Identifying Common Pathogens in Periprosthetic Joint Infection and Testing Drug‐resistance Rate for Different Antibiotics: A Prospective, Single Center Study in Beijing

Abstract

Objective

The aim of this study is to identify the common microorganisms causing PJI as well as the drug‐resistant spectrum for each microorganism, to help orthopaedic surgeons to choose appropriate antibiotics.

Method

One hundred and sixty patients who suffered from failure of primary or revision total hip or knee arthroplasty for different reasons were prospectively recruited. These patients underwent revision or re‐revision total hip or knee arthroplasty in our institution between August 2013 to August 2016. The details of patients’ medical history and comprehensive physical examination, as well as demographic data were recorded precisely. Routine blood test results, erythrocyte sedimentation rate (ESR), C‐reactive protein (CRP), high sensitive C‐reactive protein (hs‐CRP), interleukin‐6 (IL‐6) levels, and synovial leukocyte counts were collected. Additionally, aspiration was conducted during surgery to avoid pollution unless when PJI was strongly suspected, in which case, joint puncture and aspiration were conducted before surgery. Intraoperatively, the implant‐surrounding tissue and the prosthesis were collected under aseptic conditions. Postoperatively, the prosthesis, implant‐surrounding tissue and synovium were sent to the laboratory immediately. The sonicate extraction (the prosthesis was sent for ultrasound sonication first), implant surrounding tissue and synovium were sent for microbiologic culture, and the implant‐surrounding tissue was also sent for pathological examination. The isolated bacteria strains and drug‐resistance rates for each pathogen for different antibiotics were presented.

Results

There were 59 PJI cases in the infectious group and 101 cases in the non‐infectious group (PJI is diagnosed according to the diagnosing criteria from the Workgroup of the Musculoskeletal Infection Society). Of 69 strains of pathogens isolated, Gram‐positive bacterium is the most common pathogenic bacteria causing PJI (60, 86.96%). Staphylococcus epidermidis and Staphylococcus aureus played an important role as well, followed by Gram‐negative bacteria (8, 11.59%) and fungus (1, 1.45%). Penicillin (78.57%), erythromycin (66.67%) and clindamycin (44.74%) showed high antibiotic resistance rate. In addition, the second‐generation cephalosporin, usually as the prophylactic antibiotic, resistance rate was high (20%) as well. Fortunately, no vancomycin‐resistant bacteria were discovered in the current study.

Conclusion

This study provides some information on the most common pathogens in our institution and the selection of antibiotics in the perioperative period in northern China. Cefuroxime and clindamycin might not be appropriate for use as prophylactic antibiotics in revision total knee or hip arthroplasty. Vancomycin is ideal for empiric antibiotic use in suspected PJI cases because of the low drug‐resistance rate.

Introduction

Total joint arthroplasty (TJA) surgery has been recognized in recent decades as successful and cost‐effective1; the surgery could help patients with arthrosis of different etiology by relieving pain, restoring joint function, as well as improving quality of life. The number of total joint arthroplasties in the United States is appropriately 800 000 annually, and the number of total hip arthroplasties (THA) and total knee arthroplasties (TKA) will reach 572 000 and 3 480 000, respectively, by 20302. As the total amount increases steadily, the number of revision THA and TKA will increase as well. Although the survivorship of THA or TKA at 10‐year follow‐up could be higher than 95%2, aseptic loosening and infection are still major reasons for revision. Periprosthetic joint infection (PJI) is a severe and sophisticated complication for surgeons and patients following the surgery, and PJI was reported to account for 25.2% of revision THA and 14.8% of TKA3, 4. It was the most common etiology for revision TKA according to a report in the United States published in 20175. The prevalence of PJI among primary or revision total joint arthroplasty was approximately 0.7%6. With such a huge number of patients undergoing THA and TKA, many patients will continue to suffer from PJI, and will require subsequent one‐stage or two‐stage revision, which means prolonged treatment duration, much higher cost, and compromised treatment outcomes compared to those without PJI. Indeed, the rate of infection might be higher according to a cohort study of patients who underwent primary THA or TKA from 2006 to 20097.

