Coloured bone cements have been introduced to make the removal of cement debris easier at the time of primary and revision joint replacement. We evaluated the physical, mechanical and pharmacological effects of adding methylene blue to bone cement with or without antibiotics (gentamicin, vancomycin or both). The addition of methylene blue to plain cement significantly decreased its mean setting time (570 seconds (sd 4) vs 775 seconds (sd 11), p = 0.01), mean compression strength (95.4 MPa (sd 3) vs 100.1 MPa (sd 6), p = 0.03), and mean bending strength (65.2 MPa (sd 5) vs 76.6 MPa (sd 4), p < 0.001) as well as its mean elastic modulus (2744 MPa (sd 97) vs 3281 MPa (sd 110), p < 0.001). The supplementation of the coloured cement with vancomycin and gentamicin decreased its mean bending resistance (55.7 MPa (sd 4) vs 65.2 MPa (sd 5), p < 0.001).The methylene blue significantly decreased the mean release of gentamicin alone (228.2 µg (sd 24) vs 385.5 µg (sd 26), p < 0.001) or in combination with vancomycin (498.5 µg (sd 70) vs 613 µg (sd 25), p = 0.018) from the bone cement. This study demonstrates several theoretical disadvantages of the antibiotic-loaded bone cement coloured with methylene blue.
A decrease in the rate of infection following total joint replacement has been reported in large retrospective studies with the use of antibiotic-loaded bone cement, mainly in total hip replacement (THR),1-3 following which the incidence of infection is currently about 1.0%.4 Accordingly, the routine use of antibiotic-loaded bone cement in primary THR has been supported.5 However, one problem associated with cement in primary procedures is its complete removal at the time of the revision surgery to prevent possible third-body wear and secondary loosening of components.6 Removal of cement can be a demanding task and occasionally results in increased loss of bone stock. Therefore, coloured cements have been introduced to facilitate their identification during removal. However, there is a risk that such additives might alter the characteristics of bone cement. To date there are only a limited number of studies concerning coloured products.7-11
A simple and cost-effective technique for creating coloured bone cement with7 or without antibiotics8 was introduced in the 1980s. This involves adding aqueous methylene blue to the cement, and has been recently recommended by Graves and Sands.9 The blue staining allows intra-operative recognition of the cement (Fig. 1), making its removal from bone much easier. There is evidence that the addition of the methylene blue to plain bone cement does not reduce its fatigue strength,10 fracture toughness11 or resistance to tension, compression or bending forces.8 However, no mechanical data on the addition of methylene blue to antibiotic-loaded cement have so far been published, other than with the supplementation with tobramycin.7 Furthermore, the effect of colourants on the elution properties of antibiotics from cement has not been investigated. Therefore we performed the present study to: 1) evaluate the chemical, physical and mechanical characteristics of different plain and antibiotic loaded cements with and without methylene blue; and 2) determine the kinetics of elution of antibiotics from cements with and without methylene blue.
Materials and Methods
Specimens were prepared from polymerised cement Cemex XL (Tecres SpA, Verona, Italy) with a mould under aseptic conditions. This cement is a 50 g radio-opaque low-viscosity material with a powder-to-liquid ratio of 2.7:1. Vancomycin chlorhydrate (1.25 g; Apharma Ltd, Barnstaple, United Kingdom) or gentamicin sulphate (1.25 g; Fujian Fukang Pharmaceutical Co., Fuzhou, China) in powder form, alone or in combination, were mixed manually with polymethylmethacrylate (PMMA) copolymer powder with a plastic spatula for 120 seconds to achieve a uniform powder prior to the addition of the liquid polymer under laminar flow. At that time, 1.25 cm3 of 1% aqueous methylene blue (Monico SpA, Venice, Italy) was added to half of the specimens, thoroughly mixing the components for an additional 90 seconds. Specimens with evident defects were discarded. There were eight experimental groups with different formulations (Table I). The mixing and test equipment was maintained at 23°C (standard deviation (sd) 1°C) for at least two hours before beginning the mixing procedure or the test. Specimens were aged in air at 23°C (sd 1) and > 40% relative humidity for 24 hours before being tested. Tests were performed at 23°C (sd 1).
The physical assays were conducted according to the ISO 5833:2002 policy for assessing acrylic resin cements.12 The setting time and the maximum polymerisation temperature (°C) were recorded using a TBX-68T terminal block and LabVIEW application software (National Instruments, Austin, Texas). The setting temperature was calculated using the formula (Tamb + Tmax)/2, where Tamb is the ambient temperature and Tmax is the maximum temperature recorded during the test. The test was repeated twice and the mean value was calculated.
