Journal of Biomaterials Applications http://jba.sagepub.com/ Effect of calcium phosphate-based fillers on the structure and bonding strength of novel gelatin− alginate bioadhesives Benny Cohen, Maoz Panker, Eyal Zuckerman, Maytal Foox and Meital Zilberman J Biomater Appl 2014 28: 1366 originally published online 31 October 2013 DOI: 10.1177/0885328213509502 The online version of this article can be found at: http://jba.sagepub.com/content/28/9/1366
Published by: http://www.sagepublications.com
Additional services and information for Journal of Biomaterials Applications can be found at: Email Alerts: http://jba.sagepub.com/cgi/alerts Subscriptions: http://jba.sagepub.com/subscriptions Reprints: http://www.sagepub.com/journalsReprints.nav Permissions: http://www.sagepub.com/journalsPermissions.nav Citations: http://jba.sagepub.com/content/28/9/1366.refs.html
>> Version of Record - Apr 17, 2014 OnlineFirst Version of Record - Oct 31, 2013 What is This?
Downloaded from jba.sagepub.com at UCSF LIBRARY & CKM on August 30, 2014
Article
Effect of calcium phosphate-based fillers on the structure and bonding strength of novel gelatin–alginate bioadhesives
Journal of Biomaterials Applications 2014, Vol. 28(9) 1366–1375 ! The Author(s) 2013 Reprints and permissions: sagepub.co.uk/journalsPermissions.nav DOI: 10.1177/0885328213509502 jba.sagepub.com
Benny Cohen, Maoz Panker, Eyal Zuckerman, Maytal Foox and Meital Zilberman
Abstract Interest in soft and hard tissue adhesives as alternatives for conventional wound closing and bone fixation applications has increased in recent decades as a result of numerous possible advantages such as better comfort and lower cost. A novel bioadhesive based on the natural polymers GA has recently been developed and studied in our laboratory. Hydroxyapatite and tricalcium phosphate are two bioactive ceramics known for their ability to enhance bone regeneration. In the current study, these two bioactive fillers were incorporated into the bioadhesive at concentrations of 0.125, 0.25 and 0.5% w/v, and their effects on the resulting adherence properties to soft and hard tissues were studied. Porcine skin and cortical portions of bovine femurs were used as soft and hard tissue specimens, respectively. The bonding strength was evaluated using an Instron universal testing machine in tensile mode, and the microstructure analysis was based on environmental scanning electron microscope observations. Both bioactive fillers were found to have a reinforcing effect on the adhesives, significantly improving their adhesion to soft tissues in certain concentrations. The best bonding strength results were obtained for 0.25% hydroxyapatite and 0.5% w/v tricalcium phosphate–18.1 4.0 and 15.2 2.6 kPa, respectively, compared with 8.4 2.3 kPa for adhesive with no fillers. The improved adherence is probably related to the stiffness of the insoluble hydroxyapatite and tricalcium phosphate particles which reinforce the adhesive. These particles can clearly be observed in the environmental scanning electron microscope analysis. The potential of these fillers to increase the bonding strength of the adhesive to hard tissues was also demonstrated. Hydroxyapatite and tricalcium phosphate thus improve our new gelatin–alginate bioadhesives, which can be used for both soft and hard tissue adhesive applications. Keywords Soft tissue adhesives, hard tissue adhesives, hydroxyapatite, tricalcium phosphate
Introduction Lacerations and traumatic wounds are considered to be among the most prevalent scenarios that are treated in hospitals and emergency rooms.1 Re-attachment of the margins of lacerated soft tissues is traditionally performed with sutures or staples. Several major advantages have led, in the past few decades, to increased interest in the use of tissue adhesives, that is substances that have the ability to firmly re-attach lacerated tissues, as an alternative to these conventional applications.2 These advantages are that tissue adhesives can be applied more rapidly and may require less equipment. Employment of tissue adhesives prevents the painful procedure involved in using sharp instruments and was proven to be less expensive, without compromising the cosmetic outcome.2,3 Fixing fractured hard tissues using an appropriate adhesive instead of the
traditional nailing and plating methods is also considered an attractive technique. The advantages include providing an optimal load transfer from one fracture side to the other, avoiding stress-shielding phenomena and the ability to repair small or thin bone fragments.4 Ideal soft and hard tissue adhesives have not yet been developed, even though extensive efforts have been made in the past, probably due to the various stiff requirements that a substance must fulfill in order to serve as a medical tissue adhesive for clinical use. The major demands include efficient bonding Department of Biomedical Engineering, Faculty of Engineering, Tel-Aviv University, Tel-Aviv, Israel Corresponding author: Meital Zilberman, Department of Biomedical Engineering, Faculty of Engineering, Tel-Aviv University, Tel-Aviv 69978, Israel. Email: [email protected]
Downloaded from jba.sagepub.com at UCSF LIBRARY & CKM on August 30, 2014
Cohen et al.
1367
strength to the tissue in a moist environment, sufficient biocompatibility to the tissue and to its surrounding, ease and convenience of application, sufficient flexibility for ensuring that it will remain adhered to the tissue, and penetrability to cell migration. The substance should preferably be biodegradable, stable during storage and economical.4,5 Nonetheless, several soft tissue adhesive products have been approved for medical use – cyanoacrylates, fibrin, and gelatin-based adhesives. However, these products were approved for restricted use only, due to biocompatibility problems (for cyanoacrylates and formaldehyde or glutaraldehyde which were traditionally used for crosslinking gelatin-based adhesives) or low mechanical strength (fibrin).6,7 In spite of the clear medical necessity, there is no hard tissue adhesive product that is available for clinical use to date.8 Gelatin is a natural water-soluble polymer that is derived from collagen. It has become one of the most extensively investigated materials for tissue adhesives, due to its suitable natural qualities. Gelatin is considered to be biocompatible, biodegradable, and nonimmunogenic.9 It can also form physically crosslinked hydrogel structures,10 has a natural tacky behavior in solution and is highly accessible in nature.11 All of these characteristics have turned gelatin not only into a promising candidate for tissue adhesives but also for a wide range of other medical applications, such as sealants,12 hydrogels,13 and microspheres.14 In spite of its promising qualities, the mechanical strength of physically crosslinked gelatin adhesives is not sufficient for using it as an adhering substance on its own.11 Therefore, a chemical crosslinking agent and a polymeric additive (with appropriate available functional groups for the crosslinking reaction) were usually added to the gelatin solution in a wide range of published attempts, in order to create gelatin-based hydrogel formulations with mechanical properties suitable for soft tissue adhesion.11,15–18 A novel tissue adhesive based on a combination of gelatin with alginate as a polymeric additive, crosslinked by carbodiimide,19,20 was recently developed and studied by our research group. Carbodiimide, which is mainly used for modification and conjunction of proteins and other biological macrostructures, was chosen as the crosslinking agent since carbodiimides and their crosslinking byproducts have been reported as being less cytotoxic than other conventional crosslinking agents such as formaldehyde and glutaraldehyde.21 Alginate is a natural polysaccharide which is extracted from marine algae and is used extensively in the food and beverage industry as a gelling agent, stabilizer, and emulsifier.22 It is also applied in the medical and the pharmaceutical industries as a drug-delivery vehicle,23 a dental impression material,24 part of a
synthetic extracellular matrix for cell immobilization25 and for wound dressing.26 The main advantages of this new tissue adhesive is that it combines good strength with high biocompatibility, adequate viscosity before curing, and proper curing time.20,21 In the current study, two types of bioactive fillers – hydroxyapatite (HA) and tricalcium phosphate (TCP) – were loaded into the adhesive in order to investigate their effects on the structure and properties of our novel bioadhesive. These two fillers were chosen not only due to their good biocompatibility and lack of toxicity but also due to their ability to promote bone formation and osseointegration, which can be attractive qualities for hard tissue adhesives. Both HA and TCP are considered insoluble in water. However, the solubility of TCP tends to be higher than HA (0.0005 gr/L as opposed to 0.0003 gr/L, respectively).27 The effect of loading both HA and TCP into several other bone adhesives and cements was investigated and the results have shown that they improve the mechanical and adherence properties of the adhesives and promote bone growth.28–33 In the current study, we examined the ability of the bioactive fillers HA and TCP to improve the in vitro adhering properties of our gelatin–alginate-based adhesives to both soft and hard tissues.
Materials and methods Materials Gelatin ‘‘type A’’ from porcine skin (90–100 bloom), alginic acid sodium salt (viscosity 250 cP, 2% (25 C)(lit)), N-(3-dimethylaminopropyl)-N0 ethylcarbodiimide hydrochloride (EDC), HA (powder, synthetic, average particle size: 50 mm), and b-TCP (sintered powder, average particle size: 10 mm) were purchased from Sigma-Aldrich, Rehovot, Israel.
