ORTHOPEDICS | Healio.com/Orthopedics

n Feature Article

abstractFull article available online at Healio.com/Orthopedics. Search:20140401-61

In this study, the difference in expression of human beta-defensin-3 in periprosthetic tissue and cancellous bone was observed in the periprosthetic tissue and cancellous bone of pa-tients in the periprosthetic joint infection group, the aseptic loosening group, and the spacer treatment group as well as the synovial membrane and ilium of the normal control group. Hematoxylin and eosin staining of the synovial tissue showed different levels of neutrophil infiltration in all groups except for the normal control group. Immunofluorescence staining of periprosthetic tissue and cancellous bone showed the most positive cells expressing hu-man beta-defensin-3 and the largest mean optical density in the periprosthetic joint infection group, followed by the aseptic loosening group, the spacer treatment group, and the normal control group, with a significant difference in comparison between the periprosthetic joint infection group and the other 3 groups (P<.01). White blood cell count, erythrocyte sedi-mentation rate, and C-reactive protein level were highest in the periprosthetic joint infection group, whereas no difference was found between the other 3 groups. Taken together, high levels of human beta-defensin-3 protein expression were found in the periprosthetic tissue and cancellous bone of patients with periprosthetic joint infection and aseptic loosening, but there are differential expressions of human beta-defensin-3, and this may provide a new marker for the differential diagnosis of infection and loosening of the artificial joint.

The authors are from Department 8 (G-DL), Research Institute of Surgery, Daping Hospital, Third Military Medical University, Chongqing; the Department of Rehabilitation (H-JY), Southwest Hospital, Third Military Medical University, Chongqing; the Department of Anesthesiology (SO), General Hospital of Chengdu Mili-tary Command, Chengdu, Sichuan; the Department of Orthopedics (XL, W-DN), First Affiliated Hospital of Chongqing Medical University, Chongqing; the Traumatic Center (X-KH, JF), Research Institute of Surgery, Daping Hospital, Third Military Medical University, Chongqing; the Department of Orthopedics (J-YC, YW), General Hospital of Chinese People’s Liberation Army, Beijing, China; and the Rothman Institute at Thomas Jefferson University (PJ, JF), Philadelphia, Pennsylvania.

The authors have no relevant financial relationships to disclose.This study was supported by grants from the Natural Science Foundation of China (30700177, 81071459),

Science and Technology Projects of Chongqing (CSTC, 2012gg-yyjs10024), China Postdoctoral Science Foun-dation (20090460108, 201003775), and Science and Technology Achievements Transformation Fund of Third Military Medical University and Military Twelfth Five Key Projects (BWS11J038).

Correspondence should be addressed to: Jun Fei, MD, Traumatic Center, Research Institute of Sur-gery, Daping Hospital, Third Military Medical University, Changjiang Rd, Chongqing 400042, China (feijundoctor@sohu.com).

Received: June 4, 2013; Accepted: November 22, 2013; Posted: April 15, 2014.doi: 10.3928/01477447-20140401-61

Human Beta-Defensin-3 for the Diagnosis of Periprosthetic Joint Infection and Loosening Guo-DonG Liu, MD; HonG-Jun Yu, MD; SHan ou, MD; Xi Luo, MD; Wei-DonG ni, MD; Xian-Kai HuanG, MD; Ji-YinG CHen, MD; Yan WanG, MD; Parvizi JavarD, MD, FrCS; Jun Fei, MD

e384

APRIL 2014 | Volume 37 • Number 4

n Feature Article

Artificial joint replacement has been widely used clinically. In 2006, artificial joint replace-

