• Users Online: 88
  • Print this page
  • Email this page


 
 Table of Contents  
ORIGINAL ARTICLE
Year : 2019  |  Volume : 52  |  Issue : 2  |  Page : 45-51

Biodegradation patterns of injected composite bone cements in porcine vertebral bodies: A study using quantitative computed tomography


1 Department of Radiology, En Chu Kong Hospital, New Taipei City; Department of Medical Imaging and Radiological, Yuanpei University of Medical Technology, Hsinchu City, Taiwan
2 Department of Medical Imaging and Radiological, Yuanpei University of Medical Technology, Hsinchu City, Taiwan
3 School of Dental Technology, College of Oral Medicine, Taipei Medical University, Taipei, Taiwan
4 Department of Surgery, En Chu Kong Hospital, New Taipei City; Department of Biomedical Engineering, Yuanpei University of Medical Technology, Hsinchu City, Taiwan

Date of Submission04-Jun-2018
Date of Decision19-Jul-2018
Date of Acceptance18-Nov-2018
Date of Web Publication18-Apr-2019

Correspondence Address:
Dr. Chang-Chin Wu
Department of Surgery, En Chu Kong Hospital, No. 399, Fuxing Rd., Sanxia District, New Taipei City 23702
Taiwan
Login to access the Email id

Source of Support: None, Conflict of Interest: None


DOI: 10.4103/fjs.fjs_60_18

Rights and Permissions
  Abstract 

Background: For vertebroplasty, newly synthesized bone cements are proposed to replace traditional polymethylmethacrylate (PMMA). Most inventors initially evaluated these newly developed cements in animal spine models. However, even these time- and work-consuming histological inspections performed meticulously by experienced hands, there are still lots of specimen lost during the processing procedures. Although the histological sections can reveal new bone formations and surrounding tissue reactions to implanted materials, it is difficult to identify the degradation processes of the injected cement. In fact, there is no standard method to quantify the volume changes of injected substitutes postoperation.
Methods: Previously, we developed two new biodegradable cements and evaluated performances in fixed-volume and fixed-shaped holes in vertebral bodies of porcine lumbar spine. The animals were sacrificed and the retrieved spines were analyzed after 3 and 6 months. Herein, we further used computed tomography (CT) and three-dimensional CT (3D-CT) to quantitate volumes and biodegradation of cements inside vertebral bodies after previous attestation of CT findings. Exteriors of controls and injected materials were reconstructed with different Hounsfield units (HU); changes of HU as well as cement volumes were later calculated.
Results: The results revealed that the volumes and shapes of these biodegradable cements can be determined by 3D-CT. After meticulous comparisons among gross specimens, histologies, and CT images, the different patterns observed in CT implied consistency among all three observations. Gradual reductions of HU and volumes of newly synthesized cements showed the degradability. Meanwhile, consistent HU and volumes of PMMA meant its inertness.
Conclusion: CT imaging may be a preliminary, quantitative, and liable way for evaluating injectable bone cements in the vertebral bodies.

Keywords: Composite bone cement, vertebroplasty


How to cite this article:
Yeh CC, Chen CJ, Tang Y, Yang KC, Wu CC. Biodegradation patterns of injected composite bone cements in porcine vertebral bodies: A study using quantitative computed tomography. Formos J Surg 2019;52:45-51

How to cite this URL:
Yeh CC, Chen CJ, Tang Y, Yang KC, Wu CC. Biodegradation patterns of injected composite bone cements in porcine vertebral bodies: A study using quantitative computed tomography. Formos J Surg [serial online] 2019 [cited 2019 Nov 19];52:45-51. Available from: http://www.e-fjs.org/text.asp?2019/52/2/45/256534


  Introduction Top


The current commercial polymethylmethacrylate (PMMA) bone cement applied in vertebroplasty and kyphoplasty has many disadvantages such as exothermal reactions, release of toxic monomers and PMMA debris, low viscosity, nonbiodegradability, and significantly high mechanical strength.[1] Many new synthesized bone cements are developed and claimed to be better than traditional PMMA.[2],[3],[4] Most inventors initially evaluated the in vivo performances of newly synthesized cements in animal spine models.[5],[6],[7],[8],[9] To mimic vertebral augmentation procedure in regular clinical setting, the tested cements are injected into animal vertebral bodies posteriorly through pedicles under fluoroscopic guide. However, the volumes, shapes, and positions of these intraosseous cement blocks are not consistent; accurately evaluating the characteristics of the cements in both qualitative and quantitative comparisons is difficult.