Preparation of the filler-loaded adhesives Preparation of the adhesives was based on dissolving gelatin and alginate (GA) in distilled water, at constant concentrations of 200 and 40 mg/mL, respectively, at 60 C. HA and TCP fillers powders were mixed with the GA solutions at concentrations of 0.125, 0.25 and 0.5% w/v (GAF). The entire mixing process lasted 30–45 min. In the relatively low concentrations range that was chosen, both HA and b-TCP, which are basically insoluble in water, could still be loaded inside the adhesive hydrogel with only minimal precipitation of each filler. The chosen constant concentrations of GA were based on previous results, which show that these concentrations enables a relatively high in vitro bonding strength and easy handling.20 Immediately prior to the
Downloaded from jba.sagepub.com at UCSF LIBRARY & CKM on August 30, 2014
1368
Journal of Biomaterials Applications 28(9)
Table 1. The studied HA and TCP-loaded formulations for soft and hard tissue adhesion.
For in vitro soft tissue bonding strength measurements
Reference adhesive HA-loaded adhesive
Gelatin concentration (mg/mL)
Alginate concentration (mg/mL)
EDC concentration (mg/mL)
Filler concentration (% w/v)
200 200
40 40
20 20
0 0.125 0.25 0.5 0.125 0.25 0.5
10
TCP-loaded adhesive
200
40
20
10
For in vitro hard tissue bonding strength measurements
HA-loaded adhesive TCP-loaded adhesive
200 200
40 40
20 20
0.125 0.25 0.5 0.125 0.25 0.5 0.25 0.5
HA: hydroxyapatite; TCP: tricalcium phosphate.
adhesive’s use, the crosslinking agent (EDC) was added to the various GAF solutions at concentrations of 10 and 20 mg/mL. All studied formulations are presented in Table 1.
In vitro bonding strength measurements Soft tissue measurements. Porcine skin (Kibbutz Lahav, Israel) was used as a soft tissue model for evaluating the effect of the fillers on the bonding strength of the adhesive. The porcine skin was cut into 2 2 cm square-shaped pieces and their epidermis side was attached firmly to metal testing holders with a matching surface area (all dimensions of the holders are specified in Figure 1). About 140 mL of the adhesive was then spread uniformly on the dermis side of two porcine skin pieces (that were attached to the testing holders) which were immediately attached by applying a 1.25 N load on the pieces and placed in a 37 C and 100% humidity environment. After 30 min, the bonding strength was measured in tension mode at room temperature using a 5500 Instron universal testing machine (Instron Engineering Corp., Massachusetts, USA) and a 10 N load cell. The two parts of the joint were strained at a constant velocity of 2 mm/min until separation was achieved. The mechanical testing procedure was inspired by the standard test method ASTM F-2258-03. The bonding strength was defined as the maximum strength in the stress–strain curve, measured by the Instron Merlin software (Massachusetts, USA).
Hard tissue measurements. The cortical portions of bovine femurs (purchased at a local abattoir) were used as a hard tissue model in order to evaluate the fillers’ effect on the bonding strength of the adhesive. The femurs were sawed into 2 2 0.2 cm rectangular cuboid specimens using an ‘‘FMB’’ (Dalmine BG, Italy) minor portable band saw. The specimens were firmly attached to metal testing holders with a matching surface area. The entire bonding system was based on the bonding system for the soft tissue adhesion (Figure 1) and has similar dimensions. About 140 mL of the adhesive (with various concentrations of the fillers) was then spread uniformly on the exposed side of two femur specimens. The specimens were then immediately attached to each other by applying a load of 7.5 N and placed in a 37 C and 100% humidity environment. After 30 min, the bonding strength was measured in tension mode at room temperature using a 5500 Instron universal testing machine (Instron Engineering Corp.) and a 2 kN load cell. The two parts of the joint were strained at a constant velocity of 2 mm/min until separation was achieved. The mechanical testing procedure was inspired by the standard test method ASTM F-2258-03. The bonding strength was defined as the maximum strength in the stress– strain curve, measured by the Instron Merlin software.
Microstructural characterization About 420 mL of the adhesive was injected into 24 24 3 mm rectangular silicone molds. After initial
Downloaded from jba.sagepub.com at UCSF LIBRARY & CKM on August 30, 2014
Cohen et al.
1369
Figure 1. Illustration of the bonding system for soft tissue adhesion measurements. Specimen 2 2 0.2 cm, applied load 7.5 Nm at 37 C and 100% humidity environment for 30 min. The bonding strength was measured in tension mode at room temperature using a 5500 Instron universal testing machine (Instron Engineering Corp.) and a 2 kN load cell, velocity 2 mm/min (according to standard test method ASTM F-2258-03). Five repetitions were carried out for each formulation.
curing, the rectangular cuboid-shaped adhesive specimens were pressed out of their silicone molds, airdried in a chemical hood, freeze-fractured, and their cross-section was observed using an environmental scanning electron microscope (ESEM; FEI Quanta 200 FEG, Oregon, USA) in a high vacuum mode, with an accelerating voltage of 10 kV.
Statistical analysis Statistical comparison between more than two groups was carried out using the analysis of variance (Tukey Kramer) method via ‘‘IBM’’ (New York, USA) SPSS (V. 15) software. A value of p < 0.05 was considered statistically significant.
Results In vitro soft tissue bonding strength In order to evaluate the effect of HA and TCP on the bonding strength to soft tissues was measured and compared with the bonding strength of adhesive not loaded
with filler (control group). The concentrations of gelatin, alginate, and EDC were kept constant – 200, 40, and 20 mg/mL, respectively. Fifteen repetitions were carried out for each formulation. The results are presented in Figure 2. Incorporation of both HA and TCP was found to improve the bonding strength of the adhesive. Improvement of the bonding strength with HA was effective at concentrations of 0.25% w/v and above (18.1 4.0 kPa vs. 8.4 2.3 kPa for the 0.25% w/v HA loaded and the non-loaded adhesive, respectively). Although adhesive loaded with 0.5% w/v HA exhibited a significant lower bonding strength than adhesive with 0.25% HA (13.4 1.6 kPa), as stated, it was still higher than the bonding strength of the nonloaded adhesive. On the other hand, all examined TCP concentrations were found to improve the bonding strength, but no significant difference was observed between the different concentrations. The highest average value was measured for a TCP concentration of 0.5% w/v (15.2 2.6 kPa). A relatively low concentration of the crosslinking agent is beneficial when high biocompatibility is required. We therefore decided to also examine the
Downloaded from jba.sagepub.com at UCSF LIBRARY & CKM on August 30, 2014
1370
Journal of Biomaterials Applications 28(9)
Figure 2. Effect of HA (a) and TCP (b) on the bonding strength of bioadhesives composed of gelatin (200 mg/mL), alginate (40 mg/mL), and EDC (20 mg/mL) to soft tissues. Significant differences are marked with ‘‘*’’. HA: hydroxyapatite; TCP: tricalcium phosphate.
Figure 3. Effect of HA (a) and TCP (b) on the bonding strength of bioadhesives composed of gelatin (200 mg/mL), alginate (40 mg/mL), and a reduced concentration of EDC (10 mg/mL) to soft tissues, compared with the reference formulation. Significant differences are marked with ‘‘*’’. HA: hydroxyapatite; TCP: tricalcium phosphate.
effect of the fillers on the bonding strength of an adhesive with a reduced concentration of EDC (10 mg/mL). The same three concentrations of the fillers were loaded (0.125, 0.25, and 0.5% w/v) and the concentrations of the polymeric components were kept constant as before. Fifteen repetitions were carried out for each formulation. The results are presented in Figure 3. As can be seen from these results, HA and TCP-loaded adhesives with a reduced concentration of EDC had bonding strengths similar to the bonding strength of the nonloaded adhesive with non-reduced EDC concentration (20 mg/mL), and specific concentrations (0.5 and 0.125% w/v for HA and TCP, respectively) were found to have even higher bonding strength values.
chosen for the hard tissue bonding strength experiment. We assumed that formulations which perform best for soft tissues will also perform best for hard tissues. Hence, 0.25% w/v HA and 0.5% w/v TCP were loaded into the adhesive and the bonding strength results were compared with a non-loaded adhesive. The concentrations of the polymeric components were kept constant as before. The EDC concentration was set to 20 mg/mL. Three repetitions were carried out for each formulation. The results are presented in Figure 4. Loading 0.25% w/v HA almost tripled the bonding strength of the adhesive from 26.6 9.2 kPa to 71.4 28.2 kPa. No significant improvement in the bonding strength was observed for the 0.5% w/v TCP-loaded adhesive (24.0 13.0 kPa).
