IN VITRO AND IN VIVO EVALUATION OF ETOPOSIDE - SILK WAFERS FOR NEUROBLASTOMA TREATMENT

Abstract

High-risk neuroblastoma requires surgical resection and multi-drug chemotherapy. This study aimed to develop an extended release, implantable and degradable delivery system for etoposide, commonly used for neuroblastoma treatment. Different concentrations of silk, a biodegradable, non-toxic, non-immunogenic material were employed to prepare etoposide-loaded wafer formulations. Secondary structure of silk in the formulations was characterized using Fourier Transform Infrared (FTIR) spectroscopy and optimized based on the crystalline structure. Accelerated in vitro degradation studies under different conditions such as acidic, alkaline, oxidizing mediums and high temperature, were performed. The integrity of the silk wafer structure was maintained unless exposed to 0. 1N NaOH for 24 hours. In vitro release of etoposide was performed in PBS (phosphate buffered saline) at 37°C. Silk coated 6% wafers released the drug up to 45 days, while uncoated wafers released the drug for 30 days. Cytotoxicity study was performed on KELLY cells to evaluate the etoposide cytotoxicity (LC50) and the long-term efficacy of the etoposide wafer formulations. The results showed that etoposide killed 50% of the cells at 1 μg/mL concentration and the wafer formulations demonstrated significant cytotoxicity up to 22 days when compared to untreated cells. Using an orthotopic neuroblastoma mouse model, intratumoral implantation of the coated 6%, uncoated 6%, or uncoated 3% silk wafers were all effective at decreasing tumor growth. Histological examination revealed tumor cell necrosis adjacent to the drug-loaded silk wafer.

Graphics Abstract

INTRODUCTION

Neuroblastoma, the most common extracranial solid tumor of childhood, originates from neural crest cells and contributes to 15% of cancer related deaths under the age of 15 [1]. Approximately 40% of the patients are classified as high-risk, and despite aggressive treatments including surgical resection, systemic high-dose combination chemotherapy, autologous stem cell transplantation, immunotherapy and radiation therapy, high risk neuroblastoma patients have a high recurrence and mortality rate [2, 3]. Even with new dose reduction protocols for high-risk neuroblastoma treatment, intensive chemotherapy remains a key approach, predisposing patients to long- and short-term toxicities such as infection, cardiotoxicity, renal toxicity, myelosuppression and gastrointestinal symptoms [4–6]. The local application of chemotherapeutics offers effective drug delivery while avoiding systemic toxicity. Local chemotherapeutic implants are promising, since they are able to deliver the drug into the tumor or tumor resection bed and present high bioavailability [7]. Gliadel wafer^®, a carmustine implant is an example of a local delivery system for chemotherapeutics that has been approved for clinical use for the treatment of malignant glioma [8].

Etoposide, a topoisomerase II inhibitor and a commonly used drug for the treatment of high-risk neuroblastoma, causes apoptotic cell death by preventing re-ligation of DNA strands during DNA duplication [9]. Systemic delivery of etoposide is limited due to low water solubility and thus slow infusion is required up to 60 minutes with formulations containing solubilizers such alcohol or surfactant [10]. In addition, etoposide has serious systemic side effects such as myelosuppression, nephrotoxicity, acute hypotension and development of leukemia [11]. Several approaches to address these limitations have been pursued, such as nanoparticles [12], microemulsions [13], immunoliposomes [14] to improve systemic application of etoposide. However, due to the low solubility and bioavailability, and toxicity problems, there remains a need for an alternative system to deliver etoposide for successful neuroblastoma treatment.

Silk fibroin has been widely employed for biomedical applications such as tissue regeneration, drug delivery, and bioactive coating due to favorable biocompatibility, controlled biodegradability, low in vivo immune response, high drug loading capability and versatile formulations into hydrogels, sponges, fibers, films and tubes [15]. Silk has already been approved by the Food and Drug Administration (FDA) as medical sutures and more recently as a soft tissue support scaffold [16] and has been used for the delivery of neoplastic drugs [17–22]. Hydrophobic molecules have high affinity to silk fibroin, which makes this protein a suitable candidate for controlled delivery of chemotherapeutic drugs. Doxorubicin and vincristine have been loaded onto silk sponges and gels individually or in combination and demonstrated decreased tumor growth after intra-tumoral implantation into an orthotopic neuroblastoma tumor [20, 23, 24].