The tests for compressive and bending strength were performed according to the aforementioned ISO recommendations. A total of ten specimens for each formulation and each type of test were used. The specimens for the compressive test (Fig. 2) were cylinders of 6 mm diameter and 12 mm height, produced in moulds. The deformation rate was set at 20 mm/min. The specimens for the four-point bending test were strips of 3.3 mm thickness, 10 mm width, and 75 mm length, produced in moulds (Fig. 3). The deformation rate was set at 5 mm/min and to avoid hysteresis error, a preload was applied in the set-up. A Galdabini SUN 1000 (Cesare Galdabini SpA, Varese, Italy) testing machine with a 10 kN cell load was used. The load and position accuracy of the testing machine were ± 1% of indicated load and ± 0.02 mm, respectively. The stroke resolution to measure the displacement of specimens was 0.001 mm.
Deflection for the bending evaluation was detected with an analogical deflection probe with a maximum extension of 10 mm, an accuracy of 1 µm and a resolution of 0.1 µm. The formula used for converting the force applied on the specimens to the stress levels was B = 3Fa/(bh2) where B (in MPa) is the bending strength, F (N) is the force at break, a is the distance between the inner and outer loading points (20 mm), b (mm) is the width of specimens and h (mm) is the thickness of the specimens.
Elution of antibiotics
Five cylinders of 6 mm diameter and 12 mm height for each formulation were separately immersed in glass tubes with a defined volume in an isotonic sterile saline solution (pH = 7.5) at 37°C in an aerobic incubator. The saline was completely removed and replaced with the same volume of fresh saline after one, four, 24, 48 and 72 hours, and seven, 14, 21 and 28 days of immersion, and multiple aliquots of the elution samples were frozen at 24°C until the final analysis. The elution samples were analysed in duplicate in the same experiment. Recovery of the antibiotics was calculated as the total eluted micrograms. Standard concentrations of antibiotics alone and in combination (1:1) were prepared in saline and processed along with the samples.
Antibiotic concentrations alone and in combination in the eluted samples and the standards were analysed using the standard large plate agar-well diffusion method13 according to the Clinical and Laboratory Standards Institute.14Bacillus subtilis spore suspension (Merck KGaA, Darmstadt, Germany) ATCC 6633 (American Type Culture Collection, Manassas, Virginia) with a final concentration of 0.02% seeded in Isosensitest Agar (Oxoid Ltd, Basingstoke, United Kingdom) was used as the test micro-organism, since it is susceptible to both gentamicin and vancomycin. Inoculated agar was poured into sterile, level 245 mm × 245 mm plastic bioassay dishes (Corning Inc., Corning, New York). After solidification, wells of 80 mm diameter were punched out and filled with 50 μl of solution (samples and standards). After overnight incubation at 37°C, the diameters of the inhibition zones for each of the standard and eluted samples were measured (Fig. 4). The unknown antibiotic concentrations in the samples were determined by standard curves created with known antibiotic concentrations and plotted on a semi-logarithmic scale (log concentration vs zone of inhibition). All samples and standard concentrations were assayed in duplicate. Gentamicin and vancomycin assays were linear over a range of 0.6 mg/l to 20.0 mg/l (R2 = 0.99 and R2 = 0.98, respectively) against B. subtilis, whereas the combination of the two antibiotics was linear over a range of 0.6 mg/l to 40.0 mg/l (R2= 0.95) against B. subtilis. In order to check the antimicrobial activity of methylene blue, serial concentrations of methylene blue (0.5 mg to 0.03 mg) were seeded in the same agar media.
All data were normally distributed and parametric tests were used. Two-tailed t-tests and analysis of variance (ANOVA) were used to evaluate the differences between two or more means, respectively. Once we determined that differences existed among multiple means, a Bonferroni test was used to determine which means differed.
When a significant difference between the means failed to emerge at the t-test or ANOVA, we did a post-hoc analysis to calculate the power (1 – β error probability) achieved by our statistical tests, given an α-value = 0.05, the sample size, and the effect size, calculated entering the mean for each group in the statistical program. Significance was set at p < 0.05. The SPSS software (SPSS Inc., Chicago, Illinois) and G*Power software (Institut fur Experimentelle Psychologie, Heinrich Heine Universitat, Dusseldorf, Germany) was used for the statistical analyses.
In all formulations except the gentamicin-loaded cement, where a purple colour was encountered, the addition of methylene blue imparted a blue colour to the cement. The setting times and maximum polymerisation temperatures of the different formulations of cement are shown in Table II. The presence of gentamicin, vancomycin or both antibiotics increased the setting time of plain cement (p = 0.001, p < 0.001 and p < 0.001, respectively, ANOVA and Bonferroni post-hoc test). The setting time decreased when aqueous methylene blue was added to acrylic cement with or without antibiotics. The polymerisation temperature of the methylene blue cements was higher in comparison to the contrast-free cements. We did not find any differences between the polymerisation temperature of plain or contrast cements loaded with different antibiotics. The post hoc analysis detected a power of 0.21 for these mean comparisons. Regarding the physical characteristics, all cement formulations except the plain cement loaded with both vancomycin and gentamicin (due to the higher setting time) satisfied the requirements of ISO 5833:2002.
