Eur J Orthop Surg Traumatol DOI 10.1007/s00590-013-1408-6
ORIGINAL ARTICLE
Liquid dextran does not increase the elution rate of different antibiotics from bone cement K. E. Roth • B. Krause • E. Siegel • G. Maier C. Schoellner • P. M. Rommens
•
Received: 6 November 2013 / Accepted: 30 December 2013 Ó Springer-Verlag France 2014
Abstract Purpose To investigate the possibility of increasing elution of fosfomycin, gentamicin, clindamycin, and vancomycin by the addition of dextran fluid during the cementmixing phase. Methods In 12 test series, we produced standardized, antibiotic-loaded test specimens of cement, with and without addition of dextran, and determined their effectiveness against three reference pathogens in agar diffusion and elution tests. Results In the test series using combined agents, RefobacinÒ-PalacosÒR plus fosfomycin continuously produced the largest zone of inhibition, both against methicillinsensitive Staphylococcus aureus (p = 0.009) and against methicillin-resistant Staphylococcus aureus (p = 0.009). The addition of dextran to the various test series had no useful effect on the size of the zone of inhibition for any of the antibiotics tested. Conclusions Dextran supplementation in RefobacinÒPalacosÒR bone cement did not have the hope for positive effect on the elution rate of bound antibiotics. Keywords Dextran
Bone cement Antibiotics Elution rate
K. E. Roth (&) B. Krause G. Maier C. Schoellner P. M. Rommens Department of Orthopaedic Surgery, Johannes Gutenberg University, Langenbeckstraße 1, 55131 Mainz, Germany e-mail: [email protected] E. Siegel Department of Microbiology, Johannes Gutenberg University, Mainz, Germany
Introduction Bone cement has been used successfully for many decades to anchor endoprostheses and for localized delivery of antibiotics. In this function, they act, applied in the form of placeholders or chains, as a drug-delivery system that deploys its effects directly at the surgical wound site. Spacer systems are used in septic prosthesis surgery when sequential therapy procedures are preferred [1]. For the successful treatment of the infection, particularly in the use of glycosides [2], the amount of the active agent released from the bone cement is decisive, in order that the minimum inhibitory concentration (MIC) of the agent is exceeded over the whole period that the spacer remains in place [3]. It is considered that a concentration 8–10 times higher than the MIC is required for an antibiotic to function effectively [4]. The concentration in the surroundings of the spacer depends not only on the amount of the active agent bound in the cement [5]. Release kinetics have been described in a study by van der Belt [6] as a combination of surface properties and porosity, exhibiting a biphasic course (burst-sustained release) and whose size can be calculated mathematically using the Higuchi equation [7]. Previous literature has described different variables influencing an improved rate of release from the cement matrix. Among other things, this is influenced by the ability of dissolution fluids to penetrate into the pores and cracks, a property that is closely linked to the wettability of the polymer matrix [6]. A study by Kuechle et al. [8] has suggested that the addition of additives such as T-DextranÒ during the mixing and polymerization phases of polymethylacrylate (PMMA) cement is beneficial with respect to the release of a variety of antibiotics. Given the simple and versatile applicability of dextran, this therapeutic approach would seem to be of
123
Eur J Orthop Surg Traumatol
interest, which led us to perform the modified scientific analysis presented here. As the use of the dry powder T-DextranÒ used by Kuechle is only permitted in Germany for study or technical uses, we approached the question of whether fluid dextran, widely used in clinical routine, showed a comparably good effect on release of different antibiotics as that demonstrated by Kuechle for T-DextranÒ.
Materials and Methods Twelve systematic test series were applied, each loaded with different components (Table 1). To this end, we produced standardized cement test specimens loaded with test agents, which were then subjected to both an agar diffusion test and an elution test. We used three bone cements in the production of the test specimens: Refobacin PalacosR40Ò (0.8 g gentamicin sulfate–0.5 g gentamicin), CopalÒ cement (1.7 g gentamicin sulfate and 1.2 g clindamycin hydrochloride–1 g gentamicin, molecular weight: M477, 6 g/mol and 1 g clindamycin, molecular weight: M461, 5 g/mol), and Palacos R40Ò (all Biomet-Europe). Dextran was not added to either the gentamicin–clindamycin and vancomycin test specimens, or those specimens loaded with saline, and these therefore functioned as reference groups for the dextran-enriched samples. Production of the test specimens The production and processing of the test specimens (Fig. 1) took place according to the ISO 5833 standard. In Table 1 Overview of the 12 test series Cement series
Cement
Supplement Ò
Ò
RP
Refobacin -Palacos R
None
RP10*
RefobacinÒ-PalacosÒR
10 ml NaCl
Ò
Ò
RP5*
Refobacin -Palacos R
5 ml NaCl
RP10#
RefobacinÒ-PalacosÒR
10 ml Dextran
RP5#
RefobacinÒ-PalacosÒR
5 ml Dextran
RPF
RefobacinÒ-PalacosÒR
2 g Fosfomycin
RPF10*
RefobacinÒ-PalacosÒR
2 g Fosfomycin and 10 ml NaCl
RPF5*
RefobacinÒ-PalacosÒR
2 g Fosfomycin and 5 ml NaCl
RPF10#
RefobacinÒ-PalacosÒR
2 g Fosfomycin and 10 ml dextran
RPF5#
RefobacinÒ-PalacosÒR
2 g Fosfomycin and 5 ml dextran
CPL
CopalÒ
None
PV
PalacosÒR
2 g Vancomycin
123
Fig. 1 Test specimens
order to provide an equal starting point for all test specimens without the influence of variation in the number and dimension of pores, all specimens were produced under vacuum conditions (Optivac Total Hip Kit, Biomet-Europe) and the polymerization time was measured. The powdered components were mixed using a sterile spatula (fractional preparation of the antibiotic) before addition of the monomer in order to achieve a homogenous dispersal of the applied antibiotics fosfomycin (InfectofosÒ 2 g, molecular weight: M:182 g/mol InfectoPharm, Arzneimittel und Consilium GmbH, Heppenheim, Germany) and vancomycin (Vanco-saarÒ 1 g, molecular weight: M:1,486 g/mol, MIP Pharma GmbH, Blieskastel-Niederwu¨rzbach, Germany) and the polymer. According to the test protocol, the liquids (NaCl sterile solution 0.9 %, B. Braun Medical AG, Sempach, Switzerland, and Dextran 40-DeltadexÒ 40, active agent 10 % Dextran 40 in 0.9 % NaCl sterile solution, DeltaSelect GmbH, Pfullingen, Germany) were first mixed with the fluid monomer by stirring with a spatula, at the end of which the polymer was added. Four minutes after initiation of the polymerization reaction, the polymerizate was applied to the lower border plate of a sterile standardized perforated plate using a cement spray gun (Optigun, Biomet-Europe). The plate had 18 identical holes, each with a diameter of 10 mm and a depth of 2.5 mm. The upper plate was laid in place, and the structure was compressed for 14 min using a vise (Reco Werkzeugmaschinen Handels GmbH, Naumburg, Germany). This exerted a pressure of around 4 tons. After 20 min, the resulting completely hardened test specimen was removed from the perforated plate using sterile forceps. All test specimens were tested against three bacterial strains: methicillin-sensitive S. aureus (MSSA, strain DSM 1104); methicillin-resistant S. aureus (MRSA, strain ATCC 33591); as well as Bacillus subtilis acting as index organism as it posses a very good response rate to antibiotics. For
Eur J Orthop Surg Traumatol