In vitro hard tissue bonding strength The HA and TCP-loaded adhesives were also examined for their potential use in hard tissue adhesion. One concentration of each filler, based on the best results of the previous bonding strength tests for soft tissues, was
Microstructural analysis The bulk cross-sections of air-dried adhesive specimens loaded with 0.125 and 0.5% w/v HA and TCP were examined using ESEM. The concentrations of gelatin
Downloaded from jba.sagepub.com at UCSF LIBRARY & CKM on August 30, 2014
Cohen et al.
1371
Figure 4. Effect of HA and TCP on the bonding strength of selected bioadhesives composed of gelatin (200 mg/mL), alginate (40 mg/mL), and EDC (20 mg/mL) to hard tissue (bone). Significant differences are marked with ‘‘*’’. HA: hydroxyapatite; TCP: tricalcium phosphate.
(200 mg/mL), alginate (40 mg/mL), and EDC (20 mg/mL) were kept constant. Fractographs of the HA and TCP-loaded adhesives are presented in Figure 5. The fractographs provide a clear perspective on the crystallization of the fillers in the adhesive matrix. The HA particles exhibit a 50–60 mm diameter round core and a multi-shell structure, while the TCP particles are organized as 5–20 mm diameter aggregates. The fractographs also demonstrate that at the lower concentration, the HA and TCP particles are not uniformly dispersed in the adhesive. Higher concentrations of the fillers seem to provide a more uniform dispersion. It should be mentioned that the cracks in the crosssection of the specimens result from the fracturing process.
Discussion The soft tissue bonding strength experiments demonstrated that both HA and TCP can be regarded as effective reinforcement fillers, since loading them into the adhesive improves its bonding strength. The improvement in the bonding strength is probably achieved due to the stiffness of the insoluble HA and TCP particles which restrain some of the movement of the adhesive’s strings in their vicinity. As a result, the mechanical properties of the adhesive are improved and the bonding strength is therefore also improved. However, the HA and TCP concentration had a different effect on the bonding strength improvement (for an EDC concentration of 20 mg/mL). The TCP-loaded adhesives exhibited similar bonding strength values for all three examined concentrations, which is significantly higher than the bonding strength of the nonloaded adhesive. On the other hand, while it was still significantly higher than the non-loaded adhesive for both concentrations, increasing the HA concentration
from 0.25 to 0.5% w/v decreased the bonding strength. Loading lower concentrations of HA (0.125% w/v) was found to have no significant effect on the bonding strength of the adhesive. ESEM examination of the HA and TCP particles’ characteristics and dispersion inside the adhesive can provide a possible explanation for this phenomenon. Previous microstructural analysis of a non-loaded adhesive has shown that no phase separation occurs in the presence of only the three basic components of the adhesive – gelatin, alginate, and EDC.20 However, the incorporation of HA and TCP in the adhesive can clearly be noticed. Since HA is practically insoluble in aqueous solutions, its dispersion inside the water-based adhesive is not uniform. As can be seen from the ESEM fractographs (Figure 5(a) and (b)), the non-uniformity is more significant at the lower concentrations, where the density of the particles inside the adhesive is lower. As a result, no significant improvement in the bonding strength was observed for the lowest examined HA concentration. TCP is also insoluble in aqueous solutions. However, since its particles are significantly smaller than the HA particles (diameter of 5–20 mm compared with 50–60 mm, respectively), the dispersion of the TCP particles inside the adhesive matrix is likely to be more uniform than the dispersion of the HA particles at every given concentration. This is probably why an improvement in the bonding strength for TCP could be achieved even at a concentration of 0.125% w/v (although a uniform dispersion could hardly be noticed in the ESEM fractograph of that TCP concentration). The unique core and shells structure of the HA particles (Figure 5(c)) indicates that the surface of the HA particle probably reacts with the adhesive matrix either chemically or physically, which causes the shells’ formation. It is known from the literature that HA molecules have a high affinity to proteins when dispersed in an aqueous environment. This phenomenon occurs due to the dissolution of negative hydroxyl ions from the surface of the HA to the surrounding, exposing positive calcium charges which tend to bind to the proteins’ acidic groups, for example, carboxylic groups.34 The ability of the HA’s exposed positive calcium charges to bind to carboxylic groups originating from the GA may be the cause for a noticeable decrease in the bonding strength of the adhesive from a certain HA-loading concentration. The binding of positive charges originating from the HA particles’ surface to the gelatin and probably also to the alginate carboxylic groups decreases the density of the carboxylic groups which remain available for crosslinking of the adhesive in the close surrounding of the particles. As a result, a mechanical failure can develop from these areas, due to the lower crosslinking density. Evidence that supports this theory can be found in Figure 5(c), where it
Downloaded from jba.sagepub.com at UCSF LIBRARY & CKM on August 30, 2014
1372
Journal of Biomaterials Applications 28(9)
Figure 5. ESEM fractographs of adhesives (gelatin 200 mg/mL; alginate 40 mg/mL; and EDC 20 mg/mL) loaded with 0.125% w/ v HA (a), 0.5% w/v HA (b, c), 0.125% w/v TCP (d) and 0.5% w/v TCP (e, f). HA: hydroxyapatite; TCP: tricalcium phosphate; ESEM: environmental scanning electron microscope.
Downloaded from jba.sagepub.com at UCSF LIBRARY & CKM on August 30, 2014
Cohen et al.
1373
can be seen that cracks (originally from the specimens fracturing process) have a tendency to progress through the shells and stop at the core of the HA particle. This finding gives another indication for the relatively poor mechanical properties of the shells as opposed to the core and the matrix. Increasing the HA concentration increases not only the HA particles’ density but also the density of these weak mechanical areas, and it seems that the effect of these ‘‘weak’’ areas becomes more dominant than the effect of the particles from a certain HA concentration, and the bonding strength of the adhesive therefore begins to decrease. TCP, in contradistinction to HA, does not contain chemical groups that may dissolve in an aqueous environment and expose positive or negative charges on its surface. Therefore, the phenomenon of shells creation around its particles was not observed in the ESEM fractographs (Figure 5(f)), and no decrease in the bonding strength was observed for the higher loading concentrations. On the other hand, neither could an increase in the bonding strength be achieved when increasing the TCP concentration. This phenomenon may be related to the fact that due to the negligible solubility of TCP, its loading concentration range was not wide enough for achieving a noticeable consistent improvement in the bonding strength while increasing its concentration. Another possible reason can be the TCP particle size, which is relatively small compared with HA, and therefore does not enable a consistent increase in the bonding strength with the increase in its loading concentration. HA, which exhibited larger particles, improved the bonding strength of the adhesive to soft tissues more significantly than TCP. The crosslinking agent concentration has a significant effect on the bonding strength since it controls the crosslinking density in the adhesive. In a previous research, we saw that decreasing the crosslinking agent concentration by half, from 20 to 10 mg/mL, caused an almost two-fold reduction in the bonding strength of an adhesive with the same formulation.19 Comparing the bonding strength results of HA and TCP-loaded adhesives with a reduced concentration of EDC (10 mg/mL) to that of the non-loaded reference adhesive (20 mg/mL EDC) indicated that both HA and TCP have a compensatory effect on the bonding strength when decreasing the crosslinking agent concentration. Since EDC was shown to have some cytotoxic effects,19,20 this compensatory effect also has clinical and medical importance because it enables improving the biocompatibility of the adhesive without compromising its bonding strength and even improving it. The bonding strength of EvicelTM (commercial fibrin glue) was recently measured in our lab using the same bonding strength system.19 A comparison between the bonding strength of EvicelTM (2.5 2.3 kPa) and our novel HA and
TCP-loaded bioadhesives shows that even when the EDC concentration is reduced by half, incorporation of these fillers enables the achievement of up to 7 times higher bonding strength to soft tissues than with EvicelTM. Higher bonding strength values which reached more than 70 KPa were obtained when bonded to hard tissue. In the latter’s case, the contribution of the chemical adsorption mechanism is increased when HA is added to the bioadhesive. It should be noted that for the reduced EDC concentration (10 mg/mL) adhesives, the effect of the HA concentration on the bonding strength is different from its effect on the bonding strength of adhesive with the non-reduced EDC concentration. The reason for this might be the fact that although carboxylic groups originally from the GA still react with the positive charges on the surface of the HA particles, there are sufficient available groups left for reacting with all the EDC, due to its reduced concentration. Therefore, the effect of domains with lower mechanical properties around the HA particles on the bonding strength may be negligible. Due to the difficulty of preparing consistent and uniform hard tissue specimens, the bonding strength of HA and TCP-loaded adhesives on hard tissues was examined for only one concentration of each filler (0.25 and 0.5% w/v for HA and TCP, respectively) and for a non-loaded adhesive as control. For the same reason, only three repetitions were carried out for each formulation. A significant improvement in the bonding strength of the adhesive to hard tissues was observed only for the HA-loaded adhesive (Figure 4). Incorporation of TCP had practically no effect on the bonding strength to hard tissue. The adherence ability of other bioadhesives to hard tissues was evaluated and it was found that under similar test conditions, fibrin and gelatin-resorcinol-formaldehyde exhibit in vitro bonding strengths of 11 and 200 kPa, respectively.35 Our HA-loaded adhesive exhibits a bonding strength to hard tissues that is 6.5 times higher than that of fibrin (71.4 28.2 kPa), but is lower than that of the gelatin-resorcinol-formaldehyde adhesive.35 However, the latter is considered to be less biocompatible than our adhesive, since it is based on formaldehyde as the crosslinking agent, which is significantly more cytotoxic than EDC.21 In conclusion, in the current study we examined the effect of loading the bioactive fillers HA and TCP into our novel gelatin–alginate bioadhesives on the latter’s adherence properties. Our thought was that incorporating such fillers, which are known for their ability to promote bone formation and osseointegration, will enable us to extend the use of this bioadhesive to hard tissue adhesion. Incorporating both HA and TCP was found to significantly improve the bonding strength of the bioadhesive to soft tissues.
