Abstract
Background
Interindividual pharmacokinetic variability may influence the clinical benefit or toxicity of cabozantinib in metastatic renal cell carcinoma (mRCC). We aimed to investigate the exposure-toxicity and exposure-response relationship of cabozantinib in unselected mRCC patients treated in routine care.
Methods
This ambispective multicenter study enrolled consecutive patients receiving cabozantinib in monotherapy. Steady-state trough concentration (Cmin,ss) within the first 3 months after treatment initiation was used for the PK/PD analysis with dose-limiting toxicity (DLT) and survival outcomes. Logistic regression and Cox proportional-hazards models were used to identify the risk factors of DLT and inefficacy in patients, respectively.
Results
Seventy-eight mRCC patients were eligible for the statistical analysis. Fifty-two patients (67%) experienced DLT with a median onset of 2.1 months (95%CI 0.7–8.2). In multivariate analysis, Cmin,ss was identified as an independent risk factor of DLT (OR 1.46, 95%CI [1.04–2.04]; p = 0.029). PFS and OS were not statistically associated with the starting dose (p = 0.81 and p = 0.98, respectively). In the multivariate analysis of PFS, Cmin, ss > 336 ng/mL resulted in a hazard ratio of 0.28 (95%CI, 0.10–0.77, p = 0.014). By contrast, Cmin, ss > 336 ng/mL was not statistically associated with longer OS.
Conclusion
Early plasma drug monitoring may be useful to optimise cabozantinib treatment in mRCC patients treated in monotherapy, especially in frail patients starting at a lower than standard dose.
Subject terms: Renal cancer, Cancer therapy
Background
The management of metastatic renal cell carcinoma (mRCC) has dramatically improved over the last 15 years as a result of the development of increasingly effective VEGFR-targeted tyrosine kinase inhibitors (TKIs) [1]. Cabozantinib belongs to this generation of powerful VEGFR2 inhibitors but also targets cMET and AXL, two signalling pathways involved in resistance to first-generation TKIs such as sunitinib. Following the results of the METEOR trial, cabozantinib was approved in the US and Europe for the treatment of mRCC in patients who progressed on first-line VEGF-targeting TKI [2]. More recently, it has also been approved as first-line therapy in mRCC in combination with nivolumab, an immune checkpoint inhibitor targeting PD-1 [3].
While the efficacy of cabozantinib as a monotherapy has been largely demonstrated, its toxicity remains severe. In the pivotal METEOR study, 60 mg of cabozantinib daily was responsible for directly related grade 3 or 4 events in 75% of patients, and dose reduction and permanent discontinuation were reported in 60% and 9% of patients, respectively [2]. Evaluating the pharmacokinetic/pharmacodynamic (PK/PD) relationship is of value in optimising the safety/efficacy balance of TKIs [4, 5]. Only a few reports on the PK/PD relationship of cabozantinib are available. A population PK model developed on data from the METEOR trial reported increasing individual mean plasma concentrations according to the cabozantinib starting dose: 375 ng/mL (20 mg), 750 ng/mL (40 mg) and 1125 ng/mL (60 mg), respectively [6]. The authors further reported that starting cabozantinib at a lower dose of 20 or 40 mg/day was associated with an increased hazard of disease progression or death compared with a 60 mg/day dose (HR 1.39 [1.29–1.49] and HR 1.10 [1.07–1.12] respectively), as well as with lower objective response rates (19.1% vs 15.6% and 8.7%, respectively). They also reported a potentially higher risk for selected adverse events with the 60 mg exposure relative to the 20 mg exposure. Given the modest increase in the risk of death (HR 1.10) and a better safety profile with the 40 mg starting dose, a threshold value of 750 ng/mL for efficacy was suggested for steady-state average plasma concentration (Cav,ss) [7]. However, this threshold value was not confirmed in two recent retrospective PK/PD studies [8, 9]. Cerbone et al. attempted to define an optimal range for limiting toxicity without impairing efficacy in 76 mRCC patients [9]. The proposed threshold of steady-state trough concentration (Cmin,ss) for efficacy was lower than that of the pivotal study (536 vs 750 ng/mL, respectively), and the therapeutic range proposed (i.e., 536–617 ng/mL) was too narrow for any clinical application in routine care.
To better define the optimal thresholds of cabozantinib exposure to improve the safety/efficacy balance, we conducted an ambispective real-world study evaluating exposure-response relationships for efficacy and toxicity in patients treated for mRCC with cabozantinib in a real-world setting.
Methods
Study population and data collection
Between December 2016 and May 2021, we routinely monitored plasma concentration within the first 3 months of treatment in all consecutive mRCC patients treated with single-agent cabozantinib in three French university hospitals (Georges Pompidou European Hospital (Paris), Cochin University Hospital (Paris) and Henri Mondor Hospital (Créteil)). Baseline demographic, biological and pathological data, as well as clinical events (toxicity, disease progression or death), were retrospectively retrieved from the electronic medical records. This study was conducted in accordance with the 2008 Declaration of Helsinki. It was approved by the local ethics committee in oncology (CLEP number: AAA-2022–08055). Informed consent was obtained from all patients prior to inclusion. Additionally, an external validation cohort enrolling mRCC patients from two other French cancer centres, Gustave Roussy Hospital (Villejuif, France) and IUCT-Oncopole Claudius-Regaud (Toulouse, France), was created. Unselected mRCC patients treated with single-agent cabozantinib and at least one plasma concentration available within the first 3 months of treatment were eligible.
Procedures
The recommended starting dose of cabozantinib is 60 mg/day. However, a lower starting dose (20 to 40 mg/day) could be prescribed for fragile or elderly patients at the physician’s discretion. Subsequently, doses could be adjusted in 20 mg increments or decrements based on tolerance. During the treatment period, toxicities and laboratory analyses (blood cell count, renal and liver function) were assessed every 2 weeks for the first 3 months and then monthly. All adverse events were prospectively graded according to the National Cancer Institute – Common Toxicity Criteria for Adverse Events (NCI-CTCAE) version 5.0. In the event of grade 3 or 4 toxicity, cabozantinib was suspended until improvement to grade 1–2. A periodical radiological assessment was performed every 3 months, earlier if a symptomatic disease progression was suspected. Cabozantinib was administered until clinical or radiological disease progression, unacceptable adverse events or death.
