Infection after shoulder arthroplasty is relatively rare but

Transcription

1 Chinese Medical Journal 2014;127 (22) 3975 Review article Unexpected positive cultures including isolation of Propionibacterium acnes in revision shoulder arthroplasty Seung Ju Kim and Jong Hun Kim Keywords: revision shoulder arthroplasty; unexpected positive culture; Propionibacterium acnes Objective Little information is available to guide treatment strategies regarding patients with unexpected positive cultures (UPC), including Propionibacterium acnes (PA), without overt signs of infection in revision shoulder arthroplasty (RSA). The purpose of our study was to analyze the prevalence, clinical meaning, treatment and prognosis of UPC in RSA. Data sources and study selection We performed a systematic review of the literature between 1950 and 2013 for all studies reporting on UPC for PA. Studies with the prevalence and prognosis of patients with UPC in RSA were analyzed. Results Six studies meeting our inclusion criteria yielded data for patients who underwent a total of 1405 RSA. Based on the available data, following RSA 235 shoulders had UPCs with a pooled percentage of 16.7% (235/1405). The most commonly isolated bacteria from shoulders following RSA with UPCs was PA with pooled percentages of 63.4% (149/235). Occurrence of true infection from UPCs after RSA was seen in 24 shoulders (24/235, 10.2%). Antibiotic use did not influence the rate of the occurrence of true infection from UPCs (P=0.498). Conclusions Our study showed a low risk of having a true infection from UPCs after RSA without clinical signs of preoperative infection at the time of the surgery. Therefore, prolonged antibiotic therapy may not be necessary in these patients. Chin Med J 2014;127 (22): Infection after shoulder arthroplasty is relatively rare but is nonetheless a devastating complication. Currently, there have been only a few reports of patients without overt infection who had an unexpected positive culture (UPC) during revision shoulder arthroplasty (RSA). 1,2 This subgroup of patients is one of the most challenging clinical dilemmas for making an appropriate management decision, because most of these patients do well clinically in the postoperative period and UPCs are often unexpectedly found to be positive after their hospital discharge. 3 Until quite recently, Propionibacterium acnes (PA) was considered to be a contaminant organism rather than a true pathogen. 3 It has been thought that the shoulder has a tendency for positive cultures of PA as the organism is the dominant bacterium that is often isolated from normal skin in moist areas such as the axilla. 3-5 Nevertheless, PA has been identified to be a causative pathogen in true infections following shoulder instability surgery, 6 rotator cuff repair, 7 and arthroplasty. 8 Therefore, an UPC for PA at the time of RSA cannot be simply overlooked as a contaminant. 9 For these reasons, practical guidelines in relation to the evaluation and management of UPC including PA in RSA would seem to be of clinical value. However, our current knowledge about the role of UPC for PA is confined to a few reviews from single institutions. Therefore, we performed a systematic review of the literature regarding UPC in RSA and asked the following questions: What are the prevalence of UPC, isolated pathogens, and rates of occurrence of true infection for patients undergoing RSA? Are intensive antimicrobial treatment strategies necessary to reduce occurrence of true infection in patients without overt clinical signs of infection before or during the revision surgery? Our hypothesis was that intensive antimicrobial treatment after surgery might not be necessary to reduce occurrence of true infection in patients with UPC. Data sources and analysis Search strategy and criteria We searched PubMed, MEDLINE, Cumulative Index to Nursing and Allied Health (CINAHL), and the Cochrane Library on May 15, 2013, for articles that were published between 1950 and Then, we performed a second search on July 15, 2013, to identify additional articles. The search terms were arthroplasty, shoulder, infection, and propionibacterium. After discarding all duplicate articles and restricting to literature available in the English language and pertaining to human subjects, we further narrowed the search to include only those articles that focused on RSA. A flow- DOI: /cma.j.issn Department of Orthopaedics, Korea Electric Power Corporation (KEPCO) Medical Foundation, KEPCO Medical center, Seoul, Republic of Korea (Kim SJ) Division of Infectious Diseases, Department of Internal Medicine, University of Utah, Salt Lake City, UT 84132, USA (Kim JH) Correspondence to: Jong Hun Kim, M.D., Division of Infectious Diseases, Department of Internal Medicine, University of Utah SOM 4B N 1900 E, Salt Lake City, UT 84132, USA (Tel: 1 (801) Fax: 1 (801) smonti1976@hotmail.com)

