Thoughts and Progress

A New Absorbable Synthetic SubstituteWith Biomimetic Design for Dural

Tissue Repair

*Zhidong Shi, †Tao Xu, ‡Yuyu Yuan,†Kunxue Deng, ‡Man Liu, §Yiquan Ke,

§Chengyi Luo, **Tun Yuan, and ††Ali Ayyad*Department of Neurosurgery, Third AffiliatedHospital, Sun Yat-sen University; ‡School ofBioscience & Bioengineering, South ChinaUniversity of Technology; §Department of

Neurosurgery, Zhujiang Hospital, South MedicalUniversity, Guangzhou; †Bio-Manufacturing Center,

Department of Mechanical Engineering, TsinghuaUniversity, Beijing; **National EngineeringResearch Center for Biomaterials, Sichuan

University, Chengdu, China; and ††Department ofNeurosurgery, University Medical Centre Mainz,Johannes Gutenberg University of Mainz, Mainz,

Germany

Abstract: Dural repair products are evolving fromanimal tissue–derived materials to synthetic materials aswell as from inert to absorbable features; most of themlack functional and structural characteristics comparedwith the natural dura mater. In the present study, weevaluated the properties and tissue repair performance ofa new dural repair product with biomimetic design. Thebiomimetic patch exhibits unique three-dimensionalnonwoven microfiber structure with good mechanicalstrength and biocompatibility. The animal study showedthat the biomimetic patch and commercially syntheticmaterial group presented new subdural regenerationat 90 days, with low level inflammatory response andminimal to no adhesion formation detected at eachstage. In the biological material group, no new subduralregeneration was observed and severe adhesion betweenthe implant and the cortex occurred at each stage.In clinical case study, there was no cerebrospinal fluidleakage, and all the postoperation observations werenormal. The biomimetic structure and proper rate of deg-radation of the new absorbable dura substitute can guidethe meaningful reconstruction of the dura mater, which

may provide a novel approach for dural defect re-pair. Key Words: Biomimetic—Synthetic—Dural repair—Absorbable.

Dural defect is a common problem encounteredduring neurosurgical procedures. Open cranio-cerebral injuries (industrial, traffic, or war-related),tumor invasion, congenital meninges defects, orother cranial diseases can lead to defects of the duramater. Such defects should be repaired in a timelymanner to prevent leakage of the cerebrospinal fluid(CSF), development of encephalocele, and increasein stress from the barometric pressure.

Four types dural substitutes are commonly used inclinic including autologous fascia, allograft, biologicalmaterial derived from animal tissue collagen and syn-thetic material (1–3). However, clinical applicationsof autologous fascia, allograft, and biological materialcan lead to problems such as limited availability,transmission of animal pathogens, low mechanicalstrength, and immunological problems (4–6). Syn-thetic materials pose no risk of transmitting diseaseand can be easily processed into the required shapesand sizes. Generally, synthetic substitutes can bedivided into nonabsorbable and absorbable materials.Regenerated dural tissues can gradually replace theabsorbable construct during the degradation process.Various absorbable synthetic materials are widelyapplied in clinic such as Ethisorb (Ethicon GmbH,Norderstedt, Germany) with polyglactin 910 (vicryl)and poly-p-dioxanone (PDS) (7). However, currentsynthetic dural substitutes lack the complex biologicalfactors that help to drive dural regeneration, oftenresulting in incomplete dura mater healing and higherCSF leakage rates than that of collagen matrices (8,9).Besides, grafts made of bioabsorbable artificial duramater have not been widely used in clinical practicedue to their mechanical incompatibility and an inad-equate degradation rate.

To avoid previously encountered complications,Medprin Biotech GmbH developed a biodegradablenonwoven substitute composed of Poly-L-lactide(PLLA) fibers with open, highly interconnectedpores to create the unique 3D structure that facili-tates cell infiltration and tissue generation, and isgradually absorbed by the body. Besides, the PLLAhas been shown to be fully compatible with dural

doi:10.1111/aor.12568

Received February 2015; revised May 2015.Address correspondence and reprint requests to Dr. Tao Xu,

Bio-Manufacturing Center, Department of Mechanical Engineer-ing, Tsinghua University, Beijing 100084, China. E-mail: taoxu@mail.tsinghua.edu.cn

bs_bs_banner

Copyright © 2015 International Center for Artificial Organs and Transplantation and Wiley Periodicals, Inc.

