Irinotecan: A Review of the Initial Phase I Trials

ABSTRACT: The unique mechanism of action of irinotecan (CPT-11 [Camptosar]), topoisomerase I inhibition, together with the results of preclinical studies, suggest that the drug’s antitumor and toxicologic effects may be schedule-dependent. To further explore this possibility, we reviewed the initial phase I studies of various administration schedules that have been conducted in Japan, France, and the United States. This review showed toxicities to be fairly consistent across dosing schedules, although the severity and extent of diarrhea and neutropenia differed somewhat. The institution of intensive loperamide therapy and perhaps myeloid growth factors may have allowed for further dose escalation on some schedules, although it is unclear whether dosing intensity should be pursued without regard to dosing frequency. Preliminary antitumor activity of irinotecan noted in a study of leukemia and lymphoma supports the theory that the drug may exhibit schedule-dependent antitumor activity. The results of these early studies of irinotecan should be taken into account when designing subsequent trials of the agent alone or in combination with other chemotherapeutics in specific tumor types. [ONCOLOGY 12(Suppl 6):31-38, 1998]

Introduction

Irinotecan (CPT-11 [Camptosar]) is a novel semisynthetic analog of camptothecin that has recently become commercially available and continues to undergo extensive clinical evaluations. The unique mechanism of action of this agent, topoisomerase I inhibition, combined with preclinical studies assessing its efficacy and toxicologic profile, suggested the possibility of schedule-dependent antitumor and toxicologic effects.

In order to gain further insight into the potential schedule-dependency of irinotecan, this article will review the initial phase I studies that were conducted (Table 1). For simplicity, the studies have been grouped according to country and then by administration schedule. Phase I schedules that were explored initially can be grouped into single-dose, intermittent, and prolonged-infusion regimens. Interestingly, the irinotecan schedule favored for subsequent efficacy trials differed among the regions: Whereas the single-dose infusion every 3 weeks was used in France, both the United States and Japan selected a weekly regimen.

The principal toxicities noted for all of the administration schedules were diarrhea and neutropenia (Table 2, Table 3, Table 4, and Table 5). In the single-dose study in France, the institution of an intensive loperamide regimen resulted in fewer episodes of serious diarrhea requiring hospitalization, and this regimen has since become part of standard supportive care.

Although the pharmacologic behavior of irinotecan will not be reviewed in detail in this article, pharmacodynamic relationships varied among the different studies and schedules (Table 6). Preliminary antitumor activity was noted in colon cancer refractory to fluorouracil (5-FU) and advanced non-small-cell lung cancer, as well as other tumors. An interesting phase I/II trial conducted in Japan in patients with refractory or relapsed leukemia or lymphoma revealed evidence of a schedule-dependent antitumor effect.

The results of the phase I trials of irinotecan reflect the unique properties of this agent and should be used in choosing the optimal schedule for future efficacy and combination studies.

Single-Dose Schedule

Japanese Study

One of the earliest schedules of irinotecan explored was a single intravenous (IV) dose administered every 3 to 4 weeks. In a small phase I study conducted in 17 patients in Japan, doses ranged from 25 to 350 mg/m² administered by a 30-minute IV infusion.[1]

The principal dose-limiting toxicity (DLT) was leukopenia, which appeared at doses ³ 100 mg/m² and became dose-limiting at 350 mg/m². The leukopenia occurred at approximately 7 days, with a median time to recovery of 16 days after the nadir. In addition, diarrhea was reported. The recommended phase II dose of 200 mg/m² was based on concerns about the combination of leukopenia and diarrhea at doses ³ 250 mg/m².

French Study

A second phase I study of irinotecan, conducted in France, extensively explored a single-dose administration schedule.[2] The initial dose-finding portion of this trial included 64 patients (median age, 51 years), most of whom (60/64, 94%) had received prior therapy. The doses studied ranged from 100 to 750 mg/m² administered by a 30-minute IV infusion every 3 weeks.

