|
Advertisement |
|
|
||
 |
|
|||||
|
|||||
|
 |
||||||
|
||||||
|
||||||
|
© 2000 American Society for Clinical Oncology Phase I and Pharmacokinetic Study of Irofulven, a Novel Mushroom-Derived Cytotoxin, Administered for Five Consecutive Days Every Four Weeks in Patients With Advanced Solid MalignanciesFrom the Institute for Drug Development, Cancer Therapy and Research Center, and Department of Medicine, Division of Oncology, University of Texas Health Science Center at San Antonio, San Antonio, TX, and MGI Pharma, Inc, Bloomington, MN. Address reprint requests to S. Gail Eckhardt, MD, Division of Oncology, University of Colorado Health Science Center, 4200 East 9th Ave, B171, Denver, CO 80262; email:gail.eckhardt{at}uchsc.edu
PURPOSE: To evaluate the toxicity and pharmacologic behavior of the novel mushroom-derived cytotoxin irofulven administered as a 5-minute intravenous (IV) infusion daily for 5 days every 4 weeks to patients with advanced solid malignancies. PATIENTS AND METHODS: In this phase I trial, 46 patients were treated with irofulven doses ranging from 1.0 to 17.69 mg/m2 as a 5-minute IV infusion (two patients received a 1-hour infusion) daily for 5 days every 4 weeks. The modified continual reassessment method was used for dose escalation. Pharmacokinetic studies were performed on days 1 and 5 to characterize the plasma disposition of irofulven. RESULTS: Forty-six patients were treated with 92 courses of irofulven. The dose-limiting toxicities on this schedule were myelosuppression and renal dysfunction. At the 14.15-mg/m2 dose level, renal dysfunction resembling renal tubular acidosis occurred in four of 10 patients and was ameliorated by prophylactic IV hydration. The 17.69-mg/m2 dose level was not tolerated because of grade 4 neutropenia and renal toxicity, whereas the 14.15-mg/m2 dose level was not tolerable with repetitive dosing because of persistent thrombocytopenia. Other common toxicities included mild to moderate nausea, vomiting, facial erythema, and fatigue. One partial response occurred in a patient with advanced, refractory metastatic pancreatic cancer lasting 7 months. Pharmacokinetic studies of irofulven revealed dose-proportional increases in both maximum plasma concentrations and area under the concentration-time curve, while the agent exhibited a rapid elimination half-life of 2 to 10 minutes. CONCLUSION: Given the results of this study, the recommended dose of irofulven is 10.64 mg/m2 as a 5-minute IV infusion daily for 5 days every 4 weeks. The preliminary antitumor activity documented in a patient with advanced pancreatic cancer and the striking preclinical antitumor effects of irofulven observed on intermittent dosing schedules support further disease-directed evaluations of this agent on the schedule evaluated in this study.
IROFULVEN (MGI 114, 6-hydroxmethylacylfulvene [HMAF]; Fig 1) is a semisynthetic analog of illudin S, a sesquiterpene derived from the fungus Omphalotus olearius. The impetus to develop the illudins was based on the results of early studies of illudin S that demonstrated notable activity against murine leukemias, but illudin S was only minimally active and induced significant toxicity in murine solid tumor models.1 Several unique features of the illudins, including rapid intracellular accumulation and cytotoxicity in multidrug-resistant tumor cell lines, led to structure-activity studies and screening to identify analogs with greater therapeutic indices.2-4 These efforts identified the acylfulvene illudin analog, irofulven, which had a greater therapeutic index than the parent compound, illudin S.5
Irofulven rapidly enters tumor cells, where it binds to cellular macromolecules and inhibits DNA synthesis.6-8 Treatment of the human CEM leukemic cell line with irofulven results in cell-cycle arrest in the S phase and inhibition of DNA, RNA, and protein synthesis with 50% inhibitory concentration values of 2, 20, and 70 µmol/L, respectively.8 Radiolabeled irofulven is localized predominantly in the nuclear compartment followed by the cytosol and membrane compartments; more than 60%, 27%, and 11% of drug binds to proteins, DNA, and RNA, respectively.7 Irofulven induces cytotoxicity principally by generating DNA strand breakage, as exposure of CEM cells to the drug does not result in the formation of DNA intrastrand cross-links or DNA-protein cross-links.7 This finding contrasts with the observation that illudin S can act as a bifunctional alkylator in vitro.9 The most unique aspect of irofulven’s antitumor activity seems to be its ability to act as a selective inducer of apoptosis in human tumor cell lines, and in contrast to conventional antitumor agents, irofulven retains this activity against tumor cell lines regardless of their p53 or p21 expression.10,11 The illudins and acylfulvenes also differ from other agents with alkylating activity in their preferential cytotoxicity toward cell lines deficient in the excision repair cross-complementing (ERCC) DNA repair enzymes ERCC2 and ERCC3, which indicates that the repair of irofulven-induced DNA damage requires a full complement of DNA repair mechanisms that may be absent in sensitive cell lines.12-14 Irofulven has demonstrated impressive antitumor activity against