|Home | About | Journals | Submit | Contact Us | Français|
Achieving effective treatment outcomes for patients with glioblastoma (GBM) has been impeded by many obstacles, including the pharmacokinetic limitations of antitumor agents, such as topotecan (TPT). Here, we demonstrate that intravenous administration of a novel nanoliposomal formulation of TPT (nLS-TPT) extends the survival of mice with intracranial GBM xenografts, relative to administration of free TPT, because of improved biodistribution and pharmacokinetics of the liposome-formulated drug. In 3 distinct orthotopic GBM models, 3 weeks of biweekly intravenous therapy with nLS-TPT was sufficient to delay tumor growth and significantly extend animal survival, compared with treatment with free TPT (P ≤ .03 for each tumor tested). Analysis of intracranial tumors showed increased activation of cleaved caspase-3 and increased DNA fragmentation, both indicators of apoptotic response to treatment with nLS-TPT. These results demonstrate that intravenous delivery of nLS-TPT is a promising strategy in the treatment of GBM and support clinical investigation of this therapeutic approach.
Topotecan (TPT), an analogue of the chemotherapeutic camptothecin, is an approved treatment for many types of cancer, including recurrent small cell lung and ovarian tumors.1–4 With respect to brain tumors, TPT has shown significant activity against the most common and malignant type of primary brain tumor, glioblastoma (GBM), both in vitro and in subcutaneous xenograft models of GBM.5,6 Results from preclinical studies, in combination with encouraging results regarding the distribution of TPT in cerebrospinal fluid (CSF) after systemic TPT administration, prompted clinical investigations of TPT monotherapy efficacy in treating patients with GBM.5–7 Unfortunately, results from phase II clinical trials involving TPT treatment of both newly diagnosed and recurrent GBM revealed only modest patient responses.8–10 Although no longer being investigated as a monotherapy for treating GBM, TPT continues to be used in GBM clinical trials as a part of combination therapy approaches.
The minor efficacy of TPT as a monotherapy for GBM has been attributed to a rapid rate of TPT clearance from the CSF (t1/2CSF = 4.8 h) and rapid TPT inactivation in plasma (active form t1/2 = 23 min).11 Liposomal encapsulation is a strategy being explored for improving the antitumor efficacy of camptothecin derivatives,12,13 including TPT,14 and has been shown to increase drug circulatory half-life while helping maintain drug activity by providing an appropriate pH.12,15
In fact, systemic injection of nanoliposomal TPT has shown enhanced activity, relative to free TPT, against subcutaneous xenografts for several types of human cancer.13,16 Moreover, nanoliposomal TPT has demonstrated superior efficacy in treating orthotopic (intracranial) xenograft models of GBM when administered directly into the tumor by convection-enhanced delivery (CED).12,14 Specifically, in a rat orthotopic GBM xenograft model, CED of nanoliposomal TPT increased TPT half-life in the brain, relative to free TPT, and conferred a highly significant survival advantage. In spite of such demonstrations, the invasive nature of CED, combined with the limited success of CED-associated clinical trials to date, has restrained enthusiasm for more widespread use of this route of therapeutic administration in treating patients with brain tumor.
Here, we show for the first time, to our knowledge, the enhanced efficacy of systemically administered nanoliposomal TPT in 3 orthotopic xenograft models of GBM. The interpretation of enhanced efficacy is based on bioluminescence monitoring of tumor growth and therapeutic response, survival benefit to animal subjects, and immunohistochemical analysis of tumor apoptotic response to therapy, with the results from each consistent in their support for clinical investigation of this approach for treating GBM.
U87-MG cells were maintained as monolayers in high-glucose Dulbecco's minimal essential medium (DMEM) supplemented with 10% fetal calf serum, 1% nonessential amino acids, penicillin, and streptomycin. Cells were cultured at 37°C with 5% carbon dioxide. GBM43 and GBM6 cells were maintained as serially passaged subcutaneous xenografts in athymic mice. U87-MG, GBM43, and GBM6 were each modified by lentiviral infection for stable expression of firefly luciferase.17 For intracranial injection of GBM43 and GBM6, a subcutaneous tumor was removed, minced with a scalpel, and subjected to 3 rounds of passage through a 40- µm pore filter with centrifugation after each round of filtering (increasing speed: 158 g, 355 g, 631 g for 10 min each). After the final round of centrifugation, the cells were resuspended in 1 mL of sterile DMEM media (without antibiotics), counted, and diluted to 1 × 108 cells/mL for intracranial injection. U87-MG cells were harvested for intracranial injection by monolayer trypsinization and resuspended in DMEM at 1 × 108 cells/mL.