ment was performed in 500,000 cases in the United States and 130,000 cases in the United Kingdom. However, data from multiple centers have shown that the in-fection rate associated with artificial joint replacement was approximately 1.5% to 2.5% and as high as 9% for elbow joint re-placement, and even up to 20% for a revi-sion. The cost for each patient with infec-tion was at least US$30,000 to $50,000.1 There were 2 million cases of nosocomial infection, including periprosthetic joint infection, in the United States each year, and approximately 50% of these were re-lated to the implant.2 This complication not only had a cost of up to US$11 bil-lion but also led to serious consequences, such as bacterial resistance, amputation, and even death.3 Therefore, periprosthetic joint infection has become the most feared complication of joint replacement because of its difficult diagnosis and treatment. Chronic infection caused by bacteria with lower virulence, such as Propionbacteri-um acnes, is usually difficult to diagnose and differentiate from aseptic loosening.4 The indicators of this infection, such as neutrophils in the periprosthetic tissue, radionuclide imaging, C-reactive protein (CRP), and erythrocyte sedimentation rate (ESR), have been obtained clinically during surgery. The American Academy of Orthopaedic Surgeons has put forward appropriate diagnostic guidelines and evi-dence.5

As a biologic component of innate im-munity, antimicrobial peptides have been found and understood since the 1980s. Antimicrobial peptides are found in a variety of organisms and are considered the future of anti-infective therapy drugs. Antimicrobial peptides have active small-molecule peptides with broad-spectrum antimicrobial activity and are a first line of defense for various living organisms,6,7 with the advantages of high efficiency, broad spectrum, and little drug resistance.

On the basis of previous research showing that human beta-defensin-3 had definite inhibition on the formation of Staphylo-coccus aureus biofilm,8 the authors stud-ied the differential expression of human beta-defensin-3 in periprosthetic soft tis-sue and cancellous bone as well as normal synovial tissue and cancellous bone of pa-tients in the periprosthetic joint infection group, the spacer treatment group, and the aseptic loosening group. The study was conducted to examine the relationship be-tween human beta-defensin-3 expression and infection and loosening of the artifi-cial joint and explore its possible use as a new marker and means for the differential diagnosis of periprosthetic joint infection and aseptic loosening to further improve the success rate of artificial joint revision.

Materials and MethodsClinical Materials

Patient specimens from cases with in-fection and loosening of artificial joints in the authors’ hospital treated from No-vember 2009 to September 2012 were collected. Periprosthetic tissue and can-cellous bone of patients with peripros-thetic joint infection (periprosthetic joint infection group, n=26) and aseptic loos-ening (aseptic loosening group, n=18) were removed during surgery. Two-stage revision and placement of vancomycin-loaded spacers were performed on pa-tients diagnosed with periprosthetic joint infection. Tissue and cancellous bone around the spacers were collected from patients who had undergone reimplanta-tion of an artificial joint after effective control of infection. Knee joint synovium from 13 patients with meniscus injury and hip joint synovium from 2 patients with femoral neck fracture were included as a normal control group of periprosthetic tis-sue. Normal ilium from patients with limb fracture who had undergone ilium implan-tation was considered as a normal control group of cancellous bone. Preoperative white blood cell count, ESR, and CRP for all patients were obtained. The study pro-

tocol, including access to and the use of the medical records of the study subjects, was reviewed and approved by the insti-tutional ethics committee of Daping Hos-pital, Third Military Medical University.

Clinical symptoms in the peripros-thetic joint infection group included vari-ous degrees of persistent pain more than 2 months after the initial replacement. Inclu-sion criteria were the methods of diagnos-ing periprosthetic joint infection released by the Musculoskeletal Infection Society in 2012: (a) the sinus tract was connected with the prosthesis; (b) the same patho-gens were found in the periprosthetic tissue for 2 or multiple samplings or in bacterial culture of the synovial fluid; and (c) 4 or more of the following 6 criteria were met: (1) increasing CRP and ESR in the serum; (2) increasing white blood cell count in the synovial fluid; (3) an in-creasing percentage of neutrophils in the synovial fluid; (4) pus in the joints; (5) bacteria found on the 1st bacterial culture of periprosthetic tissue or synovial fluid; and (6) more than 5 neutrophils found in the periprosthetic tissue and at least 5 in-dependent views under a high-power lens (×400).9 Besides these criteria, infection and loosening needed to be confirmed in all patients through surgery and bacterial culture. This study included 26 patients, including 6 cases of Staphylococcus au-reus, 10 of coagulase-negative staphylo-cocci, 4 of Staphylococcus epidermidis, 3 of Pseudomonas aeruginosa, and 3 of uncultured bacteria (Table 1).