An acute cholinergic syndrome was reported, consisting predominantly of gastrointestinal cramps and diaphoresis at doses ³ 260 mg/m², as well as salivation, visual disturbances, and lacrimation at doses of 300 to 750 mg/m². This syndrome occurred during and for 1 hour after the drug infusion. Atropine (0.25 to 0.50 mg) administered subcutaneously ameliorated these symptoms in two patients.

Noncumulative diarrhea, which began approximately 6 days after the administration of irinotecan, became dose-limiting at 260 mg/m², and a regimen of intensive loperamide was instituted to allow for further dose escalation. The high-dose loperamide regimen, which is now standard, consisted of 2 mg orally at the first diarrheal episode, followed by 2 mg every 2 hours until the patient had been free of diarrhea for a 12-hour period. With the advent of this regimen, the incidence of grade 3 or 4 diarrhea, which had been 50% at 260 mg/m², remained < 20% until the 750-mg/m² dose level, at which point neutropenia became dose-limiting.

Severe (grade 3 or 4) nausea and vomiting were uncommon (9%) in this study. The only other common nonhematologic toxicity was asthenia, which was reported in > 50% of patients at doses ³ 600 mg/m².

Hematologic toxicity, consisting of neutropenia, was the DLT of this schedule. Neutropenia developed between days 6 and 9 of therapy, and white blood cell count recovered in approximately 5 days. The neutropenia exhibited both interpatient and intrapatient variability, was noncumulative and resulted in complications primarily in patients who were heavily pretreated or had bone metastases. Thrombocytopenia was rare, and only two patients experienced grade 4 events.

The dose recommended for subsequent phase II studies was 350 mg/m², although it was suggested that a dose of 500 mg/m² be explored further in a feasibility study using high-dose loperamide.

A total of 34 patients were assessable for response in this study. Two complete responses (cervical and head and neck cancers) and six partial responses (5-FU-refractory colon cancer) occurred, all in previously treated patients.

The pharmacologic profile of irinotecan and its active metabolite, SN-38, showed linear pharmacokinetics, long terminal half-lives (~14 hours) for both compounds, and a statistically significant correlation between the area under the curve (AUC) of irinotecan and that of SN-38 (r = 0.60; P < .001). Interestingly, irinotecan was detected in saliva, sweat, and pleural fluid. In addition, the appearance of plasma rebound concentrations of both irinotecan and SN-38 suggested enterohepatic recirculation.

Pharmacodynamic analysis revealed a relationship between the extent of neutropenia for both irinotecan and SN-38, whereas only diarrhea up to grade 2 correlated with the AUCs of irinotecan and SN-38. Of note, in the extension of this trial, performed to gain experience at the 500- and 600-mg/m² dose levels, it was determined that 600 mg/m2 was not tolerated, resulting in grade 3 or 4 diarrhea or neutropenia in 64% and 78% of patients, respectively.[3] The 500-mg/m² dose was considered acceptable in minimally pretreated patients, resulting in grade 3 or 4 diarrhea or neutropenia in 24% and 41% of patients, respectively.

US Study

In a third single-dose study performed in the United States, 32 patients (median age, 49 years) were treated with 100 to 345 mg/m² of irinotecan administered as a 90-minute infusion.[4] Of the 32 patients, 30 had received prior therapy, and 13 of these were considered heavily pretreated.

In contrast to the French and Japanese studies, the DLT in the US study consisted of a constellation of severe hematologic and nonhematologic events at doses ³ 290 mg/m². Gastrointestinal toxicity, consisting of nausea, vomiting, anorexia, abdominal cramps, or diarrhea, became particularly prominent at doses ³ 240 mg/m². The nausea and vomiting occurred and resolved primarily in the peritreatment period, whereas anorexia persisted for 2 to 3 weeks in some patients, resulting in significant weight loss and a decline in performance status.

