Improving the diagnosis of peri-prosthetic joint infections

Prosthetic joint infections are a devastating complication of arthroplasty. Despite this, current culture techniques lack sensitivity and not standardized. We examined the inoculation of peri-prosthetic tissue specimens into blood culture bottles, demonstrating improved sensitivity compared to conventional methods. These findings influence patient care, allowing accurate diagnosis of infection.

by Dr T. N. Peel, B. L. Dylla, J. G. Hughes, D. T. Lynch, K. E. Greenwood-Quaintance,  Prof. A. C. Cheng,  Dr J. N. Mandrekar and  Prof. R. Patel

Background
It is estimated that over 140 000 patients in the USA underwent revision arthroplasty in 2015 [1]. At the time of revision, it is recommended that biopsies of the tissues surrounding the prosthesis be obtained to diagnose or exclude prosthetic joint infection (PJI), a potential cause of arthroplasty failure [2]. PJIs are associated with significant patient morbidity and substantial healthcare costs [1, 3, 4]. In the setting of an ageing population and increasing demand for arthroplasty, the negative impact of PJIs will continue to compound [1, 4]. Despite the recognized impact of PJIs, diagnosis remains challenging owing to two major issues: (i) the lack of a ‘gold standard’ definition for infection and (ii) current imperfect diagnostic techniques [4].

PJI diagnosis and culture conditions
The isolation of an identical microorganism (or microorganisms) from two or more aseptically obtained peri-prosthetic specimens confirms the diagnosis of PJI [3]. In addition to confirmation of infection, microbiological cultures enable antimicrobial susceptibility testing to ensure institution of appropriate, targeted antimicrobial therapy. There are limited studies examining the optimal media for culturing these peri-prosthetic tissue biopsies [4]. Conventional microbiological techniques frequently include culture on various types of aerobic and anaerobic agar plates, and inoculation into broths such as thioglycollate broth [4]. These conventional techniques, alongside various durations of incubation, are not standardized and may have a low sensitivity, as low as 39% in some published studies, particularly in the setting chronic infection [5–7]. Recent research has focused on strategies that optimize the diagnosis of PJI, including the inoculation of peri-prosthetic tissue biopsies into blood culture bottles (Fig. 1).

Assessment of culture conditions
Previous research
Three previous studies examined inoculation of peri-prosthetic tissue specimens into blood culture bottles for the diagnosis of PJI. Baker and colleagues examined inoculation of synovial fluid aspirates and tissue specimens into blood culture bottles using the Sentinel system (Difco Laboratories) automated blood culture system. There were limitations of this study’s design and methodology, including lack of a control group and lack of comparison with conventional techniques [8]. In the second study, by Hughes and colleagues, four tissue culture techniques were examined – culture on blood and chocolate agars and in Robertson’s cooked meat broth, fastidious anaerobic broth and aerobic and anaerobic blood culture bottles. The aerobic and anaerobic blood culture bottles were incubated on the BACTEC 960 platform for 5 days (BD Diagnostic Systems). The authors defined PJI on the basis of histopathological findings, with 23 subjects meeting this criterion for infection from the 141 subjects included in the cohort [9]. Blood culture bottles were more sensitive compared to direct tissue culture and fastidious anaerobic broth (87% [95% confidence intervals {CI}, 72–100%] versus 39% [95% CI, 18–61%]; P = 0.007 and 57% [95% CI, 35–78%]; P = 0.016, respectively). There was no difference in sensitivity between culture in blood culture bottles and cooked meat broth (83% [95%CI, 66–99%]; P = 0.74) [5]. These previous studies used paired-design methodologies which may lead to the miscalculation of the true sensitivity of tests [10]. In a follow up study, Minassian and colleagues examined the optimal duration of incubation for peri-prosthetic tissue samples in blood culture bottles. The majority of blood culture bottles flagged positive within 3 days of incubation. In contrast, Propionibacterium species was isolated after a median of five days incubation (range 3–13 days). Terminal sub-culture of 1000 blood culture bottles at the end of the 14-day incubation period did not isolate clinically significant microorganisms. The authors concluded that prolonged culture for 14 days was not indicated, suggesting an optimal duration of 4 days incubation based on receiver operator characteristics (ROC) analysis [11].

