Abstract
Aim: To determine the benefits of irradiation at the time of maximum linking of oxaliplatin to the DNA of tumor cells, and evaluate the potential of its liposomal formulation, Lipoxal™, for chemoradiation therapy. Materials and Methods: Nude mice implanted with human colorectal carcinoma HCT116 cells were injected with oxaliplatin or Lipoxal™. The amount of platinum in tumor, tumoral DNA, normal tissues and blood was measured 4, 24, 48, 72 and 96 h later by inductively coupled plasma mass spectrometry. The effect of concomitant radiotherapy was assessed as tumor growth delay resulting from irradiation 4, 24 and 48 h after drug administration. Results: While the amount of platinum in the tumor reached a peak at 4 h after injection and declined over time, the concentration of oxaliplatin–DNA adducts reached two maxima observed at 4 h and 48 h after drug administration, a behavior not observed with Lipoxal™. The greatest combined effect was obtained when radiation was given at 48 h after drug injection, resulting in an increase of tumor growth delay by factors of 3.71 and 3.33 for treatments with oxaliplatin and Lipoxal™, respectively. Conclusion: Our results confirm the importance of irradiating a tumor when the concentration of oxaliplatin bound to tumor DNA is maximal. This finding should have a significant impact on the design of more efficient chemoradiation treatment protocols and should be further explored in clinical studies.
Advances in concomitant chemoradiotherapy have been attempted to improve the treatment of rectal cancer (1) by using chemotherapeutic drugs as radiosensitizers (2). Platinum-based drugs such as oxaliplatin have been studied and integrated into treatment strategies against rectal cancer (3). Although such drugs can act as radiosensitizers, their radiosensitization is limited by a narrow therapeutic index which excludes dose escalation. In recently completed clinical phase III trials (ACCORD-0405, STAR-01 and NSABPR-04) oxaliplatin was added to 5-fluorouracil (5-FU) or capacitabine and combined with radiation (4, 5). With the exception of the German CAO/ARO/AIO-04 study where oxaliplatin was added to 5-FU plus radiation (3), these trials failed to improve combined treatment. Furthermore, some difficulties in managing side-effects in combined treatment, such as hematological and gastrointestinal toxicity, are still commonly found in clinical practice (6).
Considering the limited number of publications on the use of oxaliplatin in combination with radiation, Flatmark and Ree have suggested the need for testing the radiosensitizing activity of oxaliplatin alone in pre-clinical settings prior to its clinical use as a local radiosensitizer (7). Moreover, to obtain the full benefit of the drug–radiation interactions, while maximizing tumor response, the radiosensitizing activity and the most efficient schedule of combined treatment remain to be determined.
The role of oxaliplatin as a radiosensitizer has been described recently (8). Like other platinum compounds, the main mechanism of anti-neoplastic and radiosensitizing activity of oxaliplatin is thought to occur mostly via the formation of platinum–DNA adducts (9, 10). The oxalate bidentate ligand of oxaliplatin preferentially reacts with the highly nucleophilic N7 position on guanine or adenine, and forms coordinated covalent bonds. Oxaliplatin can bind to two sites in DNA with the following order of preference: -GG-> -AG->> -GA-. The resulting bi-adduct is composed of approximately 60-65% 1,2-intrastrand GG, 25-30% 1,2-intrastrand AG, 5-10% 1,3-intrastrand GXG and 1-3% interstrand GG cross links (11). Platinum binds to DNA and causes critical structural changes, such as a bend of 45 degrees, unwinding of DNA, and disruption of the purine bases. The formation of platinum–DNA adducts consequently interferes with cellular repair and DNA replication, and triggers a chain of cell-regulatory events, which ultimately lead to cell death (12).
