Even FDA approval was not as rapid as some critics wished; the Administration held back for a time while toxicity reports from Japanese experience with the drug were explored. The main issue seems to have been a complication known as interstitial lung disease, affecting 3-4% of patients in Japan, with a third of cases proving fatal. But, as James Schultz made clear in his account of the delayed U.S. approval (J. Natl. Cancer Inst., 95[8]: 577-9, 2003), clinicians expect to see severe respiratory complications in patients of this sort.

Science Watch contacts with gefitinib trialists outside Japan were reassuring. Dr. Roy S. Herbst (M.D. Anderson Cancer Center, Houston, Texas) believes that "the efficacy—response, stable disease and symptom improvement—far outweighs toxicity," and in the U.S.A. most patients with advanced incurable NSCLC are likely to be given the drug. Dr. Giuseppe Giaccone (Free University of Amsterdam, Netherlands) also felt that interstitial lung disease in the West was less of a problem, with a frequency of 1% or less and a lower case fatality rate than in Japan. However, the discovery of EGFR gene mutations and other markers of drug sensitivity means that giving gefitinib to unselected patients "is less scientifically sound," he added.

In citation terms, articles on gefitinib are not yet the stuff of headlines but this time round we do have at #16 the report of a phase II study by Dr. Masahiro Fukuoka and colleagues, with Giaccone as a co-author (J. Clin. Oncol., 21[12]: 2237-46, 2003; with 36 citations recorded during this period and 103 overall). This was a straight comparison of two doses of gefitinib, 250 mg and 500 mg daily. For the lower dose, with a "favorable" adverse-event profile, the symptom improvement rate was 40% with a median survival time of over seven months. More recent papers include studies by Giaccone, Herbst, and others (J. Clin. Oncol., 22[5]: 777-84 and 785-94, 2004).

When paper #16 was published in June, 2003, an accompanying editorial by Dr. David H. Johnson and Dr. Carlos L. Arteaga (Vanderbilt-Ingram Comprehensive Cancer Center, Nashville, Tennessee) noted that expression of the EGFR gene, as assessed by methods then available, did not appear to be a valid predictor of the drug’s activity (J. Clin. Oncol., 21[12]: 2227-9. 2003). However, clinical development of the drug should not be held up for that reason, they concluded. A year later screening for gefitinib sensitivity by looking for EGFR mutations was firmly on the agenda. In May, 2004, came a report of somatic mutations in eight of nine patients with gefitinib-responsive lung cancer but in none of seven non-responders (T.J. Lynch, et al., New Eng.l J. Med., 350[21]: 2129-39, 2004). In vitro the mutants showed increased sensitivity to the drug. Soon afterwards came confirmation of this possibility, via a study that, interestingly, also suggested that EGFR mutations might be more common in Japanese patients with lung cancer than in patients from the U.S. Both studies were on small numbers, and the specificity and sensitivity of this screening test to identify those most likely to benefit from gefitinib are not yet known.