Clinical and Pharmacologic Study of the Novel Prodrug Delimotecan (MEN 4901/T-0128) in Patients with Solid Tumors

Purpose: To investigate i.v. administration of delimotecan (MEN 4901/T-0128), a carboxymethyldextran polymer prodrug of the active camptothecin derivative T-2513, and to assess the maximum tolerated dose, safety profile, clinical pharmacology, and antitumor activity of delimotecan and metabolites.

Experimental Design: Patients with solid tumors refractory to standard therapy received i.v. delimotecan as 3-hour infusion once every 6 weeks. The starting dose was 150 mg/m², followed by an accelerated dose escalation with at least one patient per dose level. The pharmacokinetics of delimotecan, T-2513, and its metabolites, SN-38, SN-38G, T-1335, T-0055, and T-3921, were assessed in plasma and urine, and their pharmacodynamics were determined by measuring the effect of the treatment on hematologic and nonhematologic toxicity.

Results: Twenty-two patients received 35 courses. Dose-limiting toxicities were observed at 5,400 mg/m² (n = 1), 3,600 mg/m² (n = 1), and 2,400 mg/m² (n = 2). The dose level of 1,800 mg/m² was determined as maximum tolerated dose. Two partial responses were observed in patients with anal cancer (1800 mg/m²) and head and neck cancer (2400 mg/m²). Delimotecan had a long terminal half-life of 109 h, and relatively high exposures to T-2513 and SN-38 were obtained. The percentage decrease in WBC and absolute neutrophil count significantly correlated with the dose of delimotecan.

Conclusions: Based on its preliminary antitumor activity, safety profile, and pharmacokinetic profile, we recommend to evaluate delimotecan given as 3-hour infusion once every 6 weeks at a dose level of 1,800 mg/m² in a phase II study.

Delimotecan (MEN 4901/T-0128) is a new polysaccharide prodrug comprising the active moiety 10-(3′-amino-propyloxy)-7-ethyl-(20S)-camptothecin (T-2513) bound to carboxymethyldextran via a triglycyl chain (Fig. 1). This prodrug with high solubility was designed to improve the pharmacologic profile (higher efficacy/lower toxicity) compared with clinically available camptothecin (CPT) analogues, such as irinotecan (CPT-11), which is used in the treatment of advanced colorectal cancer (1, 2). T-2513 is a selective inhibitor of topoisomerase I, responsible for relaxation of torsionally strained supercoiled DNA (3–5). T-2513 binds covalently to and stabilizes the topoisomerase I–DNA complex and inhibits DNA replication and RNA synthesis, ultimately leading to cell death (5–7). The cytotoxic effects of T-2513 and metabolites are specific for the S-phase of the cell cycle (8). Therefore, prolonged exposure to delimotecan could increase antitumor activity, as was shown for other topoisomerase I inhibitors (9, 10). Delimotecan given i.v. as a single dose and once weekly for 4 weeks showed marked antitumor activity in human lung, esophagal, gastric, colon, and pancreatic tumor xenografts in nude mice (11). A single i.v. administration of delimotecan at a dose of 80 mg/kg (dose equivalent to T-2513) resulted in higher antitumor activity in human lung, colorectal, and esophagal carcinomas in mice compared with i.v. CPT-11 given every fourth day for four times at a total dose of 240 mg/kg.⁹