the tests against the pathogens MSSA and MRSA, it was necessary to produce a bacterial suspension for inoculation of the agar dishes (Mu¨ller Hinton II Agar, BD DiagnosticSystems, Heidelberg, Germany). Industrially, pre-inoculated non-incubated agar plates are available for the tests against Bacilus subtilis (Bacillus-subtilis-Agar pH 8.0, Ref.381e, Heipha Dr. Mu¨ller GmbH, Eppelheim, Germany) with a pathogen density of 106–107 KBE/ml. In line with the Clinical Laboratory Standards (CLSI) guidelines (Institute CaLS. Performance Standards for Antimicrobial Disk Susceptibility Tests), the bacterial suspensions of the MSSA and MRSA strains were diluted to obtain a pathogenic density of 1–5 9 106. The cultures so obtained were then placed within 15 min on Mu¨ller-Hinton agar plates that had been stored in the dark at 6 °C until use. Daily control of suspension, culture, and methodology was performed using susceptibility test disks with gentamicin and vancomycin (BD Sensi-DiscTM gentamicin 10 lg, BD Sensi-DiscTM vancomycin 30 lg, BD Diagnostic-Systems, Heidelberg, Germany). All plates were incubated at 36 °C. In the agar diffusion test, the test specimen was pressed into the cultured agar plate using sterile forceps. After 24 h, the resulting zone of inhibition was measured, and subsequently, the pellet was transferred to a new agar plate [9]. This procedure continued for 10 days. Each step was replicated five times in order to increase the accuracy of the results. In the second part of the study (elution test), the test specimens were placed in a sterile, screw-topped plastic container using sterile forceps. This was then filled with 10 ml 0.9 % saline and placed for 24 h in an incubator at 36 °C. A sterile pipette was then used to instill 0.5 ll eluate on a pre-prepared agar plate. The test specimen itself was then transferred to a new saline-filled container and returned to the incubator further for a 24 h, after which eluate was again removed as previously described. The whole procedure was carried out for up to 10 days, each time with 5 replicates. After each incubation on an agar plate, the resulting zone of inhibition was measured using an electronic caliper (Digimatic Solar Digital Caliper, Miyutoyo Europe GmbH, Neuss, Germany).
grounds, only the results of the elution tests were included in the statistical analysis. The results of the elution tests were analyzed with the SPSS program 17.0, using the Mann–Whitney U test, and are presented in the form of p value tables. The Mann– Whitney U test compares the central tendency (median) of two populations with values on an ordinal scale. This takes into account the entire ranking information. All questions were subject to explorative evaluation. No adjustment was made to the level of significance with regard to multiple testing. All p values are therefore to be taken as purely descriptive.
Results All test series showed an initially higher release of the active agent, as revealed by the relevant zone of inhibition, among supplemented antibiotics, which faded rapidly. In some test series, a relevant zone of inhibition could no longer be found for only a few days after the start of the experiment. The size of the zone of inhibition among the test specimens without supplement with liquid components (RPF, PV; RP, CPL) was markedly lower for the monantibiotic musters vanomycin (PV) and gentamicin (RP) than that seen in antibiotic combinations. The smallest zones of inhibition obtained in the tests of effectiveness against Bacillus subtilis, with an average radius of 11 mm, were seen in the test series PalacosÒR plus vancomycin (PV). RefobacinÒ-PalacosÒR plus fosofomycin (RPF), in contrast, showed the best results statistically (p = 0.009),
Statistics The values from the plates (agar diffusion) and the elution tests were graphically presented as box-plot diagrams. By this means, the median, interquartile range, and the extreme values were compared. The interquartile range in the agar diffusion tests was consistently larger than in the elution tests, which indicates a greater variation in the results. Reproducible results were therefore only to be expected in the elution tests. On these
Fig. 2 Elution test; MRSA; cement with powder supplement; only RPF and CPL were effective over the entire period of measurement, while PV already showed a marked reduction in effectiveness on the second day
123
Eur J Orthop Surg Traumatol Table 2 p value Mann–Whitney U Test; elution test; MRSA; cement with powder supplement Test series
Day 1
2
3
4
5
6
7
8
9
10
RP–CPL
0.009
0.009
0.008
0.007
0.005
0.005
0.005
0.005
0.005
0.005
RP–RPF
0.009
0.009
0.009
0.007
0.005
0.005
0.005
0.005
0.005
0.005
RP–PV
0.009
0.465
0.028
0.317
1.000
1.000
1.000
1.000
1.000
1.000
RPF–CPL
0.009
0.009
0.008
0.009
0.009
0.028
0.009
0.009
0.009
0.009
RPF–PV
0.175
0.009
0.009
0.005
0.005
0.005
0.005
0.005
0.005
0.005
CPL–PV
0.047
0.009
0.008
0.005
0.005
0.005
0.005
0.005
0.005
0.005
followed by CopalÒ cement. RefobacinÒ-PalacosÒR plus fosfomycin (RPF) also produced consistently the largest zones of inhibition, against both MSSA (p = 0.009) and MRSA (p = 0.009) (Fig. 2; Table 2). Constant effectiveness against MRSA in both agar diffusion and elution testing was achieved only by RefobacinÒ-PalacosÒR plus fosfomycin (RPF) and by CopalÒ (CPL). PalacosÒR plus vancomycin (PV), in contrast, was effective against MSSA and MRSA on the first 2 or 3 days, respectively. The zones of inhibition were also clearly smaller (10 mm on day 1, 23 mm on day 2) than those seen with RefobacinÒ-PalacosÒR plus fosfomycin (RPF). After day 3, PalacosÒR plus vancomycin (PV) showed no further inhibition (p = 0.005). In the fosfomycin-free test series with the fluids NaCl and dextran (RP10*, RP5*, RP10#, RP5#), the size of zone of inhibition against Bacilus subtilis followed an undulant course. The zone of inhibition against MSSA exhibited by the reference assay RefobacinÒ-PalacosÒR (RP) was smallest on the first and third day, whereas the largest zones were seen on the second day. From day 6, a clear development was visible (p = 0.047), and from this point on, RefobacinÒ-PalacosÒR (RP) consistently produced the largest zones of inhibition. On day 4, RefobacinÒ-PalacosÒR plus 5 ml dextran (RP5#) produced the smallest zone of inhibition seen in any of the test series. Apart from the first 2 days, no statistical difference could be seen in the size of the zone of inhibition between RefobacinÒPalacosÒR (RP) and cement with a 5 or 10 ml NaCl supplement. The Mann–Whitney test revealed the superiority of RefobacinÒ-PalacosÒR (RP) as reference cement to the assays supplemented with dextran (p = 0.006). In elution testing against MRSA, effectiveness in all of the test series was restricted to a few days. All of the tested cements showed an inhibitory effect on the first 2 days. From day 3, only RefobacinÒ-PalacosÒR (RP) showed a small inhibition of growth, while the other test series showed no further zones of inhibition. In the test series with fosfomycin and fluid supplements (RPF, RPF10*, RPF5*, RPF10#, RPF5#), RefobacinÒPalacosÒR plus fosfomycin (RPF) produced the largest
123
Fig. 3 Elution test; MSSA; cement with fluid supplement; a reduction of the zone of inhibition over time was seen in all the experimental series, wherein samples mixed with dextran initially showed comparable results, but from the fourth day had the least inhibitory effect
zones of inhibition over the entire time course, irrespective of the pathogen concerned. The exception was test series against MRSA, in which RefobacinÒ-PalacosÒR plus fosfomycin and 5 ml dextran (RPF5#) predominated on day 5 (p = 0.147). Larger zones of inhibition were also visible against MSSA (Fig. 3; Table 3) on days 3–6. The test series with 10-ml fluid supplement produced on average 3.5–5 mm smaller zones of inhibition than RefobacinÒPalacosÒR plus fosfomycin and 5 ml dextran (RPF5#) over the entire course of testing, with NaCl supplementation leading to smaller zones of inhibition than dextran.