Pharmacokinetics
Blood samples were collected at steady-state concentrations during routine follow-up visits to the outpatient clinic. The mean elimination half-life of cabozantinib is 55 h [10]; therefore, the steady state was considered to occur at least 11 days after treatment initiation or dosing adjustment. Blood was drawn into 5 mL lithium heparinised Vacutainer tubes at any time over the administration interval. After centrifugation (3000 rpm for 5 min at 4 °C), plasma was transferred to polypropylene tubes and kept at −20 °C until assay. Plasma cabozantinib concentrations were measured using validated high-performance liquid chromatography coupled with tandem mass spectrometry (LC-MS/MS). A detailed description of the bioanalytical method is available in Supplemental Material 1. The method’s accuracy was ensured by the external quality assessment scheme provided by the Group of Clinical Pharmacology in Oncology (Unicancer, Paris, France) and Asqualab (Paris, France). The area under the concentration-time curve over the dosing interval (AUCss), Cav,ss and Cmin,ss were estimated using the Bayesian method from Lacy et al.’s population pharmacokinetic model [10]. This PK model was updated in our study using the learning dataset to consider the specificities of our population; apparent clearance (CL/F) and the associated inter-individual variability (IIV) were re-estimated. In the Bayesian method, individual patient concentrations and the PK parameters’ statistical distribution allow for estimating the conditional distribution of the individual PK parameters and, thus, the associated PK metrics (AUCss, Cav,ss, Cmin,ss), enabling us to propose the most probable value. All the calculations were performed with Monolix Suite 2021R1 (Lixoft, Antony, France).
Study endpoints
The primary endpoint was the onset of early dose-limiting toxicity (DLT) in the first 3 months of treatment. DLT was defined as any clinical or biological toxicity leading to dose reduction or temporary or permanent discontinuation. The main objective was to determine whether high plasma cabozantinib exposure (AUCss, Cmin,ss) or Cav,ss ≥ 750 ng/mL within the first 3 months of treatment was associated with an increased frequency of early DLT at 3 months. Regarding efficacy, progression-free survival (PFS) was the time between the first day of cabozantinib treatment and tumour progression or death. The best overall response (BOR) was assessed using the Response Evaluation Criterion in Solid Tumours (RECIST) 1.1 criteria [11]. Overall survival (OS) was the time between the first day of cabozantinib treatment and death (all causes included). Patients with no tumour progression or death ad at time of data collection were censored at the last date of evaluation.
Statistical analysis
For descriptive analyses, qualitative variables were expressed as numbers (%) and quantitative ones as mean ± standard deviation or median (interquartile range, IQR). Comparisons between two groups (with or without DLT) were performed using the student’s t-test for quantitative variables and the chi-square test for qualitative variables. Comparisons between the three groups were performed using non-parametric tests, with the Kruskal-Wallis test for quantitative variables and Fisher’s test for qualitative variables. The first PK sampling was taken into account for all PK statistical analyses. Given the linearity of pharmacokinetics over the 20 to 60 mg dose range [12], interindividual variability in plasma exposure (AUCss) could be investigated using dose-normalised exposure. Multiple linear regression was used to identify determinants of interindividual variability in AUCss (per 1000 ng/mL.h increase). Regarding exposure-toxicity analysis, univariate and multivariate logistic regression models were used to test the association between DLT onset and biological and clinical variables. Among these latter, three PK parameters were tested: AUCss, Cmin,ss and Cav,ss ≥ 750 ng/mL. Results were expressed as odds ratios (OR) with a 95% confidence interval (95%CI). Regarding exposure-efficacy analysis, only patients treated in the second or later line were eligible. Survival curves were obtained with Kaplan–Meier estimates and compared with the log-rank test. The proportional hazards assumption was checked for each model using graphical methods based on Kaplan-Meier curves and the scaled Schoenfeld residuals. Univariate and multivariate Cox proportional hazards regression was used to identify risk factors (including AUCss, Cmin,ss and Cav,ss ≥ 750 ng/mL) of disease progression or death. Results were expressed as hazard ratio (HR) with 95%CI. All p-values were two-sided, and the level of significance was set at p < 0.05. Statistical analyses were performed in R software (version 4.3.0, http://www.r-project.org, accessed on 21 April 2023) with RStudio (version 2023.03.0).
Results
Patients and treatment in the training cohort
Eighty patients were included in the training cohort. Two patients were excluded from statistical analysis because DLT occurred before the first PK sampling. Overall, 78 patients were included in the statistical analysis (Supplemental Fig. 1). The baseline characteristics of the training cohort are presented in Table 1. The mean patient age was 61.8 ± 13.2 years, with a 3:1 male/female sex ratio. Most patients (83%) were treated for clear cell renal cell carcinoma (ccRCC) and were in the second or later line of treatment (91%). Seventy-six per cent of patients exhibited Karnofsky performance status (KPS) ≥ 80. The cabozantinib starting dose was 60 mg/day in 36 patients (46%). Patients who started cabozantinib at 40 mg (N = 30) or 20 mg (N = 12) were significantly older, had a poorer KPS and had a lower estimated glomerular filtration rate (Table 1). Thirty per cent of patients were concomitantly treated with proton pump inhibitors (PPI). Before the PK sampling, the daily dose was increased from 20 to 40 mg/day in two patients and from 40 to 60 mg/day in four patients. The median follow-up was 17.9 months (95%CI 1.7–40.4). At the data cut-off in March 2022, 16 patients (21%) were still being treated with cabozantinib. The only reason for permanent treatment discontinuation was disease progression over the entire follow-up duration. Finally, 24 patients (31%) were still alive.
Table 1.
Demographic and baseline characteristics of study cohort.
| Variable | N | Overall | 60 mg/day | 20-40 mg/day | p-valuea |
|---|---|---|---|---|---|
| Age (years), mean ± SD | 78 | 61.8 ± 13.2 | 56.6 ± 12.4 | 66.3 ± 12.3 | 0.0009 |
| Gender, n (%) | 76 | 0.887 | |||
| Female | 19 (24) | 8 (22) | 11 (26) | ||
| Male | 59 (76) | 28 (78) | 31 (74) | ||
| Body mass index (kg/m²), mean ± SD | 73 | 26.0 ± 4.9 | 26.1 ± 5.0 | 25.9 ± 4.9 | 0.904 |
| Cell histology, n (%) | 76 | 0.766 | |||
| Clear cell | 63 (83) | 30 (86) | 33 (80) | ||
| Other | 13 (17) | 5 (14) | 8 (20) | ||
| IMDC status | 74 | 0.746 | |||
| Favourable | 20 (27) | 8 (23) | 12 (31) | ||
| Intermediate | 38 (51) | 19 (54) | 19 (49) | ||
| Poor | 16 (22) | 8 (23) | 8 (21) | ||
| Karnofsky performance scale, n (%) | 74 | 0.040 | |||
| <80 | 18 (24) | 4 (12) | 14 (35) | ||
| ≥80 | 56 (76) | 30 (88) | 26 (65) | ||
| Previous treatment line, n (%) | 78 | 0.0971 | |||
| 0 | 7 (9) | 3 (8) | 4 (10) | ||
| 1 | 28 (36) | 18 (50) | 10 (24) | ||
| 2 | 26 (33) | 10 (28) | 16 (38) | ||
| ≥3 | 17 (22) | 5 (14) | 12 (28) | ||
| Baseline albumin (g/L), mean ± SD | 61 | 37.1 ± 7.3 | 38.8 ± 6.4 | 35.5 ± 7.7 | 0.067 |
| eGFR (mL/min)*, mean ± SD | 62 | 67.9 ± 23.1 | 76.6 ± 24.2 | 61.6 ± 20.4 | 0.012 |
| Proton Pump Inhibitor intake, n (%) | 74 | 1 | |||
| Yes | 22 (30) | 10 (30) | 12 (29) | ||
| No | 52 (70) | 23 (70) | 29 (71) |
IMDC International Metastatic Renal Cell Carcinoma Database Consortium, eGFR estimated glomerular filtration rate, SD standard deviation.
