Antitumor activity of PEGylated biodegradable nanoparticles for sustained release of docetaxel in triple-negative breast cancer

International Journal of Pharmaceutics 473 (2014) 55–63

Contents lists available at ScienceDirect

International Journal of Pharmaceutics journal homepage: www.elsevier.com/locate/ijpharm

Pharmaceutical nanotechnology

Antitumor activity of PEGylated biodegradable nanoparticles for sustained release of docetaxel in triple-negative breast cancer Giuseppe Palma a,b,1, Claudia Conte c,1, Antonio Barbieri a , Sabrina Bimonte a , Antonio Luciano a , Domenica Rea a , Francesca Ungaro c , Pasquale Tirino d, Fabiana Quaglia c, * , Claudio Arra a a

Animal Facility, National Cancer Institute – Foundation “G. Pascale”, Via Mariano Semmola, 80131 Napoli, Italy Institute of Experimental Endocrinology and Oncology, CNR, Via Pansini, 80131 Napoli, Italy Drug Delivery Laboratory, Department of Pharmacy, University of Napoli Federico II, Via Domenico Montesano 49, 80131 Napoli, Italy d Department of Chemistry Paolo Corradini, University of Napoli Federico II, Via Cintia, 80126 Napoli, Italy b c

A R T I C L E I N F O

A B S T R A C T

Article history: Received 15 May 2014 Received in revised form 27 June 2014 Accepted 28 June 2014 Available online 30 June 2014

With the aim to find novel therapeutical approaches for triple-negative breast cancer (TNBC) treatment, we have developed a powder for i.v. injection based on cyclodextrins and docetaxel (DTX)-loaded polyethyleneglycol-poly(epsilon-caprolactone) nanoparticles (DTX-NPs). Nanoparticles are designed to concentrate at tumor level by enhanced permeability and retention effect and release drug cargo at a sustained rate in the blood and in tumor interstitium. DTX-NPs of around 70 nm, shielding proteins and allowing a sustained DTX release for about 30 days, were produced by melting sonication technique. DTX-NPs were associated to hydroxypropyl-b-cyclodextrin to give a powder for injection with excellent dispersibility and suitable for i.v. administration. DTX-NPs were as efficient as free DTX in inhibiting cell growth of MDA-MB231 cells, even at low concentrations, and displayed a comparable in vivo antitumor efficacy and better survival in a TNBC animal model as compared with DTX commercial formulation (Taxotere1). In conclusion, PEGylated biodegradable DTX-NPs highlighted their potential in the treatment of aggressive TNBC providing a foundation for future clinical studies. ã 2014 Published by Elsevier B.V.

Keywords: PEGylated nanoparticles Triple-negative breast cancer Docetaxel Powder for injection Sustained release

1. Introduction Triple-negative (basal-like) breast cancer (TNBC) is a clinically relevant term referring to a group of breast tumors that do not express the estrogen receptor (ER), progesterone receptor (PR) and human epidermal growth factor receptor type 2 (HER2). This subgroup accounting for 17–21% of all breast carcinoma has been reported to be more aggressive and have worse prognosis, due to a higher risk of relapse, metastasis, and shorter survival (Bauer et al., 2007; Sorlie et al., 2003). Due to lacking of both hormone receptors and HER2 expression, they have no chance to benefit from the endocrine therapy and HER2 targeted therapy. Consistently with its aggressive nature, TNBC is characterized by high rates of distant recurrence within the first five years after diagnosis despite adjuvant chemotherapy. Hence, development of new therapeutic strategies for these clinically intractable tumors is highly desirable.

* Corresponding author. Tel.: +39 81 678707; fax: +39 81 678707. E-mail address: [email protected] (F. Quaglia). 1 These authors contributed equally to the work.

Liedke et al. (Liedtke et al., 2008) retrospectively evaluated the effectiveness of neoadjuvant chemotherapy for TNBC. They reported that the pathological complete response (pCR) rate was 20% for anthracyclines, 28% for anthracyclines in combination with taxanes, 12% for taxanes, and 14% for other regimens, thus concluding that the administration of anthracyclines plus taxanes is the most effective treatment modality for TNBC. Moreover, neoadjuvant therapy with docetaxel (DTX) produced a pCR rate of 23% in a phase II trial (Yagata et al., 2011). Adjuvant chemotherapy with taxanes reduces the risk of cancer recurrence and death in patients with early or operable breast cancer and has shown benefits in patients with high-risk node-negative breast cancer (Isakoff, 2010). Currently, the clinical DTX formulation Taxotere1 is associated to several adverse effects, including hypersensitivity reactions, hemolysis and peripheral neuropathy, due to the absence of selectivity for target tissues (Fumoleau, 1997). In order to extend the therapeutic application of DTX and to optimize its pharmacokinetic behavior as well as to ameliorate its toxicity profile, part of research efforts are nowadays devoted to the development of injectable nanomedicines (d'Angelo et al., 2014; Jabir et al., 2012; Yang et al., 2013). Nanocarriers such as

