Phase I Clinical and Pharmacokinetic Study of NSC 655649, a Rebeccamycin Analogue, Given in Both Single-Dose and Multiple-Dose Formats¹

Indolocarbazole compounds, which are characterized by a carbohydrate moiety linked to an aromatic chromophore, represent a new class of antineoplastic drugs. Subgroups within this class include the rebeccamycin family of agents, in which the carbohydrate residue is linked to one indole nitrogen, and the staurosporine family of agents, which have two indole nitrogens linked to a carbohydrate residue. Representatives from both of these subgroups have been the subjects of clinical trials.

Rebeccamycin is an antitumor antibiotic produced by the actinomycete Saccharotrix aerocolonigenes with activity against several human tumor cell lines, including A549 (lung adenocarcinoma), HCT-116 (colon carcinoma), and KB (nasopharyngeal carcinoma; Refs. 1, 2). In contrast to staurosporine, which has potent inhibitory effects on the protein kinases A, C, and K, rebeccamycin appears to exert its primary antineoplastic effect by poisoning topoisomerase I and has negligible effect on protein kinase C and topoisomerase II (3). Rebeccamycin inhibits topoisomerase I in a manner comparable with the camptothecins, i.e., by stabilizing the topoisomerase-DNA covalent complex such that DNA may be cleaved but not resealed (4). However, in contrast to the camptothecins, which only bind to DNA in the presence of topoisomerase I, rebeccamycin and its analogues appear capable of intercalating DNA in the absence of topoisomerase I.

Greater than 100 compounds related to rebeccamycin have been derived to date. Because rebeccamycin is insoluble in water, semisynthetic, water-soluble derivatives such as NSC 655649, the tartrate salt of rebeccamycin, have been created (5). The mechanism of action of NSC 655649 is different from its parent compound in that it has no significant activity against topoisomerase I but does strongly inhibit topoisomerase II (6). It appears that the mechanism of this inhibition may differ from that of the epipodophyllotoxins such as etoposide by inhibiting the strand-passing activity of topoisomerase II rather than causing stabilization of the transiently formed tertiary intermediate complex. NSC 655649 has shown antitumor activity against the P388, B16 melanoma, M5076 reticulum cell sarcoma, M109 lung cancer, HCT116 human colon carcinoma, and A549 human lung cancer cell lines (7, 8, 9, 10).

Two Phase I trials of NSC 655649 have been published recently. Tolcher et al. (11) treated 45 patients at doses between 20 and 744 mg/m² given as a single dose every 3 weeks. Antitumor activity was noted in two ovarian cancer patients and one patient with soft tissue sarcoma. Dose-dependent myelosuppression was the principal toxicity noted. The recommended dose for further study was 500 mg/m² for heavily pretreated patients and 572 mg/m² for minimally pretreated patients. Dowlati et al. (12) treated 31 patients with doses between 60 and 188 mg/m²/day × 5 consecutive days. A long terminal half-life (154 ± 55 h) was noted, and it appeared that the drug was primarily hepatically metabolized. High concentrations of NSC 655649 were noted in the bile fluid of one patient, and intriguing activity was seen in patients with hepatobiliary tumors. The dose-limiting toxicity in this study was also neutropenia. On the basis of the results of these Phase I studies, additional Phase II evaluations of NSC 655649 in ovarian and hepatobiliary cancers are ongoing.

Herein we submit results of a third Phase I dose escalation and pharmacokinetic trial of NSC 655649.

Patients with histologically confirmed metastatic epithelial and mesenchymal tumors or malignant lymphomas refractory to conventional therapy or for whom no effective therapy existed were candidates for this study. Eligibility criteria included: (a) age 18 or older; (b) performance status of 0, 1, or 2 by National Cancer Institute criteria; (c) life expectancy 12 weeks or greater; (d) no prior chemotherapy or radiotherapy within 4 weeks of the initiation of study drug treatment and at least 2 weeks from any recent surgery; (e) adequate hematopoietic function (WBC >4,000/mm³ and platelets >100,000/mm³), hepatic function (total bilirubin <1.6 mg/dl and aspartate aminotransferase <3× normal), and renal function (creatinine <1.6 mg%). Measurable disease, although desirable, was not a requirement for entry into the protocol, although all patients were required to have at least evaluable disease. All patients gave written informed consent in accordance with federal and institutional guidelines before treatment. Patients were enrolled between 8/14/95 and 4/21/99 in two cohorts as outlined below.