In this study, the goal was to use silk fibroin to prepare and characterize etoposide loaded sustained release wafer formulations for direct implantation in neuroblastoma tumors. In contrast to doxorubicin and vincristine, etoposide is very hydrophobic which makes it difficult to achieve high drug loading. Thus a novel carrier platform, silk wafers, were designed to achieve effective dosing and controlled release of etoposide. We hypothesized that a silk wafer system could deliver etoposide in a long-term, sustained release manner and decrease neuroblastoma tumor growth.

MATERIALS AND METHODS

Silk Fibroin Isolation

Silk fibroin was isolated from Bombyx mori cocoons as previously described [15]. Cocoons were cut into pieces and boiled for 30 minutes in 0.02 M Na₂CO₃ for degumming and then rinsed with distilled water to remove sericin proteins. Following overnight air-drying, 1 g of the silk fibroin was dissolved in 4 mL of 9.3 M lithium bromide for 4 hours at 60° C. This solution was dialyzed (Pierce 3.4 kDa MWCO dialysis cassette; Fisher Scientific, Pittsburg, PA) against distilled water for 2 days to remove the salt. The final concentration of the silk fibroin solution (silk) was calculated by weighing before and after drying and diluted to 3% and 6% with ultra purified water to prepare formulations. Silk solutions were stored at 4°C until use.

Etoposide – Silk Wafer Preparation

The 3% and 6% silk solutions were prepared by diluting the 30 minutes extracted silk stock. A 100 mg/mL etopos ide stock solution was prepared in DMSO and diluted with the 3% and 6% sterile filtered silk solution to obtain 1 mg/mL etoposide concentration with 1% DMSO. Then 100 μL of the prepared solutions were cast into each well of 96 well plates and frozen at −80 °C before lyophilization overnight. The silk foams that were obtained were compressed using a bench-top press to form wafers (3mm diameter & 1mm thickness, Figure 1) and the wafers were then water vapor annealed [25] to increase crystalline structure to obtain water insoluble implants. One group of the 6% silk wafers was coated with 20 layers of glycerin:6% silk solution (glycerin:silk ratio was 3:10 wt/wt) to evaluate the effects of the coating on etoposide release. Wafers were dipped into the coating solution individually and then left to air-dry to form each layer of coating. The process was repeated 20 times to obtain 20 layers. All formulation parameters and formulation abbreviations are given in Table 1 and schematic representation of the process flow for wafer presentation is presented in Figure 1.

Figure 1.

Schematic of silk wafer preparation and representative picture of a silk wafer with dimensions.

Table 1.

Abbreviations for Etoposide - Silk Wafer Formulations

Wafer Type	Silk Amount	Coating	Etoposide Amount	DM SO Amount	Loading Efficiency
EtoU-3%	3%	Uncoated	100 μg	1%	93.2%±5.5
EtoU-6%	6%	Uncoated	100 μg	1%	77.4%±6.2
Eto20×-6%	6%	20× Glycerin-Silk (3:10)	100 μg	1%	77.5%±25.8
B-3%	3%	Uncoated	-	1%	-
B-6%	6%	Uncoated	-	1%	-
B20×-6%	6%	20× Glycerin-Silk (3:10)	-	1%	-

Characterization

The structural features of the wafers with drug were evaluated using Fourier Transform Infrared (FTIR) Spectroscopy (JASCO FTIR 6200 spectrometer, Jasco, USA) and β-sheet content was determined as described previously [26]. OPUS 5.0 software (Bruker Optics, USA) was employed to deconvolute the amide I region (1605 − 1705 cm-¹) for silk fibroin using the Fourier Self Deconvolution method with Lorentzian line, half bandwidth of 27cm⁻¹ and a noise reduction factor of 0.3. IBM SPSS Statistics 22 Software (New York, USA) was used to perform One Way ANOVA for statistical analysis (p<0.05) of the FTIR results.

Drug loading efficiency was calculated to estimate the amount of etoposide in the formulations. Following in vitro release studies, the wafers were dissolved in 9.3 M lithium bromide solution and residual etoposide amount was quantified using ultraviolet/visible light spectroscopy (SpectraMax M2e, Molecular Devices, Sunnyvale, CA, USA). Absorbance values of the samples were read at 285 nm and blank silk wafer samples were used for background correction of silk. Total entrapped drug was calculated by adding the maximum released amount of drug and the residual amount of etoposide. The loading efficiency percentages were calculated in comparison with the theoretical drug amount per wafer.