To evaluate the host responses of bony tissue to injected bone cements as well as the biodegradation of transplanted materials, histological examinations are widely conducted. Using computer-based image analysis techniques, histomorphometry illustrates the characteristics of implanted cement even including some quantitative data, such as percentages of cement degradation and volume of new bone formations.[10],[11] However, despite these time- and work-consuming histological analyses performed meticulously by experienced hands, there are still many specimen losses during the processing procedures.[12] In general, hematoxylin and eosin stain (H and E stain), Von Kossa stain, and Goldner's Masson trichrome stain are used to examine the vertebral bodies with implanted materials.[13],[14] In spite of many studies assessing material degradation and bone formation using histomorphometry, the bony samples require a decalcification process as a histological preparation.[13],[15] This procedure may cause dissociations of the residual composite and dislodgements of hard inert cement part with surrounding bone. In histology studies without decalcification, improper thickness of sample slices and scratches on the samples by the cutting diamond saws also significantly jeopardized accuracy. Defective slides, including those with inadequate stain toning, tissue sample placement errors, dislodgement of composites, and imperfect section thickness, are discarded. With the lost in specimen or mistakes taken into consideration, the statistical data will not be accurate and meaningful. Therefore, histologic studies can only be descriptive evaluations, not quantitative assessments. Furthermore, the outer appearances of the cement block cannot be illustrated by histologic method. There should be some modifications to improve these systemic errors.

Currently, there is no recommended method to evaluate the cement shape and volume changes of injected substitutes postoperation. To establish a method to evaluate the injected bone cements, we proposed to use computed tomography (CT) with three-dimensional (3D) reconstruction for the volumetric assessments of degradation and quantification of radio-opacity in a porcine study.


  Materials and Methods Top


Animal study

In our previous study, two injectable bone cements composed of 30% polymer poly (propylene fumarate) (PPF) and 70% ceramic powders (α-tricalcium/hydroxyapatite [α-TCP/HAP]; and tetracalcium phosphate/dicalcium phosphate [TtCP/DCP]) were prepared.[16] The in vivo performances of these two cements were evaluated in 12 miniature pigs (4–5 months old, Sus barbatus sumatranus). Under general anesthesia, the pigs were placed in the right decubitus position. After division of psoas muscle and ligation of segmental vessels, the lumbar vertebral bodies were exposed through lateral retroperitoneal approach. Four defects (5 mm in diameter and 10 mm in length) were created at the center of the lateral cortex of the vertebral bodies of the lumbar spine of each pig. The newly developed biodegradable cements, PPF/α-TCP/HAP and PPF/TtCP/DCP, were then injected into the holes randomly in fixed volumes. A commercial PMMA product, the most widely applied bone cement, Osteobond (Zimmer, Warsaw, Indiana, U.S.A.) was used as comparison material while one hole was left empty as a control.[12]

Postmortem assessments

Six pigs were sacrificed at 3 and 6 months by injection of overdosed thiamylal sodium solution. The whole spine with paraspinal muscles attached was harvested and analyzed by CT scan (LightSpeed VCT, GE Healthcare Global Diagnostic Imaging, Waukesha, WI, USA). The different appearances of blocks were recorded, and the Hounsfield units (HU) were calculated. Any extrablock or extraosseous leakage of cement was searched and analyzed. The 3D cement block CT (3D-CT) images were reconstructed using Vitrea Fx version 3.1 (ViTAL Images, Minnetonka, Minnesota, USA) software to calculate and analyze the volumetric changes and the HU difference. The average HU were derived from calculating the HU of central circles with 19 mm2 area (the estimated area) and measuring five slices per block. As for HU of normal bone, the same areas of circles of all lumbar vertebrae were collected, and the average HU was computed from five sagittal images of each spine. The 3D images were derived from the divergences of HU and the structures were also reconstructed based on HU variations. For the differences of HU among cement blocks (>1000), defects of control (<500) and normal bone (around 600) were significant, and the exterior appearances of the blocks could be reconstructed easily with the volumes obtained [Figure 1].
Figure 1: Three-dimensional computed tomography scans for volumetric measurement. (a) Reconstructed image for vertebral and block morphology evaluation and volumetric quantification (b) Subtraction of normal bone for better morphologic assessment. (c and d) Hounsfield units assessments of different areas of block

Click here to view


The different appearances of normal bones, injected cements, and interaction zones between cement and bone in CT images were analyzed and classified according to distinct morphologies.