NSC 655649 was supplied by the National Cancer Institute in sterile vials containing 20 ml of a solution of 10 mg/ml NSC 655649. The study was conducted in two parts to evaluate two different dosing strategies.

In the step 1 (single-dose) study, a single i.v. dose of study drug was administered every 3 weeks. The starting dose was 20 mg/m², which was based on the LD₁₀ in mice (277 mg/m²). Initially, the dose was given over 5 min for the first three dose levels, which was thereafter increased to ≥30 min because of local venous irritation. The lengthened infusion times were also accompanied by administration of more diluted infusions of NSC 655649. At least three patients were treated at each of the following successive NSC 655649 dose levels: 20, 40, 67, 100, 140, 186, 248, 330, 440, 572, and 760 mg/m².

In the step 2 (multiple-dose) study, the total dose of NSC 655649 was divided and administered i.v. over three successive days every 3 weeks. This second part of the study was undertaken to determine whether further dilution of NSC 655649 would ameliorate or eliminate the persistent venous irritation seen with the single-dose schedule. Each daily dose of NSC 655649 was given as a dilute solution over ≥30 min. In this part of the study, patients received the drug at four daily dose levels: 110, 146.7, 190.7, and 253.3 mg/m²/day (330, 440, 572, and 760 mg/m² total dose, respectively). Because of minimal toxicity seen at lower dose levels in step 1, the starting dose for step 2 was 330 mg/m².

For both parts of the study, the occurrence of DLT³ in any patient was cause for the addition of at least three additional patients at that dose level. DLT was defined as any grade 3 or 4 toxicity with the exception of nausea and vomiting. Evaluation for DLT was carried out for at least 3 patients at each level before dose escalation. The MTD was defined as the highest dose level at which fewer than two of six new patients experienced DLT. Toxicities were graded according to the National Cancer Institute Common Toxicity Criteria (version 1.0). Patients who developed DLT could continue to receive treatment at a reduced dose after recovery. No intraindividual dose escalation was allowed. At the MTD in both parts of the study, a cohort of males were accrued to assess any differences in PK parameters.

At baseline, all patients were evaluated with a complete history and physical examination as well as baseline assessment of organ function including complete blood count, differential, platelet count, urinalysis, electrolytes, creatinine clearance, calcium, phosphate, uric acid, alkaline phosphatase, aspartate aminotransferase, lactate dehydrogenase, bilirubin, chest X-ray, electrocardiogram, and pregnancy test in female patients of child-bearing potential. Evaluation of measurable disease was carried out by appropriate radiological studies at baseline and every 3–6 weeks after study entry. Patients were able to continue study participation in the absence of progressive disease. A complete remission was defined as the disappearance of all clinical evidence of active tumor and symptoms for at least 1 month, with stable performance status. A partial remission was defined as a >50% decrease in the sum of the products of the perpendicular tumor diameters of all measurable lesions documented for at least 4 weeks. Stable disease was defined as a response less than partial remission or progression lasting a minimum or 8 weeks. No simultaneous increase in the size of any lesion or the appearance of any new lesion was permissible for complete remission, partial remission, or stable disease. Progression was defined as an unequivocal increase of at least 25% in the size of any measurable lesion or the appearance or any new lesions.

In step 1, plasma samples were collected before and at the following times after the end of the infusion: 0 (end), 5, 10, 20, and 60 min and 2, 4, 6, 24, 48, and 72 h. In step 2, plasma was sampled over the 3-day period and at 48 h after the end of the day 3 infusion to determine the total exposure to NSC 655649. Urine was collected from the start of the infusion to 72 h after infusion. All samples were frozen at −70°C until assay.