Artificial Organs 2015, ••(••):••–••

tissue (10–14) and can also reduce tissue adhesion(15–18). The objective of this study was to test thesafety and effectiveness of this biomimetic patch fordural repair. We anticipate this biomimetic patchmay provide functional properties similar to alter-nate therapies. Meanwhile, the mechanical proper-ties and biocompatibility of the biomimetic patchwere characterized both in vitro and in vivo. Twomajor commercially available dural products, bio-logical material patch (NormalGEN) and syntheticabsorbable substitute (SEAMDURA), both manu-factured by using conventional fabrication methods,were selected as controls and compared with thebiomimetic patch in a canine duraplasty model. Oneclinical case is listed to present the safe and effectiveperformance of the biomimetic substitute.

MATERIALS AND METHODS

MaterialsThe biomimetic substitute was provided by

Medprin Biotech GmbH (Frankfurt am Main,Germany), which is a uniquely designed web ofbiocompatible PLLA fibers that is graduallyabsorbed by the body. An electronically controlledprocedure has been developed in which a dissolvedPLLA is sprayed from specialized jets. The PLLAfibers are deposited layer by layer as a fleece-likestructure. The fibered microstructure resembles theextracellular matrix (ECM) of human dura mater(19). Two controls were selected: NormalGEN(Grandhope Biotech. C. Ltd., Guangzhou, China), abiological dural patch made of porcine pericardium,is deantigenized by glutaraldehyde; SEAMDURA(Gunze Ltd., Kyoto, Japan), an absorbable syntheticdural substitute film consisting of copolymers ofL-lactide and ε-caprolactone (P[LA/CL]) andpolyglycolic acid (20,21), is manufactured by conven-tional weaving and/or coating methods and doesnot undergo biomimetic design. Thiazolylbluetetrazolium bromide (MTT) was purchased fromSigma-Aldrich (St. Louis, MO, USA). Dulbeccomodified eagle medium (DMEM) and fetal bovineserum were acquired from Gibco BRL(Gaithersburg, MD, USA).

Characterization of the biomimetic substitute

MicrostructureThe microstructure of the biomimetic substitute was

evaluated using a scanning electron microscope (SEM)(ESEM XL30 HY-3080, FEI, Hillsboro, OR, USA)with a standard SEM sample preparation and scanningprotocol.

Mechanical testingFor the tensile test, samples were cut into pieces of

approximately 10 × 10 mm. The tensile strength andelongation at breakage were recorded. For stitch tearstrength, 10-mm-wide pieces were tested by suturingthe sample approximately 2 mm from its edge using ano. 4-0 silk suture. The maximum tensile strengthwhen the suture was pulled out or upon destructionof the samples was recorded. The tensile and stitchtear strengths were measured at room temperature ata stretching speed of 200 mm/min by a mechanicaltesting instrument (HY-3080, Shanghai HengYi Co.,Ltd., Shanghai, China).

Cytotoxicity testAccording to the ISO10993.5-2009 standard,

cytotoxicity was evaluated with L929 mouse fibro-blast cell line using the MTT assay (22,23). Sterilebiomimetic patches were cut into 10-mm-wide pieces.A total of six samples were prepared and respectivelyextracted in DMEM without serum for 24 h. Theobtained liquid extract (1 mL per well) was separatedinto 24-well plates seeded with L929 cells. Each wellwas seeded with approximately 1 × 105 cells. At thesame time, an equivalent volume of DMEM wasadded as blank and an equivalent dilution of phenol(5%) was added as the positive control. The plateswere then kept in an incubator (37 ± 1°C, 5% CO2)for 3 days, then turned to MTT assay using amicroplate reader (Bio-Rad Model 550, Bio-Rad,Hercules, CA, USA) at a wavelength of 490 nm.Viability and cytotoxicity were expressed using theoptical density (OD) of the treated samples versusOD values of the positive control wells (untreatedcells). Both groups of wells were corrected usingblank measurements of wells without cells.

Animal experiments

Animal selectionAdult Chinese domestic dogs provided by Suibei

Medical Animal Testing Center (Guangzhou, China)were used in this study. This study was designed in fullcompliance with the policies and procedures set by Insti-tutional Animal Care and Use Committee of the ThirdAffiliated Hospital of Sun Yat-sen University(Guangzhou, China). The health and growth conditionsof the experimental dogs were monitored for at least 2weeks before surgery. A total of 24 healthy adult dogsweighing 15–20 kg and within the age range of 1.5–2years were randomly divided into three groups,biomimetic patch, biological material, and syntheticmaterial groups, with eight dogs each for the short- andmiddle-term observation (up to 180 days). Another

THOUGHTS AND PROGRESS2

Artif Organs, Vol. ••, No. ••, 2015

three dogs were selected for a 2-year long-term studyusing the three kinds of dura substitute implanted inboth sides of the dogs’ heads for dural repair.