a broad range of human malignancies in vitro and in vivo. At concentrations less than 1 µmol/L, irofulven induced 50% growth inhibition of 51 of 52 solid tumor cell lines in the National Cancer Institute’s tumor panel screen.14 Irofulven also induced cytotoxicity in tumor cell lines exhibiting resistance to a multitude of anticancer agents with 50% inhibitory concentration values of 0.4 to 8.81 µmol/L.6,14,15 Against surgically derived, primary human tumor explants in a tumor-cloning assay, irofulven exposure for 1 hour resulted in inhibition of colony formation by greater than 50% in 41 of 100 and 83 of 100 of tumors exposed to concentrations of 100 and 1,000 ng/mL, respectively.16 Irofulven demonstrated prominent activity against a wide variety of human tumor xenograft models when administered on a daily-times-five schedule, including induction of complete and partial responses in relatively chemotherapy-resistant tumors, such as HT-29 colon, MV522 lung, SK-OV-3 ovarian, MiaPaCa pancreatic, and PC-3 and DU-145 prostate carcinomas.15,17-19 Irofulven was also active against the non–small-cell lung carcinoma xenograft MV522 transfected with the human drug resistance genes mdr1 (gp170) and MRP.20,21 In preclinical toxicology studies, irofulven primarily affected lymphoid and hematopoietic tissues.14 In dogs, the species most sensitive to the effects of irofulven, intravenous (IV) administration for 5 consecutive days resulted in emesis, neutropenia, and thrombocytopenia. The hematologic toxicity was maximal at day 15 and recovered completely by days 23 to 44. Fully reversible, noncumulative cardiac toxicity, consisting of multifocal myofibrillar necrosis, was observed only in rats and seemed to be the result of enhanced uptake and/or retention of irofulven in rat myocardial tissue in comparison to canine or human myocardial tissue in vitro.14,22 Preclinical pharmacology studies revealed that irofulven was not highly bound to human plasma proteins (55% to 63%) and exhibited a very short elimination half-life (t1/2) in mice, rats, dogs, and monkeys (1.7 to 5.8 minutes).14 Studies with radiolabeled [14C]-irofulven demonstrated very rapid and wide tissue distribution, with the highest concentrations observed in liver, kidneys, stomach, and adrenal glands.14 The decision to pursue the clinical development of irofulven was based on the agent’s novel mechanism of DNA interaction and broad spectrum of antitumor activity against neoplasms exhibiting resistance to standard cytotoxic agents. A short, 5-minute, IV infusion administered daily for 5 days every 4 weeks was selected to mimic the dosing schedule associated with superior antitumor activity in preclinical studies. The principal objectives of this phase I and pharmacokinetic study were to (1) characterize the toxicities of irofulven administered daily for 5 days every 4 weeks in patients with advanced solid malignancies, (2) determine the maximum-tolerated dose (MTD) and recommended dose for subsequent phase II trials, (3) characterize the pharmacokinetic behavior of irofulven, and (4) seek preliminary evidence of antitumor activity in patients with advanced cancers.
Patient Selection Patients with histologically documented solid malignancies refractory to standard therapy or for whom no effective therapy existed were eligible for this study. Other relevant eligibility criteria included the following: (1) age  18 years; (2) Eastern Cooperative Oncology Group (ECOG) performance status of  2 (ambulatory and capable of self-care); (3) no chemotherapy or investigational agents within 4 weeks (6 weeks for nitrosoureas or mitomycin); (4) adequate hematopoietic function (absolute neutrophil count 1,500/µL, hemoglobin level 9.0 g/dL, platelet count > 100,000/µL), hepatic function (total bilirubin level 2.0 mg/dL; AST and ALT concentrations 2.0 times the institutional upper limit of normal, unless due to hepatic metastases, in which case elevations 5.0 times the institutional upper limit of normal were permitted), and renal function (creatinine concentration 2.0 mg/dL); (5) no prior radiation therapy to more than 25% of hematopoietic reserves; (6) no prior high-dose chemotherapy requiring hematopoietic stem-cell reinfusion; (7) no significant cardiac disease, including arrhythmias requiring control by medication, ischemic events within the preceding 6 months of study enrollment, significantly abnormal ECG, or prior history of cardiac failure; (8) no prior therapy with doxorubicin 300 mg/m2 or an equivalent dose of an anthracycline; and, (9) no coexisting medical problem of sufficient severity to limit full compliance with the study. Informed consent was obtained according to federal and institutional guidelines.
Drug Administration Irofulven was supplied by MGI Pharma, Inc (Bloomington, MN) in 10-mL glass vials containing 10 mg of sterile, vacuum-dried product. The drug was initially reconstituted in 0.1 mL of dehydrated alcohol (United States Pharmacopeia) followed by the addition of 9.9 mL of 5% dextrose in water. The total dose was then aspirated into a syringe and administered over 5 minutes using a Harvard pump (Harvard Apparatus, South Natick, MA). The final solution was stable for 4 hours at room temperature in a syringe.