Accelerated Degradation Studies (Stress Testing)

Accelerated degradation studies were performed to evaluate etoposide stability in silk wafer formulations for 24 hours under stress conditions. Both coated and uncoated wafers were incubated in 1 mL of PBS at 60°C, 0.1N HCl, 0. 1N NaOH or 3% H₂O₂ and compared to PBS controls at room temperature. Following 24 hours incubation, the wafers were washed with PBS once and incubated in fresh PBS at 37°C for 24 hours. PBS samples were collected and analyzed using an ultraviolet/visible light spectroscopy (SpectraMax M2e, Molecular Devices, Sunnyvale, CA, USA) and absorbance was read at 285 nm The amounts of released etoposide were compared following the accelerated degradation study. IBM SPSS Statistics 22 Software (New York, USA) has used to perform One Way ANOVA for statistical analysis (p<0.05).

In Vitro Release Studies

For drug release studies, the wafers were placed into 1 mL phosphate buffered saline (PBS), pH 7.4 (Life Technologies, Grand Island, NY) and incubated at 37°C. PBS was collected and replaced with fresh PBS at sampling time points. Absorbance of the sample solution was read at 285 nm using ultraviolet/visible light spectroscopy (SpectraMax M2e, Molecular Devices, Sunnyvale, CA, USA). Experiments were performed with 6 replicates and blank wafer release samples were used for background correction. The results of release were normalized based on the loading efficiencies of the formulations. IBM SPSS Statistics 22 Software (New York, USA) was used to perform MANOVA (multivariate analysis of variance) for statistical analysis (p<0.05).

In Vitro Cytotoxicity Studies

Etoposide cytotoxicity and long-term cytotoxicity of the extended release silk wafer formulations were evaluated via AlamarBlue® assay (Invitrogen, Carlsbad, CA) using human neuroblastoma KELLY cells (Sigma-Aldrich, St Louis, MO). The cells were maintained in RPMI 1640 with GlutaMAX (Gibco, GrandIsland, NY) supplemented with 10% fetal bovine serum, 100 IU/ml penicillin and 100 μg/ml streptomycin and incubated in a 5% CO₂ atmosphere at 37°C. Cells were passaged at 80% confluence using Trypsin-EDTA 0.25% (Gibco, GrandIsland, NY). For evaluation of etoposide cytotoxicity, KELLY cells, which are employed to induce tumors in the animal model as well, were exposed to varying concentrations of etoposide in culture medium and LC50 values were calculated (n=5). To evaluate formulation toxicity, EtoU-6% and Eto20×-6%, which provided long-term etoposide release, were tested in triplicates against blank formulations B-6% by exposing Kelly cells to the in vitro released samples. For AlamarBlue® assay, KELLY cells (P37) were seeded in 96-well plates at 10,000 cells per well and allowed to attach and adhere overnight. Medium in the plates was removed and replaced with either etoposide solutions in medium (100 μL, 10ng/mL to 100 μg/mL) or medium plus the etoposide release samples (1:1, 100 μL medium + 100 pL release sample) and incubated for two days before the AlamarBlue® assay was applied following manufacturers protocols. A 10% AlamarBlue® solution was added to the plates and etoposide samples were incubated for 4 hours while formulation samples were incubated for 24 hours at 37°C. The fluorescence was read on a plate reader (SpectraMax m2e, Molecular Devices) using excitation wavelength of 570 nm and emission wavelength of 585 nm One Way ANOVA was performed using IBM SPSS Statistics 22 Software (New York, USA) for statistical analysis (p<0.05).

In Vivo Studies

Cell Culture

Human neuroblastoma KELLY cells (Sigma-Aldrich, St Louis, MO) were maintained in RPMI 1640 (HyClone, Logan, UT) supplemented with 10% fetal bovine serum, 100 IU/ml penicillin, 100 μg/ml streptomycin, and 2 mM glutamine. All cells were maintained in a 5% CO2 atmosphere at 37°C and trypsin-passaged at 80% confluence.

