Biomaterials 32 (2011) 6381e6388 Contents lists available at ScienceDirect Biomaterials journal homepage: www.elsevier.com/locate/biomaterials Characterization of the pattern of ischemic stroke induced by artificial particle embolization in the rat brain Ming-Jun Tsai a, Yi-Hung Tsai b, **, Yu-Min Kuo a, c, * a Institute of Basic Medical Sciences, College of Medicine, National Cheng Kung University, Tainan, Taiwan Graduate Institute of Clinical Pharmacy, College of Pharmacy, Kaohsiung Medical University, Kaohsiung, Taiwan c Department of Cell Biology and Anatomy, College of Medicine, National Cheng Kung University, Tainan, Taiwan b a r t i c l e i n f o a b s t r a c t Article history: Received 26 March 2011 Accepted 16 May 2011 Available online 12 June 2011 Embolism is responsible for half of cerebral infarctions, yet few animal models were developed due to the unpredictability of the embolus-induced infarcts. We manufactured artificial embolic particles by blending chitin and poly(D,L-Lactide-co-glycolide) (chitin/PLGA) for their good biocompatibility and rapid hydration expansion property. We subdivided the chitin/PLGA microparticles into 10 size groups (from 38e45 mm to 255e350 mm) and injected them through the external carotid artery toward the bifurcation of the common carotid artery in the rat. Reduced blood flow of the ipsilateral hemisphere was evident immediately after the injection of particles. The spherical appearance of the particle was critical in occluding the cerebral vessels. Particle212e250mm produced the greatest diffuse infarction in the ipsilateral hemisphere, including the cortex, hippocampus, basal ganglion, thalamus, midbrain and cerebellum. Particle75-90mm induced single or sparse isolated infarcts mainly located in the subcortical region, resembling lacunar stroke observed in humans. Particle38e45mm frequently crossed to the contralateral hemisphere and induced diffuse infarctions in both hemispheres. The cortex infarct volumes were positively correlated to neurologic score and seizure incidence. In conclusion, we have established embolic stroke animal models, including a novel model that mainly expresses lacunar infarction, by intravenous injection of chitin/PLGA microparticles. Ó 2011 Elsevier Ltd. All rights reserved. Keywords: Microsphere Chitin Embolization Animal model 1. Introduction Ischemic stroke accounts for at least 80% of all cerebrovascular diseases [1]. There is still a lack of effective therapy for most patients who suffer from ischemic stroke. Various animal models of ischemic stroke have been designed for the development of new drugs. However, most of the developed agents, proven efficacious in animal models, failed in human clinical trials [2]. It has been argued that although damages could be induced to neurons by directly occluding extracranial or large cerebral arteries, these models mimic more to the human condition of shock than stroke [3]. Embolism is responsible for at least 20% of all stroke and half of cerebral infarctions [4]. There is evidence that any type of stroke can * Corresponding author. Department of Cell Biology and Anatomy, National Cheng Kung University, 1 Ta Hsueh Road, Tainan 70101, Taiwan. ** Corresponding author. Graduate Institute of Clinical Pharmacy, Kaohsiung Medical University, 100 Shih-chuan 1st Road, Kaohsiung, Taiwan. E-mail addresses: yhtsai@kmu.edu.tw (Y.-H. Tsai), kuoym@mail.ncku.edu.tw (Y.-M. Kuo). 0142-9612/$ e see front matter Ó 2011 Elsevier Ltd. All rights reserved. doi:10.1016/j.biomaterials.2011.05.051 be induced by embolism. For example, lacunar stroke, which accounts for about 25% of all ischemic strokes and still lacks any empirical experimental model, can be induced by embolism [5]. Tissue plasminogen activator, the only effective drug for acute stroke, was initially developed using a rat model for embolic stroke [6]. Despite embolic stroke is an important treatable cause of stroke, comparatively little research has focused on the development of animal model of embolic stroke; animal models of embolic stroke at best accounts for 10% of all animal models of ischemic stroke [7]. However, the reproducibility rates of current embolic stroke animal models were relatively low as compared with those of animal models of diffuse infarction. Therefore, a reliable model of embolic stroke, especially the lacunar stroke, is desirable for the future development of treatment agents or procedures. Recently, synthetic biodegradable polyesters such as poly(D,LLactide-co-glycolide) (PLGA) have attracted much attention in the use for drug-delivery [8e12]. One attempt to control the degradability of PLGA is to mix the polyesters with other biodegradable polymers. Chitin is a natural polysaccharide and useful material for biomedical application [13]. A chitin/PLGA blend microsphere has been developed for a special drug-delivery system [14,15]. The 6382 M.