Lung cancer is diagnosed in more than 170,000 Americans yearly. Although the mortality rate is starting to plateau in North American males, it continues to rise in American women and in both genders outside the United States, particularly in Europe and the Orient. The 5-year survival rate in the United States is 14%, compared with 8% in the United Kingdom.[1,2] The majority of patients manifest with disseminated disease at diagnosis or develop distant metastases shortly after initial local regional therapy. Hence, new agents with improved systemic activity are desperately needed. This paper will detail ongoing, randomized phase III trials in North America in extensive disease smallcell lung cancer (ED SCLC), which will serve to either confirm or refute the role of irinotecan(Drug information on irinotecan) (Camptosar) established by a critical Japanese effort. It will also address the role of irinotecan in combination with platinum up front in treatment-naive non- small-cell lung cancer (NSCLC) patients and with nonplatinum agents in both the first-line and second-line setting in advanced NSCLC. We will discuss these trials in the context of current state-of-the art therapy in both settings. Non-Small-Cell Lung Cancer Single-Agent Therapy
Multiple phase II studies, particularly in Japan, have established the utility of irinotecan in the treatment of advanced NSCLC (Table 1). Response rates have ranged as high as 30% to 35%, using conventional schedules of irinotecan at doses of 100 to 125 mg/m² weekly × 3 or 4,[3- 5] or doses of 300 to 350 mg/m² every 3 weeks, with or without hematopoietic growth factor support.[6] Toxicities have been manageable and have generally included myelosuppression and diarrhea. In the sole US trial evaluating single-agent irinotecan, Baker and colleagues noted a 15% response rate in 41 patients receiving a dose of 100 mg/m² weekly × 4 every 6 weeks; 79% had stage IV disease.[7] Kamuyama et al (Hokkaido Institute of Public Health, Sapporo, Japan) examined the activation/ detoxification of irinotecan by human lung cancer cell lines and determined the expression of the putative enzyme(s) that convert irinotecan to the active moiety, SN-38.[8] Of 25 squamous cell carcinoma cell lines, 15 were strongly positive for carbox ylesterase; of 25 adenocarcinomas, 20 were positive, including 4 that expressed strong positivity. Cisplatin/Irinotecan Combinations
Cisplatin(Drug information on cisplatin) remains the cornerstone of combination therapy in advanced NSCLC. In preclinical models, cisplatin in combination with irinotecan, has yielded synergistic cytotoxicity. With the exception of myelosuppression, irinotecan and cisplatin have nonoverlapping toxicities. Consequently, investigators have sought to assess the clinical activity of irinotecan in combination with cisplatin. The earliest efforts emanated from Japan. Masuda and colleagues evaluated cisplatin at 80 mg/m² on day 1 every 4 weeks in combination with irinotecan at 60 mg/m² on days 1, 8, and 15.[9] The overall response rate was 52%; time to progression was 4.4 months, median survival was 10.2 months, and the 1-year survival rate was 33%. Ueoka et al assessed the irinotecan and cisplatin combination at doses of 50 and 60 mg/m², respectively, on days 1 and 8 every 4 weeks.[10] The overall response rate was slightly lower at 41%. Median survival was 13 months; the 1-year survival rate proved promising at 58%. Mori et al conducted a phase II study of irinotecan at 160 mg/m² bolus on day 1 in combination with cisplatin at 20 mg/m² daily × 4 by continuous infusion with granulocyte colonystimulating factor (G-CSF) support.[ 11] Twenty-four treatment-naive advanced NSCLC patients were evaluated. The overall response was 58.5%, with a median survival time of 44.8 weeks and 1-year survival rate of 44%. Major adverse events included grade 3/4 diarrhea (23% of patients), granulocytopenia (20%), thrombocytopenia (15%), and anemia (15%), without any treatment-related deaths. In Europe, Cardenal and colleagues mounted a multi-institutional phase II trial combining infusional irinotecan at 200 mg/m² over 1 hour in combination with cisplatin at 80 mg/m² every 21 days.[12] Over an 8-month period, 48 patients were recruited; 43 were evaluable for toxicity. Grade 4 neutropenia occurred in 11.6%, grade 4 diarrhea in 9.3%, and grade 3 nausea and vomiting in 16.2%. Of 32 evaluable patients, 12 (37.5%) achieved a partial response. Phase III Randomized Trials
Two separate prospective, randomized phase III trials evaluated the role of irinotecan in advanced NSCLC. Negoro et al compared irinotecan either alone or in combination with cis- platin to a vindesine(Drug information on vindesine)/cisplatin combination.[ 13] Eligibility stipulated chemonaive stage IIIB or IV measurable NSCLC, Eastern Cooperative Oncology Group (ECOG) performance status 0 to 2, and adequate marrow, hepatic, renal, and pulmonary functions. Patients older than 75 years were excluded; 241 patients were enrolled. Demographics with respect to age, performance status, prior weight loss, albumin, and lactate dehydrogenase were well matched for each arm. Patients received one of three regimens (results are delineated in Table 2):

• Arm A:Irinotecan at 60 mg/m² days 1, 8, and 15 and cisplatin at 80 mg/m² on day 1 every 4 weeks (based on Masuda et al[9] and Negoro et al[13]).