*Glomerular filtration rate was estimated according to the Modification of Diet in Renal Disease (MDRD) equation.
ap-value : difference between 60 mg/day vs 20–40 mg/day.
Bold value means variable statistically significant i.e. p < 0.05.
Pharmacokinetics in the training cohort
The median time from treatment initiation to PK sampling was 16 days (95%CI 13–26.5). The median Cmin,ss, Cav,ss and AUCss were 625 ng/mL [IQR, 481–829], 704 ng/mL [IQR, 539–916] and 16,894 ng/mL.h [IQR, 12,946–21,975], respectively. The median Cmin,ss at 60 mg/day (824 ng/mL, IQR 700–905 ng/mL) was statistically higher than those observed for 40 mg/day (513 ng/mL, IQR 472–589 ng/mL; Wilcoxon test p-value < 0.0001) and 20 mg/day (311 ng/mL, IQR 223–336 ng/mL; Wilcoxon test p-value < 0.0001) (Fig. 1). The proposed target Cav,ss ≥ 750 ng/mL was achieved in 81% (29/36), 19% (6/32) and 0% (0/10) of patients at 60 mg, 40 mg and 20 mg/day, respectively. The interpatient pharmacokinetic variability in dose-normalised AUCss was 22%. Only the number of previous treatment lines was identified as a significant determinant of this variability (Supplemental Table 1). Patients treated in the third or later line had a higher Cmin,ss than others (p = 0.003).
Fig. 1.
Steady-state plasma trough concentration (Cmin,ss) of cabozantinib according to the daily dose.
Exposure-toxicity relationship in the training cohort
Fifty-two of 78 patients (67%) experienced DLT with a median onset of 2.1 months (95%CI 0.7–8.2); 75% of DLT events (39/52) occurred within the first 3 months of treatment (early DLT events). The median time from PK sampling to early DLT events was 34 days [IQR, 18–48]. No permanent toxicity-related discontinuation was observed over the entire follow-up duration. The incidence of dose reduction or temporary treatment interruption was higher in patients treated with a starting dose of 60 mg/day (75%, 27/36 patients) compared to that of patients starting at 40 mg/day or less (60%, 25/42 patients), but without a statistically significant difference (p = 0.23). Only the 39 early DLT events, including 31 treatment reductions (79%) and eight temporary treatment discontinuations (21%), were included in the exposure-toxicity analysis. In the univariate analysis, PK parameters, including Cmin,ss and AUCss, were statistically associated with an increased risk of DLT onset (p < 0.05) (Table 2). Early DLT events were more frequent in patients with Cav,ss ≥ 750 ng/mL compared to other patients (69% vs 35%, respectively; p = 0.006) (Fig. 2). In different multivariate models (all including starting dose and baseline albumin), Cmin,ss (OR 1.46, 95%CI [1.04–2.04]; p = 0.029), AUCss (OR 1.15, 95%CI [1.01–1.31]; p = 0.040) and Cav,ss ≥ 750 ng/mL (OR 6.87, 95%CI [1.44–32.73]; p = 0.016) were identified as independent risk factors of DLT (Table 2), while statistical significance was not reached for Cmin,ss ≥ 750 ng/mL (OR 3.99, 95%CI [0.92–17.35]; p = 0.065) (data not shown).
Table 2.
Univariate and multivariate logistic regression model for the risk factor of dose-limiting onset in the study cohort (n = 78 patients).
| Variable | No DLT | DLT | OR | 95%CI | p-value |
|---|---|---|---|---|---|
| Age (years), mean ± SD | 62.5 ± 13.6 | 61.1 ± 12.9 | 0.99 | 0.96–1.03 | 0.64 |
| Gender, n (%) | |||||
| Female | 8 (42) | 11 (58) | 1 | ||
| Male | 31 (53) | 28 (47) | 0.66 | 0.23–1.87 | 0.43 |
| Body mass index (kg/m²), mean ± SD | 25.6 ± 4.8 | 26.3 ± 5.1 | 1.03 | 0.94–1.13 | 0.54 |
| Baseline albumin (g/L), mean ± SD | 35.3 ± 7.2 | 39.2 ± 6.9 | 1.09 | 1.001–1.18 | 0.047 |
| eGFR*(mL/min), mean ± SD | 68.3 ± 26.7 | 66.0 ± 19.9 | 1.00 | 0.98–1.02 | 0.70 |
| Karnofsky performance scale, n (%) | |||||
| <80 | 11 (61) | 7 (39) | 1 | ||
| ≥80 | 27 (48) | 29 (52) | 1.69 | 0.57–4.98 | 0.34 |
| IMDC status, n (%) | |||||
| Favourable | 8 (40) | 12 (60) | 1 | ||
| Intermediate | 18 (47) | 20 (53) | 0.74 | 0.25–2.22 | 0.59 |
| Poor | 10 (62) | 6 (38) | 0.40 | 0.1–1.54 | 0.18 |
| Previous treatment lines, n (%) | |||||
| ≥2 | 18 (44) | 23 (56) | 1 | ||
| 0–1 | 19 (58) | 14 (42) | 0.58 | 0.23–1.46 | 0.24 |
| Starting dose (mg/day) | |||||
| 60 | 15 (42) | 21 (58) | 1 | ||
| 40 | 16 (53) | 14 (47) | 0.63 | 0.24–1.66 | 0.35 |
| 20 | 8 (67) | 4 (33) | 0.36 | 0.09–1.41 | 0.14 |
| Proton Pump Inhibitor intake, n (%) | |||||
| Yes | 11 (50) | 11 (50) | 1 | ||
| No | 26 (50) | 26 (50) | 1.00 | 0.37–2.71 | 1 |
| Target Cav,ss, n(%) | |||||
| <750 ng/mL | 28 (65) | 15 (35) | 1 | ||
| ≥750 ng/mL | 11 (31) | 24 (69) | 4.07 | 1.58–10.53 | 0.004 |
| Cmin,ss /100, mean ± SD | 5.6 ± 2.1 | 7.3 ± 2.5 | 1.37 | 1.11–1.70 | 0.004 |
| AUCss /1000, mean ± SD | 15.1 ± 5.5 | 19.2 ± 6.6 | 1.12 | 1.03–1.22 | 0.006 |
| Multivariate models | |||||
| Using Cmin,ss | |||||
| Cmin,ss (per 100 ng/mL increase) | 1.46 | 1.04–2.04 | 0.029 | ||
| Starting dose (mg/day) : 20–40 vs 60 | 2.27 | 0.54–9.56 | 0.26 | ||
| Baseline albumin (per 1 unit increase) | 1.08 | 0.99–1.18 | 0.07 | ||
| Using AUCss | |||||
| AUCss (per 1000 ng/mL.h increase) | 1.15 | 1.01–1.31 | 0.040 | ||
| Starting dose (mg/day) : 20–40 vs 60 | 2.21 | 0.52–9.44 | 0.29 | ||
| Baseline albumin (per 1 unit increase) | 1.08 | 0.99–1.18 | 0.071 | ||
| Using Target Cav,ss | |||||
| Cav,ss (ng/mL): ≥ 750 vs < 750 | 6.87 | 1.44–32.73 | 0.016 | ||
| Starting dose (mg/day) : 20–40 vs 60 | 2.72 | 0.57–12.92 | 0.21 | ||
| Baseline albumin (per 1 unit increase) | 1.11 | 1.01–1.22 | 0.035 | ||
95%CI 95% confidence interval, AUCss Area under the concentration-time curve over the dosing interval at steady-state, Cav,ss average concentration at steady-state, Cmin,ss trough concentration at steady-state, DLT dose-limiting toxicity, eGFR estimated glomerular filtration rate, IMDC International Metastatic Renal Cell Carcinoma Database Consortium, OR odds ratio, SD standard deviation.
