High-Dose Intermittent Treatment with the Multikinase Inhibitor Sunitinib Leads to High Intra-Tumor Drug Exposure in Patients with Advanced Solid Tumors

Simple Summary Multitargeted tyrosine kinase inhibitors (TKIs) provide clinical benefit in patients with cancer when daily dosed to continuously inhibit their designated targets. Alternative dosing strategies pursuing higher tumor drug concentrations may improve their benefit by inhibiting relevant off-target kinase activity. We studied the tumor-drug exposure in a previously established safe and feasible high-dose intermittent treatment strategy of the multikinase inhibitor sunitinib in relation to its treatment benefit. Abstract Patients with advanced cancer refractory to standard treatment were treated with sunitinib at a dose of 300 mg once every week (Q1W) or 700 mg once every two weeks (Q2W). Tumor, skin and plasma concentrations were measured and immunohistochemical staining for tumor cell proliferation (TCP), microvessel density (MVD) and T-cell infiltration was performed on tumor biopsies before and after 17 days of treatment. Oral administration of 300 mg sunitinib Q1W or 700 mg Q2W resulted in 19-fold (range 5–35×) and 37-fold higher (range 10–88×) tumor drug concentrations compared to parallel maximum plasma drug concentrations, respectively. Patients with higher tumor sunitinib concentrations had favorable progression-free and overall survival than those with lower concentrations (p = 0.046 and 0.024, respectively). In addition, immunohistochemistry of tumor biopsies revealed an induction of T-cell infiltration upon treatment. These findings provide pharmacological and biological insights in the clinical benefit from high-dose intermittent sunitinib treatment. It emphasizes the potential benefit from reaching higher tumor drug concentrations and the value of measuring TKI tumor- over plasma-concentrations. The finding that reaching higher tumor drug concentrations provides most clinical benefit in patients with treatment refractory malignancies indicates that the inhibitory potency of sunitinib may be enforced by a high-dose intermittent treatment schedule. These results provide proof of concept for testing other clinically available multitargeted tyrosine kinase inhibitors in a high-dose intermittent treatment schedule.

Despite significant clinical benefit in advanced renal cell cancer, imatinib-resistant orintolerant gastrointestinal stromal tumors and well-differentiated neuroendocrine tumors of the pancreas, treatment benefit is of limited duration due to acquired resistance. Moreover, for most common tumor types, such as colorectal cancer, prostate cancer, and breast cancer, sunitinib treatment is ineffective in the majority of patients [2].
Sunitinib is extensively distributed into various tissues, whereas only a small fraction of the drug remains in the circulation, reflected by a relatively large apparent volume of distribution (2230 L) [5]. It is metabolized by cytochrome P450 3A4 to the equipotent active metabolite N-desethyl sunitinib (SU12662) [4]. Combination of sunitinib + N-desethyl sunitinib represents the total active drug in plasma (SUM concentration), because N-desethyl sunitinib has a similar inhibitory profile and plasma protein binding to that of sunitinib in vitro [6]. Sunitinib and its metabolite are small hydrophobic molecules that can efficiently cross cellular membranes and enter tumor and tumor-supporting cells.
The hypothesis behind its daily dosing strategy is that a steady sunitinib plasma concentration will enable continuous inhibition of its designated targets [4]. Proof for this concept is lacking and knowledge regarding its exact mechanism of action at the tumor and tumor-supporting cells is limited even as the drug resistance mechanisms to sunitinib. Our recent studies indicate that drug concentrations reached in the tumor are underestimated by sunitinib plasma concentrations. Gotink et al. reported 10-times higher tumor concentrations compared to plasma in mice and 30-times higher tumor sunitinib concentrations (9.5-10.9 micromolar (µM) compared to plasma concentrations at the standard daily schedule in three patients [7]. Labots et al. confirmed these findings in patients, reporting median sunitinib tumor concentrations of 9.0 µM compared to median plasma concentrations of 0.1 µM (3.586 versus 40 ug/L) [8]. Finally, Rovithi et al. measured high sunitinib tumor concentrations for both the standard daily schedule (13.5 µM (range 7.0-37.4 µM)) and a high-dose once every week (Q1W) schedule (36 µM (range: 7.8-130 µM)) in tumors growing on the chorioallantoic membrane of the chicken embryo (CAM-assay) [9]. Based on these findings, we hypothesized that the antitumor activity of sunitinib can be improved by increasing tumor-drug exposure to significantly inhibit additional relevant (off-target) kinases with a distinct-and potentially more favorable-composite efficacy [7,8]. We established a safe and feasible high-dose intermittent schedule of 300 mg sunitinib once every week (Q1W) and 700 mg sunitinib once every two weeks (Q2W) in patients with advanced cancer [10]. Here, we measured tumor, skin and plasma concentrations of this high-dose intermittent treatment strategy with sunitinib in patients with various advanced cancer types. We focused our analyses on potential relations of these concentrations with treatment benefit and biological activity.