Both early and late diarrhea occurred. The early form was associated with cholinergic symptoms of diaphoresis and flushing, whereas the late form occurred 2 to 7 days following treatment and was not associated with these symptoms. A premedication regimen was instituted at the 240-mg/m² dose level, consisting of diphenhydramine 50 mg IV and ondansetron (Zofran) 0.15 mg/kg IV, both administered prior to treatment and then every 4 hours for 2 doses. Despite this premedication, irinotecan doses above 240 mg/m² resulted in unacceptable gastrointestinal toxicity. An intensive loperamide regimen was not used.

Both neutropenia and thrombocytopenia were prominent hematologic toxicities, although they were rarely dose-limiting. The recommended dose for subsequent trials of this schedule was 240 mg/m².

Three objective responses (all partial responses) were noted in this trial, occurring in previously treated patients with colon, renal, and cervical cancer.

The pharmacologic profile of irinotecan revealed linear pharmacokinetic behavior over the dose range studied. The AUCs for the lactone forms of irinotecan and SN-38 comprised 44% and 51% of their respective total AUCs, whereas the mean AUC values for the total and lactone forms of irinotecan were 12- to 34-fold higher than the AUC values for the total and lactone forms of SN-38.

No pharmacodynamic relationship was established between the gastrointestinal symptoms and the AUC of either forms of irinotecan or SN-38. In contrast, the decrement in neutrophils correlated with the exposure to total SN-38, but not to total or lactone irinotecan.

Weekly Administration Schedule

Japanese Study

In Japan, the rationale for a weekly administration schedule of irinotecan came from preclinical studies in mouse L1210 leukemia, which showed that the most effective antitumor regimen was injections repeated at 5-day intervals. A phase I study of irinotecan administered as a weekly (no scheduled rest weeks) 90-minute IV infusion was conducted in 17 patients (median age, 64 years) with advanced non-small-cell lung carcinoma, 15 of whom had received no previous treatment.[5] In this study, 74 weekly doses of 50 to 150 mg/m² were given.

The most prominent toxicity was gastrointestinal, manifested as nausea, vomiting, and diarrhea. Grade 3 or 4 nausea or vomiting occurred in 25% and 100% of patients treated at the 125- and 150-mg/m² dose levels, respectively, whereas grade 3 or 4 diarrhea was reported in 25% and 50% of patients at these doses, respectively. Possible drug-induced pneumonitis, which responded to steroid therapy, developed in one patient after eight doses of irinotecan.

Myelosuppression was also prominent, with the leukocyte nadir occurring at days 21 to 29 of therapy and recovering by days 27 to 34. One patient each at the 125- and 150-mg/m² dose levels experienced grade 4 neutropenia, both coinciding with severe diarrhea and complicated by sepsis. Therefore, the recommended phase II dose on this schedule was 100 mg/m².

A total of 11 patients were assessable for a response. Two partial responses occurred in previously untreated patients at the 100- and 125-mg/m² dose levels.

The pharmacokinetic profile of irinotecan in this study revealed a nonlinear increase in the AUC with dose, as well as a decrease in clearance, suggesting nonlinear pharmacokinetics. This nonlinearity was consistent with the pharmacologic behavior of irinotecan in rodents, but differed from the results of other phase I studies.

Pharmacodynamic analysis failed to reveal any significant relationships between neutropenia or diarrhea and exposure to irinotecan or SN-38.

French Study

A French study explored another weekly schedule of irinotecan. In this study, 59 patients (median age, 54 years; 38 previously untreated with chemotherapy) received doses of 50 to 145 mg/m²administered as a weekly 30- to 90-minute IV infusion for 3 consecutive weeks with a 2-week rest period.[6]

Diarrhea was the DLT, although, as noted in other studies, there was considerable interpatient susceptibility. The diarrhea became more frequent and pronounced after the second and third weekly doses of irinotecan, with grade 3 or 4 events occurring in more than 45% of patients treated with doses > 115 mg/m²/wk.