It is important for surgeons to understand the reasons for failure following THA or TKA, as the treatment for patients with PJI is quite different from those without PJI. The American Academy of Orthopaedic Surgeons (AAOS Clinical Practice Guidelines on Diagnosis of Periprosthetic Joint Infection), the Workgroup of the Musculoskeletal Infection Society (MSIS) and participants of the International Consensus Meeting on Periprosthetic Joint Infection (Proceedings of the International Consensus Meeting on Periprosthetic Joint Infection) have put forth their own documents on PJI in 2010, 2011, and 2013, respectively, to help orthopaedic surgeons manage patients with suspected infection following THA or TKA. Although there are different tools for the diagnosis of PJI, and new biomarkers such as serum interleukin‐6 (IL‐6) have improved specificity8, diagnosis of PJI could, notwithstanding, remain difficult for orthopaedic surgeons sometimes, especially when physicians are trying to find the pathogen. In addition, under some circumstances, low grade infection could be underdiagnosed, as some cases are recognized as aseptic loosening, which could cause the diagnosis of PJI to become more difficult9, 10, 11. As a matter of fact, the incidence of PJI could be even higher with more sensitive detection and diagnostic methods.

Although there are methods applied, such as the use of prophylactic antibiotics, to lower the incidence of PJI, joint infection is still one of the most disastrous reasons for TJA failure. It often leads to significant morbidity and accounts for a substantial proportion of healthcare expenditures12. PJI increases the burden of healthcare observably. Steven et al. report that the cost of infections following THA or TKA in the USA was approximately $320 million in 2001; however, in 2009 the cost was approximately $566 million, and by 2020 the cost might increase to $1.62 billion13.

The epidemiology of PJI‐associated microbiological and related drug‐resistance conditions varies among countries14. Staphylococcus aureus and Staphylococcus epidermidis are the most common pathogens in the USA6, but in Europe15, coagulase‐negative Staphylococcus spp. is the most common pathogen, followed by Staphylococcus aureus. The common pathogens in China might be different as well; thus, it is necessary to identify the common pathogens in PJI. For patients with PJI, extended use of antibiotics is often needed. In addition, drug resistance could not be overlooked. The options of antibiotics to combat pathogens have been restricted due to the slow development of new antibiotics16 and the abuse of antibiotics has changed the drug sensitivity of pathogens as well.

Results of microbiologic culture of implant‐surrounding tissue, sonicate extractions of joint prosthesis and synovium, as well as the drug‐resistance tests are essential information in diagnosis and subsequently choosing appropriate antibiotics and also treatment approach, whether one‐stage or two‐stage revision. Although treatment of PJI is challenging, it could be successfully managed with accurate microbiologic diagnosis, and when appropriate decisions are made in regards to the medical and surgical strategy. Accurate microbiologic diagnosis of the PJI is essential for the success of the whole treatment, and could permit effective use of antibiotics of the narrow spectrum17, whether patients underwent revision or not.

The present study examined the common microorganisms and the drug‐resistance status of bacteria strains, to provide additional data on PJI in China and to help surgeons choose appropriate antibiotics.

Materials and Methods

Patients

This was a descriptive study of patients who received revision or re‐revision total hip or knee arthroplasty for different reasons in our institution from August 2013 to August 2016. PJI is diagnosed according to the diagnosing criteria from the Workgroup of the Musculoskeletal Infection Society18: (i) there is a sinus tract communicating with the prosthesis; or (ii) a pathogen is isolated by culture from at least two separate tissue or fluid samples obtained from the affected prosthetic joint; or (iii) four of the following six criteria exist: (a) elevated serum erythrocyte sedimentation rate (ESR) and serum C‐reactive protein (CRP) concentration, (b) elevated synovial leukocyte count, (c) elevated synovial neutrophil percentage (PMN%), (d) presence of purulence in the affected joint, (e) isolation of a microorganism in one culture of periprosthetic tissue or fluid, or (f) greater than five neutrophils per high‐power field in five high‐power fields observed from histologic analysis of periprosthetic tissue at ×400 magnification.

Exclusion Criteria

The exclusion criteria included: (i) patients who refused to join the trial group; (ii) patients who lacked any of the essential examinations for the diagnosis of PJI; (iii) revision total hip/knee arthroplasty for periprosthetic fracture; and (iv) patients who received revision total hip/knee arthroplasty for metal allergy.

Data Acquisition

For each patient, comprehensive physical examination was conducted and medical history was collected, including the patient’s demographic data and data on the previous surgery.

