Afatinib (BIBW-2992, Tomtovok) wins a battle (PFS) but loses the war (OS) for EGFR TKI-sensitive patients

Howard (Jack) West, MD, Founder and President

Although the responses we see with a targeted therapy like the epidermal growth factor receptor (EGFR) oral tyrosine kinase inhibitors (TKIs) for patients with the precise target, an activating mutation in the EGFR gene, has redefined our hopes and expectations about what is possible to achieve for at least some patients with advanced non-small cell lung cancer (NSCLC), nearly all of these patients develop a resistance to these agents at some point months or years after often having a very significant response to one of these agents. What to do at that point, whether there is some particular intervention to pursue to get the genie back into the bottle and achieve some similar kind of benefit again, remains unknown.

In the LUX-Lung 1 trial, the irreversible EGFR and HER-2/neu inhibitor BIBW-2992, now named afatinib, with a marketing name Tomtovok (changed from what was previously Tovok), was studied in this population compared with placebo (2:1 randomization to active drug), with all patients also getting supportive care as well. The trial was open to patients with a lung adenocarcinoma, had previously received at least one line of prior chemotherapy, and had gone for at least 12 weeks without progressing on another EGFR inhibitor, either Iressa (gefitinib) or Tarceva (erlotinib). A total of 585 patients were enrolled, and these were clearly closer to the population of the IPASS trial than a general population of folks at the VA hospital: the median age was in the late 50s, 56-58% of the patients were Asian, about 1/4 with no symptoms even after several prior lines of treatment, and nearly 2/3 were never-smokers. This was clearly enriching for a population likely to have an EGFR mutation, though we haven't seen any analysis of molecular markers for the trial. The median duration of EGFR TKI therapy was about 10 months, and approximately 45% of the enrolled patients achieved an objective response (significant tumor shrinkage), so most of these patients appear to have been the target population that really benefited greatly from an EGFR TKI and then developed acquired resistance to it.

While many people might be upset with the idea of a placebo, the trial attempted to minimize the potential for patients on the placebo to go too long while progressing. Repeat scans were done as early as 4 weeks after randomization, so that patients could move on to some other intervention if they were progressing, though cross-over from placebo to Tomtovok wasn't permitted.

The response rate was significantly higher with Tomtovok: 13% vs. 0.5% (unconfirmed numbers reported), as was the disease control rate 8 weeks into treatment (including patients achieving stable disease): 58% vs. 19%. Similarly, the hazard ratio for progression-free survival (PFS) was 0.38, meaning that the active drug was associated with an over 70% improvement in (PFS). The PFS benefit was present throughout just about every subgroup they looked at (age, sex, number of prior chemo regimens, smoking status, etc.), but it was least impressive for patients who responded to a prior EGFR TKI for less than 24 weeks and those who achieved stable disease only, rather than an objective response of significant tumor shrinkage. So this suggests that the patients least likely to actually have an EGFR mutation, or more precisely those who received the least dramatic benefit from the original EGFR TKI they received also received the least benefit from Tomtovok afterward.

Recipients of Tomtovok were also more likely to demonstrate improvement in quality of life indicators like cough, shortness of breath, and pain, as well as a longer time before these symptoms worsened.

Meanwhile, side effects weren't a real surprise. The leading side effects included diarrhea, rash, mouth sores (stomatitis), nail changes, diminished appetite, and less commonly nose bleeds (also known as epistaxis) and itching (aka pruritis). These were all more common with the active drug than with placebo, but in most cases, the symptoms were mild to moderate rather than more severe.

That's the good news. The bad news, which I would consider a real surprise, especially when looking at all of the favorable results seen during the period of actual treatment, is that there was no improvement in overall survival, which was the primary endpoint of the study. Sadly, there's more: survival was numerically (though not statistically significantly) worse in recipients of the active drug (10.8 vs. 12.0 months; HR = 1.09). It's worth noting that these median survival numbers are certainly very good for patients who had received 3-4 lines of prior treatment. Although I don't have the actual survival curves to reproduce here yet, I've seen them directly, and what they show is that for the first several months, the two groups track together in an overlapping way, then separate with a clear suggestion that the placebo arm is doing better later on.

That may correspond to the higher rates of post-study therapy being delivered to the recipients of placebo, both chemotherapy and additional EGFR TKI therapy. I'll mention here again that I've had a few patients in the last few months who have had very impressive responses after re-initiating either the same EGFR TKI (Tarceva after prior Tarceva, with an 18 month hiatus for chemo and observation), so perhaps many (and disproportionately more) patients on the placebo arm received significant benefit from effective post-study therapy. Another possibility is that many patients who started with placebo developed rapid progression just from coming off of an EGFR TKI they were on previously, so they had rebound progression that was captured on the first scan. So it may be partly that the patients on Tomtovok did particularly well in terms of PFS, but also that patients randomized to placebo did particularly poorly because some had their progression accelerated by coming off of a still partially effective EGFR TKI before starting the trial.

What can we make of this? Tomtovok is clearly an active drug in this setting, but not only did the agent fail to meet its primary endpoint of a significant improvement in overall survival, the patients assigned to placebo did numerically better in terms of survival, for reasons that we can't really explain. Maybe molecular marker work will help to clarify that some patients were well served while others were not (perhaps based on a particular EGFR mutation, presence or absence of a particular mechanism of resistance, which treatments patients received after the trial, etc.). In the meantime, we need to learn more before I'd envision Tomtovok being an appealing choice in this setting if it leads to a trend toward worse survival than the placebo arm.



Join the Conversation

Please feel free to offer comments and raise questions in our
discussion forums.

 Join The conversation