Preparation and Observation of Light Microscopy Samples

After sterile harvesting of the test spec-imens, standardized histologic processing was performed. Cancellous bone samples were placed in 0.5 M ethylenediaminetet-raacetic acid pH 8.0 buffer solution that was changed weekly. Decalcification was accomplished when the needle could be inserted without resistance. After decal-cification and dehydration, the specimen blocks were cut into sections that were 5

e385

ORTHOPEDICS | Healio.com/Orthopedics

n Feature Article

μm thick with an ultra-high-performance microtome for hematoxylin and eosin staining. Histologic morphology and neutrophil infiltration of tissue sections in each group were observed under light microscopy (×200).

Immunofluorescence StainingThe sections were dewaxed and rehy-

drated routinely, and the citrate antigens were cooled to room temperature after thermal repair. Endogenous peroxidase was blocked with 3% hydrogen peroxide and 1% bovine serum albumin. Primary antibody-anti-human beta-defensin-3 (ie, mouse anti-human polyclonal antibody [Novus Biologicals Co, Littleton, Colo-rado]) was incubated at 4°C overnight. Fluorescent secondary antibody-rabbit anti-mouse fluorescent antibody (Invitro-gen Zymed Laboratories, San Francisco, California) and 4’,6-diamidino-2-phenyl-indole marked with tetramethylrhodamine isothiocyanate dye were incubated and stained with 4’,6-diamidino-2-phenylin-dole after rewarming. Phosphate-buffered saline was used as the primary antibody for negative control. Staining of positive cells was observed with an inverted-phase contrast fluorescence microscope, and 5 in-dependent views of expressions of positive human beta-defensin cells were selected at random. The mean optical density of each positive cell was measured with Image-Pro Plus 7.0c image analysis software (Media Cybernetics, Bethesda, Maryland).

Statistical Analysis Statistical analysis was performed with

SPSS version 17.0 software (SPSS Inc, Chicago, Illinois). The rank-sum test of 2 independent samples was used for com-parison between groups. P<.05 was con-sidered statistically significant.

resultsHematoxylin and Eosin Staining of Periprosthetic Tissue and Synovium

Hematoxylin and eosin staining of periprosthetic tissue and synovium when examined with the naked eye showed that the gross tissue of synovium in the nor-mal control group was soft and bright, but the periprosthetic tissue had a rough texture in other groups. Microscopic ob-servation (×200) with hematoxylin and eosin staining showed dense connective tissue, with no neutrophil or other infil-tration in the normal control group, but various degrees of connective tissue, with infiltration of a variable number of neutrophils and lymphocytes in the peri-prosthetic joint infection group, aseptic loosening group, and spacer treatment group (Figure 1).

Hematoxylin and Eosin Staining of Cancellous Bone

Hematoxylin and eosin staining showed no significant difference on gross specimens of cancellous bone in each group when examined with the naked eye. Microscopic observation (×200) of he-

matoxylin and eosin staining showed no significant difference in neutrophil infil-tration in each group (Figure 2).

Immunofluorescence Staining of Periprosthetic Tissue and Normal Synovium

After immunofluorescence staining of periprosthetic tissue and normal synovi-um, an inverted-phase contrast fluores-cence microscope (Olympus IX71, Ja-pan) showed that the red particles were positive cells (green arrows) of human beta-defensin-3 proteins stained by tetra-methylrhodamine isothiocyanate-marked fluorescent secondary antibody and that the blue particles were nuclei stained by 4’,6-diamidino-2-phenylindole (yel-low arrows) (Figure 3). Compared with the aseptic loosening group, the spacer treatment group, and the normal control group, the periprosthetic joint infection group had more positive cells and ob-viously higher fluorescence intensity. More positive cells and obviously higher fluorescence intensity were shown in the aseptic loosening group than in the spacer treatment group and the normal control group. However, no significant difference was found between the spacer treatment group and the normal control group.