56

G. Palma et al. / International Journal of Pharmaceutics 473 (2014) 55–63

liposomes, dendrimers, polymeric micelles and nanoparticles (NPs) have been developed to increase the therapeutic index of several cytotoxic drugs by prolonging their serum half-lives while boosting solid tumor accumulation through the enhanced permeability and retention (EPR) effect (Danhier et al., 2010; Hirsjarvi et al., 2011; Maeda et al., 2013). Nevertheless, NPs providing a sustained release of the entrapped drug can be of great benefit since they could not only decrease the duration/number of administration/s but also to expose continuously the drug to the tumor, although their concrete advantages over other systems have not been fully considered (Conte et al., 2014). A safe strategy to build sustained-release nanocarriers involves the use of biodegradable and biocompatible polyester derivatives with amphiphilic properties such as poly (lactic acid) (PLA) covalently linked to monomethoxy-polyethyleneglycol (PEG) (Mikhail and Allen, 2009). NPs with a lipophilic core wrapped by a hydrophilic biomimetic PEG cloud are formed after their assembly. Some nanocarriers based on this type of polymers have been proposed for delivering taxanes and are already in clinical trials or have been approved by the FDA for the treatment of ovary, lung and breast cancer (Genexol-PM1, Nanoxel-PMTM) (Lammers et al., 2012; Lim et al., 2010). Results obtained in clinical studies suggest the possibility to employ this system as an alternative, less toxic and efficacious Tween 80-free DTX vehicle (Kim et al., 2007; Lee et al., 2008, 2011). However, a great drawback of these nanocarriers relies on their fast disassembly once in contact with blood components (Chen et al., 2008) and, as a consequence, poor control over time of drug release rate. To overcome this drawback, we have proposed recently NPs based on amphiphilic and biodegradable block copolymers of poly (epsilon-caprolactone) and PEG (PEG–PCL) to deliver DTX in the body (Ungaro et al., 2012a). NPs were prepared by a meltingsonication technique (MeSo) developed in our labs, which allows formation of drug-loaded NPs with high PEG coverage by exploiting low melting temperature and amphiphilic properties of PEG–PCL (Quaglia et al., 2006). These systems are considered very promising for the delivery of poorly water soluble chemotherapeutic agents such as taxanes (Conte et al., 2013; Liu et al., 2012; Shahin and Lavasanifar, 2010; Zhang et al., 2012) due to their high affinity for PCL core (Conte et al., 2013; Quaglia et al., 2008; Savic et al., 2006a). Nevertheless, PCL-based nanocarriers are less prone to disassembly in the bio-environment due to the high crystallinity of PCL core (Mikhail and Allen, 2009). We have demonstrated that PEG–PCL NPs loaded with DTX (DTX-NPs) exhibited a milder toxicity profile in healthy mice as compared to the commercial DTX formulation (Taxotere1) (Ungaro et al., 2012b). In perspective, such a system could also mimic metronomic administration of a chemotherapeutic agent where frequent exposition of the body at a low non-toxic level of drugs can be of help in optimizing administration schedule in cancer patients (Pasquier et al., 2010). On this basis, here we develop an i.v. formulation for DTX based on PEG–PCL NPs obtained by MeSo and investigate its activity in TNBC. NPs were fully characterized to predict their biological behavior after i.v. administration. Antitumor activity of the formulation was investigated in vitro and in vivo in comparison with free DTX and its commercial formulation, respectively. 2. Materials and methods 2.1. Materials Monomethoxy-polyethyleneglycol–poly( e-caprolactone) diblock copolymer (PEG–PCL with PEG = 2000 Da and PCL = 4300 Da) was prepared as previously described (Ungaro et al., 2012a). Docetaxel (DTX) (99%) was purchased from LC laboratories (USA).

Human serum albumin (HSA) (99%), (2-hydroxypropyl)-b-cyclodextrin (HPbCD, DS = 0.6), EZBlueTM Gel Staining Reagent, polysorbate 80, nile red (NR) (99%), 1-D sodium dodecyl sulfate (98%), polyacrylamide (99%), potassium phosphate dibasic (98%), potassium phosphate monobasic (98%), sodium azide (99%) and sodium chloride (98%) were purchased from Sigma–Aldrich. Sodium hydroxide was provided from Delchimica Scientific Glassware. Ethanol (96%), phosphoric acid (85%), acetonitrile (99.9%) and tetrahydrofuran (99.9%) were purchased from Carlo Erba Reagenti (Milan, Italy). Human plasma was obtained from the EDTA-treated blood of one healthy individual according to institutional bioethics approval. DMEM enriched with FBS 10% was used as cell culture medium. 2.2. Preparation of nanoparticles Empty and DTX-loaded NPs (DTX-NPs) were prepared by a MeSo technique as previously described (Ungaro et al., 2012a). Briefly, 10 mg of PEG–PCL were added to 1.8 mL of water and heated at 72  2  C. DTX (around 10% of copolymer weight) was dissolved in 0.2 mL of ethanol and added to melted copolymer/water. The mixture was sonicated for 10 min at 3 W (Sonicator 3000, Misonix, USA) and finally cooled at room temperature. NPs were filtered through 0.45 mm filters (RC, Chemtek, Italy), added with HPbCD as cryoprotectant (10:1 mass ratio with copolymer) and freeze-dried for 24 h to give a DTX-NPs/CD powder. Recovery yield of production process was evaluated on an aliquot of NP dispersion (without cryoprotectant) by weighting the solid residue after freeze-drying. Results are expressed as the ratio of the actual NP weight to the theoretical polymer/drug weight  100. For cell uptake experiments, fluorescent NPs containing NR (NR-NPs) at a theoretical loading of 0.07 mg of NR per 100 mg of loaded NPs were produced analogously. 2.3. Characterization of nanoparticles 2.3.1. Colloidal properties The hydrodynamic diameter (DH) and polydispersity index (PI) of NP dispersion were determined by photon correlation spectroscopy (PCS) at 25  C on 90 angle using a N5 Submicron Particle Size Analyzer (Beckman-Coulter, USA). Zeta potential was determined by analysing a NP dispersion on a Zetasizer Nano Z (Malvern Instruments Ltd., UK). The morphology of NPs was examined by transmission electron microscopy (CM 12 Philips, The Netherlands) using samples stained with a 2% phosphotungstic acid solution. 2.3.2. Drug loading DTX loading inside NPs was assessed by dissolving 1 mg of NPs in 500 mL of acetonitrile under stirring for 1 h. Thereafter, 500 mL of water were added and after stirring for further 1 h, the sample was filtered through a 0.45 mm filter (RC, Chemtek, Italy). DTX was analyzed by HPLC on a Shimadzu apparatus equipped with a LC-10ADvp pump, a SIL-10ADvp autoinjector, a SPD-10Avp UV–vis detector and a C-R6 integrator. The analysis was performed on a Phenosphere-NEXT 5 mm, C18 column (250 mm  4.6 mm, Å) (Phenomenex, USA). The mobile phase was a 40:60 (v/v) mixture of 20 mM phosphate buffer at pH 4.5 and acetonitrile pumped at a flow rate of 1 mL/min. The UV detector was set at 227 nm. A calibration curve for DTX in ethanol was generated in the concentration range 0.980–196 mg/mL. The limits of quantification (LOQ) and detection (LOD) were 1.29 and 0.39 mg/mL. 2.3.3. Colloidal stability The colloidal stability of NPs was evaluated from their resistance to NaCl- and HSA-induced aggregation. A known amount of

