A phase I pharmacokinetic and safety study of cabazitaxel in adult cancer patients with normal and impaired renal function

This was an open-label, multicenter, phase I study (NCT01527929) in patients with advanced solid tumors and varying degrees of stable, chronic renal impairment or normal renal function. Patients were enrolled into one of three cohorts at seven institutions across five countries. Renal function cohorts were defined by creatinine clearance (CrCL), calculated using the Chronic Kidney Disease Epidemiology Collaboration equation (CKD-EPI) formula [18‐20]. Patients were enrolled into Cohort A: normal renal function (CrCL >80 mL/min/1.73 m²), Cohort B: moderate renal impairment (CrCL ≥30 to <50 mL/min/1.73 m²) or Cohort C: severe renal impairment (CrCL <30 mL/min/1.73 m²).

Cabazitaxel was provided as a sterile non-pyrogenic solution in 60-mg vials, diluted into a premix solution prior to use and administered within 8 h of preparation. All patients received 1-h intravenous (IV) infusions of cabazitaxel on Day 1 of 3-weekly cycles until unacceptable toxicity, disease progression, withdrawal of consent, investigator decision or study cutoff. Cabazitaxel starting doses were based on the renal function stratification. Based on available PK data for patients with normal renal function and moderate renal impairment, no significant changes of cabazitaxel PK were expected in these patients. Therefore, the approved cabazitaxel dose of 25 mg/m² was administered to patients with normal renal function (Cohort A) or moderate renal impairment (Cohort B). For patients with severe renal impairment, cabazitaxel starting dose was 20 mg/m², escalated to 25 mg/m² at later cycles if no dose-limiting toxicities (DLTs) were observed during Cycle 1. For Cohort C, PK assessment was carried out at the two dose levels for patients who received 25 mg/m² at later cycles. The 20 mg/m² starting dose was based on a possible increase in the free fraction of cabazitaxel in patients with severe renal impairment due to its high plasma protein binding (91.6%, mostly to albumin and lipoproteins) and the frequency of hypoalbuminemia in such patients. DLTs were defined as the following cabazitaxel-related adverse events (AEs, as assessed by the investigator): grade 2 vomiting and/or diarrhea; grade 3–4 non-hematologic AE (excluding grade 3 fatigue and transaminase or bilirubin elevation that returned to baseline prior to next treatment cycle); hematologic toxicity, defined as neutropenic infection, febrile neutropenia (fever of unknown origin without documented infection, with grade 3–4 neutropenia), grade 4 neutropenia lasting >7 days, or grade 3–4 thrombocytopenia. Toxicity and AEs were graded and recorded according to National Cancer Institute Common Terminology Criteria for AEs (NCI CTCAE) v4.03 [21]. Granulocyte-colony stimulating factor (G-CSF) was allowed with therapeutic or prophylactic intent and left to the investigator's judgement.

Patients had a diagnosis of a histologically or cytologically proven non-hematologic malignancy that was refractory to standard therapy or for which no standard therapy was available, and for which cabazitaxel was judged to be an adequate treatment option by the investigator. Eligible patients were ≥18 years of age with a life expectancy of >3 months, an Eastern Cooperative Oncology Group (ECOG) performance status (PS) of 0, 1 or 2, stable renal function (defined as CrCL within a range of ±10% during a 3-month period with ≥3 measurements performed), and adequate liver and bone marrow function (defined as absolute neutrophil count ≥1.5 × 10⁹/L, platelets ≥100 × 10⁹/L, total bilirubin ≤1.0 × institution upper limit of normality [ULN], transaminases ≤2.5 × ULN, and alkaline phosphatase ≤2.5 × ULN). Patients must have completed prior anticancer therapy ≥4 weeks before study entry. Key exclusion criteria included neurotoxicity of grade ≥2, acute renal failure or dialysis that would be required during the study, history of hypersensitivity to docetaxel or polysorbate 80, known brain metastases, and any treatments known to strongly induce CYP3A isoenzymes or to strongly inhibit CYP3A4 activity within 2 weeks before or during the test period for PK sampling. The institutional review board at each participating institution approved the protocol. Written informed consent was obtained from all patients. The study was conducted in accordance with the ethical principles stated in the Declaration of Helsinki and good clinical practice.