Downloaded from jba.sagepub.com at UCSF LIBRARY & CKM on August 30, 2014
1374
Journal of Biomaterials Applications 28(9)
Combining these fillers in the bioadhesive can also improve its biocompatibility, since their presence enables decreasing the concentration of the relatively cytotoxic crosslinking agent without compromising the mechanical properties. The potential of these fillers to improve the bonding strength of the adhesive to hard tissues was also demonstrated. Thus, HA and TCP are beneficial to our new gelatin–alginate bioadhesives, which can be used for both soft and hard tissue adhesives applications. Acknowledgments The authors are grateful to the Office of the Chief Scientist in the Israel Ministry of Industry, Trade and Labor, and Kodesz Foundation (Tel-Aviv University) for supporting this research.
Conflict of interest None declared.
References 1. Sibert JR, Maddocks GB and Brown BM. Childhood accidents – an endemic of epidemic proportion. Arch Dis Child 1981; 56: 225–227. 2. Osmond MH. Wound repair and tissue adhesives. In: Moyer V, Elliott E and Davis R (eds) Evidencebased Pediatrics and Child Health. 1 ed. London: BMJ Books, 2000. 3. Singer AJ, Kinariwala M, Lirov R, et al. Patterns of use of topical skin adhesives in the emergency department. Acad Emerg Med 2010; 17: 670–672. 4. Kraus R, Schluckebier D, Stiller A-C, et al. Bone adhesives in trauma and orthopedic surgery. Eur J Trauma 2006; 32: 141–148. 5. Donkerwolcke M, Burny F and Muster D. Tissues and bone adhesives—historical aspects. Biomaterials 1998; 19: 1461–1466. 6. Reece TB, Maxey TS and Kron IL. A prospectus on tissue adhesives. Am J Surg 2001; 182: S40–S44. 7. Ryou M and Thompson CC. Tissue adhesives: a review. Techniques Gastrointes Endosc 2006; 8: 33–37. 8. Farrar DF. Bone adhesives for trauma surgery: a review of challenges and developments. Int J Adhes Adhesiv 2011. 9. Kuijpers AJ, Engbers GHM, Feijen J, et al. Characterization of the network structure of carbodiimide cross-linked gelatin gels. Macromolecules 1999; 32: 3325–3333. 10. Djabourov M and Papon P. Influence of thermal treatments on the structure and stability of gelatin gels. Polymer 1983; 24: 537–542. 11. Bae SK, Sung TH and Kim JD. A soft-tissue gelatin bioadhesive reinforced with a proteinoid. J Adhes Sci Technol 2002; 16: 361–372. 12. Jonas RA, Ziemer G, Schoen FJ, et al. A new sealant for knitted Dacron prostheses: minimally cross-linked gelatin. J Vas Surg 1988; 7: 414–419.
13. Tabata Y, Hijikata S and Ikada Y. Enhanced vascularization and tissue granulation by basic fibroblast growth factor impregnated in gelatin hydrogels. J Control Rel 1994; 31: 189–199. 14. Narayani R and Rao KP. Controlled release of anticancer drug methotrexate from biodegradable gelatin microspheres. J Microencapsul 1994; 11: 69–77. 15. Otani Y, Tabata Y and Ikada Y. A new biological glue from gelatin and poly (L-glutamic acid). J Biomed Mater Res 1996; 31: 157–166. 16. Sung HW, Huang DM, Chang WH, et al. Evaluation of gelatin hydrogel crosslinked with various crosslinking agents as bioadhesives: in vitro study. J Biomed Mater Res 1999; 46: 520–530. 17. Otani Y, Tabata Y and Ikada Y. Rapidly curable biological glue composed of gelatin and poly (L-glutamic acid). Biomaterials 1996; 17: 1387–1391. 18. Matsuda S, Iwata H, Se N, et al. Bioadhesion of gelatin films crosslinked with glutaraldehyde. J Biomed Mater Res 1999; 45: 20–27. 19. Shefy-Peleg A, Foox M, Cohen B and Zilberman M. Novel Antibiotic-eluting gelatin-alginate soft tissue adhesives for various wound closing applications. Int J Poly Mater and Poly Biomater (in press). 20. Cohen B, Shefy-Peleg A and Zilberman M. Gelatin-based soft tissue adhesive loaded with analgesic drugs: structure and properties. J Biomater Sci, Polym Ed (in press). 21. Nishi C, Nakajima N and Ikada Y. In vitro evaluation of cytotoxicity of diepoxy compounds used for biomaterial modification. J Biomed Mater Res 1995; 29: 829–834. 22. Draget KI, Smidsrød O and Skjak-Brak G. Alginates from algae. Biopolym Online 2005: 215–244. 23. Liew CV, Chan LW, Ching AL, et al. Evaluation of sodium alginate as drug release modifier in matrix tablets. Int J Pharm 2006; 309: 25–37. 24. Ashley M, McCullagh A and Sweet C. Making a good impression: (a ‘how to’ paper on dental alginate). Dent Update 2005; 32: 169. 25. Smidsrød O and Skjaok-Brk G. Alginate as immobilization matrix for cells. Trends Biotechnol 1990; 8: 71–78. 26. Matthew IR, Browne RM, Frame JW, et al. Subperiosteal behaviour of alginate and cellulose wound dressing materials. Biomaterials 1995; 16: 275–278. 27. Dorozhkin SV. Calcium orthophosphates: occurrence, properties, biomineralization, pathological calcification and biomimetic applications. Biomatter 2011; 1: 121–164. 28. Harper EJ, Behiri JC and Bonfield W. Flexural and fatigue properties of a bone cement based upon polyethylmethacrylate and hydroxyapatite. J Mater Sci: Mater Med 1995; 6: 799–803. 29. Liebendo¨rfer A, Schmitz B, Wenz R, et al. Experimental studies on a new bone cement: hydroxyapatite composite resin. In: The 21st annual meeting of the society for biomaterials, San Francisco, USA, 1995, p. 335. 30. Lee HB, Khang G, Lee JH, et al. Polymeric biomaterials. Boca Raton, FL: CRC Press, 2003. 31. Gellynck K, Abou Neel EA, Li HY, et al. Cell attachment and response to photocured, degradable bone
Downloaded from jba.sagepub.com at UCSF LIBRARY & CKM on August 30, 2014
Cohen et al.
1375
adhesives containing tricalcium phosphate and purmorphamine. Acta Biomater 2011; 7: 2672–2677. 32. Gerhart TN, Renshaw AA, Miller RL, et al. In vivo histologic and biomechanical characterization of a biodegradable particulate composite bone cement. J Biomed Mater Res 1989; 23: 1–16. 33. Collinge C, Merk B and Lautenschlager EP. Mechanical evaluation of fracture fixation augmented with tricalcium phosphate bone cement in a porous osteoporotic
cancellous bone model. J Orthopaed Trauma 2007; 21: 124–128. 34. Kandori K, Miyagawa K and Ishikawa T. Adsorption of immunogamma globulin onto various synthetic calcium hydroxyapatite particles. J Colloid Interf Sci 2004; 273: 406–413. 35. Chivers RA and Wolowacz RG. The strength of adhesive-bonded tissue joints. Int J Adhes Adhesiv 1997; 17: 127–132.