G. Palma et al. / International Journal of Pharmaceutics 473 (2014) 55–63

freeze-dried powder was dispersed in different NaCl and HSA solutions at pH 7.4 to reach the desired polymer concentration of 1 mg/mL and incubated at 37  C. After 1, 24, 48 and 72 h of incubation, turbidity was measured at 500 nm on an UV spectrophotometer (Shimadzu UV 1800, Japan). Data reported are the mean value of at least three measurements  standard deviation. 2.4. Interaction of NPs with human plasma Stability of DTX-NPs in human plasma was assessed by different methods, including evaluation of protein interaction on NP surface, turbidity and size measurements. DTX-NPs dispersed in water were incubated with different amounts of plasma (10, 25, 50 and 100 mL) at 37  C to give a total volume of 0.35 mL ([NPs] = 3.5 mg/mL). After 24 h, the samples were centrifuged at 278,000  g for 40 min to pellet the particle– protein complexes. The pellet was suspended in water, transferred to a new vial and centrifuged again; this procedure was repeated three times. Quantification of serum protein adsorbed onto NP surface was performed by the Bradford protein assay. Washed NPs were solubilised in DCM/water 1:1 (400 mL), centrifuged and the aqueous phase collected. Samples were treated with Bradford reagent and incubated for 15 min before reading the absorption of the solution at 595 nm on an UV spectrophotometer (Shimadzu UV 1800, Italy). The absorbance of the samples was measured and compared to a calibration curve where plasma was employed as protein standard. Separation of plasma proteins bound to NP surface was conducted by using 1-D sodium dodecyl sulfate–polyacrylamide gel electrophoresis (SDS–PAGE). Aqueous phase (20 mL) was loaded onto a vertical slab gel consisting of a 10% stacking gel and 12% resolving gel. All gels were run at a constant voltage mode of 90 V for 90 min in a Tris/glycine/SDS buffer. Proteins on the SDS– PAGE gel were stained with EZBlueTM Gel Staining Reagent. As controls, we performed the experiment on human plasma (10 mL) without NPs and on polystyrene NPs of comparable size (positive control). For turbidity and size studies, a known amount of freeze-dried powder (corresponding to 2.5 mg of NPs) was dispersed in 0.5 mL of human plasma and incubated at 37  C. Size measurements and turbidity were taken as reported in 2.3.1 and 2.3.3 after 24, 48 and 72 h of incubation. 2.5. DTX release from nanoparticles In vitro release of DTX from NPs was assessed in 10 mM phosphate buffer saline (PBS) at pH 7.4 containing NaCl (137 mM) and KCl (2.7 mM) by a dialysis method. A known amount of freezedried powder (corresponding to 2 mg of NPs) was dispersed in 0.5 mL of PBS or 0.3 mL of human plasma and placed in a dialysis bag (MWCO = 3500 Da, Spectra/Por1). The sample was plunged in 5 mL of PBS (sink condition) and kept at 37  C up to 72 h. At selected time intervals, 1 mL of release medium was withdrawn and replaced with an equal volume of fresh medium. As control, release profiles of DTX dissolved in EtOH (10 mg/mL) and added to PBS or human plasma were assessed too. DTX quantitative analysis in the release medium was performed as described in Section 2.3.2. The results are expressed as release % over time  SD of three experiments. 2.6. Cell culture studies 2.6.1. Cell cultures Human breast cancer MDA-MB231 cells (American Type Tissue culture Collection, Rockville, Maryland) were grown in DMEM with heat-inactivated 10% FBS, 20 mM HEPES, 100 mg/mL penicillin,

57

100 mg/mL streptomycin, 1% L-glutamine and 1% sodium pyruvate. Cells were grown in a humidified atmosphere of 95% air/5% CO2 at 37  C. 2.6.2. Cell uptake studies Cells were seeded into 24-well plates at a density of 2  105 cells/well in 1 mL of medium. A NR solution (in DMSO 1% v/v) or NR-NPs in PBS (NR concentration = 0.07 mg/mL, NP concentration = 0.1 mg/mL) was incubated with MDA-MB231 cells for 4 h and 24 h at 37  C. Then, the medium was removed and the cells were washed three times with fresh PBS. Fluorescence intensity inside cells was evaluated after cell lysis by a 1% solution of Triton X-100 in PBS. NR was quantified by fluorescence after construction of a calibration curve of NR dissolved in the 1% PBS solution of Triton X-100 (Ex = 546 nm; Em = 590 nm). NR accumulation was normalized per milligram of protein as evaluated by Lowry assay. Percent uptake was calculated as the ratio between protein-normalized amount of NR internalized and NR amount initially added to each well. 2.6.3. Cell proliferation assay and western blot analysis of p53 Analysis of cell proliferation was performed in the presence of free DTX, DTX-NPs and empty NPs on cell lines seeded in 96-well plates at the density of 1–2  103 cells/well in serum-containing media. After 24 h of incubation at 37  C, cells were treated with increasing concentrations of free DTX (in DMSO 1%) and DTX-NPs (1–500 ng/mL of DTX corresponding to 0.01–5 mg/mL of NPs) as well as empty NPs. Cell viability was assessed with MTT method as described previously after 72 h. Cell lysate were prepared by adding ice-cold lysis buffer (0.5% Triton X100 – 50 nmol/Tris pH 7.2, 140 nmol of NaCl, 10 mmol of EDTA) containing the protease inhibitor cocktail Complete Mini (Roche, Germany). Protein samples were mixed with an equal volume of 2% SDS–PAGE sample buffer, boiled for 5 min, and then separated using 10% SDS–PAGE. After electrophoresis, proteins were transferred to PVDF membranes by semi-dry electrophoretic transfer. The membranes were blocked in 5% dry milk, rinsed and then incubated with primary antibody of p53 (Sigma) overnight at 4  C. The primary antibody was removed, membranes were washed four times and followed by HRP-conjugated secondary antibody. Detection was then performed using an enhanced chemoluminescence kit and exposed to X-ray film. Anti a-actin was used as loading control. 2.6.4. Colony formation assay Cells were trypsinized and plated in 6-well dishes at density of 500 cells/well. Cells were allowed to attach overnight and then treated with empty NPs, free DTX and DTX-NPs at DTX concentration of 0.5, 1, 10, 50, 100 ng/mL and then incubated at 37  C. Fourteen days later, the cells were fixed in ethanol and stained with crystal violet at 0.2%. The numbers of colonies were counted and plotted as composite results from three independent experiments. 2.7. In vivo antitumor activity The antitumor effects of DTX-NPs were tested in a heterotopic mouse model of triple-negative breast cancer. The experimental procedures performed in this study (from 01/06/2012 to 01/09/ 2012) followed the specific guidelines of the Italian (No. 116/1992) and European Council law (No. 86/609/CEE) for animal care. All the experiments performed on animal models were in compliance with the guidelines for the Care and Use of Laboratory Animals of the National Cancer Institute – IRCCS – “Fondazione G. Pascale”.

