Mechanical Properties and Porosity of Acrylic Cement Bone Loaded with Alendronate Powder.

Aseptic loosening is the most common complication of joint replacement. Previous studies showed that acrylic bone cement loaded with a commercially-available alendronate powder (APAC) had good promise against wear debris-mediated osteolysis for prevention of aseptic loosening. The purpose of the present study was to investigate the effect of adding alendronate powder to an acrylic bone cement on quasi-static mechanical properties (namely, compressive strength, compressive modulus, tensile strength, and flexural strength), fatigue life, porosity, and microstructure of the cement. The results showed that adding up to 1 wt./wt.% alendronate powder exerted no detrimental effect on any of the quasi-static mechanical properties. However, the fatigue life of APAC decreased by between ~17% and ~27 % and its porosity increased by between ~ 5-7 times compared with corresponding values for the control cement (no alendronate powder added). Fatigue life was negatively and significantly correlated with porosity. Considering that fatigue life of the cement plays a significant role in joint replacement survival, clinical use of APAC cannot be recommended.

Introduction

Aseptic loosening is the most common complication of joint replacement 1. Currently, there is no other treatment of aseptic loosening except revision arthroplasty 2. The effective prevention is critical in the control of aseptic loosening. It has been shown that oral-administered bisphosphonate, especially alendronate, effectively inhibits wear debris-induced osteolysis and reduces the risk of aseptic loosening 3. However, adverse events, notably, upper gastrointestinal ulcer, and a low bioavailability of 0.7% were frequently observed with oral-administered alendronate 4. Furthermore, poor drug compliance in patients compromised the therapeutic efficacy of alendronate 5.

In 1994, Hamish et al. put forward the idea of a bisphosphonate-loaded acrylic bone cement as a local drug delivery system 6. This proposal is similar to the addition of antibiotic(s) to an acrylic bone cement 7. The local high administration of bisphosphonate is achieved through sustained drug elution from the cement 8, 9. Our previous in vivo study showed that the local release of alendronate from an acrylic bone cement inhibited wear debris-mediated osteolysis and enhanced the mechanical strength of the host bone in a cemented arthroplasty 10. The proposed mechanism was that the eluted bisphosphonate stimulated an increase in osteoprotegerin ligand and a decrease in receptor activator of nuclear factor kappa-B ligand 11.

There have been a number of reports on bisphosphonate-loaded acrylic bone cement 12-15. Yu et al. 12 found loading liquid pamidronate reduced the bending strength of the cement by 39%. Lewis et al. 13 demonstrated that the fatigue life of the cured cement was reduced by 97% and 39% when alendronate was added in liquid form and in the form of tablet-ground powder, respectively (both values relative to that for the plain cement). These results were attributed to the interference of the polymerization process by the liquid alendronate and the excipient in the alendronate tablet 12, 13. Matuszewski et al. 14 found that addition of 60 mg of pamidronate powder to 40 g of cement did not have a detrimental effect on either the compressive strength or the flexural strength of the cement. Song et al. 15 showed that addition of no more than 500 mg of alendronate powder to 50 g of cement did not significantly decrease either the compressive strength or the bending strength of the cement. To the best of our knowledge, there are no studies that involve determination of both quasi-static mechanical properties and fatigue life of an acrylic bone cement loaded with the powder of a bisphosphonate.

The purpose of the present study was to determine the effect of adding a commercially-available alendronate powder to an acrylic bone cement (APAC) on the compressive strength, tensile strength, flexural strength, fatigue life, porosity, and microstructure of the cement.

Materials and Methods

Materials and preparation of test specimens

The alendronate powder (Merck, Sharp & Dohme, Hangzhou, China; particle size range of 4.67 - 44.21 µm and mean = 20.77 µm) and the powder of the bone cement (CMW Endurance; DePuy, London, England) were mixed using the open-bowl technique in ambient laboratory conditions, as described by Lewis et al. 13. The liquid of the cement was added to the blended powder mixture. After that, the cement dough was injected into different stainless steel molds to produce specimens for different tests (Fig. 1). Then, the cured cement specimens were taken out of the molds and polished with a 600-grit sandpaper. The specimens were examined by visual inspection, digital calipering, high-light exposure, and X-radiography. Specimens were discarded if any of the following problems existed: surface flaw(s) > 0.25 mm on the body of the specimen and/or in transition sections; pore size > 1 mm in diameter within the specimen and/or in transition sections; and any dimension being more than 5% within its nominal value 13.

