KANAGAWA, Japan—According to the results of a randomized, multicenter phase III trial, irinotecan(Drug information on irinotecan) (Camptosar) plus cisplatin(Drug information on cisplatin) (Platinol) was associated with a highly significant improvement in survival, with less myelosuppression, in patients with extensive-disease small-cell lung cancer (SCLC), compared with the standard therapy of etoposide(Drug information on etoposide)/cisplatin. The findings were presented at the 36th Annual Meeting of the American Society of Clinical Oncology (ASCO).
The overall response rate was 83% in the irinotecan/cisplatin arm vs 68% in the etoposide/cisplatin arm, said Kazumasa Noda, MD, of the Kanagawa Cancer Center, Japan. The study was conducted by the Japan Clinical Oncology Group (JCOG 9511).
Irinotecan/cisplatin every 4 weeks is considered a new standard treatment regimen against extensive-disease SCLC. Irinotecan/cisplatin is now being used as a standard regimen in Japan, Dr. Noda told ONI.
The large difference in survival observed at the first interim analysis prompted the second interim analysis to be performed earlier than originally planned. The significant difference in survival continued at the second interim analysis, done November 19, 1998.
“The irinotecan/cisplatin arm was highly superior to the etoposide/cisplatin arm in overall survival (P = .00025),” Dr. Noda said. This difference triggered early termination of the study.
“In accordance with the recommendation of the independent monitoring committee, enrollment into this study was terminated in December 1998,” Dr. Noda said. The study, which ended with 154 patients enrolled, had originally planned a sample size of 230 patients.
Much enthusiasm has greeted the Japanese study, and efforts are underway to replicate its results in North America.
“That’s a fabulous study,” Alan Sandler, MD, of Vanderbilt University, said in an ONI interview. “It was a nicely performed study, and it ended early because the results were so disparate . . . really some of the most amazing results seen for SCLC. Obviously it needs to be confirmed in other trials.”
Two studies are planned to replicate the Japanese study. Dr. Sandler, together with Paul Bunn, MD, of the University of Colorado Cancer Center, Denver, is planning to repeat the study using a modified dosing scheme.
“About 50% of Dr. Noda’s patients didn’t receive day 15 of irinotecan because of hematologic toxicities,” he said. “We’re planning on doing a weekly two-out-of-three regimen [for both platinum and irinotecan] rather than three out of four.” Dr. Sandler hopes to begin enrolling patients by the fall.
In addition, the Southwest Oncology Group (SWOG) is planning a confirmatory study using doses identical to the regimen Dr. Noda described.
Giuseppe Giaccone, MD, PhD, of Free University Hospital, Amsterdam, The Netherlands, who commented on Dr. Noda’s report at the ASCO session, expressed a few reservations.
He recommended that a multivariate analysis be performed. He also recommended that if studies intended to replicate the results find similarly promising results, they should nevertheless not be prematurely terminated.
“This is extensive-disease SCLC,” he said. “For 20 years we did not see an improvement in survival with chemotherapy by changing different cocktails. I wonder whether an interim analysis should not have looked just simply at safety and response rate and things like this, and not really have given advice about stopping the study when a difference in survival is seen before the target accrual is reached.”
He added, “I think the study is clearly interesting but requires repetition by another group.”
Between November 1995 and November 1998, 154 patients were randomized between the two arms, with 77 patients entered into each arm. The regimen for the irinotecan/cisplatin arm was irinotecan 60 mg/m² on days 1, 8, and 15 and cisplatin 60 mg/m² on day 1 every 4 weeks for four cycles. The etoposide/cisplatin arm regimen consisted of etoposide 100 mg/m² on days 1, 2, and 3 and cisplatin 80 mg/m² on day 1 every 3 weeks for four cycles. Both arms required hydration and administration of antiemetics and G-CSF (Neupogen) if necessary until recovery.
The primary endpoint for the study was survival. The complete remission rate was 2.7% in the irinotecan/cisplatin arm and 9.1% in the etoposide/cisplatin arm. Partial remission rate was 80.5% in the irinotecan/cisplatin arm
and 58.4% in the etoposide/cis-platin arm.
According to updated survival data presented by Dr. Noda at the conference, median survival and 1-year survival were 390 days and 58.4% in the irinotecan/cisplatin arm vs 287 days and 37.7% in the etoposide/cisplatin arm.
“The irinotecan/cisplatin regimen is expected to prolong median survival by up to 40%, compared to the etoposide/cisplatin regimen,” Dr. Noda said.
A difference in hematologic toxicities was seen between the two arms. JCOG grade 3-4 leukopenia occurred in 27% of the patients in the irinotecan/cisplatin arm and 52% of patients in the etoposide/cisplatin arm (P = .003). Grade 3-4 neutropenia developed in 66% of those in the irinotecan/cisplatin arm vs 92% in the etoposide/cisplatin arm (P = .0002). Grade 3-4 thrombocytopenia occurred in 5% of patients in the irinotecan/cisplatin arm and 19% of those in the etoposide/cisplatin arm (P = .01).
“Severe adverse events were observed more frequently in the irinotecan/cisplatin arm; however, those were not beyond the expected level,” Dr. Noda said.
Except for diarrhea, nonhematologic toxicities occurred with a similar severity in each arm. Grade 3-4 diarrhea was seen in 16% of the irinotecan/cisplatin arm and in none of the patients in the etoposide/cisplatin arm (P = .0001).
Four treatment-related deaths occurred, three in the irinotecan/cisplatin arm and one in the etoposide/cisplatin arm. Two deaths in the irinotecan/cisplatin arm were due to infection accompanied by neutropenia. The third death in the irinotecan/cisplatin arm was caused by bleeding from the metastatic site. The death in the etoposide/cisplatin arm was due to pneumonitis.
A second randomization for complete or good partial responders was also planned. It was designed to compare subsequent radiotherapy with observation. “However the second randomization was terminated early because of poor accrual,” Dr. Noda said. Only 20 patients had been enrolled, he told ONI.