58

G. Palma et al. / International Journal of Pharmaceutics 473 (2014) 55–63

Table 1 Properties of NPs. SD were calculated on three different batches.

HPbCD (NP weight) Yield (%) Mean DH (nm  SD) Polydispersity index Zeta potential (mV  SD) DTX actual loading (mg of drug per 100 mg of powder) Entrapment efficiencyc (%  SD) a b c

3. Results

DTX-NPs

DTX-NPs/CD

– 85 54.0  3a 0.228a 8.8  4.0a 9.49  0.5

10 – 60.5  1.4b 0.217b 10.8  2.1b 0.86  0.03

96.1  1.7

–

NPs dispersed in water. Freeze-dried powder after dispersion in water. Ratio between actual and theoretical loading  100.

Moreover, all the experiments were performed by following the European Directive 63/2010/UE and the Italian Law (DL 116/1992). Mice used in experiments were anesthetized with zolazepam (50 mg/kg i.p.), xylazine (20 mg/kg), and atropine sulfate (0.04 mg/kg). All efforts were made to minimize animal suffering. The animals were euthanized by cervical dislocation when tumor volume was 1500 mm3 (ethic cut-off) and at the end of the study. MDA-MB231 cells (2.5  106) in 0.2 mL PBS were implanted subcutaneously into the right-side flank area in 8-week old female nude mice (Harlan, Italy). When tumors reached 30–60 mm3, the animals were randomized into control and treatment groups (10 animals per group). Mice received one i.v. administration into the caudal vein of NaCl 0.9% (control), freeze-dried powder with empty NPs/CD, a DTX commercial formulation (Taxotere1) and freezedried DTX-NPs/CD at a DTX dose was 10 mg/kg. Powders were reconstituted in saline. The tumor size was measured using calipers, and the tumor volume was estimated by the formula: tumor volume (mm3) = (W  L) 2  1/2, where L is the length and W is the width of the tumor. Normally distributed data were represented as mean  S.E.M. Two-way ANOVA and Bonferroni post-hoc analysis were used to examine the significance of differences among groups (Graph pad Prism 5.0). A probability value with *P < 0.05 and **P < 0.01 was considered to be statistically significant.

A 0.180

1h

24 h

48 h

3.1. Nanoparticle properties MeSo consists in the nanoemulsification of a fluid, non watermiscible PEG–PCL (at a temperature higher than melting temperature) in water under sonication. The copolymer is then hardened by cooling at room temperature giving spherical nonaggregated NPs with a high degree of PEG coating and a hydrophobic polymeric core embedding the drug. Properties of the NP formulations produced by MeSo are reported in Table 1. A hydrodynamic diameter compatible with an i.v. administration, slightly negative zeta potential and satisfactory DTX actual loading were found. After freeze-drying, NPs underwent aggregation upon dispersion in water (aggregates >1500 nm were measured by laser light scattering, data not shown). In order to prevent the collapse of NPs during the freeze-drying step, addition of HPbCD as cryoprotectant at 10:1 mass ratio with NPs was suitable to overcome aggregation. Fast NP dispersion in both water and saline was observed for DTX-NPs/CD, giving size comparable to that of “as prepared” NPs (Table 1). Long term stability of the carrier in buffer at different ionic strength and in the presence of HSA was found (Fig. 1). The onset of a dramatic increase in turbidity was considered as aggregation concentration. The increase of the ionic strength induced a decrease of the PEG steric effect due to a breakdown of hydrogen bonding between PEG ether oxygens and water molecules and a subsequent dehydration of the outer shell. The spherical morphology and the absence of aggregation phenomena of NPs in the presence of HPbCD was also evidenced by TEM images of DTX-NPs/CD after dispersion in water (Fig. 2A). Furthermore, NP size was in line with that obtained by PCS. In order to identify and quantify eventual proteins attracted on NP surface, different amounts of human plasma were mixed with NPs (3.5 mg/mL). After 24 h of incubation and extensive washing, no proteins were detected by both Bradford assay (data not shown) and as band in the gel (Fig. 2B), independently by the initial amount of plasma mixed with NPs. These results were also in line with the fact that no weight change of NPs after incubation with human plasma, washing and freeze-drying was observed (Supplementary material, Fig. S1).

B 0.200

72 h

1h

24 h

48 h

72 h

0.180 0.160

ABS

ABS

0.160

0.140

0.140

0.120 0.120

0.100 0.100 0

30

60 90 NaCl (mM)

120

150

0.080 0

5

10 15 HSA (mg/ml)

20

25

Fig. 1. Turbidity of freeze-dried DTX-NPs/CD (1 mg/mL) reconstituted in a NaCl (panel A) and in a HSA solution (panel B) at different concentrations. Optical density of the samples at 500 nm was monitored. Data are reported as mean of three independent experiments  SD.

G. Palma et al. / International Journal of Pharmaceutics 473 (2014) 55–63

59

Fig. 2. (A) TEM image of DTX-NPs/CD; (B) protein adsorption on NPs evaluated by SDS–PAGE. Lane 1, human plasma (10 mL); lane 2, polystyrene NPs treated with human plasma (10 mL); lanes 3–6, DTX-NPs (3.5 mg/mL) treated with different amounts of human plasma (10–25–50–100 mL); (C) Hydrodynamic diameter and turbidity analysis of DTX-NPs/CD incubated in human plasma (5 mg/mL) up to 72 h. Data are reported as mean of three independent experiments  SD. (D) Release profile of DTX from DTX-NPs in 10 mM phosphate buffer at pH 7.4 (PBS) or in human plasma at 37  C. Release profile of free DTX in the same conditions is reported as comparison. Data are reported as mean of three independent experiments  SD.