Mouse Orthotopic Neuroblastoma Model

All mouse procedures were performed in accordance with University of Illinois’ recommendations for the care and use of animals and were maintained and handled under protocols approved by the Institutional Animal Care and Use Committee. The establishment of orthotopic neuroblastoma xenograft in a seven-week old female NCr nude mouse (Harlan, Indianapolis, IN, USA) was previously described [27]. Briefly, the mouse was anesthetized with inhalational isoflurane. A transverse incision was made on the left flank to locate the left adrenal gland, and 2 μL of phosphate buffered saline (PBS) containing one million KELLY cells were injected into the adrenal gland. The fascia and skin were closed in separate layers. Tumor formation was followed by non-invasive ultrasound measurements, and the animals euthanized when the tumor volume exceeded 1000 mm³. A full necropsy was performed at the time of tumor harvest.

Implantation of Silk Wafers

This procedure was previously described [27]. Briefly, once the tumor volume reached >100 mm³ as determined via high frequency ultrasound imaging, animals underwent treatment. Through a left lateral transverse incision the tumor was identified and dissected free from surrounding structures. Using cautery, the tumor capsule was incised, and the silk wafer was inserted into the tumor. The fascia and the skin were then closed in layers. The control wafers implanted for the in vivo experiments were the corresponding silk wafer formulation in the experimental groups (B-3%, B-6% or B20×-6%).

High Frequency Ultrasound

After anesthetizing with inhalational isoflurane, the mouse was secured in a prone position, and a VisualSonics Vevo 2100 sonographic probe (Toronto, Ontario, Canada) was placed over the left flank to locate the left adrenal gland and the tumor. Serial cross-sectional images (0.076 mm between images) were taken, and 3-D reconstruction tool (Vevo Software v1.6.0, Toronto, Ontario, Canada) was used to calculate the tumor volume. All tumors were measured every three to four days after the implantation of the silk wafer. Throughout this entire project, ultrasound of the animals and tumor volume calculations were performed by the same person (JZ) to ensure consistency.

Statistical Analysis for In Vivo Studies

Tumor sizes and post-operative days were entered into a scatter plot, and a curve of best fit with the associated equation was created for each animal. By entering the desired tumor size (500mm³, 600 mm³, 700 mm³, or 800 mm³) into the equation, we solved for the post-operative day using the “goal seek” function under the “what-if analysis” (Microsoft Excel, 2011, version 14.7.4, Redmond, Washington). The post-operative days obtained were compared between different groups using Student ‘s t test, and a p value <0.05 was considered statistically significant.

Histology

After the animals were euthanized, the tumors were removed, fixed in 10% buffered formalin, serially dehydrated and embedded in paraffin. Five-micron thick sections were collected onto glass slides. Sections were stained with hematoxylin and eosin. All sections were subjected to morphological examination by light microscopy.

RESULTS

Characterization of Formulations

β-sheet crystalline fractions define the stability and mechanical properties of silk, with a higher β-sheet content resulting in a stiffer silk material which also impacts drug release [28]. Crystallinity was induced by water vapor annealing which is a compatible method to preserve drug stability, β-sheet bands were observed between 1622 and 1637 cm⁻¹ and charted against the other fractions in silk structure (Figure 2). Deconvolution showed the β-sheet content of wafers was between 24.4 to 27.6 % and crystalline structure of coated wafers was significantly higher than the uncoated wafers (p<0.05). These findings correlated to more rapid drug release from the uncoated formulations as expected. The FTIR data also indicated that the addition of etoposide or DMSO did not have a significant effect on silk β-sheet content.

Figure 2.

Structural properties of silk wafers based on FTIR spectroscopy (n=3). (a) FTIR spectra of silk wafer formulations, (b) percentage of the silk content fractions in wafer formulations (statistical significance within groups has shown with *, (p<0.05)). See Table 1 for formulation codes.

Following in vitro release studies, the wafers were dissolved in 9.3 M lithium bromide solution and the total drug amount in the formulation was calculated (n=6). The 3% silk formulations had a higher loading than the 6% silk formulations where EtoU-3%, EtoU-6% and Eto20×-6% showed 93.2%±5.5, 77.4%±6.2 and 77.5%±25.8 loading efficiency, respectively (Table 1). A high standard deviation of Eto20×-6% samples suggested that the coating process might cause variability in drug loading or impact premature drug loss during processing steps.