2 3976 Chin Med J 2014;127 (22) Statistical analysis Data were recorded using Microsoft Excel 2007 version (Microsoft Corp, Redmond, WA, USA) and analyzed using SPSS software (SPSS Inc, Chicago, IL, USA). The χ 2 test was used to evaluate differences in the recurrence rates after RSA between group I (with antibiotic treatment) and group II (without antibiotic treatment). P <0.05 was regarded as significant. Results Figure 1. Flow diagram of literature search. chart depicting our search criteria and outcome is shown in the Figure. The inclusion criteria included: articles published from January 1, 1950 to May 15, 2013, English language articles with human species, electronic publications that reported cases of UPC, both retrospective and prospective series, cases with no clinical sign or diagnostic marker of infection at the time of revision, and articles that evaluated occurrence of true infection following the revision arthroplasty. The exclusion criteria included: cases of primary arthroplasty, articles focusing on the preoperative factors of UPC, articles without data of positive or negative manifested pathogen(s), evaluation of any joint other than the shoulder (such as elbow, knee, and hip), and articles without treatment methods. Limits for the number of patients in each study or the minimum duration of follow-up were not used. All study designs from Level I to IV were eligible. Bibliographies of the included studies were also searched to identify relevant studies. Based on the above search criteria, six clinical studies were available for the final analysis. 1,2,10-13 Owing to a lack of prospective randomized studies, most of the larger cohorts that resolved, or at least gained an insight to, clinical problems were selected for this review. Data extraction Relevant information such as the level of evidence, patient age at surgery, gender, preoperative blood tests, microorganism isolated during the UPC, prevalence of UPC, any reported complications, treatment methods including antibiotic therapy, occurrence of true infection after RSA and the mean follow-up period were extracted from each of these studies as measurement of clinical outcomes. Literature review Six level IV studies 1,2,10-13 meeting our inclusion criteria yielded data from 1402 patients who underwent a total of 1405 RSA procedures. No Level I or II studies were identified with our search. While the numbers of RSAs, positive cultures, pathogens and occurrence of true infections were available for all subjects, other variables such as patient age, gender, preoperative diagnosis with laboratory markers, and postoperative care were not found on a consistent basis across all studies. Uniform criteria have not typically been used for the diagnosis of infection of prosthetic joints caused by PA. Scoring systems were not consistently used to assess the clinical outcomes. 2,10 Moreover, untoward events, such as complications and the need for unplanned surgical procedures, were either not mentioned or documented in an inconsistent fashion. Baseline characteristics The average age of the patients at revision surgery was 62.9 years old with a mean follow-up of 3.9 years. The number of patients in each study ranged from 27 to 678. Detailed information regarding demographics of patients is shown in Table 1. No shoulder had clinical evidence of an active infection prior to the revision surgery. The reasons for revision were stiffness, pain, fracture, dislocation, or loosening. Preoperatively, most patients had normal values for C-reactive protein (CRP) (0 to 10 mg/dl), erythrocyte sedimentation rate (ESR) (0 to 15 mm/h), and white bloodcell count (WBC) (4.3 to /L) according to most articles. 1,2,12,13 Of note, mean preoperative CRP and ESR for patients who were found to have unexpected positive cultures were 0.6 mg/dl and 11.9 mm/h in the majority of articles. 1,2,12,13 Unexpected positive cultures Based on the available data, 235 shoulders from 1405 RSA procedures had intraoperative positive cultures with a pooled percentage of 16.7% (235/1 405). The most commonly isolated bacterial pathogen was PA, which was identified in 149 shoulders with a pooled percentage of 63.4% (149/235) (Table 2). Coagulase-negative