• Arm B: Vindesine at 3 mg/m² on days 1, 8, and 15 and cisplatin at 80 mg/m² on day 1 every 4 weeks (the standard).
• Arm C: Irinotecan at 100 mg/m² on days 1, 8, and 15 every 4 weeks.

Leukopenia was much more pronounced in the vindesine/cisplatin combination, while thrombocytopenia was more common in the irinotecan and cisplatin combination. Both irinotecan arms resulted in substantially more grade 3 and 4 diarrhea. The cisplatin arms yielded considerably more nausea and vomiting. The incidence of anemia was driven by cisplatin: 39% grade ≤ 3 for irinotecan and cisplatin compared with 23% for vindesine and cisplatin, and 6% for irinotecan alone. Survival among stage IV patients was significantly better for those receiving irinotecan (arm A or C) compared to those receiving vindesine and cisplatin (Figure 1). In contrast, a second phase III trial reported by Niho et al[14] and Kunitoh and colleagues[15] comparing the two platinum combinations alone did not reveal a significant advantage or disadvantage for irinotecan. Baseline demographics were well matched in the 199 randomized patients. Grade ≥3 neutropenia was more common for the vindesine and cisplatin arm (83%) compared with the irinotecan and cisplatin arm (64%). Grade ≥ 3 anemia was more common for the irinotecan combination (24%) compared with the vindesine and cisplatin arm (17%). Likewise, grade ≥ 3 nausea/vomiting and diarrhea were more common at 19% and 15%, respectively, for irinotecan and cisplatin, compared with 12% and 2%, respectively, for the vindesine and cisplatin combination. Median survival for patients enrolled in the irinotecan and cisplatin arm was 10.6 months, with 1- and 2- year survival rates of 36.4% and 8.7%, respectively, in stage IV patients. Median survival for vindesine and cisplatin was 11.5 months, with 1- and 2-year survival rates of 41.4% and 10.3%, respectively (P = .668). Significant negative prognostic factors included male gender (P = .0028) and performance status of 2 (vs 0/1) (P = .0003). It is unclear why one trial showed an apparent survival advantage for the irinotecan and cisplatin combination in stage IV patients, while the other did not. Nevertheless, irinotecan joins the pantheon of other modern agents (vinorelbine, gemcitabine(Drug information on gemcitabine), the taxanes) which, in combination with a platinum, have demonstrated comparable if not superior activity to older second-generation agents. North American Trials
The initial combination effort in North America reported by DeVore and colleagues[16] recapitulated the Japanese study of Negoro and colleagues.[ 13] Cisplatin was dosed at 80 mg/m² on day 1, and irinotecan at 60 mg/m² weekly × 3 every 4 weeks.[16] A total of 52 patients were enrolled. Overall response rate was 29%, with a median time to progression of 5.1 months, a median survival of 9.9 months; and a 1-year survival of 37%. Grade ≥ 3 neutropenia occurred in 46% of patients, with an overall 11.5% incidence of febrile neutropenia. Nausea and vomiting and asthenia were driven by cisplatin. The relative grade ≥ 3 incidences were 32.7% and 23.1%, respectively. Grade 3 or 4 diarrhea occurred in 17.3% of patients. The relative dose intensity of irinotecan was 75.5%. Dose reductions of irinotecan were required by 73% of patients. Most patients ultimately had their irinotecan dose reduced to ≤40 mg/m² weekly. This trial has been criticized for overly stringent dose modification criteria. Consequently, a follow-up study mounted through Vanderbilt Cancer Network (VCCAN) and Fox Chase Cancer Center (FCCC) combined weekly irinotecan with weekly cisplatin.[ 17] There were three justifications for this move: (1) same-day administration could better exploit the putative in vitro synergy of these two agents; (2) improved sequencing of irinotecan and cisplatin might potentially improve efficacy; and (3) decreasing the cisplatin dose per administration could potentially diminish toxicity. The regimen was modeled after phase I data of Saltz and colleagues.