After CT scans done, calcified evaluation (Von Kossa staining) and noncalcified histologic evaluations (H and E stain and Goldner's stain) of each vertebral body were performed. The histologic characteristics of normal bones, injected cements, and interaction zones between the cement blocks and surrounding bone were compared to those images identified in CT scan. Based on previous classification, meticulous analyses of results of gross morphologies of cement blocks and surrounding bony structures, histological findings, and CT images, different CT scan pictures implied different gross and histologic findings. Clear thin bone cement junctions revealed in CT scans represented histologically good incorporations between cement and surrounding bone; hyposignal radiolucent junctions along hypersignal cement blocks in the CT images symbolized fibrous tissue interposition with adverse host reactions to decomposed cements, and in some vertebral bodies with fabulous bony incorporation, there is blurred boundary between the cement blocks and ingrown newly formed bone. The comparison data had been collected and published.[12]

Statistical methods

Data were expressed as mean ± standard error of the mean. Statistical analyses were made by analysis of variance with post hoc Dunnett's multiple comparison tests. Statistically, significant difference was set at P < 0.05.


  Results Top


Computed tomography and three-dimensional reconstruction of computed tomography image

During the rearing course, plain X-rays of the pigs were taken. However, any chronological difference was difficult to be determined by plain X-ray films, which could be influenced by the sizes of the growing animal in this study (data not shown). Thus, the volumetric or morphologic changes of cement blocks could not be measured accurately using plain X-ray films.

Conversely, cement leakage into the spinal canal through a posterior cortical breakage and another posterior cortex breakage in the control group were identified by CT scan clearly. Except these two, varying degrees of drilled hole size reductions could be illustrated distinctly, and one complete union of the hole was identified. Consistent block configuration was observed in all PMMA blocks, and well-defined interface between bone and cement block was revealed in almost all PMMA vertebrae. As for composite cements, dissimilar block configurations were revealed by CT scans varying from good incorporation between cement and surrounding bone to cement block breakdown circled by a wide radiolucent halo, even with vertebral body deformity.

The 3D-reconstructed images of control group were unable to be obtained because the linear HU discrepancies of newly grown bone at different stages of mineralization gradually filled empty holes and jeopardized imaging. Meanwhile, the contours of all inert PMMA block remaining cylindrical without significant divergences, and changeable morphologies of degradable cement blocks were shapely displayed after simultaneous reduction of surround bone pixels [Figure 2].
Figure 2: The exterior contours of different cement blocks inside the vertebral bodies of a sacrified pig after 6 months' breading were presented in three different views and in different colors including (a) anterioposterior view, (b) oblique view, (c) lateral view

Click here to view


Hounsfield units measurement

The average HU of normal vertebral bodies in this study was 602 ± 80 at 3 months and 624 ± 70 at 6 months. There was no significant difference in the HU of normal vertebral bodies between 3 and 6 months. The HU of PMMA (1423 ± 70 and 1439 ± 65) remained almost constant. In contrast, defects in the control group increased significantly, from 265 ± 144 to 520 ± 118 (96% increase; P < 0.05). In the meantime, the HU of the α-TCP/HAP group (1316 ± 112 and 1270 ± 65, 3% decrease; P > 0.05) and the TtCP/DCP group (1226 ± 107 and 1046 ± 177, 15% decrease; P > 0.05) decreased after 3 more months. There was no significant difference in the HU of α-TCP/HAP and TtCP/DCP groups at 3 and 6 months. However, the HU of PMMA was significantly higher than that of the TtCP/DCP group at 3 months (P < 0.05) and 6 months (P < 0.01) [Figure 3] and [Figure 4].
Figure 3: Hounsfield units of postoperative 3 months: Significantly higher Hounsfield units in polymethylmethacrylate group than composite, defect and normal bone groups

Click here to view
Figure 4: Hounsfield units of postoperative 6 months: Hounsfield units of polymethylmethacrylate and normal bone groups remained almost constant, while significant increase in control defect group