-J. Tsai et al. / Biomaterials 32 (2011) 6381e6388 chitin/PLGA blend microspheres are the phase-separated microspheres consisting of numerous PLGA particulates dispersed in the chitin matrix. The biocompatibility, biodegradability and hydration expansion properties of the chitin/PLGA blend microspheres make them ideal emboli. The purpose of this study is to develop a rat model of embolic stroke using the chitin/PLGA blend microspheres. The chitin/PLGA particles were prepared and subdivided by gradation sieving into various size groups ranging from <38 mm to >355 mm. The ranges of the particle size were narrowed as much as possible. The effects of the size, amount and injection location of the chitin/PLGA particles on the pattern and severity of stroke induction were evaluated. 2. Materials and methods 2.1. Materials PLGA 50/50 (MW: 40 kDa, Lactide/glicolide ratio: 50/50) and IR-780 iodide were purchased from SigmaeAldrich Co. (St. Louis, MO). Tetrazolium Red (2.3.5triphenyltetrazolium chloride, TTC) was obtained from Alfa Aesar Co. (Ward Hill, MA). Chitin was purchased from Tokyo Chemical Industry Co. (Tokyo, Japan). All other chemicals and solvents were of analytical grade and purchased from Sigmae Aldrich Co. unless other specified. 2.2. Preparation of chitin/PLGA mixed solution Chitin solution 1% (W/V) was prepared by suspending chitin powder in dimethylacetamide (DMAC) solution containing 5% (W/V) LiCl [14]. The chitin/DMAC-LiCl mixed suspension was stirred and refluxed at 130  C to dissolve the chitin powder, until a brown solution was obtained. The chitin/PLGA 50/50 mixed solution was prepared by directly dissolving the PLGA 50/50 powder in the prepared chitin solution. The blending ratio of chitin: PLGA 50/50 finally prepared solution was 1:1. particle size analyzer were performed in every lot of particles to ensure the range of particle size. Furthermore, the swelling ratio in water was also evaluated in each group. 2.6. Animal preparation The animal experiment protocol was reviewed and approved by the Institutional Animal Care and Use Committee of Kaohsiung Medical University. All animals were given with free food and water intake and housed in a temperature and light controlled animal care facility. Male Wister albino rats (3-month-old, 300e350 g) were starved overnight before anesthetized by an intra-peritoneal injection of 300 mg/kg chloral hydrate. During operation, the body temperature of the rat was maintained at 37  C by the automated temperature regular system. A rectal thermometer was used to monitor the body temperature. The rat was fixed in the supine position on an operation plate and midline excision of ventral neck was made to explore the bifurcation of right common carotid artery (CCA) and right external carotid artery (ECA). The PE-50 catheter (Becton Dickinson, Franklin Lakes, NJ) was either inserted from the right ECA to the bifurcation of the right CCA or from the right CCA to the right internal carotid artery (ICA). 2.7. Laser Doppler imaging for mapping cerebral blood flow To measure regional cerebral blood flow, a scanning laser Doppler imaging (Moor Instruments Ltd., Axminster, UK) was used to map the dorsal surface of the skull 2 mm caudal and 2.5 mm lateral to bregma. 2.8. Neurologic test All rats received neurologic test at 2, 4, 6, 8 and 24 h after operation. We graded the behavior of the rats using the following criteria: 0 e no neurologic defects, 1 e one paw clumsiness, 2 e tilt, 3 e rounding only a unilateral circle, 4 e akinesia, 5 e seizure, 6 e stupor or lack of any spontaneous movement, 7 e death. The rats died within 3 h after the operation were withdrew from the study. All neurologic behaviors were evaluated by the same investigator. 