A Bioactive Sol-Gel Glass Implant for In Vivo Gentamicin Release. Experimental Model in Rabbit L. Meseguer-Olmo,1 MJ. Ros-Nicola´s,1 V. Vicente-Ortega,2 M. Alcaraz-Ban˜os,3 M. Clavel-Sainz,1 D. Arcos,4 C.V. Ragel,4 M. Vallet-Regı´,4 Cl. Meseguer-Ortiz1 1
Laboratorio de Cirugı´a Ortope´dica Experimental, Facultad de Medicina, Universidad de Murcia, Spain
2
Departamento de Anatomı´a Patolo´gica, Facultad de Medicina, Universidad de Murcia, Spain
3
Departamento de Medicina Fı´sica y Radiologı´a, Facultad de Medicina, Universidad de Murcia, Spain
4
Departamento de Quı´mica Inorga´nica y Bioinorga´nica, Facultad de Farmacia, Universidad Complutense, 28040 Madrid, Spain
Received 7 February 2005; accepted 15 August 2005 Published online 31 January 2006 in Wiley InterScience (www.interscience.wiley.com). DOI 10.1002/jor.20064
ABSTRACT: Biomaterial pieces with osteogenic properties, suitable for use in the treatment of bone defects, were synthesized. The materials, which avoid bone infections, are exclusively composed of gentamicin sulfate and bioactive SiO2-CaO-P2O5 sol-gel glass (synthesized previously), and were manufactured by means of uniaxial and isostatic pressure of the mixed components. After implanting the pieces into rabbit femur, we studied (1) antibiotic release, determining the concentration in proximal and distal bone, liver, kidney, and lung as a function of time, and (2) bone growth as a consequence of the glass reactivity in the biological environment. The results demonstrated that the implants are good carriers for local gentamicin release into the local osseous tissue, where they show excellent biocompatibility and bone integration. Moreover, these implants are able to promote bone growth during the resorption process. ß 2006 Orthopaedic Research Society. Published by Wiley Periodicals, Inc. J Orthop Res 24:454–460, 2006
Keywords: in vivo gentamicin release; sol-gel glasses; bioactive glasses; in vivo bioactivity; SiO2-CaO-P2O5 system
INTRODUCTION The use of carriers for local antibiotic release is a very important aspect in therapeutic and orthopedic surgery, because neither meticulous and surgical precision is able to ensure the absence of infectious microorganisms. In fact, in the case of osteomielitis, the implant must be removed if further complications, such as loss of function and septicemia, are to be avoided. The systemic administration of antibiotics is not always allow for efficient concentrations mainly due to poor blood flow in bone tissue. This means that large antibiotic doses have to be administed to obtain acceptable concentrations in the affected Correspondence to: L. Meseguer-Olmo (Puerta Nueva, 18, 28A, 30008 Murcia, Spain. Tel.: þ34-968-241826; Fax: þ34968-241826; E-mail:
[email protected]) ß 2006 Orthopaedic Research Society. Published by Wiley Periodicals, Inc.