Medical histories were recorded at baseline and assessment of vital signs, physical examinations, ECOG PS, and electrocardiograms were performed at baseline prior to cabazitaxel administration, during study treatment as required and ≥30 days after the last administration of study treatment. All signs and symptoms observed from the time of informed consent were recorded as AEs. All AEs were recorded until 30 days after last administration of study treatment. After this follow-up period, only new or ongoing treatment-related AEs were recorded, except for ongoing serious AEs, which were assessed until resolution or stabilization, regardless of relationship to study treatment. Weekly laboratory evaluations were performed including cell blood counts with differentials, and analysis of coagulation, liver function, plasma electrolytes, glucose, albumin, total proteins, blood urea nitrogen, creatinine and urinalysis. An eye examination was performed at baseline and end of study. CrCL was determined using the CKD-EPI formula and if CrCL decreased by 50% from baseline during the study, or if dialysis was required, treatment was discontinued. The maximum dose delay allowed, due to acute toxicity, was 2 weeks. If the treatment gap was longer than 2 weeks, the patient was discontinued from study treatment. If the dose was reduced due to toxicity, then it was not re-escalated. Up to a maximum of two dose reductions were allowed per patient. Radiologic studies for disease assessment were conducted pretreatment and according to the site practice. Tumor response was assessed according to the Response Evaluation Criteria in Solid Tumors (RECIST 1.1).

Heparinized blood samples were collected from all patients for cabazitaxel concentration measurement at Cycle 1 (and Cycle 2, or Cycle 3 for one patient, following dose escalation in Cohort C) before start of infusion, 5 min before the end of infusion and then 5, 15 and 30 min, and 1, 2, 3, 5 and 8 h post-infusion, and at approximately 24 (Day 2), 48 (Day 3), 72 (Day 4), 120 (Day 6), 168 (Day 8) and 216 (Day 10) h after the end of infusion. The timing of treatment administration and timing of sampling were precisely recorded for each patient. Total cabazitaxel concentrations in the plasma were determined using a validated liquid chromatography with tandem mass spectrometry method (LC–MS/MS) with a lower limit of quantification (LLOQ) of 1 ng/mL [22, 23]. In addition, blood samples were collected to determine the free, unbound fraction of cabazitaxel in all patients at Cycle 1 before start of infusion, 5 min before the end of infusion, and 3 and 24 h after end of infusion. Free cabazitaxel concentrations were determined after equilibrium dialysis in buffer using a validated LC–MS/MS method with a LLOQ of 0.1 ng/mL.

The primary PK endpoints were area under the plasma concentration versus time curve (AUC) and cabazitaxel clearance (CL). Secondary PK endpoints included observed maximum plasma concentration (C _max), volume of distribution at steady state (V _ss) and elimination half-life (t _1/2ʎ3). V _ss and CL normalized by body surface area (BSA; V _ss/BSA and CL/BSA) were also calculated. Total plasma concentrations of cabazitaxel and relative actual time values (as well as actual dose) were used to calculate the PK parameters using non-compartmental analysis (for C _max) and individual modeling using a three-compartment open model with first-order elimination (for CL, AUC, V _ss and t _1/2ʎ3). The calculation of PK parameters was performed using validated softwares (PKDMS version 2 running with WinNonlin Professional, version 5.2.1, Pharsight and WinNonlin Professional, version 6.3, Phoenix, Pharsight), as described previously [24, 25].

Sample size for this study was based on empirical considerations and on the experiences in previous population studies; no formal sample size calculation was performed. A total of 24 to 27 patients were expected to be enrolled (at least 8 patients enrolled and evaluable for final PK evaluation in each cohort). One patient with a major deviation (missing PK sample at a critical time point) was excluded from the PK analysis. Statistical analysis evaluated the effect of population group on cabazitaxel PK parameters by modeling the relationship between measures of renal function (CrCL) and the PK parameters with a regression model as recommended by the United States Food and Drug Administration (FDA) and European Medicines Agency (EMA) [12, 13]. Log-transformed PK parameters were analyzed using a linear regression model with the independent variable being log-transformed CrCL at screening. Log-transformed BSA was a covariate in the models except in models using BSA normalized parameters. The effect of renal impairment on cabazitaxel PK parameters was analyzed using a linear fixed effects model. Geometric mean estimates were determined using log CrCL values corresponding to the mean boundaries of the CrCL interval covering the patient cohorts with moderate renal impairment (40 mL/min/1.73 m²) and severe renal impairment (15 mL/min/1.73 m²), along with a value representing the normal population and defined as the control (90 mL/min/1.73 m²). Using the regression model parameter estimates, point estimates for PK parameters corresponding to CrCL of 90, 40 and 15 mL/min/1.73 m² were calculated after converting these values to the log scale. Geometric mean estimates were computed, and estimates for the geometric mean ratio of each population group (40 and 15 mL/min/1.73 m²) versus the control population group (90 mL/min/1.73 m²) were calculated. A similar analysis was performed for the models using BSA normalized parameters except log BSA was not used. Cabazitaxel unbound fraction was measured longitudinally in each patient. To estimate cabazitaxel unbound fraction, a linear mixed model was used with log BSA, log CrCL at screening and time as fixed effects, and random intercept and time slope as random effects.