The PK parameters for NSC 655649 were measured as the free base (NSC 640199) in plasma and urine using a validated high performance liquid chromatographic assay. Pure NSC 655649 for preparation of standards was obtained from the Pharmaceutical Management Branch (Cancer Therapy Evaluation Program, Division of Cancer Treatment and Diagnosis, National Cancer Institute, Bethesda, MD). The high-performance liquid chromatographic system consisted of a Thermo Separations P4000 pump, UV2000 detector set to 318 nm, and AS3000 autosampler (ThermoQuest, San Jose, CA) and a Shimadzu 4271 Integrator (Shimazdu, Columbia, MD). Chromatographic separations were achieved on two NovaPak 15-cm C₁₈ reverse-phase columns hooked in series (Waters/Millipore, Milford, MA). A gradient solvent system was used, starting at 60% acetonitrile/40% ammonium phosphate buffer (50 nm) pumped at 0.8 ml/min for 6 min and increasing over 10 min to 80% acetonitrile pumped at 1.6 ml/min. The gradient was regenerated over 10 min. In this system, the retention time for the free base was 4 min.

For plasma, 300 μl patient samples or standards prepared from drug-free plasma were extracted with 600 μl of acetonitrile. After vortexing and centrifugation, 70 μl of the resulting extract was taken for chromatography. Samples were quantified from standard curves constructed from chromatographic peak height of standards versus concentration. All samples were assayed in duplicate. Urine was assayed by direct injection after dilution with 1.5 volumes of mobile phase and quantified using a standard curve constructed from standards of free drug in blank urine.

The PK parameters for NSC 655649 measured as the free base in individual patients were examined by both compartmental and noncompartmental methods using PKAnalyst (MicroMath Research, Salt Lake City, UT). The AUC was calculated by the linear trapezoidal rule, extrapolated to infinity. Clearance (Cl_t) was determined by (total dose)/(extrapolated AUC).

The characteristics of the patients enrolled are outlined in Table 1. Sixty-nine patients received a total of 152 courses of treatment on protocol. There were no significant differences between patients in the two cohorts. Men and women were roughly evenly represented in this study. Two-thirds of our patients had cancer of the colon, kidney, lung, or breast, although 19 tumor types were represented in all. Patients were heavily pretreated, with the majority having had surgery and multiple chemotherapy regimens and almost half having had prior radiotherapy. Only 2 patients had received no prior treatment. The dose escalation schema, numbers of patients per level, courses at each level, and number of patients experiencing DLT at each level for both steps 1 and 2 are shown in Table 2.

To be fully evaluable, patients were required by the protocol to have been treated and then observed for at least 21 days. Six patients (1 on step 1 and 5 on step 2) did not meet these criteria. Three patients had progressive disease before the conclusion of the observation period. One patient had expressive aphasia noted on the evening after the third day of drug infusion at level 4 on the step 2 protocol. He was found to have brain metastases by head computed tomography, requiring removal from the study. One patient experienced hemoptysis and new neurological signs and symptoms during day 2 of treatment on step 2. He was found to have brain metastases and bleeding pulmonary metastases and required urgent pulmonary brachytherapy and whole brain radiotherapy. One patient was removed from study because of a deviation from the required infusion duration.

The median number of cycles received in both steps 1 and 2 was two. Dose modifications per protocol were required in a total of 11 patients, 4 in step 1 and 7 in step 2. Only patients in dose cohorts at or above 440 mg/m² required dose reduction. Neutropenia was the indication for all dose modifications. When only those patients who received doses of NSC 655649 at doses of 440 mg/m² or higher are considered, 4 of 18 (22%) patients in step 1 and 7 of 19 (37%) patients in step 2 required dose reduction because of neutropenia.

In step 1, 23 men received 59 cycles and 19 women received 49 cycles of NSC 655649. In step 2, 9 men received 14 cycles and 12 women received 27 cycles. Overall, 76 cycles were administered to women and 73 cycles to men. There were no apparent differences in toxicity at any grade between men and women on study. In addition, 6 patients received drug but were unevaluable, as outlined above. The toxicities experienced by these ineligible patients are included in this analysis.

Grade 1–2 and grade 3–4 toxicities noted during both step 1 and step 2 drug administration are portrayed in Tables 3 and 4. The most frequently observed toxicities were nausea, vomiting, fatigue, venous irritation at the drug administration site, and myelosuppression.