Pretreatment and Follow-Up Studies A formal assessment of disease was performed before treatment and then after every other course. A complete response was defined as the disappearance of all measurable or assessable disease for at least two measurement periods separated by at least 4 weeks, without worsening of cancer-related symptoms or performance status. A partial response required at least a 50% reduction in the sum of the bidimensional products of all measurable lesions documented by at least two measurements separated by at least 4 weeks. Patients were allowed to continue treatment in the absence of disease progression or intolerable toxicity. Patients were stratified according to the extent of prior myelosuppressive therapy prospectively after the fifth patient was enrolled at the 14.15-mg/m2 dose level and retrospectively in all patients at the 8.0- to 17.69-mg/m2 dose levels, for the purposes of analysis. Patients were considered heavily pretreated if they had received (1) 25% irradiation of bone marrow–containing areas, (2) more than six courses of alkylating agent–containing combination chemotherapy, (3) any prior therapy with a nitrosourea, or (4) more than two courses of treatment with mitomycin or carboplatin. Additionally, patients with extensive bone metastases were considered to be heavily pretreated.
Pharmacokinetic Sampling and Assay Plasma was assayed for irofulven by reverse-phase HPLC. The HPLC system consisted of an Accu-Flow III pump (Fisher Scientific, Inc, Itasca, IL) and a model AI-1A auto-injector (Rainin, Ridgefield, NJ) with an injection volume of 100 µL. An Alltima C18 analytic column (4.6 x 25 cm; particle size, 5 µm; Alltech, Deerfield, IL), which was maintained at 40°C, and an RP-18 prefilter (15 x 3.2 mm; particle size, 7 µmol/L; Alltech) were used. The mobile phase consisted of a mixture of acetonitrile, methanol, and water (25:15:60, v/v/v), and the flow rate was set at 1.0 mL/min. Detection was performed with a model 773 ultraviolet absorbance detector (Kratos, Ramsey, NJ) operating at 330 nm. The data-processing system consisted of an HP 1000, model A990, with the 3350A Laboratory Automation System (Hewlett-Packard, Palo Alto, CA). Immediately after thawing of plasma samples, the plasma was acidified with 0.01 mol/L sodium acetate buffer solution (pH 5), and the internal standard, illudin S, was then added. Next, 0.7 g of solid sodium chloride was added to the mixture. Irofulven was extracted from plasma by the addition of a solution of methylene chloride and pentane (1:1, v/v). The organic layer was evaporated to dryness under a stream of nitrogen at room temperature, and the residue was reconstituted with mobile phase and analyzed by HPLC. The recoveries of irofulven from plasma (means) ranged from 84.2% to 88.6%. The average recovery of internal standard from plasma was 46.3%. Calibration curves were generated by adding irofulven to 1.5 mL of blank plasma, which resulted in final concentrations that ranged from 2 to 2,500 ng/mL. The mixtures were then processed and extracted as described previously. The intra-assay precision (coefficient of variation) and accuracy (bias as percentage of nominal) of 3-day validation ranged from 1.2% to 4.8% and 2.0% to 11.0%, respectively, for the 10.0, 100, and 1,000 ng/mL quality control samples.
Pharmacokinetic Analysis Irofulven pharmacokinetic parameters were summarized using descriptive statistics. The relationships between irofulven dose and both Cmax and AUC were assessed by univariate linear regression. The paired t test was used to compare pharmacokinetic parameters between days 1 and 5. Univariate analysis of variance was used to determine the relationship between (1) irofulven pharmacokinetic parameters and dose, (2) irofulven exposure and toxicity, and (3) estimated creatinine Cl and toxicity. When the means of more than two groups were being compared, the Tukey-Kramer test was used to determine which means were significantly different. A significance level of .05 was set for all analysis. Statistical analysis was performed using the JMP version 3.1 statistical software program (SAS Institute, Cary, NC).
Forty-six patients received a total of 92 courses of irofulven at doses ranging from 1.0 to 17.69 mg/m2/d. All patients were assessable for toxicity. Four patients failed to complete the 5 days of therapy, but drug toxicity (grade 3 nausea) was the cause in only one of these patients. Patient characteristics are listed in Table 1. The dose escalation scheme, as well as the number of patients and courses administered as a function of dose level, is reported in Table 2. The median number of courses administered per patient was two (range, one to seven). Three new patients were treated at the initial dose level of 1.0 mg/m2/d followed by a single new patient per dose level until the 3.38-mg/m2 dose level was reached, at which point three patients were enrolled due to grade 3 vomiting. Additional patients were enrolled at the 4.5-mg/m2 and 8.0-mg/m2 dose levels due to grade 2 phlebitis and grade 3 fatigue, respectively. Initially, only one patient was treated at the 10.64-mg/m2 dose level; however, the enrollment at this dose level was subsequently expanded to include 12 patients at the recommended phase II dose. Fourteen new patients were treated at the 14.15-mg/m2 dose level due to renal toxicity, which is described in detail below, and only one patient was treated at the 17.69-mg/m2 dose level owing to concerns about renal toxicity and dose reductions. At doses of 14.15 mg/m2/d and above, five of 13 assessable patients (38%) required dose reductions by one to three dose levels for persistent nausea (one patient), renal toxicity (two patients), or persistent grade 3 thrombocytopenia (two patients). No new patients at the 10.64-mg/m2 dose level required dose reduction.
