Anticancer Activity of the Combination of Cabozantinib and Temozolomide in Uterine Sarcoma

Sarcomas are heterogeneous and aggressive soft-tissue cancers with more than 100 histological subtypes that occur at various anatomical sites (1, 2). Uterine leiomyosarcoma is an invasive subtype that accounts for 24% of all soft-tissue sarcomas and has high metastatic potential (3–5). Previous studies have shown the 5-year survival rate for women with metastatic disease at initial diagnosis between 10% and 15% (6). At present, surgical resection is the main treatment for leiomyosarcoma. For unresectable lesions, systemic chemotherapy or radiotherapy is considered, but their clinical efficacy is poor, so effective treatment agents need to be developed.

Cabozantinib is a multitarget tyrosine kinase inhibitor that inhibits c-MET, VEGFR-2, Ret, Kit, FLT-1, FLT-3, FLT-4, TIE-2, and AXL (7). It has shown clinical activity in targeting several types of tumors and been approved for treatment of advanced renal cell carcinoma, advanced hepatocellular carcinoma, and progressive, metastatic medullary thyroid cancer (8–11). Cabozantinib inhibits the activity of a variety of tyrosine kinases that are expressed in soft-tissue sarcomas as well, including MET, VEGFR, AXL, and TYRO3 (12–16). The relevance of those molecular targets of cabozantinib in sarcomas and the results of previous in vitro and phase II trials indicate that cabozantinib is a promising candidate for treatment of various diseases (17–20). When cabozantinib was added to temozolomide and bevacizumab in patients with heavily pretreated relapsed uterine leiomyosarcoma, the response rate and clinical benefit rate were shown to be 50% and 100%, respectively, whereas those who received temozolomide and bevacizumab only demonstrated 35% of response rate and 78% of clinical benefit rate (20).

Temozolomide is an alkylating prodrug with active metabolites similar to dacarbazine (21). Its primary mechanism of action involves methylation of the O6-position of guanine to form DNA adducts that interfere with replication (22). Temozolomide has demonstrated activity against several soft-tissue sarcomas, including modest activity against leiomyosarcoma, in several phase II trials (23–27). When given to patients with previously treated unresectable or metastatic leiomyosarcoma, temozolomide demonstrated the objective response rate (ORR) of 18%, with disease stabilization occurring in 27% (23). In another phase II trial that included 45 patients with soft-tissue sarcoma, an overall response rate of 15.5% was demonstrated. These responses were seen in 5 of 11 patients who had gynecologic leiomyosarcoma (24).

In an effort to add further evidence to the body of literature suggesting anticancer effects of cabozantinib and temozolomide and their potential synergistic effect when combined, the present study was designed to investigate the effectiveness of the drugs in sarcoma cell lines and in vivo experiments, including patient-derived xenografts (PDX) of uterine sarcoma.

The human sarcoma cell lines SK-LMS-1 (Homo sapiens vulva, leiomyosarcoma), SK-UT-1 (Homo sapiens uterus, mesodermal tumor mixed), MES-SA (Homo sapiens uterus, sarcoma), and SKN (Homo sapiens uterus, leiomyosarcoma) were used. The SK-LMS-1, SK-UT-1, and MES-SA cells were obtained from the ATCC. SKN cells were purchased from the Japanese Collection of Research Bioresources. Cells were maintained in EMEM (30-2003; ATCC), MEM (11095098; Gibco), and Ham's F12 (11765-062; Gibco) medium containing 10% FBS (S001-07; Welgen) with 100 U/mL penicillin and 100 μg/mL streptomycin (Sigma-Aldrich). They were grown at 37°C in a 5% CO₂ incubator. Cabozantinib and temozolomide were obtained from CSN Pharm (CSN-15710 and CSN-12046; Arlington Heights).

The cells used for the PDX model were separated from the tumor tissue of a patient with uterine leiomyosarcoma. After cutting and mincing the tumor tissue, the cells were filtered out of the strainer, centrifuged to collect the cell pellet, and then released with EMEM and placed in a culture dish.