Discussion Some authors suggest that the movement of antibiotics through a structure such as bone cement is dependent on the Fick principle relying on the concentration gradient, the thickness of the membrane, and the diffusion coefficient of the substance. The concept of a diffusion model
Eur J Orthop Surg Traumatol Table 3 p value Mann–Whitney U Test; elution test; MSSA; cement with fluid supplement Test series
Day 1
2
3
4
5
6
7
8
9
10
RP–RP10*
0.009
0.009
0.009
0.009
0.346
0.047
0.009
0.009
0.028
0.028
RP–RP5*
0.009
0.009
0.009
0.009
0.754
0.016
0.012
0.009
0.016
0.009
RP–RP10#
0.009
0.009
0.602
0.251
0.076
0.047
0.016
0.016
0.007
0.007
RP–RP5#
0.009
0.600
0.009
0.009
0.009
0.009
0.117
0.009
0.007
0.005
also circulates widely, in which the porosity of the cement is responsible for the magnitude of release. According to this thesis, the surface-to-volume ratio is considered as decisive for the extent of the release of the agent from the cement matrix, with the size and dimensions of the pores acting as an indicator of the emission capacity [10]. It is agreed that increasing the surface area of the cement and the improved transport of fluids that results, along with the permeability, has an influence on the release kinetics of antibiotics [11]. In the present study, we analyzed the extent to which the addition of liquid dextran brought about an improvement in the release of antibiotics. The majority of publications [12, 13] do not recommend liquid additives, citing their negative influence on cement stability, but other authors have reported promising results using liquid additives [14]. We compared fosfomycin, an antibiotic with a relatively small molecular mass, with established cement-antibiotic composites. CopalÒ, which has been shown to act against biofilm formation and to possess a high antibiotic release rate [15], and vancomycin cement were used as reference products. Studies carried out in the Far East on the thermostability of the epoxide-antibiotic fosfomycin [16] acted as the foundations of our study. In the test series without addition of dextran, the largest zones of inhibition against all the tested bacteria (Bacillus species, MSSA, MRSA) were produced by RefobacinPalacosÒR40 (currently PalacosÒR ? G-Heraeus-Medical) plus fosfomycin (RPF), followed by CopalÒ. PalacosÒR plus vancomycin cement (PV), in contrast, was the least effective option. A significant aspect of these observations is the fact that the cement-bound antibiotic combinations (RPF,CPL) influence each others release from the cement matrix [17] so that the elution of each of the two agents is more marked than when they are applied alone (PV) [15]. The second antibiotic thus acts as a kind of soluble filler, similar to the role assumed to be played by dextran. The majority of the studies concerned with this matter conclude that the biantibiotic reinforcement of PMMA leads to better release rates than is the case in monotherapy [18]. The comparatively low elution rate of vancomycin could, however, have other causes connected to the
chemical properties of the agent, its molecular size, its stability in the presence of the exothermic cement polymerization, and the technical properties of the cement [19]. In most studies, however, the poor release of the agent is attributed to the design of the cement medium rather than the structure of vancomycin itself. New developments in the cement field show that changes in cement composition can achieve improved elution rates for vancomycin. The replacement of zirconium dioxide with calcium carbonate in CopalÒ-Spacem cement (Hereus-Medical, Wehrheim, Germany), a cement specially conceived for spacers, for example, due to its hydrophilic properties, makes it capable of transporting fluids into the core of the element and not just through the surface, which can also bring advantages in the use of mono-antibiotics [20]. It is an interesting observation that in our study, fosfomycin proved more effective against all pathogens than CopalÒ. With the same starting point, with respect to the carrier polymer, it showed a better release over the whole period of observation. It is probable that the fact that the size of the agent is a parameter in calculating the Higuchi equation plays a role here, with reduction in the mass of the antibiotic leading to higher rates of elution. In this context, the superior solubility in water of fosfomycin (4.69e?01 g/l) compared to the other antibiotics (clindamycin 3.10e?00 g/l, vancomycin 2.25e-01 g/l, gentamicin 1.26e?01 g/l) probably also plays a central role. Under the conditions of our study, the admixture of dextran and RefobacinÒ-PalacosÒR (RP) had no positive effects on the size of the zone of inhibition. Throughout the samples with 5 or 10 ml supplementary dextran were even less effective than the control series with NaCl or those free of fluid supplements. This result stands in stark contradiction to the results published by Kuechle. We think that we have a plausible explanation why our results are so far from those obtained by Kuechle [8]: A change in the polymer-to-monomer ratio, which is given as R = 1.83 for Refobacin PalacosÒ [20], has a lasting influence on the porosity of the cement matrix. A surplus at the polymer has a positive effect on the emission of antibiotic, whereas the opposite produces the inverse relationship [12]. Supplementation with dextran has a negative effect on this relationship and reduces the porosity. A range of data in the
123
Eur J Orthop Surg Traumatol
literature suggests that in this conformation, the resultant change in the duration of polymerization influences the elution of medication [21]. This interplay also underlies the slow phase of the release of medication, which means that the sustained release must be a function of the porosity of the cement. The reports of an improved release of antibiotics in the studies of Virto [13] can probably be viewed similarly and, in the end, are to be explained by the massive reduction (50 %) in the monomer content. Another explanation for the ineffectiveness of liquid additives lays in the size, configuration, and confluence of the sides of the pores generated by the filler. McLaren, for example, concluded that the creation of smaller pores by low-volume ‘‘soluble filler’’ could lead, through interconnectivity, to the production of larger effective surface areas [11, 22]. He suggested that the so-called interconnecting porosity positively influences the release of the contained antibiotic [23]. Dextran, with its stately Stokes radius of 4.45 mm, will necessarily produce coarsely porous structures due to its molecular size. Dunne has postulated that such a molecular size (chitosan in his study) can block the release of smaller antibiotic molecules [22]. The summary of the results shows that dextran in liquid form has not been proven successful in improving the release rate of antibiotics from PMMA cement. Along with standard cements such as CopalÒ, fosfomycin, due to its good elution rate from acrylic cement and its wide pathogenic spectrum against grampositive and gram-negative pathogens, has shown good potential for application as an antibiotic against problem pathogens. Newer cement with a favorable diffusion rate, such as CopalÒSpacem (Heraeus Medical, Wehrheim, Germany), should be examined in further studies to establish their potential for releasing fosfomycin. The comparison of different PMMA bone cements is a limiting factor in this study. But having said that, the usage of ISO 5833 guidelines in the manufacture of the pellets ensures thorough comparability. Acknowledgments We thank Dr. Rainer Specht (Biomet Merck Biomaterials GmbH, Darmstadt, Germany) for technical support, for providing materials, and for allowing us to use their laboratory facilities. This paper contains data from an as yet unpublished thesis by the author BK. The experiments comply with the current laws of Germany. Conflict of interest
None.