Figure 1

Dimensioned drawings of the specimens used in the mechanical properties tests (All dimensions are in mm).

Study groups

The accepted specimens were divided randomly into eight groups based on the alendronate loading 6 and the aging condition of the cured cement (in ambient laboratory atmosphere or immersed in the phosphate buffered saline solution (PBS), pH 7.4 (Gibco Invitrogen Corp, Grand Island, NY) for 30 days) 16. Thus, the groups were: the atmospheric control group (GA0), the immersed control group (GI0), the atmospheric 0.1 wt./wt.%-alendronate group (GA0.1), the immersed 0.1 wt./wt.%-alendronate group (GI0.1), the atmospheric 0.5 wt./wt.%-alendronate group (GA0.5), the immersed 0.1 wt./wt.%-alendronate group (GI0.5), the atmospheric 1 wt./wt.%-alendronate group (GA1), and the immersed 1 wt./wt.%-alendronate group (GI1).

Mechanical tests

The compressive strength, compressive modulus, tensile strength, and flexural strength were determined from tests carried out in ambient laboratory conditions (air, at 22 ± 1 °C), in accordance with ASTM F451 17 (n = 6-7), ISO 527 18 (n = 6-8), and ISO 5833 19 (n = 6-9), respectively. The fatigue tests were conducted in PBS, at 37 ± 2 °C 13, in accordance with ASTM F2118 20, with applied load corresponding to stress of ± 15 MPa and frequency of 5 Hz (n = 7). The results of the fatigue tests (number of cycles to fracture (Nf)) were analyzed using the three-parameter Weibull relationship 21-23, leading to values of the minimum or guaranteed fatigue life (No), the characteristic fatigue life (Na), and the Weibull modulus (m) 21-23. The fatigue performance index was computed as the Weibull median (Nwm) thus 21

Morphology and porosity of fracture surfaces

Following a fatigue test, the fracture surface of the specimen was cleaned ultrasonically, dried, sputter-coated with a gold-palladium layer, and inspected at 25.0 kV with a scanning electron microscope (SEM; SIRION 100, FEI, United States) at a magnification of 93×. Five sections of a specimen were selected at random and, then, porosity was determined. This involved using a commercially-available software package (Image-pro plus 6.0, Media Cybernetics, Inc., Silver Spring, MD, USA) to select a region of interest (ROI) within a segment and, then, digitizing it. For each image, the number of pores, the mean pore size, and the number of pixels counted were determined. Porosity (P) was calculated thus 10

where Ap is the total area occupied by all the pores and Ai is the area of the whole image.

Statistical analysis

Quantitative results were expressed as mean ± standard deviation. Each set of quantitative results was first tested for normality using the variance homogeneity and normal distribution tests. When this was found to be the case, a parametric test of significance of difference in means (one-way analysis of variance) was used. Otherwise, a non-parametric test of significance was used. Intergroup differences were compared with the method of the least significant difference. Regression analyses were performed between P and each of the quasi-static mechanical properties as well as between P and Nf 24. Statistical significance was denoted if p < 0.05. Analyses were performed with a commercially-available software package (SPSS 16.0 for Windows; SPSS Inc., Chicago, IL, USA).

Results

Quasi-static mechanical properties

The influence of alendronate powder loading (ALN) on compressive strength was not significant (F = 4.177, p = 0.094), but immersion in PBS produced a significant reduction in this strength (F = 7.140, p = 0.0100), and, for each cement group, this strength exceeded the minimum level stipulated in ASTM F451 (70 MPa) (Fig. 2).The same trends were found for (a) compressive modulus in terms of influence of ALN (F = 4.306, p = 0.090) and of immersion in PBS (F = 15.833, p < 0.001) (Fig. 2); and (b) tensile strength in terms of influence of ALN (F = 6.732, p = 0.126) and of immersion in PBS (F = 23.125, p < 0.001) (Fig. 2). Neither ALN nor immersion in PBS exerted a significant influence on flexural strength (F = 0.005, p = 0.999 and F = 15.833, p < 0.001, respectively) (Fig. 2). Furthermore, for each cement group, flexural strength was higher than the minimum level stipulated in ISO 5833 (50 MPa) (Fig. 2).