All animal experiments were conducted using protocols approved by the University of California, San Francisco, Institutional Animal Care and Use Committee. Four to six-week-old female athymic nu/nu mice (Simonsen Labs) were anesthetized by intraperitoneal injection of ketamine (100 mg/kg) and xylazine (10 mg/kg). A 1-cm sagittal incision was made along the scalp, and the skull suture lines were exposed. A small hole was created by puncture with a 25-g needle at 2 mm to the right of the bregma and 0.5 mm anterior of the coronal suture. With use of a sterile Hamilton syringe (Stoelting), 3 × 105 cells in 3 μL were injected at a depth of 3 mm over a 60-s period. After injection, the syringe was held in place for 1 min and then slowly removed. The skull was cleaned with 3% hydrogen peroxide and then sealed with bone wax. The scalp was closed using 7-mm surgical staples (Stoelting). Mice were monitored daily and euthanized when exhibiting significant neurological deficit or >15% reduction from their initial body weight.
A detailed description of nanoliposomal TPT preparation is given elsewhere.13 In brief, liposomes for packaging TPT consisted of hydrogenated soy phosphocholine and 1,2-distearoyl-sn-glycero-3-phosphoethanolamine-N- [methoxy(polyethylene glycol)-2000] at a molar ratio of 3:2:0.015. A sucrose octasulfate gradient was used for loading TPT into the liposomes. Unencapsulated TPT was removed by Sephadex G-25 gel filtration chromatography, eluting with HEPES buffered saline (pH, 6.5). The phospholipid concentration of the purified solution was measured spectrophotometrically using a standard phosphate assay.18 Before HPLC column injection for TPT quantitation, liposome samples were solubilized in a solution containing 1% trifluoroacetic acid in methanol. TPT was detected by fluorescence (excitation, 370 nm; emission, 535 nm) with a retention time of 11.5 min. The nanoliposomal TPT (nLS-TPT) had a final drug to phospholipid ratio of 327.7 ± 5.31 g TPT HCl per mol phospholipid. Liposome size was determined by photon correlation spectroscopy using a Coulter N4 Plus particle size analyzer (Beckman Coulter) and was reported as the volume-weighted mean diameter ± standard deviation of the liposome size distribution (100.1 ± 8.4 nm).
Mice with intracranial U87-MG tumors were injected with 5 mg/kg of either free TPT or nLS-TPT via tail vein, then euthanized at 1, 8, 24, or 48 h after injection by trans-cardiac perfusion with phosphate-buffered saline (PBS). Blood samples were obtained immediately before perfusion via cardiac puncture and were centrifuged for 10 min at 8154 g. The plasma was then removed and frozen at −80°C. After perfusion with PBS, the tumor-bearing brain hemispheres were resected and frozen at −80°C. TPT was extracted from 10–20 µL of plasma with 980–990 µL of 1% trifluoroacetic acid in methanol. The mixture was vortexed for 5 s and stored at −80°C for 2 h, followed by centrifugation at 13 400 g for 10 min. The supernatant was transferred to an HPLC autosampler vial and stored at 4°C until analyzed for TPT content. HPLC analysis was performed as described in “Preparation of Nanoliposomal TPT.” Brain tissues were processed with a mechanical homogenizer in water at 20% wt/wt ratio; 100 µL of the tissue homogenate was added to 400 µL of 1% trifluoroacetic acid in methanol and subjected to the same extraction and analysis procedure that was used for plasma. The spiked extraction recovery from tumor-containing brain tissue was 97.8%. The standard curve linearity was a mean of >0.998 with a detection limit of 0.1 ng/mL. TPT concentrations shown include both free and encapsulated TPT. Area under curve (AUC) calculations used the trapezoidal rule.19
Athymic mice were treated with 1 mg/kg doses of either free TPT or nLS-TPT as described in “TPT Treatment.” Blood samples were collected weekly from the submandibular vein into EDTA-coated tubes, according to the University of California, San Francisco Institutional Animal Care and Use Committee/Laboratory Animal Resource Center (UCSF IACUC/LARC) standard procedure,20 and were analyzed using a Hemavet 850 (Drew Scientific) within 2 h after collection.