References 1. Lieberman JR, Callaway GH, Salvati EA, Pellicci PM, Brause BD (1994) Treatment of the infected total hip arthroplasty with a two-stage reimplantation protocol. Clin Orthop Relat Res 301:205–212 [Epub 1994/04/01]
123
2. Springer BD, Lee GC, Osmon D, Haidukewych GJ, Hanssen AD, Jacofsky DJ (2004) Systemic safety of high-dose antibiotic-loaded cement spacers after resection of an infected total knee arthroplasty. Clin Orthop Relat Res 427:47–51 [Epub 2004/11/24] 3. Fink B, Vogt S, Reinsch M, Buchner H (2011) Sufficient release of antibiotic by a spacer 6 weeks after implantation in two-stage revision of infected hip prostheses. Clin Orthop Relat Res 469(11):3141–3147 [Epub 2011/06/17] 4. Nelson CL, Griffin FM, Harrison BH, Cooper RE (1992) In vitro elution characteristics of commercially and noncommercially prepared antibiotic PMMA beads. Clin Orthop Relat Res 284:303–309 [Epub 1992/11/01] 5. McLaren RL, McLaren AC, Vernon BL (2008) Generic tobramycin elutes from bone cement faster than proprietary tobramycin. Clin Orthop Relat Res 466(6):1372–1376 [Epub 2008/03/15] 6. van de Belt H, Neut D, Schenk W, van Horn JR, van der Mei HC, Busscher HJ (2000) Gentamicin release from polymethylmethacrylate bone cements and Staphylococcus aureus biofilm formation. Acta Orthop Scand 71(6):625–629 [Epub 2001/01/06] 7. Higuchi T (1961) Rate of release of medicaments from ointment bases containing drugs in suspension. J Pharm Sci 50:874–875 [Epub 1961/10/01] 8. Kuechle DK, Landon GC, Musher DM, Noble PC (1991) Elution of vancomycin, daptomycin, and amikacin from acrylic bone cement. Clin Orthop Relat Res 264:302–308 [Epub 1991/03/01] 9. Institute CaLS (2012) Performance standards for antimicrobial disk susceptibility tests; Approved Standard. 950 West Valley Road, Suite 2500, Wayne, PA 19087, USA, M02-A11 Vol. 32 No. 1 Replaces M02-A10, Vol. 29 No. 1(Eleventh Edition):1–15 10. Shiramizu K, Lovric V, Leung A, Walsh WR (2008) How do porosity-inducing techniques affect antibiotic elution from bone cement? An in vitro comparison between hydrogen peroxide and a mechanical mixer. J Orthop Traumatol Off J Italian Soc Orthop Traumatol 9(1):17–22 [Epub 2009/04/23] 11. McLaren AC, McLaren SG, McLemore R, Vernon BL (2007) Particle size of fillers affects permeability of polymethylmethacrylate. Clin Orthop Relat Res 461:64–67 [Epub 2007/09/01] 12. Downes S (1991) Methods for improving drug release from poly(methyl)methacrylate bone cement. Clin Mater 7(3):227–231 [Epub 1990/12/10] 13. Virto MR, Frutos P, Torrado S, Frutos G (2003) Gentamicin release from modified acrylic bone cements with lactose and hydroxypropylmethylcellulose. Biomaterials, 24(1):7–87 [Epub 2002/11/06]. doi:10.1016/S0142-9612(02)00254-5 14. Seldes RM, Winiarsky R, Jordan LC, Baldini T, Brause B, Zodda F et al (2005) Liquid gentamicin in bone cement: a laboratory study of a potentially more cost-effective cement spacer. J Bone Joint Surg Am 87(2):268–272 [Epub 2005/02/03] 15. Ensing GT, van Horn JR, van der Mei HC, Busscher HJ, Neut D (2008) Copal bone cement is more effective in preventing biofilm formation than Palacos R-G. Clin Orthop Relat Res 466(6):1492–1498 [Epub 2008/03/14] 16. Buranapanitkit B, Srinilta V, Ingviga N, Oungbho K, Geater A, Ovatlarnporn C (2004) The efficacy of a hydroxyapatite composite as a biodegradable antibiotic delivery system. Clin Orthop Relat Res 424:244–252 [Epub 2004/07/09] 17. Cui Q, Mihalko WM, Shields JS, Ries M, Saleh KJ (2007) Antibiotic-impregnated cement spacers for the treatment of infection associated with total hip or knee arthroplasty. J Bone Joint Surg Am 89(4):871–882 [Epub 2007/04/04] 18. Anagnostakos K, Kelm J, Regitz T, Schmitt E, Jung W (2005) In vitro evaluation of antibiotic release from and bacteria growth inhibition by antibiotic-loaded acrylic bone cement spacers. J Biomed Mater Res B Appl Biomater 72(2):373–378 [Epub 2004/12/04]
Eur J Orthop Surg Traumatol 19. Seligson D, Popham GJ, Voos K, Henry SL, Faghri M (1993) Antibiotic-leaching from polymethylmethacrylate beads. J Bone Joint Surg Am 75(5):714–720 [Epub 1993/05/01] 20. Neut D, Kluin OS, Thompson J, van der Mei HC, Busscher HJ (2010) Gentamicin release from commercially-available gentamicin-loaded PMMA bone cements in a prosthesis-related interfacial gap model and their antibacterial efficacy. BMC Musculoskelet Disord 11:258 [Epub 2010/11/12] 21. Shinsako K, Okui Y, Matsuda Y, Kunimasa J, Otsuka M (2008) Effects of bead size and polymerization in PMMA bone cement
on vancomycin release. Bio-Med Mater Eng 18(6):377–385 [Epub 2009/02/07] 22. Dunne N, Buchanan F, Hill J, Newe C, Tunney M, Brady A et al (2008) In vitro testing of chitosan in gentamicin-loaded bone cement: no antimicrobial effect and reduced mechanical performance. Acta Orthop 79(6):851–860 [Epub 2008/12/17] 23. McLaren AC, Nelson CL, McLaren SG, De CGR (2004) The effect of glycine filler on the elution rate of gentamicin from acrylic bone cement: a pilot study. Clin Orthop Relat Res 427:25–27 [Epub 2004/11/24]
123
ORIGINAL ARTICLE
Liquid dextran does not increase the elution rate of different antibiotics from bone cement K. E. Roth • B. Krause • E. Siegel • G. Maier C. Schoellner • P. M. Rommens
•
Received: 6 November 2013 / Accepted: 30 December 2013 Ó Springer-Verlag France 2014
Abstract Purpose To investigate the possibility of increasing elution of fosfomycin, gentamicin, clindamycin, and vancomycin by the addition of dextran fluid during the cementmixing phase. Methods In 12 test series, we produced standardized, antibiotic-loaded test specimens of cement, with and without addition of dextran, and determined their effectiveness against three reference pathogens in agar diffusion and elution tests. Results In the test series using combined agents, RefobacinÒ-PalacosÒR plus fosfomycin continuously produced the largest zone of inhibition, both against methicillinsensitive Staphylococcus aureus (p = 0.009) and against methicillin-resistant Staphylococcus aureus (p = 0.009). The addition of dextran to the various test series had no useful effect on the size of the zone of inhibition for any of the antibiotics tested. Conclusions Dextran supplementation in RefobacinÒPalacosÒR bone cement did not have the hope for positive effect on the elution rate of bound antibiotics. Keywords Dextran
Bone cement Antibiotics Elution rate
K. E. Roth (&) B. Krause G. Maier C. Schoellner P. M. Rommens Department of Orthopaedic Surgery, Johannes Gutenberg University, Langenbeckstraße 1, 55131 Mainz, Germany e-mail: [email protected] E. Siegel Department of Microbiology, Johannes Gutenberg University, Mainz, Germany