Click here to view


Three-dimensional volumetric measurement

After 3 months, the average block volumes were 318 ± 44 mm3, 317 ± 39 mm3, and 308 ± 53 mm3 in the α-TCP/HAP, TtCP/DCP, and PMMA groups, respectively. There were no significant differences between the study groups and the controls (280 ± 44 mm3), and the defects in the controls were smaller than those with cements, but not statistically significant (P > 0.05). At 6 months, volume defects in the controls shrank significantly to 180 ± 40 mm3 (36% decrease; P < 0.01). Similar results were observed in the α-TCP/HAP group (255 ± 43 mm3, 20% decrease; P < 0.05) and in the TtCP/DCP group (237 ± 34 mm3, 25% decrease; P < 0.01), with volumes reducing significantly in composite cements with defects. However, those of PMMA (313 ± 48 mm3) were not lessened and remained constant after 6 months [Figure 5] and [Figure 6].
Figure 5: Volumetric measurement of cement blocks and the hole of control at postoperative 3 months

Click here to view
Figure 6: Volumetric measurement of cement blocks and the hole of control at postoperative 6 months

Click here to view



  Discussion Top


Traditional histological inspections are used to evaluate the responses of host tissue to biomaterials.[12],[13],[14],[15] Regarding the in vivo evaluations for the effects of new bone cements in vertebral augmentation procedures (VAPs), many animal models have been proposed to mimic regular VAP procedures with cements injected posteriorly through catheters via pedicles under fluoroscopic guidance. A major shortcoming of these studies is that most of them only provide qualitative assessments of cements and reactions between the cement and surrounding host bone tissue. Furthermore, histomorphometric analyses may prove to be difficult to quantify the volume change of implanted materials. Otherwise, Zhu et al. reported 15.47% new bone formation and 16.48% calcium phosphate cement (CPC) degradation 24 weeks after implantation by histomorphometric assessment of undecalcified vertebral bodies.[5] Similar method was reported by Galovich et al.[6] However, the possibility of specimen loss during sample preparation, staining problems, or geographic bias during processing, and difficult validations of different kinds of tissues by the naked eye or even by recognition software cannot be ignored. Using CT with 3D reconstruction, we tried to establish a digital method to assess the volume changes and the radio-opacity of injectable bone cements in a porcine model. Similarly, Sariibrahimoglu et al. used a computer-based image analysis technique to evaluate the new bone formation surrounding a composite cement.[17] Likewise, Alt et al. used high-resolution peripheral quantitative CT to determine the percentages of material degradation, new bone formations, and surrounding tissue reactions.[18]

Since reconstruction of the block's 3D configurations is based on the HU discrepancies among different structures, accurate and sharp outward appearances of cement blocks are generated in this study for the significant HU divergences between cement blocks (>1000) and normal bone (about 600). The obtained volumes of cement blocks and control defects are significantly higher than the estimates (π × 2.52 × 10 = 196.4 mm3) and the volume of all cement blocks volume is >300 mm3. The assumed reasons for the increased volumes are inaccuracies during surgical procedures, leakage of cement, volume expansions into the surrounding bone marrow spaces during polymerization in wet field environments, or higher HU expressions of surrounding bone stocks after rearing.

Cement volume and morphology assessment using 3D reconstruction is more reasonable and accurate than traditional histomorphometric assessment. However, the current study has a shortcoming. Experimental data comparisons between different rearing intervals were among different sacrificed individuals, thereby weakening the statistical strengths. The analyses of CT scan data collected at different rearing intervals without sacrifice from the same individual are more accurate and statistically powerful.

The CT imaging data of PMMA remain stationary at 3 and 6 months, with few variations of HU, volume, and surrounding bone reactions. Histological specimens reveal similar results, with only direct bone–cement contacts or a thin layer of fibrous tissues between the surrounding bone and PMMA cement. Compared to composite cements, the results of PMMA are very reliable and predictable. In our previous in vitro study, there was no significant difference (P = 0.45) between HU of 70% α-TCP/HAP group (1456.00 ± 80.64 HU) and PMMA (1461.78 ± 6.62 HU).[16] However, in this in vivo study, the HU of the α-TCP/HAP (1316 ± 112 and 1270 ± 65) and TtCP/DCP groups (1226 ± 107 and 1046 ± 177) were lower than that of PMMA (1423 ± 70 and 1439 ± 65). This illustrated the degradation abilities of composite cements after 3 and 6 months, as well as different results in an in vivo wet field environment.