For the cell viability assays, as previously described (28), cells were plated in culture medium in 96-well plates at 3 × 10³ cells/well. The cells were stained with 3-(4, 5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT; Amresco). After 4 hours of incubation, the medium was discarded, 100 μL of acidic isopropanol (0.1 N HCL in absolute isopropanol) was added, and the plate was shaken gently. Absorbance was measured on an enzyme linked immunosorbent assay reader at a wavelength of 540 nm. Each sample was assayed in triplicate, and the experiment was repeated three times.

The relative proportion of apoptotic cells was evaluated at 48 hours using an Annexin V-FITC apoptosis detection kit-1 (BD Pharmingen) according to the manufacturer's protocol. Briefly, cells were washed twice in PBS, and the pellet was resuspended in annexin V binding buffer at a concentration of 10⁶ cells/mL. Annexin V-FITC and propidium iodide (PI) were added (5 μL per 10⁵ cells), and the samples were mixed and cultured for 15 minutes at room temperature in the dark before FACS analysis. Each sample was assayed in triplicate, and the experiment was repeated three times.

For analyzing active Caspase-3 levels, a commercial active Caspase-3 ELISA assay was used (#KHO1091, Invitrogen). Cells were seeded in a 6-well plate (1×10⁴ cells in 3 mL of media per well), and incubated overnight to allow the cells to attach to the plate. After 24 hours of treatment with cabozantinib and temozolomide, the medium was removed by suction. The cells were lysed with lysis buffer. The apoptotic activity was determined for each well according to the manufacturer's protocols.

Cells were lysed in PRO-PRE Protein Extraction Solution (Intron Biotechnology). Protein concentration was determined using a BCA protein assay kit (Thermo Fisher Scientific). Cell or tissue lysates (50 μg of total protein) were separated into 7.5%, 10%, or 15% acrylamide gels by SDS-PAGE (Bio-Rad) and transferred to Immobilon-P transfer membranes (Merck Millipore). The membranes were blocked for 1 hour with 5% BSA in TBS containing 0.1% Tween-20 at room temperature. Protein bands were probed with phospho-c-MET (ab5662; Abcam), c-MET (ab74217; Abcam), STAT3 (ab32500; Abcam), phospho-STAT3, phospho-AKT, total AKT, phospho-ERK (p-ERK), total ERK, phospho-AMPKα, and AMPKα (Cell Signaling Technology) antibodies at 1:1,000 dilution or with β-actin (Santa Cruz Biotechnology) antibody at a 1:3,000 dilution and then tagged with horseradish peroxidase–conjugated anti-rabbit antibody (7074V; Cell Signaling Technology) and anti-mouse antibody (sc-2025; Santa Cruz Biotechnology). Bands were envisioned with enhanced chemiluminescence using an ECL kit (Amersham Biosciences) according to the manufacturer's protocol.

To establish SK-LMS-1 xenograft tumors, female BALB/c nude mice were purchased from Orient Bio (Seongnam). Autoclaved water and food were available to the mice ad libitum. A total of 2×10⁶ cells in 200 mcl of Hanks’ Balanced Salt Solution (Biocompare) was inoculated subcutaneously into the flanks of the animals bilaterally. The tumor volume was calculated using a standard formula (length × width² × 0.52), and growth curves were drawn.