*Glomerular filtration rate was estimated according to the Modification of Diet in Renal Disease (MDRD) equation.
Bold value means variable statistically significant i.e. p < 0.05.
Fig. 2.
Frequency of dose-limiting toxicity (DLT) according to the threshold value of 750 ng/mL for steady-state plasma trough concentration (Cav,ss) within the first 3 months of treatment.
Exposure-efficacy relationship in the training cohort
Seven (9%) out of 78 patients treated in the first-line setting were excluded from the statistical analysis to be consistent with the European marketing authorisation label of cabozantinib. Overall, 71 patients, including 59 ccRCC patients (83%), were eligible for the exposure-efficacy analysis (Supplemental Figure 1). Twenty-two (31%), 39 (56%) and nine (13%) patients had an objective response (PR or CR), SD, and PD, respectively, as BOR. One patient (1%) was non-evaluable, according to RECIST. No statistical difference in terms of Cmin,ss (Kruskal–Wallis test p = 0.51) (Supplemental Fig. 2A) and starting dose (Fisher test p = 0.17) (Supplemental Fig. 2B) was observed among PD, SD, PR and CR patients. Regarding survival, the median PFS and OS were 9.9 months (95%CI, 8.6–12.0) and 18.4 months (95%CI, 17.0–23.2), respectively (Supplemental Fig. 3). PFS and OS were not significantly different among patients starting at a 60, 40, or 20 mg dose (log rank-test p = 0.81 and p = 0.98 for PFS and OS, respectively) (Supplemental Fig. 4). Median PFS in patients with Cav,ss ≥ 750 ng/mL was 11.3 months (95%CI, 8.5-14.7), compared with 9.7 months (95%CI, 6.8–12.6) in patients with Cav,ss < 750 ng/mL (log rank-test p = 0.92). No statistical difference for OS was observed between these two groups (19.2 [95% CI, 17.0–25.6] vs 18.4 [95%CI, 12.7–not reached] months, respectively; log rank-test p = 0.75) (Supplemental Fig. 5). However, patients with Cmin,ss ≥ 336 ng/mL (first decile) exhibited longer PFS compared to patients with Cmin,ss < 336 ng/mL (11.3 [95% CI, 8.9–13.1] vs 4.9 [95%CI, 2.5–not reached] months, respectively; log rank-test p = 0.025) (Fig. 3a). In multivariate analysis, Cmin,ss ≥ 336 ng/mL resulted in an HR of 0.28 (95%CI, 0.10–0.77, p = 0.014) (Table 3). By contrast, Cmin,ss ≥ 336 ng/mL was not significantly associated with longer OS (Fig. 3b). In the 59 patients with clear cell histology (sensitivity analysis), the median PFS and OS were 8.9 months (95%CI, 6.6–11.3) and 18.4 months (95%CI, 17.0–31.2), respectively. Finally, the multivariate Cox proportional hazards models of the sensitivity analysis indicated that Cmin,ss ≥ 336 ng/mL may be an independent protective factor of disease progression, although statistical significance was not achieved by using a bilateral test (HR 0.26, 95%CI, 0.06–1.07, p = 0.061) (Supplemental Table 2).
Fig. 3. Survival in patients treated in the second-line or later-line (n = 71) according to cabaozantinib plasma exposure.
Progression-free survival (a) and overall survival (b) according to the threshold value of 336 ng/mL (first decile) for steady-state plasma trough concentration (Cmin,ss).
Table 3.
Univariate and multivariate Cox proportional hazards models for risk factors of death and disease progression in patients treated in the second-line or later-line (n = 71).