Collected laboratory data included routine blood test results, C‐reactive protein (CRP), erythrocyte sedimentation rate (ESR), high sensitive C‐reactive protein (hs‐CRP), interleukin ‐6 (IL‐6), synovial leukocyte counts, as well as microorganisms isolated from periprosthetic tissues, articular fluid and the prosthesis. When patients had abnormal erythrocyte sedimentation rates and/or C‐reactive protein results, and PJI was strongly suspected, joint puncture and aspiration was conducted before surgery. In other circumstances, aspiration was conducted during surgery to avoid pollution and iatrogenic infection, according to the American Academy of Orthopaedic Surgeons (AAOS) guidelines19. Periprosthetic tissues and articular fluid were kept in sterile glass bottles and the prosthesis in sterile casing. The aspirated fluid was then sent for microbiologic culture. Patients did not use any types of antibiotics for a minimum of 2 weeks prior to obtaining the intra‐articular culture20. All processes were carried out without pollution.

According to a recent meta‐analysis, sonication is a promising diagnostic tool for PJI21, so for each patient, implant sonication was regularly conducted, after which sonicate fluid and the implant‐surrounding tissue were sent for microbiologic culture and a subsequent drug‐resistance test22, 23. If necessary, the microbiologic culture would continue for 4 weeks. We used the frozen and paraffin sections of peri‐implant tissues24, 25.

Statistical Methods

Patients were divided into two groups according to the diagnosis criterion. Patients who met the criteria for PJI were identified as the infectious group; the others were identified as the non‐infectious group. Continuous variables are reported as the mean and the standard deviation (SD), and categorical variables are reported as proportions. Differences between groups at the baseline were assessed using the two‐sided t‐test and the χ²‐test for continuous and categorical variables, respectively. Statistical significance was set as P < 0.05. All statistical analyses were performed using IBM SPSS software for Windows version 20.0 (SPSS, Chicago, IL, USA).

Results

Patients’ History

The demographic characteristics of 160 participants are presented in Table 1. No significant differences were found in sex and age between the two groups. The median survival duration of prostheses in the infectious group was 3.0 (1.0, 8.0) years, and 7.0 (2.0, 12.0) years in the non‐infectious group. There was statistical difference between the two groups (χ² = 118.83, P < 0.01).

Table 1.

Baseline characteristics

Groups	n	Gender (male/female)	Age (years, mean ±SD)	Hip/Knee
Infectious group	59	27/32	65.1 ±10.8	32/27
Non‐infectious group	101	36/65	64.6 ±10.5	51/50
χ2/t‐value	—	0.66	1.02	1.89
P‐value	—	0.209	0.778	0.650

Pathogens

In total, 42/59 patients’ microbiological culture in the infectious group and 17/101 patients’ microbiological culture in the non‐infectious group showed positive results, and 69 strains of pathogens were isolated, of which 60 were Gram‐positive (G+) bacterium (86.96%), 8 were Gram‐negative (G−) bacterium (11.59%), and one was fungus (1.45%). Among the G+ bacterium, Staphylococcus epidermidis was the most common pathogen (19, 27.54%), followed by Corynebacterium (7, 27.54%). Five cases of mixed‐pathogen infection were observed, of which 2 pathogens were isolated in 4 cases (Staphylococcus epidermidis and Ralstonia pickettii, Staphylococcus epidermidis and Ralstonia pickettii, Staphylococcus haemolyticus and Micrococcus luteus, Staphylococcus epidermidis and Enterococcus, respectively), and in the other case, 3 pathogens were isolated (Sphingomonas paucimobilis, Staphylococcus aureus, and Candida parapsilosis). One case of Enterococcus faecium infection and 2 cases of Enterococcus faecalis infection were established (Table 2).

Table 2.