According to fundamental considerations (13, 14), the most favorable concomitant effect should occur when the amount of the radiosensitizer in the DNA of cancer cells is at a maximum value. In practice, however, the activity of radiosensitizers is often linked to the concentration of the drug accumulated in the tumor tissue (15-17). To date, very few publications have reported studies on pre-clinical combined treatment of oxaliplatin and radiation for rectal cancer. Furthermore, information underlying the assessment of radiotherapy at the maximum formation of oxaliplatin–DNA adducts is not yet clarified in both pre-clinical and clinical settings. In the present study, we hypothesized that the potential of combined treatment depends on the interaction of radiation at different times after drug administration, which correspond to different levels of platinum–DNA adducts.
Previous studies have demonstrated the efficacy of oxaliplatin in the treatment of colorectal cancer; however, the antitumor activity of oxaliplatin may be limited by its severe systemic side-effects, such as neurotoxicity, hematological and gastrointestinal toxicity (18). Liposomal oxaliplatin (Lipoxal™) was developed to reduce the systemic toxicity of oxaliplatin while attempting to improve its anticancer efficacy (19). As a single agent, Lipoxal™ has significant cytotoxicity towards human colorectal cancer cells and showed adequate efficacy in pre-treated patients with colorectal cancer in a phase I study (20). Lipoxal™ is currently under study in phase II treatment of colorectal cancer (19). The combination of radiation and Lipoxal™ is, therefore, an interesting pre-clinical research subject that might shed some light on new clinical treatment strategies. Insufficient data exist concerning pharmacokinetics, anticancer activity, and combined treatment efficacy of Lipoxal™ compared to its parent compound oxaliplatin. In addition, studying its combination with radiotherapy is important in defining the role of Lipoxal™ as a radiosensitizer.
In the present work, variations of platinum accumulation with time were measured in blood, and different normal and tumoral tissue compartments, as well as platinum bound to tumoral DNA. Thereafter, we explored the relationships between different levels of platinum–DNA adducts and the potential of combined treatment of oxaliplatin and Lipoxal™ with radiation. The results of this study should provide a better understanding over the mechanism of platinum-induced radiosensitization and should have an impact on the design of more efficient chemoradiation treatment protocols.
Materials and Methods
Animal xenograft model. Animal experimentations were approved by the Université de Sherbrooke Animal Care and Use Committee (Approval number 235-10). HCT116 colorectal carcinoma cell line (American Type Culture Collection, Manassas, VA, USA) was inoculated (2×106, 0.1 ml) subcutaneously (s.c.) into each rear flank of outbred male nu/nu nude mice at 4-6 weeks of age (n=3-5) (Charles River Laboratories, Saint-Constant, QC, Canada). Tumor measurements began one week post-injection and continued biweekly. Tumor volumes were calculated with the formula: V (mm3)=π/6 × a (mm) × b2 (mm2), where a and b were the largest and smallest perpendicular tumor diameters, respectively. The pharmacokinetic studies began when tumor volumes reached the range of 30-60 mm3. For all procedures (implantation, chemotherapy and radiotherapy), mice were anesthetized with an intraperitoneal injection of ketamine/xylazine (87/13 mg/ml) at 1 ml/kg.
Pharmacokinetics studies. Oxaliplatin was obtained through the pharmacy of the Centre Hospitalier Universitaire de Sherbrooke. Lipoxal™, the liposomal formulation of oxaliplatin, was generously provided by Regulon Inc. (Athens, Greece). Drugs were dissolved in 5% dextrose for a final concentration of 1 mg/ml to be given intravenously (i.v.) as a single dose of 10 mg/kg into the tail vein of mice. Three groups of mice were studied: i) oxaliplatin-treated, ii) Lipoxal™-treated, and iii) treated with 5% dextrose only, as a control group. Three to five animals from each group were sacrificed at 4, 24, 48, 72, and 96 h post-injection. Blood samples were obtained via cardiac puncture for whole-blood and serum platinum assessments. Tumors and samples of different tissues were taken to determine platinum accumulation in the cytoplasm, and the nucleus, as well as in DNA extractions (Figure 1A).