Hematologic Toxicity Thrombocytopenia was the principal hematologic toxicity of irofulven on this administration schedule. The median, range, and distribution of platelet counts, as well as the National Cancer Institute grade of the nadir platelet and neutrophil counts as a function of dose level ( 8.0 mg/m2) and prior therapy classification, are listed in Table 3. Grade 3 and 4 thrombocytopenia and persistent thrombocytopenia were not more frequent in patients with extensive prior therapy. The platelet nadirs typically occurred between days 19 and 26 and resolved by day 28 in 63 (74%) of 85 assessable courses. A treatment delay of 1 to 3 weeks was required in 17 (50%) of 34 subsequent courses at doses of 8.0 mg/m2/d and above; two (25%) of eight, three (33%) of nine, and three (43%) of seven courses were delayed in new patients treated at the 8.0-, 10.64-, and 14.15-mg/m2 dose levels, respectively. Treatment delays occurred in another 11 courses after dose reductions from the 17.69-, 14.15-, and 10.64-mg/m2 dose levels. A trend toward cumulative thrombocytopenia was suggested, and although thrombocytopenia was prominent in this study, it was a DLT (grade 4) in only one patient treated at the 14.15-mg/m2 dose level.
Severe neutropenia was uncommon with irofulven. At the 8.0-, 10.64-, and 14.15-mg/m2 dose levels, severe (grade 3 or 4) neutropenia occurred in only 0%, 14%, and 12% of courses, respectively, and was dose-limiting (grade 4) in only two patients. The first patient received irofulven 14.15 mg/m2 over 1 hour and developed grade 4 neutropenia concurrent with grade 4 thrombocytopenia, whereas the second episode occurred in the single minimally pretreated patient at the 17.69-mg/m2 dose level and was not associated with fever. Anemia resulting in a greater than 4% decrement in hemoglobin was observed in 51 courses (55%). The anemia seemed to be dose-related and cumulative, and transfusions (in the absence of bleeding from underlying disease) were required in eight patients (17%).
Renal Toxicity
Fluid retention, manifested by weight gain and/or facial/lower extremity edema, occurred in 22 (32%) of 68 courses at doses 8.0 mg/m2. Patients generally reported peripheral edema during or just after the 5 days of treatment, with resolution during the following week.
Other Toxicities
The two patients who were treated with a dose of 14.15 mg/m2 over a 1-hour IV infusion experienced profound nausea, vomiting, and anorexia. One patient refused further therapy after two doses of irofulven because of intolerable nausea (grade 3), despite additional IV and oral antiemetics. The other patient experienced significant nausea, vomiting, and anorexia and was hospitalized on days 5 through 12 for IV hydration and supportive care. One episode of grade 3 fatigue occurred at the 8.0-mg/m2 dose level, whereas dose-related grade 1 or 2 fatigue was reported in 57 (62%) of 92 courses and was most prominent during the week of treatment and the week after treatment. Irofulven induced facial erythema in 22 (24%) of 92 courses which generally appeared during the first day of treatment and persisted throughout the 5 days of therapy. Grade 1 or 2 diarrhea was reported in 18 (20%) of 92 courses, occurred between days 5 and 12, and was ameliorated with loperamide in six courses. Mild or moderate phlebitis was noted after five (29%) of 17 irofulven infusions in which the agent was administered through a peripheral vein. The phlebitis was dose-related, associated with erythema, pain, and induration, and resolved over 1 to 2 weeks. Patients enrolled at doses greater than 8.0 mg/m2 were urged, although not required, to have a central line placed.
Antitumor Activity
Pharmacokinetic Studies Mean plasma concentration-time profiles on days 1 and 5 after administration of irofulven 10.64 mg/m2 and 14.15 mg/m2 as a 5-minute infusion are shown in Fig 2. The average time of the last measurable irofulven concentration was 35 minutes. The pharmacokinetics of irofulven on day 1 were characterized by a mean t1/2 of 4.2 minutes, Vss of 54 L, and Cl of 9.7 L/min (Table 6). Irofulven Cmax and AUC increased proportionally with dose, but substantial interpatient variability (five- to nine-fold) was observed at the 8.0-, 10.64-, and 14.15-mg/m2 dose levels (Fig 3). Day 5 pharmacokinetic parameters were not assessable for the one patient who received irofulven 17.69 mg/m2. At the dose levels with three or more observations, irofulven pharmacokinetic parameters were similar on days 1 and 5 and were independent of dose. After administration of irofulven 14.15 mg/m2 as a 1-hour infusion, the average values were Cmax of 63 ng/mL, t1/2 of 7.9 minutes, Vss of 268 L, Cl of 11 L/min, and AUC of 2,557 ng·min/mL, and concentrations remained above 2.0 ng/mL, the assay limit of quantification, for 75 to 90 minutes.