Accelerated Degradation Studies (Stress Testing)

Following accelerated degradation, 24-hours release samples were collected and analyzed in order to compare with PBS controls to evaluate the integrity of wafer formulations. In accelerated degradation studies, the expectation is to achieve drug release as close as possible to the control group (PBS, 37°C) since the goal is to show that the extreme storage conditions do not have a significant effect on the in vitro release profile of the formulations. The results (Figure 3a) indicated that the formulations were intact and etoposide release was statistically similar (p<0.05) to the PBS control when incubated in PBS at 60°C, 0.1N HCl and 3% H₂O₂. However, incubation in 0.1N NaOH resulted in significantly lower etoposide release when compared to PBS controls and the coating did not have a protective effect in alkali conditions. In alkali formulations, silk starts to dissolve [29] and free drug is more soluble in alkali conditions, thus the expectation would be that higher amounts of drug would be released than from the controls. However, against this expectation of a higher in vitro release, and despite the wafer erosion, there was a significantly lower amount of free drug available. This data indicated that the drug was not stabilized in alkali conditions, leading a significant change in the in vitro release profile. Released etoposide decreased with increased temperature, however the difference was not statistically different when compared to PBS controls at room temperature. Based on the data, formulations are resistant to high temperature, acidic and oxidative conditions, however they should be protected from alkali conditions to preserve etoposide stability.

Figure 3.

a) Released etoposide following accelerated degradation (*significance for Etoposide Uncoated (EtoU-6%), #Significance for Etoposide 20× (Eto20×-6%) in comparison with PBS control (p<0.05), n=6). b) Cumulative etoposide released from silk wafers as a function of time (n=6). See Table 1 for formulation codes.

In Vitro Release Studies

The EtoU-3% formulation was designed for rapid etoposide release to provide an initial higher dose, while EtoU-6% and Eto20×-6% formulations had longer-term release profiles to provide the maintenance dose. In vitro release studies (Figure 3b) showed that EtoU-3% released 72% of the loaded drug (67.1 μg) in 24 hours and 92.6% (86.3 μg) in 2 days. These data meet initial dosing needs. The results also indicated that release from EtoU-6% and Eto20×-6% continued for 30 days and 45 days, respectively, which means both the coated and uncoated 6% wafer formulations provided extended release of etoposide and the coating slowed the release. MANOVA results showed that released etoposide amount from EtoU-3% wafers was statistically different than EtoU-6% and Eto20×-6% wafers at every sampling time point. The release profile of EtoU-6% and Eto20×-6% formulations were significantly different from each other at every time point except 2 hours, 6 hours and 45 days (p<0.05).

In Vitro Cytotoxicity Studies

Etoposide cytotoxicity on KELLY cells was investigated using AlamarBlue® assay and LC50 value was calculated based on cell viability data (Figure 4a). The data showed that etoposide killed 50% of the KELLY cells at a concentration of 1 μg/mL. Based on the LC50 data, 1 μg of etoposide killed 50,000 KELLY cells. To determine the long-term effect of the etoposide loaded silk wafer formulations, KELLY cells were exposed to the in vitro release samples of EtoU-6% and Eto20×-6% formulations in comparison with blank silk wafer formulations B-6% and B20×-6%. Cell viability data (Figure 4b) indicated that EtoU-6% killed 95% of the cells in the first 2 hours and more than 30% of the cells at every time point for 22 days. A peak in cytotoxicity was observed on day 15 with 21% cell survival possibly because of the degradation of the wafer. In contrast, Eto20×-6% was less cytotoxic than EtoU-6% due to the slower drug release rate where it killed 76% of the cells on day 7. The results showed that cells treated with both formulations returned to 100% cell survival at day 30. The silk carrier did not cause any significant toxicity on KELLY cells when compared to the untreated cells and PBS treated cell controls.

Figure 4.

Etoposide cytotoxicity results with KELLY cells (n=3) (*statistically different than the blank formulation or PBS, p<0.05). (a) Dose dependent etoposide cytotoxicity (LC50), (b) long term cytotoxicity of etoposide loaded silk wafers. See Table 1 for formulation codes.