[18] Eligible patients received irinotecan at 65 mg/m² weekly × 4 in combination with cisplatin at 30 mg/m² weekly × 4. Treatment was repeated at 6-week intervals. A total of 50 patients were enrolled. Overall response rate was 36%; median time to progression was 6.9 months, median survival was 11.6 months, and 1-year survival was 46%. These results were the best ever observed in advanced NSCLC in the history of the VCCAN, and proved comparable to previous results seen with combination paclitaxel(Drug information on paclitaxel) and carboplatin(Drug information on carboplatin) (Paraplatin) at FCCC and its affiliated network. The overall incidence of grade ≥ 3 neutropenia was 25.5%; febrile neutropenia occurred in 6% of patients. Grade ≥ 3 thrombocytopenia occurred in 12% and grade ≥ 3 nausea and vomiting in 26% of patients. The relative dose intensity of irinotecan in this combination was 89%, and the dose intensity was fairly well maintained for both agents. Comparison of the two North American irinotecan and cisplatin regimens is depicted in Table 3. Data to date favor using weekly as opposed to monthly cisplatin. Nausea and vomiting were less pronounced. In addition, serious neutropenia was less common. Finally, the response rate and survival appeared comparable to, if not superior to, the standard established approach using monthly cisplatin and weekly irinotecan. Because of cumulative asthenia, several investigators have proposed a 2-weeks-on/ 1-week-off schedule, although this approach has never been formally studied. Interest also exists in comparing this schedule to other established cytotoxic regimens, though planned trials have not yet evolved. Compared to cisplatin, carboplatin yields considerably less nonhematologic toxicity, in particular nausea and vomiting. Fledgling phase I Japanese efforts have integrated carboplatin and irinotecan. In one effort, 17 patients, 71% of whom had stage IV NSCLC, received carboplatin at area under the concentration-time curve (AUC) of 5 every 4 weeks and irinotecan weekly on days 1, 8, and 15 every 4 weeks. The maximum tolerated dose of irinotecan was 60 mg/m² and the recommended phase II dose was 50 mg/m². Dose-limiting toxicity included neutropenia and thrombocytopenia. Median survival was 10.5 months; 1-year survival was 35%.[19] Mukohara and colleagues evaluated a similar regimen.[20] The incidence of grade ≥ 3 neutropenia, anemia, thrombocytopenia, and diarrhea was 76%, 26%, 47%, and 6%, respectively. Only 43% of the intended day 15 irinotecan dose was administered. Overall response rate was 25%. Median survival was 10.8 months and 1-year survival was 39%.[20] Sunpaweravong et al[21] at the University of Colorado evaluated this combination further. A total of 14 patients were enrolled onto this phase I/ II effort, including 10 with NSCLC and 2 with SCLC. The initial dose of carboplatin at AUC 6 and irinotecan at 60 mg/m² days 1 and 8 every 3 weeks proved too toxic. DLT included neutropenia, thrombocytopenia, diarrhea, and fatigue; hence the recommended dose was AUC 5 for carboplatin and 50 mg/m² days 1 and 8 for irinotecan. Partial responses were observed in one patient each with NSCLC and SCLC, and stable disease occurred in two patients (NSCLC). A phase II trial by Kelly and colleagues at this institution in untreated extensive-stage SCLC is under way. Taxane and Irinotecan Combinations
There are a number of reasons to consider irinotecan and taxanes in combination. Each is individually active in both NSCLC and SCLC. Both agents have nonoverlapping toxicities, malleable schedules, and relative non- cross resistance. Finally, preclinical data suggest an additive, if not synergistic, effect. Murren and colleagues mounted a phase I trial of irinotecan and paclitaxel, both agents administered weekly × 4 every 6 weeks.[22] Twenty-one patients were enrolled in this phase I study. A total of 53 cycles were administered. The maximum tolerated dose for irinotecan in this schedule was 50 mg/m² and paclitaxel 75 mg/m² weekly. Pharmacokinetics revealed no drug-drug interaction based on levels of irinotecan and its active metabolite SN-38. However, chemotherapy on weeks 3 and 4 was often dose-modified or omitted. Only 26% of cycles were administered as planned. An ongoing phase II trial by these investigators enrolling advanced NSCLC patients (performance status 0 to 2) was an abbreviated schedule: irinotecan at 50 mg/m² weekly × 2 every 3 weeks in combination with paclitaxel at 75 mg/m² weekly × 2 every 3 weeks. To date, 10 patients, ages 47 to 68, have been accrued. There has been no grade 4 neutropenia. The relative dose intensity is 97%. Two of six evaluable patients have sustained partial response (18+ weeks), three have had stable disease, and one only progressive disease. Rosen et al combined irinotecan with paclitaxel, both agents given once every 3 weeks.