Pharmacodynamic Analysis No relationship between irofulven exposure and any parameter of hematologic toxicity during course 1 was noted. The relationship between irofulven exposure and pretreatment estimated creatinine Cl and clinical (RTA) is shown in Fig 4. At the 10.64- to 17.69-mg/m2 dose levels, patients who experienced RTA (n = 5) had significantly lower estimated creatinine Cls (mean, 64 mL/min; range, 45 to 101 mL/min) than those who did not (n = 17; mean, 97 mL/min; range, 57 to 171 mL/min; P = .039). Irofulven AUC values on day 5 in patients who experienced RTA tended to be higher (mean, 4,342 ng·min/mL; range, 3,484 to 5,069 ng·min/mL) than AUC values in those who did not (mean, 3,337 ng·min/mL; range, 997 to 7,772 ng·min/mL; P = .292). Irofulven Cmax values were not related to the development of RTA (P > .7). No relationship between irofulven Cl and estimated creatinine Cl was observed. Examination of the relationship between irofulven dose level and the decline in renal function during course 1 revealed larger percentage decrements in estimated creatinine Cl at dose levels 8.0 mg/m2 (mean, 27%; range, 0% to 70%) than those observed at dose levels less than 8.0 mg/m2 (mean, 2.9%; range, 0% to 13%; P < .0001). No other irofulven exposure parameters (eg, Cmax and AUC) were related to decline in renal function.
The illudins comprise a unique class of mushroom-derived cytotoxins characterized by rapid intracellular accumulation and cytotoxicity against human tumor cell lines exhibiting the multidrug resistance phenotype. Irofulven, an acylfulvene illudin analog, was the product of a structure-activity–based screening program to identify analogs of illudin S with a greater therapeutic index. In preclinical studies of irofulven in vitro, the agent was shown to induce apoptotic strand breakage of DNA, whereas cells with specific deficiencies in DNA repair enzymes were particularly susceptible to irofulven-induced cytotoxicity.8,14 In vivo, irofulven demonstrated notable antitumor activity against a broad range of human malignancies, including those resistant to conventional chemotherapeutic agents.15,17-21 This phase I and pharmacologic study was designed to evaluate the feasibility of administering irofulven as a short, 5-minute IV infusion daily for 5 days every 4 weeks, in order to mimic the schedule that was associated with superior antitumor activity in vivo. The principal toxicities of irofulven on this administration schedule were renal dysfunction, thrombocytopenia, and refractory nausea and vomiting. Although the renal dysfunction observed at the 14.15-mg/m2 dose level could be ameliorated by the use of IV hydration, this dose level was associated with substantial nausea and vomiting, while repetitive administration resulted in persistent thrombocytopenia. For example, four of five patients enrolled at the 14.15-mg/m2 dose level required dose reductions by one to two dose levels for protracted nausea or persistent thrombocytopenia. The 10.64-mg/m2 dose level was better tolerated; no patients enrolled at this dose level experienced significant renal toxicity, and of the five patients who received subsequent courses of therapy, no dose reductions were required. Thus, the recommended dose of irofulven for subsequent disease-directed studies on this schedule is 10.64 mg/m2. Although the MCRM was used in this study, in practice, neither patient resources nor study duration was reduced. Dose selection by the MCRM was frequently overruled by the clinical decision to expand a dose level to characterize a toxicity or to stratify patients according to particular risk factors. This was best illustrated at the 14.15 mg/m2-dose level, at which 14 patients were treated and stratified according to prior myelosuppressive therapy and renal function. Additionally, the prior distribution and dose toxic-response model, which was based on preclinical studies, limited the ability of the model to assess the probability of DLT at dose levels greater than 5 mg/m2, since this dose was predicted to have a 95% probability of exceeding the MTD. Finally, not all derivations of the MTD are based on absolute presence or absence of DLT, and as illustrated in this study, the MTD may be derived based on a constellation of toxicities that do not strictly satisfy the a priori definition of DLT. In a recent assessment of the use of the MCRM in trials conducted in San Antonio, TX, we concluded that the MCRM did not result in more rapid study completion because of the mandatory waiting periods between patients to observe toxicity.25 Additionally, the ability to accomplish single-patient dose escalation was found to be most successful with agents that were noncytotoxic, where the starting dose was substantially below the MTD. These and other concerns have stimulated the discussion and evolution of alternative dose escalation schemes and further refinements of the continual reassessment method in phase I studies.26 The principal hematologic toxicity in the present study, thrombocytopenia, was similar to the major toxicity observed in preclinical studies in dogs.27 Interestingly, it was the duration of the thrombocytopenia, rather than the magnitude, that was of the greatest clinical significance. For example, prolonged (> 5 days) thrombocytopenia occurred in 30% of courses, and 46% of subsequent courses were delayed because of persistent thrombocytopenia. Moreover, in the few patients who received multiple courses, decreases in the nadir