Immunofluorescence Staining of Cancellous Bone

The fluorescence microscope (Olym-pus IX71) showed that the red particles were human beta-defensin-3 proteins in the cytoplasm of the bone cells (green ar-rows) and the blue particles were nuclei of the bone cells (yellow arrows). Com-pared with the aseptic loosening group, the spacer treatment group, and the nor-mal control group, the periprosthetic joint infection group had various degrees of higher human beta-defensin-3 protein fluorescence intensity and positive ex-pression in cells in the endosteum (white arrows). More positive expression and ob-viously higher fluorescence intensity were

n Feature Article

Table 1

Grouping of Periprosthetic Tissue and Cancellous Bone

Total No. (Male/Female)

Group Periprosthetic Tissue Cancellous Bone

Periprosthetic joint infection 26 (16/10) 26 (16/10)

Aseptic loosening 18 (10/8) 16 (10/8)

Spacer treatment 24 (12/12) 18 (11/7)

Normal control 15 (8/7) 9 (5/4)

Total 83 69

e386

APRIL 2014 | Volume 37 • Number 4

n Feature Article

observed in the aseptic loosening group than in the other 2 groups. There was a significant difference between the spacer treatment group and the normal control group (Figure 4).

Mean Optical Density of Human Beta-Defensin-3 in Periprosthetic Tissue, Synovium, and Cancellous Bone

Mean optical density of human beta-defensin-3 in the periprosthetic joint in-

fection group was the highest, followed by the aseptic loosening group, the spacer treatment group, and the normal control group. There was a significant difference between the periprosthetic joint infection

Figure 1: Hematoxylin and eosin staining (×200) of periprosthetic tissue and synovium in the periprosthetic joint infection group (A), aseptic loosening group (B), and spacer treatment group (C), all with various degrees of connective tissue and infiltration of a variable number of neutrophils and lymphocytes; and in the normal control group, with dense connective tissue but no neutrophil infiltration or other infiltration (D).

Figure 2: Hematoxylin and eosin staining (×200) of cancellous bone in the periprosthetic joint infection group (A), aseptic loosening group (B), spacer treatment group (C), and normal control group (D). There is no significant difference in neutrophil infiltration between groups.

Figure 3: Immunofluorescence staining (×200) of periprosthetic tissue and normal synovium in the periprosthetic joint infection group (A), aseptic loosening group (B), spacer treatment group (C), and normal control group (D). The green arrows show that the red particles are positive cells of human beta-defensin-3 protein stained by tetramethylrhodamine isothiocyanate-marked fluorescent secondary antibody. The yellow arrows show that the blue particles are nuclei stained by 4’,6-diamidino-2-phenylindole.

Figure 4: Immunofluorescent staining (×200) of cancellous bone in the periprosthetic joint infection group (A), aseptic loosening group (B), spacer treatment group (C), and normal control group (D). The green arrows show that the red particles are human beta-defensin-3 protein in the cytoplasm of the bone cells. The yellow arrows show that the blue particles are nuclei of the bone cells. The white arrows show human beta-defensin-3 protein fluorescence intensity and positive expression cells in the endosteum.

A B C D

A B C D

A B C D

A B C D

e387

ORTHOPEDICS | Healio.com/Orthopedics

n Feature Article

group and the other 3 groups (P<.01), between the aseptic loosening group and the spacer treatment group and the nor-mal control group (P<.01), and between the spacer treatment group and the normal control group (P<.01) (Table 2). The 95% confidence intervals for the mean opti-cal density of periprosthetic tissue in the periprosthetic joint infection group, the aseptic loosening group, the spacer treat-ment group, and the normal control group were 0.4253 to 0.4352, 0.3059 to 0.3105, 0.2305 to 0.2381, and 0.07880 to 0.07992, respectively. The 95% confidence intervals for the mean optical density of cancellous bone in the periprosthetic joint infection group, the aseptic loosening group, the

spacer treatment group, and the normal control group were 0.4661 to 0.4791, 0.3423 to 0.3480, 0.2394 to 0.2440, and 0.1357 to 0.1401, respectively.