454
JOURNAL OF ORTHOPAEDIC RESEARCH MARCH 2006
region. For this reason, one of the most important topics in medical literature1–3 is the development of biomaterials suitable for use as local drug delivery systems. Implants able to deliver a drug in a higher concentrations than the minimum inhibitory dose would eliminate pathogenic microorganisms and avoid toxic overdose. One of the first examples of this was the inclusion of antibiotics in implant fixation cements,4 and gentamicin-containing poly-methyl-metacrylate (PMMA) beads,5 methods, which have been shown to be successful in the treatment and prevention of osseous infections. However, the use of these kinds of polymer can lead to several complications. For example, gentamicin-containing PMMA beads have several disadvantages, such as low degree of biocompatibility, low release rate, and the possibility of harm to the patient during the subsequent intervention to remove the beads.6 Furthermore, mixing a drug PMMA does not always lead to solve
BIOACTIVE SOL-GEL GLASS IMPLANT FOR GENTAMICIN RELEASE
intervals under continuous stirring. A clear sol was obtained, which was introduced into Teflon containers and allowed to gelify over a period of 3 days at room temperature. The original container lid was replaced by one featuring a 1-mm diameter hole, after which the solution was allowed to dry at 1508C for 52 h. Taking into account the data reported in previous studies,10,11 we stabilized the glass by means of a thermal treatment at 7008C for 3 h under air atmospheric pressure. X-ray diffraction experiments confirmed the glassy state of the synthesized material. Preparation of the Implants The sol-gel glass was grounded for 1 h in a rotatory ballmill (Retsch), using an agate container and 20 mm agate balls. The ball/glass weight ratio was 6:1. The particles were sieved and the fraction ranging in size from 32 to 62 mm were collected. Particle size distribution was determined with a Sedigraph 5100 by suspending the glass grains in a 0.5% hexametaphosphate water solution. The results obtained are shown in Figure 1a. Finer mass percentage of sol gel glass showed that 80% (wt %) of the sol-gel glass consisted of particles with an equivalent spherical diameter of less than 29 mm, while 50% of the glass mass particles were less than 17 mm in diameter. Figure 1b shows the particle size distribution, indicating a single-
a
100 80
% mass finer
the problem of release time, local drug concentration, or the nonreleased drug ratio. An interesting alternative is the use of bioactive material/PMMA composites, which ensure total integration and bone regeneration. For example, the incorporation of sol-gel SiO2-CaO-P2O5 glasses allows the rapid formation of an apatite-like deposit over the implant surface as well as in the inner part.7 This apatite phase is similar to the mineral component of natural bone, and its presence leads to a strong and close bond between the implant and the living tissue. The growth mechanism for this apatite phase over sol-gel glasses is well known, and can be explained as follows. The high surface area and porosity leads to rapid ionic exchange with the surrounding fluids. This exchange creates a silica-gel layer that incorporates Ca and P from the solution, forming an amorphous calcium phosphate layer and, after subsequent crystallization, the apatite-like phase. Glass/drug/PMMA composites can be processed by radical polymerization of methyl-metacrylate (MMA) obtaining to provide pieces of different sizes and shapes to deliver the drug locally in a controlled way.8,9 However, the presence of acrylic cement in the bone has the same disadvantages as mentioned above. For this reason, resorbable and bioactive implants able to maintain efficient antibiotic levels for an acceptable period would represent a considerable improvement in the field of therapeutic and orthopedic surgery. This article describes the synthesis and in vivo evaluation of sol-gel/gentamicin implants, which are used in an attempt to avoid the disadvantages related with PMMA. For this purpose, were prepared pieces by means of uniaxial and isostatic pressure on an SiO2-CaO-P2O5 glass/gentamicin sulfate mixture. The behavior of this material was evaluated by studying its in vivo osteogenic performance and ability to release antibiotics after being implanted in the rabbit femur.
455
60 40 20 0
b
MATERIALS AND METHODS Synthesis of the Glass The glass was synthesized by hydrolysis and polycondensation of tetraethyl orthosilane (TEOS), triethylphosphate (TEP), and Ca(NO3)2 4H2O in the appropriate ratio to obtain a nominal composition of SiO2 58-CaO 36P2O5 6 (% mol). The reaction was carried out by maintaining an H2O/TEOS þ TEP ¼ molar ratio of 8, and was catalyzed by adding 2N HNO3. To ensure the total hydrolysis of each reagent, they were added at 1-h DOI 10.1002/jor
freq mass(%)
3
2
1 0 100
10
1 diameter ( m)
0.1
Figure 1. Particle size of synthesized glass: (a) cumulative mass percent and (b) particle size distribution. JOURNAL OF ORTHOPAEDIC RESEARCH MARCH 2006
456
MESEGUER-OLMO ET AL.