In an attempt to ameliorate the severity of the diarrhea, the infusion duration was prolonged from 30 to 90 minutes beginning at the 115-mg/m² dose level. The prolonged infusion duration appeared to decrease the frequency of grade 3 events, whereas grade 4 events remained unchanged and were not controlled with standard measures. High-dose loperamide was not used in this study. There was a suggestion that the number of prior chemotherapy regimens correlated with the frequency of grade 2 to 4 diarrhea. Grade 3 or 4 nausea or vomiting (overall frequency, 17% of patients) occurred primarily at the higher dose levels and was controlled by the use of metoclopramide or serotonin antagonists.

Neutropenia was another prominent toxicity and was characterized by a nadir occurring between days 21 and 25, and recovery within 5 days. Grade 4 neutropenia occurred in only 2 of 17 patients (12%) at the 130-mg/m²dose level and 1 of 7 patients (14%) at the 145-mg/m² level, however, and was complicated by sepsis in one patient. There was no effect of the infusion duration on the frequency of neutropenia.

One patient with cervical cancer, previously treated with surgery and radiotherapy, achieved a partial response. The recommended dose for phase II studies was 100 mg/m² in high-risk patients (more than one prior chemotherapy regimen), and 115 mg/m² in low-risk patients.

In contrast to the weekly phase I Japanese study (and the preclinical pharmacokinetic studies in rodents), the pharmacokinetic profile of irinotecan in this study exhibited linear behavior, although there was significant interpatient variability. The correlation between the AUC of irinotecan and SN-38 reached statistical significance. One patient had sampling of bile performed at the 100-mg/m² dose level, revealing mean irinotecan and SN-38 levels that were 113- and 40-fold higher, respectively, than corresponding plasma levels. The biliary secretion of irinotecan and SN-38 was thought to contribute to the relatively long terminal half-lives of 9.3 and 7.7 hours, respectively.

Pharmacodynamic relationships were significant for exposure to either irinotecan or SN-38 and the extent of neutropenia or diarrhea.

US Study

In the United States, irinotecan was studied as a 90-minute IV infusion administered weekly for 4 consecutive weeks every 6 weeks. A total of 32 patients (median age, 55 years; [range 19-78]) previously treated with chemotherapy) were treated in this study with doses of 50 to 180 mg/m²/wk.[7]

Grade 4 diarrhea was the DLT, occurring in 4 (67%) of 6 patients at the 180-mg/m² dose level. Similar to previous weekly studies, diarrhea generally began following the second or third weekly treatment. When diarrhea was of grade 4 severity, it lasted 5 to 7 days and was unresponsive to standard measures. Intensive loperamide therapy was not used in this study. Grade 3 nausea and vomiting, which occurred in 1 (17%) of 6 patients at the 100-mg/m² dose level and 2 (33%) of 6 patients at the 180-mg/m² dose level, was subsequently controlled by the use of routine prophylaxis in all but one patient.

Serious hematologic toxicity was rare, with only one episode of uncomplicated grade 4 neutropenia occurring in one patient during the first course of therapy. Even when all cycles were analyzed, only 3 of 118 courses (3%) resulted in grade 4 neutropenia, all occurring in heavily pretreated patients at doses ³ 150 mg/m².

Since this study involved weekly treatment and thus weekly determination of dosing, the protocol was modified to allow for dose reduction (rather than just omission) of irinotecan based on the toxicities observed in the preceding weeks of treatment. This dose-modification regimen was incorporated into subsequent efficacy trials and is now commonly used. An analysis of the dose intensity of this study revealed that at doses > 150 mg/m², only an increase in toxicities was observed without a further increase in dose intensity. Thus, the dose recommended for phase II trials was 150 mg/m², which was associated with dose-limiting diarrhea in 1 (17%) of 6 patients.

Two partial responses occurred, both in patients with 5-FU-refractory colon cancer.

The pharmacokinetic profile of irinotecan was linear over the doses studied, and no differences in the parameters between weeks 1 and 4 were noted in the two patients sampled. Bile was sampled from one patient in this study, revealing concentrations of irinotecan and SN-38 that were 10- to 60-fold and 2- to 9-fold higher than the corresponding plasma concentrations, respectively.