3 Chinese Medical Journal 2014;127 (22) 3977 Table 1. Data on the studies Author Year Revision period Number of shoulder Age Percentage male revisions (patients) (range) (%, male/female) Mean follow-up in years (range) Lorenzetti et al (55) 63 (51 75) 33 (18/37) 1.7 (16 days to 6.7 years) Foruria et al (678) 61.9 (27 83) 51 (348/330) 5.6 (32 days to 25 years) Grosso et al (187) 66.5 ( ) Not available 3 (22 months to 84 months) Kelly et al (27) 62 (43 81) 33 (9/18) 1.8 (13 months to 37 months) Topolski et al (389) 59.5 (27 79) Not available 5 (1 day to 18.6 years) Cheung et al (66) 65 (29 84) 60 (41/27) 5 (0.8 years to 26.2 years) Total (1402) 62.9 (27 84) 50 (416/412) 3.9 (1 day to 26.2 years) Author Number of shoulder revisions (patients) Shoulder revisions with UPC (n (%)) Table 2. Outcomes of the studies Number of PA (PA/ number of shoulder revisions with UPC, %) Occurrence of true infection (occurrence of infection/number of shoulder revisions with UPC, %) Occurrence of true infection with PA (PA/occurrence of true infection, %) Lorenzetti et al (55) 8 (14.5) 6 (75.0) 3 (37.5) 1 (33.3) Foruria et al (678) 107 (15.8) 68 (63.6) 11 (10.3) 10 (90.9) Grosso et al (187) 17 (9.1) 10 (58.8) 1 (5.9) 0 (0.0) Kelly et al (27) 8 (28.6) 6 (75.0) 2 (25.0) 2 (100.0) Topolski et al (389) 75 (19.3) 45 (60.0) 7 (9.3) 5 (71.4) Cheung et al (66) 20 (29.4) 14 (70.0) 0 (0.0) 0 (0.0) Total (1402) 235 (16.7) 149 (63.4) 24 (10.2) 18 (75.0) UPC: unexpected positive culture; PA: Propionibacterium acne staphylococcus was the second most common identified pathogen of UPCs. The pathogens of UPCs in each study are presented in Table 3. Antibiotic treatment After detection of the UPCs, patients were treated with either intravenous (IV) or oral antibiotic therapy (group I) or without antibiotic therapy except for 2 to 3 doses of IV antibiotics given postoperatively (group II). Group I (32.0%, 75 of 235) received IV or oral antibiotic therapy in addition to routine postoperative dosing after the confirmation of UPCs. Of note, one patient was treated with oral doxycycline for three weeks for cellulitis overlying the skin incision for an iliac crest bone graft. 11 The type and duration of antibiotics varied and were not completely documented in all cases. Group II (68.0%, 160 of 235) received the standard 2 to 3 doses of IV antibiotic therapy postoperative without further antibiotic therapy afterwards. Reasons for not using antibiotic after discovery of UPCs included the interpretation of the positive culture as a false-positive or that the positive culture was not noticed. Occurrence of true infection Occurrence of true infection following the discovery of UPCs after RSA occurred in 24 shoulders (10.2%, 24 of 235). There were no significant differences in the rates of occurrence of true infection between group I and group II (odds ratio = 1.45, 95% CI: , P=0.498). Of note, occurrence of true infection caused by PA from UPCs after RSA was similar (12.1%, 18/149) to that of the overall occurrence of true infection from UPCs after RSA. However, PA was the most significant major pathogen in the cases of occurrence of true infection (75.0%, 18 of 24). Most cases of occurrence of true infection required surgical treatment ranging from debridement, 2 removal of prosthesis followed by resection arthroplasty, 11 or rerevision arthroplasty. 1,12,13 Table 3. Pathogens (unexpected positive culture) in revision shoulder arthroplasty Pathogens Lorenzetti et al 12 Foruria et al 9 Grosso et al 10 Kelly et al 12 Topolski et al 12 Cheung et al 6 Total Propionibacterium acne Coagulase-negative Staphylococcus Methicillin-sensitive Staphylococcus aureus Methicillin-resistant Staphylococcus aureus Staphylococcus saccharolyticus Other Staphylococcus spp Streptococcus viridans Peptostrepcoccus spp Corynebacterium spp Bacillus spp Veillonella spp Pseudomonas spp Proteus spp Alcaligenes spp Other Gram-negative bacilli Aspergillus spp Candida spp Multiple * 0 4 * 0 15 * Overlapping with one or more pathogens identified in the individual study.