mode distribution, with the maximum size placed at 22 mm. For each implant, 200 mg glass and 7 mg gentamicin sulfate (B. Braun Medical S.A. Labs, Barcelona, Spain) were gently mixed for each implant. The resulting mixture was shaped into 6 mm (diameter) 6 mm (length) cylinders in two steps: (1) 70 MPa uniaxial pressure, and (2) 355 MPa isostatic pressure. The second step provides a higher consistency to the implants and a more homogeneous stress distribution. The isostatic pressure was carried out by covering the cylinders with latex and soaking them into a water/oil mixture. The force applied to the liquid exerted pressure over the entire cylinder surface. At the end of the process, cylinders were obtained with the above-mentioned dimensions. Animals Model Twenty New Zealand rabbits of both sexes and ranging in weight from 3.5 to 4 kg were randomly used in this experiment. The animals were distributed into five groups corresponding to the foreseen times of sacrifice (1, 2, 4, 8, and 12 weeks postimplant). The animals were placed into individual cages, fed with a full rabbit’s special diet RSD (code 112, Panlab, Barcelona, Spain), water ad libitum, and illumination with a 12 h/12 h light–dark photoperiod, in agreement with the animal experimentation rules of the EEC 86/609. Anesthetic and Surgical Method The animals were anesthetized with ketamin sulfate (50 mg/kg) and chlorpromazine (10 mg/kg). Antibiotic prophylaxis was carried out with penicillin (240,000 IU) and streptomycin (300 mg) in a single dose before surgery. Next, the left rear extremity was shaved and the surgery region was disinfected with BetadineTM and covered by sterile adhesive film (OpSiteTM). An incision was made in the lateral face of the lower third of the leg; the metaphysial area close to the distal epiphysis of the femur was exposed through the intramuscular wall of the quadriceps–fascia lata muscle and the femoral biceps (in the lower plane). An osseous defect was created with a 5-mm surgical drill. The defect was gently rinsed and the implant was placed. To prevent the antibiotic from being released outside the bone, the cortical hole was sealed with a thin layer of osseous wax. The wound was stitched up in anatomic planes. The animals were treated with local subcutaneous mepivacain (4 mg/kg) as postoperative analgesic. The animals were euthanized, after being anesthetized, with a lethal intracardiac injection of sodium pentobarbital. Radiological and Histological Studies Femurs were extracted and X-rayed with a Mammodiagnost UC device (Philips) with 28 kV, 25 mA, using coarse focus and Kodak Min-R Screen film. Standard JOURNAL OF ORTHOPAEDIC RESEARCH MARCH 2006
Figure 2. Schema of the osseous sections (distal and proximal) obtained after implantation.
plain were carried out. Next, the images were recorded with a video camera, focusing on the implant outline and a constant peripherical area. The images were processed and analyzed by MICH software (Microm Image Processing). Osseous sections of 0.5-cm thickness, perpendicular to the diaphysis axis, were collected from regions proximal and distal to the implantation zones (Fig. 2). Some were fixed with 10% buffered formol, decalcified, and stained with hematoxylin eosin for observation with a Leitz Orthoplan FSA optical polarized light microscope. Other sections were kept frozen at 408C to determine the local antibiotic concentration. Tissue samples from the liver, lung, kidney, and iliac lymphatic nodes were extracted, frozen with liquid nitrogen, and preserved at 408C. Samples of peripheral blood were collected and the serum was extracted for later analysis. Antibiotic Diffusion Study The osseous sections (free of bone marrow remains and soft parts), lung, liver, and kidney samples were weighed and gently crushed. The samples were homogenized in 3 mL of phosphate buffer saline (PBS) and stirred for 4 h. Next, they were centrifuged at 4000 rpm for 20 min. The extract was kept frozen at 408C. The gentamicin concentration was assessed by fluorescence polarization immune assays (FPIA) with a TDx/TDxFLx system (Abbott1). Standards in the range of 0.0–10 mg/mL were used to calibrate the system, which was set as particular conditions required, for example, for determining concentrations outside this DOI 10.1002/jor
BIOACTIVE SOL-GEL GLASS IMPLANT FOR GENTAMICIN RELEASE
457