Current study objectives and design
We undertook a larger, prospective cohort study at the Mayo Clinic, Rochester, MN, USA over a 9-month period (August 2013 – April 2014). The study included 369 consecutive subjects undergoing revision arthroplasty, of which 71 subjects (19%) underwent upper limb revision arthroplasty (49 shoulder joint and 22 elbow joint revisions). Overall, 117 study subjects met Infectious Diseases Society of America (IDSA) criteria for PJI. Applying the definitions proposed by Tsukayama and colleagues, the majority of subjects had late chronic infections (82%), 7% had early post-operative infections and 11% had hematogenous infections [7]. The majority (49%) of infections were caused by staphylococci, including Staphylococcus aureus which was isolated in 24% of culture positive cases. In contrast, Propionibacterium acnes was isolated in 41% of shoulder arthroplasty infections.

The study compared the sensitivity and specificity of inoculation of peri-prosthetic tissue specimens into blood culture bottles with standard agar and thioglycollate broth culture. The performance of the different media was analysed using Nemar’s test of paired proportion. In addition, Bayesian latent class modelling was undertaken; this statistical method overcomes potential limitations of traditional analysis as it assumes that no gold standard exists and that the true disease prevalence is unknown, both germane to PJI [10].

Current study results
Inoculation of tissues into blood culture improved sensitivity by 47% compared to conventional agar and broth cultures applying Bayesian latent class modelling (92.1% [95% credible interval, 84.9–97.0%] versus 62.6% [95% credible interval, 51.7–72.5%], respectively); this magnitude of difference was similar when IDSA criteria were applied (60.7% [95% CI, 51.2–69.6%] versus 44.4% [95% CI, 35.3–53.9%], respectively; P = 0.003). The specificity of culture in blood culture bottles was similar to conventional media, if two or more cultures were required to be positive. In 13 subjects (11%) of the 117 subjects meeting the IDSA criteria for PJI, inoculation of peri-prosthetic tissue samples into blood culture bottles detected microorganisms not found using other culture media. However, inoculation of peri-prosthetic tissue specimens into blood culture bottles failed to identify the pathogen that was detected by conventional methods in five PJI subjects (4%).

Microorganisms were detected earlier using blood culture bottles compared to conventional media; aerobic and anaerobic blood culture bottles flagged positive at a median of 21 and 23 hours (interquartile range [IQR] 14–45 and 16–47 hours, respectively) compared to 41 hours (IQR 21–63) for aerobic agar, 62 hours (IQR 43–144) for anaerobic agar and 65 hours (IQR 43–92) for thioglycollate broth.

Discussion
Culture using blood culture bottles has several advantages: it provides a semi-automated method, with positive results automatically flagging by the instrument, potentially yielding faster time-to-positivity than conventional methods; no technologist intervention is required for negative results except for removal of the bottles at the end of the incubation period; furthermore, the technology is available and used in most microbiology laboratories. Use of blood culture bottles overcomes a current limitation of total laboratory automation, the inability to handle anaerobic cultures.

The timely detection of microorganisms in blood culture bottles compared to conventional methods also has a number of clinical advantages particularly as detection of microorganisms in blood culture bottles may be combined with direct species identification using rapid diagnostics, such as matrix-assisted laser desorption ionization time-of-flight mass spectrometry (MALDI-TOF MS) or rapid nucleic acid amplification tests. Given the speed to detection of microorganisms in blood culture bottles, it is theoretically possible that the causative pathogen could be identified with resistance characterization within as few as 24 hours of revision surgery. This would facilitate optimization of antimicrobial therapy and minimize unnecessary antimicrobial use [12].

Conclusion
This study informs clinical care for patients undergoing revision surgery, demonstrating improved sensitivity for diagnosis of PJI with inoculation of peri-prosthetic tissue specimens into blood culture bottles compared to conventional culture techniques. The use of automated blood culture systems also yields faster results with the potential for pathogen identification within the first day of surgery.