Materials & Methods
Patients and study design: From 2013-2018, a dose escalation phase I, single-institution clinical trial, was conducted at Amsterdam University Medical Center, location VUmc, Amsterdam, the Netherlands (ClinicalTrials.gov identifier 02058901) to establish an intermittent dosing schedule of sunitinib (Q1W or Q2W). All patients provided written informed consent before entering the study. The study was conducted in accordance with the Declaration of Helsinki and with the principles of the International Conference on Harmonization Guidelines for Good Clinical Practice. The study protocol was approved by the local institutional board and an independent Ethics Committee.
Sunitinib was administered orally according to one of two schedules: Q1W or Q2W. A standard Phase 1 "3 + 3 design" was used with a starting dose of 200 mg sunitinib administered orally once every week and escalating in steps of 100 mg. Patients continued treatment until progression, intolerance, or withdrawal of consent. The primary outcomes of the phase I clinical trial were described previously [10].
Food: An extra cohort was opened to evaluate the effect of food on the interpatient variation in bioavailability and to study whether concomitant intake with food could further increase the maximum plasma concentration (C max ) . (See Supplementary Data S1 for details regarding the food cohort) Study pharmacokinetic assessment in plasma: Ethylenediaminetetraacetic acid (EDTA) blood was collected pretreatment and subsequently at multiple time points in all patients (0, 2, 4, 6, 8, 10, and 24 h post-dose on day 1 for both time schedules and thereafter at days 3, 8, 10, 15, 17, and 22 for the once weekly schedule and at days 3, 15, 17, and 29 for the once every 2 weeks schedule). All samples were centrifuged within 1 h after collection at 2000 g for 10 min (at 4 degrees Celsius ( • C)), plasma was separated and stored at −80 • C until the day of analysis.
Study pharmacokinetic assessment in tumor and skin tissue: Patients who gave additional informed consent for tissue collection, underwent a tumor needle biopsy before and after 17 days of study treatment taken by an interventional radiologist. An optional skin biopsy (from healthy skin tissue at the abdomen) was taken after 17 days of treatment. All tissue samples were immediately snap frozen (within 1 min) in liquid nitrogen and cryopreserved at −80 • C until the day of analysis.
Tissue biopsies were weighed accurately before analysis. Tissue homogenate was prepared by using a mechanical homogenizer. The tumor tissue was homogenized with the precellys lysing kit (bertin instruments ® ) in the Roche MagNA lyzer ® . Proteins in the homogenate were precipitated with methanol primed with internal standards (the stable isotopes 2H10-sunitinib and 2H5-desethylsunitinib). Subsequently, sunitinib and N-desethyl sunitinib were quantified against a calibration curve with known concentrations. Tissue homogenate samples were directly analyzed.
Drug concentration measurements: Sunitinib and N-desethylsunitinib were measured in all samples; using a validated liquid chromatography-tandem mass spectrometry (LC-MS-MS) as previously described (see Supplementary Data S2-methods).
Immunohistochemistry: Immunohistochemical staining was performed on 3-5 µm thick sections of the (freeze) tumor biopsies for CD3, CD31 and Ki67 according to standard methods (see Supplementary Data S2). Morphological characteristics were assessed in hematoxylin-eosin (H&E)-stained tissue sections. Evaluation of staining was performed by four investigators using light microscopy of serial tissue sections. Ki-67 was scored based on the percentage of positively stained malignant nuclei, using the following ranges: 0% to 20%, >20% to 40%, >40% to 60%, >60% to 80%, and >80% to 100%. Semi-quantification of CD31 staining was scored manually by a pathologist. The number of cells positive for CD3, CD31 was expressed as an estimated proportion of the total number of cells.
Statistics: Descriptive statistics were used to summarize patient characteristics and sunitinib and N-desethyl sunitinib concentrations. Two-sided paired Student's t-test on logtransformed data was used to compare concentrations between sunitinib and N-desethyl sunitinib concentrations in plasma, tissue, and skin tissue and pre-and on treatment immunohistochemistry analysis. An independent t-test on log-transformed data was used to compare sunitinib concentrations between the two MTD-dose levels (300 mg Q1W and 700 mg Q2W). Sunitinib and N-desethyl sunitinib concentrations in plasma, tumor and skin and their correlations were calculated using the Spearman's rank correlation coefficient (ρ). Efficacy was assessed in patients who completed at least two weeks of treatment. Kaplan-Meier curves and log-rank tests were used to evaluate differences in PFS and OS between two groups dichotomized by geometric mean sunitinib and Ndesethyl sunitinib concentration. Sunitinib and N-desethyl sunitinib concentrations in plasma, tumor and skin and their correlation with PFS and OS were calculated using the Spearman's rank correlation coefficient (ρ). Statistical significance was set at p < 0.05. All statistical analyses were performed in SPSS, version 25.
Pharmacokinetic analysis: The concentration-time data for sunitinib and N-desethyl sunitinib were analyzed by means of nonlinear mixed-effects modeling (NONMEM, V7.4.3). Since there is no evidence for the exact relationship between plasma and tumor concentrations, we chose to use the concentration-time data calculated by NONMEM for C max , the average concentration over time (C average ) and the trough concentrations just before the next dose was administered (C trough ) based on rich PK-data of each individual patient. As a starting point for the analysis, the previously developed population pharmacokinetic model for sunitinib and its active metabolite by Yu et al. was used [6]. The model was extended with a sequential zero-and first-order absorption model, to describe the dissolution and subsequent absorption of the high sunitinib dose in the gastrointestinal tract. Inter-occasion variability in bioavailability and absorption duration were allowed in the model. The control stream of the final model and the goodness-of-fit plots are supplied in the Supplementary Material (Supplementary Data S3). The developed model was used to derive the individual predictions for the C max , C average , and C trough . SUM concentrations were calculated. correlation coefficient (ρ). Efficacy was assessed in patients who completed at least two weeks of treatment. Kaplan-Meier curves and log-rank tests were used to evaluate differences in PFS and OS between two groups dichotomized by geometric mean sunitinib and N-desethyl sunitinib concentration. Sunitinib and N-desethyl sunitinib concentrations in plasma, tumor and skin and their correlation with PFS and OS were calculated using the Spearman's rank correlation coefficient (ρ). Statistical significance was set at p < 0.05. All statistical analyses were performed in SPSS, version 25.