[23] The maximum tolerated doses were 225 and 100 mg/m², respectively. Grade ≥3 diarrhea was observed in only one of nine treatment courses at maximum tolerated dose, whereas at higher doses, grade ≥ 3 diarrhea occurred in five of seven patients. Objective responses were seen in NSCLC and in occult primary squamous malignancy. As in the Murren trial, paclitaxel coadministration did not alter irinotecan pharmacokinetics (irinotecan and SN-38 levels). Socinski and colleagues have mounted phase I and phase II studies of irinotecan in combination with standard paclitaxel and carboplatin.[24] Thirty-three NSCLC patients were enrolled in the initial phase I effort. The maximum tolerated doses for these three agents using an every-3- week schedule were paclitaxel at 175 mg/m² on day 1, irinotecan at 100 mg/m² on day 1, and carboplatin at AUC 5 on day 1. Myelosuppression precluded more frequent weekly dosing of irinotecan. The overall response rate was 39% (10% complete response). Median time to progression was 6.8 months; median survival was 11 months, and 1-year survival was 46%. Forty patients were ultimately enrolled in the subsequent phase II effort using the maximum tolerated doses established in the phase I trial. The response rate of 32% proved comparable to the phase I effort. Time to progression at 5.6 months was similar; median survival exceeded 1 year.[25] The incidence of neutropenic fever, significant fatigue, and late diarrhea was 20% each. Based on these promising results, a randomized phase II study in fit stage IIIB/IV NSCLC patients assessing the triple-agent regimen and its constituent double-agent regimens (irinotecan and carboplatin and paclitaxel and irinotecan) using a 1:1:1 assignment was initiated. Unfortunately, because of sluggish accrual, this trial was shelved. Phase I trials have also assessed irinotecan in combination with docetaxel(Drug information on docetaxel) (Taxotere).[26] Thirty-two treatment- naive advanced NSCLC patients (22 stage IV) received docetaxel on day 2 and irinotecan on days 1, 8, and 15. Eligibility stipulated ECOG performance status 0 to 2, ≤ 75 years of age, and adequate marrow, hepatic, and renal function. Median age was 55 years (range: 46 to 71 years); 17 were male. Doses were escalated across sequential cohorts. The maximum tolerated dose for each agent was 50 mg/m². Higher doses led to dose-limiting toxicity, primarily neutropenia. The overall response rate was 37%, median survival was 48 weeks, and 1-year survival was 45%. There was no effect of irinotecan on docetaxel pharmacokinetics. Murren and colleagues evaluated weekly therapy using both agents.[27] Cohort A received weekly therapy × 4 every 6 weeks; cohort B received weekly therapy × 2 every 3 weeks. A total of 44 patients were accrued; 26 were previously treated with chemotherapy; 20 were diagnosed with NSCLC, 12 with pancreatic and biliary primaries, and 14 with other diagnoses. Seventy-five percent were male. All but seven were performance status 0 or 1. The maximum tolerated doses for irinotecan and docetaxel using the weekly × 4 regimen were 50 and 35 mg/m², respectively. The maximum tolerated doses for the weekly × 2 regimen were 60 and 35 mg/m², respectively.[25] From the standpoint of cumulative toxicity and dose intensity, the latter schedule was considered preferable. Adjei and colleagues[28] adopted an opposite tactic, dosing each agent at 3-week intervals. Irinotecan was escalated from 160 to 200 mg/m², and docetaxel from 50 to 75 mg/m² Eighteen patients were accrued, and 85 courses were administered. Dose-limiting toxicity proved to be myelosuppression; neutropenic fever occurred in three individuals. Diarrhea was dose-dependent, but severe only in those who neglected to take appropriate antidiarrheals. Nonhematologic toxicity was mild to moderate with grade 3 anorexia and nausea observed in one patient each. Five of 16 patients responded, including three with NSCLC and one each with soft tissue sarcoma and cholangiocarcinoma. The maximum tolerated doses and recommended phase II doses for irinotecan and docetaxel were 160 and 65 mg/m², respectively.[28] In a critical randomized phase II study conducted in Japan, this combination of docetaxel and irinotecan appeared equivalent to docetaxel and cisplatin[29] in the therapy of advanced NSCLC (Table 4). Treatment was cycled at 3-week intervals; patients received a fixed dose of docetaxel 60 mg/m² and were randomized to either cisplatin at 80 mg/m² or to irinotecan at 60 mg/m² on days 1 and 8.