Mild to moderate nausea and/or vomiting were seen in over half of the patients in both arms of the study. Severe nausea and vomiting was uncommon, occurring in <10% of patients in step 1 and in no patients in arm 2. Nausea was noted primarily at higher doses, with 32 of 48 total grade 1–2 and 4 of 5 total grade 3–4 nausea events occurring in the two highest dose groups (572 and 760 mg/m²).

Grade 1–2 fatigue was commonly described in 24% of step 1 patients and 45% of step 2 patients. Fatigue, like nausea, was a dose-dependent toxicity, with the majority of patients experiencing this symptom coming from the highest dose cohorts, which is probably why step 2 patients had a higher incidence of fatigue; step 2 patients were enrolled only at the four highest dose levels.

Venous irritation was a common but uniformly low-grade toxicity seen in all dose cohorts in step 1. Overall, 36% of step 1 patients experienced this toxicity, and the protocol was amended after the first three dose cohorts were completed to require the drug be further diluted and administered over ≥30 min instead of the previously stipulated 5 min. Despite additional dilution, venous irritation was still seen in a substantial minority of step 2 patients (22%). In the end, we required all patients to receive drug through central i.v. access, a decision that effectively prevented this toxicity.

Increased alkaline phosphatase and transaminases were seen in 19 and 14% of patients in step 1 but not in patients in step 2, suggesting that dividing the dose over 3 days reduced low-grade effects of NSC 655649 on the liver.

Dose-dependent myelosuppression was the most common serious toxicity seen in patients on both arms of the study. As noted above, neutropenia was the only toxicity that met criteria for dose modification during the course of the study. Data on neutrophil nadir were examined in detail only for patients treated at the highest three dose levels in each arm, because no significant myelosuppression was seen at lower doses. In step 1, the median neutrophil nadir was 1200/mm³ (range, 196-4900), and this occurred at a median of 14 days from receiving study drug. Recovery of neutrophil count for step 1 patients took a median of 7 days. For step 2 patients, the median neutrophil nadir was 1900/mm³ (range, 196-5200), and this occurred at a median of 20 days. Recovery was prompt for step 2 patients, taking a median of 5 days. Severe thrombocytopenia or anemia were seen by only one patient each on either arm.

The DLT for both arms of this study was grade 3–4 neutropenia. In step 1, one of the first three patients at the 572 mg/m² dose level experienced grade 3 neutropenia. As a result, an additional 3 patients were added at that dose level. None of these three additional patients experienced DLT. In the 760 mg/m² cohort, 2 of 3 enrolled patients had grade 4 neutropenia in cycle 1. As a result, the 572 mg/m² dose level was declared the MTD for step 1, and the emphasis of the study shifted to carrying out step 2.

No patient on step 2 experienced DLT during dose escalation until the final dose level of 760 mg/m². At this level, 3 of 4 eligible patients experienced grade 4 neutropenia. Therefore, the 572 mg/m² was also the MTD for step 2. By chance, all 6 patients treated at the MTD on step 2 were female, and all 6 patients treated at the MTD on step 1 were male.

For this reason, additional patients were added to achieve gender equity at the MTD for both step 1 and step 2. Of the 6 female patients added in step 1, 3 experienced DLT (grade 3–4 neutropenia). There was no difference in level of prior treatment to explain this difference in toxicity between men and women at this dose level. Also, no difference was seen between the median neutrophil nadir for men and women treated at the highest three dose levels on step 1.

Of the 6 male patients added in step 2, one had grade 4 neutropenia. Thus, only 1 of 12 patients in step 2 experienced DLT at the MTD of 572 mg/m². There were no treatment-related deaths, although 4 patients on step 1 died of progressive disease within 30 days of coming off study.

One patient with carcinoma of unknown primary in the liver achieved a partial response. This patient had not received any previous chemotherapy and received the MTD of 572 mg/m² on step 1 (single-dose protocol). This patient received a total of 7 cycles of treatment, with the last five cycles given at a 50% dose reduction because the protocol required dose reductions for neutropenia.