Introduction Bone cement has been used successfully for many decades to anchor endoprostheses and for localized delivery of antibiotics. In this function, they act, applied in the form of placeholders or chains, as a drug-delivery system that deploys its effects directly at the surgical wound site. Spacer systems are used in septic prosthesis surgery when sequential therapy procedures are preferred [1]. For the successful treatment of the infection, particularly in the use of glycosides [2], the amount of the active agent released from the bone cement is decisive, in order that the minimum inhibitory concentration (MIC) of the agent is exceeded over the whole period that the spacer remains in place [3]. It is considered that a concentration 8–10 times higher than the MIC is required for an antibiotic to function effectively [4]. The concentration in the surroundings of the spacer depends not only on the amount of the active agent bound in the cement [5]. Release kinetics have been described in a study by van der Belt [6] as a combination of surface properties and porosity, exhibiting a biphasic course (burst-sustained release) and whose size can be calculated mathematically using the Higuchi equation [7]. Previous literature has described different variables influencing an improved rate of release from the cement matrix. Among other things, this is influenced by the ability of dissolution fluids to penetrate into the pores and cracks, a property that is closely linked to the wettability of the polymer matrix [6]. A study by Kuechle et al. [8] has suggested that the addition of additives such as T-DextranÒ during the mixing and polymerization phases of polymethylacrylate (PMMA) cement is beneficial with respect to the release of a variety of antibiotics. Given the simple and versatile applicability of dextran, this therapeutic approach would seem to be of
123
Eur J Orthop Surg Traumatol
interest, which led us to perform the modified scientific analysis presented here. As the use of the dry powder T-DextranÒ used by Kuechle is only permitted in Germany for study or technical uses, we approached the question of whether fluid dextran, widely used in clinical routine, showed a comparably good effect on release of different antibiotics as that demonstrated by Kuechle for T-DextranÒ.
Materials and Methods Twelve systematic test series were applied, each loaded with different components (Table 1). To this end, we produced standardized cement test specimens loaded with test agents, which were then subjected to both an agar diffusion test and an elution test. We used three bone cements in the production of the test specimens: Refobacin PalacosR40Ò (0.8 g gentamicin sulfate–0.5 g gentamicin), CopalÒ cement (1.7 g gentamicin sulfate and 1.2 g clindamycin hydrochloride–1 g gentamicin, molecular weight: M477, 6 g/mol and 1 g clindamycin, molecular weight: M461, 5 g/mol), and Palacos R40Ò (all Biomet-Europe). Dextran was not added to either the gentamicin–clindamycin and vancomycin test specimens, or those specimens loaded with saline, and these therefore functioned as reference groups for the dextran-enriched samples. Production of the test specimens The production and processing of the test specimens (Fig. 1) took place according to the ISO 5833 standard. In Table 1 Overview of the 12 test series Cement series
Cement
Supplement Ò
Ò
RP
Refobacin -Palacos R
None
RP10*
RefobacinÒ-PalacosÒR
10 ml NaCl
Ò
Ò
RP5*
Refobacin -Palacos R
5 ml NaCl
RP10#
RefobacinÒ-PalacosÒR
10 ml Dextran
RP5#
RefobacinÒ-PalacosÒR
5 ml Dextran
RPF
RefobacinÒ-PalacosÒR
2 g Fosfomycin
RPF10*
RefobacinÒ-PalacosÒR
2 g Fosfomycin and 10 ml NaCl
RPF5*
RefobacinÒ-PalacosÒR
2 g Fosfomycin and 5 ml NaCl
RPF10#
RefobacinÒ-PalacosÒR
2 g Fosfomycin and 10 ml dextran
RPF5#
RefobacinÒ-PalacosÒR
2 g Fosfomycin and 5 ml dextran
CPL
CopalÒ
None
PV
PalacosÒR
2 g Vancomycin
123
Fig. 1 Test specimens
order to provide an equal starting point for all test specimens without the influence of variation in the number and dimension of pores, all specimens were produced under vacuum conditions (Optivac Total Hip Kit, Biomet-Europe) and the polymerization time was measured. The powdered components were mixed using a sterile spatula (fractional preparation of the antibiotic) before addition of the monomer in order to achieve a homogenous dispersal of the applied antibiotics fosfomycin (InfectofosÒ 2 g, molecular weight: M:182 g/mol InfectoPharm, Arzneimittel und Consilium GmbH, Heppenheim, Germany) and vancomycin (Vanco-saarÒ 1 g, molecular weight: M:1,486 g/mol, MIP Pharma GmbH, Blieskastel-Niederwu¨rzbach, Germany) and the polymer. According to the test protocol, the liquids (NaCl sterile solution 0.9 %, B. Braun Medical AG, Sempach, Switzerland, and Dextran 40-DeltadexÒ 40, active agent 10 % Dextran 40 in 0.9 % NaCl sterile solution, DeltaSelect GmbH, Pfullingen, Germany) were first mixed with the fluid monomer by stirring with a spatula, at the end of which the polymer was added. Four minutes after initiation of the polymerization reaction, the polymerizate was applied to the lower border plate of a sterile standardized perforated plate using a cement spray gun (Optigun, Biomet-Europe). The plate had 18 identical holes, each with a diameter of 10 mm and a depth of 2.5 mm. The upper plate was laid in place, and the structure was compressed for 14 min using a vise (Reco Werkzeugmaschinen Handels GmbH, Naumburg, Germany). This exerted a pressure of around 4 tons. After 20 min, the resulting completely hardened test specimen was removed from the perforated plate using sterile forceps. All test specimens were tested against three bacterial strains: methicillin-sensitive S. aureus (MSSA, strain DSM 1104); methicillin-resistant S. aureus (MRSA, strain ATCC 33591); as well as Bacillus subtilis acting as index organism as it posses a very good response rate to antibiotics. For
Eur J Orthop Surg Traumatol
the tests against the pathogens MSSA and MRSA, it was necessary to produce a bacterial suspension for inoculation of the agar dishes (Mu¨ller Hinton II Agar, BD DiagnosticSystems, Heidelberg, Germany). Industrially, pre-inoculated non-incubated agar plates are available for the tests against Bacilus subtilis (Bacillus-subtilis-Agar pH 8.0, Ref.381e, Heipha Dr. Mu¨ller GmbH, Eppelheim, Germany) with a pathogen density of 106–107 KBE/ml. In line with the Clinical Laboratory Standards (CLSI) guidelines (Institute CaLS. Performance Standards for Antimicrobial Disk Susceptibility Tests), the bacterial suspensions of the MSSA and MRSA strains were diluted to obtain a pathogenic density of 1–5 9 106. The cultures so obtained were then placed within 15 min on Mu¨ller-Hinton agar plates that had been stored in the dark at 6 °C until use. Daily control of suspension, culture, and methodology was performed using susceptibility test disks with gentamicin and vancomycin (BD Sensi-DiscTM gentamicin 10 lg, BD Sensi-DiscTM vancomycin 30 lg, BD Diagnostic-Systems, Heidelberg, Germany). All plates were incubated at 36 °C. In the agar diffusion test, the test specimen was