platelet values were observed over time. Patients who initially received doses of irofulven greater than 10.64 mg/m2 continued to exhibit persistent thrombocytopenia despite dose reductions by one to two dose levels, which suggests a dose-related irreversible effect on platelets. By contrast, the neutropenia associated with irofulven treatment was not clinically significant, and only two courses were complicated by grade 4 neutropenia. The selective late-occurring toxic effects of irofulven on platelets suggest that the target cell is likely at the level of the committed megakaryocyte precursor, rather than a more primitive pluripotent stem cell. Although patients at the four highest dose levels were stratified according to prior myelosuppressive therapy, there was no clear relationship between extent of prior therapy and myelosuppression. However, the analysis was limited by the small number of patients who received multiple cycles of therapy, as the principal hematologic toxicity of irofulven was prolonged and cumulative thrombocytopenia. For example, of the patients treated at the four highest dose levels who received subsequent courses, six patients were heavily pretreated but received only 15 total cycles (2.5 cycles per patient), whereas the six minimally pretreated patients received a total of 34 courses (5.6 courses per patient). Clinically significant renal toxicity from irofulven did not become evident until the 14.15-mg/m2 dose level and was characterized by mild to moderate elevations in the serum creatinine and blood urea nitrogen. The clinical picture was consistent with acute nonoliguric renal failure with prompt resolution. In contrast to typical drug-induced acute tubular necrosis, no urinary casts were observed, and the rapid recovery of renal function suggests a more transient effect of irofulven on glomerular filtration. In five patients, a clinical syndrome resembling RTA also occurred, which suggests that there may have been dual effects of irofulven on both glomerular filtration and renal tubular function. The nephrotoxic effects of irofulven may be compatible with the 1.7-fold higher concentration of [14C]-irofulven demonstrated in the renal tissue of rats compared with plasma, which indicates that irofulven may preferentially distribute into renal tissue. IV hydration may provide optimal amelioration by essentially clearing parent compound and/or metabolites from the kidney while maintaining adequate glomerular filtration. The other principal toxicities of irofulven were significant nausea and vomiting. The pattern of the early-onset emesis was characteristic of that which occurs after activation of the chemoreceptor trigger zone in the medulla, and the emesis was likewise controlled by the administration of IV prophylaxis consisting of a serotonin antagonist and dexamethasone. However, at doses greater than 10.64 mg/m2, a different pattern of toxicity occurred that was characterized by protracted nausea, vomiting, and anorexia, which were poorly managed with antiemetics. These data illustrate the emetogenic potential of irofulven and support the use of IV antiemetic prophylaxis before dosing to prevent acute nausea and vomiting in addition to careful monitoring of patients for delayed nausea and vomiting. Interestingly, studies of the uptake of irofulven labeled with 14C in male rats demonstrated a two-fold greater distribution of irofulven into the stomach compared with plasma, suggesting that irofulven or a metabolite may preferentially distribute into gastric tissue, thereby inducing gastric receptor afferent–mediated nausea and vomiting. The one objective response that occurred in this trial was in a patient with refractory pancreatic cancer and extensive hepatic metastases. Consistent with this observed clinical activity in pancreatic adenocarcinoma, irofulven has also demonstrated in vitro and in vivo effects against the human pancreatic cell lines MiaPaca and Panc-1. The activity of this agent against this refractory human malignancy has also prompted the initiation of phase II trials of irofulven in gemcitabine-refractory advanced pancreatic cancer and suggests that the agent may be active against other refractory tumors. The pharmacokinetic profile of irofulven after the administration of a single dose was characterized by a t1/2 of 4.2 (± 2.3) minutes and rapid plasma Cl. This rapid elimination is consistent with a compound that undergoes rapid intracellular accumulation and/or metabolic conversion in blood or other tissues. Irofulven exhibited dose-proportional increases in Cmax and AUC, although substantial interpatient variability was demonstrated in these exposure parameters (five- to nine-fold). The extent of interindividual variability observed may be due, in part, to the difficulty in obtaining consistent end-of-infusion samples with a drug that is eliminated so rapidly from the plasma, or it may be due to substantial interpatient differences in the metabolism of irofulven. Such variability could represent a challenge during phase II studies, resulting in unanticipated toxicity and/or suboptimal efficacy. Despite this variability, parameters were strikingly similar between days 1 and 5 at the 10.64- and 14.15-mg/m2 dose levels, which resulted in almost superimposable mean concentration-time profiles. Thus, there was no evidence of plasma accumulation of irofulven. Interestingly, in three patients, irofulven plasma concentrations were measurable at baseline on day 