Nevertheless, we’ve seen evidence that Tarceva was convincingly inferior to standard chemotherapy for poor performance status patients (at least “unselected” patients who hadn’t been particularly singled out by having an EGFR mutation, being a never-smoker, etc.); another study compared Iressa to the widely used single agent approach of Navelbine in elderly patients (not synonymous with poor performance status) and showed very comparable efficacy. Patients with both good and poor performance status were included in the important clinical trial that showed a survival benefit for Tarceva compared with placebo as a second or third line therapy, dicussed further below. We get some additional information from a newly reported trial that randomized 201 poor performance status patients (considered to be “unfit for chemotherapy”) to either Iressa or placebo (plus general supportive care for all patients). The study was conducted in several centers in North America and Europe (so very few Asian patients were enrolled, a significant issue because results with EGFR inhibitors are often very different in Asian vs. Western populations). The majority of patients were also older, about half 75 or older, and most of the rest in the 65-74 age range. Importantly, this trial included not only patients with a PS of 2, who are still up and out of bed more than 50% of the time, but also PS 3 patients, who are pretty debilitated and spend more than half of the day in bed (enrollment was an approximately 60/40 split of PS 2 vs. 3 patients). Our experience in studying such frail patients is extremely limited.

The study found that there was a trend toward improved progression-free survival and overall survival that favored the recipients of Iressa, as well as a higher response rate with Iressa than the placebo. There was an approximately 15% relative improvement in progression-free and overall survival with Iressa, but the magnitude of benefit just wasn’t enough to translate to a statistically significant benefit. Both groups had a disappointing median survival of only about 3 months, and there was not a significant benefit with Iressa in any of these efficacy measures. Not surprisingly, the patients with a better PS of 2 did better overall than the patients with a PS of 3, regardless of what treatment they received. Interestingly, though, the patients with a PS of 2 didn’t show any real improvement with Iressa, while the more debilitated patients with a PS of 3 actually seemed to get more of a benefit with the active drug. This result is extremely hard for me to explain, and it’s definitely not consistent with prior work with frail patients, so I wonder if it is just a random finding that is sometimes seen when we study small numbers of patients. Importantly, there were no real surprises in terms of side effects, with diarrhea and rash being commonly reported but usually mild to moderate in severity; nausea, vomiting, diminished appetite, and fatigue were also reported, but these symptoms were also commonly seen in the recipients of the placebo and can be understood as common cancer-related symptoms as well. Quality of life was measured and wasn’t really different between recipients of Iressa vs. placebo. I don’t think we should be surprised that Iressa showed a trend toward modestly improved efficacy but no significant improvement. In a much larger trial of Iressa vs. placebo in previously treated patients with advanced NSCLC, Iressa showed only a non-significant improvement in survival: this negative trial led to the withdrawal of Iressa from the US market, while a very similar trial of Tarceva vs. placebo had demonstrated a significant survival benefit. But how helpful is Tarceva in a patient with a poor performance status? We can look at the trial called BR.21, of Tarceva vs. placebo in previously treated patients with advanced NSCLC, which enrolled some patients with a PS of 2 (about 25% of the study population) or even 3 (9% of the population). Tarceva showed an approximately 20% relative improvement in the PS 2 and 3 patients, translating to a two month improvement in median survival (4.6 vs. 6.8 months), which is comparable to what was seen in the better PS patients, and only a little more of a benefit than seen in the Iressa vs. placebo trial described above. The Southwest Oncology Group also recently reported results of a single arm trial of 81 PS 2 patients who received Tarceva at the standard starting dose of 150 mg daily and reported a response rate of 8% and a median survival of 5 months. Moderate to severe side effects were reported in 40% of patients (fatigue, rash, diarrhea, decreased appetite). These numbers were in the general ballpark of what many of the chemo-based trials have shown, and SWOG is primarily interested in whether it’s possible to select optimal candidates for EGFR-based therapy based on molecular characteristics. In fact, this last point is an important one. I had previously posted on the Japanese experience of giving Iressa to frail elderly patients with EGFR mutations and noted that this was a very effective therapy. In contrast, from the results I describe above, I think it’s fair to say that the EGFR inhibitors have a potential role, with some conflicting evidence, but it’s not clear that these agents are an ideal substitute for standard chemotherapy in an unselected population. Though only about 5-10% of Western populations have a NSCLC tumor with an EGFR mutation, a much larger proportion have EGFR gene amplification by a test called FISH. Whether EGFR FISH testing is likely to be very helpful in predicting benefit from the oral EGFR inhibitors in the elderly and poor risk populations is a subject we’ll cover soon.