Due to the small numbers of patients with grade 3/4 neutropenia or diarrhea, no pharmacodynamic relationships were established, although there was a trend toward an association between exposure to the SN-38 lactone and the extent of diarrhea.

Five-Day Continuous-Infusion Schedule

A 5-day continuous-infusion schedule of irinotecan was explored in Japan in an attempt to maximize the antitumor effect of this S-phase-specific agent, while perhaps minimizing its toxicity. A total of 36 patients (mean age, 57 years; 17 previously treated with chemotherapy) received doses of 5 to 40 mg/m²/d.[8]

Diarrhea was the DLT and appeared to be dose-dependent. Grade 3 or 4 events occurred in 2 of 6 patients (33%) at the 25-mg/m² dose level, as compared with 4 of 6 patients (67%) at the 40-mg/m² dose level. Leukopenia was also prominent but variable, with 2 of 6 patients (33%) experiencing grade 4 events at the 25-mg/m² dose level, whereas only grade 3 events were noted at the 40 mg/m2 dose level in 2 of 6 patients (33%). The recommended dose on this schedule was 30 mg/m²/d. No responses were reported.

The pharmacokinetic analysis of irinotecan and SN-38 in this study displayed great interpatient variability, and no correlation between the AUCs of irinotecan and SN-38 was noted. The pharmacodynamic analysis revealed a correlation between the severity of diarrhea and exposure to irinotecan but not SN-38, whereas leukopenia correlated with exposure to SN-38 but not irinotecan.

Daily for 3 Consecutive Days Schedule

The third dosing schedule explored in France was irinotecan administered by a 30-minute IV infusion for 3 consecutive days every 3 weeks.[9] In this study, 46 patients (median age, 56 years; 43 previously treated with chemotherapy) received 150 cycles at doses of 33 to 115 mg/m²/d.

Once again, diarrhea was determined to be dose-limiting, although the early and late forms were indistinguishable with this schedule. Severe diarrhea (grade 3 or 4) occurred in 8 of 11 patients (73%) at the 100-mg/m² dose level and 5 of 10 patients (50%) at the 115-mg/m² dose level, respectively. Grade 3 or 4 nausea and vomiting were also quite prominent, occurring in 9 of 21 patients (43%) at the two highest dose levels, although routine prophylaxis was not used.

Hematologic toxicity, consisting of neutropenia, was relatively uncommon at the lower dose levels. Grade 3 or 4 neutropenia was reported in only 1 of 16 patients (6%) treated with doses £ 75 mg/m², whereas at higher dose levels it became more prominent, occurring in 8 of 30 patients (27%) treated with ³ 85 mg/m². The leukocyte nadir occurred on day 8, with recovery usually by day 20. In this study, both the incidence and severity of neutropenia appeared to increase with repeated administration of irinotecan.

The recommended dose was 100 mg/m²/d. Two partial responses were reported in previously treated patients with breast cancer and carcinoma of unknown primary.

 Pharmacologic analysis revealed linear pharmacokinetic behavior and a significant correlation between the AUCs of irinotecan and SN-38. Plasma rebound concentrations of SN-38 were observed, suggesting enterohepatic recycling. The only statistically significant pharmacodynamic correlation observed was between exposure to irinotecan and the severity of diarrhea.

A Phase I/II Study in Leukemia and Lymphoma

A small but very interesting study was performed in Japan in patients with either primary refractory or relapsed Hodgkin’s lymphoma, non-Hodgkin’s lymphoma, or leukemia.[10] This trial enrolled 62 patients who had received previous chemotherapy or radiotherapy (median age, 46 years), 58 of whom were assessable for efficacy and 59 for toxicity. Of the 58 patients, 32 had lymphoma (multiple histologies) and 26 had either acute lymphoblastic leukemia (11 patients) or acute myelogenous leukemia (15 patients).