5, and further analysis revealed the possibility of a coeluting substance. These patients were also noted to be taking concomitant medications with the potential to interact with the cytochrome P450 system. The mean day 1 and day 5 Cmax concentrations achieved in this study at the recommended dose level (320 ± 219 and 556 ± 26 ng/mL, respectively) are within the range of those required for in vitro cytotoxicity (ie, 39 to 295 ng/mL [0.16 to 1.2 µmol/L]). To more closely mimic the in vitro studies of irofulven, a 1-hour infusion of 14.15 mg/m2 was attempted in two patients, resulting in unacceptable renal and gastrointestinal toxicity. Since the Cmax values were lower and the AUC values were similar to those achieved after administration of irofulven over 5 minutes, the severe toxicity observed after the 1-hour infusion was most likely not related to peak plasma concentration but possibly to maintaining parent drug or metabolite plasma concentrations above a cytotoxic threshold concentration for a longer period of time. Although the preclinical in vitro studies with irofulven demonstrated antitumor effects with longer exposure durations (1 to 48 hours) than that observed in this study, direct comparisons are difficult because of the rapid intracellular penetration of the compound and the fact that plasma concentrations may not reflect intracellular concentrations and disposition.6-8 Additionally, irofulven demonstrated substantial in vivo antitumor efficacy in rodent models despite a rapid t1/2 of 5 to 5.7 minutes.15,17-19 The ability to analyze relationships between the pharmacokinetic parameters of irofulven and standard pharmacodynamic end points may have been hampered by the rapid disappearance of irofulven from plasma, the difficulty in obtaining reliable end-of-infusion samples, and thus difficulty in determining accurate measurements of AUC. No relationship was established between any pharmacologic parameter of irofulven and the hematologic toxicities observed in this trial. However, a relationship was demonstrated between the occurrence of RTA in five patients and baseline creatinine Cl, although the Cl of the parent compound, irofulven, was not related to renal function. These observations suggest that there may be a renally excreted, potentially nephrotoxic metabolite. The analysis and confirmation of such relationships may be facilitated by pharmacokinetic studies involving larger numbers of patients and by further investigation of irofulven metabolites. The results of this phase I and pharmacologic study demonstrate that irofulven on a daily-times-five schedule every 4 weeks results in a complex, albeit manageable, toxicity profile and plasma concentrations associated with preclinical activity. The dose of irofulven on this schedule that is recommended for subsequent disease-directed studies is 10.64 mg/m2, based on tolerable nonhematologic effects noted at this dose level and the ability to administer this dose repetitively without dose reductions or delays caused by thrombocytopenia. Given the antitumor activity observed in a patient with advanced gemcitabine-refractory pancreatic cancer, further evaluation of irofulven in this disease is appropriate.
Supported in part by a grant from MGI Pharma Inc and grant no. CA 54174 from the San Antonio Cancer Institute.
1. Kelner MJ, McMorris TC, Beck WT, et al: Preclinical evaluation of illudins as anticancer agents. Cancer Res 47: 3186-3189, 1987
2.
Kelner MJ, McMorris TC, Taetle R: Preclinical evaluation of illudins as anticancer agents: Basis for selective cytotoxicity. J Natl Cancer Inst 82: 1562-1565, 1990 3. McMorris TC, Kelner MJ, Wang W, et al: Structure-activity relationships of illudins: Analogs with improved therapeutic index. J Org Chem 57: 6876-6883, 1992 4. Kelner MJ, McMorris TC, Taetle R: In vitro and in vivo studies on the anticancer activity of dehydroilludin M. Anticancer Res 15: 873-878, 1995[Medline] 5. McMorris TC, Kelner MJ, Wang W, et al: Hydroxymethylacylfulvene (HMAF), an illudin derivative with superior antitumor properties. J Nat Prod 59: 896-899, 1996[Medline] 6. Kelner MJ, McMorris TC, Montoya MA, et al: Characterization of MGI 114 (HMAF) histiospecific toxicity in human tumor cell lines. Cancer Chemother Pharmacol 44: 235-240, 1999[Medline] 7. Herzig MC, Arnett B, MacDonald JR, et al: Drug uptake and cellular targets of hydroxy methylacylfulvene (HMAF). Biochem Pharmacol 58: 217-225, 1999[Medline] 8. Woynarowski JM, Napier C, Koester S, et al: Effects on DNA integrity and apoptosis induction by a novel antitumor sesquiterpene drug, 6-hydroxymethylacylfulvene (HMAF, irofulven). Biochem Pharmacol 54: 1181-1193, 1997[Medline] 9. McMorris TC, Kelner MJ, Chandra RK, et al: Structure and reactivity of illudins. Tetrahedron 45: 5433-5440, 1989 10. Woynarowska BA, Woynarowski JM, Herzig MCS, et al: Differential cytotoxicity and induction of apoptosis in tumor and normal cells by hydroxymethylacylfulvene (HMAF). Biochem Pharmacol 59: 1217-1226, 2000[Medline] 11. Izbicka E, Davidson K, Von Hoff DD: MGI-114 antiproliferative effects against human tumor cell lines with different p53 and p21waf expression. Anticancer Res 19: 1299-1308, 1999[Medline] 12. Kelner MJ, McMorris TC, Estes L: Characterization of illudin S sensitivity in DNA repair-deficient Chinese hamster cells: Unusually high sensitivity in ERCC2 and ERCC3 DNA helicase-deficient mutants compared to other chemotherapeutic agents. Biochem Pharmacol 48: 403-409, 1994[Medline]
13.