2.9. Calculation of infarction volume 2.3. Preparation of chitin/PLGA microparticles The wet phase inversion by dropping method was used to produce the chitin/ PLGA microparticles as described previously with minor modifications [14]. To prepare microparticles, the chitin/PLGA mixed solution was kept at 70  C and dropped through a syringe (27 gauge) into 1% sodium lauryl sulfate water bath. The temperature of water bath was kept at 25  C, which provided a sink of coagulation for completely replacing of DMAc-LiCl solution from the chitin/PLGA blend droplets. The gelled microparticles were allowed to harden in the cool water bath (25  C) for 12h. After hardening, the microparticles were filtered, rinsed with deionized water and air dried overnight before sieving. We used U.S. Standard Sieve Series (size mesh from 40 to 400 mesh, Analytical Test Sieve, Retsch, Germany) to sieve the particles. The sieves were stacked on the sieve shaker (Ro-Tap Sieve Shaker, Laval Lab Inc., Canada) and the particles were placed on the top sieve. The whole nest of sieves was shaken for 30 min. To trace the location of embolization, we prepared chitin/PLGA microparticles loaded with fluorescent dyes IR-780. IR-780 iodide was dissolved in the chitin/PLGA 50/50 mixed solution and the microparticles were prepared as aforementioned. The IR-780 microparticles were rinsed, air dried overnight and submitted to size sieving as described above. We used 2,3,5-triphenyltetrazolium chloride (TTC) staining to measure infarct size. After deep anesthesia, the rat brains were removed and positioned in a rat brain matrix (Activational Systems, Ann Arbor, MI). For those rats died before the end of evaluation, their brains were removed within 30 min. Each brain was cut into 12 coronal sections, 2 mm thick. The sections were incubated in a 0.05% TTC solution for 30 min at 30  C before fixed in 4% buffered formaldehyde solution for 24 h. The infarct areas were traced and quantified by an image analysis system (ImageJ v. 1.33u, NIH, USA). The infarct areas of all sections were summed and multiplied by the section thickness to get the total infarct volume. 2.10. Statistical analysis All data were expressed as mean  SEM. Statistical analyses of more than two groups were performed with the one-way analysis of variances (ANOVAs) followed by post-hoc NewmaneKeuls Multiple Comparisons if appropriate (p < 0.05). The Chi-square test was used to compare the rates (%) of inducing stroke and seizure. The blood flows of two hemispheres were analyzed using a mixed model ANOVA with time as within-subject factor and hemisphere as between subject factor. The univariate correlations between infarct volume and neurologic score and seizure rate were assessed by Pearson’s correlation. 2.4. Scanning electronic microscopy (SEM) The chitin/PLGA 50/50 blend microparticles were attached onto a double-sided adhesive tape and fixed to an aluminum stage. The microparticles were sputtercoated with gold using a Hitachi coating unit and the surface of the microparticles were examined using a Hitachi S-2300 SEM (Japan). 2.5. Swelling rate of chitin/PLGA microparticles The swelling behavior of chitin/PLGA 50/50 blend microparticles was determined by immersing the microparticles in deionized water at room temperature for 10 min. The mean particle size of the microparticles were measured at 0, 1, 3, 5 and 10 min by photon correlation spectroscopy (Zetasizer 2000HS, Malvern, Worcestershire, UK) using a helium-neon laser with a wavelength of 633 nm at 25  C. The swelling ratio was calculated using the equation shown as below: PSR ð%Þ ¼ ½ðDs Dd Þ=Dd Š  100 where PSR is the percentage of swelling size of the chitin/PLGA 50/50 blend microparticle at equilibrium, Dd and Ds are the diameter size of the samples in the dry and swollen states, respectively. The microscope evaluations were performed in every lot of particles to ensure the homogeneous and round appearance. The particle size estimations by the 3. Results 3.1. Characterization of the chitin/PLGA microparticles The chitin/PLGA microparticles produced by the wet dropping method were smooth and non-porous with PLGA in the middle and chitin surrounding outside (Fig. 1A, left panel). The gradation sieving, by standard U.S. size meshes (from 40 to 400 mesh), yielded 14 sizes of particles (Fig. 1B). The distribution of the amount of each size was skewed; more than 95% of the particles produced by our method were >90 mm (Fig. 1B). Changing the chitin: PLGA ratio would alter the distribution of the amount of each particle size (data not show). The water-absorbing property of chitin invited us to survey the hydration expanding rate of each size of particles. We found that the particle size in all groups increased approximately 45% (43.0e46.5%) within 1 min of dropping them into water and M.