| Variable | Progression-free survival | Overall survival | ||||
|---|---|---|---|---|---|---|
| HR | 95%CI | p-value | HR | 95%CI | p-value | |
| Univariate analysis | ||||||
| Age (per 1 unit increase) | 0.990 | 0.964–1.017 | 0.46 | 1.007 | 0.979–1.037 | 0.62 |
| Gender (male vs female) | 0.690 | 0.356–1.340 | 0.27 | 0.393 | 0.207–0.746 | 0.0043 |
| Body mass index (per 1 unit increase) | 0.949 | 0.889–1.014 | 0.12 | 0.968 | 0.905–1.036 | 0.35 |
| Baseline albumin (per 1 unit increase) | 0.965 | 0.922–1.010 | 0.13 | 0.922 | 0.878–0.967 | 0.0009 |
| eGFR* (per 1 unit increase) | 1.007 | 0.993–1.020 | 0.32 | 1.004 | 0.990–1.017 | 0.59 |
| IMDC status | ||||||
| Intermediate vs Favourable | 1.236 | 0.667–2.290 | 0.50 | 1.675 | 0.810–3.464 | 0.16 |
| Poor vs Favourable | 1.077 | 0.442–2.626 | 0.87 | 1.814 | 0.736–4.471 | 0.20 |
| Karnofsky performance scale | ||||||
| ≥80 vs <80 | 0.508 | 0.263–0.979 | 0.043 | 0.395 | 0.205–0.759 | 0.0053 |
| Previous treatment line | ||||||
| 1 vs ≥2 | 0.871 | 0.505–1.502 | 0.62 | 0.689 | 0.385–1.231 | 0.21 |
| Starting dose (mg/day) | ||||||
| 40 vs 60 | 0.837 | 0.469–1.493 | 0.55 | 1.063 | 0.581–1.945 | 0.84 |
| 20 vs 60 | 1.015 | 0.460–2.240 | 0.97 | 1.052 | 0.452–2.448 | 0.90 |
| Cmin,ss (100 ng/mL per increase) | 0.999 | 0.888–1.125 | 0.99 | 0.993 | 0.881–1.119 | 0.91 |
| AUCss (1000 ng/mL.h per increase) | 1.002 | 0.958–1.049 | 0.92 | 0.998 | 0.953–1.046 | 0.94 |
| Target Cav,ss | ||||||
| ≥750 vs <750 ng/mL | 1.095 | 0.628–1.909 | 0.75 | 1.028 | 0.605–1.748 | 0.92 |
| Cmin,ss (First decile) | ||||||
| ≥336 vs <336 ng/mL | 0.411 | 0.184–0.916 | 0.030 | 0.76 | 0.300–1.923 | 0.56 |
| Proton Pump Inhibitor intake | ||||||
| No vs yes | 1.030 | 0.565–1.879 | 0.92 | 0.811 | 0.438–1.502 | 0.51 |
| Multivariate analysis | ||||||
| IMDC status | ||||||
| Intermediate vs Favourable | 1.881 | 0.912–3.881 | 0.09 | 1.950 | 0.845–4.498 | 0.12 |
| Poor vs Favourable | 1.259 | 0.456–3.479 | 0.66 | 1.578 | 0.551–4.517 | 0.40 |
| Baseline albumin (per 1 unit increase) | 0.960 | 0.911–1.012 | 0.131 | 0.920 | 0.871–0.972 | 0.003 |
| Cmin,ss (First decile) | ||||||
| ≥336 vs <336 ng/mL | 0.283 | 0.104–0.774 | 0.014 | 0.627 | 0.199–1.973 | 0.43 |
| Starting dose (mg/day) | ||||||
| 20-40 vs 60 | 0.747 | 0.373–1.495 | 0.41 | 0.969 | 0.456–2.060 | 0.94 |
95%CI 95% confidence interval, AUCss Area under the concentration-time curve over the dosing interval at steady-state, Cav,ss average concentration at steady-state, Cmin,ss trough concentration at steady-state; eGFR, estimated glomerular filtration rate; HR, hazard, ratio; IMDC, International Metastatic Renal Cell Carcinoma Database Consortium.
*Glomerular filtration rate was estimated according to the Modification of Diet in Renal Disease (MDRD) equation.
External validation cohort
Thirty-eight patients were enrolled in the external validation cohort (from Gustave Roussy, Villejuif, and IUCT-Oncopole Claudius-Regaud Toulouse, France). Twenty-six patients (68%) experienced early DLT with a median onset of 1.8 months (95%CI 1.3–2.3). In this external validation cohort, the frequency of early DLT in patients with Cav,ss ≥ 750 ng/mL was similar to that reported in the training cohort (69%), while the starting dose, Cmin,ss, Cav,ss and AUCss were statistically higher (Supplemental Table 3). The frequency of early DLT in patients with Cav,ss < 750 ng/mL was numerically higher in the validation cohort compared to the training cohort (42% vs 35%, respectively). No statistical difference in dose-normalised Cmin,ss was observed between the two cohorts (p = 0.19). The univariate logistic regression analysis showed a tendency for a higher risk of early DLT in patients with increased Cmin,ss (OR 1.23, 95%CI [0.96–1.57]; p = 0.10) or Cav,ss ≥ 750 ng/mL (OR 3.15, 95%CI [0.76–13.01]; p = 0.11). Finally, no patient exhibited Cmin,ss below 336 ng/mL, which did not allow confirming this threshold value in regard to PFS.
Discussion
Cabozantinib monotherapy has shown consistent clinical benefits for treatment-naïve, intermediate- or poor-risk patients with mRCC or those who have received prior VEGF-targeted therapy [13]. However, dose reduction was required in 40–62% of patients enrolled in phase 3 trials [2, 13] and in 57% of patients in the CABOREAL study [14]. Therefore, the optimal starting dose of cabozantinib is still questioned, especially in fragile patients treated in routine care. As previously suggested for other VEGFR-TKIs [15, 16], plasma drug monitoring could be helpful in addressing the interpatient variability in PK. The main findings of the present ambispective real-world study are the association between early DLT onset and increased plasma exposure, as well as the relationship between shorter PFS and a low plasma exposure (Cmin,ss < 336 ng/mL). These results, taken together, suggest that mRCC patients treated with cabozantinib might benefit from reduced toxicity and increased efficacy through pharmacokinetically-guided dosing.
Compared to the patients enrolled in the cabozantinib registration clinical trial (namely METEOR) [13], the patients in our training cohort were in poorer general condition, as evidenced by a KPS < 80% in 25% of them. Moreover, 54% of our patients started cabozantinib at a lower dose than that recommended (60 mg/day), reflecting the greater frailty of this real-life population. Nevertheless, the efficacy of cabozantinib in our training cohort is comparable to that reported in the METEOR trial in terms of objective response rate, PFS and OS [13] (Supplemental Table 4). In the same way, the dose reduction rate was comparable between the two studies. However, we did not report any discontinuation of treatment due to toxicity, compared with 12% in the pivotal study, which suggests that starting with a lower dose in real-life situations enables treatment to be maintained until progression. In the retrospective study from the International Metastatic Renal Cell Carcinoma Data Consortium (IMDC) [17], the patients’ characteristics were comparable to those in our study. However, Gan et al. reported a slightly shorter survival (PFS and OS) but a higher related toxicity discontinuation rate (23%) [17], which could explain the lower survival observed in their study (Supplemental Table 4). Interestingly, they observed a median time-to-dose reduction of 1.2 months, which highlights that the early onset of severe toxicities may have an impact on the further dose intensity of cabozantinib and, therefore, on its efficacy. For this reason, our study focused on the relationship between plasma drug exposure and the onset of early toxicities within the first 3 months. The CABOREAL study, the largest European study on the use of cabozantinib in real-life settings, reported data from a frail and heavily pretreated population (Supplemental Table 4) [14]. In agreement with our results, the starting dose of cabozantinib did not have any impact on PFS. However, a longer median OS was observed in patients initiating at 60 mg (15.4 versus 11.8 months, respectively, p = 0.0314), and the starting dose was identified as an independent risk factor of death. Nevertheless, this result should be interpreted with caution since the authors did not include the ECOG-PS variable in the multivariate analysis. Indeed, it seems obvious that the impact of the starting dose on survival could be related to the general condition of the patients, especially as the authors reported a similar time of exposure to cabozantinib whatever the starting dose (7.9 vs 7.1 months for doses of 60 vs 40 or 20 mg, respectively). In the present study, the starting dose was not independently associated with OS because its dose intensity was based on KPS, a well-known prognostic factor of death. Finally, patients in the CABOREAL study experienced more frequent permanent discontinuation due to toxicity (21%) compared to those included in our study (0%).