Kelner MJ, McMorris TC, Estes L, et al: Efficacy of acylfulvene illudin analogues against a metastatic lung carcinoma MV522 xenograft nonresponsive to traditional anticancer agents: Retention of activity against various mdr phenotypes and unusual cytotoxicity against ERCC2 and ERCC3 DNA helicase-deficient cells. Cancer Res 55: 4936-4940, 1995 14. MGI 114 [clinical investigator’s brochure]. Bloomington, MN, MGI PHARMA, Inc, 1999
15.
MacDonald JR, Muscoplat CC, Dexter DL, et al: Preclinical antitumor activity of 6-hydroxymethlacylfulvene, a semisynthetic derivative of the mushroom toxin illudin S. Cancer Res 57: 279-283, 1997 16. Hidalgo M, Izbicka E, Cerna C, et al: Antitumor activity of 6-hydroxymethylacylfulvene (HMAF), a semisynthetic derivative of illudin S, against adult and pediatric human tumor colony-forming units. Anticancer Drugs 10: 837-844 1999[Medline] 17. Kelner MJ, McMorris TC, Estes L, et al: Efficacy of HMAF (MGI 114) in the MV522 metastatic lung carcinoma xenograft model nonresponsive to traditional anticancer agents. Invest New Drugs 14: 161-167, 1996[Medline] 18. Marty J, MacDonald JR, Von Hoff DD, et al: Antitumor activity of MGI 114 against human ovarian and pancreatic tumor xenograft models. Proc Am Assoc Cancer Res 40: 123, 1999 (abstr)
19.
Mangold G, MacDonald JR, Von Hoff DD, et al: Antitumor activity of MGI 114 against two human prostate tumor xenograft models. Proceedings of the 10th NCI-EORTC Symposium on New Drugs in Cancer Therapy. Ann Oncol 9: 37, 1998 (suppl 2, abstr) 20. Kelner MJ, McMorris TC, Estes L, et al: Efficacy of MGI 114 (6-hydroxymethylacylfulvene, HMAF) against the mdr1/gp170 metastatic MV522 lung carcinoma xenograft. Eur J Cancer 34: 908-913, 1998 21. Kelner MJ, McMorris TC, Estes L, et al: Efficacy of MGI 114 (HMAF) against the MRP+ metastatic MV522 lung carcinoma xenograft. Anticancer Drugs 11: 217-224, 2000[Medline] 22. Kiorpes AL, Marshall RF, MacDonald JR: Single and multiple-dose intravenous toxicity of MGI 114, a novel antitumor agent, in rats. Toxicologist 36: 180, 1997 (abstr) 23. Goodman SN, Zahurak ML, Piantadosi S: Some practical improvements in the continual reassessment method for phase I studies. Stat Med 14: 1149-1161, 1995[Medline] 24. Gibaldi M, Perrier D. Pharmacokinetics, ed 2. New York, NY, Marcel Dekker, 1982, pp 409-417 25. Eckhardt SG, Siu LL, Clark G, et al: The continual reassessment method (CRM) for dose escalation in phase I trials in San Antonio does not result in more rapid study completion. Proc Am Soc Clin Oncol 18: 163a, 1999 (abstr 627)
26.
Eisenhauer EA, O’Dwyer PJ, Christian M, et al: Phase I clinical trial design in cancer drug development. J Clin Oncol 18: 684-692, 2000 27. Marshall RF, Kiorpes AL, MacDonald JR: Single and multiple-dose intravenous toxicity of MGI 114, a novel antitumor agent, in beagle dogs. Toxicologist 36: 180, 1997 (abstr) Submitted January 31, 2000; accepted July 6, 2000. This article has been cited by other articles:
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
TOP
ABSTRACT
INTRODUCTION
z), which was determined from the slope of the terminal phase of the plasma concentration–time curve. Systemic clearance (Cl) was determined by dividing the dose by AUCinf. The volume of distribution at steady state (Vss) was calculated using standard noncompartmental methods.
) at the (A) 10.64-mg/m2 and (B) 14.15-mg/m2 dose levels.
, 10.64 mg/m2; 
, 17.69 mg/m2; and •, 14.15 mg/m2 administered as a 1-hour infusion. The dashed horizontal line shows the overall mean of all the observations. The P values represent the statistical difference between the mean values of patients who experienced clinical RTA and those who did not.