Response was demonstrated by a liver ultrasound showing 63% shrinkage of the dominant lesion after 77 days on treatment. A repeat ultrasound confirmed this finding on day 115 of treatment. This patient’s partial response lasted 91 days, ending when disease progression was noted in the liver.

Twenty-six patients had a best response of stable disease, with 4 patients having a disease stabilization period of >100 days. The median period of SD in this group of patients was 83 days (range, 41–280 days). Twenty-nine patients had progressive disease, noted after a median of 43 days (range, 21–70 days).

In 13 patients, no formal disease response data were collected. Two patients experienced clinical deterioration without radiographic evidence of progressive disease and were taken off study at their physician’s discretion after 21 and 23 days, respectively. Three patients were removed from study after only one cycle for reasons other than disease progression: in two cases, neutrophil counts did not recover in time for cycle 2 to be given; and in one, grade 3 transaminase elevation was noted on the day cycle 2 was to be given. The protocol did not allow for treatment delay for this lab result, and the patient was taken off study. Two patients were removed at their own request after one cycle of treatment each, in one case because of persistent grade 1 nausea and in the other case because of the occurrence of grade 2 deep venous thrombosis at the conclusion of cycle 1. Lastly, there were 6 patients who were not fully evaluable for reasons discussed above. These 6 patients were included in analysis of toxicity but not in analysis of response or in PK analysis.

Absolute recovery of drug in plasma standard extracts averaged 88 ± 7% and did not vary over the concentration range. Recovery of drug in patient samples by standard addition averaged 100 ± 7% (n = 6). The standard curve was linear from 0 to 20 μg/ml in plasma (r² = 0.9994 ± 0.001 for five curves over 6 weeks), and the limit of quantitation was 20 ng/ml plasma concentration (0.46 ng injected). Intraday variability of three replicate determinations of each plasma standard over the concentration range was <6%. Triplicate determinations of patient samples had a coefficient of variation of 2.2 ± 1.1% (n = 4, each run in triplicate). Four repetitive assays of a control patient sample over a 6-week period showed a variability of 4.1%.

Patients received single doses ranging from 20 to 760 mg/m². In all patients at dose level 3 (67 mg/m²) and above, the drug was detectable for 72 h after infusion. The postinfusion concentration-time data sets fit best to a triexponential model. NSC 655649 exhibited linear PK over this dose range; both the total AUC and the maximum concentration at the end of the infusion (C_max) were linearly related to dose (Fig. 2; no other parameter was dose related. The drug had a long terminal elimination half-life and relatively high volume of distribution (Vss). Less than 2% of the drug as the free base was excreted intact in the urine from the start of the infusion to 72 h after dosing. Thus, the principal mechanism of NSC 655649 elimination appears to be hepatic metabolism and biliary excretion. Table 5 lists the PK parameters by dose level. The first three dose levels were not included in Fig. 1 B because of the short (5-min) infusion used. At the MTD of the single-dose schedule, 572 mg/m², cohorts of 6 males and 6 females were accrued to assess any difference in PK parameters. There was no statistical difference in the clearance or the half-life for the male cohort versus the female cohort.

Patients received the drug as three divided doses, at four daily dose levels: 110, 146.7, 190.7, and 253.3 mg/m²/day (330, 440, 572, and 760 mg/m² total dose, respectively). The PK parameters compared with the single-dose study are shown in Table 6 and in Fig. 2. Fig. 3 shows the complete concentration-time curves for a patient on the single-dose schedule and a patient on the divided-dose schedule at the same total dose.

NSC 655649 exhibited significant accumulation over the 3 days of infusion because of the long terminal half-life. The 24-h AUC on day 3 was 141 ± 26% of that on day 1. Over the dose range used for the 3-day schedule, there was no significant difference in total clearance between the two schedules. The mean terminal half-life after day 3 appeared somewhat lower for the divided-dose schedule, which was probably because plasma sampling continued for only 48 h after day 3. If the half-life for the single-dose schedule is calculated using only the time points to 48 h, the half-life is essentially the same as that for the divided dose schedule (25.5 ± 4.4 h versus 22.0 ± 8.2 h; P = 0.100). At the MTD for the divided-dose schedule, there was no statistical difference in the clearance or the half-life for a male cohort versus a female cohort, as was true for the single-dose schedule.