Surgical methodUpon achieving anesthesia by intravenous admin-

istration, the hair of each dog was shaved off, and theanimal was placed on the operation table in a proneposition. After the top and mid scalp was cut verti-cally, a unipolar electrotome was used to make a0.5-cm lateral cut along the sagittal suture. Then,muscle tissues on both sides of the head weredetached, and the periosteum was detached toexpose a 4 × 3-cm region of the skull plates on twosides of the top skull. A high-speed drill was used togrind the skull plates to form two bone windows(measuring approximately 2.5 × 2 cm). A bipolarcoagulator was used to ablate the blood vessels of theexposed dura mater, and then an oval-shaped pieceof dura (2 × 1.5 cm) on both sides was cut using amicroscopic scissor to create a dural defect on the topskull. The repair material was cut into the appropri-ate size to completely cover the dural defect. Thewound was seamed with 3/0 notched interruptedsuture, with one stitch on each side. The suture wasknotted to eliminate gaps between the patch and theoriginal dura. Penicillin powder was then sprinkledevenly across the cut to prevent infection. Themuscles were sutured using a round pin and no. 1thread. Rubber drains were placed on the right andleft sides outside the dura, and the scalp was discon-tinuously sutured using no. 4 thread.

Postoperative observationAt different time points up to 2 years after surgery,

postoperative complications such as eating or drink-ing, animal activity, infection or CSF leakage wererecorded, the animals were then sacrificed, and theimplanted samples were retrieved. In brief, after theskulls of the dogs were surgically opened, adhesionwas observed, and the implants and surroundingtissues were collected. The retrieved sampleswere visually assessed, and two sections measuring10 × 10 mm were excised. One tissue sample wasplaced into a clean specimen bottle for mechanicaltesting, whereas the other was fixed in formalin solu-tion for histological analyses using hematoxylin andeosin (H&E) and Masson’s trichrome staining.

Tissue biomechanicsTo determine the mechanical changes in the

implant, the tensile strength of the biomimetic patchwas tested before implantation and after implanta-

tion at the time points of 1, 4, 10, and 12 weeks. Thetesting method was the same as described previously.

Grading systemA quantitative grading system was used to score

the tissue response to the implants (24–26). Amacroscopic assessment for adhesions to cerebralcortex, and inflammation, collagen fiber formation ofthe graft were evaluated: 0 = none, 1 = minimal,2 = moderate, and 3 = extensive.

Integration of the device: 0 = no integration,1 = ingress of fibroblasts into the device, 2 = onlyparts of the device remaining detectable, 3 = com-plete integration.

Clinical case studyTo further confirm the safety and efficacy of this

new device, a clinical case was assessed in terms ofthe applicability of biomimetic substitute to a patientwho required dural repair. The China Food and DrugAdministration and the medical ethics committee ofthe Zhujiang Hospital of South Medical Universityapproved the neurosurgical procedure of the clinicalcase using this biomimetic substitute in July 2011.Written informed consent was obtained from thepatient. The Declaration of Helsinki guidelines forresearch on human subjects were followed.

PatientThe patient was a 25-year-old female with a dural

defect after brain tumor surgery. She was selectedand ascertained to have none of the following severeconditions: diseases of the heart, liver, kidney, bloodsystem, or other vital organs; unstable emotion; andpregnant or lactating.

Before surgery, the patient received radiationtreatment for astrocytic glioma in the right frontaland parietal lobe, but the tumor recurred. Thepatient underwent computed tomography (CT),which revealed enlargement of the focus volume andedematous area.

Surgical procedureThe patient underwent surgery under general

anesthesia. After removing the space-occupyingtumors, one 4 × 6-cm biomimetic patch was cutaccording to the dimensions of the dural defect andsutured to the remaining dura mater to achieve sat-isfactory closure. The subcutaneous layer was rinsedwith diluted povidone-iodine solution, and a drain-age tube was installed. Finally, the scalp with fullthickness was sutured.

Outcome measurementsAfter surgery, the patient’s condition was routinely

monitored at 1, 3, 5, 8, and 11 days, including body

THOUGHTS AND PROGRESS 3

Artif Organs, Vol. ••, No. ••, 2015

temperature, signs of meningeal irritation, CSFleakage, hydrocephalus, nausea and vomiting, epi-lepsy, infection, and wound healing. On postopera-tive day 11, the patient underwent a CT scan with alumbar puncture to assess the neurolymph conditionand intracranial pressure. The patient was thentested by routine clinical assays, including blood test,liver and kidney function, cell-mediated immunity,and humoral immunity. The CT scan examined theCSF and hydrocephalus. The patient was followed up90 and 180 days after surgery to determine the pres-ence of any abnormal effects.

Besides this clinical case, a multicenter clinical trialof this new product is in progress. The results of thisclinical trial will be reported upon completion of thestudy.