Chemotherapy treatment. Platinum drugs (10 mg/kg) were i.v. injected into the tail vein at 4, 24 and 48 h prior to radiation treatment (Figure 1B). The selected drug concentration of oxaliplatin and Lipoxal™ did not induce significant side-effects, since the body weight of all groups of animals did not change between the initial and final time of combined treatment (data not shown).
Tumor irradiation. Mice were anaesthetized and positioned in our in-house constructed stereotactic frame designed for the 4C Gamma Knife (GK) (Elekta Instruments AB, Stockholm, Sweden) (Figure 1C) (23, 24). The radiation treatment (15 Gy, dose rate of 3.6 Gy/min) was delivered using 8-mm collimators at predetermined coordinates targeting the tumor, which had a diameter less than 7 mm. Radiation was applied to one side of the rear flank, whereas the other side was kept as the non-irradiated control tumor. The radiation dose of 15 Gy was taken from a previous study (25) as the dose that induces temporary tumor growth delay but does not lead to complete tumor cure. This radiation dose therefore allows assessment of the combined effect of chemotherapy with radiation. On the other hand, a radiation protocol of 25 fractions of 2 Gy as used in clinic cannot be applied in an animal model because daily anesthesia before each irradiation would be lethal for the animal. In addition, a single dose of 15 Gy was considered to be equivalent to the 25 fractions of 2 Gy used for rectal cancer (3, 26).
Determination of platinum accumulation. Cytoplasm and nucleus of tumor samples were isolated using the Activemotif extraction kit (ActiveMotif, Carlsbad, CA, USA). DNA in tumor samples was extracted according to a salting-out procedure (21). The quantity of platinum accumulated in tissue samples and in the cytoplasm, nucleus and DNA of tumor was determined with an inductively coupled plasma mass spectrometer (ICP-MS) (ELAN DRC-II; PerkinElmer, Waltham, MA, USA) (22, 23) after treatment with 23% nitric acid and 8% H2O2. As internal controls, platinum and thallium (m/z 195 and 205, respectively) were quantified in untreated animals. Cytoplasmic and nuclear uptake of platinum in tumor tissue samples were expressed as nanograms of platinum per gram of cytoplasm or nucleus. The concentration of platinum in whole blood and in serum was expressed as micrograms of platinum per milliliter of whole blood and serum. Platinum–DNA adducts were expressed as picograms of platinum per μg of DNA. Muscle, liver and kidney uptake was expressed as micrograms of platinum per gram of tissue.
Apoptosis analysis. Groups of three mice bearing HCT116 tumors were treated with platinum drugs 24 or 48 h before the radiation treatment, as described above. Mice were sacrificed two days after irradiation. Tumors were removed, fixed in 10% buffered formalin and then embedded in paraffin. Tumor sections of 4 μm were taken at three different levels. For histological studies, tumor sample sections were dewaxed with xylene and rehydrated with hematoxylin and eosin. To detect apoptotic cells within tumor tissues, tumor sections were analyzed with a commercially available ApopTag® Peroxidase In Situ Apoptosis Detection Kit (Millipore, Temecula, CA, USA), following the manufacturer's protocol. A positive control slide was prepared by nicking DNA with DNaseI and treated similarly (data not shown). Samples treated similarly but without enzyme treatment served as negative controls. To determine the amount of cellular area stained with diaminobenzidine (positive for apoptosis) and the amount of tumor area counterstained with methyl green (negative for apoptosis), the tumor sections were scanned with an Olympus FSX100 microscope (Olympus, Center Valley, PA, USA) and the values were quantified using Adobe Photoshop CS5 analysis software (Adobe Systems Canada, Ottawa, ON, Canada). Briefly, the percentage of apoptotic cells was calculated as the quotient of the diaminobenzidine-positive area over the total area (diaminobenzidine-positive plus methyl green-negative), multiplied by 100.
