Dolastatin 10 was isolated in 1987 from the Indian Ocean marine mollusc, the sea hare (Dolabella auricularia) and was found to possess both in vitro and in vivo anti-tumor activity (₁,₂,₃,₄). TZT-1027 (Soblidotin), a synthetic tetrapeptide derivative of dolastatin 10, is a mitotic spindle poison that inhibit microtubule polymerization by interacting with tubulin in a similar domain as the vinca alkaloid binding region (₁, ₂).

TZT-1027 has a broader range of anti-tumor activity in vitro and in vivo against a variety of tumor types including those that are taxane-resistant and vincristine-resistant (₅). This may be because TZT-1027 also inhibits monosodium glutamate-induced tubulin polymerization and thus implicates two binding sites, one of high affinity and the other of low affinity. TZT-1027 inhibits both guanosine 5'-triphosphate (GTP) binding to tubulin as well as GTP hydrolysis on tubulin while vinblastine only inhibits GTP hydrolysis. Thus, the binding sites are not completely identical (₆). The poisoning of microtubules causes cell cycle arrest at the G₂/M transition phase resulting in apoptosis (₇).

TZT-1027 is approximately 30 times more potent than vincristine and about 2,000 times more potent than cisplatin against 8 human tumor cell types. The cytotoxic action of TZT-1027 in tumor cells is time-dependent as activity is enhanced with longer exposure periods rather than by increases in concentration, with cytotoxicity peaking with exposures of 12 hours or more (₈, ₉). TZT-1027 cytotoxicity is less affected by the over-expression of multi-drug resistant proteins (e.g. P-glycoprotein, BCRP, MRP) compared to other tubulin inhibitors (vincristine, paclitaxel, or docetaxel).

Dose intensity is best achieved with repeat intermittent weekly dosing. Intravenous doses of TZT-1027 produce a greater than 80% tumor growth inhibitory response and increased the life span of syngeneic mice implanted with colon 26 adenocarcinoma, M5076 reticulum cell sarcoma, or B16 melanoma. TZT-1027 also showed significant anti-tumor activity against vincristine-resistant, cisplatin-resistant, and 5-FU-resistant P388 leukemia cells inoculated into mice (₁₀).

Early Phase I clinical trials of TZT-1027 lead to the recommendation for Phase II studies of doses of 2.4 mg/m ² on day 1 and day 8 every 21 days (₁₁) or 2.7 mg/m ² every 3 weeks (₁₂). More recently, in combination with carboplatin area under the plasma concentration versus time curve (AUC) 5 on day 1, the recommended dose is 1.6 mg/m ² on day 1 and 8 every 3 weeks. (₁₃).

The reason to combine TZT-1027 with gemcitabine is based on non-overlapping mechanisms of action of the two drugs, their potential for broad activity, and the prolongation of survival seen in animal models (₁₄). Based on the early Phase I studies, a weekly schedule with doses delivered on days 1 and 8, every 21-days of administration was selected for the combination of TZT-1027 and gemcitabine. Myelosuppression was anticipated to be the dose limiting toxicity (DLT) of this combination.

• To determine the maximum tolerated dose (MTD) and the DLTs of TZT-1027 and gemcitabine in combination on day 1 and day 8 of a 21-day course.

• To determine the toxicity profile,

• To characterize possible pharmacokinetic interactions between the two drugs,

• To observe any anti-tumor activity.

• Patients > 18 years old

• Histologically confirmed locally advanced or metastatic cancer

• ECOG performance status 0 - 2

• Life expectancy > 12 weeks

• Hematology: neutrophils >1.5 x 10 3 /mm 3 and platelet counts >100 x 10 3 /mm 3

• Chemistry: serum creatinine and a total bilirubin level of < 1.5 x upper limit of normal; a transaminase level of < 2.5 x upper limit of normal (< 5 in case of metastases in the liver)

• At least four weeks had elapsed since prior surgery, radiation or chemotherapy

• Minimally pretreated (MP), i.e., had received fewer than 6 courses of an alkylating agent-containing chemotherapy, six or fewer courses of carboplatin, fewer than 2 courses of mitomycin C or a nitrosourea as a part of a single regimen, and had not been treated with prior bone marrow transplantation or radiation therapy to >25% of hematopoietic reserves.

• Signed informed consent.

• No progression of disease during earlier treatment with gemcitabine

• No concurrent radiotherapy, surgery or chemotherapy

• No women with the potential of becoming pregnant, unless utilizing birth control

• No symptomatic brain metastases

• No previous or current malignancies (except in situ carcinoma of the cervix or non-melanoma cancer of the skin)

• No mental disability or incompetence to give informed consent

• No concurrent life threatening illness

• No administration of any investigational drug within 28 days prior to receiving TZT-1027

• No history of hypersensitivity to gemcitabine

• No neuropathy at baseline ≥ grade 2

• No cardiac ejection fraction of < 40% by multiple gated acquisition cardiac scan (MUGA).

