In the same issue of the New England Journal of Medicine that contained the IPASS trial results, Dr. Rosell and colleagues reported results of their effort to institute large-scale EGFR mutation testing in lung cancer patients in Spain, who then received erlotinib (Tarceva).
Patients with non-small lung cancer were recruited from selected public hospitals in Spain. Although 100 sites were chosen (at random), the program was so popular that an additional 29 centers requested and were granted inclusion in the study. Patients who were determined to have EGFR mutations in their tumors were offered Tarceva. The summary of the trial is as shown below:
Out of the 2105 patients enrolled, 17% had EGFR mutations. Consistent with previous observations, patients with EGFR mutations were more likely to be female, never-smokers with adenocarcinomas. About 10% of the patients with EGFR mutations had large cell carcinoma, a histology not usually associated with EGFR mutations. In fact, a similar study previously found no EGFR mutations in a series of patients with large cell carcinoma. I’m intrigued by this but also would interpret it with caution. Although the EGFR testing was done centrally, it is not clear to me that the histology of the specimens was also reviewed by the study pathologists. I wonder if some of these tumors may have been more poorly differentiated and perhaps misclassified as large cell carcinomas. I’d like to see this finding replicated in another study before I start sending off EGFR testing on my patients with a large cell NSCLC histology. While the majority of patients with mutations were never-smokers, EGFR mutations did occur in 13 current smokers and in 56 ex-smokers.
Of the patients who received Tarceva, about half got it as first-line therapy and the other half as second- or third-line therapy. The majority of patients who received Tarceva benefited with only 10% of patients with EGFR mutations experiencing progressive disease while on the drug. Median progression-free survival was 14 months and median overall survival was 27 months! These results are impressive, given that the trial allowed patients with a performance status of 2 and patients who had had received 2 lines of prior therapy. There was no difference in outcome whether patients received Tarceva as first-line or second-line therapy.
(click on image to enlarge)
There was a substantial difference in outcome according to patient sex. For women, median PFS was 16 months, compared to 9 months for men. Similarly, overall survival was 29 months in women and 18 months in men. These differences were statistically significant. When the authors analyzed the variables that affected prognosis, male gender, the L858R mutation (compared to deletion 19), BAC histology, performance status (PS) of 1 (compared to 0, which is the best, completely unlimited) and the presence of brain metastases emerged as variables associated with inferior survival on Tarceva. It is important to remember that this does NOT mean that these patients did not benefit from Tarceva -- just that they did not appear to do as well as other groups. I’d take these results with a big dash of salt, though, as some of the numbers involved here were quite small (only 34 patients with BAC histology) and thus less statistically robust. This is evident in the fact that the analysis suggested that PS 1 patients did worse than PS 2 patients, which doesn’t make a lot of sense. While it has been a consistent finding that women with lung cancer have a better prognosis, it is disheartening that even in patients who all had EGFR mutations and access to Tarceva, men appeared to do substantially worse. It’s frustrating to me that the sex differences we observe in lung cancer are still sort of a black box.
Based on their results, the authors conclude that women, those who have never smoked, and those with non-squamous cancers should be screened for EGFR mutations. They also conclude that large-scale screening is feasible. Although I agree that it makes sense to screen certain populations for EGFR mutations and to make treatment decisions based on these results, I think that doing this in the US may present certain challenges. In Spain, the health care system is a public one: all of the hospitals in the study were part of the same health care system. When patients have more fragmented care, this whole process takes time and in some cases, delaying the patient’s treatment for these results could be harmful. As a practitioner in the US at a tertiary referral center, most patients come to me with biopsies done at a private hospital in their community. Just obtaining the tissue can range from a nuisance to a herculean task. We have had hospitals refuse to release slides and in a few cases, they have called the patients and demanded payment in order to release the patient’s own tissue. Luckily, as this study confirms, it does not seem to impact survival whether a patient with an EGFR mutation receives Tarceva in the first- or second-line setting. I’d feel comfortable starting a patient on front-line chemotherapy even if I thought they had an EGFR mutation but based on the IPASS data, I would not feel comfortable starting a patient on front-line Tarceva without knowing their EGFR mutation status, even if she fit the clinical profile.