Baseline Patient Characteristics and Covariates
Pharmacokinetic analysis: The concentration-time data for sunitinib and N-desethyl sunitinib were analyzed by means of nonlinear mixed-effects modeling (NONMEM, V7.4.3). Since there is no evidence for the exact relationship between plasma and tumor concentrations, we chose to use the concentration-time data calculated by NONMEM for Cmax, the average concentration over time (Caverage) and the trough concentrations just before the next dose was administered (Ctrough) based on rich PK-data of each individual patient. As a starting point for the analysis, the previously developed population pharmacokinetic model for sunitinib and its active metabolite by Yu et al. was used [6]. The model was extended with a sequential zero-and first-order absorption model, to describe the dissolution and subsequent absorption of the high sunitinib dose in the gastrointestinal tract. Inter-occasion variability in bioavailability and absorption duration were allowed in the model. The control stream of the final model and the goodness-of-fit plots are supplied in the Supplementary Material (Supplementary Data S3). The developed model was used to derive the individual predictions for the Cmax, Caverage, and Ctrough. SUM concentrations were calculated.   Tumor drug concentration: To determine tumor drug concentrations with this treatment schedule, non-mandatory tumor biopsies from 22 patients at day 17 after start of treatment were obtained (after 3 cycles Q1W or 2 cycles Q2W). Twenty of 22 tumor biopsies were taken from patients treated at the MTD of 300 mg Q1W or 700 mg Q2W. The geometric mean of the sunitinib + N-desethyl sunitinib (SUM) tumor concentration of patients treated with 300 mg Q1W (n = 12) was 6.656 µg/L (coefficient of variation (CV) 111%) versus 17.376 µg/L (CV 109%) in those treated with sunitinib 700 mg Q2W (n = 8, p-Value (p) = 0.07) ( Table 1 and Supplementary Data S4). The other two patients received a dose of 400 mg Q1W and 800 mg Q2W resulting in tumor drug concentrations of 1.369 and 515 µg/L respectively. Tumor concentrations obtained from patients treated with 300 mg sunitinib Q1W were approximately 19-times higher than achieved maximum plasma concentration (C max ) (geometric mean of 6656 µg/L, range 877-48.647 µg/L versus 349 µg/L, range 183-1396 µg/L, respectively; p = 0.001) and approximately 3 times higher than skin concentrations (a geometric mean of 1957 µg/L, range 319-8772 µg/L; p = 0.001) on day 17. Tumor concentrations obtained from patients treated with 700 mg sunitinib Q2W were approximately 37-times higher than achieved plasma C max (geometric mean of 17.376 µg/L, range 2791-69.781 µg/L versus 473 µg/L, range 283-790 µg/L, respectively; p = 0.001) and approximately 9-times higher than skin concentrations (a geometric mean of 1.857 µg/L, range 379-19.539 µg/L; p = 0.001) on day 17 (Table 1). No correlation was observed between tumor concentrations and plasma-or skin concentrations and food did not affect the oral uptake of sunitinib (Figure 2A and Supplementary Data S1).