The degradability of composite cements also explains significant volumetric shrinkages and more variable soft tissue reactions. In the current study, the injected PMMA is a commercial product with proper designs for intraoperative premixed applications. Thus, its qualities are relatively stable, with reliable results. In contrast, the components of our newly designed cements, including CPC powder, Benzoyl Peroxide (BP) powder, PPF liquid, and N-Vinylpyrrolidone (N-VP) liquid are manufactured and packaged by the investigators, with limited volumes in different containers. After packaging, those components are also sterilized separately and stored for studies. During the premixing process, high percentages of the remnants of each component are attached to the inner side of the containers. This phenomenon may influence the assumed compositions, resulting in divergent data of the study groups. Attempts to solve the problem are made by setting large-scale component manufacturing equipment and standardizing the package and sterilization procedures.

Except for one with surgical errors, defects in the control group apparently heal, as shown by significant reductions in defect volume and increases in HU. Since the study has been performed on young porcine for reasons mentioned above, the control group is not compared with the cement groups.

Based on the results of combined radiologic and histologic comparisons in our previous study,[12] different CT scan imaging implies different gross and histologic findings. These findings make the above-mentioned CT scan method for analysis of animal model in bone filler study both feasible and meaningful. Individuals can be scanned chronologically to obtain a series of data with scarification. This CT scan method reduces labor and bias from scarification, histological preparations, and image processing which only revealed regional qualitative findings.

However, histological slices cannot be completely identical to CT scan slice images. Further delicate comparisons should be performed to verify this theory. By gross observation and CT scan, the appearances of composite blocks are more varied compared to the reliable results of nondegradable PMMA block morphologies. It can be hypothesized that both ceramic and polymeric components in cement are degraded after implantation, even if the size and morphology changes are expected.

This study also has some limitations, including the very short rearing period and the significant diversities between the two composite cements. There are also no chronologic evaluations of CT scan on each tested porcine. A series of imaging data of the same test subject, from the immediate postoperation to postmortem, can provide more accurate information for comparison. The second shortage was that young pigs had been used in the current study with significantly decreased volumes of defects of controls and increased HU numbers, indicating excellent growth and healing potentials of these young immature porcine which was different from the old and osteoporotic subjects met clinically.


  Conclusion Top


The CT imaging may be an alternative way for analyzing the animal model in bone filler study, especially since CT scanning can repeatedly, chronologically, and easily evaluate injectable bone cements and host bone reactions in vertebral bodies. In this in vivo porcine model, there are still a lot of adjustments and corrections needed to improve these novel fillers. Nonetheless, CT scan and 3D reconstruction may be an alternative way for analyzing animal models in bone filler study.

Financial support and sponsorship

Nil.

Conflicts of interest

There are no conflicts of interest.

 
  References Top

1.
Page JM, Harmata AJ, Guelcher SA. Design and development of reactive injectable and settable polymeric biomaterials. J Biomed Mater Res A 2013;101:3630-45.  Back to cited text no. 1
    
2.
Schröder C, Nguyen M, Kraxenberger M, Chevalier Y, Melcher C, Wegener B, et al. Modification of PMMA vertebroplasty cement for reduced stiffness by addition of normal saline: A material properties evaluation. Eur Spine J 2017;26:3209-15.  Back to cited text no. 2
    
3.
Aghyarian S, Hu X, Haddas R, Lieberman IH, Kosmopoulos V, Kim HK, et al. Biomechanical behavior of novel composite PMMA-CaP bone cements in an anatomically accurate cadaveric vertebroplasty model. J Orthop Res 2017;35:2067-74.  Back to cited text no. 3
    
4.
Zhu T, Ren H, Li A, Liu B, Cui C, Dong Y, et al. Novel bioactive glass based injectable bone cement with improved osteoinductivity and its in vivo evaluation. Sci Rep 2017;7:3622.  Back to cited text no. 4
    
5.
Zhu XS, Zhang ZM, Mao HQ, Geng DC, Zou J, Wang GL, et al. A novel sheep vertebral bone defect model for injectable bioactive vertebral augmentation materials. J Mater Sci Mater Med 2011;22:159-64.  Back to cited text no. 5
    
6.
Galovich LA, Perez-Higueras A, Altonaga JR, Orden JM, Barba ML, Morillo MT, et al. Biomechanical, histological and histomorphometric analyses of calcium phosphate cement compared to PMMA for vertebral augmentation in a validated animal model. Eur Spine J 2011;20 Suppl 3:376-82.  Back to cited text no. 6
    
7.
Sheng SR, Wang XY, Xu HZ, Zhu GQ, Zhou YF. Anatomy of large animal spines and its comparison to the human spine: A systematic review. Eur Spine J 2010;19:46-56.  Back to cited text no. 7
    