Reversible leukocytopenia and anemia, decrease in body, thymus, and testes weight, and lymphoid depletion were observed after single and repeated i.v. dosing of delimotecan. Preclinical studies showed that the single dose of i.v. delimotecan severely toxic to 10% of the animals (STD₁₀), was 40 mg/kg (∼800 mg/m²) in dogs and 2,000 mg/kg (∼6,200 mg/m²) in mice after single i.v. administration. Delimotecan had a long terminal half-life (t_1/2) of 40 to 129 h. The area under the plasma concentration time curve (AUC) of delimotecan seemed to increase linearly with dose. Preclinical studies suggested that delimotecan can permeate and accumulate in tumor tissue and is cleaved to T-2513 by cathepsin B, which is overexpressed in tumors. These studies further showed that tumor-associated macrophages may play a key role in the uptake of delimotecan and cleavage and local release of T-2513 (12, 13). T-2513 can be metabolized to the CPT analogues, T-0055, T-1335, T-3921, and SN-38 (the active metabolite of CPT-11), which is glucuronidated to SN-38 glucuronide (SN-38G; Fig. 1). T-0055, T-1335, and T-3921 are formed out of T-2513, in which the terminal amino group is metabolized to a hydroxyl group (T-0055), carboxylacid group (T-1335), and aminocarboxymethyl group (T-3921). Preclinical studies determining the in vitro cytotoxicity [concentration-inhibiting cell growth by 50% (GI₅₀)] after 24 hours of drug incubation in a variety of human tumor cell lines showed that SN-38 had the highest cytotoxicity (GI₅₀ ∼ 1.5-15 ng/mL), followed by T-0055 (GI₅₀ ∼ 3.6-20 ng/mL), T-2513 (GI₅₀ ∼ 15-111 ng/mL), T-3921 (GI₅₀ ∼ 30-180 ng/mL), and T-1335 (GI₅₀ ∼ 73-1195 ng/mL; ref. 14).

Based on the preclinical studies, which showed marked antitumor activity, moderate toxicity, and a long t_1/2 of delimotecan after single i.v. dosing, a phase I study was initiated to investigate the feasibility of administering delimotecan as 3-hour infusion once every 6 weeks in patients with solid tumors. A starting dose of 150 mg/m² delimotecan was chosen based on the STD₁₀ in dogs. The objectives of the study were to assess the maximum tolerated dose, safety profile, and the pharmacokinetics, pharmacodynamics, and antitumor activity of delimotecan.

Patients. Patients with histologically or cytologically confirmed solid tumors refractory to standard therapy were eligible. Other criteria were age of ≥18 y, a WHO performance status of ≤2 on the Eastern Cooperative Oncology Group scale, and an estimated life expectancy of ≥3 mo. Previous chemotherapy and radiotherapy had to be discontinued for at least 4 wk before study entry and 6 wk in the case of mitomycin-C, nitrosourea, melphalan, and high-dose carboplatin. Patients were allowed to be pretreated with any kind of therapy without further limits. Patients had to have adequate bone marrow function (WBC, ≥3.0 × 10⁹/L; absolute neutrophil count, ≥1.5 × 10⁹/L; platelets, ≥100 × 10⁹/L), kidney function (creatinine clearance, ≥50 mL/min), and liver function (total bilirubin, <1.5 mg/dL and/or alanine aminotransferase and aspartate aminotransferase, ≤1.5 times the ULN or ≤2.5 times the ULN in the case of tumor involvement in the liver). The study protocol was approved by the medical ethics committee of both participating hospitals, and all patients signed informed consent.

Treatment plan, schedule of administration, and study design. The first patient was treated at a dose of 150 mg/m² delimotecan, given as 3-h infusion once every 6 wk. Toxicities were graded according to the National Cancer Institute-Common Toxicity Criteria version 3.0.¹⁰

If none or minimal (≤grade 1) toxicity was observed, dose escalation was done with one patient per dose level and 100% dose increment. When ≥grade 2 toxicity was observed, the dose was escalated by 67%, 50%, 40%, or 33% increments, relative to the preceding dose level, dependent on the toxicity profile, in a cohort of three patients per dose level. If one patient experienced a dose-limiting toxicity (DLT), the cohort was to be expanded to six patients. However, the escalation of doses, as well as the interval between doses, was driven by the overall clinical judgment of the investigators that was mainly based on the ongoing evaluation of the toxicity profile and pharmacokinetic results. DLTs were defined as grade 4 neutropenia lasting >5 d or complicated neutropenia (absolute neutrophil count < 0.5 × 10⁹/L and T > 38.5°C or clinical signs of infection), grade 4 thrombocytopenia and/or hemorrhagic complications or any bleeding episode requiring platelet transfusion, any ≥grade 3 nonhematologic toxicity (except alopecia and vomiting in case toxicity was grade 4 with maximal antiemetic support), and diarrhea ≥grade 2 and lasting ≥3 d, all during cycle 1, and the inability to receive the second cycle within 2 wk after day 43. The maximum tolerated dose was defined as one dose level below the dose level that induced DLTs in two of six patients. Each patient had to be followed for at least 6 wk before entering the next dose level. Patients started a new cycle in case of complete recovery from any toxicity of the previous cycle. In case of failure of recovery, treatment was to be delayed for a maximum of 2 wk. Patients with partial response, complete response, or stabilization of disease who experienced any DLT were allowed to be retreated at a lower dose level if this was in the best interest of the patient. In case there was no recovery after 2 wk, the patient was taken off the study.