Four dosing schedules were explored: (1) 200 mg/m² IV every 3 to 4 weeks, (2) 40 mg/m² IV daily for 5 consecutive days every 3 to 4 weeks, (3) 40 mg/m² IV for 3 consecutive days every week, and (4) 20 mg/m² IV two times per day for 7 days every 3 to 4 weeks. Hematologic and nonhematologic toxicities were compatible with other studies and consisted primarily of neutropenia and diarrhea.

Of particular interest was the pattern of responses observed in this trial. Whereas the single-dose schedule was completely inactive against lymphoma, both the daily × 5 and daily × 3 weekly schedules were active, with response rates of 31% (2 complete responses, 3 partial responses) and 33% (2 complete responses, 1 partial response), respectively. In leukemia, the first three schedules were inactive (0 responses), whereas the twice-daily regimen resulted in a response rate of 25% (1 complete response, 2 partial responses) in a very treatment-resistant population. Although certainly a complex study with mixed populations, these preliminary results suggest that irinotecan may exhibit tumor-specific schedule-dependent efficacy--a theory that deserves further study.

Conclusions

In the phase I studies of irinotecan performed in Japan, France, and the United States, the toxicities appear to have been fairly consistent across dosing schedules, although the severity and extent of diarrhea and neutropenia may have differed. The institution of intensive loperamide therapy and perhaps myeloid growth factors may have allowed for further dose escalation on some schedules, although it is unclear whether dosing intensity should be pursued without regard to dosing frequency.

The preliminary antitumor activity noted in the study of leukemia and lymphoma lends support to the theory that irinotecan may exhibit schedule-dependent antitumor activity. This hypothesis should be explored further in phase II trials.

References:

1. Taguchi T, Wakui A, Hasegawa K: Phase I clinical study of CPT-11. Jpn J Chemother 17:115-120, 1990.

2. Abigerges D, Chabot GG, Armand JP, et al: Phase I and pharmacologic studies of the camptothecin analog irinotecan administered every 3 weeks in cancer patients. J Clin Oncol 13(1):210-221, 1995.

3. Merrouche Y, Extra JM, Abigerges D, et al: High-dose intensity of irinotecan administered every 3 weeks in advanced cancer patients: A feasibility study. J Clin Oncol 15(3):1080-1086, 1997.

4. Rowinsky EK, Grochow LB, Ettinger DS, et al: Phase I and pharmacological study of the novel topoisomerase I inhibitor 7-ethyl-10-[4-(1-piperidino)-1] carbonyloxycamptothecin (CPT-11) administered as a ninety-minute infusion every 3 weeks . Cancer Res 54:427-436, 1994.

5. Negoro S, Fukuoka M, Masuda N, et al: Phase I study of weekly intravenous infusions of CPT-11, a new derivative of camptothecin, in the treatment of advanced non-small-cell lung cancer. J Natl Cancer Inst 83(16):1164-1165, 1991.

6. de Forni M, Bugat R, Chabot GG, et al: Phase I and pharmacokinetic study of the camptothecin derivative irinotecan, administered on a weekly schedule in cancer patients. Cancer Res 54:4347-4354, 1994.

7. Rothenberg ML, Kuhn JG, Burris HA, et al: Phase I and pharmacokinetic trial of weekly CPT-11. J Clin Oncol 11(11):2194-2204, 1993.

8. Ohe Y, Sasaki Y, Shinkai T, et al: Phase I study and pharmacokinetics of CPT-11 with 5-day continuous infusion. J Natl Cancer Inst 84(12): 972-973, 1992.

9. Catimel G, Chabot GG, Guastalla JP, et al: Phase I and pharmacokinetic study of irinotecan (CPT-11) administered daily for three consecutive days every three weeks in patients with advanced solid tumors. Ann Oncol 6:133-140, 1995.

10. Ohno R, Okada K, Masaoka T, et al: An early phase II study of CPT-11: A new derivative of camptothecin for the treatment of leukemia and lymphoma. J Clin Oncol 8 (11):1907-1912, 1990.