The MTD of the divided dose schedule was 190.7 mg/m² daily, or 572 mg/m² total dose, which was the same MTD as the single-dose schedule. With reference to Table 6, comparison of PK parameters between the two schedules at the MTD showed no significant difference in AUC or clearance. A slight difference in half-life was noted, as discussed above, and there was also a significant difference in the single-dose C_max versus the day 3 C_max (14.5 ± 3.2 versus 5.3 ± 0.9 μg/ml; P < 0.001). This was to be expected because the daily dose for the divided-dose schedule was one-third of the single dose.

The mean clearances reported here are much lower than the preclinical values for dogs and mice, 210 ml/min/m². The mouse LD₁₀ was 277 mg/m², about half our determined MTD.

Indolocarbazoles represent a new class of topoisomerase-inhibiting antineoplastic agents. NSC 655649, the tartrate salt of rebeccamycin, is one of the first representatives from this class of compounds to come into clinical trial. Inhibition of both DNA topoisomerase I and II has been a fruitful strategy in the treatment of malignancy (13). Topoisomerases I and II are enzymes that alter the topological state of DNA by generating transient breaks in the sugar-phosphate backbone of DNA, allowing DNA relaxation when needed for proper cellular function. Topoisomerase enzymes all possess three key functions: (a) the cleavage of DNA; (b) the passage of intact DNA through the cleaved strand; and (c) the religation of cleaved segments. Epipodophyllotoxins, such as etoposide, and camptothecins, such as irinotecan or topotecan, inhibit the third function, the religation of cleaved DNA by topoisomerases II and I, respectively. NSC 655649, however, exerts its effect by inhibiting the second function, the strand passing activity of topoisomerase II (1, 7).

This Phase I study examined the administration of NSC 655649 to patients with advanced malignancy in two dose formats. In the first dose format (step 1), patients were given a single dose of NSC 655649, whereas in the second dose format (step 2), the dose was divided into three equal doses over 3 days.

A total of 69 patients were treated with 6 being unevaluable. Forty-two evaluable patients received 108 cycles on step 1, and 21 evaluable patients received 41 cycles on step 2. There were no differences in the characteristics of patients in the two dose cohorts. A wide variety of malignancies were represented, although colon, renal, lung, and breast cancer represented two-thirds of all patients. Our patients were extensively pretreated, as would be expected for a Phase I clinical trial.

Venous irritation, although not a DLT, proved to be a difficult toxicity to characterize and control. In attempting to determine whether this reaction was concentration dependent, the infusion vehicle (normal saline) was increased at designated dose levels and reactions monitored. The incidence of venous irritation appeared to be independent of the infusion concentration. We found NSC 655649 to be peripherally intolerable despite appropriate drug administration manipulations. After central venous access was mandated, no further infusional toxicity was noted. Mild phlebitis was also seen in patients treated on the two other Phase I trials of NSC 655649. Both of these studies eventually required patients to have central venous access devices because of this toxicity (11, 12).

Very little toxicity and no responses were seen at dose levels below 330 mg/m². The MTD identified in each arm of the study was identical: 572 mg/m². The primary DLT in both arms was neutropenia, as expected from animal studies. Grade 3–4 neutropenia was seen exclusively in patients treated at the highest dose cohorts on both steps 1 and 2. We identified the 572 mg/m² dose as MTD in step 1. When we added 6 additional female patients to that dose level to achieve gender equity, 3 of 6 added patients experienced DLT (neutropenia). Thus, overall 4 of 12 patients treated at the MTD on step 1 experienced DLT (3 of 6 women versus 1 of 6 men). Because there was no difference by gender in the median neutrophil nadir, time to nadir, or time to recovery, we doubt that a real difference in myelosuppressive effect exists between men and women treated with NSC 655649. Neither of the other phase I studies of NSC 655649 noted such a difference, and no such difference was seen in patients in our study treated on step 2.