Microbiologic culture details

Bacteria	n	Percentage (%)
Gram‐positive bacterium	60	86.96
Staphylococcus epidermidis	19	27.54
Corynebacterium	7	10.14
Staphylococcus aureus	5	7.25
Staphylococcus capitis	3	4.35
Staphylococcus cohnii	3	4.35
Staphylococcus hominis	3	4.35
Staphylococcus haemolyticus	3	4.35
Micrococcus luteus	3	4.35
Enterococcus	3	4.35
Staphylococcus caprae	2	2.90
Staphylococcus auricularis	1	1.45
Staphylococcus chromogenes	1	1.45
Staphylococcus warneri	1	1.45
Streptococcus parasanguis	1	1.45
Kocuria kristinae	1	1.45
Streptococcus dysgalactiae	1	1.45
Arthrobacterium	1	1.45
Solution of thiamine bacillus	1	1.45
Finegoldia magna	1	1.45
Gram‐negative bacterium	8	11.59
Propionibacterium acnes	2	2.90
Pseudomonas aeruginosa	2	2.90
Sphingomonas paucimobilis	1	1.45
Ralstonia pickettii	1	1.45
Serratia marcescens	1	1.45
Genus aromatic	1	1.45
Fungus	1	1.45
Candida parapsilosis	1	1.45

Drug Resistance

The drug‐resistance rate is an important factor in the diagnosis and treatment of PJI, and greatly influences surgeons’ decision‐making regarding antibiotics use. Penicillins, cephalosporins, macrolides, lincosamides, and antituberculotic showed high drug‐resistance rates (Table 3); Among penicillins, 33 of 42 strains (78.57%) showed penicillin resistance while 22 of 37 strains (59.46%) showed oxacillin resistance. The drug‐resistance rates for erythromycin and clindamycin were also high (66.67% and 44.74%, respectively). We discovered 1 strain of imipenem‐resistant Pseudomonas aeruginosa. Of 19 strains of Staphylococcus epidermidis observed, almost all of them were penicillin‐resistant and erythromycin‐resistant; meanwhile, approximately half of them were clindamycin‐resistant and cotrimoxazole‐resistant, and one‐quarter of them were ciprofloxacin‐resistant. Moreover, 6 strains of Staphylococcus epidermidis were methicillin‐resistant coagulase‐negative staphylococci (MRSCoN). One strain of Enterococcus faecium and 2 strains of Enterococcus faecalis were found, and high‐level aminoglycoside resistance (HLAR) was identified in 1 of the Enterococcus faecalis and Enterococcus faecium. No vancomycin‐resistant bacteria and Mycobacterium tuberculosis were found.

Table 3.

Drug resistance of each strain

Drugs	Total number	Drug‐resistant strain	Percentage (%)
Penicillins
Oxacillin	37	22	59.46
Penicillin	42	33	78.57
Ampicillin	5	2	40.00
Piperacillin	5	0	0.00
Oxacillin	1	1	100.00
Cephalosporins
Ceftriaxone	6	2	33.33
Cefotaxime	2	0	0.00
Cefazolin	5	2	40.00
Cefepime	8	0	0.00
Ceftazidime	5	1	20.00
Cefoxitin	3	1	33.33
Cefuroxime	5	1	20.00
Cefotetan	3	2	66.67
Carbapenems
Imipenem	5	1	20.00
Meropenem	4	1	25.00
Ertapenem	1	0	0.00
Monobactams
Aztreonam	3	1	33.33
β‐lactams/β‐lactamase inhibitor
Ampicillin/Sulbactam	5	1	20.00
Piperacillin/Tazobactam	4	0	0.00
Cefoperazone/Sulbactam	3	0	0.00
Aminoglycoside
Gentamicin	45	12	26.67
Amikacin	4	0	0.00
Tobramycin	5	0	0.00
Tetracyclines
Tetracycline	39	7	17.95
Tigecycline	38	0	0.00
Chloramphenicols
Chloramphenicol	1	0	0.00
Macrolides
Erythromycin	42	28	66.67
Lincosamides
Clindamycin	38	17	44.74
Polypeptide
Vancomycin	43	0	0.00
Teicoplanin	40	0	0.00
Polymyxin B	1	0	0.00
Fluoroquinolone
Levofloxacin	45	9	20.00
Moxifloxacin	36	5	13.89
Antituberculotic
Rifampicin	37	7	18.92
Others
Nitrofurantoin	2	0	0.00
Cotrimoxazole	37	12	32.43
Linezolid	42	0	0.00
KuiNu leptin/dafoe leptin	21	1	4.76

Discussion

Although the usage of prophylactic antibiotics has reduced the rate of revision TJA26, aseptic loosening and infection are still two key reasons2, 27, 28, 29. In some published studies, PJI is cited as the most common reason for revision5, but diagnosing PJI is sometimes difficult. Articular fluid culture is important and necessary when the patients’ CRP and ESR are high. In addition, the drug‐resistance test can provide information to surgeons in choosing the proper antibiotics in the treatment of PJI.