Downloaded from jba.sagepub.com at UCSF LIBRARY & CKM on August 30, 2014
Published by: http://www.sagepublications.com
Additional services and information for Journal of Biomaterials Applications can be found at: Email Alerts: http://jba.sagepub.com/cgi/alerts Subscriptions: http://jba.sagepub.com/subscriptions Reprints: http://www.sagepub.com/journalsReprints.nav Permissions: http://www.sagepub.com/journalsPermissions.nav Citations: http://jba.sagepub.com/content/28/9/1366.refs.html
>> Version of Record - Apr 17, 2014 OnlineFirst Version of Record - Oct 31, 2013 What is This?
Downloaded from jba.sagepub.com at UCSF LIBRARY & CKM on August 30, 2014
Article
Effect of calcium phosphate-based fillers on the structure and bonding strength of novel gelatin–alginate bioadhesives
Journal of Biomaterials Applications 2014, Vol. 28(9) 1366–1375 ! The Author(s) 2013 Reprints and permissions: sagepub.co.uk/journalsPermissions.nav DOI: 10.1177/0885328213509502 jba.sagepub.com
Benny Cohen, Maoz Panker, Eyal Zuckerman, Maytal Foox and Meital Zilberman
Abstract Interest in soft and hard tissue adhesives as alternatives for conventional wound closing and bone fixation applications has increased in recent decades as a result of numerous possible advantages such as better comfort and lower cost. A novel bioadhesive based on the natural polymers GA has recently been developed and studied in our laboratory. Hydroxyapatite and tricalcium phosphate are two bioactive ceramics known for their ability to enhance bone regeneration. In the current study, these two bioactive fillers were incorporated into the bioadhesive at concentrations of 0.125, 0.25 and 0.5% w/v, and their effects on the resulting adherence properties to soft and hard tissues were studied. Porcine skin and cortical portions of bovine femurs were used as soft and hard tissue specimens, respectively. The bonding strength was evaluated using an Instron universal testing machine in tensile mode, and the microstructure analysis was based on environmental scanning electron microscope observations. Both bioactive fillers were found to have a reinforcing effect on the adhesives, significantly improving their adhesion to soft tissues in certain concentrations. The best bonding strength results were obtained for 0.25% hydroxyapatite and 0.5% w/v tricalcium phosphate–18.1 4.0 and 15.2 2.6 kPa, respectively, compared with 8.4 2.3 kPa for adhesive with no fillers. The improved adherence is probably related to the stiffness of the insoluble hydroxyapatite and tricalcium phosphate particles which reinforce the adhesive. These particles can clearly be observed in the environmental scanning electron microscope analysis. The potential of these fillers to increase the bonding strength of the adhesive to hard tissues was also demonstrated. Hydroxyapatite and tricalcium phosphate thus improve our new gelatin–alginate bioadhesives, which can be used for both soft and hard tissue adhesive applications. Keywords Soft tissue adhesives, hard tissue adhesives, hydroxyapatite, tricalcium phosphate
Introduction Lacerations and traumatic wounds are considered to be among the most prevalent scenarios that are treated in hospitals and emergency rooms.1 Re-attachment of the margins of lacerated soft tissues is traditionally performed with sutures or staples. Several major advantages have led, in the past few decades, to increased interest in the use of tissue adhesives, that is substances that have the ability to firmly re-attach lacerated tissues, as an alternative to these conventional applications.2 These advantages are that tissue adhesives can be applied more rapidly and may require less equipment. Employment of tissue adhesives prevents the painful procedure involved in using sharp instruments and was proven to be less expensive, without compromising the cosmetic outcome.2,3 Fixing fractured hard tissues using an appropriate adhesive instead of the
traditional nailing and plating methods is also considered an attractive technique. The advantages include providing an optimal load transfer from one fracture side to the other, avoiding stress-shielding phenomena and the ability to repair small or thin bone fragments.4 Ideal soft and hard tissue adhesives have not yet been developed, even though extensive efforts have been made in the past, probably due to the various stiff requirements that a substance must fulfill in order to serve as a medical tissue adhesive for clinical use. The major demands include efficient bonding Department of Biomedical Engineering, Faculty of Engineering, Tel-Aviv University, Tel-Aviv, Israel Corresponding author: Meital Zilberman, Department of Biomedical Engineering, Faculty of Engineering, Tel-Aviv University, Tel-Aviv 69978, Israel. Email: [email protected]
Downloaded from jba.sagepub.com at UCSF LIBRARY & CKM on August 30, 2014
Cohen et al.
1367
strength to the tissue in a moist environment, sufficient biocompatibility to the tissue and to its surrounding, ease and convenience of application, sufficient flexibility for ensuring that it will remain adhered to the tissue, and penetrability to cell migration. The substance should preferably be biodegradable, stable during storage and economical.4,5 Nonetheless, several soft tissue adhesive products have been approved for medical use – cyanoacrylates, fibrin, and gelatin-based adhesives. However, these products were approved for restricted use only, due to biocompatibility problems (for cyanoacrylates and formaldehyde or glutaraldehyde which were traditionally used for crosslinking gelatin-based adhesives) or low mechanical strength (fibrin).6,7 In spite of the clear medical necessity, there is no hard tissue adhesive product that is available for clinical use to date.8 Gelatin is a natural water-soluble polymer that is derived from collagen. It has become one of the most extensively investigated materials for tissue adhesives, due to its suitable natural qualities. Gelatin is considered to be biocompatible, biodegradable, and nonimmunogenic.9 It can also form physically crosslinked hydrogel structures,10 has a natural tacky behavior in solution and is highly accessible in nature.11 All of these characteristics have turned gelatin not only into a promising candidate for tissue adhesives but also for a wide range of other medical applications, such as sealants,12 hydrogels,13 and microspheres.14 In spite of its promising qualities, the mechanical strength of physically crosslinked gelatin adhesives is not sufficient for using it as an adhering substance on its own.11 Therefore, a chemical crosslinking agent and a polymeric additive (with appropriate available functional groups for the crosslinking reaction) were usually added to the gelatin solution in a wide range of published attempts, in order to create gelatin-based hydrogel formulations with mechanical properties suitable for soft tissue adhesion.11,15–18 A novel tissue adhesive based on a combination of gelatin with alginate as a polymeric additive, crosslinked by carbodiimide,19,20 was recently developed and studied by our research group. Carbodiimide, which is mainly used for modification and conjunction of proteins and other biological macrostructures, was chosen as the crosslinking agent since carbodiimides and their crosslinking byproducts have been reported as being less cytotoxic than other conventional crosslinking agents such as formaldehyde and glutaraldehyde.21 Alginate is a natural polysaccharide which is extracted from marine algae and is used extensively in the food and beverage industry as a gelling agent, stabilizer, and emulsifier.22 It is also applied in the medical and the pharmaceutical industries as a drug-delivery vehicle,23 a dental impression material,24 part of a
synthetic extracellular matrix for cell immobilization25 and for wound dressing.26 The main advantages of this new tissue adhesive is that it combines good strength with high biocompatibility, adequate viscosity before curing, and proper curing time.20,21 In the current study, two types of bioactive fillers – hydroxyapatite (HA) and tricalcium phosphate (TCP) – were loaded into the adhesive in order to investigate their effects on the structure and properties of our novel bioadhesive. These two fillers were chosen not only due to their good biocompatibility and lack of toxicity but also due to their ability to promote bone formation and osseointegration, which can be attractive qualities for hard tissue adhesives. Both HA and TCP are considered insoluble in water. However, the solubility of TCP tends to be higher than HA (0.0005 gr/L as opposed to 0.0003 gr/L, respectively).27 The effect of loading both HA and TCP into several other bone adhesives and cements was investigated and the results have shown that they improve the mechanical and adherence properties of the adhesives and promote bone growth.28–33 In the current study, we examined the ability of the bioactive fillers HA and TCP to improve the in vitro adhering properties of our gelatin–alginate-based adhesives to both soft and hard tissues.
Materials and methods Materials Gelatin ‘‘type A’’ from porcine skin (90–100 bloom), alginic acid sodium salt (viscosity 250 cP, 2% (25 C)(lit)), N-(3-dimethylaminopropyl)-N0 ethylcarbodiimide hydrochloride (EDC), HA (powder, synthetic, average particle size: 50 mm), and b-TCP (sintered powder, average particle size: 10 mm) were purchased from Sigma-Aldrich, Rehovot, Israel.