pressed into the cultured agar plate using sterile forceps. After 24 h, the resulting zone of inhibition was measured, and subsequently, the pellet was transferred to a new agar plate [9]. This procedure continued for 10 days. Each step was replicated five times in order to increase the accuracy of the results. In the second part of the study (elution test), the test specimens were placed in a sterile, screw-topped plastic container using sterile forceps. This was then filled with 10 ml 0.9 % saline and placed for 24 h in an incubator at 36 °C. A sterile pipette was then used to instill 0.5 ll eluate on a pre-prepared agar plate. The test specimen itself was then transferred to a new saline-filled container and returned to the incubator further for a 24 h, after which eluate was again removed as previously described. The whole procedure was carried out for up to 10 days, each time with 5 replicates. After each incubation on an agar plate, the resulting zone of inhibition was measured using an electronic caliper (Digimatic Solar Digital Caliper, Miyutoyo Europe GmbH, Neuss, Germany).
grounds, only the results of the elution tests were included in the statistical analysis. The results of the elution tests were analyzed with the SPSS program 17.0, using the Mann–Whitney U test, and are presented in the form of p value tables. The Mann– Whitney U test compares the central tendency (median) of two populations with values on an ordinal scale. This takes into account the entire ranking information. All questions were subject to explorative evaluation. No adjustment was made to the level of significance with regard to multiple testing. All p values are therefore to be taken as purely descriptive.
Results All test series showed an initially higher release of the active agent, as revealed by the relevant zone of inhibition, among supplemented antibiotics, which faded rapidly. In some test series, a relevant zone of inhibition could no longer be found for only a few days after the start of the experiment. The size of the zone of inhibition among the test specimens without supplement with liquid components (RPF, PV; RP, CPL) was markedly lower for the monantibiotic musters vanomycin (PV) and gentamicin (RP) than that seen in antibiotic combinations. The smallest zones of inhibition obtained in the tests of effectiveness against Bacillus subtilis, with an average radius of 11 mm, were seen in the test series PalacosÒR plus vancomycin (PV). RefobacinÒ-PalacosÒR plus fosofomycin (RPF), in contrast, showed the best results statistically (p = 0.009),
Statistics The values from the plates (agar diffusion) and the elution tests were graphically presented as box-plot diagrams. By this means, the median, interquartile range, and the extreme values were compared. The interquartile range in the agar diffusion tests was consistently larger than in the elution tests, which indicates a greater variation in the results. Reproducible results were therefore only to be expected in the elution tests. On these
Fig. 2 Elution test; MRSA; cement with powder supplement; only RPF and CPL were effective over the entire period of measurement, while PV already showed a marked reduction in effectiveness on the second day
123
Eur J Orthop Surg Traumatol Table 2 p value Mann–Whitney U Test; elution test; MRSA; cement with powder supplement Test series
Day 1
2
3
4
5
6
7
8
9
10
RP–CPL
0.009
0.009
0.008
0.007
0.005
0.005
0.005
0.005
0.005
0.005
RP–RPF
0.009
0.009
0.009
0.007
0.005
0.005
0.005
0.005
0.005
0.005
RP–PV
0.009
0.465
0.028
0.317
1.000
1.000
1.000
1.000
1.000
1.000
RPF–CPL
0.009
0.009
0.008
0.009
0.009
0.028
0.009
0.009
0.009
0.009
RPF–PV
0.175
0.009
0.009
0.005
0.005
0.005
0.005
0.005
0.005
0.005
CPL–PV
0.047
0.009
0.008
0.005
0.005
0.005
0.005
0.005
0.005
0.005
followed by CopalÒ cement. RefobacinÒ-PalacosÒR plus fosfomycin (RPF) also produced consistently the largest zones of inhibition, against both MSSA (p = 0.009) and MRSA (p = 0.009) (Fig. 2; Table 2). Constant effectiveness against MRSA in both agar diffusion and elution testing was achieved only by RefobacinÒ-PalacosÒR plus fosfomycin (RPF) and by CopalÒ (CPL). PalacosÒR plus vancomycin (PV), in contrast, was effective against MSSA and MRSA on the first 2 or 3 days, respectively. The zones of inhibition were also clearly smaller (10 mm on day 1, 23 mm on day 2) than those seen with RefobacinÒ-PalacosÒR plus fosfomycin (RPF). After day 3, PalacosÒR plus vancomycin (PV) showed no further inhibition (p = 0.005). In the fosfomycin-free test series with the fluids NaCl and dextran (RP10*, RP5*, RP10#, RP5#), the size of zone of inhibition against Bacilus subtilis followed an undulant course. The zone of inhibition against MSSA exhibited by the reference assay RefobacinÒ-PalacosÒR (RP) was smallest on the first and third day, whereas the largest zones were seen on the second day. From day 6, a clear development was visible (p = 0.047), and from this point on, RefobacinÒ-PalacosÒR (RP) consistently produced the largest zones of inhibition. On day 4, RefobacinÒ-PalacosÒR plus 5 ml dextran (RP5#) produced the smallest zone of inhibition seen in any of the test series. Apart from the first 2 days, no statistical difference could be seen in the size of the zone of inhibition between RefobacinÒPalacosÒR (RP) and cement with a 5 or 10 ml NaCl supplement. The Mann–Whitney test revealed the superiority of RefobacinÒ-PalacosÒR (RP) as reference cement to the assays supplemented with dextran (p = 0.006). In elution testing against MRSA, effectiveness in all of the test series was restricted to a few days. All of the tested cements showed an inhibitory effect on the first 2 days. From day 3, only RefobacinÒ-PalacosÒR (RP) showed a small inhibition of growth, while the other test series showed no further zones of inhibition. In the test series with fosfomycin and fluid supplements (RPF, RPF10*, RPF5*, RPF10#, RPF5#), RefobacinÒPalacosÒR plus fosfomycin (RPF) produced the largest
123
Fig. 3 Elution test; MSSA; cement with fluid supplement; a reduction of the zone of inhibition over time was seen in all the experimental series, wherein samples mixed with dextran initially showed comparable results, but from the fourth day had the least inhibitory effect
zones of inhibition over the entire time course, irrespective of the pathogen concerned. The exception was test series against MRSA, in which RefobacinÒ-PalacosÒR plus fosfomycin and 5 ml dextran (RPF5#) predominated on day 5 (p = 0.147). Larger zones of inhibition were also visible against MSSA (Fig. 3; Table 3) on days 3–6. The test series with 10-ml fluid supplement produced on average 3.5–5 mm smaller zones of inhibition than RefobacinÒPalacosÒR plus fosfomycin and 5 ml dextran (RPF5#) over the entire course of testing, with NaCl supplementation leading to smaller zones of inhibition than dextran.