The present study highlights that an increased plasma exposure (AUCss, Cmin,ss) is independently associated with a higher risk of DLT onset within the first 3 months of treatment. In accordance with previous studies [6, 8, 9, 18, 19], this study reinforces the need to monitor plasma concentration to address interpatient PK variability and prevent the occurrence of severe toxicity. In an exploratory analysis, Krens et al. reported a higher incidence of dose reduction in mRCC patients with Cmin,ss ≥ 750 ng/mL compared to those with Cmin,ss < 750 ng/mL (78.6 vs 38.7%, p = 0.003) [8]. In our training cohort, Cav,ss ≥ 750 ng/mL was identified as an independent risk factor of early DLT onset. However, this threshold value was not confirmed in the external validation cohort, but univariate logistic regression analysis showed a tendency for a higher risk of early DLT in patients with Cav,ss > 750 ng/mL (p = 0.11). The limited size of the training and external validation cohorts could partly explain these results, as could other inter-centre differences, including bioanalytical assays and clinical management of toxicity. Even if the upper limit of the therapeutic range (Cav,ss < 750 ng/mL) is not clearly validated, high cabozantinib plasma exposure can lead to severe toxicity and potential adherence issues. In this context, treatment reduction could improve outcomes by providing longer exposure to cabozantinib. In this line, different real-world studies have reported that a reduction in cabozantinib dosing related to toxicity improves time to treatment failure and OS [8, 17, 20]. In accordance with Krens et al. [8], patients with Cav,s ≥ 750 ng/mL demonstrated similar PFS compared to patients with Cav,ss ≤ 750 ng/mL. Taken together, these data suggest that the dose of cabozantinib could be adjusted in 20 mg decrements in patients with quality-of-life affecting toxicity and Cav,ss ≥ 750 ng/mL to improve tolerability without adversely affecting efficacy. This intervention should be associated with plasma drug monitoring of cabozantinib to ensure that Cmin,ss remains superior to 336 ng/mL after dosing adjustment.
Regarding the exposure-efficacy relationship, a target Cav,ss ≥ 750 ng/mL for efficacy has been proposed from the phase III METEOR trial’s PK/PD data [6, 7]. However, this threshold value is not confirmed in our study or in previous studies [8, 9]. Interestingly, we demonstrated that Cmin,ss ≥ 336 ng/mL seems to be an independent protective factor of disease progression contrary to the starting daily dose in the patients treated in the second or later line. This efficacy threshold was not confirmed in the sensitivity analysis, probably because of a lack of statistical power. This result could not be explored in the external validation cohort since no patients exhibited Cmin,ss < 336 ng/mL. In the phase III METEOR trial [6], an average Cav,ss of 750 ng/mL (40 mg/day) and 375 ng/ml (20 mg/day) were associated with a 1.1- and 1.4-fold increased risk of disease progression compared to Cav,ss of 1125 ng/mL (60 mg/day), which supports our threshold value of 336 ng/mL. Cerbone et al. proposed targeting Cmin,ss > 537 ng/mL [9]; however, this value is not comparable with ours for two main reasons: the use of BOR as the primary efficacy endpoint and a later median time from treatment start to blood draw (38.3 weeks vs 16 days, respectively). Overall, these data highlight the need to identify underexposed patients treated in monotherapy to prevent disease progression. In the present study, 87% of underexposed patients (Cmin,ss < 336 ng/mL) were treated at 20 mg/day, suggesting that this starting dose is too low for most patients. Therefore, early plasma drug monitoring should be recommended in fragile patients starting cabozantinib at 20 mg/day. Furthermore, cabozantinib plasma exposure is known to decrease over time [19], as previously reported for other VEGFR-TKIs [21, 22]. This phenomenon might decrease the magnitude of adverse events over time but conversely affect the drug efficacy. In patients underexposed at the time of progression, a dose escalation could restore adequate plasma exposure. Different studies have reported the clinical benefit of dose escalation in some patients with progressive disease [23], but no PK data were available before dose escalation. We believe that plasma drug monitoring could be helpful in clinical decision-making to identify underexposed patients for whom dose escalation could be beneficial.
The co-administration of PPIs can significantly decrease the bioavailability of VEGFR-TKIs with pH-dependent solubility, potentially impacting their efficacy [20, 24, 25]. Regarding cabozantinib, the concomitant administration of PPI is known to not significantly impact its bioavailability and steady-state plasma exposure [26, 27]. In agreement with these findings, we did not identify the concomitant intake of PPI as a determinant of the interindividual variability in AUCss. On the other hand, the impact of the concomitant intake of PPI on cabozantinib efficacy in mRCC patients remains controversial [20, 27, 28]. In the present study, we found that the concomitant use of PPI does not significantly impact the efficacy of cabozantinib, which aligns with the PK data [26, 27]. However, Buti et al. recently showed that the use of PPI was independently associated with both shorter PFS (HR, 1.64; p < 0.004) and poorer OS (HR, 1.92; p = 0.015) in mRCC patients treated with cabozantinib [28]. Different methodological characteristics can be pointed out to explain these differing results: the sample size (122 vs 71 patients, respectively) and the proportion of PPI users (51.6% vs 30.0%, respectively). Besides, the overlapping ratio (defined as the percentage of the TKI administration period during which there was an overlapping administration of PPI therapy) is known to influence the deleterious effect of PPI therapy on survival in advanced lung cancer patients treated with epidermal growth factor receptor-TKIs with pH-dependent solubility (erlotinib, gefitinib) [29]. In this context, the overlapping ratio with cabozantinib should be investigated in future clinical studies addressing the drug-drug interaction between PPI and cabozantinib. Overall, the current data do not allow for drawing a firm conclusion regarding a clinically meaningful drug-drug interaction between PPI and cabozantinib.
The present study has several limitations, such as its retrospective nature in terms of the collection of clinical data and the limited number of patients in each cohort (training and external validation). Another limitation of our study is the absence of a prospective safety evaluation to precisely investigate the exposure-toxicity relationship. However, we believe that the choice of DLT onset as the primary safety endpoint reflects clinical use in a real-world setting. Given that cabozantinb plasma exposure decreases over time [19], assaying drug exposure at the time of disease progression would also have been worthwhile to refine the exposure-efficacy relationship. All these limitations highlight the methodological issues to clearly define threshold values for toxicity and efficacy in a real-world cohort. Nevertheless, early blood sampling within the first 3 months represents a methodological strength compared to previous PK/PD studies for which no predefined sampling moments were set [8, 9].