3.2. Cell studies MDA-MB431 triple-negative breast cancer cells were treated with NPs tagged with NR and the percent uptake evaluated as the

ratio between protein-normalized amount of NR internalized and NR amount initially added to cells (Fig. 3). It is worth of note that 0.0020 free NR NR-NPs 0.0015

Uptake (%)

Possible aggregation of NPs in human plasma up to 72 h was monitored through turbidity (UV absorption at 500 nm) and size measurments. As evidenced in Fig. 2C, DTX-NPs/CD displayed a good stability in plasma up to 24 h, followed by a slight tendency to size increase after 48 and 72 h of incubation. In spite of this behavior, DTX-NP size remained always lower than 150 nm, which is still compatible with NP circulation in the body. Macroscopic aggregation was not observed in all the tested samples. The stability of NPs was demonstrated by unchanged particle size and constant turbidity of the dispersion. Release profiles of DTX from DTX-NPs at 37  C, in both PBS at pH 7.4 and human plasma are reported in Fig. 2D and compared with those of the free drug. Results showed a sustained release of DTX from DTX-NPs in both PBS and plasma, highlighting also the stability of DTX-NPs in biologically-relevant conditions. Nevertheless, release of free DTX was very different according to the suspending medium inside the dialysis bag. In fact, while free DTX was quantitatively released from a DTX dispersion in PBS after 6 h, a sustained release was evident after DTX dispersion in human plasma.

0.0010

0.0005

0.0000 4

time (hours)

24

Fig. 3. Percent NP uptake in MDA-MB431 cells compared to free NR after 4 and 24 h of incubation at 37  C. Key legend: nile Red (NR), NR-loaded NPs (NR-NPs). Data are reported as mean of three independent experiments  SD.

60

G. Palma et al. / International Journal of Pharmaceutics 473 (2014) 55–63

increasing DTX-NPs concentrations to inhibit the migration of cancer cells (Supplementary material S3). A prolonged exposure (14 days) of cells at concentration ranging between 5 and 1000 ng/mL of NPs (corresponding to DTX concentrations in the range 0.5–100 ng/mL) showed that empty NPs were not toxic, with only a slight reduction of cell proliferation comparable to control (Fig. 3B). Conversely, cell ability to produce colony was reduced by free DTX and DTX-NPs treatment on continuous exposure for 14 days. The number of colonies formed at each concentration between DTX and DTX-NPs was not statistically different. In western blot analysis, p53 expression level was higher in DTX-NPs and free DTX in a dose dependent manner as compared to the control (Fig. 5). 3.3. In vivo studies In vivo activity of DTX-NPs was evaluated in a heterotopic mouse model of TNBC (Fig. 6). Mice received a single i.v. injection of saline (CTR), empty NPs/CD, free DTX and DTX-NPs/CD at DTX dose of 10 mg/kg. At 48 days post-injection, tumour growth in animals that received free DTX was significantly slower than that in the control group (P < 0.05). DTX-NPs/CD displayed tumour growth inhibition similar to that of free DTX at the same drug dose. Group treated with empty NPs did not show statistically different effects as compared with the control group. Variation in body weight of nude mice in DTX-NPs/CD group as other groups was not found (data not show). Survival of animals closely followed the tumour growth profile. Kaplan–Meier results showed an increased median survival for DTX-NPs treated mice with respect to the control groups (CTR and empty NPs, P = 0.008 determined using log-rank test) as well as to free DTX group at 48 days. 4. Discussion TNBC is a heterogeneous disease including different orphan breast cancers simply defined by the absence of ER/PR/HER-2. Fig. 4. (A) Growth of MDA-MB431 cells treated with free DTX (DMSO 1%), DTX- NPs and empty NPs for 72 h as evaluated by the MTT assay and expressed as percentage of untreated cells; (B) MDA-MB431 colony formation after incubation with different concentration of free DTX, DTX-NPs and empty NPs for 14 days. Data are reported as mean of three independent experiments. Each point is the mean of four measurements  SD.

NR-labeled NPs revealed a remarkable increase in cell internalization as compared to free NR, especially after 24 h of treatment, suggesting that the extent of NP uptake in this cell line is a timedependent process. Nevertheless, the extent of NP uptake was very low as expected for a PEGylated system. In vitro cytotoxicity of free DTX or DTX-NPs was evaluated by incubating MDA-MB231 with increasing concentrations of free DTX (1–500 ng/mL), DTX-NPs and empty NPs (10–5000 ng/mL corresponding to DTX concentrations in the range 1–500 ng/mL) for 72 h (Fig. 4A). Cell growth was then assessed by using the MTT assay as described in the Section 2. Empty NPs did not induce significant cell growth inhibition at each concentration and time interval tested. Free DTX caused a significant cell growth inhibition after 72 h of treatment in a range from 10 to 30%, and dose/effect curves of both free-DTX and DTX-NPs showed a clear dosedependent effect. Furthermore, DTX-NPs induced cell growth inhibition not statistically different when compared to free DTX (P = 0.20). Nevertheless, the obtained data suggest that DTX-NPs caused a more effective growth inhibition at low DTX doses (from 10 to 100 ng/mL) as compared to free DTX. The overall data were in line with cell viability results (Supplementary material S2). Furthermore, wound healing assay demonstrated the capacity of

Fig. 5. Protein quantification of p53 expression in MDA-MB231 cells by western blotting after 72 h incubation. (A) p53 and b-actin as loading control expression: bar 1: control; bar 2–5–8: empty NPs [NPs] = 10–50–250 ng/mL; bar 3–6–9: free DTX [DTX] = 1–50–250 ng/mL; bar 4–7–10: DTX-NPs [DTX] = 0.01–0.5–2.5 mg/mL); (B) densitometry of p-53 protein content corrected for loading with actin (n = 16 vs 16). p53 expression in the DTX-NPs showed a statistical trend toward an increase (*P < 0.1).

G. Palma et al. / International Journal of Pharmaceutics 473 (2014) 55–63

Fig. 6. (A) Tumor growth inhibition and (B) Kaplan–Meier survival plot of female athymic Nu/Nu nude mice bearing subcutaneous MDA-MB231 tumor. Mice received an i.v. injection of saline (CTR), empty NPs (NPs), DTX commercial formulation (Taxotere1) and DTX-NPs/CD NP samples were dispersed in saline. DTX dose was 10 mg/kg. *P < 0.05 DTX vs CTR at 47 days post injection; **P < 0.01 DTX-NPs vs CTR at 47 days post injection.