Drug formulation. Delimotecan was supplied as a lyophilized powder in 24 mL glass vials by AG-Menarini Group (Galenical Department, Berlin-Chemie AG). Each vial contained 200 mg of delimotecan sodium (composed of 4.5-5.25% T-2513) and Na₂HPO₄ dihydrate and citric acid as inactive ingredients. Before use, it was constituted in 10 mL of 0.9% NaCl, diluted with 0.9% NaCl to a final volume of at least 250 mL, and given via a peripheral line as i.v. infusion. The reconstituted solution was stable at room temperature for 6 h in glass or PVC bags.

Patient evaluation and follow-up. Pretreatment evaluation was done within 2 wk before start of treatment and included a complete medical history and physical examination, vital signs, ECG, chest X-ray, tumor measurements, hematology (hemoglobin, WBC and differential count, platelets), serum chemistry (bilirubin, alkaline phosphatase, aspartate aminotransferase, alanine aminotransferase, total protein, albumin, sodium, potassium, calcium, phosphate, γ-glutamyltranspeptidase, lactate dehydrogenase, glucose, creatinine, and β-human chorionic gonadotrophin; in females with child-bearing potential before the first dose and afterwards only on indication), and urinalysis. Creatinine clearance had to be determined in case of creatinine at >1.5 × ULN. Before each cycle, the physical examination and vital signs were repeated, and hematology and serum chemistry were checked. During each cycle, hematology and serum chemistry were checked once weekly. Urinalysis was done only during the first week of each subsequent cycle or more often if clinically indicated. Tumor assessments, by clinical and instrumental evaluation (X-ray, CT scan, spiral CT scan, MRI), were done after the first and second cycle and, thereafter, every two cycles. Response to treatment was evaluated according to the RECIST criteria (15).

Pharmacokinetic studies. Pharmacokinetics of delimotecan, total and free T-2513, and free metabolites in plasma and urine were determined during the first cycle of therapy. Blood samples of 3 mL of venous blood were drawn into sodium-heparinized tubes at the following time points: on day 1, before dosing, and at 0.5 and 1.5 h after the start of infusion, at the end of infusion and 4, 7, 8, 10, 24, 48, and 72 h after the start of infusion, and on days 7, 10, 14, 21, 28, 35, and 42. After collection of a blood sample, the tube was gently shaken, placed on ice, and immediately centrifuged at 1,500 × g for 5 min at 4°C. Then, plasma was transferred into a polypropylene tube and stored at −20°C until analysis. Urine was collected before dosing and during 0 to 8, 8 to 24, and 24 to 48 h after the start of infusion and stored at 4°C. For each collection interval, the total volume of urine was determined and ∼2 mL of each urine collection were transferred into a polypropylene tube and immediately stored at −20°C. In addition, one urine sample was collected on days 4, 7, 10, 14, 21, 28, and 35 and handled as described above. Analysis of T-2513 and metabolites in plasma, urine, and tissues was done using a sensitive validated high-performance liquid chromatographic method with fluorescence detection. For analysis of total T-2513 in plasma and urine, 0.1 mL of sample was mixed with 0.2 mL of 6 mol/L HCl and kept for 4 h at 100°C to hydrolyze delimotecan. Then, 0.2 mL of 5.5 mol/L NaOH was added, the mixture was vortex mixed, and 0.85 mL of formic ammonium buffer and 50 μL of internal standard solution [10-(2-hydroxyethyloxy)-20(s)-camptothecin, 3 μg/mL] were added to 0.1 mL hydrolysate. After vortex mixing for 15 s, 0.1 mL of the content was taken and 0.3 mL formic ammonium buffer and 0.1 mL acetonitrile were added and vortex mixed for 10 s; the sample was centrifuged at 14,900 × g for 5 min at 4°C, and the supernatant was transferred into a sampler vial. For analysis of free T-2513, T-1335, SN-38, SN-38G, T-3921, and T-0055 in plasma and urine, 50 μL of working solution (0.2 μg/mL) of the IS were added to 0.1 mL of sample. After gentle mixing, 0.3 mL acetonitril and 0.1 mL 0.1 mol/L HCl were added. Each sample was vortex mixed for 10 s and centrifuged at 14,900 × g for 15 min at 4°C. Then, 0.2 mL of the supernatant was transferred into a sampler vial containing 0.3 mL of 35 mmol/L formic ammonium buffer (pH 4.9).