Statistical analysisStatistical differences were determined using

analysis of variance, followed by the Tukey–Kramermultiple comparison test. Differences were regardedas significant at P < 0.05.

RESULTS

In vitro characterization of biomimeticdural substitute

Gross and microstructure observationAs shown in Fig. 1A, the biomimetic patch appears

as a white membrane with an even thickness ofapproximately 0.3 mm. The SEM image (Fig. 1B)shows that the membrane is a nonwoven fabric com-posed of PLLA fibers, with an average diameter of0.7–2 μm. Absorbable PLLA fibers form a nonwovenweb with open, highly interconnected pores to createthe unique 3D structure of the biomimetic substitute.Its 3D matrix resembles the ECM of human duramater (Fig. 1C) (19), which may provide a favorableenvironment for cell migration and proliferation andthus lead to rapid repair and regeneration of duraltissues. This will be evaluated in the animal experi-ment study.

Mechanical testThe basic mechanical performance of the

biomimetic patch showed that among transverse and

longitudinal directions, the tensile strength rangedfrom 2.8 to 4.3 Mpa and showed no significant differ-ence (P = 0.4716, n = 8). Combining the transverseand longitudinal results, the average tensile strengthwas 4.14 ± 0.18 Mpa (n = 8), the average stretchingelongation was 60.5 ± 13.2% (n = 8), and the averagestitch tear load was 3.71 ± 0.46 N (n = 8), with allsamples having a stitch tear load of >1 N. Comparedwith the low mechanical properties of collagenmatrix substitutes (15,27), the tensile strength of thisbiomimetic patch can be sutured to the dogs’ duralmater easily without CSF leakage.

Cytotoxicity testThe results of the cytotoxicity test showed that the

OD value and cell proliferation rate were 0.57 and101.15%, respectively, for the membrane liquidextracts, 0.58 and 105.55% for the standard tissueculture plate, and 0.1 and 17.87% for the positivecontrol, respectively. No significant difference wasobserved between the construct and the blank (stan-dard tissue culture plate) (P > 0.05). However, sig-nificant differences were detected between patchesand the positive control (P < 0.05). These findingsindicate that the biomimetic patch did not affect cellproliferation and had no cytotoxicity.

Animal experiment studyTo evaluate whether this biomimetic patch

could enhance dural repair effect in vivo, we comparethe feasibility of repairing dura mater of biomi-metic patch with biological material NormalGEN(Grandhope Biotech. C. Ltd.) and synthetic material:SEAMDURA (Gunze Ltd.) in a canine duraplastymodel.

Short- or mid-term observation of the implants (upto 180 days)

After 90 and 180 days of implantation, theimplants were retrieved. Figure 2 shows the repairingeffect of the implants upon gross observation,whereas Fig. 3 presents the interface between theimplant and the cerebral cortex. Figure 4 shows theresults of histological evaluation using H&E stainingand Masson’s trichrome staining, respectively. The

FIG. 1. Appearance (A) of the biomimeticsubstitute, SEM microstructure ofbiomimetic substitute (B) and human duramater (C).

THOUGHTS AND PROGRESS4

Artif Organs, Vol. ••, No. ••, 2015

summary of the grades stratified by macroscopic andhistological findings are provided in Table 1.

After 90 and 180 days of implantation, all animalsin the three groups recovered well and demonstratedcomplete closure of the defects. No postoperativecomplications, such as reduced eating or drinkingand reduced activity, CSF leakage, or infection wasobserved, which indicated good tolerance of thethree materials in vivo. The biomimetic patch andsynthetic material groups integrated well with thesurrounding tissues without leaving any traces oftheir boundaries, whereas the biological materialgroup did not integrate, and the boundary of thepatch could be clearly distinguished (Fig. 2).

To determine whether the patch adhered to thebrain tissues, the interfaces between the implant andthe cerebral cortex were examined. Figure 3 andTable 1 show that most of the areas of the biomimeticpatch and synthetic material implants had no adhe-

sion with clear and smooth surfaces, and only fewspots were determined to be mild adhesions. Theseadhesions could be easily separated byslightly detaching the implant from the brain tissue.However, the biological material group demon-strated severe adhesions between the implant and thecortex. Separating these adhesions can often result indamage of brain tissues.

For histological analyses, the implants and the sur-rounding tissues were isolated (Fig. 4). After 90 daysof implantation, the biomimetic patch and syntheticmaterial were tightly wrapped by the surroundingtissues and replaced with connective dura-like tissueextensively infiltrated by fibroblasts and showedmassive neovascularization (Table 1). The materialswere separated into several fragments and approxi-mately 30–40% of the residual materials wereretained. The biological material samples were alsotightly wrapped by the surrounding tissues. However,

FIG. 2. Observation of the retrievedimplants after 90 and 180 days of implan-tation. Biomimetic patch group: 90th day(A-90d) and 180th day (A-180d); the bio-logical material group: 90th day (B-90d)and 180th day (B-180d); the syntheticmaterial group: 90th day (C-90d) and180th day (C-180d).