-J. Tsai et al. / Biomaterials 32 (2011) 6381e6388 6383 imaging. Immediately after the injection of particles, the blood flow decreased dramatically in the ipsilateral (right) hemisphere but remained relatively constant in the contralateral hemisphere (Fig. 2). The blood flow of the ipsilateral hemisphere stayed low for at least 24 h (Fig. 2). The blood flow of the contralateral hemisphere marginally reduced at 0.5 h after the injection and remained stable afterward (Fig. 2, p ¼ 0.18). We selected two injection sites near the bifurcation of CCA to test the efficacy of inducing embolic stroke by the particles: 1) from the ECA toward the bifurcation of CCA and 2) from the CCA to the ICA. Our results showed that injection from ECA gave higher rates of stroke than that via CCA in two tested sizes of particles: particle150e180mm: 100% (n ¼ 5) vs. 38% (n ¼ 16), p < 0.001; particle180e212mm: 100% (n ¼ 10) vs. 63% (n ¼ 8), p < 0.001; ECA vs. CCA. The amount of injected particles was also tested. Three amounts (1, 2 and 3 mg) of particle125e150mm were injected from ECA and the rates of stroke were 100%, 71% and 80% (p < 0.001), respectively. The volumes of infarction were 259.0  60.5, 214.4  116.6 and 288.3  135.3 mm3 for 1, 2 and 3 mg of particle125e150mm, respectively (p > 0.5). Fig. 1. Physical and chemical characters of the chitin/PLGA microparticles. A) Representative scanning electron micrographs show the surface morphology of particle212e250mm produced by dropping method (left panel, bar ¼ 100 mm) and particle125e150mm produced by grinding method (right panel, bar ¼ 50 mm). B) Distribution of sizes of chitin/PLGA microparticles produced by dropping method. Particles size were classified by different sizes of mesh. The distribution tends to have a skewed type (n ¼ 3). C) Time-passed water expanding rate of the chitin/PLGA microparticles (n ¼ 3). remained at the same size for at least 10 min (Fig. 1C). In the following experiments, we excluded the particles of two extreme sizes (<38 and >355 mm) and used only those particle sizes in between (from 38 to 45 mm to 250e355 mm) to induce embolic stroke. 3.2. Embolic stroke animal model induced by the chitin/PLGA microparticles In this study, we defined “stroke” as decreased cerebral blood flow combined with changes of motor behavior. The change of blood flow in the brain was monitored using Laser Doppler Fig. 2. Change of time-passed blood flow before and after the injection of chitin/PLGA microparticles into cerebral circulation via ECA. A) Representative Laser Doppler Repeat images. B) Quantitative analyses of blood flow before, during and after the injection of particle. Green vertical line: inserting tube; red arrow: injecting particles. Mixed model ANOVA: F ¼ 172.3, p < 0.001. n ¼ 4e6. ** (p < 0.01), *** (p < 0.001): verses contralateral hemisphere. 6384 M.-J. Tsai et al. / Biomaterials 32 (2011) 6381e6388 Fig. 4. Effect of different sizes of Chitin/PLGA microparticles on the incidence of single small infarct in rats. A) Representative TTC-stained sections (2 mm thickness) display different locations of infarcts. Panels: upper left- cortex; upper middle- basal ganglion; upper right- thalamus; lower left- hippocampus; lower middle- midbrain; lower rightsparse small infarcts in cortex, basal ganglion and thalamus. B) The incidence of single small infarcts in rats received different sizes of particles. The isolated infarct involved more than one brain slide was excluded. n ¼ 4e6. C) Representative fluorescent images demonstrate the location of embolization. The embolizations induced by the fluorescent dye IR-780-loaded particle75e90mm (blue color) are sparse, and mainly located in nearby subcortical region including hippocampus, basal ganglion & thalamus and parietal cortex. Serial sections in 2 mm thickness. 