For analysis of T-2513 and metabolites in tissues, each tissue was homogenized and an aliquot was extracted and analyzed after procedure as described above.

The injection volume was 50 μL. The high-performance liquid chromatography system consisted of a model 125 solvent delivery pump (Beckman), automatic sample injector model 232 (Gilson), an Inertsil ODS-2 (5 μm) 250 × 4.6 mm column with a Lichrospher 100-RP-18 guard column (Ghrom), and a RF-551 fluorescence detector (Shimadzu). The excitation and emission wavelength of the detector was 377 and 420 nm, respectively. Column temperature was kept at 40°C, and the flow rate was 1 mL/min. The mobile phase consistent of gradient A (35 mmol/L formic ammonium buffer, pH 4.9) and gradient B [35 mmol/L formic ammonium buffer (pH 4.9)/acetonitril (1:1, v/v)]. The following gradients were used: t = 0-5 min, 70% A; t = 5-12 min, 70-40% A; t = 15-22 min, 40% A; t = 22-29 min, 40-70% A; and t = 29-33 min, 70% A. The analytes had the following retention times: SN-38G, t = 10.7 min; T-2513, t = 14.9 min; T-1335, t = 17.8 min; SN-38, t = 19.3 min; T-3921, t = 20.5 min; T-0055, t = 22.2 min. The intraassay and interassay accuracy for T-2513 and metabolites ranged from −8.8% to 6.3% and the intraassay and interassay precision ranged from 1.2% to 9.0 %. The recovery for T-2513 and metabolites was above 90% for each analyte. The lower limit of quantification for each analyte was 2 ng/mL.

Pharmacokinetic analysis. The plasma pharmacokinetic variables of delimotecan, T-2513, and metabolites were determined by noncompartmental analysis, using WinNonLin (version 5.0, Pharsight Corporation). AUC was determined using the linear logarithmic trapezoidal method up to the last measured concentration-time point and extrapolated to infinity (AUC_0-∞). Furthermore, terminal half-life (t_1/2), maximal drug concentration (C_max), time to maximal drug concentration (T_max), and apparent clearance (Cl; for delimotecan only), and apparent volume of distribution (V_d; for delimotecan only) were determined. The extent of metabolism of total T-2513 to SN-38 was expressed by a metabolic ratio value, defined as the AUC ratio of SN-38 + SN-38G over total T-2513 (based on molar exposure values). The extent of SN-38 glucuronidation was expressed by a glucuronidation ratio value, defined as the AUC ratio of SN-38G over SN-38. In addition, the percentage of the given T-2513 dose recovered in the urine as T-2513 (total and free), T-1335, SN-38, SN-38G, T-0055, and T-3921 during the first 48 h after the start of infusion (U_{excr,t = 0-48 h}) were calculated. Data were reported as mean ± SD. Relationships between the dose of delimotecan and the AUC of total T-2513, SN-38, and T-1335 and between body surface area and weight with Cl of delimotecan were explored by scatter plots.

Pharmacokinetic-pharmacodynamic analysis. Relationships between the dose of delimotecan and the AUC of total T-2513, SN-38, and T-1335 with the percentage decrease in WBC count and absolute neutrophil count were explored with scatter plots. The percentage decrease in blood cells is defined as [100 × (pretreatment value − nadir)] / pretreatment value. Furthermore, relationships between the AUC of total T-2513, SN-38, and T-1335 and categorical toxicity data (intensity of diarrhea, nausea, vomiting, fatigue, and skin rash) were determined. Only data obtained in the first course were used.