FIG. 3. The interfaces between theimplants and cerebral cortex after 90 and180 days of implantation. Biomimeticpatch group: 90th day (A-90d) and 180thday (A-180d); the biological materialgroup: 90th day (B-90d) and 180th day(B-180d); the synthetic material group:90th day (C-90d) and 180th day (C-180d).

THOUGHTS AND PROGRESS 5

Artif Organs, Vol. ••, No. ••, 2015

FIG. 4. HE staining and Masson’strichrome staining of the implants after 90days, 180 days, and 2 years of implanta-tion (HE staining—the first rows; Masson’strichrome staining—the last two rows). A:biomimetic patch group, B: biologicalmaterial group, C: synthetic materialgroup. (Magnification: × 100; →: regener-ated collagen fiber and fibroblast cells; *:muscle fiber; **: cerebral pia mater; ***:brain tissue).

TABLE 1. Summary of grades stratified by macroscopic and histological findings*

Biomimetic patch Biological material Synthetic material

90 days 180 days 2 years 90 days 180 days 2 years 90 days 180 days 2 years

Adhesion of graft to cerebral cortex 0.2 0.3 0 2.6 2.8 2.9 0.2 0.3 0Inflammation 1.4 1.2 0 1.0 0.9 0.8 1.2 0.9 0Integration of the device 1.5 2.0 2.9 0 0.2 0.5 1.3 1.9 3.0Collagen fiber formation of the graft —† 2.5 3.0 — 1.0 1.5 — 1.8 3.0

* Scores are detailed in the Materials and Methods section.† No data.

THOUGHTS AND PROGRESS6

Artif Organs, Vol. ••, No. ••, 2015

there was no indication that the material was beingresorbed and no infiltrated fibroblasts or bloodvessels because of its dense structure that wasderived from bovine pericardium and no newsubdura was present. The inflammatory reactions ofthe three groups were similar, which were deter-mined to be mild and involved only a few lympho-cytes and foreign giant cells (Table 1). Localcalcification was observed in both the biologicalmaterial and synthetic material groups.

After 180 days of implantation, all three groupsshowed a reduction in the number of inflammatorycells, including macrophage and lymphocytes (Fig. 4and Table 1). For the biomimetic patch group, thePLLA material was mostly degraded, with only20–30% remaining, and the degraded part wasreplaced by new highly dividing fibrotic tissues, thedura-like tissue became thicker and was composedof a dense network of well-oriented collagen fibersand capillary neovessels settling in the dural defect.In the biological material group, minimal fibroblastswere observed growing into the dense structure,showing minimal degradation of the material. Adhe-sion between the new fibrosis tissue and brain tissuewas clearly detected. On the surface of the braintissue, a cerebral pia mater defect was detected, withsome cerebral pia integrating with new fibrosistissues. For the synthetic material group, severalfragments of the material were distributed acrossthe new dural tissues, accompanied with minimalcalcification.

Grades from Masson’s trichrome staining data(Table 1) showed that all the three groups had under-gone collagen formation. The newly formed collagenfibers in the biological material group was minimal.In particular, the biomimetic patch group showed theformation of a collagen layer on the surface of the

cerebral pia mater, consisting of highly organized col-lagen fibers.

Long-term observation of the implants(up to 2 years)

Figure 5 shows a 2-year repairing effect of thethree patches. After 2 years of implantation, thethree groups were found to be safe and effective inrepairing the created defect in dura mater but withdifferent biological response, device resorption,cell penetration, vascularization, and collagenremodeling.

Figure 4 and Table 1 show the histological resultsof long-term implantation. The biomimetic patchgroup showed complete degradation of materials,which were then replaced by regularly aligned fibro-blasts and collagen fibers. The new dural tissueconsisted of a large amount of fibroblasts, and noinflammatory cells were observed. In the syntheticmaterial group, only minimal material was remain-ing, and the cerebral pia mater was intact with anormal cerebral cortex. The materials showed amassive deposition of fibroblasts and collagen,without signs of focal calcification. Most of the mate-rial in the biological material group showed slowdegradation, thus still showed the original collagenfibers of the biological patch, which originated fromporcine pericardium, along with a few newly depos-ited fibroblasts and collagen fibers, and showed fila-mentous adhesion between the outer surface of themesh and brain musculature, which was difficult toseparate. Minimal focal calcification and a few carti-lage cells were observed in the fibroblast tissues. Afew inflammatory cells had infiltrated into one side ofthe cerebral pia mater.