Luciferase-modified U87-MG cells were plated in a 96-well plate at a concentration of 5000 cells/well. Twenty-four hour later, cells were treated for either 24 h or 48 h with 2.67 × 10−3 mg/mL free TPT or nLS-TPT. After incubation, the cells were washed in prewarmed PBS, and fresh media without phenol red was added to the cells. D-luciferin was added to each well at a final concentration of 0.6 mg/mL. After a 10-min incubation, cells were examined for luminescence on the IVIS Lumina System (Caliper Life Sciences), as described in the next section.
Mice were anesthetized by intraperitoneal injection of ketamine (100 mg/kg) and xylazine (10 mg/kg) and then were injected intraperitoneally with 33.3 mg of D-luciferin (potassium salt; Gold Biotechnology) dissolved in sterile saline. Tumor bioluminescence was determined 10 min after luciferin injection using the IVIS Lumina System (Caliper Life Sciences) and LivingImage software, as the sum of photon counts per second in regions of interest defined by a lower threshold value of 25% of peak pixel intensity. Imaging was performed biweekly from 3 days after tumor implantation to completion of study.
TPT (Fisher) was dissolved in pyrogen-free sterile water and stored at −20°C. Before treatment, mice were randomized according to bioluminescence imaging score at most recent imaging. Both nLS-TPT and free TPT were administered to mice via tail vein injection.21 Immediately before injection, nLS-TPT and free TPT were diluted to a concentration of 0.2 mg/mL in either 5 mM HEPES (nLS-TPT) or sterile water (free TPT). Mice received TPT administrations twice weekly (Tuesday and Friday), for a maximum of 6 administrations.
After resection, mouse brains were fixed for 48 h in 10% buffered formalin. Brains were then paraffin-embedded and sectioned (10 µm) for hematoxylin and eosin staining and immunohistochemical analysis. To assay DNA fragmentation, TUNEL staining was performed using the DeadEnd Colorimetric TUNEL system (Promega) according to the manufacturer's protocol for paraffin-embedded tissues. For the determination of cleaved caspase-3 reactivity, unstained sections were processed using a Ventana BenchMark XT automated system and a protocol consisting of pretreatment with 3% ethanolic hydrogen peroxide for 16 min at room temperature, epitope retrieval in Tris buffer (pH, 8) for 8 min at 90°C, and incubation with primary antibody to cleaved caspase-3 (Cell Signaling Tech) at 0.2 µg/mL for 32 min at 37°C.
PRISM 5, Version 5.03 (GraphPad Software) was used to conduct all statistical analyses. For survival analysis, significance was determined by the log-rank (Mantel-Cox) test. For all other analyses, a 2-tailed unpaired t test was applied.
Athymic mice with intracranial U87-MG tumors were injected with a single dose of nLS-TPT or free TPT and sacrificed at 1, 8, 24, and 48 h after injection, with tissue samples removed for HPLC analysis of TPT content. At all times after injection, there was significantly more TPT in tumor-bearing brain from mice treated with nLS-TPT than in brain from mice treated with free TPT (P < .05, all time points) (Fig. 1A). In fact, after 24 h, mice treated with free TPT had no detectable TPT in tumor-bearing brain samples (limit of detection = 1.67 ng TPT/g tissue), whereas mice treated with nLS-TPT continued to show detectable TPT in brain samples at 48 h after administration of nLS-TPT, the last time point assayed. Furthermore, AUC calculations, indicating total drug exposure, revealed that brain tissue from mice receiving nLS-TPT was 5-fold greater than for mice receiving free TPT (Table 1).
Corresponding results for the analysis of plasma TPT showed that nLS-TPT treatment significantly increased the duration of and total exposure to TPT in circulation (Fig. 1B, Table 1). The plasma AUC∞ for mice treated with nLS-TPT was 197 µg*h/mL, over 200-fold higher than the AUC∞ in mice treated with free TPT (0.655 µg*h/mL). With respect to duration, TPT was not detectable in plasma samples from mice 48 h after administration of free TPT, whereas the 48-h concentration of TPT in mice receiving nLS-TPT was readily detectable and significantly greater than the 8- and 24-h plasma concentrations for mice receiving free TPT.