range. The specificity of the system meant that the possibility of crossed reactivity with another aminoglycoside of similar chemical structure was always lower than 1%. Sensitivity was 95% with a value of 0.27 mg/mL, while the precision, according to the EP5-T protocol (National Committee for Clinical Laboratory Standards), showed variation coefficients lower than 6%. Statistical Analysis A descriptive statistical study was performed to calculate the characteristic parameters of the concentrations: mean, standard mean error, and typical deviation. A comparison between groups was achieved with a two-way analysis variance (ANOVA), complemented by equality contrast in pairs of measurements (Pairwise test), a Student t-test, and Bonferroni correction. In cases where unequal variance was obtained by the Levene test, ANOVA was not carried out; instead, Welch’s approximate and Brow-Forsythe test were applied. ANOVA was also repeated following data transformation by the logarithmic function: y ¼ lnð1 þ concentrationÞ
RESULTS One animal was excluded from the experiment due to a iatrogenic fracture in the area of implantation as a result of accidental manipulation. No signs of infection were detected in the remaining animals. Radiologic and Histologic Studies of Bone Ingrowth At the beginning of the study (1 week), the implant showed a cylindrical morphology, with a homogeneous X-ray intensity greater than that of surrounding osseous tissue (Fig. 3). The lateral cortical bone defect made during the implantation could be clearly observed. After 4 weeks, linear reinforcements were observed in the osseous tissue surrounding the implant, and in close contact with it. Cortical bone repair was observed as small areas still without osseous tissue, probably as a consequence of the remains of the osseous wax used to seal the cortical hole. A moderate decrease in the density of the implant outline could be observed, together with changes in the morphology and the presence of numerous lines of calcium density of irregular arrangement and in contact with the implanted material. After 8 weeks, irregularities in the implant outline DOI 10.1002/jor
Figure 3. X-ray images of the extracted femurs: (a) After 1 week, a rectangular image of homogeneous density is observed, which corresponds to the implant (*) and the lateral cortical bone defect performed for implantation. (b) After 4 weeks, the image shows full cortical restoration (big arrow), a slight decrease in density at the edges of the material and a certain change in morphology. (c) After 12 weeks, the implant size decreases and shows a rounder shape due to the resorption process. Several lines of calcium density can be observed (arrow) in direct contact with the implanted material.
showed that the resorption process had started. Moreover, newly formed bone extended from the lateral cortical bone covering the implant. At the end of the study (12 weeks), the morphology of the implant had totally changed, and it now appeared smaller and had with a more rounded outline. Small isolated fragments could also be observed. The nearest bone tissue was directly arranged over the implant, with no radiolucent lines around it. The image analysis corroborated the X-ray finding, pointing to a continuous decrease in the implant surface, while a radio-dense tissue appeared at the implant periphery. The histologic sections (Fig. 4), which included the peripheral and central portions of the implanted material, showed an inflammatory response around the implant after 1–2 weeks. Numerous multinucleated giant cells showing phagocytic activity appeared around the implant. Areas of fibroblastic proliferation and hematic material, together with trabecular bone, could be observed surrounding the implant. Fibroblastic proliferation around the implant, next to adipose bone marrow, was also observed. After 4 weeks, osteogenic activity was clearly observed. This was characterized by the presence of newly formed bone trabecules with a continuous coating of osteoblastic cells. The implant region could be JOURNAL OF ORTHOPAEDIC RESEARCH MARCH 2006
458
MESEGUER-OLMO ET AL.
Figure 4. Optical micrographs from the histological sections: (a) detail of implanted material (*) after 1 week, with trabecular bone (arrow head) and areas of fibroblastic proliferation and hematic material (arrow) (HE, original magnification: 125). (b) After 2 weeks, at the bone–implant interface, the beginning of the osteogenesis process is manifested by the presence of fibroblasts (arrow head) and neoformed trabecular bone (arrow) (HE, original maginification 125). (c) After 1 month, fatty bone marrow close to the implant area: detail of the foreign body giant cells (arrow head) with granulated material (arrow) (HE, original magnification 412). (d) Granular material (*), surrounded by multinucleated giant cells (arrow head), fibroblastic proliferation and newly formed bone trabecules in inner areas of the implant (arrow) (HE, original magnification 312.5).