Statistical analysis. Relationships between dose with AUC and between body surface area and weight with Cl were assessed by Spearman and Pearson correlation coefficients. The relationship between dose and percentage decrease in blood cells was fit to a sigmoidal maximum effect (E_max) model using WinNonLin according to the following formula: E = E_max [(DE)^γ / (DE₅₀)^γ + (DE)^γ], wherein E represents the observed effect (i.e., percentage decrease), produced by drug exposure (DE); E_max denotes the maximal elicitable effect; DE is a measure of drug exposure (i.e., the dose); DE₅₀ represents the drug exposure associated with 50% of E_max; and γ is the Hill coefficient, which describes the sigmoidity of the curve. Relationships between the AUCs and categorical toxicity data were explored by Spearman rank correlation tests. Statistical analysis was done with the software package Statistical Product and Service Solutions (version 15.0 for Windows, SPSS, Inc.). Differences were considered to be statistically significant at P < 0.05.

Patient demographics. Twenty-two patients were enrolled onto the study, and all patients were evaluable for toxicity. Patient characteristics are outlined in Table 1.

Dose levels and main drug-related toxicities. At the lower dose levels, patients tolerated therapy well and the dose was escalated by 100% with one patient per dose level from 150 to 1,200 mg/m². The number of patients treated at each dose level is depicted in Table 2. Overall, 35 cycles of therapy over eight dose levels were given. At the dose level of 2,400 mg/m², a male with an adenocarcinoma of unknown primary developed grade 3 elevated alkaline phosphatase on day 13 of the first cycle, which was considered as disease-related. No other drug-related toxicities were observed. The patient did not receive a second cycle and was taken off the study because of progressive disease. A lower dose increment of 50% was applied, leading to a dose of 3,600 mg/m² for the next cohort. A male patient with colorectal carcinoma and lung and liver metastasis developed grade 3 pneumonia and grade 4 dyspnea within 6 days after start of the first cycle, which were considered unrelated to the study drug and which resolved within 10 days. Because no DLTs were observed at this dose level, the dose was further escalated by 50% to 5,400 mg/m². At this dose level, a 47-year-old female with colon carcinoma and liver metastasis developed grade 3 skin rash on day 3 of the first cycle, followed by grade 3 fever, grade 3/grade 4 hematologic toxicity, grade 3 mucositis, and grade 4 oliguria, elevated aspartate aminotransferase/alanine aminotransferase, multiorgan failure, and respiratory insufficiency. Plasmapheresis did not improve her clinical situation, and she died 12 days after the start of treatment. Based on the severe toxicity and the pharmacokinetic data (see pharmacokinetics/pharmacodynamics of delimotecan, T-2513, and metabolites) in this patient, the dose was deescalated to the previous level of 3,600 mg/m², at which two more patients were treated. A female with colorectal carcinoma experienced a grade 3 rash on day 10, stomatitis on day 22 (DLTs), and grade 4 neutropenia for <5 days on day 17. She developed a progressive disease, after which she was taken off the study. The other patient, a female with adenoid cystic carcinoma and lung metastasis, experienced grade 3 neutropenia, and she was not given a second cycle because of the toxicities observed in the other patients treated at this dose level. Subsequently, the dose was deescalated to 2,400 mg/m², at which, in total, seven patients were treated and DLTs occurred in two patients. In two of seven patients, the second cycle was given at a 25% reduced dose of 1,800 mg/m². The dose was deescalated to 1,800 mg/m², at which, in total, seven patients were treated. Antiemetics were used by eight patients at the following dose levels: 150 mg/m², n = 1; 1,800 mg/m², n = 3; 2,400 mg/m², n = 2; 3,600 mg/m², n = 1; and 5,400 mg/m², n = 1.