8.
Urrutia J, Bono CM, Mery P, Rojas C. Early histologic changes following polymethylmethacrylate injection (vertebroplasty) in rabbit lumbar vertebrae. Spine (Phila Pa 1976) 2008;33:877-82.  Back to cited text no. 8
    
9.
Togawa D, Kovacic JJ, Bauer TW, Reinhardt MK, Brodke DS, Lieberman IH, et al. Radiographic and histologic findings of vertebral augmentation using polymethylmethacrylate in the primate spine: Percutaneous vertebroplasty versus kyphoplasty. Spine (Phila Pa 1976) 2006;31:E4-10.  Back to cited text no. 9
    
10.
Manrique E, Chaparro D, Cebrián JL, López-Durán L.In vivo tricalcium phosphate, bone morphogenetic protein and autologous bone marrow biomechanical enhancement in vertebral fractures in a porcine model. Int Orthop 2014;38:1993-9.  Back to cited text no. 10
    
11.
Qin Y, Ye J, Wang P, Gao L, Jiang J, Wang S, et al. Evaluation of the biphasic calcium composite (BCC), a novel bone cement, in a minipig model of pulmonary embolism. J Biomater Sci Polym Ed 2016;27:317-26.  Back to cited text no. 11
    
12.
Wu CC, Hsu LH, Sumi S, Yang KC, Yang SH. Injectable and biodegradable composite bone filler composed of poly (propylene fumarate) and calcium phosphate ceramic for vertebral augmentation procedure: An in vivo porcine study. J Biomed Mater Res B Appl Biomater 2017;105:2232-43.  Back to cited text no. 12
    
13.
Grosfeld EC, Hoekstra JW, Herber RP, Ulrich DJ, Jansen JA, van den Beucken JJ, et al. Long-term biological performance of injectable and degradable calcium phosphate cement. Biomed Mater 2016;12:015009.  Back to cited text no. 13
    
14.
Wu CC, Tsai YF, Hsu LH, Chen JP, Sumi S, Yang KC, et al. A self-reinforcing biodegradable implant made of poly(ɛ-caprolactone)/calcium phosphate ceramic composite for craniomaxillofacial fracture fixation. J Craniomaxillofac Surg 2016;44:1333-41.  Back to cited text no. 14
    
15.
El Khassawna T, Daghma DE, Stoetzel S, Ray S, Kern S, Malhan D, et al. Postembedding decalcification of mineralized tissue sections preserves the integrity of implanted biomaterials and minimizes number of experimental animals. Biomed Res Int 2017;2017:2023853.  Back to cited text no. 15
    
16.
Wu CC, Yang KC, Yang SH, Lin MH, Kuo TF, Lin FH, et al. In vitro studies of composite bone filler based on poly (propylene fumarate) and biphasic α-tricalcium phosphate/hydroxyapatite ceramic powder. Artif Organs 2012;36:418-28.  Back to cited text no. 16
    
17.
Sariibrahimoglu K, An J, van Oirschot BA, Nijhuis AW, Eman RM, Alblas J, et al. Tuning the degradation rate of calcium phosphate cements by incorporating mixtures of polylactic-co-glycolic acid microspheres and glucono-delta-lactone microparticles. Tissue Eng Part A 2014;20:2870-82.  Back to cited text no. 17
    
18.
Alt V, Cheung WH, Chow SK, Thormann U, Cheung EN, Lips KS, et al. Bone formation and degradation behavior of nanocrystalline hydroxyapatite with or without collagen-type 1 in osteoporotic bone defects – An experimental study in osteoporotic goats. Injury 2016;47 Suppl 2:S58-65.  Back to cited text no. 18
    


    Figures

  [Figure 1], [Figure 2], [Figure 3], [Figure 4], [Figure 5], [Figure 6]



 

Top
 
 
  Search
 
Similar in PUBMED
   Search Pubmed for
   Search in Google Scholar for
 Related articles
Access Statistics
Email Alert *
Add to My List *
* Registration required (free)

 
  In this article
Abstract
Introduction
Materials and Me...
Results
Discussion
Conclusion
References
Article Figures

 Article Access Statistics
    Viewed584    
    Printed63    
    Emailed0    
    PDF Downloaded112    
    Comments [Add]    

Recommend this journal


[TAG2]
[TAG3]
[TAG4]