The combination of nivolumab (240 mg every two weeks) and cabozantinib (40 mg/day) has been recently approved as first-line treatment for patients with advanced RCC. PK/PD analysis of the CheckMate 9ER trial showed no association between PFS and the predicted average cabozantinib concentration over the entire treatment period [19]. Regarding toxicity, the rate of grade 3–4 adverse events in the nivolumab plus cabozantinib arm was 60.6% [3]. The safety-exposure relationship was flat for nivolumab, contrary to cabozantinib [19]. The management of adverse events in mRCC patients treated with nivolumab plus cabozantinib is a challenge for physicians because these two drugs exhibit multiple overlapping adverse events [30]. Furthermore, dose reductions are not permitted for nivolumab, and 6.6% of patients discontinued nivolumab in the CheckMate 9ER trial [3]. In patients with overlapping adverse events and for whom nivolumab discontinuation is discussed, plasma drug monitoring could help determine the potential role of cabozantinib exposure in the occurrence of adverse events. In this case, high exposure to cabozantinib could be an argument for its imputability.
Conclusion
The present multicenter ambispective study confirms that starting with a dose of cabozantinib lower than 60 mg/day does not represent a loss of clinical benefit in a real-life population. Furthermore, it shows a significant PK/PD relationship for both toxicity and efficacy. In terms of insights for clinical practice, our results highlight the need to identify either underexposed patients who are at higher risk of rapid disease progression or overexposed patients to limit the occurrence of DLT. Early plasma drug monitoring could be helpful to identify these patients and refine their dosing to achieve a tolerated plasma exposure that also yields optimal clinical efficacy. Therefore, determining a therapeutic range of cabozantinib is of strong clinical interest, and further studies with larger cohorts should be conducted to address this issue. Overall, starting with a lower dose associated with early pharmacokinetically-guided dose adjustment could optimise the clinical management of patients in a real-world setting.
Supplementary information
Author contributions
BB, AJ, JM, and YV designed the study. All authors participated in the collection and assembly of data. AJ performed the statistical analysis. BB, AJ, and YV drafted the manuscript. All authors revised the work, approved its final version, and agreed to be accountable for all its aspects.
Data availability
The datasets used and analysed during the current study are available from the corresponding author upon reasonable request.
Competing interests
BB has received consulting and speaking fees (Bristol Myers Squibb, Clovis Oncology, Janssen, Pierre Fabre, Pfizer and Promise). SO has received honoraria (Sanofi, Pfizer, Bristol Myers Squibb, Eisai, Merck, Novartis, Ipsen, Astellas, Janssen and Bayer), travel and accommodation expenses (Sanofi, Pfizer, Bristol Myers Squibb, Eisai, Merck, Novartis, Ipsen, Astellas, Janssen and Bayer) and research grants (Sanofi, Astra Zeneca, Pfizer, Novartis, Janssen, Bayer and Roche). YV has received honoraria (Bristol Myers Squibb, Merck, Ipsen, Pfizer and Eisai) and research grants (Bristol Myers Squibb and Ipsen). CJ has received honoraria (Ipsen, Astellas, Bayer and Janssen). RF has received honoraria (Bayer, Astellas, Janssen, Ipsen, Merck, Bristol Myers Squibb, Pfizer and Merck). AP has received speaking fees (Bristol Myers Squibb and Pierre Fabre). LM has received honoraria (Sanofi, Astellas, Janssen, MSD, Bristol Myers Squibb, Ipsen, Astra Zeneca and Merck, Novartis) and travel and accommodation expenses (Sanofi, Astellas, Janssen, Bristol Myers, Ipsen, Astra Zeneca, Pfizer, MSD). LA has received honoraria (Novartis, Astellas, Janssen, MSD, Bristol Myers Squibb, Ipsen, Eisai, Pfizer and Merck, Roche) and travel and accommodation expenses (Bristol Myers Squibb, Ipsen, Pfizer, MSD). OH has received honoraria (Sanofi, Bayer, MSD, Bristol Myers Squibb, Ipsen, Pfizer, Eisai, Janssen, Astra Zeneca and Merck). XD has received consulting honoraria (Inflectis Biosciences, MedDay Pharmaceuticals, MAPREG and Merck). JM has received consulting honoraria (Daiichi Sankyo, Gilead, Lilly Eli, MSD and Pfizer) and travel and accommodation expenses (Lilly Eli, Gilead and Seattle Genetics).AXV, AJ, FP, DC, FLL, FT, MT, TP, CT, JC, EC, MV and FG declare no conflict of interest.
Ethics approval and consent to participate
The study was approved by the local ethics committee in oncology (CLEP number: AAA-2022–08055). Informed consent was obtained from all patients prior to inclusion.
Footnotes
The original online version of this article was revised: In this article the author’s name Alexandre Xu-Vuillard was incorrectly written as Alexandre Xu-Vuilard.
Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Change history
12/29/2025
The original online version of this article was revised: In this article the author’s name Alexandre Xu-Vuillard was incorrectly written as Alexandre Xu-Vuilard.
Change history
1/20/2026
A Correction to this paper has been published: 10.1038/s41416-025-03335-4
Supplementary information
The online version contains supplementary material available at 10.1038/s41416-024-02585-y.