Hematologic and nonhematologic toxicity. Hematologic and nonhematologic toxicities are outlined in Table 2. A total of 98 hematologic drug–related adverse events occurred in 14 patients (63.6%), of which 76 (77.5%) were ≤grade 2 and 22 (22.5%) were grade 3/grade 4 (nine leukocytopenia, six neutropenia, five thrombocytopenia, two anemia). In total 217 nonhematologic drug–related adverse events occurred in 21 patients (95.5%), of which 187 (86.2%) were ≤grade 2, 29 (13.3%) were grade 3/grade 4, and 1 (0.5%) was with unclassified grade. Sixty-three gastrointestinal-related toxicities (i.e., diarrhea, nausea, vomiting, constipation, abdominal pain, stomatitis, dry mouth, dyspepsia, and pancreatitis) occurred in 13 (59.1%) patients of which 59 (93.7%) were ≤grade 2 and 4 (6.3%) were grade 3/grade 4. Treatment-related nonhematologic and hematologic toxicity of grade ≥3 occurred only at doses equal to or higher than 1,800 mg/m². At this dose level, the following grade 3 adverse events were reported: anemia (n = 1), anorexia (n = 1), diarrhea (n = 1), fatigue (n = 2), infection (n = 1), and vomiting (n = 1). Diarrhea grade 4 was not considered as DLT because it lasted only 1 day. None of the other nonhematologic toxicities were considered as DLTs and were likely disease-related.

A total of 26 serious adverse events were experienced by 11 patients (50%), of which six serious adverse events experienced by four patients (18.2%) were related to the study drug and included: grade 3 diarrhea (n = 1 at 1,800 mg/m², lasting only 1 day), grade 2 anorexia (n = 1 at 2,400 mg/m²), grade 3 stomatitis (n = 1 at 3,600 mg/m²), and grade 3 rash, grade 2 elevated creatinine, and grade 4 multiorgan failure (n = 1 at 5,400 mg/m²).

Eight patient received more than one cycle of therapy at the following dose levels: 300 mg/m² (one patient on three cycles), 600 mg/m² (one patient on three cycles), 1,800 mg/m² (one patient on five cycles and two patients on two cycles each), 2,400 mg/m² (two patients on two cycles each), and 3,600 mg/m² (one patient on two cycles). Almost all hematologic and nonhematologic toxicities started during the first cycle and resolved or improved within 1 to 14 days. Toxicities did not significantly increase in number and/or severity after multiple cycles of therapy. Thus, no signs of cumulative toxicity were observed. These data should be interpreted with caution because of the low number of cycles that were given and the low number of patients treated.

DLTs were observed at 5,400 mg/m² (n = 1 with grade 3 rash and grade 4 multiorgan failure), 3,600 mg/m² (n = 1 with grade 3 stomatitis and grade 3 rash), and 2,400 mg/m² (n = 1 with grade 4 thrombocytopenia; n = 1 with grade 3 stomatitis, grade 3 elevated γ-glutamyltranspeptidase, and grade 3 elevated alkaline phosphatase). At a dose equal or below 1,800 mg/m², no DLT was observed. As two of the first six treated patients at the 2,400 mg/m² dose level experienced DLTs, the dose of 1,800 mg/m² was overall considered the maximum tolerated dose of delimotecan. The safety assessment of the 1,800 mg/m² dose level was extended by two patients who initially received 2,400 mg/m² at first cycle.

Response. Two patients had a partial response: a patient with head and neck cancer (primary tumor, parotis) treated at 2,400 mg/m² and a patient with anal cancer treated at 1,800 mg/m². The duration of the partial response in the patient with head and neck cancer was 36 days. Patients had been heavily pretreated before inclusion in this study. The patient with head and neck cancer had been treated with cisplatin (one course), capecitabine (two courses), E7070 (two courses), and gemcitabine (one course), and the patient with anal cancer had received mitomycine (one course) and 5-FU (two courses) before the start of delimotecan treatment. Stable disease was observed in nine patients with a median duration of 77 days (range, 34-202 days) at the following dose levels: 300 mg/m² (n = 1; three cycles), 600 mg/m² (n = 1; three cycles), 1,800 mg/m² (n = 3; one, five, and two cycles, respectively), 2,400 mg/m² (n = 2; one and two cycles, respectively), and 3,600 mg/m² (n = 2; two and one cycles, respectively). The patients with stable disease had the following tumor types: mesothelioma (n = 3), colorectal cancer (n = 2), adenoid cystic carcinoma (n = 1), bronchial carcinoma (n = 1), head and neck cancer (n = 1), and adenocarcinoma of unknown primary (n = 1). Eight patients had disease progression, and three patients were not evaluable for tumor response.