4 3978 Discussion Orthopedic surgeons often recognize positive cultures taken from presumably uninfected shoulders during revision arthroplasty. The significance, optimal treatment strategy, and outcome for patients with positive intraoperative cultures and no overt signs of infection are largely unknown and have recently been the issue of debate 5,14 with extensive discussion in regard to hip and knee arthroplasty Despite the difficulties mentioned earlier, we were able to gather some clinically relevant information regarding UPC in shoulder arthroplasty by pooling the extracted data from the literature. We recognized a trend of positive cultures taken from presumably uninfected shoulders during RSA. However, the results of preoperative blood work and intraoperative pathologic analysis are often negative in this setting 13 and not useful for the diagnosis. Several reports on so-called false-positive intraoperative culture specimens obtained during revision arthroplasties have questioned the importance of positive intraoperative cultures in the absence of other indicators of infection. 18,19 As a consequence, the clinical meaning of these UPCs is controversial. One major concern for this subset of patients with UPCs without signs of overt clinical infection is that a true infection may occur in the future with devastating complications. 2 Our results show that UPCs can be expected in 16.7% of patients undergoing RSA without clinical signs of pre-operative infection. The most common pathogen cultured at revision surgery was PA, followed by coagulase-negative staphylococcus. However, the presence of a positive culture during RSA in patients without overt clinical signs of infection at the time of RSA was only weakly associated with occurrence of true infection afterwards (10.2%). Therefore, an UPC from the RSA in a patient without clinical signs of having active infection may be less likely to be associated with subsequent active infection. One thing to be noted is the role of PA from UPC in the occurrence of infection after RSA. Although the rate of progression from UPC to a true infection is low (12.1%), PA constituted significant proportion of confirmed cases of true infection (75.0%). Typical clinical or laboratory manifestations of infection may not be so apparent in cases of PA infection and the result of intraoperative culture may take weeks due to the slow-growing nature of PA. 20,21 Taken together, more careful and close monitoring of patients in the post-operative period, who are found to have PA from UPCs following RSA, needs to be considered. As the interpretation of the culture result is uncertain, the choice of treatment among patients with UPC having revision surgery becomes the question. There is no consensus on how to manage UPC when infection was not clinically suspected. Although it has been recommended that all cases of positive intraoperative cultures should be treated with antibiotics in the cases of hip arthroplasty, 22 no prospective study has looked at the outcomes of treatment with antibiotics as compared with no additional antibiotic treatment following the RSA with perioperative antibiotics. The practical options are the use of antibiotics Chin Med J 2014;127 (22) or observation. UPCs have been reported in other joints, with similar prevalence and results compared with the present study, including a wide variety of postoperative managements, and a low incidence of subsequent septic complications despite short periods of antibiotic treatment or no treatment at all. 1,2,17,23 In our pooled analysis, there were no significant differences in the occurrence of true infection rates between group I and group II. The majority of patients were successfully treated without prolonged antibiotic therapy with a low rate of occurrence of true infection afterwards. Therefore, re-intervention or prolonged postoperative antimicrobial treatment strategies do not seem to be generally necessary in this patient population. We noted several limitations within the existing literature on UPC in RSA that led to our inability to draw firm ideal conclusions or perform sophisticated statistical analyses. First, this was a pooled analysis of a several international databases that were not designed specifically for clinical research regarding UPCs following the RSA. Due to the heterogeneity of the literature reports covered in the systematic review, it was not possible to conduct a metaanalysis. Consequently, not all the data that we sought to include in our analysis was consistently available in the chosen studies. Certain variables, such as patient age and gender, length of follow-up, microbiologic culture incubation time, and period of antibiotic treatment were not always specified or were not reported in enough detail to allow advanced statistical inferences and comparisons. Second, only six studies were available as a basis for the review due to the paucity and suboptimal quality of the existing literature on this topic after application of our inclusion and exclusion criteria. However, a