White Blood Cell Count, Erythrocyte Sedimentation Rate, and C-reactive Protein Level

Comparison between any 2 groups showed no significant difference in white blood cell count, but ESR and CRP showed a more obvious difference be-tween the periprosthetic joint infection group and the other 3 groups. There was no significant difference between the aseptic loosening group and the spacer treatment group and the normal control group or be-

tween the spacer treatment group and the normal control group (Table 3).

discussion It is not difficult to diagnose acute

periprosthetic infection because of its obvious clinical symptoms. Chronic peri-prosthetic infection is mainly caused by pathogens of lower virulence, such as coagulase-negative staphylococci and P acnes, and therefore it is difficult to dif-ferentiate chronic periprosthetic infection from aseptic loosening. Because there have been no uniform standards for the differential diagnosis of periprosthetic infection, patients diagnosed with aseptic loosening were infected with bacterium of lower virulence and underwent revisions directly because of the positive postopera-tive culture results. Therefore, the search for new markers to help clinicians to dif-ferentiate artificial joint infection from aseptic loosening preoperatively and choose the correct treatment has become a focus in periprosthetic joint infection.

White blood cell count is a commonly used indicator of infection. It increases to varying degrees in patients with local infection, systemic infection, and other noninfectious diseases. However, white blood cell count is not helpful in the di-agnosis of periprosthetic joint infection. Di Cesare et al10 found no statistical sig-nificance between the infection group and the noninfection group with regard to white blood cell count preoperatively. The results of the current study also showed no difference in white blood cell count among groups. However, infection is still possible when the white blood cell count increases obviously in patients with acute periprosthetic joint infection.

The most common indicators in the differential diagnosis of periprosthetic joint infection and aseptic loosening are ESR and CRP. Della Valle et al11 reported an accuracy rate of 77% for ESR and 84% for CRP in the diagnosis of periprosthetic joint infection. ESR and CRP still have a high diagnostic efficacy in the diagno-

n Feature Article

Table 2

Mean Optical Density of Human Beta-Defensin-3 in Periprosthetic Tissue and Cancellous Bone

Mean±SD

Group Periprosthetic Tissue Cancellous Bone

Periprosthetic joint infection 0.43±0.013 0.473±0.017

Aseptic loosening 0.308±0.005a 0.345±0.006a

Spacer treatment 0.234±0.009a,b 0.242±0.005a,b

Normal control 0.089±0.019a,b,c 0.138±0.003a,b,c

aP<.01 compared with periprosthetic joint infection group. bP<.01 compared with aseptic loosening group. cP<.01 compared with spacer treatment group.

Table 3

Comparison of White Blood Cell Count, Erythrocyte Sedimentation Rate, and C-reactive Protein

GroupWhite Blood Cell

Count, ×109a

Erythrocyte Sedimentation Rate,

mm/hbC-reactive Protein,

mg/dLb

Periprosthetic joint infection

6.19±1.98 35.15±23.47 2.39±2.51

Aseptic loosening 6.15±0.88 8±2.5c 0.4±0.09c

Spacer treatment 5.32±1.04 9.83±4.97c 0.45±0.23c

Normal control 5.84±1.04 11.07±4.5c 0.41±0.09c

aP>.05 between all groups. bP>.05 between aseptic loosening, spacer treatment, and normal control groups. cP<.01 compared with periprosthetic joint infection group.

e388

APRIL 2014 | Volume 37 • Number 4

n Feature Article

sis of periprosthetic joint infection after interference factors are eliminated.10,11 Generally, the difference in ESR and CRP between the periprosthetic joint infection group and the aseptic loosening group were statistically significant in this study, but the possibility of periprosthetic joint infection cannot not be precluded, even if ESR and CRP are normal, especially in infection caused by bacterium of lower virulence, such as P acnes, when both in-dicators might be normal.