This is a descriptive study investigating common pathogens in PJI and drug‐resistance rates for different commonly‐used antibiotics. We also included patients with aseptic loosening because some patients might have low‐grade infections in this group. Gram‐positive bacterium is still the most common pathogenic bacteria in PJI, accounting for over 50% of cases30, 31, 32, and in our study the proportion of isolated G+ bacteria was 86.9% (60/69); Staphylococcus epidermidis and Staphylococcus aureus were the largest in number.

Two‐stage revision surgery and one‐stage revision are treatment options for PJI. In the perioperative period, antibiotics use is crucial and is longer compared to primary total joint arthroplasty; therefore, microbiological culture and drug sensitive tests are important, whether one‐stage or two‐stage revision is undertaken33. Cefuroxime is the most commonly used prophylactic antibiotic for elective orthopaedic surgery. Clindamycin is prescribed instead if the patient has a history of being allergic to cephalosporins; however, we found that both second generation cefuroxime (20%) and clindamycin (44.7%) exhibited high resistance rates in this group of patients.

Cefuroxime and clindamycin as the prophylactic antibiotic might be efficient for primary TJA; however, because of the high drug‐resistance rate of cefuroxime (20%) and clindamycin (44.74%), for patients who have undergone revision or re‐revision TJA, especially when PJI was strongly suspected, they might not be appropriate. Fortunately, no vancomycin‐resistant bacteria were identified, and had not been observed in any other institution in China either34. Therefore, vancomycin can be selected as the empiric antibiotic treatment for PJI before the results of drug‐resistance tests come out, and for prophylaxis in cases with high risk of PJI, because G+ bacterium is most common in PJI. However, for G− bacterium, imipenem might be more applicable; unfortunately, we found 1 strain of imipenem‐resistant Pseudomonas aeruginosa.

The failure of microbiological culture makes the diagnosis of PJI even more difficult, and culture‐negative PJI accounted for 10% of all cases35, 36. In the PJI group (59 patients), only 42 patients were culture‐positive (71.19%). One possibility is that the harvest method of periprosthetic tissue samples was not appropriate; for example, using the electrotome to resect the tissues could sterilize the bacteria and affect the culture rate.

Pathogens were isolated from 17 patients in the non‐infectious group (17%). One possibility is that the specimen was polluted. Moreover, low‐grade infections cannot be overlooked. The method of 16S polymerase chain reaction (16S PCR) has been demonstrated to confirm the existence of pathogens when the results of microbial cultivation are negative in low‐grade infection37, 38, and the analysis of antimicro peptide expression in synovial fluids could also provide evidence of low‐grade infection39. Clarke et al. report that bacterial DNA could be found in patients with aseptic loosening, although they identified bacterial DNA in 21% of the tissue samples from primary total hips as well; this could possibly indicate a high contamination rate40. In this study, we did not use the method of 16S PCR; further studies should examine more detection methods to help diagnose PJI. Fortunately, it might not affect implant survival for one‐stage revision surgery, although in cases of underdiagnosis of low‐grade infection in presumed aseptic revision THA, as reported by Boot et al. 41.

Our study has some limitations. First, some patients’ synovial failed to be collected, and the tests such as 16S PCR and bone scan imaging were not adopted for every patient; therefore, some cases of low‐grade infection may have been ignored. Second, the sample size of the current study is relatively large compared to others and it was conducted in a prospective manner. Even so, we believe that our study could provide some information on current treatment of PJI. Previous studies have found that decolonization plays an essential role in postoperative infection, and the use of prophylactic antibiotics against the decolonization might decrease the rate of PJI30, 42, 43. Further study should focus on finding common decolonizing bacterium in patients undertaking elective orthopaedic surgery in China, and further test the appropriateness of using cefuroxime and clindamycin in prophylaxis in primary TJA.

Conclusion

The current study provides some information on the most common pathogens in revision TJA in our institution. Use of cefuroxime and clindamycin in revision TJA patients should proceed with caution, because of the high drug‐resistance rate. Vancomycin is ideal for infection prevention in cases with high risk of PJI and empiric use for the treatment of PJI.