Pharmacokinetics-pharmacodynamics of delimotecan, T-2513, and metabolites. Blood and urine sampling for pharmacokinetic analysis was done in all patients. The plasma concentration-time curves and AUC values for T-2513 (total and free), SN-38, SN-38G, and T-1335 are depicted in Figs. 2 and 3, respectively. The pharmacokinetic variables of T-2513 and metabolites are presented in Table 3. Delimotecan was slowly eliminated from plasma (Fig. 2) with a median t_1/2 of 109 h (22-203 h). The median apparent Cl and V_d were 21 mL/h (10-35 mL/h) and 3.3 L (0.8-5.8 L), respectively. No relationships were found between Cl and body surface area (r = 0.273, P = 0.220) and between Cl and weight (r = −0.165, P = 0.462). The interpatient variability (coefficient of variation, % CV) of the AUC of total T-2513 was 16%, 14%, and 17% at the dose levels of 1,800, 2,400, and 3,600 mg/m², respectively. The plasma AUC of free T-2513 corresponded to 0.01% (0-0.05%) of the given dose of T-2513, indicating a very low exposure to free T-2513 in plasma. As depicted in Fig. 3, the AUC of total T-2513, SN-38, and T-1335 increased with the dose of delimotecan (r = 0.857, P < 0.001 for T-2513; r = 0.513, P = 0.018 for SN-38; r = 0.760, P < 0.001 for T-1335). The AUC of SN-38 did not increase linearly with the AUC of delimotecan (r = 0.224, P = 0.389). Delimotecan was mainly excreted in the form of T-2513 conjugate in the urine (Table 3). Metabolites were present in urine with peak concentrations between 7 and 10 days. The ratio of the amount of metabolite excreted in the urine during the first 48 h over the given dose of T-2513 was 23% (12-52%) for total T-2513, 0.1% (0.03-0.3%) for free T-2513, 0.05% (0.02-0.2%) for SN-38, 0.21% (0.03-2.0%) for SN-38G, and 0.4% (0.1-1.1%) for T-1335. The decrease in WBC and absolute neutrophil count significantly correlated with the dose of delimotecan (P < 0.01 for each), and this relationships could fit best a sigmoidal E_max model (E_max = 100%, DE₅₀ = 2,525 mg/m², γ = 1.8 for WBC; E_max = 100%; DE₅₀ = 2,129 mg/m², γ = 2.0 for absolute neutrophil count). The relationship between the decrease in WBC and the dose of delimotecan is presented in Supplementary Fig. S4. No significant relationships were found between the AUC of parent drug and metabolites and the intensity of diarrhea, nausea, vomiting, fatigue, and skin rash. In the female patient treated at 5,400 mg/m², who developed lethal hepatic and renal failure, the dose-corrected AUC for SN-38, SN-38G, and T-1335 were significantly (>3 × SD) higher than in the other patients (Fig. 3). Tissue samples were obtained at autopsy with the approval of the family. Ex vivo tissue pharmacokinetic evaluation revealed high levels of free T-2513, SN-38, and T-1335 in kidney (734, 942, and 3,748 ng/g) and liver (583, 216, and 718 ng/g) compared with plasma (7, 38, and 595 ng/mL). The ratio of the concentration of free T-2513 over the concentration of total T-2513 was 2% to 15% in organs and 0.02% in plasma, indicating a predominant cleavage of delimotecan in tissues.

This is the first study investigating the safety and pharmacology of the novel prodrug delimotecan in patients with advanced solid tumors. During the past decades, synthetic drug polymers, including CPT-related compounds, have been developed to increase solubility and targeting of anticancer drugs (16, 17), and prodrugs of doxorubicin, paclitaxel, and camptothecin have been used in the clinic (18–20). For S-phase cell cycle–specific cytotoxic agents, prolonged low-dose exposures have been suggested more effective than short-lasting exposures at high concentrations (21). In addition, prolonged low-dose infusion of CPT-11 showed good tolerability and clinical activity and resulted in enhanced conversion of CPT-11 to SN-38 (22).