Pathogenic diagnosis is one of the most powerful means for diagnosing infectious diseases. Tissue with suspected peripros-thetic infection is often collected during surgery for bacterial culture in the treat-ment of patients with periprosthetic joint infection. Bacterial culture is always af-fected by sampling sites, the number of samples, inspection methods, and time limit as well as culture methods and time. Therefore, the bacterium is not necessarily cultured in all of the infected tissue. Most periprosthetic joint infections are chronic infections caused by bacterium of lower virulence and can form biofilms. The bac-terium inside the biofilms is dormant and shows atypia,12,13 which can lead to false-negative results on bacterial culture.14,15 In this study, even if obvious necrosis and pus formation could be seen with the naked eye during surgery, pathogens were cul-tured from only 20 of 26 infected patients. Negative results on bacterial culture do not necessarily preclude periprosthetic joint in-fection. However, if the same pathogen is cultured in samples from many sites, this could not only improve the diagnosis rate but also play an important role in guiding the clinical use of antibiotics.

Human defensin is a family of an-timicrobial peptides.16 Compared with human beta-defensin-1 and human beta-defensin-2, human beta-defensin-3 has stronger antibacterial activity against En-terococcus faecalis, such as gram-positive bacteria, gram-negative bacteria, and van-comycin-resistant enterococcus.17 Studies have shown that the antibacterial activity

of human beta defensins (except for hu-man beta-defensin-3) is strongly inhibited under physiologic salt concentration or in the presence of divalent cations.18,19 Hu-man beta-defensin-3 can express at vari-ous degrees in the skin, respiratory tract, oral mucosa, trachea, and other tissue18 and is related to conditions such as infec-tions and cancer.20,21 Warnke et al22 found human beta-defensin-3 expression on the endosteum beside the bone trabecula and in the cytoplasm of bone cells rather than in the cytoplasm of bone cells of jawbone under normal conditions. However, there is no finding of similar human beta-defen-sin-3 expression in periprosthetic tissue and cancellous bone in the periprosthetic joint infection group and the aseptic loos-ening group.

The mean optical density of human beta-defensin-3 is its expression in peri-prosthetic tissue of patients with peri-prosthetic joint infection. In this study, the mean optical density of human beta-defensin-3 in the periprosthetic joint in-fection group was significantly higher than that in the other groups. The mean optical density of human beta-defensin-3 in the aseptic loosening group was higher than in the normal control group, with statistical difference between groups. This difference might be caused by the local increase and degree of human beta- defensin-3 expression induced by in-fection or inflammation. Human beta-defensin-3 expression was not found in the normal synovium, but was found in the synovium of patients with osteoar-thritis. The normal synovium in the test was obtained from patients with menis-cus injury and femoral neck fracture, but the possibility of joint degeneration (ie, osteoarthritis) could not be precluded. In addition, the fluorescence antibody used in the test was tetramethylrhodamine isothiocyanate of red fluorescence that, to some extent, was of autofluorescence. Therefore, the results might not be the same with human beta-defensin-3 expres-sion on normal synovium, and this might

be one of the limitations of this study. Paulsen et al17 performed experimental studies on 4 groups (10 cases of synovial tissue samples for each group), including patients with pyogenic arthritis, osteoar-thritis, and rheumatoid arthritis, as well as a normal control group, using human beta-defensin-1, human beta-defensin-2, human beta-defensin-3, human alpha-defensin-5, and human alpha-defensin-6, respectively, with reverse transcription-polymerase chain reaction. The results showed positive rates of human beta- defensin-3 in the 4 groups of 100%, 100%, 0%, and 0%, respectively. These findings showed that human beta-defensin-3 ex-pression could increase in inflammatory disease of the joints, such as pyogenic ar-thritis and osteoarthritis. Until now, there has been no research on the expression of human beta-defensin-3 in periprosthetic tissue and cancellous bone in relation to diagnosis and treatment outcomes in pa-tients with periprosthetic joint infection.

Aseptic loosening can also result in inflammation, mainly because of chronic aseptic inflammation caused by stimula-tion of worn polyethylene particles and other foreign matter.23,24 Although infec-tion and aseptic loosening showed differ-ent degrees of inflammation in peripros-thetic tissue, more neutrophil infiltration was found on frozen pathologic sections during surgery,25 indicating that the de-gree of inflammatory reaction in infected patients was higher than that in patients with aseptic loosening. The mean optical density of human beta-defensin-3 in the infection group showed statistical differ-ence compared with the other 3 groups, especially the aseptic loosening group, which coincided with the results of clini-cal inflammation and pathologic examina-tion. The infection had been controlled to a certain extent or effectively after com-prehensive treatment with debridement, antibiotics, and implantation of spacers for patients in the spacer treatment group.