Preparation of the filler-loaded adhesives Preparation of the adhesives was based on dissolving gelatin and alginate (GA) in distilled water, at constant concentrations of 200 and 40 mg/mL, respectively, at 60 C. HA and TCP fillers powders were mixed with the GA solutions at concentrations of 0.125, 0.25 and 0.5% w/v (GAF). The entire mixing process lasted 30–45 min. In the relatively low concentrations range that was chosen, both HA and b-TCP, which are basically insoluble in water, could still be loaded inside the adhesive hydrogel with only minimal precipitation of each filler. The chosen constant concentrations of GA were based on previous results, which show that these concentrations enables a relatively high in vitro bonding strength and easy handling.20 Immediately prior to the
Downloaded from jba.sagepub.com at UCSF LIBRARY & CKM on August 30, 2014
1368
Journal of Biomaterials Applications 28(9)
Table 1. The studied HA and TCP-loaded formulations for soft and hard tissue adhesion.
For in vitro soft tissue bonding strength measurements
Reference adhesive HA-loaded adhesive
Gelatin concentration (mg/mL)
Alginate concentration (mg/mL)
EDC concentration (mg/mL)
Filler concentration (% w/v)
200 200
40 40
20 20
0 0.125 0.25 0.5 0.125 0.25 0.5
10
TCP-loaded adhesive
200
40
20
10
For in vitro hard tissue bonding strength measurements
HA-loaded adhesive TCP-loaded adhesive
200 200
40 40
20 20
0.125 0.25 0.5 0.125 0.25 0.5 0.25 0.5
HA: hydroxyapatite; TCP: tricalcium phosphate.
adhesive’s use, the crosslinking agent (EDC) was added to the various GAF solutions at concentrations of 10 and 20 mg/mL. All studied formulations are presented in Table 1.
In vitro bonding strength measurements Soft tissue measurements. Porcine skin (Kibbutz Lahav, Israel) was used as a soft tissue model for evaluating the effect of the fillers on the bonding strength of the adhesive. The porcine skin was cut into 2 2 cm square-shaped pieces and their epidermis side was attached firmly to metal testing holders with a matching surface area (all dimensions of the holders are specified in Figure 1). About 140 mL of the adhesive was then spread uniformly on the dermis side of two porcine skin pieces (that were attached to the testing holders) which were immediately attached by applying a 1.25 N load on the pieces and placed in a 37 C and 100% humidity environment. After 30 min, the bonding strength was measured in tension mode at room temperature using a 5500 Instron universal testing machine (Instron Engineering Corp., Massachusetts, USA) and a 10 N load cell. The two parts of the joint were strained at a constant velocity of 2 mm/min until separation was achieved. The mechanical testing procedure was inspired by the standard test method ASTM F-2258-03. The bonding strength was defined as the maximum strength in the stress–strain curve, measured by the Instron Merlin software (Massachusetts, USA).
Hard tissue measurements. The cortical portions of bovine femurs (purchased at a local abattoir) were used as a hard tissue model in order to evaluate the fillers’ effect on the bonding strength of the adhesive. The femurs were sawed into 2 2 0.2 cm rectangular cuboid specimens using an ‘‘FMB’’ (Dalmine BG, Italy) minor portable band saw. The specimens were firmly attached to metal testing holders with a matching surface area. The entire bonding system was based on the bonding system for the soft tissue adhesion (Figure 1) and has similar dimensions. About 140 mL of the adhesive (with various concentrations of the fillers) was then spread uniformly on the exposed side of two femur specimens. The specimens were then immediately attached to each other by applying a load of 7.5 N and placed in a 37 C and 100% humidity environment. After 30 min, the bonding strength was measured in tension mode at room temperature using a 5500 Instron universal testing machine (Instron Engineering Corp.) and a 2 kN load cell. The two parts of the joint were strained at a constant velocity of 2 mm/min until separation was achieved. The mechanical testing procedure was inspired by the standard test method ASTM F-2258-03. The bonding strength was defined as the maximum strength in the stress– strain curve, measured by the Instron Merlin software.
Microstructural characterization About 420 mL of the adhesive was injected into 24 24 3 mm rectangular silicone molds. After initial
Downloaded from jba.sagepub.com at UCSF LIBRARY & CKM on August 30, 2014
Cohen et al.
1369
Figure 1. Illustration of the bonding system for soft tissue adhesion measurements. Specimen 2 2 0.2 cm, applied load 7.5 Nm at 37 C and 100% humidity environment for 30 min. The bonding strength was measured in tension mode at room temperature using a 5500 Instron universal testing machine (Instron Engineering Corp.) and a 2 kN load cell, velocity 2 mm/min (according to standard test method ASTM F-2258-03). Five repetitions were carried out for each formulation.
curing, the rectangular cuboid-shaped adhesive specimens were pressed out of their silicone molds, airdried in a chemical hood, freeze-fractured, and their cross-section was observed using an environmental scanning electron microscope (ESEM; FEI Quanta 200 FEG, Oregon, USA) in a high vacuum mode, with an accelerating voltage of 10 kV.
Statistical analysis Statistical comparison between more than two groups was carried out using the analysis of variance (Tukey Kramer) method via ‘‘IBM’’ (New York, USA) SPSS (V. 15) software. A value of p < 0.05 was considered statistically significant.
Results In vitro soft tissue bonding strength In order to evaluate the effect of HA and TCP on the bonding strength to soft tissues was measured and compared with the bonding strength of adhesive not loaded
with filler (control group). The concentrations of gelatin, alginate, and EDC were kept constant – 200, 40, and 20 mg/mL, respectively. Fifteen repetitions were carried out for each formulation. The results are presented in Figure 2. Incorporation of both HA and TCP was found to improve the bonding strength of the adhesive. Improvement of the bonding strength with HA was effective at concentrations of 0.25% w/v and above (18.1 4.0 kPa vs. 8.4 2.3 kPa for the 0.25% w/v HA loaded and the non-loaded adhesive, respectively). Although adhesive loaded with 0.5% w/v HA exhibited a significant lower bonding strength than adhesive with 0.25% HA (13.4 1.6 kPa), as stated, it was still higher than the bonding strength of the nonloaded adhesive. On the other hand, all examined TCP concentrations were found to improve the bonding strength, but no significant difference was observed between the different concentrations. The highest average value was measured for a TCP concentration of 0.5% w/v (15.2 2.6 kPa). A relatively low concentration of the crosslinking agent is beneficial when high biocompatibility is required. We therefore decided to also examine the
Downloaded from jba.sagepub.com at UCSF LIBRARY & CKM on August 30, 2014
1370
Journal of Biomaterials Applications 28(9)
Figure 2. Effect of HA (a) and TCP (b) on the bonding strength of bioadhesives composed of gelatin (200 mg/mL), alginate (40 mg/mL), and EDC (20 mg/mL) to soft tissues. Significant differences are marked with ‘‘*’’. HA: hydroxyapatite; TCP: tricalcium phosphate.
Figure 3. Effect of HA (a) and TCP (b) on the bonding strength of bioadhesives composed of gelatin (200 mg/mL), alginate (40 mg/mL), and a reduced concentration of EDC (10 mg/mL) to soft tissues, compared with the reference formulation. Significant differences are marked with ‘‘*’’. HA: hydroxyapatite; TCP: tricalcium phosphate.
effect of the fillers on the bonding strength of an adhesive with a reduced concentration of EDC (10 mg/mL). The same three concentrations of the fillers were loaded (0.125, 0.25, and 0.5% w/v) and the concentrations of the polymeric components were kept constant as before. Fifteen repetitions were carried out for each formulation. The results are presented in Figure 3. As can be seen from these results, HA and TCP-loaded adhesives with a reduced concentration of EDC had bonding strengths similar to the bonding strength of the nonloaded adhesive with non-reduced EDC concentration (20 mg/mL), and specific concentrations (0.5 and 0.125% w/v for HA and TCP, respectively) were found to have even higher bonding strength values.
chosen for the hard tissue bonding strength experiment. We assumed that formulations which perform best for soft tissues will also perform best for hard tissues. Hence, 0.25% w/v HA and 0.5% w/v TCP were loaded into the adhesive and the bonding strength results were compared with a non-loaded adhesive. The concentrations of the polymeric components were kept constant as before. The EDC concentration was set to 20 mg/mL. Three repetitions were carried out for each formulation. The results are presented in Figure 4. Loading 0.25% w/v HA almost tripled the bonding strength of the adhesive from 26.6 9.2 kPa to 71.4 28.2 kPa. No significant improvement in the bonding strength was observed for the 0.5% w/v TCP-loaded adhesive (24.0 13.0 kPa).