easily observed because of the small accumulations of granular materials surrounded by multinucleated giant cells with granulated material inside the cytoplasm, fibroblastic proliferation, and newly formed bone trabecules inside the implant. After 8 weeks, the osseous tissue had colonized the implant from the peripheral regions towards the interior with no interposition of fibrous tissue. The region of enchondral ossification was visible in the implanted material (optically empty). The histologic sections of the lymphatic nodes showed follicular hyperplasia and extensive sinusal histiocytosis. The above-mentioned granular material was not observed in the macrophage cytoplams by polarized light microspcopy. Neither did the optical microscopy study carried out in the organs (liver, lung, and kidney) show the presence of the above particles or any histological alteraJOURNAL OF ORTHOPAEDIC RESEARCH MARCH 2006
tions (such as inflammatory signals or necrosis), which might point to organ dysfunction (images no provided). Gentamicin Levels Of note during the first week was the higher gentamicin concentration (p < 0.001) in kidney (2.00 0.03 mg/mL) than in lung (0.26 0.02 mg/ mL) and liver (0.48 0.02 mg/mL), which as seen in Figure 5. This can be expressed as [kidney] > [liver] & [lung]. This ratio changed at 2 weeks, when the gentamicin concentration had decreased to half in the kidney (0.9 0.1 mg/mL), while the concentration in lung (0.407 0.009 mg/ mL) and liver (2.41 0.02 mg/mL) had increased so that [liver] > [kidney] > [lung] (p < 0.01). The concentration decreased from 2 weeks until the end of the test in these organs. The very low DOI 10.1002/jor
BIOACTIVE SOL-GEL GLASS IMPLANT FOR GENTAMICIN RELEASE
gentamicin (µg/mL)
3
levels in DB increasing progressively from 0.8 0.3 mg/mL (1 week) to 6 1 mg/mL (12 weeks), although without statistical significance. All the concentrations measured in proximal bone and distal bone were higher than the lowest inhibitory concentration (LIC) for Staphiloccoccus aureus, which is 0.12–1 mg/mL.
liver lung kidney
2
1
0
DISCUSSION 0
2
4
6
8
10
12
14
time (weeks)
Figure 5. Gentamicin levels measured in several organs as a function of implantation time (explanation in the text).
gentamicin concentration in serum did not allow us to detect the antibiotic in serum. The gentamicin values detected in cortical bone (Fig. 6), corresponding to proximal osseous sections (PB), increased quickly and constantly, reaching a maximum mean value of 13 6 mg/mL at 4 weeks, after which, the drug concentration in PB fell to reach 0.8 0.2 mg/mL by the end of the study. Therefore, significant differences can be established for [PB-4 weeks] > [PB-8 weeks] > [PB-12 weeks] (p < 0.05). The above-mentioned gentamicin concentration in PB after 4 weeks was much greater than that observed in distal bone (DB) at the same time (p ¼ 0.28), a situation had been reversed by the end of the study (p < 0.5) (Fig. 6), because the drug
20 gentamicin (µg/mL)
459
DB PB
16 12 8 4
The local gentamicin levels detected in osseous tissue were higher than the LIC throughout the assay and, as such, these levels would be effective against most microorganisms, even the most resistant. The progressive decrease of gentamicin values in the bone was not sufficient to bring its concentration below the LIC at any stage during the assay, meaning that there was a wide local area in which concentrations were sufficient to inhibit microorganism growth. The local blood flow in the defect was partially restored after 1 week, as can be seen by the high antibiotic concentration in organs such as kidney. However, the concentration in serum could not be measured for the above-mentioned reasons. The blood flow into the bone improved after the second month, coinciding with a higher gentamicin level in the distal bone. Therapeutic gentamicin concentrations in humans were been established by Barza et al.12 and Noone et al.13 as 5–10 mg/mL. Persistent concentrations higher than 10 mg/mL lead to alterations in the VIII craneal nerve and nephrotoxicity, although lower levels may also be associated with toxic effects. In our study, only the kidney showed slightly higher levels during the first week, although this did not lead to histologic alterations in the organs analyzed. The integration and regeneration of osseous tissue has been widely studied in the case of the bioactive glasses.6,14,15 The sol-gel glass used in our work allowed the growth of osseous tissue from the peripheral region toward the implant. The newly formed bone penetrated into the implant and, at the same time, the glass was slowly but progressively resorbed by phagocytosis.