Mice receiving treatment with free TPT or nLS-TPT and bearing intracranial U87-MG xenografts experienced a reduced tumor growth rate relative to untreated controls (Fig. 2A), with nLS-TPT significantly outperforming free TPT. To address the consistency of nLS-TPT efficacy in vivo, we applied the same experimental design as used with the U87-MG experiment to orthotopic GBM models in which the tumor cell sources were obtained from serially propagated subcutaneous xenografts (Fig. 2B and C). This approach for tumor maintenance has been previously shown to promote retention of patient GBM molecular and biologic characteristics.22,23 Bioluminescence monitoring of response to treatment for xenografts GBM43 (proneural subtype) and GBM6 (classical subtype)24 (Fig. 2B and C) revealed antitumor activity of nLS-TPT consistent with that indicated for the U87-MG model, supporting the finding that nLS-TPT is more effective in suppressing tumor growth.
Consistent with our prior experience in using bioluminescence imaging for monitoring intracranial tumor response to therapy,17,25 corresponding survival analysis for all 3 GBM models tested (U87-MG, GBM43, GBM6) showed that suppression of tumor growth is predictive of survival benefit. Mice treated with free TPT experienced increased median and mean survival, relative to control group mice, but of lesser extent than mice treated with nLS-TPT (Fig. 3). Of importance, in all 3 xenograft models, there was a statistically significant survival benefit from nLS-TPT treatment, compared with free TPT treatment (P ≤ .03 for all models).
To rule out any intrinsic differences in TPT efficacy related to liposomal packaging, we compared free TPT and nLS-TPT antitumor effects in vitro employing the same luciferase-modified U87-MG cells as used in vivo (Figs 2A and and3A).3A). Results for 24- and 48-h incubations indicated similar activities for free TPT and nLS-TPT (Supplemental Figure 2), supporting the differential efficacy observed in vivo as being attributable to improved drug distribution and stability from nanoliposomal formulation.
To assess mode of action and biologic consequences of nLS-TPT treatment in vivo, athymic mice bearing GBM43 tumors were injected with a single dose of either free TPT or nLS-TPT and were sacrificed 24 h after treatment. Brains were removed, fixed, embedded, and then sectioned for molecular analysis of intracranial xenograft response to therapy. In contrast to tumors from untreated mice, tumors from mice treated with free TPT or nLS-TPT showed significantly higher levels of TUNEL staining (Fig. 4A–C), indicative of more extensive DNA fragmentation, with cellular positivity in tumors exposed to nLS-TPT significantly higher than in tumors exposed to free TPT (P < .01 for all 2-way comparisons). Cleaved caspase-3 staining showed a similar pattern of results, confirming the highest tumor apoptotic response from nLS-TPT (Supplemental Figure 1). Staining for Ki67, to address the effects of therapy on tumor cell proliferation, showed no significant difference in cellular positivity between treatment groups (data not shown).
For each in vivo experiment, the body weight of all animal subjects was monitored. There was no significant difference in mean body weights between free TPT– and nLS-TPT–treated mice during the period in which mice remained asymptomatic of tumor burden, irrespective of tumor model being tested (Supplemental Figure 3).
To compare the myelosuppressive effects of free TPT and liposomal TPT and to contrast TPT myelosuppressive effects against the most frequently used cytotoxic chemotherapeutic for treating GBM, athymic mice were either given 10 mg/kg of temozolomide (TMZ) for 5 consecutive days or TPT using the same free TPT and nLS-TPT regimens as in the efficacy experiments (Figs 2 and and3).3). Peripheral blood samples were collected weekly, and blood cell counts were determined. For all time points tested, neither free TPT nor nLS-TPT treatment had a significant effect on hematocrit or hemoglobin levels (Supplemental Figure 4), suggesting that nLS-TPT treatment is unlikely to cause anemia. In contrast, mice receiving TMZ experienced a near-significant decrease in hemoglobin levels (control vs TMZ: P = .08). Platelet levels were unaffected by TPT or TMZ treatments (Supplemental Figure 5). There was a significant decrease in neutrophil levels at the first time point after initiation of TMZ or TPT therapy (day 4: control vs free TPT, P = .01; control vs nLS-TPT, P = .05; control vs TMZ, P = .01) (Fig. 5). Of importance, neutrophil levels recovered rapidly after completion of the treatment cycle, with baseline levels achieved within 7 days after the last administration of TPT. In addition, there was no significant difference in neutrophil levels between nLS-TPTand free TPT–treated mice at any time point.
After recovery of blood cell counts to pretreatment levels for mice receiving TMZ, the same mice were administered nLS-TPT to investigate myelosuppressive effects of nLS-TPT when used as a salvage therapy for recurrent GBM. Results show that administration of nLS-TPT after blood cell recovery from TMZ monotherapy has a lesser myelosuppressive effect than does the initial TMZ treatment (Fig. 6).
