In my last post, I described the novel oral agent PF299804 (PF299), an irreversible "pan-HER" inhibitor not only of the epidermal growth factor receptor but of other members of the human epithelial growth factor receptor (HER) family. We covered some small studies in patients previously treated with chemotherapy and an EGFR inhibitor, with these studies demonstrating that this new agent clearly has activity, shrinking a minority of tumors, with many other patients demonstrating prolonged stable disease even after several lines of prior therapy. This suggests that this agent can provide some additional benefit beyond that conferred by a reversible EGFR inhibitor that we already have available, namely Tarceva (erlotinib) or Iressa (gefitinib).
But another way to ask about how much PF299 offers over a currently available standard oral EGFR inhibitor is to compare them directly in patients who have never received prior therapy against EGFR. That study has been done and was presented at ASCO 2010.
Michael Boyer, out of Sydney, Australia led a randomized phase II trial that was conducted at a handful of centers around the world. The study enrolled 188 patients with advanced NSCLC and tissue available for molecular marker studies, who had previously received one or two prior lines of chemotherapy but hadn't received EGFR inhibitor therapy. Patients were randomized to receive either daily Tarceva at the standard 150 mg daily dose, or PF299 at 45 mg by mouth daily. The primary endpoint was progression-free survival (PFS), and the investigators also looked at tumor response rate, side effects, and several other measures. They planned subset analyses of PFS results based on clinical and molecular variables as well. It's worth noting that while the two arms were well balanced for many variables, it just happened to work out that the PF299 arm had a higher proportion of patients with an EGFR mutation (20% vs. 12%). On the other hand, more patients on the PF299 arm had a marginal performance status of 2 (19% vs. 3%), which would be expected to disfavor the PF299 arm.
The trial demonstrated a significantly higher response rate with PF299 vs. Tarceva (17% vs. 4%, p = 0.009) as well as a significantly improved PFS in recipients of PF299 in the overall trial population (p - 0.019).
What was especially interesting was that there didn't seem to be any isolated subgroup that received a much greater benefit with PF299 vs. Tarceva, or vice versa. Instead, just about all subgroups had similar degree of modestly superior PFS in the recipients of PF299. In the figure below, the blue boxes situated to the left of the vertical line on the right side represent improvement with PF299 over erlotinib (and the more to the left, the stronger the benefit). There are no blue boxes that fall to the right of the bar to signify better results with Tarceva.
In these plots, the width of the various boxes is proportional to the size of the group, and the horizontal lines extending from it represent the degree of variability in the results. So we see that whether patients have an EGFR mutation or wild type, KRAS mutation or wild type, adenocarcinoma or non-adenocarcinoma, men or women, never-smoker or ever-smoker, and Asian or another race, the trend favored PF299 to a similar degree of about 30-40%. The only subset that didn't show that trend was comprised of patients 65 or older, which was a small group, and in whom the results were remarkably similar between PF299 and Tarceva.
This study also provided an opportunity to directly compare PF299 in terms of side effects to an agent we've become very familiar with. These results showed that side effects may well be a concern and potential limitation for PF299, at least at the dose studied. "Dermatitis acneiform", as opposed to "rash", was more common with PF299 (though I think it may be splitting hairs to try to understand how they categorized dermatitis acneiform compared to rash). Also, nailbed infections, mouth sores, and diarrhea were all more common and tended to be more severe with PF299 vs. Tarceva. Since there are certainly patients for whom side effects of Tarceva can be a real challenge, it sounds to me like PF299 may be more difficult for many patients to navigate, even if side effects were most often in the mild to moderate range.
The authors concluded that this work is promising enough to warrant a larger, randomized phase III trial with the same design, and I certainly agree. Other studies with PF299, both in EGFR inhibitor-treated and naive patients are likely to be pursued in the next few years, and I certainly would be eager to have my patients participate. It may well prove to be an agent that not only can provide additional benefit to the minority of patients who have an initial response to an EGFR inhibitor and then become resistant, but also appears likely to provide a meaningfully greater efficacy than Tarceva in a much broader population of patients, in various clinically and molecularly defined subgroups. It will be important for future trials and overall clinical experience to better define that advantage and explore the potentially increased toxicity challenges that may be a real limitation for some patients.