To establish the PDX model, surgically removed patient tumor specimens were reduced to small pieces (less than 2–3 mm), implanted into the subrenal capsules of the left kidneys of BALB/c nude mice, and propagated by serial transplantation (29–31). Clinical information for the patient is provided in Supplementary Table S1. The mice used in these experiments were 6 to 8 weeks old. The drugs were orally administered (cabozantinib 5-mg/kg daily and temozolomide 5-mg/kg daily). The mice (n = 10 per group) were monitored daily for tumor development and sacrificed when any appeared moribund. The body weight, tumor weight, and number of tumor nodules were recorded. Tumors were fixed in formalin and embedded in paraffin or snap-frozen in optimal cutting temperature compound (Sakura Finetek Japan) in liquid nitrogen. This study was reviewed and approved by the Institutional Animal Care and Use Committee (IACUC) of Samsung Biomedical Research Institute (protocol number H-A9-003). The IACUC is accredited by the Association for the Assessment and Accreditation of Laboratory Animal Care International and abides by the guidelines of the Institute of Laboratory Animal Resources. The study of the PDX model for leiomyosarcoma was approved by the Samsung Medical Center Institutional Review Board (IRB file number 2010-04-004), and experiments were performed in accordance with the approved guidelines and regulations.

Immunohistochemical studies were performed on formalin-fixed, paraffin-embedded, 4-μm-thick tissue sections, as previously described (28). Immunostaining for Ki-67 (NB600-1252; NOVUS Biologicals) was performed with a BOND-MAXTM automated immune-staining device (Leica Biosystems) and a BONDTM polymer purification detection kit (Vision Biosystems). Briefly, antigen retrieval was carried out in ER1 buffer at 97°C for 20 minutes. The endogenous peroxidase activity was blocked with 3% hydrogen peroxidase for 10 minutes, and the antibody was diluted to 1:200 at room temperature for 15 minutes. Apoptotic-positive cells were analyzed by TUNEL assay using an ApopTag peroxidase in situ apoptosis detection kit (Merck Millipore; ref. 32).

Total RNA was isolated from IOSE, SK-LMS-1, SKN, SK-UT-1, MES-SA, and MCF-7 cells using TRizol (Invitrogen) and cDNA was synthesized using Superscript III reverse transcriptase. RT-PCR was performed using HotStart Pfu PCR PreMix (Bioneer). The primers used for MGMT (F: 5′-GATGCCGTGGAGGTCCCAGCC-3′/R: 5′-AGCCCGAGGTAGCTCCCGCT-3′) were obtained from Bioneer (Daejeon). The PCR products were visualized under ultraviolet light (Bio-Rad).

Cells were treated with drugs in 60-mm dishes for 48 hours. Cells were trypsinized and 1×10⁶ cells were fixed in 70% ethanol overnight at −20°C. Cells washed with PBS were stained with PI solution [50 μg/mL PI; Sigma-Aldrich), 0.1 mg/mL RNase A (Sigma-Aldrich), and 0.05% Triton X-100 in PBS] for 30 minutes at 37°C. Cell-cycle analysis was performed using FACSVerse (BD Biosciences).

The Mann–Whitney U test was used to compare differences between groups in both the in vitro and in vivo assays. All statistical tests were two-sided, and P values less than 0.05 were considered statistically significant. SPSS software (Version 27.0) was used for all statistical analyses.

The data generated in this study are available upon request from the corresponding author.

To evaluate the separate effects of cabozantinib and temozolomide on cell proliferation, an MTT assay was performed in the SK-LMS-1, SK-UT-1, MES-SA, and SKN sarcoma cell lines. In all four sarcoma cell lines, cell viability decreased as the concentration of cabozantinib or temozolomide increased (Fig. 1). The IC₅₀ values for each cell line are presented in Table 1. The inhibitory effects of cabozantinib or temozolomide were similar with 48 and 72 hours of treatment.

The effects of administering cabozantinib and temozolomide together to sarcoma cell lines were investigated. When the cell lines were treated with the drugs for 48 hours, the SK-LMS-1, MES-SA, and SKN cell lines showed synergistic inhibitory effects, whereas the SK-UT-1 cells did not show any change in cell proliferation (Fig. 2A). When the cell lines were treated for 72 hours, however, all cell lines showed a synergistic inhibitory effect with statistical significance (Fig. 2B).