Statistical analysis. The mean±SD were calculated. A value of p<0.05 was considered statistically significant (two-tailed paired Student's t-test).
Results
Platinum concentration in whole blood, serum, tumor and normal tissues. After i.v. administration, the total blood levels of oxaliplatin and Lipoxal™ declined steadily, with a tendency for later stabilization, or even slightly increased levels. These changes were more pronounced for oxaliplatin than Lipoxal™ and for whole blood than serum. Serum levels were about 10% of the corresponding levels in whole blood (Figure 2A and B). At 4 h, the concentration of platinum in tumor tissues after treatment with oxaliplatin was about twice that found in the Lipoxal™-treated group (Figure 2C). For both drugs, the total tumoral platinum content declined gradually over time. Similar results were obtained in muscle, kidney and liver tissues (Figure 2D-F).
Platinum accumulation in cytoplasm and nucleus of tumor cells and tumoral DNA-bound platinum. After peaking initially at 4 h, the amount of platinum in the cytoplasm and the nucleus declined gradually over time for both oxaliplatin and Lipoxal™ (Figure 3A and 3B). The level of tumoral DNA-bound platinum stabilized over time in the case of Lipoxal™; while for oxaliplatin, the level of tumoral DNA-bound platinum reached a first peak at 4 h, and then gradually declined to reach a first nadir at 24 h, then rose again to reach a second and higher peak at 48 h before re-declining. Even at 72 h and 96 h, levels of tumoral DNA-bound platinum were found to be closer to those at 4 h (Figure 3C).
Tumor growth delay. Tumor response was assessed as the time necessary to reach five-times the initial tumor volume (5Td) and the tumor growth delay (TGD=5Td value for a treated group minus 5Td for the control group) (Figure 4) (Table I). A single oxaliplatin or Lipoxal™ administration induced a significant tumor growth delay compared to non-treated animals (p<0.05). Treatment with radiation alone had considerable effects, with a 5Td of 19±1.0 days compared to 10.5±0.71 days compared to the non-treated control group (p<0.01).
When oxaliplatin or Lipoxal™ was combined with radiation, the tumor growth delay increased significantly compared to the control group (p<0.01) (Figure 4A). Combined treatment with oxaliplatin 4, 24 and 48 h prior to radiation treatment induced a dramatic reduction in tumor growth compared to oxaliplatin alone (p<0.01). Similar increase of tumor growth delay was observed after combined treatment with Lipoxal™ injected 4, 24 and 48 h before tumor irradiation compared to Lipoxal™ alone (p<0.01) (Figure 4B). A significant difference in tumor growth delay between drug treatments with and without radiation was measured five and seven days after irradiation for Lipoxal™ and oxaliplatin, respectively.
Better treatment was obtained with oxaliplatin and Lipoxal™ in combination with radiation than in the group treated only with radiation (p<0.01), with the exception of radiation 24 h after oxaliplatin administration (p=0.08). Efficacy was similar in the groups irradiated 4 h and 24 h after oxaliplatin injection, while the best combined effect was obtained with the group irradiated at 48 h. The higher efficacy of the latter treatment was observed starting 20 days post-treatment (Figure 4A). For Lipoxal™, the tumor growth delay was not significantly different among combined treatment schedules (Figure 4B).
Enhancement factor. The enhancement factor (EF) was calculated to assess the benefit of combining radiation with oxaliplatin or Lipoxal™ (Table I). When radiation was given 48 h after oxaliplatin, the EF was 3.71, suggesting an important and significant improvement; while lower EFs were measured when radiation treatment was given 4 h or 24 h after oxaliplatin administration. The encapsulation of oxaliplatin in a liposome, i.e. Lipoxal™, resulted in an overall significantly improvement of radiosensitivity with EFs greater than 3 at time intervals of 4, 24 and 48 h between drug administration and tumor irradiation. As for oxaliplatin, the lowest EF with Lipoxal™ was measured at 24 h post drug administration.