3.3. Effect of the sizes of chitin/PLGA microparticles on the severity of stroke Fig. 3. Effect of different sizes of Chitin/PLGA microparticles on the pattern and volume of cerebral infarction in rats. A) Representative TTC-stained serial sections (2 mm thickness) display different infarction patterns induced by different size of particles. B) Quantitative analyses of infarct volumes in various brain regions induced by different sizes of particles. n ¼ 4e6. *, significantly different from particle212e250mm; #, significantly different from particle38e45mm, particle45e75mm and particle212e250mm. Based on previous findings, we decided to inject 1 mg of chitin/ PLGA microparticles from the ECA to the bifurcation of CCA to induce embolic stroke. Our results showed that nearly all sizes of particles had an 80% or higher success rate of inducing stroke with the exception of particle250e355mm which induced stroke in 60% of injected rats. However, different sizes of particles induced dramatic differences in infarct volumes (Fig. 3). Among all the particles, particle212e250mm induced the largest whole brain infarct volumes (>600 mm3) with large diffuse infarction overlying most cortex and hippocampus along with variable damage in the basal ganglion, thalamus and midbrain (Fig. 3A,B). The two smallest tested particles, particle38e45mm and particle45e75mm, also induced large diffuse infarctions (>500 mm3) similar to that of particle212e250mm (Fig. 3A,B). Particle38e45mm induced diffuse infarction in both hemispheres in 60% of rats (Fig. 3A). Although the whole brain M.-J. Tsai et al. / Biomaterials 32 (2011) 6381e6388 6385 Fig. 5. Neurologic scores of rats injected with different sizes of particles and their correlations with infarct volumes in various brain regions. *, significantly different from other groups except particle106e125mm; #, significantly different from particle38e45mm, particle150-180mm and particle212e250mm. n ¼ 4e6. infarct volume (<100 mm3) was the smallest among all groups (Fig. 3B), particle75e90mm frequently induced an interesting pattern of infarction: single or sparse small isolated infarcts in the subcortical region and occasionally in the cortex (Fig. 3A). 3.4. Effect of the sizes of chitin/PLGA microparticles on the single small infarct The unique lacunar stroke-like pattern induced by the chitin/ PLGA microparticles rendered us to further investigate these single small infarcts. Most single small infarcts were round or oval shape, although sometimes they could be cubic or pin-point in the cortex (Fig. 4A). Among all particle size, particle75e90mm and particle90e106mm induced the highest incidence of single small infarct (Fig. 4B). The average volume of a single small infarct were 5.9 mm3 in the cortex, 7.3 mm3 in the basal ganglion & thalamus, 6.2 mm3 in the hippocampus, and 7.6 mm3 in the midbrain & cerebellum. To investigate the location of embolization, we produced fluorescent particles75e90mm loaded with fluorescent dyes IR-780. The IR-780 particles75e90mm were injected through the ECA and the fluorescent images were visualized using an in-vivo optical imaging system (NightOWL II LB983, Berthold Technologies, Bad Wildbad, Germany). The results showed that the embolizations were sparse 6386 M.-J. Tsai et al. / Biomaterials 32 (2011) 6381e6388 Fig. 6. Seizure incidences of rats injected with different sizes of particles and their correlations with infarct volumes in various brain regions. n ¼ 4e6. and mainly located in the subcortical region (Fig. 4C), compatible with the sparse isolated infarctions in the TTC staining (Fig. 4A). 3.5. Effect of the infarct volumes on the neurologic deficits The neurologic behaviors of rats were evaluated repeatedly within 24 h after the injection of the particles and only the highest rating point was used to score the animals. Rats that received particle75e90mm injection displayed the mildest neurologic deficits (Fig. 5). Correlation analyses showed that neurologic score correlated well to the infarct volumes of whole brain and cortex, but not to that of other investigated brain regions (Fig. 5). Some rats that received chitin/PLGA microparticle injection developed seizure, evident by sudden onset of tonic and clonic movement. However, the incidences of seizure varied dramatically, from 0% to more than 60%, among different particle sizes (Fig. 6). The incidence of seizure significantly correlated to the infarct volumes of whole brain and cortex, but not to that of basal ganglion & thalamus, hippocampus or midbrain & cerebellum (Fig. 6). 3.6. Effect of the structure of the chitin/PLGA microparticles on the embolic stroke To clarify the influence of the structure of the particles, we prepared another set of particles by grinding large size particles M.