The effect of combined cabozantinib and temozolomide on cellular apoptosis was evaluated by measuring annexin V, a calcium-binding protein that binds to phosphatidylserine, which is exposed on the outer leaflet of the plasma membrane, when apoptosis occurs. In all four cell lines, cellular apoptosis increased upon treatment with cabozantinib or temozolomide (1 μmol/L cabozantinib and 1 mmol/L temozolomide). The apoptosis rates further increased in all four cell lines when the two drugs were administered together in all four cell lines compared with treatment with each drug alone (Fig. 3A). Specifically, in SK-LMS-1 cells, the mean apoptosis rates were 6.3% versus 14.5% versus 24.4% in the cabozantinib versus temozolomide versus cabozantinib with temozolomide groups, respectively. In SK-UT-1 cells, the rates were 10.2% versus 20.4% versus 25.7%, respectively. In MES-SA cells, the rates were 14.3% versus 23.3% versus 70.0%, respectively, and in SKN cells, the rates were 20.8% versus 27.9% versus 54.5%. The synergistic effect was statistically significant compared with the treatment of each drug alone in all four cell lines. This synergistic effect was also demonstrated by the increased levels of active Caspase-3 when the cells were treated with cabozantinib and temozolomide together compared with the treatment of each drug alone (Fig. 3B).

Tumor cells from the PDX model mice were isolated, and the MTT assay and annexin V investigations were repeated. The MTT assay revealed a significant decrease in cell proliferation when the cells were treated with cabozantinib and temozolomide together for 48 hours compared with each drug administered alone. The same pattern of results was observed when the drugs were administered for 72 hours (Fig. 4A). The apoptosis rates of cells isolated from PDX model mice also increased when they were treated with cabozantinib and temozolomide together, consistent with the results from the previous experiments with the sarcoma cancer cell lines (Fig. 4B). The tumor cells used in this experiment mimicked the biological characteristics of the original tumor cells isolated from the patient. Therefore, these results indicate that the combination of cabozantinib and temozolomide has a synergistic inhibitory effect on tumor growth.

Cell line (SK-LMS-1) xenograft model mice were used to examine the effects of cabozantinib and temozolomide. The tumor volume in the model mice that received cabozantinib or temozolomide was significantly smaller than that in the control group (Fig. 5A). When the model mice received cabozantinib and temozolomide together, the mean tumor volume was again significantly smaller than when the drugs were administered alone, demonstrating the synergistic inhibitory effect of the drugs on tumor growth (1201.2 ± 875.5 mm³ vs. 819.0 ± 428.0 mm³ vs. 211.6 ± 126.9 mm³ vs. 121.6 ± 61.9 mm³ in control vs. cabozantinib vs. temozolomide vs. cabozantinib with temozolomide, measured on the 32nd day after tumor implantation). The tumor volume in the model mice that received cabozantinib and temozolomide together was almost 85% smaller than the tumor volume of the model mice that received cabozantinib alone and 43% smaller than the tumor volume of the model mice that received temozolomide alone (Fig. 5B). The body weight of the model mice was not affected by the drug treatment. PDX tumors of uterine leiomyosarcoma were also examined. Four groups of 10 mice were used. The cabozantinib group received 5 mg/kg of drug, and the temozolomide group received the same dose. The combination group received 5 mg/kg of cabozantinib and 5 mg/kg of temozolomide together. The mean tumor weight of the control group was 3.6 ± 0.7 g, and those of the cabozantinib and temozolomide groups were 2.9 ± 0.5 g and 2.9 ± 0.6 g, respectively. In comparison, the mice that received cabozantinib and temozolomide together had statistically significantly lighter tumors (2.0 ± 0.9 g, P < 0.001; Fig. 5C). The body weight of mice that received cabozantinib, temozolomide, or the combination did not differ among the groups after the administration. The TUNEL and Ki-67 proteins in the tumors harvested from the PDX model mice were analyzed by IHC. As shown in Fig. 5D, the number of TUNEL-positive cells was significantly greater in the model mice that received cabozantinib and temozolomide together than in those that received each drug alone (Fig. 5D). The number of Ki-67–positive cells was also significantly lower in the model mice that received cabozantinib and temozolomide together.