TNBC represents 10–20% of all mammary tumours and is characterized by its unique molecular profile, aggressive behavior, distinct patterns of metastasis, and lack of targeted therapy. It is often found in younger women and has been associated with poor prognosis, due to aggressive tumor phenotype(s), early metastasis to visceral organ or brain after chemotherapy. Several studies have demonstrated that TNBC has significantly higher pCR rate compared to hormone receptor positive breast cancer when treated with neoadjuvant chemotherapy with taxanes and anthracycline. The NSABP trial randomized 2400 women to one of three arms to evaluate the response to neoadjuvant therapy and long term outcomes (36). The administration of preoperative DTX nearly doubled the pCR rate from 13% to 21%. Interestingly, subgroup analysis showed that the pCR rate nearly doubled with the addition of DTX for both ER+ and ER- tumors, from 6–14% to 14–23%, respectively (Rastogi et al., 2008; Wu et al., 2011). In order to optimize the antitumor profile of DTX as well as to ameliorate its toxicity profile, we designed an alternative and safe DTX formulation based on NPs suitable for the i.v. therapy of TNBC with DTX. NPs were designed to be long-circulating and to reach passively solid tumours through EPR effect. In details, we have developed an injectable powder made up of biodegradable DTXNPs dispersed in a HPbCD carrier to be suspended in saline before administration. DTX-NPs were prepared by an in-house melting sonication technique on the basis of an our previous study demonstrating a better PEG coverage on the NP surface as compared to classical nanoprecipitation technique (Quaglia et al., 2006) and an improved tolerability profile of DTX-NPs as compared to free DTX (Ungaro et al., 2012b). In this study, we devoted our effort to translational aspects focusing on characterization of NPs from a biological standpoint in simulated in vitro conditions, in order to fully understand their in vivo therapeutic potential. In fact, a critical issue to consider after

61

NP i.v. injection is their possible interaction with plasma proteins and blood components that can strongly modify the biodistribution as well as the pharmacological effects of a nanomedicine (d'Angelo et al., 2014; Saptarshi et al., 2013). To this respect, it has been shown that NP surface and size need to be engineered in order to prevent opsonisation, which is responsible of clearance mechanisms overall affecting circulation time (d'Angelo et al., 2014; Owens and Peppas, 2006). The low adsorption of human plasma proteins on the surface of DTX-NPs indicated that the presence of a hydrophilic PEG fringe confers protein-repelling properties while allowing effective NP dispersion also in proteincontaining complex media. Nevertheless, their sustained release of DTX in human plasma suggested the integrity of DTX-NPs in simulated biological conditions and highlighted the higher stability of tailored PCL-based nanocarriers as compared to similar NPs made of a PLLA core (Letchford and Burt, 2012; Savic et al., 2006b). It is equally noted that free DTX was slowly released in the presence of serum proteins, highlighting the formation of strong plasma protein–drug binding which can in turn affect biodistribution and activity in the body. On the basis of these evidences, we can speculate that the occurrence of an equilibrium in the blood circulation between free DTX and DTX bound to plasma proteins, especially albumin, contributes to drug accumulation in cancer tissues via specific interactions with gp60 (Gradishar, 2006). In vitro cell inhibition studies of DTX-NPs in MDA-MB231 cells showed that unloaded NPs are highly biocompatible and do not exert any cytotoxicity also after 15 days of cell exposure. The fact that DTX-NPs show cytotoxicity similar to that of free DTX at the same dose is in line with the previous findings on larger PEG–PCL micelles tested in hepatic cancer cell lines (Liu et al., 2012) although it is surprising considering that NPs are releasing only a fraction of the entrapped drug. It can be speculated that this effect is related to the mechanism of DTX entry inside cancer cells. Indeed, DTX fraction released from DTX-NPs outside cells can bind to albumin which is taken up by cells in growing tumor tissue (Frei, 2011; Joshi et al., 2013), and in part passively diffuse inside cells. On the other hand, NP-uptake inside cells can be considered a supplementary pathway to transport drug. In the case of free DTX, its incubation with cells and interaction with medium proteins may induce a passive diffusion and receptor-mediated endocytosis via albumin binding. The overall result is that asides from the mode of entry of DTX in cancer cells, similar cytotoxic effects between free DTX and DTX-NPs occur. Thus, internalization mechanism, interaction with proteins and release properties contribute all to cytotoxicity of DTX at very low released doses. Finally, it is demonstrated that the inhibitory effect of free DTX and DTX-NPs on cancer cell growth can be still ascribed to an induction of concentration-dependent apoptosis. A great concern in developing i.v. formulations for NPs, refers to their poor dispersibility in a suitable medium to allow administration. Despite of the presence of a hydrophilic coating, PEGylated NPs underwent collapse under freeze-drying conditions thus demanding stabilization strategies. It was recently demonstrated that HPbCD act as a cryoprotectant for PCL–PEG–PCL NPs allowing to redisperse micelles up to 6% in water and giving poorly hygroscopic powders (Moretton et al., 2012). Remarkably, DTX-NPs were readily dispersed in saline up to around 100 mg/mL when HPbCD, an excipient already approved in marketed i.v. formulations (Stella and He, 2008), was employed as cryoprotectant. Indeed, it has been recently demonstrated that HPbCD is widely compatible with DTX and can represent a new vehicle for its parenteral administration (Kim et al., 2012) as demonstrated also on the basis of hemolysis results (Supplementary material S4). We can thereby conclude that the i.v. administration of DTX-NPs in a HPbCD carrier may be well tolerated also in humans due to high stability in biological fluids, and at the same time strong limitation