Before entry

• Complete blood cell counts

• Medical history

• Physical examination with routine neurological examination, weight, vital signs, performance status

• Blood chemistry

• Adverse event collection. Adverse events were graded according to the National Cancer Institute Common Toxicity Criteria (NCI-CTCAE - Version 3.0, revised 2003).

• Chest x-ray and other indicated imaging studies

• MUGA scan

• Survival status

Gemcitabine was obtained from a commercial source. TZT-1027 Injection for parenteral use in clear glass vials containing 2.5 mg of TZT 1027 in 5 mL of sterile, clear, colorless, aqueous solution was supplied Daiichi Pharmaceutical Corporation.

The dose was administered in 250 mL 0.9% sodium chloride as an intravenous infusion over one hour into a peripheral vein or through a central venous catheter using a programmed peristaltic pump system. To allow optimal distribution of gemcitabine to the tumor, TZT-1027 which has an antivascular effect in animal models was administered after gemcitabine (₁₆). One course consisted of a 30-minute infusion of gemcitabine followed by a 60-minute infusion of TZT-1027 weekly for 2 consecutive weeks every 3 weeks. The starting dose of gemcitabine and TZT-1027 were 800 mg/m ² and 1.6 mg/m ² , respectively. The TZT-1027 dose was chosen to be 2/3 of the dose determined in single agent phase I study (₁₁). Increments of 0.4 mg/m ² of TZT-1027 were planned into a 3+3 phase I design. Re-treatment on day 8 per protocol required neutrophils > 0.75 x 10 ³ /mm ³ , platelet counts >75 x 10 ³ /mm ³ , and for all non-hematologic toxicities to have recovered to baseline level or grade < 1, except alopecia. Treatment delays to allow recovery from toxic effects could not exceed 2 weeks. Reduction of doses to 1.2 mg/m ² of TZT-1027 with gemcitabine 800 mg/m ² was allowed if patients were being treated at dose level one). Dose re-escalation was not permitted. Continuation of treatment was dependent on the response and tolerance, usually until progression of disease occurred. Patients who experienced further DLTs despite dose delay or dose reduction were discontinued from therapy.

DLT definition : A 3+3 Simon design was utilized.

• Grade 4 neutropenia either lasting more than 5 days or with a fever above 38.5 o C, thrombocytopenia (< 25 x 10 3 /mm 3 )

• Grade 4 vomiting with maximum supportive care

• Grade 3 or 4 neurotoxicity

• Other grade 3 adverse events (excluding nausea and vomiting) requiring an adjustment of treatment administration, and inability to receive the second administration (Day 8) in the first course because of persistent toxicity or to start a second course of treatment after a 1 week delay (Day 29).

Doses of gemcitabine and TZT-1027, just below the dose level at which 2 patients of a cohort (of 3 or 6 patients) experience DLT during the first course. Additional patients were treated at the MTD.

Blood samples were collected on day 1 during the first course of the study. Samples were taken before the start of the infusion, then post infusion at 30 minutes (end of gemcitabine infusion), at 1.0, 1.5 (end of TZT-1027 infusion), 2.0, 3.0, 5.5, 7.5, and 24 hours.

Separate heparinized samples (5.0 ml/time point) were collected for analysis of gemcitabine, TZT-1027 and metabolites Ma (a hydroxyl TZT-1027 compound), and Mc (a demethylated TZT-1027 compound). For gemcitabine specimens, 50 µL of tetrahydrouridine was added to each heparinized 5 mL Vacutainer tube. All samples were centrifuged and plasma stored at -20ºC before shipping to MDS Pharma Services (Montreal, Canada).

Plasma levels of TZT-1027 and gemcitabine were measured by a validated high performance liquid chromatography (HPLC) method and mass spectrometric detection, with the lower limit of quantitation for TZT and gemcitabine being 0.250 mcg/L and 10.0 mcg/L, respectively (₁₇, ₁₈). The plasma collections were used to determine the pharmacokinetic parameters by non-compartmental analyses by using the computer program Kinetica (version 4.3) (₁₉).

The pharmacokinetic parameter calculations were normalized with the individual body surface area measurements. The following parameters were evaluated: (maximum plasma concentration (C_max), time of maximum plasma concentration (T_max), AUC`, elimination half-life (T_1/2), volume of distribution (Vd), and clearance (CL), and mean residence time (MRT). The linearity of the TZT-1027 and gemcitabine plasma kinetics in relation to dose and the relationship between toxicity and the pharmacokinetic parameters C_max, AUC, and CL were assessed.

In human plasma, TZT-1027 is predominantly bound to α1-AGP (₂₀). The correlation between plasma α1-AGP concentrations and TZT-1027 clearance were measured by MDS Pharma Services Central Lab in Mississauga, Ontario, Canada.

Patient characteristics were tabulated for safety and efficacy data.