The MTD identified in our study was similar to that found by both other Phase I studies of NSC 655649. Tolcher’s group found an MTD of 500 mg/m² for heavily pretreated patients and 572 mg/m² for minimally pretreated patients. By dividing their total dose over 5 days, Dowlati et al. (12) were able to achieve a higher total MTD of 165 mg/m² (825 mg/m²) for patients without prior chemotherapy and 141 mg/m²/day (700 mg/m²) for pretreated patients. We did not perform separate analysis according to degree of prior treatment because our patient population was almost uniformly heavily pretreated.

The most common nonhematological toxicity in both arms was nausea. Interestingly, although grade 3–4 nausea occurred in 10% of patients in step 1, this degree of nausea was not seen when the dose was divided over three days in step 2. Grade 1–2 fatigue was the second most common toxicity noted by patients. This, like nausea and myelosuppression, was a dose-dependent toxicity, occurring predominantly in higher dose cohorts. A minority of patients on step 1 had mild, reversible liver function test abnormalities.

Pharmacokinetic data showed that NSC 655649 exhibited linear PKs over the dose range tested in both arms of the study. Both AUC and C_max were linearly related to dose. The drug had a long terminal elimination half-life and large volume of distribution. In the multiple-dose portion of the study, significant drug accumulation was noted over the 3 days of drug administration, although there were no significant differences in total clearance or AUC noted between the two schedules. Significant interpatient variability was noted on both arms of the study with regard to pharmacokinetic data. The long half-life, large volume of distribution, and high interpatient variability reported here is quite similar to that reported by Tolcher et al. (9). The evidence of drug accumulation attributable to the drug’s long terminal half-life in the multiple-dose part of our study closely resembles results obtained by Dowlati et al. (12) in their Phase I pharmacokinetic analysis of NSC 655649 given over 5 consecutive days.

Only one patient had a response to NSC 655649. This patient had carcinoma of unknown primary in the liver and had received no prior therapy before entry into this study. In this respect, he was an unusual patient, because the vast majority of our patients had been heavily pretreated. A sizable minority of patients had periods of stable disease on study, including four patients whose disease remained stable over 100 days. It is impossible to be certain whether the drug was active in these patients.

The design of this trial allowed us to compare toxicity and PK data from two different dosing schemas of NSC 655649. The toxicity profile was different for single-dose versus multiple-dose administration. Patients on step 1 (single dose) were more likely to experience neutropenia and nausea, whereas patients on step 2 had more fatigue. Both methods of administration required placement of a central venous access device because of phlebitis. No conclusions regarding which dosing method is more efficacious can be drawn from this trial. Because the multiple-dose regimen was less myelosuppressive, it would be appropriate to use this dosing schema if NSC 655649 is used in combination in future trials. On the basis of our experience with the single-dose arm of this study (4 of 12 patients at MTD having dose limiting neutropenia), we would concur with Tolcher et al. (11) in choosing a lower dose of 500 mg/m² of NSC 655649 for heavily pretreated patients and 572 mg/m² for minimally pretreated patients involved in Phase II study of this drug if it is administered by a single dose.

More responses were seen in the two other Phase I trials of this agent. In Tolcher’s study, 2 of 3 responding patients had ovarian cancer, a disease type which made up almost 20% of patients in that trial. In our trial, however, only 1 ovarian cancer patient was included. This patient experienced stable disease for 112 days, one of the longest durations of stable disease seen. Also, in trial by Dowlati et al. (12), a large number of treated patients were chemotherapy naïve, unlike our patients.

NSC 655649 represents a new class of topoisomerase inhibitors that have their effect on the strand-passing activity of the topoisomerase II enzyme. Our study adds to the growing literature regarding this agent. NSC 655649 was safe, although intolerable for patients when peripherally administered. Limited evidence of antitumor effect was observed, although such an assessment was not the primary goal of the study. Study of this compound at the Phase II level is warranted, in view of activity demonstrated in 1 patient and in a larger number of patients in two other Phase I studies. Consideration should be given to further study of NSC 655649 in combination with non cross-resistant such as weekly low-dose taxane therapy or cisplatin. Studies of this drug in ovarian cancer and hepatobiliary cancer are already underway. Because of the common occurrence of local irritation at the site of i.v. drug delivery, we recommend that NSC 655649 be administered only through central i.v. access in future studies.