Western blot analyses of MET, AMPK, AKT, mTOR, S6K1, ERK, STAT3, and their corresponding phosphorylated forms were performed after treating SK-LMS-1 cells with cabozantinib, temozolomide, or both. Activation of c-MET and AMPK was inhibited by both cabozantinib and temozolomide whereas AKT, mTOR, S6K1, and ERK activity was inhibited by cabozantinib and STAT3 activity was inhibited by temozolomide (Fig. 6). Taken together, the expression of p-MET and p-AMPK decreased significantly when the cells were treated with cabozantinib and temozolomide together compared with treatment of each drug alone, indicating the potential downstream signaling mechanism by which the two drugs exert their synergistic effect. Because cabozantinib inhibited c-MET, we evaluated whether a c-MET–specific inhibitor, SU11274, could be used instead of cabozantinib. An MTT assay was performed in SK-LMS-1 with SU11274 and temozolomide. As shown in Supplementary Fig. S1A, cell viability decreased significantly when it was treated with SU11274 and temozolomide together compared with when it was treated by each drug alone. Western blot analysis demonstrated that p-MET levels were suppressed when the cells were treated with SU11274, but no additional inhibitory effect was seen when they were treated with SU11274 and temozolomide together (Supplementary Fig. S1B). Furthermore, the synergistic inhibitory effect of cabozantinib and temozolomide on AMPK levels was not seen when we treated the cells with SU11274 and temozolomide. To determine the expression levels of c-MET and AMPK in uterine leiomyosarcoma cell lines as well as normal ovary cell line, IOSE, Western blot analysis was performed. We found that the expression levels of MET were lower in SKN and MES-SA cell lines than in IOSE, whereas AMPK was highly expressed in all four uterine leiomyosarcoma cell lines compared with IOSE (Supplementary Fig. S2). The expression of phosphorylated proteins was consistent with the expression levels of their corresponding total proteins. On the basis of these findings, the synergistic inhibition by cabozantinib and temozolomide was thought to be mediated by the AMPK signaling pathway. It has been reported that temozolomide induces cell death by inducing G₂–M cell-cycle arrest (33). Therefore, cell-cycle analysis was also performed by flow cytometry (Supplementary Fig. S3) to determine whether the synergistic inhibition by cabozantinib and temozolomide was mediated by cell-cycle arrest. The results demonstrated that temozolomide induced G₂–M arrest in all four uterine leiomyosarcoma cell lines and S-phase arrest in SK-LMS-1 cell line. Cabozantinib had no significant effect on cell-cycle distribution in uterine leiomyosarcoma cell lines. Therefore, the cell-cycle arrest was thought to be mainly due to the administration of temozolomide.

This study demonstrated that the combination of cabozantinib and temozolomide synergistically inhibited tumor growth in a PDX mouse model of uterine sarcoma in vivo, and that result is supported by in vitro experiments with sarcoma cell lines. Despite recent advances, leiomyosarcoma continues to carry a dismal prognosis, and high recurrence rates have prompted research to find a more effective treatment modality. Previous studies have examined anthracycline-based regimens with or without ifosfamide, but no highly effective chemotherapeutic regimen has been identified. Numerous preclinical and early-phase clinical trials have indicated the potential therapeutic benefits of cabozantinib and temozolomide in soft-tissue sarcomas. Currently, there is an ongoing phase II clinical trial that is investigating the role of cabozantinib and temozolomide in patients with unresectable soft-tissue sarcoma (NCT04200443). The study is expected to be completed by the end of 2022 according to the researchers, and results are awaited.