Baseline
Tumor drug concentrations in relation to outcome: As previously reported, median progression free survival (PFS) for all patients was 2.5 months (range 0.5-11 months); 2.2 months for 300 mg Q1W (range 0.7-11) and 2.6 months (range 0.5-11) for 700 mg Q2W. Median overall survival (OS) for all patients was 4.9 months (range 0.8-27 months); 3.1 months for 300 mg Q1W (range 1.  Tumor drug concentrations in relation to outcome: As previously reported, median progression free survival (PFS) for all patients was 2.5 months (range 0.5-11 months); 2.2 months for 300 mg Q1 W (range 0.7-11) and 2.6 months (range 0.5-11) for 700 mg Q2 W. Median overall survival (OS) for all patients was 4.9 months (range 0.8-27 months); 3.1 months for 300 mg Q1 W (range 1. Immunohistochemical staining: Immunohistochemical staining for tumor cell proliferation (TCP), microvessel density (MVD) and T-cell infiltration was performed in 10 available paired tumor biopsies before (at baseline) and after 17 days of treatment ( Figure  2A). While no significant differences in pre-versus on treatment biopsies were found for TCP and MVD, a significant increase in T cell infiltration, assessed by CD3 staining, was found (paired t-test, mean increase 3.55 (from 8.6% to 12.2%), 95% CI 1. 15-5. Immunohistochemical staining: Immunohistochemical staining for tumor cell proliferation (TCP), microvessel density (MVD) and T-cell infiltration was performed in 10 available paired tumor biopsies before (at baseline) and after 17 days of treatment (Figure 2A). While no significant differences in pre-versus on treatment biopsies were found for TCP and MVD, a significant increase in T cell infiltration, assessed by CD3 staining, was found (paired t-test, mean increase 3.55 (from 8.6% to 12.2%), 95% CI 1.15-5.95, p-Value 0.009) (Figure 3 and Supplementary Data S6).