The mechanical characteristics of the different formulations of bone cement are summarised in Table III; all formulations satisfied the ISO requirements. Overall, methylene blue cements exhibited lower compressive and bending strengths and a lower elasticity modulus in comparison with the no contrast cements. As shown in the table, supplementation with antibiotic resulted in a lower bending strength of the cement with respect to the plain formulations with and without methylene blue. Among the contrast-free cements, Bonferroni tests revealed that PMMA + vancomycin + gentamicin had lower flexural strength compared to both plain PMMA (p < 0.001) and PMMA + gentamicin (p < 0.05). Among the methylene blue cements, the bending strength and elasticity modulus of the vancomycin-loaded samples were lower and higher, respectively, compared with the vancomycin-free cements (p < 0.001). Among the contrast-free cements loaded with different antibiotics, no differences in the compressive strength and elastic modulus emerged. The power of the mean comparison tests for compressive strength and elastic modulus was 0.25 and 0.10, respectively.
The cements supplemented with combined gentamicin and vancomycin released a higher absolute quantity of antibiotic than the cements loaded with a single antibiotic (p < 0.001) (ANOVA and Bonferroni post-hoc test). Moreover, the two antibiotics acted synergistically to increase the elution rate; thus the amount of eluted gentamicin and vancomycin when they were jointly added to the cement was higher with respect to the amount eluted when they were added separately to the plain or contrast cement, although the differences were not significant (Table IV). Both the plain and contrast cement showed a burst release of antibiotics in the first day, followed by a prolonged low level release (Fig. 5). The release of antibiotics over time from the different formulations of cement with and without methylene blue is summarised in Table IV. The presence of contrast significantly decreased the total amount of eluted antibiotic in all groups except the vancomycin-PMMA group. The aqueous methylene blue alone showed a mild antimicrobial activity (Table V).
Development of new biomaterials and modifications to existing ones are the subject of ongoing research to try to improve function or increase reliability. After the alteration of a biomaterial, it must be re-evaluated to confirm that the alteration does not adversely affect the performance characteristics. Few studies dealing with the use of aqueous methylene blue to make a coloured bone cement with7,9 and without8,10,11 antibiotics have been published. Because of the lack of available preclinical data on this topic, we evaluated the physical, mechanical and pharmacological effects of adding a visual contrast agent to cement with or without antibiotics. Despite the small sample size and thus the low power of some statistical test results, we found several differences in physical, mechanical and pharmacological characteristics between the experimental groups. The setting time decreased when aqueous methylene blue was added to the cement, thus confirming previous reports.8,9 We noted that the addition of methylene blue to plain cement decreased its compression and bending strength as well as its elastic modulus. One previous study8 also demonstrated that the elastic modulus of a specific cement decreased when methylene blue was added. However, those authors found that the tension, compression and bending strength were not influenced by the supplementation with dye. The use of different cements and evaluation methods may help to explain this discrepancy. Indeed, the latter study used the American Society for Testing and Materials specifications for acrylic bone cements,15 whereas we followed the most recent requirements for mechanical testing of acrylic cements.12 Other studies have provided evidence that fracture toughness11 and fatigue strength10 are not modified when aqueous methylene blue is added to plain cement. However, the mechanical properties of cement are critically dependent on a large number of factors such as test parameters, mixing methods, ageing conditions and formulations.16 We did not evaluate the fatigue strength or fracture toughness, but only the flexural and compression strength of cement. Indeed, flexural testing is a suitable method to assess the mechanical performance of cement,8 because it combines elements of compression, tension and shear, which mimic the in vivo stresses. Recently, other options for preparing coloured cement have been described.17 This latter study showed that the addition of pigment did not affect the mechanical properties of plain cement. However, the effect on cement’s mechanical properties of adding dye and antibiotics was not investigated. Notably, we found that the supplementation with vancomycin and gentamicin decreases bending and compression resistance, respectively, of the contrast cement. To the best of our knowledge, no data on this topic have been published previously. However, the clinical relevance of these mechanical data is uncertain and remains to be further evaluated. Indeed, differences that are small in magnitude typically lack practical relevance. Despite the observed detrimental effect of methylene blue on the mechanical characteristics of acrylic cement, all the formulations satisfied the ISO requirements.