References
- 1.Hsieh JJ, Purdue MP, Signoretti S, Swanton C, Albiges L, Schmidinger M, et al. Renal cell carcinoma. Nat Rev Dis Prim. 2017;3:17009. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Choueiri TK, Escudier B, Powles T, Mainwaring PN, Rini BI, Donskov F, et al. Cabozantinib versus everolimus in advanced renal-cell carcinoma. N. Engl J Med. 2015;373:1814–23. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Choueiri TK, Powles T, Burotto M, Escudier B, Bourlon MT, Zurawski B, et al. Nivolumab plus cabozantinib versus sunitinib for advanced renal-cell carcinoma. N. Engl J Med. 2021;384:829–41. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Buil-Bruna N, López-Picazo J-M, Martín-Algarra S, Trocóniz IF. Bringing model-based prediction to oncology clinical practice: a review of pharmacometrics principles and applications. Oncologist. 2016;21:220–32. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Schmidinger M, Danesi R, Jones R, McDermott R, Pyle L, Rini B, et al. Individualized dosing with axitinib: rationale and practical guidance. Future Oncol Lond Engl. 2018;14:861–75. [DOI] [PubMed] [Google Scholar]
- 6.Lacy S, Nielsen J, Yang B, Miles D, Nguyen L, Hutmacher M. Population exposure-response analysis of cabozantinib efficacy and safety endpoints in patients with renal cell carcinoma. Cancer Chemother Pharm. 2018;81:1061–70. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Castellano D, Pablo Maroto J, Benzaghou F, Taguieva N, Nguyen L, Clary DO, et al. Exposure-response modeling of cabozantinib in patients with renal cell carcinoma: Implications for patient care. Cancer Treat Rev. 2020;89:102062. [DOI] [PubMed] [Google Scholar]
- 8.Krens SD, van Erp NP, Groenland SL, Moes DJAR, Mulder SF, Desar IME, et al. Exposure-response analyses of cabozantinib in patients with metastatic renal cell cancer. BMC Cancer. 2022;22:228. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Cerbone L, Combarel D, Geraud A, Auclin E, Foulon S, Alves Costa Silva C, et al. Association of cabozantinib pharmacokinetics, progression and toxicity in metastatic renal cell carcinoma patients: results from a pharmacokinetics/pharmacodynamics study. ESMO Open. 2021;6:100312. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Lacy S, Yang B, Nielsen J, Miles D, Nguyen L, Hutmacher M. A population pharmacokinetic model of cabozantinib in healthy volunteers and patients with various cancer types. Cancer Chemother Pharm. 2018;81:1071–82. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Eisenhauer EA, Therasse P, Bogaerts J, Schwartz LH, Sargent D, Ford R, et al. New response evaluation criteria in solid tumours: revised RECIST guideline (version 1.1). Eur J Cancer Oxf Engl 1990. 2009;45:228–47. [DOI] [PubMed] [Google Scholar]
- 12.Lacy SA, Miles DR, Nguyen LT. Clinical pharmacokinetics and pharmacodynamics of cabozantinib. Clin Pharmacokinet. 2017;56:477–91. [DOI] [PubMed] [Google Scholar]
- 13.Choueiri TK, Escudier B, Powles T, Tannir NM, Mainwaring PN, Rini BI, et al. Cabozantinib versus everolimus in advanced renal cell carcinoma (METEOR): final results from a randomised, open-label, phase 3 trial. Lancet Oncol. 2016;17:917–27. [DOI] [PubMed] [Google Scholar]
- 14.Albiges L, Fléchon A, Chevreau C, Topart D, Gravis G, Oudard S, et al. Real-world evidence of cabozantinib in patients with metastatic renal cell carcinoma: results from the CABOREAL Early Access Program. Eur J Cancer Oxf Engl 1990. 2021;142:102–11. [DOI] [PubMed] [Google Scholar]
- 15.Mueller-Schoell A, Groenland SL, Scherf-Clavel O, van Dyk M, Huisinga W, Michelet R, et al. Therapeutic drug monitoring of oral targeted antineoplastic drugs. Eur J Clin Pharm. 2021;77:441–64. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Puisset F, Mseddi M, Mourey L, Pouessel D, Blanchet B, Chatelut E, et al. Therapeutic drug monitoring of tyrosine kinase inhibitors in the treatment of advanced renal cancer. Cancers. 2023;15:313. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Gan CL, Dudani S, Wells JC, Donskov F, Pal SK, Dizman N, et al. Cabozantinib real-world effectiveness in the first-through fourth-line settings for the treatment of metastatic renal cell carcinoma: results from the international metastatic renal cell carcinoma database consortium. Cancer Med. 2021;10:1212–21. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Krens SD, van Boxtel W, Uijen MJM, Jansman FGA, Desar IME, Mulder SF, et al. Exposure-toxicity relationship of cabozantinib in patients with renal cell cancer and salivary gland cancer. Int J Cancer. 2022;150:308–16. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Tran BD, Li J, Ly N, Faggioni R, Roskos L. Cabozantinib exposure-response analysis for the phase 3 CheckMate 9ER trial of nivolumab plus cabozantinib versus sunitinib in first-line advanced renal cell carcinoma. Cancer Chemother Pharm. 2023;91:179–89. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Corianò M, Giannarelli D, Scartabellati G, De Giorgi U, Brighi N, Fornarini G, et al. Tailoring treatment with cabozantinib or pazopanib in patients with metastatic renal cell carcinoma: does it affect outcome? Expert Rev Anticancer Ther. 2023;23:545–54. [DOI] [PubMed] [Google Scholar]
- 21.Arrondeau J, Mir O, Boudou-Rouquette P, Coriat R, Ropert S, Dumas G, et al. Sorafenib exposure decreases over time in patients with hepatocellular carcinoma. Invest N. Drugs. 2012;30:2046–9. [DOI] [PubMed] [Google Scholar]
- 22.Ozbey AC, Combarel D, Poinsignon V, Lovera C, Saada E, Mir O, et al. Population pharmacokinetic analysis of pazopanib in patients and determination of target AUC. Pharm Basel Switz. 2021;14:927. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Sharma A, Elias R, Christie A, Williams NS, Pedrosa I, Bjarnason GA, et al. Extended disease control with unconventional cabozantinib dose increase in metastatic renal cell carcinoma. Kidney Cancer Clifton Va. 2022;6:69–79. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Mir O, Touati N, Lia M, Litière S, Le Cesne A, Sleijfer S, et al. Impact of concomitant administration of gastric acid-suppressive agents and pazopanib on outcomes in soft-tissue sarcoma patients treated within the EORTC 62043/62072 trials. Clin Cancer Res J Am Assoc Cancer Res. 2019;25:1479–85. [DOI] [PubMed] [Google Scholar]
- 25.Ha VH, Ngo M, Chu MP, Ghosh S, Sawyer MB, Chambers CR. Does gastric acid suppression affect sunitinib efficacy in patients with advanced or metastatic renal cell cancer? J Oncol Pharm Pr Publ Int Soc Oncol Pharm Pr. 2015;21:194–200. [DOI] [PubMed] [Google Scholar]
- 26.Nguyen L, Holland J, Mamelok R, Laberge M-K, Grenier J, Swearingen D, et al. Evaluation of the effect of food and gastric pH on the single-dose pharmacokinetics of cabozantinib in healthy adult subjects. J Clin Pharm. 2015;55:1293–302. [DOI] [PubMed] [Google Scholar]
- 27.Rassy E, Cerbone L, Auclin E, Benchimoll-Zouari A, Flippot R, Alves Costa Silva C, et al. The effect of concomitant proton pump inhibitor and cabozantinib on the outcomes of patients with metastatic renal cell carcinoma. Oncologist. 2021;26:389–96. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Buti S, Tommasi C, Scartabellati G, De Giorgi U, Brighi N, Rebuzzi SE, et al. The impact of proton-pump inhibitors administered with tyrosine kinase inhibitors in patients with metastatic renal cell carcinoma. Anticancer Drugs. 2023;34:178–86. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.Lee C-H, Shen M-C, Tsai M-J, Chang J-S, Huang Y-B, Yang Y-H, et al. Proton pump inhibitors reduce the survival of advanced lung cancer patients with therapy of gefitinib or erlotinib. Sci Rep. 2022;12:7002. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30.McGregor B, Mortazavi A, Cordes L, Salabao C, Vandlik S, Apolo AB. Management of adverse events associated with cabozantinib plus nivolumab in renal cell carcinoma: a review. Cancer Treat Rev. 2022;103:102333. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Data Availability Statement
The datasets used and analysed during the current study are available from the corresponding author upon reasonable request.