Discussion
We herein report remarkably high tumor drug exposure from high-dose intermittent sunitinib treatment in patients with advanced cancer lacking standard treatment options. A more favorable PFS and OS were observed in patients with the highest sunitinib + Ndesethyl sunitinib tumor concentrations. These tumor sunitinib concentrations are 2-5times higher than reported for the tumor concentrations derived from daily treatment [8]. Importantly, these concentrations exceeded the concentration required for inhibition of its 'designated angiogenesis-related targets' indicative for potential inhibition of a wide relevant spectrum of the human kinome (Supplementary Data S7 and S8) [4,5,[11][12][13]. Clinical trials with other TKIs (e.g., erlotinib and sorafenib) at high-dose intermittent treatment schedules also showed higher plasma Cmax concentrations compared to daily dosing, but no data on their tumor concentrations have been reported [14][15][16].
The significantly higher tumor drug concentrations compared to plasma drug concentrations might be due to the chemical characteristics of sunitinib, trafficking easily through membranes in its neutral state but trapped in its protonated form under acidic circumstances (in cellular vesicles) [7]. We measured total sunitinib tumor tissue concentrations, while only the intracellular unbound, available free drug can interact and inhibit multiple kinases [11]. Determination of the exact fraction of intracellular unbound, available free drug from the total drug concentration is not feasible since no assays are available

Discussion
We herein report remarkably high tumor drug exposure from high-dose intermittent sunitinib treatment in patients with advanced cancer lacking standard treatment options. A more favorable PFS and OS were observed in patients with the highest sunitinib + N-desethyl sunitinib tumor concentrations. These tumor sunitinib concentrations are 2-5-times higher than reported for the tumor concentrations derived from daily treatment [8]. Importantly, these concentrations exceeded the concentration required for inhibition of its 'designated angiogenesis-related targets' indicative for potential inhibition of a wide relevant spectrum of the human kinome (Supplementary Data S7 and S8) [4,5,[11][12][13]. Clinical trials with other TKIs (e.g., erlotinib and sorafenib) at high-dose intermittent treatment schedules also showed higher plasma C max concentrations compared to daily dosing, but no data on their tumor concentrations have been reported [14][15][16].
The significantly higher tumor drug concentrations compared to plasma drug concentrations might be due to the chemical characteristics of sunitinib, trafficking easily through membranes in its neutral state but trapped in its protonated form under acidic circumstances (in cellular vesicles) [7]. We measured total sunitinib tumor tissue concentrations, while only the intracellular unbound, available free drug can interact and inhibit multiple kinases [11]. Determination of the exact fraction of intracellular unbound, available free drug from the total drug concentration is not feasible since no assays are available to measure intracellular unbound drug concentrations. When considering this relative shortcoming, it is striking that a positive relation between the total tumor drug concentration and clinical outcome was found. This supports our hypothesis that high dosing contributes to inhibition of additional relevant 'off' targets in tumor and supporting cells of the tumor microenvironment [12,17,18]. Sunitinib has been implicated in impairing T cell activation and proliferation in vitro and in vivo. Guislain et al. found improved tumor-infiltrating lymphocyte expansion in tumor digests of renal cell cancer patients who were pre-treated with sunitinib. This was associated with reduced intratumoral myeloid-derived suppressor cells. Exploratory analyses of 10 paired tumor biopsies in our population showed a significant increase in T-cell infiltration over time. This increase provides a potential lead for combination strategies with immune checkpoint inhibitors and high-dose intermittent sunitinib [18,19].
In conclusion, this study reveals that high-dose intermittent treatment with sunitinib leads to high tumor drug concentrations compared to prior reported tumor drug concentrations of daily dosing. These high tumor drug concentrations are favorably related to an improved PFS and OS in a small and heterogeneous group of patients with advanced cancer. This improved outcome might be due to additional off-target kinase inhibition with a distinct-and potentially more favorable-composite efficacy. The results of this study emphasize the importance of measuring drug concentrations at the target site and underscores the limitations of regular plasma pharmacokinetics in determining optimal doses and schedules. Further evaluation of a potential benefit from high-dose intermittent sunitinib treatment compared to standard dosing is of interest in registered indications such as renal cell cancer and gastrointestinal stroma cell tumors (GIST), and in potential new indications such as the current ongoing trial in colorectal cancer (NCT03909724).

Informed Consent Statement:
All patients provided written informed consent before entering the study.

Data Availability Statement:
The authors declare that all data used in the conduct of the analyses are available within the article and (supplementary) tables and figures. To protect the privacy and confidentiality of patients in this study, clinical data are not publicly available in a repository or in the supplementary material of the article, but they can made available upon reasonable request to the corresponding author. Those requests will be reviewed by a study steering committee to verify whether the request is subject to any intellectual property or confidentiality obligations. All data shared will be de-identified.