In vitro hard tissue bonding strength The HA and TCP-loaded adhesives were also examined for their potential use in hard tissue adhesion. One concentration of each filler, based on the best results of the previous bonding strength tests for soft tissues, was
Microstructural analysis The bulk cross-sections of air-dried adhesive specimens loaded with 0.125 and 0.5% w/v HA and TCP were examined using ESEM. The concentrations of gelatin
Downloaded from jba.sagepub.com at UCSF LIBRARY & CKM on August 30, 2014
Cohen et al.
1371
Figure 4. Effect of HA and TCP on the bonding strength of selected bioadhesives composed of gelatin (200 mg/mL), alginate (40 mg/mL), and EDC (20 mg/mL) to hard tissue (bone). Significant differences are marked with ‘‘*’’. HA: hydroxyapatite; TCP: tricalcium phosphate.
(200 mg/mL), alginate (40 mg/mL), and EDC (20 mg/mL) were kept constant. Fractographs of the HA and TCP-loaded adhesives are presented in Figure 5. The fractographs provide a clear perspective on the crystallization of the fillers in the adhesive matrix. The HA particles exhibit a 50–60 mm diameter round core and a multi-shell structure, while the TCP particles are organized as 5–20 mm diameter aggregates. The fractographs also demonstrate that at the lower concentration, the HA and TCP particles are not uniformly dispersed in the adhesive. Higher concentrations of the fillers seem to provide a more uniform dispersion. It should be mentioned that the cracks in the crosssection of the specimens result from the fracturing process.
Discussion The soft tissue bonding strength experiments demonstrated that both HA and TCP can be regarded as effective reinforcement fillers, since loading them into the adhesive improves its bonding strength. The improvement in the bonding strength is probably achieved due to the stiffness of the insoluble HA and TCP particles which restrain some of the movement of the adhesive’s strings in their vicinity. As a result, the mechanical properties of the adhesive are improved and the bonding strength is therefore also improved. However, the HA and TCP concentration had a different effect on the bonding strength improvement (for an EDC concentration of 20 mg/mL). The TCP-loaded adhesives exhibited similar bonding strength values for all three examined concentrations, which is significantly higher than the bonding strength of the nonloaded adhesive. On the other hand, while it was still significantly higher than the non-loaded adhesive for both concentrations, increasing the HA concentration
from 0.25 to 0.5% w/v decreased the bonding strength. Loading lower concentrations of HA (0.125% w/v) was found to have no significant effect on the bonding strength of the adhesive. ESEM examination of the HA and TCP particles’ characteristics and dispersion inside the adhesive can provide a possible explanation for this phenomenon. Previous microstructural analysis of a non-loaded adhesive has shown that no phase separation occurs in the presence of only the three basic components of the adhesive – gelatin, alginate, and EDC.20 However, the incorporation of HA and TCP in the adhesive can clearly be noticed. Since HA is practically insoluble in aqueous solutions, its dispersion inside the water-based adhesive is not uniform. As can be seen from the ESEM fractographs (Figure 5(a) and (b)), the non-uniformity is more significant at the lower concentrations, where the density of the particles inside the adhesive is lower. As a result, no significant improvement in the bonding strength was observed for the lowest examined HA concentration. TCP is also insoluble in aqueous solutions. However, since its particles are significantly smaller than the HA particles (diameter of 5–20 mm compared with 50–60 mm, respectively), the dispersion of the TCP particles inside the adhesive matrix is likely to be more uniform than the dispersion of the HA particles at every given concentration. This is probably why an improvement in the bonding strength for TCP could be achieved even at a concentration of 0.125% w/v (although a uniform dispersion could hardly be noticed in the ESEM fractograph of that TCP concentration). The unique core and shells structure of the HA particles (Figure 5(c)) indicates that the surface of the HA particle probably reacts with the adhesive matrix either chemically or physically, which causes the shells’ formation. It is known from the literature that HA molecules have a high affinity to proteins when dispersed in an aqueous environment. This phenomenon occurs due to the dissolution of negative hydroxyl ions from the surface of the HA to the surrounding, exposing positive calcium charges which tend to bind to the proteins’ acidic groups, for example, carboxylic groups.34 The ability of the HA’s exposed positive calcium charges to bind to carboxylic groups originating from the GA may be the cause for a noticeable decrease in the bonding strength of the adhesive from a certain HA-loading concentration. The binding of positive charges originating from the HA particles’ surface to the gelatin and probably also to the alginate carboxylic groups decreases the density of the carboxylic groups which remain available for crosslinking of the adhesive in the close surrounding of the particles. As a result, a mechanical failure can develop from these areas, due to the lower crosslinking density. Evidence that supports this theory can be found in Figure 5(c), where it
Downloaded from jba.sagepub.com at UCSF LIBRARY & CKM on August 30, 2014
1372
Journal of Biomaterials Applications 28(9)
Figure 5. ESEM fractographs of adhesives (gelatin 200 mg/mL; alginate 40 mg/mL; and EDC 20 mg/mL) loaded with 0.125% w/ v HA (a), 0.5% w/v HA (b, c), 0.125% w/v TCP (d) and 0.5% w/v TCP (e, f). HA: hydroxyapatite; TCP: tricalcium phosphate; ESEM: environmental scanning electron microscope.
Downloaded from jba.sagepub.com at UCSF LIBRARY & CKM on August 30, 2014
Cohen et al.
1373
can be seen that cracks (originally from the specimens fracturing process) have a tendency to progress through the shells and stop at the core of the HA particle. This finding gives another indication for the relatively poor mechanical properties of the shells as opposed to the core and the matrix. Increasing the HA concentration increases not only the HA particles’ density but also the density of these weak mechanical areas, and it seems that the effect of these ‘‘weak’’ areas becomes more dominant than the effect of the particles from a certain HA concentration, and the bonding strength of the adhesive therefore begins to decrease. TCP, in contradistinction to HA, does not contain chemical groups that may dissolve in an aqueous environment and expose positive or negative charges on its surface. Therefore, the phenomenon of shells creation around its particles was not observed in the ESEM fractographs (Figure 5(f)), and no decrease in the bonding strength was observed for the higher loading concentrations. On the other hand, neither could an increase in the bonding strength be achieved when increasing the TCP concentration. This phenomenon may be related to the fact that due to the negligible solubility of TCP, its loading concentration range was not wide enough for achieving a noticeable consistent improvement in the bonding strength while increasing its concentration. Another possible reason can be the TCP particle size, which is relatively small compared with HA, and therefore does not enable a consistent increase in the bonding strength with the increase in its loading concentration. HA, which exhibited larger particles, improved the bonding strength of the adhesive to soft tissues more significantly than TCP. The crosslinking agent concentration has a significant effect on the bonding strength since it controls the crosslinking density in the adhesive. In a previous research, we saw that decreasing the crosslinking agent concentration by half, from 20 to 10 mg/mL, caused an almost two-fold reduction in the bonding strength of an adhesive with the same formulation.19 Comparing the bonding strength results of HA and TCP-loaded adhesives with a reduced concentration of EDC (10 mg/mL) to that of the non-loaded reference adhesive (20 mg/mL EDC) indicated that both HA and TCP have a compensatory effect on the bonding strength when decreasing the crosslinking agent concentration. Since EDC was shown to have some cytotoxic effects,19,20 this compensatory effect also has clinical and medical importance because it enables improving the biocompatibility of the adhesive without compromising its bonding strength and even improving it. The bonding strength of EvicelTM (commercial fibrin glue) was recently measured in our lab using the same bonding strength system.19 A comparison between the bonding strength of EvicelTM (2.5 2.3 kPa) and our novel HA and
TCP-loaded bioadhesives shows that even when the EDC concentration is reduced by half, incorporation of these fillers enables the achievement of up to 7 times higher bonding strength to soft tissues than with EvicelTM. Higher bonding strength values which reached more than 70 KPa were obtained when bonded to hard tissue. In the latter’s case, the contribution of the chemical adsorption mechanism is increased when HA is added to the bioadhesive. It should be noted that for the reduced EDC concentration (10 mg/mL) adhesives, the effect of the HA concentration on the bonding strength is different from its effect on the bonding strength of adhesive with the non-reduced EDC concentration. The reason for this might be the fact that although carboxylic groups originally from the GA still react with the positive charges on the surface of the HA particles, there are sufficient available groups left for reacting with all the EDC, due to its reduced concentration. Therefore, the effect of domains with lower mechanical properties around the HA particles on the bonding strength may be negligible. Due to the difficulty of preparing consistent and uniform hard tissue specimens, the bonding strength of HA and TCP-loaded adhesives on hard tissues was examined for only one concentration of each filler (0.25 and 0.5% w/v for HA and TCP, respectively) and for a non-loaded adhesive as control. For the same reason, only three repetitions were carried out for each formulation. A significant improvement in the bonding strength of the adhesive to hard tissues was observed only for the HA-loaded adhesive (Figure 4). Incorporation of TCP had practically no effect on the bonding strength to hard tissue. The adherence ability of other bioadhesives to hard tissues was evaluated and it was found that under similar test conditions, fibrin and gelatin-resorcinol-formaldehyde exhibit in vitro bonding strengths of 11 and 200 kPa, respectively.35 Our HA-loaded adhesive exhibits a bonding strength to hard tissues that is 6.5 times higher than that of fibrin (71.4 28.2 kPa), but is lower than that of the gelatin-resorcinol-formaldehyde adhesive.35 However, the latter is considered to be less biocompatible than our adhesive, since it is based on formaldehyde as the crosslinking agent, which is significantly more cytotoxic than EDC.21 In conclusion, in the current study we examined the effect of loading the bioactive fillers HA and TCP into our novel gelatin–alginate bioadhesives on the latter’s adherence properties. Our thought was that incorporating such fillers, which are known for their ability to promote bone formation and osseointegration, will enable us to extend the use of this bioadhesive to hard tissue adhesion. Incorporating both HA and TCP was found to significantly improve the bonding strength of the bioadhesive to soft tissues.