62

G. Palma et al. / International Journal of Pharmaceutics 473 (2014) 55–63

of haemolytic events connected to the conventional therapy based on Taxotere1 (Baker et al., 2009). Analogously to results collected in cell lines, treatment with DTX-NPs at 10 mg/kg of DTX in a heterotopic mouse model of TNBC induced an inhibition of tumor growth not statistically different to free DTX while maintaining a mild toxicity profile. These results are in line with a previous study on Nanoxel-PMTM (DTX delivered in PEG–PLLA micelles) where comparable activity of Taxotere1 and micellar formulation were found in athymic nude mice bearing human lung cancer after 3 consecutive days treatment at a higher DTX dose (Lee et al., 2011). Nevertheless, sustained release effects of the developed formulation after a single administration shows as a major outcome an improved animal survival rather than overall tumor growth. To further highlight this benefit we plan to extend the study to an orthotopic mouse model of breast cancer which takes into account also stroma interaction with tumor cells to emphasize obtained data show if sustained release reduce the tumor growth and metastasis formation in mouse model. As alleviation of side effects of drug-loaded NPs is one of the main concerns when developing novel drug-delivery vehicles, the low toxicity profile associated to preservation of antitumor effects and increase in mice survival should be considered as an excellent basis to translate this formulation to the clinic. 5. Conclusion Core–shell PEG-b-PCL NPs associated to a cyclodextrin carrier were found to give a powder product successfully delivering docetaxel in TNBC models. DTX-loaded NPs prepared by a melting sonication technique displayed excellent stability in plasma and undetectable surface absorption of plasma proteins. DTX delivered from NPs strongly associated with albumin, suggesting multiple mode of entry of the drug inside cancer cells. As an overall result, cytotoxicity of DTX in DMSO and loaded inside NPs were found to be not statistically different. Nevertheless, in vivo activity in a mice xenograft highlighted that DTX-loaded NPs increase significantly mice survival as compared to a commercial DTX product. On the basis of these data, the formulation developed here turns to be of high potential for translation in humans. Acknowledgment The financial support of Italian Ministry of Education, University and Research (PRIN2010H834LS) is gratefully acknowledged. Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at http://dx.doi.org/10.1016/j.ijpharm. 2014.06.058. References Baker, J., Ajani, J., Scotte, F., Winther, D., Martin, M., Aapro, M.S., von, M.G., 2009. Docetaxel-related side effects and their management. Eur. J. Oncol. Nurs. 13, 49–59. Bauer, K.R., Brown, M., Cress, R.D., Parise, C.A., Caggiano, V., 2007. Descriptive analysis of estrogen receptor (ER)-negative, progesterone receptor (PR)negative, and HER2-negative invasive breast cancer, the so-called triplenegative phenotype: a population-based study from the California cancer registry. Cancer 109, 1721–1728. Chen, H., Kim, S., He, W., Wang, H., Low, P.S., Park, K., Cheng, J.X., 2008. Fast release of lipophilic agents from circulating PEG–PDLLA micelles revealed by in vivo forster resonance energy transfer imaging. Langmuir 24, 5213–5217. Conte, C., Ungaro, F., Maglio, G., Tirino, P., Siracusano, G., Sciortino, M.T., Leone, N., Palma, G., Barbieri, A., Arra, C., Mazzaglia, A., Quaglia, F., 2013. Biodegradable core–shell nanoassemblies for the delivery of docetaxel and Zn(II)-phthalocyanine inspired by combination therapy for cancer. J. Control. Release 167, 40–52.

Conte, C., d'Angelo, I., Miro, A., Ungaro, F., Quaglia, F., 2014. PEGylated polyesterbased nanoncologicals. Curr. Top. Med. Chem. 14, 1097–1114. d'Angelo, I., Conte, C., Miro, A., Quaglia, F., Ungaro, F., 2014. Core–shell nanocarriers for cancer. Part I: biologically-oriented design rules. Expert Opin. Drug Deliv. 11, 283–297. Danhier, F., Feron, O., Preat, V., 2010. To exploit the tumor microenvironment: passive and active tumor targeting of nanocarriers for anti-cancer drug delivery. J. Control. Release 148, 135–146. Frei, E., 2011. Albumin binding ligands and albumin conjugate uptake by cancer cells. Diabetol. Metab. Syndr. 3, 11. Fumoleau, P., 1997. Efficacy and safety of docetaxel in clinical trials. Am. J. Health Syst. Pharm. 54, S19–S24. Gradishar, W.J., 2006. Albumin-bound paclitaxel: a next-generation taxane. Expert Opin. Pharmacother. 7, 1041–1053. Hirsjarvi, S., Passirani, C., Benoit, J.P., 2011. Passive and active tumour targeting with nanocarriers. Curr. Drug Discov. Technol. 8, 188–196. Isakoff, S.J., 2010. Triple-negative breast cancer: role of specific chemotherapy agents. Cancer J. 16, 53–61. Jabir, N.R., Tabrez, S., Ashraf, G.M., Shakil, S., Damanhouri, G.A., Kamal, M.A., 2012. Nanotechnology-based approaches in anticancer research. Int. J. Nanomed. 7, 4391–4408. Joshi, M.R., Yao, N., Myers, K.A., Li, Z., 2013. Human serum albumin and p53activating peptide fusion protein is able to promote apoptosis and deliver fatty acid-modified molecules. PLoS One 8, e80926. Kim, D.W., Kim, S.Y., Kim, H.K., Kim, S.W., Shin, S.W., Kim, J.S., Park, K., Lee, M.Y., Heo, D.S., 2007. Multicenter phase II trial of Genexol-PM, a novel Cremophor-free, polymeric micelle formulation of paclitaxel, with cisplatin in patients with advanced non-small-cell lung cancer. Ann. Oncol. 18, 2009–2014. Kim, T.K., Yoo, H.H., Kim, E.J., Lee, B.Y., Park, J.H., 2012. Pharmacokinetic equivalence of Taxotere and SID530, a novel docetaxel formulation containing hydroxypropyl-beta-cyclodextrin in monkeys. Arzneimittelforschung 62, 280–284. Lammers, T., Kiessling, F., Hennink, W.E., Storm, G., 2012. Drug targeting to tumors: principles, pitfalls and (pre-) clinical progress. J. Control. Release 161, 175–187. Lee, K.S., Chung, H.C., Im, S.A., Park, Y.H., Kim, C.S., Kim, S.B., Rha, S.Y., Lee, M.Y., Ro, J., 2008. Multicenter phase II trial of Genexol-PM, a Cremophor-free, polymeric micelle formulation of paclitaxel, in patients with metastatic breast cancer. Breast Cancer Res. Treat. 108, 241–250. Lee, S.W., Yun, M.H., Jeong, S.W., In, C.H., Kim, J.Y., Seo, M.H., Pai, C.M., Kim, S.O., 2011. Development of docetaxel-loaded intravenous formulation, Nanoxel-PM using polymer-based delivery system. J. Control. Release 155, 262–271. Letchford, K., Burt, H.M., 2012. Copolymer micelles and nanospheres with different in vitro stability demonstrate similar paclitaxel pharmacokinetics. Mol. Pharm. 9, 248–260. Liedtke, C., Mazouni, C., Hess, K.R., Andre, F., Tordai, A., Mejia, J.A., Symmans, W.F., Gonzalez-Angulo, A.M., Hennessy, B., Green, M., Cristofanilli, M., Hortobagyi, G. N., Pusztai, L., 2008. Response to neoadjuvant therapy and long-term survival in patients with triple-negative breast cancer. J. Clin. Oncol. 26, 1275–1281. Lim, W.T., Tan, E.H., Toh, C.K., Hee, S.W., Leong, S.S., Ang, P.C., Wong, N.S., Chowbay, B., 2010. Phase I pharmacokinetic study of a weekly liposomal paclitaxel formulation (Genexol-PM) in patients with solid tumors. Ann. Oncol. 21, 382–388. Liu, Q., Li, R., Zhu, Z., Qian, X., Guan, W., Yu, L., Yang, M., Jiang, X., Liu, B., 2012. Enhanced antitumor efficacy, biodistribution and penetration of docetaxelloaded biodegradable nanoparticles. Int. J. Pharm. 430, 350–358. Maeda, H., Nakamura, H., Fang, J., 2013. The EPR effect for macromolecular drug delivery to solid tumors: improvement of tumor uptake, lowering of systemic toxicity, and distinct tumor imaging in vivo. Adv. Drug Deliv. Rev. 65, 71–79. Mikhail, A.S., Allen, C., 2009. Block copolymer micelles for delivery of cancer therapy: transport at the whole body, tissue and cellular levels. J. Control. Release 138, 214–223. Moretton, M.A., Chiappetta, D.A., Sosnik, A., 2012. Cryoprotection-lyophilization and physical stabilization of rifampicin-loaded flower-like polymeric micelles. J. R. Soc. Interface 9, 487–502. Owens III, D.E., Peppas, N.A., 2006. Opsonization, biodistribution, and pharmacokinetics of polymeric nanoparticles. Int. J. Pharm. 307, 93–102. Pasquier, E., Kavallaris, M., Andre, N., 2010. Metronomic chemotherapy: new rationale for new directions. Nat. Rev. Clin. Oncol. 7, 455–465. Quaglia, F., Ostacolo, L., De, R.G., La Rotonda, M.I., Ammendola, M., Nese, G., Maglio, G., Palumbo, R., Vauthier, C., 2006. Nanoscopic core–shell drug carriers made of amphiphilic triblock and star-diblock copolymers. Int. J. Pharm. 324, 56–66. Quaglia, F., Ostacolo, L., Nese, G., Canciello, M., De, R.G., Ungaro, F., Palumbo, R., La Rotonda, M.I., Maglio, G., 2008. Micelles based on amphiphilic PCL–PEO triblock and star-shaped diblock copolymers: potential in drug delivery applications. J. Biomed. Mater. Res. A 87, 563–574. Rastogi, P., Anderson, S.J., Bear, H.D., Geyer, C.E., Kahlenberg, M.S., Robidoux, A., Margolese, R.G., Hoehn, J.L., Vogel, V.G., Dakhil, S.R., Tamkus, D., King, K.M., Pajon, E.R., Wright, M.J., Robert, J., Paik, S., Mamounas, E.P., Wolmark, N., 2008. Preoperative chemotherapy: updates of National Surgical Adjuvant Breast and Bowel Project Protocols B-18 and B-27. J. Clin. Oncol. 26, 778–785. Saptarshi, S.R., Duschl, A., Lopata, A.L., 2013. Interaction of nanoparticles with proteins: relation to bio-reactivity of the nanoparticle. J. Nanobiotechnol. 11, 26. Savic, R., Azzam, T., Eisenberg, A., Maysinger, D., 2006a. Assessment of the integrity of poly(caprolactone)-b-poly(ethylene oxide) micelles under biological conditions: a fluorogenic-based approach. Langmuir 22, 3570–3578.