In summary, comparing with the biological mate-rial with slow degradation, biomimetic patch and

FIG. 5. Observation of the retrievedimplants for the long-term period up to 2years. A: biomimetic patch group, B: bio-logical material, C: synthetic material.

THOUGHTS AND PROGRESS 7

Artif Organs, Vol. ••, No. ••, 2015

synthetic material had the advantage of beingabsorbed after guiding the real regeneration of thedura mater.

Tissue biomechanicsTo monitor the tensile-resistant strength of the

regenerated dura-like tissue after implantation at thedura defect, the tensile strength of the biomimeticpatch was compared before and after implantation atdifferent time points. The biomimetic patch has amaximum tensile load of 11.86 ± 2.24 N beforeimplantation. As the material degraded, themaximum tensile load decreased to 7.68 ± 0.43 Nafter 1 week and 7.28 ± 0.78 N at 4 weeks of implan-tation. However, with the formation of new tissues,the maximum load then gradually increased to19.83 ± 1.05 N after 8 weeks and 34.42 ± 0.74 N after12 weeks of implantation (Fig. 6).

This phenomenon of an initial decrease duringthe early phase of implantation followed by anincrease in strength highly correlated with the spe-cific material and structural properties of thebiomimetic scaffold. The biomimetic substitute iscomposed of biodegradable PLLA, which willgradually degrade after implantation in the body.The tensile strength decreased with the degradationof PLLA at the early stage (4 weeks) while at thesame time it can also sustain leak preventionwhile the tissue is healing. Furthermore, the tensilestrength rapidly increased (after 8 weeks), reachinglevels of the natural dural tissue. The tissue biome-chanics was consistent with the results of histo-logical observation. The unique 3D biomimeticstructure facilitates the rapid growth of new duraltissues, whose strength compensates the mechanicalloss due to material degradation. Induction of rapidgrowth of the dural tissues by the presence of thepatch also ensures that no CSF leaks out during andafter degradation of the fabricated material anddemonstrates the equivalence of the degradationrate of the materials to that of the new tissues.

Clinical researchOne clinical case with a 6-month follow-up was

studied. The body temperature of the patient wasnormal, and there was no obvious infection or CSFleakage after surgery as shown in Table 2. The CTimage on the third day confirmed that the sizes of theintracranial pneumatocele and the hematocele in theright temporal lobe were significantly reduced com-pared with the preoperative sizes (Fig. 7). The righttemporal lobe consisting of the gray matter showed alarge lamellar low-density shadow, and the area wasthe same as that observed preoperatively, althoughwith a low density. The same findings were alsoobserved in the suspected brain infarction foci. Onthe 3rd and 11th day, no CSF leakage was observedby CT (Fig. 7C,D). The hematocele and edema in theright temporal lobe focus were absorbed, and clinicalimprovement was seen in the suspected brain infarc-tion foci. The other features remained the same aspreoperation.

At 90- and 180-day clinical follow-ups, the patientshowed no fever, headache, nausea, vomiting,

FIG. 6. The mechanical analysis of the biomimetic patch afterimplantation at different time points.

TABLE 2. Clinical postoperation observation

Postoperation(days) Headache Fever

Nausea &vomiting

CSF &subcutaneous

fluidsWound

heal

1 Mild No Yes No —3 Slight No Yes No Good5 No No No No Good8 Intracranial pressure and CSF normal*

* From lumbar puncture examination.

THOUGHTS AND PROGRESS8

Artif Organs, Vol. ••, No. ••, 2015

meningeal irritation, epilepsy, or CSF leakage, andgood healing was observed.

DISCUSSION

Dural substitutes have evolved from animalsources to synthetic materials to avoid the transmis-sion of animal diseases as well as from inert tobiodegradable features to prevent chronic inflam-matory reactions. Moreover, biomimicry-inspireddesign is becoming an important trend in tissuescaffold including artificial dural substitutes toenhance the formation of new tissues. Althoughthere have been extensive efforts to develop syn-thetic biodegradable dural products with improvedproperties in the past decade, these products arebased on conventional fabrication methods such asweaving, knitting, coating, and solvent casting andpossess structural limitations such as being too stiffand tough for handling and less biocompatible.More importantly, these products lack a biomimeticdesign, leading to less tissue repair and regeneration

because their microstructures are markedly differ-ent from the natural dural tissues. To address theseissues, new strategies and fabrication methods aredesired.