PK parameters were summarized for each dose group and compared to other pharmacokinetics studies of TZT1027 administered alone (1027E-PRT001 (₁₂), 1027E-PRT002 (₁₁) and 1027A-PRT009 (₁₃), and with published pharmacokinetics results for gemcitabine (₂₁, ₂₂). Analysis of variance (ANOVA) was performed using these studies as a fixed effect on the ln-transformed AUC_inf/Dose, Cmax/Dose, CL and CL by body surface area (BSA) for TZT-1027. Each ANOVA included calculation of study least-squares means (LSM). The above statistical analyses were done using the SAS ^® general linear model (GLM) procedure.

For each patient, the observed α1-AGP values were plotted against TZT-1027 CL and the correlation between those values was investigated using linear regression, Emax and sigmoidal Emax models. The goodness of fit of the predicted values from each model was based on the Akaike information criterion methodology (₂₃).

All adverse events are summarized together because of the small dose increment (25%) between cohort 1 and 2 and the small number of patients. A total of 66 courses were administered. Adverse events possibly related to drug administration are presented in the table

Figure 2

Most common drug-related adverse event observed during all courses in both cohorts. If a patient experienced multiple occurrences of an individual adverse event, it is counted only once, at the highest observed intensity.

There were 12 instances of dose delay, 8 were due to persistent neutropenia, to levels. No dose delay exceeded one week.

One patient developed a DLT of grade 4 neutropenia lasting >5 days. Another patient developed grade 3 fatigue which was attributed by the investigator to rapidly progressing disease. This patient was replaced. In all courses, grade 3 or 4 neutropenia was seen in a total of 6 patients (2 grade 4 and 4 grade 3). Grade 1 neutropenia was seen in 2 patients and grade 1 thrombocytopenia in 1 patient. Anemia was seen in 2 patients (1 grade 2 and 1 grade 3). Non-hematologic events included 3 instances of grade 2 fatigue, 2 instances of grade 2 diarrhea, 1 instance each of grade 1 and 2 emesis, and 2 and 1 instances of grade 1 and 2 nausea, respectively. There were two serious adverse events in course 1, including one patient hospitalized for an ileus unrelated to the treatment and the patient described above with grade 3 fatigue, who had progressive disease after the first course of treatment. Therefore, only 1 patient in this cohort experienced a DLT that was unequivocally related to TZT-1027.

There were 2 episodes of grade 3-4 neutropenia and one episode of grade 3 thrombocytopenia. The grade 4 neutropenia occurred in course 1 and was considered a DLT. Fatigue, nausea and vomiting were limited to grades 1-2. No diarrhea was reported at this dose level. The level of myelosuppression seen in cohort 2 made it clear that the weekly administration was not feasible at these doses. This cohort was not expanded because the sponsoring company decided to stop the development of this TZT-1027.

Fourteen patients had a PK profile on day 1.

Figure 3

These results are consistent with previous pk reports of tzt-1027 demonstrating a linear behavior (₁₁, ₁₂) as described in the table below which shows a comparison of pharmacokinetics parameters between 4 studies, in patients treated at similar doses. Each parameter is a function of the dose.

Figure 4

The p-values derived from the ANOVA did not show a statistically significant difference (p-value >0.05) between the ln-transformed PK parameters AUC_inf/Dose, Cmax/Dose, CL and CL/BSA of TZT-1027 when given concomitantly with gemcitabine (this study) versus administered alone (Daiichi 1027E-PRT001, 1027E-PRT002, and 1027E-PRT009 (course 1, day 8)). Furthermore, the PK of gemcitabine, as assessed by comparison of the PK parameters, CL and t½, to values previously reported in the literature, did not appear to be affected by the concurrent administration of TZT-1027

Based on these results, the PK disposition of TZT-1027 and gemcitabine did not appear to be altered when they were coadministered as a single dose. As shown in the next table.

Mean (Range) Pharmacokinetic Parameters Unadjusted for BSA for Gemcitabine in Plasma – Course 1, Day 1 (with TZT-1027)

Figure 5

One patient with transitional cell carcinoma of the bladder experienced a confirmed partial response. This patient was in cohort 1 (gemcitabine 800 mg/m ² and TZT-1027 1.6 mg/m ² ). At baseline, this patient had 4 measurable lesions: 2 lymph nodes, 1 lung and 1 liver metastases. After the second course the total measurement of these lesions shrunk by 31%. This was confirmed after the fourth course. After the sixth course, the size of these lesions increased by 18% and the patient was removed from the study. The duration of partial response was 12 weeks. Stable disease was seen in patients with the following primary tumor types: non-small cell lung cancer (2 patients), breast (2 patients), rectum, adrenal, uterine, colon, ovarian, soft tissue sarcoma and unknown primary (1 patient in each). In the patient with colon cancer, the disease remained stable for 5.4 months and in the patient with uterine cancer, for 6.5 months. Two patients with pancreas and ovarian cancer had immediate progressive disease.