-J. Tsai et al. / Biomaterials 32 (2011) 6381e6388 (>300 mm) into smaller particles (125e150 mm). This method destroyed the spherical appearance of the particles produced by the dropping method (Fig. 1A, right panel). Compared to the infarct volumes and neurologic scores induced by the original particle125e150mm produced by dropping methods, all rats received the same size particles prepared by the grinding method exhibited smaller infarct volumes (24.5  11.6 vs. 259.0  60.5, p < 0.05) and milder neurologic deficit (1.6  0.7 vs. 5.7  1.0, p < 0.05). 4. Discussion Previous animal models designed for cerebral infarcts by embolism obtained unpredictable infarction outcomes. Most studies of the animal model of embolic stroke seldom discussed the characteristics of embolic particles they selected. Among the few cases showed the amount and size of particles selected for embolism, the chemical and physical characters of the particles are seldom discussed [16e21]. Furthermore, most studies reported only one range of particle size without explaining the selection criteria. Therefore, it is not surprising that previous animal models attained unreliable cerebral infarcts by embolism. In this study, we selected chitin/PLGA blending microparticles as artificial emboli for their good biocompatibility and rapid expansion in aqueous solution. The water-absorbing character of chitin makes the chitin/PLGA microparticle expand rapidly in an aqueous environment; hence it permits fast occlusion of cerebral vessels. The similar expanding rates of all sizes of particle indicate the homogeneity of the microparticles. Furthermore, the chitin/PLGA microparticles produced by the dropping method were round across all sizes. The spherical appearance of the particle is critical in occluding the cerebral vessels as the stroke severity induced by the ground particles with cracked surface were only 1/10 of that induced by the same size but sphere-shaped particles. Artificial embolic particles have been synthesized in two size categories: macrospheres and microspheres. The macrosphere procedure installed large 300e400 mm particles into the ICA, which resulted in lodging of the macrospheres in the MCA and produced an infarct of similar volume and location as the permanent sutureocclusion of the MCA [20,22]. The microsphere model selected about 50 mm diameter particles and instilled into the MCA or ICA, that produced diffuse infarctions involving the cortex, hippocampus and thalamus [16,21]. In parallel with these studies, we also found that particle212e250mm and particle38e45mm induced large diffuse infarctions within the MCA territory. Unlike any of previous study, we noticed a unique infarct pattern in the particle-injected rats that resembled lacunar stroke in humans. The single or multiple isolated small infarcts resemble lacunar stroke in many ways. First, human lacunar strokes are often small in size, ranging from 28 to 80 mm3 [7]. Using human hemisphere volumetric measurements, this volume translates to 4.5e14% of a hemisphere. The volumes of single isolated small infarcts detected in this study accounted for no more than 10% of the hemisphere, similar to that of lacunar strokes. Second, the most frequent location of the single isolated small infarcts was the subcortical area, matching that of lacunar infarcts in humans. Third, the shapes of most single or multiple isolated small infarcts were round or oval-shaped, mimicking that of lacunar strokes in humans. The discovery of lacunar stroke-like small infarct is clinically significant, because lacunar strokes are more common in human strokes than large infarcts and are possibly reversible and treatable. Our animal model provides a wonderful opportunity to evaluate future treatment agents and/or strategies for lacunar stroke. The incidence of ischemic stroke-related seizure was evaluated in this study because seizure is related to ischemic stroke in 6387 humans. As expected, the incidences of seizure correlated well to the infarct volume of cortex, but not that of subcortical regions. The expression of seizure in the particle-injected rats further values the animal model of embolic stroke developed by our method. 