total of 1405 revision shoulder arthroplasties were included in this study, which might represent the current trend. In conclusion, UPC can be expected in 16.7% of RSA and the most prevalent bacterium is PA. When the intraoperative culture was read as positive, prolonged IV and/or oral antibiotic therapy may not be necessary in patients with little likelihood of having an infection. We believe that a decision can be made to observe and follow up these patients for development of subsequent infection. Careful preparation and draping can theoretically help to prevent UPC. However, as the ideal management pathway for UPC following RSA without clinical signs of preoperative infection is still unclear, surgeons treating these patients must maintain a high level of suspicion, especially for cases of PA isolation from UPCs. Additional prospective and retrospective studies with the involvement of larger numbers of patients with UPCs following RSA are needed to better define the optimal management strategy. REFERENCES 1. Foruria AM, Fox TJ, Sperling JW, Cofield RH. Clinical meaning of unexpected positive cultures (UPC) in revision shoulder arthroplasty. J Shoulder Elbow Surg 2013; 22:

5 Chinese Medical Journal 2014;127 (22) Grosso MJ, Sabesan VJ, Ho JC, Ricchetti ET, Iannotti JP. Reinfection rates after 1-stage revision shoulder arthroplasty for patients with unexpected positive intraoperative cultures. J Shoulder Elbow Surg 2012; 21: Dodson CC, Craig EV, Cordasco FA, Dines DM, Dines JS, Dicarlo E, et al. Propionibacterium acnes infection after shoulder arthroplasty: a diagnostic challenge. J Shoulder Elbow Surg 2010; 19: Athwal GS, Sperling JW, Rispoli DM, Cofield RH. Deep infection after rotator cuff repair. J Shoulder Elbow Surg 2007; 16: Patel A, Calfee RP, Plante M, Fischer SA, Green A. Propionibacterium acnes colonization of the human shoulder. J Shoulder Elbow Surg 2009; 18: Chandler D, King JD, Bernstein SM, Marrero G, Koh J, Hambrecht H. Results of 21 Wagner limb lengthenings in 20 patients. Clin Orthop Relat Res 1988; 230: Herrera MF, Bauer G, Reynolds F, Wilk RM, Bigliani LU, Levine WN. Infection after mini-open rotator cuff repair. J Shoulder Elbow Surg 2002; 11: Sperling JW, Kozak TK, Hanssen AD, Cofield RH. Infection after shoulder arthroplasty. Clin Orthop Relat Res 2001; 382: Levy PY, Fenollar F, Stein A, Borrione F, Cohen E, Lebail B, et al. Propionibacterium acnes postoperative shoulder arthritis: an emerging clinical entity. Clin Infect Dis 2008; 46: Cheung EV, Sperling JW, Cofield RH. Revision shoulder arthroplasty for glenoid component loosening. J Shoulder Elbow Surg 2008; 17: Kelly JD 2nd, Hobgood ER. Positive culture rate in revision shoulder arthroplasty. Clin Orthop Relat Res 2009; 467: Lorenzetti AJ, Wongworawat MD, Jobe CM, Phipatanakul WP. Cyanoacrylate microbial sealant may reduce the prevalence of positive cultures in revision shoulder arthroplasty. Clin Orthop Relat Res 2013; 471: Topolski MS, Chin PY, Sperling JW, Cofield RH. Revision shoulder arthroplasty with positive intraoperative cultures: the value of preoperative studies and intraoperative histology. J Shoulder Elbow Surg 2006; 15: Dramis A, Aldlyami E, Grimer RJ, Dunlop DJ, O'Connell N, Elliott T. What is the significance of a positive Propionibacterium acnes culture around a joint replacement? Int Orthop 2009; 33: Barrack RL, Aggarwal A, Burnett RS, Clohisy JC, Ghanem E, Sharkey P, et al. The fate of the unexpected positive intraoperative cultures after revision total knee arthroplasty. J Arthroplasty 2007; 22(Suppl 2): Berend KR, Lombardi AV Jr, Adams JB. Unexpected positive intraoperative cultures and gram stain in revision total hip arthroplasty for presumed aseptic failure. Orthopedics 2007; 30: Padgett DE, Silverman A, Sachjowicz F, Simpson RB, Rosenberg AG, Galante JO. Efficacy of intraoperative cultures obtained during revision total hip arthroplasty. J Arthroplasty 1995; 10: Athanasou NA, Pandey R, de Steiger R, Crook D, Smith PM. Diagnosis of infection by frozen section during revision arthroplasty. J Bone Joint Surg Br 1995; 77: Lonner JH, Desai P, Dicesare PE, Steiner G, Zuckerman JD. The reliability of analysis of intraoperative frozen sections for identifying active infection during revision hip or knee arthroplasty. J Bone Joint Surg Am 1996; 78: Schäfer P, Fink B, Sandow D, Margull A, Berger I, Frommelt L. Prolonged bacterial culture to identify late periprosthetic joint infection: a promising strategy. Clin Infect Dis 2008; 47: Bayston R, Nuradeen B, Ashraf W, Freeman BJ. Antibiotics for the eradication of Propionibacterium acnes biofilms in surgical infection. J Antimicrob Chemother 2007; 60: Tsukayama DT, Estrada R, Gustilo RB. Infection after total hip arthroplasty. A study of the treatment of one hundred and six infections. J Bone Joint Surg Am 1996; 78: Marculescu CE, Berbari EF, Hanssen AD, Steckelberg JM, Osmon DR. Prosthetic joint infection diagnosed postoperatively by intraoperative culture. Clin Orthop Relat Res 2005; 439: (Received March 6, 2014) Edited by Wang De