Thus, the difference between the peri-prosthetic joint infection group and the

e389

ORTHOPEDICS | Healio.com/Orthopedics

n Feature Article

spacer treatment group was statistically significant, suggesting successful infec-tion control. These analyses proved that human beta-defensin-3 expression is as-sociated with the degree of inflammatory reaction of tissue, indicating that human beta-defensin-3 expression in local infect-ed areas might be an indicator that could better reflect the degree of local inflamma-tion. In recent years, studies of the diag-nosis of periprosthetic joint infection have focused mainly on local infection.26,27 The tissue in this study, including peripros-thetic tissue and cancellous bone belong-ing to infected local tissue, can reflect the strength of local inflammation. Currently, white blood cell count, ESR, and CRP are commonly used results clinically obtained from peripheral venous blood and reflect systemic reaction of the body against in-fections and other diseases. These mea-sures may be easily affected by many factors, such as infection and connective tissue in other parts of the body. To make a diagnosis of periprosthetic joint infec-tion, a number of complex systemic and local diseases and other factors that cause increases in white blood cell count, ESR, and CRP must be ruled out. Therefore, hu-man beta-defensin-3 expression, detected partly in this study, can more accurately reflect the degree of inflammation or periprosthetic joint infection, suggesting that higher expression of human beta- defensin-3 also occurs in periprosthetic tissue and cancellous bone as a result of infection or inflammation and correlates with the severity of infection.

conclusionExpression of human beta-defensin-3

can reflect the severity of infection in peri-prosthetic tissue and cancellous bone of patients with periprosthetic joint infection. Human beta-defensin-3 expression may be a good reference marker for the differential diagnosis of periprosthetic joint infection and aseptic loosening and hence may play an important role in improving the success rate of artificial joint revision.

references 1. Esposito S, Leone S. Prosthetic joint infec-

tions: microbiology, diagnosis, management and prevention. Int J Antimicrob Agents. 2008; 32(4):287-293.

2. Harris LG, Richards RG. Staphylococci and implant surfaces: a review. Injury. 2006; 37 (suppl 2):S3-S14.

3. Gollwitzer H, Ibrahim K, Meyer H, Mittel-meier W, Busch R, Stemberger A. Antibacte-rial poly(D,L-lactic acid) coating of medical implants using a biodegradable drug delivery technology. J Antimicrob Chemother. 2003; 51(3):585-591.

4. Zimmerli W, Ochsner PE. Management of infection associated with prosthetic joints. Infection. 2003; 31(2):99-108.

5. Della Valle C, Parvizi J, Bauer TW, et al. American Academy of Orthopaedic Sur-geons clinical practice guideline on: the diag-nosis of periprosthetic joint infections of the hip and knee. J Bone Joint Surg Am. 2011; 93(14):1355-1357.

6. Beisswenger C, Bals R. Functions of antimi-crobial peptides in host defense and immunity. Curr Protein Pept Sci. 2005; 6(3):255-264.

7. Dubin A, Mak P, Dubin G, et al. New genera-tion of peptide antibiotics. Acta Biochim Pol. 2005; 52(3):633-638.

8. Huang Q, Yu HJ, Liu GD, et al. Comparison of the effects of human beta-defensin 3, van-comycin, and clindamycin on Staphylococ-cus aureus biofilm formation. Orthopedics. 2012; 35(1):e53-e60.

9. Parvizi J, Zmistowski B, Berbari EF, et al. New definition for periprosthetic joint infec-tion: from the Workgroup of the Musculo-skeletal Infection Society. Clin Orthop Relat Res. 2011; 469(11):2992-2994.

10. Di Cesare PE, Chang E, Preston CF, Liu CJ. Serum interleukin-6 as a marker of peripros-thetic infection following total hip and knee arthroplasty. J Bone Joint Surg Am. 2005; 87(9):1921-1927.