Induction of apoptosis. The percentage of apoptotic cells (Figure 5) measured was 0.94±0.38% and 2.86±0.89% in the untreated group and in the group treated with radiotherapy-alone. Compared to treatment with oxaliplatin-alone (1.59±0.31%), combined treatment of oxaliplatin with radiation significantly increased the percentage of apoptotic cells to 3.89±1.05% and 12.53±2.35% when oxaliplatin was administrated 24 h and 48 h prior to radiation treatment (p<0.05). Similarly, a higher percentage (p<0.05) of apoptotic cells was observed when radiation was given 48 h after Lipoxal™ injection (16.32±4.81%) compared to 24 h after drug administration (6.41±1.72%).
Discussion
The optimal chemoradiotherapy schedules for colorectal cancer using oxaliplatin-based therapy are actively being investigated. However, pre-clinical evidence of oxaliplatin as a radiosensitizer for colorectal cancer treatment is limited and its precise mechanism of action remains undetermined. As a general rule, the radiosensitizing activity of platinum-based chemotherapeutic agents has been linked to two mechanisms of interaction between drug and radiation: first, by inducing additional DNA damage or modifying radiation damage in DNA, and secondly, by inhibiting post-irradiation repair of DNA damage (27).
According to the present study, after treatment with a single dose of the chemotherapeutic agents oxaliplatin and Lipoxal™, the platinum concentration in tumoral tissue declines gradually over time. However, the amounts of platinum drug binding to tumoral DNA followed different kinetics. Indeed, while the amount of platinum in tumor reached an initial peak at 4 h after injection and declined over time, the concentration of oxaliplatin–DNA adducts reached two maxima observed at 4 h and 48 h after drug administration, a behavior not observed with Lipoxal™. Additionally, Pieck and co-workers reported a characteristic time course of oxaliplatin–DNA adducts in white blood cells of 37 patients with cancer who received 130 mg/m2 oxaliplatin as a 2 h infusion (28). They showed that the adduct levels at 24 and 48 h, as well as at the maximum platinum:nucleotide ratio, were correlated to tumor responses. It was notable that all patients with tumor accumulating less than 4 Pt atoms per 106 nucleotides at 48 h post-injection were not responsive to oxaliplatin.
Variations in the amount of platinum bound to DNA may be explained by different types of adducts that each drug forms with DNA. The initial reduction of the amount of platinum–DNA adducts after 4 h shown in Figure 3C suggests the formation of mono platinum–DNA adducts (Pt-dG or Pt-dA), which are rapidly removed by the nucleotide excision repair system. At 24 h after oxaliplatin administration, most DNA mono-adducts have possibly been repaired, thus minimizing the amount of platinum still bound to DNA. Therefore, exposure to radiation 4 h and 24 h after oxaliplatin administration had less impact on the HCT116 xenografts, with corresponding EFs of 2.38 and 2.19, respectively. The time interval between these two modalities appeared to be important since the maximal effect of combination therapy was observed 48 h after drug administration, with an EF of 3.71. This may be explained by the increasing presence of other types of platinum–DNA adducts which form more slowly or are not rapidly removed. In fact, the amount of bi-functional adducts formed directly, or following rearrangements of monofunctional adducts, are known to continue to increase up to 48 h after drug administration (29). Furthermore, bi-functional adducts forming interstrand cross-links contribute significantly to the drug activity and are more resistant to DNA repair systems (30-32). Therefore, a maximum formation of platinum–DNA adducts, especially as interstrand cross-links, might occur after 48 h of oxaliplatin administration, leading to maximum radiosensitivity after oxaliplatin therapy. This hypothesis may contribute to a better understanding of the radiosensitizing activity of oxaliplatin, and explain the wide variation in the results of different clinical reports that describe the efficacy of oxaliplatin as a radiosensitizer (9, 32-34). In addition, the amount of platinum–DNA adducts after drug administration may be considered as an excellent indicator predicting the effect of combined treatments.