In Vivo Studies

After the tumor reached 100 mm³, EtoU-6% (n=5), Eto20×-6% (n=4), and the corresponding control wafers, B-6% and B20×-6%, respectively (n=3 each) were implanted into the tumor to determine whether coated and uncoated etoposide-loaded wafers had an effect in vivo. Following implantation, tumors treated with EtoU-6% took 8.2 ± 3.0 days to reach 500mm³, Eto20×-6% took 5.9 ± 1.9 days, and control wafers 1.6 ± 0.6 days (Figure 5a). This difference was statistically significant between EtoU-6% and control wafer (p=0.01), between Eto20×-6%, and control wafer (p=0.02) but not between EtoU-6% and Eto20×-6% (p=0.23). This pattern of difference was also observed for tumors to reach 600mm³, 700mm³, and 800mm³ (Figure 5a).

Figure 5.

Days to reach tumor volumes 500mm³, 600mm³, 700mm³, and 800mm³ after implantation of (a) coated (Eto20×-6%) and uncoated (EtoU-6%) etoposide or control wafers, (b) 3% (EtoU-3%) or control wafers, (c) 6% (EtoU-6%) etoposide or control wafers. See Table 1 for formulation codes.

To determine whether 6% and 3% uncoated etoposide-loaded wafers affected tumor growth in vivo, after the tumor reached 100mm³, EtoU-6% (n=5), EtoU-3% (n=4), and the corresponding control wafers, B-6% and B-3%, respectively (n=7 total) were implanted into the tumor. Following implantation, tumors treated with EtoU-6% took 8.6 ± 1.7 days to reach 500mm³, EtoU-3% took 8.1 ± 1.7 days, 6% control wafers took 5.7 ± 1.0 days, and 3% control wafers 4.4 ± 1.0 days (Figure 5b-5c). This difference was statistically significant between EtoU-6% and 6% control wafer (p=0.01), between EtoU-3%, and 3% control wafer (p=0.02) but not between EtoU-6% and EtoU-3% (p=0.63). This pattern of difference was also observed for tumors to reach 600mm³, 700mm³, and 800mm³ (Figure 5b-5c).

At the time of tumor harvest, there was no tumor metastasis beyond the primary tumor. Hematoxylin and eosin stains of paraffin-embedded tumor sections demonstrated tumor necrosis adjacent to the wafer in tumors implanted with etoposide-loaded silk wafers. Tumors implanted with control wafers showed viable tumor cells adjacent to the wafer. The representative images from EtoU-3% and B-3% were shown in Figure 6. There was no foreign body reaction or macrophage adjacent to the wafers.

Figure 6.

Histology of tumor sections stained with hemotoxylin and eosin. (A) Tumor treated with control wafer. Tumor cells exhibited features of classic neuroblastoma histology of “small round blue cell.” (B) Tumor treated with EtoU-3%, Tumor cells adjacent to the drug-loaded silk wafer (area denoted as ★) were necrotic (area denoted as ■).

DISCUSSION

Safe and effective treatment of solid tumors remains a challenge that requires alternative formulations to reduce systemic toxicity and increase the efficacy of chemotherapy. Conventional treatments such as oral and injectable chemotherapeutics present serious systemic side effects and insufficient efficacy due to dose limitations. As a result, alternative tumor-targeted therapies are being sought, including implantable drug delivery systems. Gliadel wafer^®, the first and only FDA approved implant for chemotherapeutics, raised interest for implantable systems for the treatment of solid tumors. Currently there is no local treatment clinically available for neuroblastoma, which led us to investigate the possibility of designing silk-based local implant systems for the treatment of neuroblastoma.