Discussion Some authors suggest that the movement of antibiotics through a structure such as bone cement is dependent on the Fick principle relying on the concentration gradient, the thickness of the membrane, and the diffusion coefficient of the substance. The concept of a diffusion model
Eur J Orthop Surg Traumatol Table 3 p value Mann–Whitney U Test; elution test; MSSA; cement with fluid supplement Test series
Day 1
2
3
4
5
6
7
8
9
10
RP–RP10*
0.009
0.009
0.009
0.009
0.346
0.047
0.009
0.009
0.028
0.028
RP–RP5*
0.009
0.009
0.009
0.009
0.754
0.016
0.012
0.009
0.016
0.009
RP–RP10#
0.009
0.009
0.602
0.251
0.076
0.047
0.016
0.016
0.007
0.007
RP–RP5#
0.009
0.600
0.009
0.009
0.009
0.009
0.117
0.009
0.007
0.005
also circulates widely, in which the porosity of the cement is responsible for the magnitude of release. According to this thesis, the surface-to-volume ratio is considered as decisive for the extent of the release of the agent from the cement matrix, with the size and dimensions of the pores acting as an indicator of the emission capacity [10]. It is agreed that increasing the surface area of the cement and the improved transport of fluids that results, along with the permeability, has an influence on the release kinetics of antibiotics [11]. In the present study, we analyzed the extent to which the addition of liquid dextran brought about an improvement in the release of antibiotics. The majority of publications [12, 13] do not recommend liquid additives, citing their negative influence on cement stability, but other authors have reported promising results using liquid additives [14]. We compared fosfomycin, an antibiotic with a relatively small molecular mass, with established cement-antibiotic composites. CopalÒ, which has been shown to act against biofilm formation and to possess a high antibiotic release rate [15], and vancomycin cement were used as reference products. Studies carried out in the Far East on the thermostability of the epoxide-antibiotic fosfomycin [16] acted as the foundations of our study. In the test series without addition of dextran, the largest zones of inhibition against all the tested bacteria (Bacillus species, MSSA, MRSA) were produced by RefobacinPalacosÒR40 (currently PalacosÒR ? G-Heraeus-Medical) plus fosfomycin (RPF), followed by CopalÒ. PalacosÒR plus vancomycin cement (PV), in contrast, was the least effective option. A significant aspect of these observations is the fact that the cement-bound antibiotic combinations (RPF,CPL) influence each others release from the cement matrix [17] so that the elution of each of the two agents is more marked than when they are applied alone (PV) [15]. The second antibiotic thus acts as a kind of soluble filler, similar to the role assumed to be played by dextran. The majority of the studies concerned with this matter conclude that the biantibiotic reinforcement of PMMA leads to better release rates than is the case in monotherapy [18]. The comparatively low elution rate of vancomycin could, however, have other causes connected to the
chemical properties of the agent, its molecular size, its stability in the presence of the exothermic cement polymerization, and the technical properties of the cement [19]. In most studies, however, the poor release of the agent is attributed to the design of the cement medium rather than the structure of vancomycin itself. New developments in the cement field show that changes in cement composition can achieve improved elution rates for vancomycin. The replacement of zirconium dioxide with calcium carbonate in CopalÒ-Spacem cement (Hereus-Medical, Wehrheim, Germany), a cement specially conceived for spacers, for example, due to its hydrophilic properties, makes it capable of transporting fluids into the core of the element and not just through the surface, which can also bring advantages in the use of mono-antibiotics [20]. It is an interesting observation that in our study, fosfomycin proved more effective against all pathogens than CopalÒ. With the same starting point, with respect to the carrier polymer, it showed a better release over the whole period of observation. It is probable that the fact that the size of the agent is a parameter in calculating the Higuchi equation plays a role here, with reduction in the mass of the antibiotic leading to higher rates of elution. In this context, the superior solubility in water of fosfomycin (4.69e?01 g/l) compared to the other antibiotics (clindamycin 3.10e?00 g/l, vancomycin 2.25e-01 g/l, gentamicin 1.26e?01 g/l) probably also plays a central role. Under the conditions of our study, the admixture of dextran and RefobacinÒ-PalacosÒR (RP) had no positive effects on the size of the zone of inhibition. Throughout the samples with 5 or 10 ml supplementary dextran were even less effective than the control series with NaCl or those free of fluid supplements. This result stands in stark contradiction to the results published by Kuechle. We think that we have a plausible explanation why our results are so far from those obtained by Kuechle [8]: A change in the polymer-to-monomer ratio, which is given as R = 1.83 for Refobacin PalacosÒ [20], has a lasting influence on the porosity of the cement matrix. A surplus at the polymer has a positive effect on the emission of antibiotic, whereas the opposite produces the inverse relationship [12]. Supplementation with dextran has a negative effect on this relationship and reduces the porosity. A range of data in the
123
Eur J Orthop Surg Traumatol
literature suggests that in this conformation, the resultant change in the duration of polymerization influences the elution of medication [21]. This interplay also underlies the slow phase of the release of medication, which means that the sustained release must be a function of the porosity of the cement. The reports of an improved release of antibiotics in the studies of Virto [13] can probably be viewed similarly and, in the end, are to be explained by the massive reduction (50 %) in the monomer content. Another explanation for the ineffectiveness of liquid additives lays in the size, configuration, and confluence of the sides of the pores generated by the filler. McLaren, for example, concluded that the creation of smaller pores by low-volume ‘‘soluble filler’’ could lead, through interconnectivity, to the production of larger effective surface areas [11, 22]. He suggested that the so-called interconnecting porosity positively influences the release of the contained antibiotic [23]. Dextran, with its stately Stokes radius of 4.45 mm, will necessarily produce coarsely porous structures due to its molecular size. Dunne has postulated that such a molecular size (chitosan in his study) can block the release of smaller antibiotic molecules [22]. The summary of the results shows that dextran in liquid form has not been proven successful in improving the release rate of antibiotics from PMMA cement. Along with standard cements such as CopalÒ, fosfomycin, due to its good elution rate from acrylic cement and its wide pathogenic spectrum against grampositive and gram-negative pathogens, has shown good potential for application as an antibiotic against problem pathogens. Newer cement with a favorable diffusion rate, such as CopalÒSpacem (Heraeus Medical, Wehrheim, Germany), should be examined in further studies to establish their potential for releasing fosfomycin. The comparison of different PMMA bone cements is a limiting factor in this study. But having said that, the usage of ISO 5833 guidelines in the manufacture of the pellets ensures thorough comparability. Acknowledgments We thank Dr. Rainer Specht (Biomet Merck Biomaterials GmbH, Darmstadt, Germany) for technical support, for providing materials, and for allowing us to use their laboratory facilities. This paper contains data from an as yet unpublished thesis by the author BK. The experiments comply with the current laws of Germany. Conflict of interest
None.