In the present study, the biomimetic PLLAimplant exhibits 3D networks of microfibers rangingfrom 0.7 to 2 μm, highly similar to the ECM environ-ment of the natural dural tissues (19), and goodtensile strength (≥4 Mpa) is expressed to withstandinitial surgical implantation and sustain leak preven-tion while the tissue is healing. PLLA is biodegrad-able material and has been approved by the Foodand Drug Administration for clinical applications.Both in vitro cytotoxicity and in vivo animal experi-ments showed excellent biocompatibility of thebiomimetic patch. In the animal study following up to180 days, the biomimetic patch and synthetic mate-rial group presented new subdural regeneration andtriggered a low level inflammatory response, indi-cated by the extremely low number of inflammatorycells detected at each stage of implantation. Mean-while there was minimal to no adhesion formation. In

FIG. 7. The CT scan images and surgicaloperation of the patient (A: CT scan imagebefore operation; B: Surgical method;C: CT scan image on the third daypostoperation; D: CT scan image on the11th day postoperation).

THOUGHTS AND PROGRESS 9

Artif Organs, Vol. ••, No. ••, 2015

the biological group, no new subdural was present,and severe adhesions between the implant and thecortex were observed. The tissue biomechanics of thebiomimetic patch after implantation was also consis-tent with the results of histological observations. Allthese results demonstrate the equivalence of the deg-radation rate of the biomimetic patch to that of newtissues. In the histological evaluation of thebiomimetic patch implantation at 2 years, thebiomimetic patch showed ideal repair of duraldefects and complete degradation that indicated itslong-term safety. The clinical application in a patientwho underwent removal of astrocytic glioma andsubsequent repair of the dural defect was alsoreported. All major clinical observations, includingabsence of CSF leakage and local infection, as well asthe observation of wound healing were indicative ofits positive effects. The surgery was completed 2years ago, and the patient is now healthy and showsnormal functional capacity.

CONCLUSION

In summary, the mechanical properties and bio-compatibility of this biomimetic absorbable PLLAdural repair patch were characterized in both invitro and in vivo studies. The animal study demon-strated that there was no significant differencebetween the tissue biocompatibility, inflammatoryreaction, absence of cerebrospinal fluid leakage ofthis biomimetic patch and those to theSEAMDURA, as well as superior anti-adhesiveproperty than NormalGEN. This biomimetic patchhas good strength that enables it to be easilysutured and sustain leak prevention while the tissueis healing. The biomimetic structure and absorbablePLLA induces a proper rate of degradation equiva-lent to the rate of formation of new tissues and thereal dural reconstruction. This biomimetic patchachieves complete degradation within a 2-yearperiod and has been assessed to be safe and effec-tive when used for a long time in animal and clinicalapplications. In addition, this biomimetic patch hasan ample source without virus transmission. Thenew absorbable synthetic dural substitute withbiomimetic design shows great potential as an arti-ficial dural substitute.

Acknowledgments: We would like to express oursincere gratitude to Medprin Biotech GmbH forproviding us this experimental opportunity. Thiswork is partly supported by China GuangdongInnovative Research Team Program (No. 2011S055)and China Shenzhen Peacock Plan Project (No.KQTD201209).

Conflict of Interest: The authors report no conflictof interest concerning the materials or methods usedin this study or the findings specified in this article.

REFERENCES

1. Costa BS, Cavalcanti-Mendes GDA, de Abreu MS, de SousaAA. Clinical experience with a novel bovine collagen duramater substitute. Arq Neuropsiquiatr 2011;69:217–20.

2. Mukai T, Shirahama N, Tominaga B, Ohno K, Koyama Y,Takakuda K. Development of watertight and bioabsorbablesynthetic dural substitutes. Artif Organs 2008;32:473–83.

3. Narotam PK, Qiao F, Nathoo N. Collagen matrix duraplastyfor posterior fossa surgery: evaluation of surgical technique in52 adult patients. Clinical article. J Neurosurg 2009;111:380–6.

4. Bejjani GK, Zabramski J, Durasis Study Group. Safety andefficacy of the porcine small intestinal submucosa dural sub-stitute: results of a prospective multicenter study and literaturereview. J Neurosurg 2007;106:1028–33.

5. Martinez-Lage JF, Perez-Espejo MA, Palazon JH,Lopez Hernandez F, Puerta P. Autologous tissues for duralgrafting in children: a report of 56 cases. Childs Nerv Syst2006;22:139–44.

6. Martinez-Lage JF, Rabano A, Bermejo J, et al. Creutzfeldt-Jakob disease acquired via a dural graft: failure of therapy withquinacrine and chlorpromazine. Surg Neurol 2005;64:542–5,discussion 545.

7. Esposito F, Cappabianca P, Fusco M, et al. Collagen-onlybiomatrix as a novel dural substitute. Examination of the effi-cacy, safety and outcome: clinical experience on a series of 208patients. Clin Neurol Neurosurg 2008;110:343–51.