Figure 2

Summary of the quasi-static mechanical properties.

Fatigue life

Increase in ALN significantly decreased Nf and NWM (F = 4.306, p = 0.0090) as did immersion in PBS (F = 11.642, p = 0.01) (Table 1).

Table 1

Summary of the fatigue test results (Nf) and the calculated Weibull parameters

Alendronate powder loading	Aging condition	Nf (cycles)	Weibull Parameters
			No (cycles)	Na (cycles)	m	NWM (cycles)
0 wt./wt.%	Atmosphere	17724±5815	10404	19232	0.74	22136
	PBS	13458±2713a	9414	15301	0.72	18595
0.1 wt./wt.%	Atmosphere	13701±2496a	8103	15550	0.46	15101
	PBS	10875±1923a	8518	14472	0.97	18504
0.5 wt./wt.%	Atmosphere	12973±2964a	8955	14309	1.09	19114
	PBS	10980±1767a	8690	11127	0.76	17562
1.0 wt./wt.%	Atmosphere	13398±2973a	8604	11300	0.87	16062
	PBS	11291±1865a	8349	11771	0.93	16236

a Compared with the atmospheric control group, p < 0.05.

Porosity measures and pore morphology

While ALN significantly increased pore number for any pore size (F = 43.463, p < 0.001), the influence of immersion in PBS on a given porosity parameter was not significant (F = 0.096, p = 0.759) (Fig. 3, Table 2, and Fig. 4). Furthermore, for each cement group, the pores were uniformly distributed from the margin area to the core area of a specimen (Fig. 5).

Figure 3

Summary of porosity measures.

Table 2

Microstructure parameters of fracture surface of specimens

Alendronate powder loading (wt./wt.%)	Aging condition	Number of pores	Total pore area (mm2)	Porosity (%)
0	Atmosphere	3.33±2.42	0.051±0.030	0.90±0.71
	PBS	2.86±1.46	0.062±0.030	0.90±0.47
0.1	Atmosphere	87.29±44.77a	0.017±0.010a	5.33±1.56a
	PBS	79.29±26.76a	0.016±0.004a	5.17±1.12a
0.5	Atmosphere	75.71±33.08a	0.019±0.007a	6.03±1.69a
	PBS	88.57±25.75a	0.014±0.004a	5.27±0.58a
1	Atmosphere	84.14±31.18a	0.018±0.007a	5.93±1.28a
	PBS	103.29±59.13a	0.017±0.011a	6.36±2.45a

aCompared with the atmospheric control group or immersed control group, p < 0.001.

Figure 4

The micrographs of the surfaces of cement specimens after fracture in the fatigue test

Figure 5

A sample of continuous scanning images of the surfaces of cement specimens after fracture in the fatigue test (GA1 cement group specimens).

SEM images showed that GA0 and GI0 specimens had few pores while specimens in each of the other six alendronate groups had a large number of small particles ranging in size from 10 µm to 200 µm (Fig. 4). The pores were distributed from the marginal area to the core area of the specimen (GA1 as an example in Fig. 5).

Correlations

While the correlation between P and each of the quasi-static mechanical properties was weak and not statistically significant (p > 0.05) (Fig. 6), porosity had a negative and significant correlation with fatigue life (R2 = 0.562, p < 0.01) (Fig. 7).

Figure 6

Correlations between porosity and each of the quasi-static mechanical properties.

Figure 7

Correlation between porosity and fatigue life (Nf).