References
1. Kurtz SM, Ong KL, Lau E, Bozic KJ. Impact of the economic downturn on total joint replacement demand in the United States: updated projections to 2021. J Bone Joint Surg Am. 2014; 96: 624-30.
2. International Consensus Meeting on Periprosthetic Joint Infection. Musculoskeletal Infection Society; 2013 (http://www.msis-na.org/wp-content/themes/msis-temp/pdf/ism-periprosthetic-joint-information.pdf).
3. Osmon D, Berbari E, Berendt A, Lew D, Zimmerli W, Steckelberg J, Rao N, Hanssen A, Wilson W, Infectious Diseases Society of America. Diagnosis and management of prosthetic joint infection: clinical practice guidelines by the Infectious Diseases Society of America. Clin Infect Dis. 2013; 56:e1–e25.
4. Tande AJ, Patel R. Prosthetic joint infection. Clin Microbiol Rev. 2014; 27: 302–345.
5. Hughes H, Newnham R, Athanasou N, Atkins B, Bejon P, Bowler I. Microbiological diagnosis of prosthetic joint infections: a prospective evaluation of four bacterial culture media in the routine laboratory. Clin Microbiol Infect. 2011; 17: 1528–1530.
6. Font-Vizcarra L, Garcia S, Martinez-Pastor JC, Sierra JM, Soriano A. Blood culture flasks for culturing synovial fluid in prosthetic joint infections. Clin Orthop Relat Res. 2010; 468: 2238–2243.
7. 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: 512–523.
8. Baker S, Fraise AP. Use of Sentinel blood culture system for analysis of specimens from potentially infected prosthetic joints. J Clin Pathol. 1994; 47: 475–476.
9. Hughes HC, Newnham R, Athanasou N, Atkins BL, Bejon P, Bowler IC. Microbiological diagnosis of prosthetic joint infections: a prospective evaluation of four bacterial culture media in the routine laboratory. Clin Microbiol Infect. 2011; 17: 1528–1530.
10. Collins J, Huynh M. Estimation of diagnostic test accuracy without full verification: a review of latent class methods. Stat Med. 2014; 33: 4141–4169.
11. Minassian A, Newnham R, Kalimeris E, Bejon P, Atkins B, Bowler I. Use of an automated blood culture system (BD BACTEC) for diagnosis of prosthetic joint infections: easy and fast. BMC Infect Dis.2014; 14: 233.
12. Tamma PD, Tan K, Nussenblatt VR, Turnbull AE, Carroll KC, Cosgrove SE. Can matrix-assisted laser desorption ionization time-of-flight mass spectrometry (MALDI-TOF) enhance antimicrobial stewardship efforts in the acute care setting? Infect Control Hosp Epidemiol. 2013; 34: 990–995.

The authors
Trisha N. Peel^1,2 MBBS, FRACP, PHD, Brenda L. Dylla¹ MT (ASCP), John G. Hughes¹ MT (ASCP), David T. Lynch¹ MT (ASCP), Kerryl E. Greenwood-Quaintance¹ MS,  Allen C. Cheng^3,4 MSSB, FRACP, MPH, PhD, MBIOSTATs,  Jayawant N. Mandrekar⁵ PhD,  Robin Patel*^1,6 MD
¹Division of Clinical Microbiology, Department of Laboratory Medicine and Pathology, Mayo Clinic, Rochester, MN, USA
²Department of Surgery, St Vincent’s Hospital Melbourne, University of Melbourne, Melbourne, Australia
³Department of Infectious Diseases, Alfred Hospital, Melbourne, Australia
⁴Department of Epidemiology and Preventive Medicine, Monash University, Melbourne, Australia
⁵Division of Biomedical Statistics and Informatics, Department of Health Sciences Research, Mayo Clinic, Rochester, MN, USA
⁶Division of Infectious Diseases, Department of Medicine Mayo Clinic, Rochester, MN, USA

*Corresponding author
E-mail: Patel.robin@mayo.edu