Cabozantinib is known to have specific MET-receptor inhibitory activity (34). Studies have reported that MET receptors play a role in inducing tumorigenesis in soft-tissue sarcomas (35–37). In the CABONE study, significantly longer progression-free survival was observed in patients with osteosarcoma with highly soluble c-MET at baseline than in patients with low concentrations of soluble c-MET, suggesting that the concentration of soluble c-MET might predict the benefit of cabozantinib (34). The results of the Western blot analyses in this study agree with those previous results. Strong inhibition of MET phosphorylation was observed in cells treated with cabozantinib alone, suggesting the high specificity of that drug for MET receptors (38). In addition to the inhibition of MET receptors, cabozantinib also inhibits other receptor tyrosine kinases that are implicated in tumor growth, metastasis, and angiogenesis, with targets such as VEGFR2, RET, AXL, KIT, and TIE-2 (39). In vitro and in vivo pharmacodynamic activity against those receptors has been evaluated in several tumor types and been associated with tumor growth inhibition and tumor regression (11, 39). As previously described, temozolomide has also shown efficacy in various sarcomas, including leiomyosarcoma (,23, 27, 40–43). As previous studies have suggested, the potential mechanism by which temozolomide exerts an inhibitory effect on tumor growth is suppression of AMPK and STAT3, as seen in the Western blot analysis in this study (44).

Leiomyosarcoma is characterized by complex and uncertain genetic variations. Compared with other types of soft-tissue sarcomas, leiomyosarcoma has greater changes in the homologous recombination pathway (45). Although the genes driving this condition are not fully understood, some signaling pathways have abnormal enrichment, which suggests the potential for using precise inhibitors with multiple pathway targets, such as antivascular therapies, PARP inhibitors, or mTOR signaling inhibitors. The combined inhibition of multiple molecular pathways could have synergistic effects that enhance and maximize the inhibitory effect of each drug.

In a phase II clinical trial performed by a Japanese group, patients with heavily treated, relapsed uterine leiomyosarcoma received temozolomide and bevacizumab with (n = 6) or without cabozantinib (n = 9; ref. 20). The addition of cabozantinib to temozolomide and bevacizumab increased the clinical benefit rate (defined as complete response, partial response, and stable disease for more than 3 months) from 67% to 100%, though the ORR was 33% for both cohorts. A complete response was achieved in 3 patients (1 with cabozantinib and 2 without cabozantinib). The authors stated that toxicity was mild and manageable.

The strengths of the present study include the use of experiments with cell lines and patient-derived tumor cells to understand the changes in multiple signaling pathways caused by cabozantinib and temozolomide. They also include the use of xenograft mouse models to validate the eventual consequences of treatment. On the other hand, the study is limited in that it used tumor cells from only one patient who responded poorly to the standard treatment regimen. In consideration of the diverse molecular characteristics of this disease, further research should be conducted with various types of sarcoma samples.

Nonetheless, the results of the present study demonstrate that the combination of cabozantinib and temozolomide was highly effective in inhibiting tumor growth both in vivo and in vitro. The molecular mechanism by which cabozantinib and temozolomide inhibits tumor growth will be further tested in our future studies. The relative ease of administration combined with a low incidence of significant toxicity and a reasonable clinical response rate makes combined cabozantinib and temozolomide an appealing option for treating advanced or recurrent uterine leiomyosarcomas when first-line agents have failed or are not appropriate for a given clinical situation. Thus, it is suggested that cabozantinib and temozolomide are an attractive regimen for a prospective clinical trial in this setting. Further translational studies investigating drug activity at the molecular level could enable more accurate predictions of the clinical benefit offered by this regimen.

No disclosures were reported.

J.J. Noh: Formal analysis, writing–original draft. Y.-J. Cho: Data curation, investigation. J.-Y. Ryu: Investigation. J.-J. Choi: Resources, data curation, software. J.R. Hwang: Investigation, visualization. J.-Y. Choi: Investigation. J.-W. Lee: Conceptualization.

The present work was supported by a Korea Medical Device Development Fund grant funded by the Korean government (Ministry of Science and ICT; Ministry of Trade, Industry, and Energy; Ministry of Health and Welfare; Ministry of Food and Drug Safety; KMDF_PR_20200901_0153-2022). The present work was also supported by a National Research Foundation of Korea (NRF) grant funded by the Korean government (MSIT; 2020R1C1C1007482).

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.