References 1. Lieberman JR, Callaway GH, Salvati EA, Pellicci PM, Brause BD (1994) Treatment of the infected total hip arthroplasty with a two-stage reimplantation protocol. Clin Orthop Relat Res 301:205–212 [Epub 1994/04/01]
123
2. Springer BD, Lee GC, Osmon D, Haidukewych GJ, Hanssen AD, Jacofsky DJ (2004) Systemic safety of high-dose antibiotic-loaded cement spacers after resection of an infected total knee arthroplasty. Clin Orthop Relat Res 427:47–51 [Epub 2004/11/24] 3. Fink B, Vogt S, Reinsch M, Buchner H (2011) Sufficient release of antibiotic by a spacer 6 weeks after implantation in two-stage revision of infected hip prostheses. Clin Orthop Relat Res 469(11):3141–3147 [Epub 2011/06/17] 4. Nelson CL, Griffin FM, Harrison BH, Cooper RE (1992) In vitro elution characteristics of commercially and noncommercially prepared antibiotic PMMA beads. Clin Orthop Relat Res 284:303–309 [Epub 1992/11/01] 5. McLaren RL, McLaren AC, Vernon BL (2008) Generic tobramycin elutes from bone cement faster than proprietary tobramycin. Clin Orthop Relat Res 466(6):1372–1376 [Epub 2008/03/15] 6. van de Belt H, Neut D, Schenk W, van Horn JR, van der Mei HC, Busscher HJ (2000) Gentamicin release from polymethylmethacrylate bone cements and Staphylococcus aureus biofilm formation. Acta Orthop Scand 71(6):625–629 [Epub 2001/01/06] 7. Higuchi T (1961) Rate of release of medicaments from ointment bases containing drugs in suspension. J Pharm Sci 50:874–875 [Epub 1961/10/01] 8. Kuechle DK, Landon GC, Musher DM, Noble PC (1991) Elution of vancomycin, daptomycin, and amikacin from acrylic bone cement. Clin Orthop Relat Res 264:302–308 [Epub 1991/03/01] 9. Institute CaLS (2012) Performance standards for antimicrobial disk susceptibility tests; Approved Standard. 950 West Valley Road, Suite 2500, Wayne, PA 19087, USA, M02-A11 Vol. 32 No. 1 Replaces M02-A10, Vol. 29 No. 1(Eleventh Edition):1–15 10. Shiramizu K, Lovric V, Leung A, Walsh WR (2008) How do porosity-inducing techniques affect antibiotic elution from bone cement? An in vitro comparison between hydrogen peroxide and a mechanical mixer. J Orthop Traumatol Off J Italian Soc Orthop Traumatol 9(1):17–22 [Epub 2009/04/23] 11. McLaren AC, McLaren SG, McLemore R, Vernon BL (2007) Particle size of fillers affects permeability of polymethylmethacrylate. Clin Orthop Relat Res 461:64–67 [Epub 2007/09/01] 12. Downes S (1991) Methods for improving drug release from poly(methyl)methacrylate bone cement. Clin Mater 7(3):227–231 [Epub 1990/12/10] 13. Virto MR, Frutos P, Torrado S, Frutos G (2003) Gentamicin release from modified acrylic bone cements with lactose and hydroxypropylmethylcellulose. Biomaterials, 24(1):7–87 [Epub 2002/11/06]. doi:10.1016/S0142-9612(02)00254-5 14. Seldes RM, Winiarsky R, Jordan LC, Baldini T, Brause B, Zodda F et al (2005) Liquid gentamicin in bone cement: a laboratory study of a potentially more cost-effective cement spacer. J Bone Joint Surg Am 87(2):268–272 [Epub 2005/02/03] 15. Ensing GT, van Horn JR, van der Mei HC, Busscher HJ, Neut D (2008) Copal bone cement is more effective in preventing biofilm formation than Palacos R-G. Clin Orthop Relat Res 466(6):1492–1498 [Epub 2008/03/14] 16. Buranapanitkit B, Srinilta V, Ingviga N, Oungbho K, Geater A, Ovatlarnporn C (2004) The efficacy of a hydroxyapatite composite as a biodegradable antibiotic delivery system. Clin Orthop Relat Res 424:244–252 [Epub 2004/07/09] 17. Cui Q, Mihalko WM, Shields JS, Ries M, Saleh KJ (2007) Antibiotic-impregnated cement spacers for the treatment of infection associated with total hip or knee arthroplasty. J Bone Joint Surg Am 89(4):871–882 [Epub 2007/04/04] 18. Anagnostakos K, Kelm J, Regitz T, Schmitt E, Jung W (2005) In vitro evaluation of antibiotic release from and bacteria growth inhibition by antibiotic-loaded acrylic bone cement spacers. J Biomed Mater Res B Appl Biomater 72(2):373–378 [Epub 2004/12/04]
Eur J Orthop Surg Traumatol 19. Seligson D, Popham GJ, Voos K, Henry SL, Faghri M (1993) Antibiotic-leaching from polymethylmethacrylate beads. J Bone Joint Surg Am 75(5):714–720 [Epub 1993/05/01] 20. Neut D, Kluin OS, Thompson J, van der Mei HC, Busscher HJ (2010) Gentamicin release from commercially-available gentamicin-loaded PMMA bone cements in a prosthesis-related interfacial gap model and their antibacterial efficacy. BMC Musculoskelet Disord 11:258 [Epub 2010/11/12] 21. Shinsako K, Okui Y, Matsuda Y, Kunimasa J, Otsuka M (2008) Effects of bead size and polymerization in PMMA bone cement
on vancomycin release. Bio-Med Mater Eng 18(6):377–385 [Epub 2009/02/07] 22. Dunne N, Buchanan F, Hill J, Newe C, Tunney M, Brady A et al (2008) In vitro testing of chitosan in gentamicin-loaded bone cement: no antimicrobial effect and reduced mechanical performance. Acta Orthop 79(6):851–860 [Epub 2008/12/17] 23. McLaren AC, Nelson CL, McLaren SG, De CGR (2004) The effect of glycine filler on the elution rate of gentamicin from acrylic bone cement: a pilot study. Clin Orthop Relat Res 427:25–27 [Epub 2004/11/24]
123