Downloaded from jba.sagepub.com at UCSF LIBRARY & CKM on August 30, 2014
1374
Journal of Biomaterials Applications 28(9)
Combining these fillers in the bioadhesive can also improve its biocompatibility, since their presence enables decreasing the concentration of the relatively cytotoxic crosslinking agent without compromising the mechanical properties. The potential of these fillers to improve the bonding strength of the adhesive to hard tissues was also demonstrated. Thus, HA and TCP are beneficial to our new gelatin–alginate bioadhesives, which can be used for both soft and hard tissue adhesives applications. Acknowledgments The authors are grateful to the Office of the Chief Scientist in the Israel Ministry of Industry, Trade and Labor, and Kodesz Foundation (Tel-Aviv University) for supporting this research.
Conflict of interest None declared.
References 1. Sibert JR, Maddocks GB and Brown BM. Childhood accidents – an endemic of epidemic proportion. Arch Dis Child 1981; 56: 225–227. 2. Osmond MH. Wound repair and tissue adhesives. In: Moyer V, Elliott E and Davis R (eds) Evidencebased Pediatrics and Child Health. 1 ed. London: BMJ Books, 2000. 3. Singer AJ, Kinariwala M, Lirov R, et al. Patterns of use of topical skin adhesives in the emergency department. Acad Emerg Med 2010; 17: 670–672. 4. Kraus R, Schluckebier D, Stiller A-C, et al. Bone adhesives in trauma and orthopedic surgery. Eur J Trauma 2006; 32: 141–148. 5. Donkerwolcke M, Burny F and Muster D. Tissues and bone adhesives—historical aspects. Biomaterials 1998; 19: 1461–1466. 6. Reece TB, Maxey TS and Kron IL. A prospectus on tissue adhesives. Am J Surg 2001; 182: S40–S44. 7. Ryou M and Thompson CC. Tissue adhesives: a review. Techniques Gastrointes Endosc 2006; 8: 33–37. 8. Farrar DF. Bone adhesives for trauma surgery: a review of challenges and developments. Int J Adhes Adhesiv 2011. 9. Kuijpers AJ, Engbers GHM, Feijen J, et al. Characterization of the network structure of carbodiimide cross-linked gelatin gels. Macromolecules 1999; 32: 3325–3333. 10. Djabourov M and Papon P. Influence of thermal treatments on the structure and stability of gelatin gels. Polymer 1983; 24: 537–542. 11. Bae SK, Sung TH and Kim JD. A soft-tissue gelatin bioadhesive reinforced with a proteinoid. J Adhes Sci Technol 2002; 16: 361–372. 12. Jonas RA, Ziemer G, Schoen FJ, et al. A new sealant for knitted Dacron prostheses: minimally cross-linked gelatin. J Vas Surg 1988; 7: 414–419.
13. Tabata Y, Hijikata S and Ikada Y. Enhanced vascularization and tissue granulation by basic fibroblast growth factor impregnated in gelatin hydrogels. J Control Rel 1994; 31: 189–199. 14. Narayani R and Rao KP. Controlled release of anticancer drug methotrexate from biodegradable gelatin microspheres. J Microencapsul 1994; 11: 69–77. 15. Otani Y, Tabata Y and Ikada Y. A new biological glue from gelatin and poly (L-glutamic acid). J Biomed Mater Res 1996; 31: 157–166. 16. Sung HW, Huang DM, Chang WH, et al. Evaluation of gelatin hydrogel crosslinked with various crosslinking agents as bioadhesives: in vitro study. J Biomed Mater Res 1999; 46: 520–530. 17. Otani Y, Tabata Y and Ikada Y. Rapidly curable biological glue composed of gelatin and poly (L-glutamic acid). Biomaterials 1996; 17: 1387–1391. 18. Matsuda S, Iwata H, Se N, et al. Bioadhesion of gelatin films crosslinked with glutaraldehyde. J Biomed Mater Res 1999; 45: 20–27. 19. Shefy-Peleg A, Foox M, Cohen B and Zilberman M. Novel Antibiotic-eluting gelatin-alginate soft tissue adhesives for various wound closing applications. Int J Poly Mater and Poly Biomater (in press). 20. Cohen B, Shefy-Peleg A and Zilberman M. Gelatin-based soft tissue adhesive loaded with analgesic drugs: structure and properties. J Biomater Sci, Polym Ed (in press). 21. Nishi C, Nakajima N and Ikada Y. In vitro evaluation of cytotoxicity of diepoxy compounds used for biomaterial modification. J Biomed Mater Res 1995; 29: 829–834. 22. Draget KI, Smidsrød O and Skjak-Brak G. Alginates from algae. Biopolym Online 2005: 215–244. 23. Liew CV, Chan LW, Ching AL, et al. Evaluation of sodium alginate as drug release modifier in matrix tablets. Int J Pharm 2006; 309: 25–37. 24. Ashley M, McCullagh A and Sweet C. Making a good impression: (a ‘how to’ paper on dental alginate). Dent Update 2005; 32: 169. 25. Smidsrød O and Skjaok-Brk G. Alginate as immobilization matrix for cells. Trends Biotechnol 1990; 8: 71–78. 26. Matthew IR, Browne RM, Frame JW, et al. Subperiosteal behaviour of alginate and cellulose wound dressing materials. Biomaterials 1995; 16: 275–278. 27. Dorozhkin SV. Calcium orthophosphates: occurrence, properties, biomineralization, pathological calcification and biomimetic applications. Biomatter 2011; 1: 121–164. 28. Harper EJ, Behiri JC and Bonfield W. Flexural and fatigue properties of a bone cement based upon polyethylmethacrylate and hydroxyapatite. J Mater Sci: Mater Med 1995; 6: 799–803. 29. Liebendo¨rfer A, Schmitz B, Wenz R, et al. Experimental studies on a new bone cement: hydroxyapatite composite resin. In: The 21st annual meeting of the society for biomaterials, San Francisco, USA, 1995, p. 335. 30. Lee HB, Khang G, Lee JH, et al. Polymeric biomaterials. Boca Raton, FL: CRC Press, 2003. 31. Gellynck K, Abou Neel EA, Li HY, et al. Cell attachment and response to photocured, degradable bone
Downloaded from jba.sagepub.com at UCSF LIBRARY & CKM on August 30, 2014
Cohen et al.
1375
adhesives containing tricalcium phosphate and purmorphamine. Acta Biomater 2011; 7: 2672–2677. 32. Gerhart TN, Renshaw AA, Miller RL, et al. In vivo histologic and biomechanical characterization of a biodegradable particulate composite bone cement. J Biomed Mater Res 1989; 23: 1–16. 33. Collinge C, Merk B and Lautenschlager EP. Mechanical evaluation of fracture fixation augmented with tricalcium phosphate bone cement in a porous osteoporotic
cancellous bone model. J Orthopaed Trauma 2007; 21: 124–128. 34. Kandori K, Miyagawa K and Ishikawa T. Adsorption of immunogamma globulin onto various synthetic calcium hydroxyapatite particles. J Colloid Interf Sci 2004; 273: 406–413. 35. Chivers RA and Wolowacz RG. The strength of adhesive-bonded tissue joints. Int J Adhes Adhesiv 1997; 17: 127–132.
Downloaded from jba.sagepub.com at UCSF LIBRARY & CKM on August 30, 2014