In our study the methylene blue showed a mild antimicrobial activity and this result supports its use as a urinary antiseptic or topical antiseptic.18
We confirmed the results of several studies on the pharmacokinetic characteristics of antibiotic-loaded cement. As in previous studies,19,20 we detected a peak in the elution rate of antibiotic within 24 to 48 hours for all formulations. The microbiological assay of vancomycin in agar medium demonstrates the poor diffusion of glycopolypeptides from bone cements, which was also reported in previous investigations in which Palacos (Haraeus Kutzer GmbH, Hanau, Germany) or Simplex (Stryker Orthopaedics, Mahwah, New Jersey) cements were used.21,22 However, the main finding of our study was that methylene blue significantly decreased the release of gentamicin alone or in combination with vancomycin from the cement.
Notably, we were able to identify a non-linear regression analysis formula to predict with good accuracy the elution of tested antibiotics over time with and without the dye, and this finding might be useful to future researchers and physicians performing joint arthroplasty surgery.
Some limitations of this study should be acknowledged. We investigated only two types of antibiotics. We chose to study gentamicin because of its broad antibacterial spectrum, heat stability, antibacterial activity at low concentrations and elution characteristics from PMMA.23 However, not all organisms associated with prosthetic infection are gentamicin-sensitive and vancomycin is often used in these cases.24 Therefore, the generalisation of our results to all antibiotics is not feasible. Future studies evaluating other antibiotics and brands of cement will be needed to confirm our findings. Additionally we only evaluated a single dose of antibiotics and methylene blue testing the supplementation of 50 g of bone cement with 1.25 g of vancomycin chlorhydrate and/or 1.25g of gentamicin sulphate. This dose was chosen because we considered the evaluation of any possible detrimental effect of methylene blue to be of the utmost importance when this visual contrast cement was used in arthroplasty. Indeed, routine prophylaxis requires an antibiotic dose as low as ≤ 1 g of powdered antibiotic per 40 g of cement.25 Notably it has been shown that higher concentrations as great as 3.6 g of antibiotic per 40 g of acrylic cement should be used to obtain effective therapeutic levels of antibiotics.26 The use of high concentration of antibiotics causes a decrease in the compressive strength of cement with a detrimental effect on implant fixation.27 However, this issue is negligible when an antibiotic-loaded cement spacer made with methylene blue is temporarily used to treat a septic joint after the removal of prosthetic implants (Fig. 1), whereas it appears to be crucial when a similar cement is prophylactically used in primary joint replacement.
We chose to evaluate the dose of 1.25 cm3 of 1% methylene blue per 50 g of cement because this proportion of dye was the one adopted in all the studies reporting on this visual contrast cement.7-9 Moreover, Davies and Harris10 evaluated three different methylene blue dye formulations added to cement and concluded that the optimum methylene blue impregnated cement was produced by adding 1 cm3 of a 1% methylene blue solution to 40 g of cement. In a preliminary investigation we performed, this dose adequately coloured the cement, thus enabling a clear distinction between this material and the surrounding bone or soft tissues to be made.
We are aware that hand-mixing could lead to clumping or inclusion voids in the cement, which may decrease its mechanical strength.28 Hand mixing is also associated with a decreased release of antibiotics in comparison with commercial premixed antibiotic-loaded cements, whereas vacuum mixing has been shown to result in only a minor reduction in antibiotic release.29,30 We used the hand mixing procedure because it represents the suggested procedure if contrast cement with aqueous methylene blue has to be prepared.7,9 An additional limitation of the study was that the cement samples were tested in isolation; future work should include how these cement samples behave when used in a bone-implant construct.
Notwithstanding these limitations, the strengths of our study include the use of reliable and valid measures. Indeed, proper comparison of results is impossible in the absence of standardised and fully described methods of testing. The comprehensive pharmacological, physical and mechanical evaluation, allowing the control of a large set of variables, is a further strength of our study.
In summary, this in vitro study demonstrates theoretical disadvantages of antibiotic-loaded bone cement coloured with methylene blue, although caution should be exercised in transferring our findings to the clinical context. However, based on our findings, we do not recommend methylene blue supplementation of PMMA for routine clinical use and believe that it should be demonstrated that the values of the relevant physical and mechanical parameters of this cement are not different from the corresponding values obtained with the plain variety of the same formulation. It should also be determined whether these differences outweigh the clinical benefit.
The authors would like to thank Tecres SpA (Verona, Italy), for providing laboratories to perform the tests of this study. No funds or grants have been received from Tecres SpA.
No benefits in any form have been received or will be received from a commercial party related directly or indirectly to the subject of this article.
- Received March 17, 2011.
- Accepted July 12, 2011.
- ©2011 British Editorial Society of Bone and Joint Surgery