Discussion

The mechanical strength of acrylic cement is highly related with the lifespan of the cemented arthroplasty 25. In previous studies, it was showed that APAC could locally release bisphosphonate, inhibit periprosthetic bone resorption, and enhance the mechanical strength of host bone while avoiding the detrimental effect of oral medication 8-11. Incorporation of liquid bisphosphonate or tablet-ground alendronate powder significantly decreased the mechanical strength of acrylic bone cement 12, 13. APAC may be a better option to maintain the mechanical strength of cement 14, 15. In the present study, we evaluated quasi-static mechanical properties, fatigue life, porosity, and microstructure of APAC.

Our microscopic results demonstrated increased porosity of APAC (relative to that of the control cement) and propose two possible mechanisms for this finding. First, previous studies found that, just like as in the case of an antibiotic-loaded acrylic bone cement, elution of a bisphosphonate powder from a bone cement is low 8. Since, in the case of APAC, the pores were distributed throughout the fracture surface, we assumed that the large number of pores could not result from powder aggregation. Second, given that the control group and the APAC group specimens were all cured under the same conditions, air inclusion could be eliminated as the reason.

Crack initiation, propagation, and accumulation is the cause of cement fracture. It results in mechanical failure of a cemented arthroplasty 26. Topoleski et al. 26 and Hoey et al. 24 proposed that, in an acrylic bone cement, pores play a dual role: on one hand, they cause stress concentration and promote crack initiation and propagation, which is detrimental for the cement mantle; but, on the other hand, they divert energy to the surroundings and prevent crack propagation, which is beneficial for the cement mantle. In the present study, a strong and significantly negative correlation was found between fatigue life and porosity, but the correlation between each of the quasi-static mechanical properties and porosity was weak and not statistically significant. These findings suggest that in the fatigue test, pores promoted cracks instead of inhibition, which agreed with the conclusions of previous studies 24, 27, but, in a quasi-static test, the effect of crack promotion is offset by that of crack inhibition.

In previous studies, it was shown that liquid bisphosphonate significantly reduced the mechanical strength of acrylic bone cement 12. It was speculated that liquid bisphosphonate dilutes the substrate and the catalyst of polymerization 12. The residual monomer acted as a stress riser, which led to mechanical failure. When tablet-ground alendronate powder was added to an acrylic bone cement, fatigue life was reduced by 39% 13. The hand-made preparation of the powder and the mixed excipient in the tablet may account for this unsatisfactory result. In the present study, we used a commercially-available alendronate powder, which does not contain any excipient mixed. Nonetheless, the fatigue life of APAC decreased by a much smaller margin compared to the case when tablet-ground alendronate powder was used 13.

We immersed all the specimens in PBS, at 37 °C, for 30 days to simulate the physiological environment 16. The mechanical strength after immersion changed due to two effects: the aging effect increased the strength in the early time and the plasticization effect of PBS uptake decreased the strength in the late time 16. In our study, all the specimens were tested after 30 days' standing to mitigate the aging effect as far as possible. Quasi-static mechanical properties decreased after PBS immersion, suggesting that the plasticization effect might play an important role.

Three study limitations are recognized. First, porosity was determined using SEM. A 3D technique, such as micro-computed tomography, may provide a more reliable and precise evaluation of number, size, and distribution of pores in bone cement than SEM 28. Second, the fatigue tests were conducted at only one stress level (± 15 MPa); as such, fatigue limits of the cements were not determined. Third, the specimens were prepared using hand-mixed, instead of vacuum mixed, cement 13. It has been suggested that vacuum mixing decreases porosity and increases fatigue life of an acrylic bone cement 29. However, Macaulay et al. 30 found that hand mixing by skilled technicians did not result in negative effects on cement porosity. Hence, hand mixing is thought to be reliable and reproducible 29, 30.

In conclusion, we found that adding a commercially-available alendronate powder to an acrylic bone cement (APAC) did not significantly affect any of the quasi-static mechanical properties of the cement determined but significantly decreased the fatigue life of the cement, which may be a consequence of the significant increase in porosity. Thus, clinical use of APAC cannot be recommended. Future work will investigate reinforcement of APAC with additives, such as bioceramic fibers 31, as means of increasing the fatigue life of the cement.