11. Della Valle CJ, Sporer SM, Jacobs JJ, Berger RA, Rosenberg AG, Paprosky WG. Preop-erative testing for sepsis before revision total knee arthroplasty. J Arthroplasty. 2007; 22(6 suppl 2):90-93.

12. Zimmerli W, Trampuz A, Ochsner PE. Pros-thetic-joint infections. N Engl J Med. 2004; 351(16):1645-1654.

13. Costerton JW, Montanaro L, Arciola CR. Biofilm in implant infections: its produc-tion and regulation. Int J Artif Organs. 2005; 28(11):1062-1068.

14. Costerton JW, Stewart PS, Greenberg EP. Bac-terial biofilms: a common cause of persistent infections. Science. 1999; 284(5418):1318-1322.

15. Tunney MM, Patrick S, Curran MD, et al. Detection of prosthetic hip infection at re-

vision arthroplasty by immunofluorescence microscopy and PCR amplification of the bacterial 16S rRNA gene. J Clin Microbiol. 1999; 37(10):3281-3290.

16. Selsted ME, Tang YQ, Morris WL, et al. Puri-fication, primary structures, and antibacterial activities of beta-defensins, a new family of antimicrobial peptides from bovine neutro-phils. J Biol Chem. 1993; 268(9):6641-6648.

17. Paulsen F, Pufe T, Conradi L, et al. Antimi-crobial peptides are expressed and produced in healthy and inflamed human synovial membranes. J Pathol. 2002; 198(3):369-377.

18. Harder J, Bartels J, Christophers E, Schro-der JM. Isolation and characterization of hu-man beta-defensin-3, a novel human induc-ible peptide antibiotic. J Biol Chem. 2001; 276(8):5707-5713.

19. Garcia JR, Jaumann F, Schulz S, et al. Iden-tification of a novel, multifunctional beta-de-fensin (human beta-defensin 3) with specific antimicrobial activity: its interaction with plasma membranes of Xenopus oocytes and the induction of macrophage chemoattrac-tion. Cell Tissue Res. 2001; 306(2):257-264.

20. Dressel S, Harder J, Cordes J, et al. Differ-ential expression of antimicrobial peptides in margins of chronic wounds. Exp Dermatol. 2010; 19(7):628-632.

21. Shuyi Y, Feng W, Jing T, et al. Human beta-defensin-3 (hBD-3) upregulated by LPS via epidermal growth factor receptor (EGFR) signaling pathways to enhance lymphatic in-vasion of oral squamous cell carcinoma. Oral Surg Oral Med Oral Pathol Oral Radiol En-dod. 2011; 112(5):616-625.

22. Warnke PH, Springer IN, Russo PA, et al. In-nate immunity in human bone. Bone. 2006; 38(3):400-408.

23. Harris WH. Wear and periprosthetic oste-olysis: the problem. Clin Orthop Relat Res. 2001; (393):66-70.

24. Matthews JB, Besong AA, Green TR, et al. Evaluation of the response of primary human peripheral blood mononuclear phagocytes to challenge with in vitro generated clini-cally relevant UHMWPE particles of known size and dose. J Biomed Mater Res. 2000; 52(2):296-307.

25. Trampuz A, Hanssen AD, Osmon DR, Man-drekar J, Steckelberg JM, Patel R. Synovial fluid leukocyte count and differential for the diagnosis of prosthetic knee infection. Am J Med. 2004; 117(8):556-562.

26. Sarda-Mantel L, Saleh-Mghir A, Welling MM, et al. Evaluation of 99mTc-UBI 29-41 scintigraphy for specific detection of ex-perimental Staphylococcus aureus prosthetic joint infections. Eur J Nucl Med Mol Imag-ing. 2007; 34(8):1302-1309.

27. Deirmengian C, Hallab N, Tarabishy A, et al. Synovial fluid biomarkers for peripros-thetic infection. Clin Orthop Relat Res. 2010; 468(8):2017-2023.

n Feature Article

e390