Silk is a suitable drug carrier with biocompatible/biodegradable structure and an ability to form a variety of carrier systems such as microparticles, gels, hydrogels, films, sponges and wafers. We have successfully formulated different silk-based chemotherapeutic delivery systems in our previous studies and tested for local delivery of doxorubicin and vincristine for the treatment of neuroblastoma [18, 19, 21, 22, 30]. Controlled release doxorubicin–silk films were implanted over the neuroblastoma tumor resection beds and local doxorubicin treatment was effective in control of tumor growth [23]. Combinations of doxorubicin and vincristine were formulated in silk sponges and these controlled release drug-loaded silk sponges were implanted into murine orthotopic neuroblastoma tumors [24]. Compared to delivering the same amount of chemotherapeutic drugs intravenously, implanting the drug-loaded silk sponges into the center of the tumor significantly decreased tumor growth, with higher intra-tumoral drug concentration and lower drug concentration in plasma. Poly(lactic-co-glycolic acid) has been used to prepare etoposide loaded intravitreal implants where drug loading was 33.3% and the implant system released 57% of the drug in 50 days in vitro [31]. Recently, poly (L-lactic acid) (PLLA)/polyethylene glycol 4000 (PEG4000) sustained release etoposide implants were developed by direct compression of 40% drug, 50% PLLA and 10% PEG4000. Implants were able to release in vitro up to 140 hours, followed by testing in a subcutaneous tumor model where tumor sizes decreased significantly [32]. These results encouraged the development of intratumoral treatment for neuroblastoma and possibly other solid tumors. When compared to other FDA approved biomaterials such as PLLA, poly (lactic-co-glycolic acid) (PLGA), silk has a higher affinity to hydrophobic molecules which results in longer release times that can be further modified by changing silk concentration or molecular weight. Furthermore silk scaffolds/implants generally degrade in 2 to 6 months in vivo, without any residuals or changes in pH [33], unlike PLGA that generates acidic monomers during degradation [34]. Thus silk fibroin is a suitable candidate for the local delivery of hydrophobic chemotherapeutics with control over drug loading, release duration and degradation time.

Higher β-sheet content usually results in longer, sustained drug release [35, 36]. We used high (6%) and low (3%) concentration silk solutions to achieve different crystalline states after water vapor annealing in the wafers to control the release rate. FTIR results confirmed that silk coating increased β-sheet content significantly. We demonstrated with in vitro release studies that EtoU-3% released 92.6% of etoposide (86.3 μg) in the first two days to serve as the initial dose whereas Eto20×-6% released the drug up to 45 days. These formulations have been applied to study the in vivo efficacy of a high burst dose or a long release period. Even though the release profile of EtoU-3% and EtoU-6% was different, the tumor response after the treatment appeared to be similar. One explanation for this observation could be due to the limitation in drug diffusion. Etoposide needed to diffuse across the tumor in order to kill the tumor cells. There could be a maximal rate that the drug could diffuse across the tumor. Once that rate was reached, additional etoposide would not result in additional tumor kill. Another explanation could be that the additional drug that was released initially from EtoU-3% compared to that from EtoU-6% was insufficient to achieve any measureable increase in tumor kill. Even though EtoU-3% had a higher initial release compared to that from EtoU-6%, EtoU-6% released the drug over a longer period of time than that of EtoU-3%. The differences in initial amount of drug release could possibly be counteracted by the differences in release period. Clinical studies showed that 0.5–2.0 μg/ml of etoposide concentration in plasma is required to achieve therapeutic activity when applied systemically [37, 38]. We can deliver up to 100μg etoposide into the tumor, and the silk concentration can be varied to tailor the drug amount released over any selected time period.

The intracellular pH is similar between tumors and normal tissues, however extracellular pH is significantly lower in tumors than normal tissues [39], There are also oxidizing molecules such as lipid peroxidases in neuroblastoma tumors [40]. These environmental differences may effect the degradation or the release properties of the wafers when implanted into tumors. Thus, we have examined the degradation of the wafers in different incubation conditions such as acidic pH and oxidizing medium The formulations retained their integrity, however, the release was higher at the lower pH and in the presence of H₂0₂.

The cytotoxicity of the wafers was investigated in vitro with KELLY cells. The LC50 data for etoposide showed that 1 μg of etoposide killed 50,000 KELLY cells in vitro. We were able to load 100 μg of etoposide per wafer, thus 1 wafer could kill about 5 million cells. It has been reported that without any blood vessels a tumor can grow up to about 10⁶ cells. However, tumors can also grow into a larger mass in the presence of new blood vessels [41]. Increasing drug loading would increase the efficacy even in larger tumors. EtoU-6% was the most effective formulation, where 100 μg of etoposide loaded in the wafers was effective for the in vivo mouse neuroblastoma model.

CONCLUSIONS

We prepared and characterized a controlled release silk-based etoposide delivery system for neuroblastoma treatment. Silk wafers with a biodegradable structure can be easily implanted within a solid tumor. Dosing control of the silk wafer can be varied by the silk content, and implantation of the etoposide-loaded silk wafers within orthotopic neuroblastoma tumors resulted in decreased tumor growth.