Although the use of camptothecins in treating primary brain tumors has been widely researched, these compounds have yet to be used routinely in clinical practice, in large part because of their toxicity and related pharmacokinetic shortcomings.8,26 Liposomal encapsulation of chemotherapeutics has been shown to increase tumor drug exposure and to reduce systemic toxicity.27–30 Because of the size of therapeutic liposomes, previous brain tumor studies have often focused on their direct, local delivery to tumors by CED.31–34 Although nLS-TPT CED has shown efficacy in preclinical investigations using rodent models of GBM,14 enthusiasm for use of this route of administration has been restrained because of the invasive nature of CED and the limited success of CED-associated clinical trials conducted to date.35
In contrast to CED, intravenous delivery is less invasive and promotes a more uniform drug distribution throughout the brain, which is an important consideration because of the invasive nature of GBM. In our study, we investigated intravenous administration of nanoliposomal TPT for efficacy against 3 types of GBM xenografts and report for the first time, to our knowledge, that liposomal encapsulation enhanced TPT concentration in the brain, resulting in increased antitumor activity. We have previously shown that empty liposomes do not have antitumor activity,36 indicating that it is the liposomal packaging of TPT, and not the liposomes, that is responsible for improved antitumor activity in relation to free TPT.
By intercalating into DNA, TPT inhibits an essential process in proliferating cells, namely the role of topoisomerase I (TOPI) in DNA replication, ultimately resulting in DNA strand breaks that initiate programmed cell death. Indeed, the results of our TUNEL analysis show increased DNA strand breaks in TPT-treated tumors (Fig. 4) and a corresponding increase in activated caspase-3 (Supplemental Figure 1), a marker of programmed cell death. Of importance, DNA strand breaks and activated caspase-3 were significantly elevated by liposomal packaging of TPT. These results complement our previously published data indicating that TPT treatment depletes topoisomerase I in tumor cells36 and, in combination, address the mode of action and biologic consequence of TPT activity.
In contrast to targeted therapeutics, such as Tarceva, for which an antitumor effect has been shown to be specific to a subclass of GBM,37 TPT is expected to have a more generalized effect against GBM because of the essential nature of topoisomerase I in tumor growth. In fact, an analysis of The Cancer Genome Atlas (TCGA) data for DNA TOPI expression shows significantly elevated TOPI mRNA levels irrespective of GBM subclassification (Supplemental Figure 6). Results from our analysis of 3 distinct GBM cell sources (U87-MG and serially propagated subcutaneous xenografts GBM6 and GBM43) for TPT treatment response (Figs 2 and and3)3) support the antitumor effect of this cytotoxic chemotherapeutic as being generalizable to most if not all subtypes of GBM.
The increased efficacy of nLS-TPT did not come at the expense of increased toxicity; there was not a significant difference in either mouse body weights or in neutrophil counts in mice treated with nLS-TPT, compared with mice treated with free TPT (Supplemental Figure 3, Fig. 5). Moreover, our analysis of blood cell counts in mice treated with TMZ followed by treatment with nLS-TPT indicate that nLS-TPT has a favorable safety profile for use in treating recurrent GBM.
Previous studies have not shown a significant advantage to adding TPT to radiation therapy.38,39 However, because of the dramatic improvements in distribution and efficacy seen with nLS-TPT, further study into nLS-TPT combined with radiation seems warranted. In addition, it would be of interest to examine the efficacy of TPT when used in combination with inhibitors of proteins that prevent apoptosis, such as obatoclax mesylate (GX15-070MS), a small-molecule pan-Bcl-2 family inhibitor recently shown to be well-tolerated when administered with TPT to patients with cancer.40
In total, our study results and previous results from others indicate several promising possibilities for maximizing benefit from intravenous administration of nLS-TPT and support additional investigation of this therapeutic agent and approach.
Conflict of interest statement. D.C.D. and D.B.K. are employees of Merrimack Pharmaceuticals. C.O.N. is a former employee of Merrimack Pharmaceuticals.
This work was supported by NS065819-02 (to C.D.J.) and CA097257 (to M.D.P., J.W.P., C.D.J.).
We thank Raquel Santos, the UCSF Brain Tumor Tissue Core, and the UCSF Mouse Pathology Core for technical assistance.