The toxicity of delimotecan included hematologic and nonhematologic adverse events, consisting mainly of grade 3/grade 4 leucocytopenia and neutropenia and grade 3 skin rash, fatigue, and diarrhea. Symptoms of the cholinergic syndrome (e.g., early onset diarrhea and flushing), which have been observed after CPT-11 therapy, as a result of rapid reversible inhibition of acetylcholinesterase by CPT-11 (23), were not reported in this study.

Clinical activity was observed in this group of heavily pretreated patients. Two partial responses were observed in a patient with anal cancer (1,800 mg/m²) and a patient with head and neck cancer (2,400 mg/m²).

The high plasma AUC of SN-38, SN-38G, and T-1335 suggests relatively high metabolism of T-2513 in the single patient at the 5,400 mg/m² dose level compared with the other patients. In this patient, pharmacokinetics was also evaluated in tissues to better understand the reason for severe toxicity. The high exposure to SN-38 and T-1335 in liver and kidney compared with the other organs in this patient could be a result of a high cleavage of delimotecan to free T-2513 and a high metabolic conversion to these metabolites. This might have contributed to the severe hepatic and renal toxicity in this single patient. Furthermore, polymorphisms in UDP glucuronosyltransferase 1A1 (UGT1A1) genes have been associated with increased levels of SN-38 and conjugated bilirubin and with severe toxicity (neutropenia, diarrhea) after CPT-11 treatment at high dose levels (24–26). However, the glucuronidation ratio was relatively high, and only a moderate increase in unconjugated bilirubin was observed in this particular patient who experienced liver toxicity, suggesting no decrease in glucuronidation capacity. The observed skin rash in the patients was generalized erythematous and nonpruritic nor desquamating. Possibly, the skin rash was related to SN-38 and/or T-1335 because delimotecan, T-2513, and carboxymethyldextran did not cause allergic or immunogenic reactions in mice. The application of a compound that inhibits cathepsin or the main metabolic enzymes involved in T-2513 biotransformation might reduce the exposure to the active CPTs, thereby reducing severe toxicity of delimotecan.

The pharmacokinetics of delimotecan, demonstrating a low Cl and V_d and a long t_1/2, suggests that delimotecan was present for prolonged periods of time, mainly in the systemic circulation. The pharmacokinetics in all patients showed a steep increase in the systemic exposure to total T-2513 at indicated doses. Furthermore, SN-38, SN-38G, and T-1335 were the main metabolites formed. SN-38 was detectable in the plasma of patients who received doses of 1,200 to 5,400 mg/m² delimotecan. The mean AUC was 1,331 h μg/L corresponding to 3,395 h nmol/L, which was 2-fold to 7-fold higher than the values of 507 to 2,080 h nmol/L reported after CPT-11 treatment at doses of 7.5 to 40 mg/m²/d (27–30). This shows that the large AUC and long t_1/2 of total T-2513, after enzymatic conversion, result in a significant systemic exposure to SN-38. In this study, the total (lactone plus carboxylate) forms of T-2513 and SN-38 were measured. Total drug concentrations highly correlated with lactone concentrations for CPT-11 and SN-38 in patients who received short drug infusions (31).

In conclusion, delimotecan, given as a 3-hour infusion at a dose of 1,800 mg/m², is tolerated with only mild to moderate gastrointestinal toxicity and skin rash without relevant hematologic toxicity. At this dose level, which is established as maximum tolerated dose, clinical activity was observed. The pharmacokinetics of delimotecan showed a long t_1/2 and a significant exposure to active CPT derivatives, including SN-38 at the 1,800 mg/m² dose level. Therefore, based on its preliminary antitumor activity, safety profile, and pharmacokinetic profile, we recommend phase II testing of delimotecan given as 3-hour infusions once every 6 weeks at a dose level of 1,800 mg/m².

No potential conflicts of interest were disclosed.

We thank Aurora Pacurar and Annelies Hiemstra for data management and the medical and nursing staff of the Netherlands Cancer Institute-Antoni van Leeuwenhoek Hospital and the University Medical Center Utrecht for the care and support of the patients in this study.