G. Palma et al. / International Journal of Pharmaceutics 473 (2014) 55–63 Savic, R., Azzam, T., Eisenberg, A., Maysinger, D., 2006b. Assessment of the integrity of poly(caprolactone)-b-poly(ethylene oxide) micelles under biological conditions: a fluorogenic-based approach. Langmuir 22, 3570– 3578. Shahin, M., Lavasanifar, A., 2010. Novel self-associating poly(ethylene oxide)-b-poly (epsilon-caprolactone) based drug conjugates and nano-containers for paclitaxel delivery. Int. J. Pharm. 389, 213–222. Sorlie, T., Tibshirani, R., Parker, J., Hastie, T., Marron, J.S., Nobel, A., Deng, S., Johnsen, H., Pesich, R., Geisler, S., Demeter, J., Perou, C.M., Lonning, P.E., Brown, P.O., Borresen-Dale, A.L., Botstein, D., 2003. Repeated observation of breast tumor subtypes in independent gene expression data sets. Proc. Natl. Acad. Sci. U. S. A. 100, 8418–8423. Stella, V.J., He, Q., 2008. Cyclodextrins. Toxicol. Pathol. 36, 30–42. Ungaro, F., Conte, C., Ostacolo, L., Maglio, G., Barbieri, A., Arra, C., Misso, G., Abbruzzese, A., Caraglia, M., Quaglia, F., 2012a. Core–shell biodegradable nanoassemblies for the passive targeting of docetaxel: features, antiproliferative activity and in vivo toxicity. Nanomedicine 8, 637–646.

63

Ungaro, F., Conte, C., Ostacolo, L., Maglio, G., Barbieri, A., Arra, C., Misso, G., Abbruzzese, A., Caraglia, M., Quaglia, F., 2012b. Core–shell biodegradable nanoassemblies for the passive targeting of docetaxel: features, antiproliferative activity and in vivo toxicity. Nanomedicine 8, 637–646. Wu, J., Li, S., Jia, W., Su, F., 2011. Response and prognosis of taxanes and anthracyclines neoadjuvant chemotherapy in patients with triple-negative breast cancer. J. Cancer Res. Clin. Oncol. 137, 1505–1510. Yagata, H., Kajiura, Y., Yamauchi, H., 2011. Current strategy for triple-negative breast cancer: appropriate combination of surgery, radiation, and chemotherapy. Breast Cancer 18, 165–173. Yang, J., Lee, E., Ku, M., Huh, Y.M., Cheong, J.H., 2013. Functional nanoplatforms for enhancement of chemotherapeutic index. Anticancer Agents Med. Chem. 13, 212–221. Zhang, L., He, Y., Ma, G., Song, C., Sun, H., 2012. Paclitaxel-loaded polymeric micelles based on poly(varepsilon-caprolactone)–poly(ethylene glycol)–poly(varepsilon-caprolactone) triblock copolymers: in vitro and in vivo evaluation. Nanomedicine 8, 925–934.