Although induction of apoptosis is not dominant after tumor irradiation, our results show that treatment with radiation alone significantly increased the level of apoptosis as compared with the untreated group. This increase was even more important when combining oxaliplatin or Lipoxal™ with radiation, suggesting that this type of cancer cell death might contribute to the antitumor effect, as observed by the increasing tumor growth delay in our animal model of colorectal cancer. The ability of platinum to induce apoptosis after irradiation was previously reported. Yang et al. suggested that DNA damage induced by radiation is usually repairable, but can become non-repairable owing to the capacity of platinum to scavenge free electrons (35), leading to apoptotic cell death. However, the specific mechanisms that trigger apoptosis in response to platinum–DNA damage have not yet been fully elucidated.
The ability of liposomes to cross leaky tumor vessels was suggested to increase its selective localization in tumor tissues (36). However, our results show that the encapsulation of oxaliplatin in liposome did not improve its accumulation in tumor cells, nor its fraction bound to DNA. As liposomal formulations are believed to reduce systemic toxicity, a higher concentration of Lipoxal™ could be administrated compared to oxaliplatin.
Lipoxal™ was previously evaluated in vitro for its cellular toxicity and potential radiosensitizing properties (37). In an animal model of glioblastoma, treatment of Lipoxal™ combined with radiation showed promising results in terms of animal survival (38); however, the animals were not irradiated at optimal time intervals after Lipoxal™ administration. In the present study, the optimal interval between Lipoxal™ administration and tumor irradiation was assessed. In contrast to oxaliplatin, efficacy after intervals of 4 and 48 h was not significantly better than that observed 24 h after Lipoxal™ treatment. On the other hand, although the encapsulation of oxaliplatin in liposome resulted in smaller amounts of platinum–DNA adducts compared to its free form at 24 h post drug administration, a greater effect when combined with radiation was observed compared to oxaliplatin (EFLipox=3.05 vs. EFox=2.19). This might be caused in part by a larger percentage of apoptotic cells induced by combining radiation with Lipoxal™ than with oxaliplatin.
The mechanisms of interaction between Lipoxal™ and radiation remain to be clarified. Previous studies have suggested that the binding of platinum drugs to cellular proteins via sulfur atoms in the cysteine or methionine residues may affect the activity of enzymes, receptors, and other proteins (39). Moreover, platinum drugs can induce a high level of mitochondrial reactive oxygen species, which would further interfere with vital cellular functions, leading to cell death (26). These data, thus, provide interesting avenues for further investigations on chemoradiation therapy combined with Lipoxal™, both in experimental models and in the clinic.
With respect to improving outcome of patients treated with oxaliplatin and radiation, our results demonstrate that the benefits of combination therapy depend on the administrative schedule of both modalities. This correlates with platinum levels in different cellular compartments, mainly with those of platinum–DNA adducts. The impressively large increase in time delay in tumor growth observed with Lipoxal™ and radiation treatment compared to Lipoxal™ alone should encourage further evaluation of the radiosensitizing activity of Lipoxal™ for application in clinical trials. A better understanding of the mechanisms of synergism between radiotherapy and oxaliplatin could lead to the design of new and more efficient chemoradiation protocols for treatment of colorectal cancer.
Acknowledgements
This research was financed by the Canadian Institute of Health Research (grant # 81356). Léon Sanche and Benoit Paquette are members of the FRQS-funded Centre de Recherche du CHUS. Sanofi-Aventis Canada also offered a partial unrestricted grant to support this project.
- Received June 12, 2014.
- Revision received July 22, 2014.
- Accepted July 24, 2014.
- Copyright© 2014 International Institute of Anticancer Research (Dr. John G. Delinassios), All rights reserved