Web-Kusuri-no-Check May 7, 2003

Gefitinib Information No3

FDA approval might overlook real toxicity in Japan
Non-clinical and clinical data as well

 AstraZeneca announced May 5, 2003 that the U.S. FDA has granted approval (accelerated approval) for IRESSATR (gefitinib) for the treatment of advanced non-small-cell lung cancer (NSCLC) where no approved treatment exist.
 AstraZeneca also announced May 1, 2003 that the Australian drug regulatory authority approved it.

 AstraZeneca said that FDA approval is based upon data from a U.S. Phase II trial that studied two doses of IRESSA in a total of 216 patients who received both platinum-based and docetaxel chemotherapies (trial 39 or IDEAL 2). Included in the FDA analysis were 142 of the 216 patients. In the group receiving the recommended dose of 250 mg/day, 13.6% (95% CI: 6.4-24.3) of the patients had their tumor shrink by at least 50%. Higher doses did not give a better response and more side effects were observed. The overall response rate for both doses combined in all 142 patients was 10.6% (95% CI: 6.0-16.8). Median duration of response was 7 months.

 But the response rate in Trial 39 was 11.8 % (95%CI: 5.5-18.0) and the median duration of response was 6 month by FDA's analysis last year. What have changed in the same trial?

Concerning with the adverse effects, AstraZeneca say as follows :

 The most frequent drug-related adverse events associated with IRESSA were diarrhea (48%) sometimes associated with dehydration, rash (43%), acne (25%), dry skin (13%), nausea (13%), and vomiting (12%). These events generally occurred within the first month of therapy and usually were mild to moderate. 2% of patients stopped taking IRESSA due to an adverse drug reaction. Infrequent cases (about 1%) of interstitial lung disease (ILD-described as interstitial pneumonia, pneumonitis, and alveolitis) have been observed in patients receiving IRESSA. Approximately 1/3 of the ILD cases were fatal. When ILD occurred, it was often accompanied by acute onset of breathing difficulty with cough or low grade fever requiring hospitalization. The reported incidences of ILD in the 23,000 patient US expanded access program was about 0.3%. In Japanese post-marketing experience the reported rate of ILD was about 2%.

But the fact is as follows:

 Death due to adverse reactions to Iressa is 3.2 % (12/381: 95%CI; 1.4-4.9) by the intensive surveillance at the major oncology hospital in Japan including Tohoku University Hospital (11.1%) and Japanese National Cancer Institute Hospital (3.6%). This rate is far more than the overall death rate due to adverse reactions to Iressa in the clinical trials conducted in Japan (trial 16 and trial v1511: 0%=0/133). The rate is significantly higher than the overall rate of phase I and phase II including Japan (0.3%=2/677) (p<0.001). This rate is also the same order of death rate due to adverse event in the clinical trials other than in Japan which rage 3.1 to 13 % (overall 6.1%). It means most of the adverse reactions were decided as non-Iressa related by investigators in clinical trials. But after the marketing of Iressa in Japan, many fatal cases (more than 246 fatal cases in the latest data at April 22, 2003) were reported. Moreover, as mentioned above, intensive monitoring in major institutes revealed that many of the cases which were considered as non-Iressa-related adverse event in clinical trials should have been classified as Iressa related death.

 AstroZeneca say that in the phase III controlled studies in combination with chemotherapy, there were similar rates of ILD (about 1%) reported in both the placebo and IRESSA arms of the study.

 But deaths rate due to adverse events was higher in the Iressa 250mg group than in the placebo group although strictly speaking it is not significant but p value might be about 0.07. If one patients died less in the control group or if one died more in the Iressa group, it might significantly different.

Further information can be browsed in our two articles;

"Iressa should have never been approved" and "Gefitinib Story".
The latter is the one in the latest ISDB Newsletter (March 2003).

Information on gefitinib (Iressa) is welcomed.