0 0
2
4
6
8
10
12
14
time (weeks) Figure 6. Gentamicin levels measured in proximal (PB) and distal bone (DB) as a function of implantation time (explanation in the text). DOI 10.1002/jor
CONCLUSIONS The implants of sol-gel glass acted as successful drug carriers and delivery systems at bone level during the study time. Moreover, the implants JOURNAL OF ORTHOPAEDIC RESEARCH MARCH 2006
460
MESEGUER-OLMO ET AL.
used showed an excellent degree of biocompatibility, a very low inflammatory response, full osteointegration with direct apposition of the newly formed bone, and no fibrous tissue around the implant; as the implant was resorbed, so new bone was formed. Such implants can be considered as a novel alternative for local gentamicin delivery, while the material of which they are made promotes the formation of new bone.
ACKNOWLEDGMENTS The authors express their appreciation to Prof. M. Canteras Jordana from the Biostatistical Department, Faculty of Medicine, for his help in the statistical analysis.
5.
6.
7.
8.
9.
10.
REFERENCES 1. Otsuka M, Matsuda Y, Kokubo T, et al. 1995. Drug release from a novel self-setting bioactive glass bone cement containing cephalexin and its physico chemical properties. J Biomed Mater Res 29:33–38. 2. P del Real R, Padilla S, Vallet-Regı´ M. 2000. Gentamicin release from hydroxyapatita/poly (ethyl methacrylate)/poly (methyl methacrylate) composites. J Biomed Mater Res 52:1–7. 3. Van Wachem PB, Van Luyn MJA, de Wit AW, et al. 1997. Tissue reactions to bacteria-inoculated rat lead samples. I. Effect of local gentamicin release through vicinal sponge of solution-diping. J Biomed Mater Res 35:217–232. 4. Buchholz HW, Engelbercht H. 1970. Uber die depotwirkung einiger antibiotica bei vermischun
JOURNAL OF ORTHOPAEDIC RESEARCH MARCH 2006
11. 12.
13.
14. 15.
mit dem kunstharz Palacos. Chirurg 41:511– 515. Vecsei V, Barquet A. 1981. Treatment of chronic osteomyelitis by necrectomy and gentamicinPMMA beads. Clin Orthop 159:201–207. Torholm C, Lidgren L, Lindberg L, et al. 1983. Total hip joint arthroplasty with gentamicin-impregnated cement. A clinical study of gentamicin excretion kinetics. Clin Orthop 181:99–106. Vallet-Regı´ M, Arcos D, Pe´rez-Pariente J. 2000. Evolution of porosity during in vitro hydroxycarbonate apatite growth in sol-gel glasses. J Biomed Mater Res 51:23–28. Arcos D, Ragel CV, Vallet-Regı´ M. 2001. Bioactivity in glass/PMMA composites used as drug delivery system. Biomaterials 22:701–708. Ragel CV, Vallet-Regı´ M. 2000. In vitro bioactivity and gentamicin release from glass–polymer–antibiotic composites. J Biomed Mater Res 51:424– 429. Balas F, Arcos D, Pe´rez-Pariente J, et al. 2001. textural properties of SiO2-CaO-P2O5 glasses obtained by the sol-gel method. J Mater Res 16:1345–1348. Vallet-Regı´ M. 2001. Ceramics for medical applications. J Chem Soc, Dalton Trans 2001:97–108. Barza M, Lauerman M. 1978. Why monitor serum level of gentamicin? Clin Pharmacokinet 3:202– 215. Noone P, Parsons TMC, Pattison JR. 1974. Experience in monitoring gentamicin therapy during treatment of seriuos gram negative sepsis. Br Med J 1:477–481. Ducheyne P. 1998. Stimulation of biological function with bioactive glass. MRS Bull 23:43–49. Hench LL, Kokubo T. 1998. Properties of bioactive glasses and glass–ceramics. In: Black J, Hastings G, editors. Handbook of biomaterial properties. London: Chapman & Hall; p 355–356.
DOI 10.1002/jor