5. Conclusion We manufactured embolic particles and systematically characterized the patterns of cerebral infarction induced by these artificial emboli with a spectrum of different sizes. Nearly all sizes of particles had an 80% or higher success rate of inducing stroke and the reproducibility rates were not inferior to those of the animal models of diffuse infarction. The small size particle38e45mm and the big size particle212e250mm effectively induced large cerebral infarction including the cortical and subcortical areas. Particles of size smaller than 45 mm crossed easily to the contralateral hemisphere and induced infarctions in both hemispheres. The infarct volumes correlated well to the neurologic deficits and the incidence of seizure. Particle75e90mm induced an unique pattern of infarction similar to that of lacunar infarcts observed in humans. We therefore suggest a rat model of ischemic stroke induced by 75e90 mm chitin/ PLGA microparticles for evaluating lacunar infarct. Acknowledgment This work was supported by the National Science Council of Taiwan (NSC95-2320-B-037-006). References [1] Harmsen P, Wilhelmsen L, Jacobsson A. Stroke incidence and mortality rates 1987 to 2006 related to secular trends of cardiovascular risk factors in Gothenburg, Sweden. Stroke 2009;40:2691e7. [2] Hoyte L, Kaur J, Buchan AM. Lost in translation: taking neuroprotection from animal models to clinical trials. Exp Neurol 2004;188:200e4. [3] Small DL, Buchan AM. Animal models. Br Med Bull 2000;56:307e17. [4] Schneider AT, Kissela B, Woo D, Kleindorfer D, Alwell K, Miller R, et al. Ischemic stroke subtypes: a population-based study of incidence rates among blacks and whites. Stroke 2004;35:1552e6. [5] Norrving B. Long-term prognosis after lacunar infarction. Lancet Neurol 2003; 2:238e45. [6] Zivin JA, Fisher M, Degirolami U, Hemenway CC, Stashak JA. Tissue plasminogen-activator reduces neurological damage after cerebral embolism. Science 1985;230:1289e92. [7] Carmichael ST. Rodent models of focal stroke: size, mechanism, and purpose. NeuroRx 2005;2:396e409. [8] Langer R, Folkman J. Polymers for the sustained release of proteins and other macromolecules. Nature 1976;263:797e800. [9] Cao X, Schoichet MS. Delivering neuroactive molecules from biodegradable microspheres for application in central nervous system disorders. Biomaterials 1999;20:329e39. [10] Sturesson C, Carlfors J. Incorporation of protein in PLG-microspheres with retention of bioactivity. J Control Release 2000;67:171e8. [11] Jain RA. The manufacturing techniques of various drug loaded biodegradable poly(lactide-co-glycolide) (PLGA) devices. Biomaterials 2000;21:2475e90. [12] Eliaz RE, Kost J. Characterization of a polymeric PLGA-injectable implant delivery system for the controlled release of proteins. J Biomed Mater Res 2000;50:388e96. [13] Muzzarelli RA. Chitin and its derivatives: new trends of applied and research. Carbohydr Polym 1993;3:52e7. [14] Mi FL, Lin YM, Wu YB, Shyu SS, Tsai YH. Chitin/PLGA blend microspheres as a biodegradable drug-delivery system: phase-separation, degradation and release behavior. Biomaterials 2002;23:3257e67. [15] Mi FL, Shyu SS, Lin YM, Wu YB, Peng CK, Tsai YH. Chitin/PLGA blend microspheres as a biodegradable drug delivery system: a new delivery system for protein. Biomaterials 2003;24:5023e36. [16] Miyake K, Takeo S, Kaijihara H. Sustained decrease in brain regional blood flow after microsphere embolism in rats. Stroke 1993;24:415e20. [17] Beech JS, Williams SC, Campbell CA, Bath PM, Parsons AA, Hunter AJ, et al. Further characterisation of a thromboembolic model of stroke in the rat. Brain Res 2001;895:18e24. [18] Wang CX, Todd KG, Yang Y, Gordon T, Shuaib A. Patency of cerebral microvessels after focal embolic stroke in the rat. J Cereb Blood Flow Metab 2001; 21:413e21. 6388 M.-J. Tsai et al. / Biomaterials 32 (2011) 6381e6388 [19] Zhang Z, Zhang RL, Jiang Q, Raman SB, Cantwell L, Chopp M. A new rat model of thrombotic focal cerebral ischemia. J Cereb Blood Flow Metab 1997;17: 123e35. [20] Gerriets T, Li F, Silva MD, Meng X, Brevard M, Sotak CH, et al. The macrosphere model: evaluation of a new stroke model for permanent middle cerebral artery occlusion in rats. J Neurosci Methods 2003;122:201e11. [21] Mayzel-Oreg O, Omae T, Kazemi M, Li FH, Fisher M, Cohen Y, et al. Microsphereinduced embolic stroke: an MRI study. Magnet Reson Med 2004;51:1232e8. [22] Gerriets T, Stolz E, Walberer M, Kaps M, Bachmann G, Fisher M. Neuroprotective effects of MK-801 in different rat stroke models for permanent middle cerebral artery occlusion: adverse effects of hypothalamic damage and strategies for its avoidance. Stroke 2003;34:2234e9.