8. Yamata K, Miyamoto S, Takayama M, et al. Clinical applica-tion of a new bioabsorbable artificial dura mater. J Neurosurg2002;96:731–5.

9. Von Wild KRH. Examination of the safety and efficacy of anabsorbable dura mater substitute (Dura Patch) in normalapplications in neurosurgery. Surg Neurol 1999;52:418–25.

10. Gautier S, Oudega M, Fraguso M, et al. Poly(a-hydroxyacids)for application in the spinal cord: resorbability and biocom-patibility with adult rat schwann cells and spinal cord.J Biomed Mater Res 1998;42:642–54.

11. Levy FE, Hollinger JO, Szachowicz EH. Effect of abioresorbable film on regeneration of cranial bone. PlastReconstr Surg 1994;93:307–11.

12. Lundgren D, Nyman S, Mathisen T, Isaksson S, Klinge B.Guided bone regeneration of cranial defects, using biodegrad-able barriers: an experimental pilot study in the rabbit. JCraniomaxillofac Surg 1992;20:257–60.

13. Nyilas E, Chiu T, Sidman R, et al. Peripheral nerve repair withbioresorbable prosthesis. Trans Am Soc Artif Intern Organs1983;29:307–13.

14. Seckel B, Chiu T, Nyilas E, Sigman R. Nerve regenerationthrough synthetic biodegradable nerve guides: regulation bythe target organ. Plast Reconstr Surg 1984;74:173–81.

15. Wang YF, Guo HF, Ying DJ. Multilayer scaffold ofelectrospun PLA–PCL–collagen nanofibers as a dural substi-tute. J Biomed Mater Res B Appl Biomater 2013;101B:1359–66.

16. Iliopoulos J, Corwall GB, Evans RON, et al. Evaluationof a bioresorbable polyactide film in a large Animal MODELfor the reduction of retrosternal adhesions. J Surg Res2004;118:144–53.

17. Klopp LS, Welch WC, Tai JW, et al. Use of polylactideresorbable film as a barrier to postoperative perdural adhesionin an ovine dorsal Laminectomy model. Neurosurg Focus2004;16:1–9.

18. Walsh WR, Evans RN, Iliopoulos J, Cornwall GB, ThomasKA. Evaluation of a bioresorbable polylactide sheet for thereduction of pelvic soft tissue attachments in porcine animalmodel. J Biomed Mater Res B Appl Biomater 2006;79B:166–75.

THOUGHTS AND PROGRESS10

Artif Organs, Vol. ••, No. ••, 2015

19. Protasoni M, Sangiorgi S, Cividini A, et al. The collagenicarchitecture of human dura mater. J Neurosurg 2011;114:1723–30.

20. Yamada K, Miyamoto S, Nagata I, et al. Development of adural substitute from synthetic bioabsorbable polymers. JNeurosurg 1997;86:1012–7.

21. Yamada K, Miyamoto S, Takayama M, et al. Clinical applica-tion of a new bioabsorbable artificial dura mater. J Neurosurg2002;96:731–5.

22. Chiba K, Makino I, Ohuchi J, et al. Interlaboratory validationof the in vitro eye irritation tests for cosmetic ingredients (9).Evaluation of cytotoxicity test on HeLa cells. Toxicol in Vitro1999;13:189–98.

23. Gociu M, Patroi D, Prejmerean C, et al. Biology andcytotoxicity of dental materials: an in vitro study. Rom JMorphol Embryol 2013;54:261–5.

24. Haq I, Cruz-Almeida Y, Siqueira EB, Norenberg M, GreenBA, Levi AD. Postoperative fibrosis after surgical treatmentof the porcine spinal cord: a comparison of dural substitutes:invited submission from the Joint Section Meeting on Disor-ders of the Spine and Peripheral Nerves, March 2004. JNeurosurg Spine 2005;2:50–4.

25. Barbolt TA, Odin M, Léger M, Kangas L, Holste J, Liu SH.Biocompatibility evaluation of dura mater substitutes in ananimal model. Neurol Res 2001;23:813–20.

26. Neulen A, Gutenberg A, Takács I, et al. Evaluation of efficacyand biocompatibility of a novel semisynthetic collagen matrixas a dural onlay graft in a large animal model. Acta Neurochir(Wien) 2011;153:2241–50.

27. Sandoval-Sánchez JH, Ramos-Zúñiga R, de Anda SL, et al. Anew bilayer chitosan scaffolding as a dural substitute: experi-mental evaluation. World Neurosurg 2012;77:577–82.

THOUGHTS AND PROGRESS 11

Artif Organs, Vol. ••, No. ••, 2015