What's New in Musculoskeletal Infection.

Abstract

This update on musculoskeletal infection presents a review of infection-related articles from January 2020 through December 2020 in English-language journals, located using the National Center for Biotechnology Information (NCBI) website with a special emphasis on periprosthetic joint infection (PJI). The additional sections cover recent salient articles in the areas of spine and trauma. Introduction PJI is a devastating complication for patients expecting pain relief and improved function. However, the negative psychological impact is not limited to the patient. There exists a concurrent negative emotional impact on the treating surgeon. In a sample of Swedish arthroplasty surgeons, many felt guilt, stress, and a sense of failure if their patient incurred a PJI. Receiving peer support was the most affective coping strategy1. Trying to identify those patients at greatest risk for PJI remains a challenge. Body mass index (BMI) has been shown to be a better predictor of PJI risk than the thickness of adipose tissue in total knee arthroplasty (TKA). The systemic effects of obesity and metabolic syndrome may be more important than local adipose factors2. Recently, the effect of socioeconomic status on PJI risk was studied. Using a claims database, the authors found that patients with Medicaid as a primary payer were at greater risk for experiencing PJI3. The importance of careful initial treatment during the primary surgical procedure was highlighted in a study evaluating PJI risk. In patients requiring aseptic revision within 1 year of the index surgical procedure, the risk of PJI was much higher compared with patients who did not require revision4. Despite improved infection prevention protocols, little progress has been made in decreasing the incidence of PJI. In a population-based cohort study, no improvement in infection rates was noted over a 15-year period5. Although the incidence of PJI remains unchanged over time, a study to evaluate treatment was similarly disappointing. The survivorships of irrigation and debridement and 2-stage revisions were comparable between 2000 and 2010 and between 2011 and 20166. Patients who have multiple prosthetic joints in place appear to be at increased risk. In a series of 197 patients with a PJI and other joint replacements in situ, 37 (19%) developed a PJI in another prosthetic joint. Careful clinical and radiographic evaluation of the other prosthetic joints is mandatory. Diagnosis remains difficult because of the concomitant rise in serologic markers and the antibiotic treatment of the infected joint7. Surgical site infection protocols have become a priority in hospitals performing arthroplasty. Stakeholders have an interest in minimizing the occurrence of PJI, and, in some cases, hospital reimbursement has been linked to surgical site infection prevention. Therefore, accurate measurement of infection outcomes and incidence is critical. Surgical site infection rates are often underestimated using short follow-up periods when hospital-based administrative data are utilized. In 1 study, if only hospital-based data were used, 158 of 926 infected cases would have been missed8. This stresses the importance of using data from both hospital and outpatient settings to obtain accurate rates. Reimbursement for PJI treatment continues to threaten access. A study comparing work-related values for aseptic compared with septic revisions found that, although septic revision procedures require longer operative time and more burdensome aftercare, reimbursement was lower in the septic cases9. In a similar study, disparities in payment were noted regardless of indication for a revision surgical procedure or case complexity. Revision procedures performed to treat PJI were associated with the highest 90-day costs, yet had similar reimbursement compared with revision procedures performed to treat other indications10. Prevention Several studies elucidated preoperative risk factors for PJI. Patients who underwent arthroplasty and had a preoperative albumin of <3.5 mg/dL had a higher rate of infection within 30 days11. In a cost-effectiveness model, non-selective vitamin D repletion was more valuable than selective testing and repletion in order to prevent PJI12. Evaluating the physical effects of obesity, greater soft-tissue thickness on the anterior and medial aspects of the knee was associated with PJI after TKA13. Morbidly obese patients were also found to be at higher risk for treatment failure after debridement, antibiotics, and implant retention (DAIR)14. Symptomatic, benign prostatic hyperplasia15 and lymphedema16 were found to predispose to PJI after total joint arthroplasty (TJA). Although a 97% survivorship after TKA was noted in patients with previous septic arthritis17, patients with previous PJI of the hip or knee had 1.5-fold higher odds of developing PJI after total hip arthroplasty (THA)17. The use of intravenous antibiotics and antibiotic-loaded bone cement in TJA was extensively studied in the British National Joint Registry. Antibiotic cement showed a modest protective effect (hazard ratio, 0.79) for hips with regard to PJI18. Similarly, U.S. veterans had a lower rate of revision for PJI after TKA when antibiotic cement was used (1.9% compared with 2.6%)19. In the Kaiser Permanente Total Joint Replacement Registry, plain cement was noninferior to antibiotic cement in TKA overall, but antibiotic cement provided protection against PJI in patients with diabetes20. Dual antibiotic cement was superior to cement with a single antibiotic in PJI prevention21, although patients receiving intravenous antibiotics other than cefazolin preoperatively had higher PJI risk22. In studying bacterial skin colonization, a high prevalence of positive cultures was seen in hip skin biopsies prior to THA. A predominance of positive culture specimens were from the anterior hip region and grew Cutibacterium on culture23. Staphylococcal screening and eradication, although beneficial in preventing Staphylococcus aureus infection, had no effect on the overall PJI risk after TJA24. Conflicting findings have been reported for povidone-iodine lavage. In a prospective randomized controlled trial, a reduction in PJI within 90 days after TJA was shown25. However, in a systematic review and meta-analysis comprising 31,213 TJAs, there was no difference in infection rate between patients who did and did not receive povidone-iodine lavage26. Another study showed no difference between chlorhexidine and povidone iodine with respect to PJI within 1 year after TJA27. A combination of powdered vancomycin and povidone iodine in the wound after TKA reduced the PJI rate by 27.8% compared with controls28. Two separate studies on the operating environment demonstrated the value of iodine in reducing bacterial contamination in both sterile water basins29 and impervious drapes30. Intraoperatively, the use of tranexamic acid reduced the rate of PJI after primary TJA31 and revision TJA32, and Foley catheters increased the odds of developing PJI after TKA (odds ratio [OR], 2.6)33. Postoperative factors such as corticosteroid injections34 and arthroscopy35 after TKA were shown to increase PJI risk. Patients who underwent aseptic reoperation within 1 year after THA had a higher risk of later development of PJI36. Increased blood glucose variability was associated with a 31% increase in reinfection rate after a 2-stage exchange37. Diagnosis of PJI Abdelaziz et al. retrospectively validated the 2018 International Consensus Meeting (ICM) minor criteria for PJI (345 cases), with an excellent area under the curve (AUC, 0.90), sensitivity (0.96), and specificity (0.84), in which the performance was greater in hips than in knees, making these criteria useful to rule out PJI38. Furthermore, if synovial fluid samples are unavailable, these criteria have been shown to be more accurate than the Musculoskeletal Infection Society (MSIS) criteria39. A multicenter study determined that the optimal thresholds to diagnose PJI within 90 days of an index arthroplasty were 6,000 cells/µL for a synovial white blood-cell (WBC) count, 39.8 mg/L for serum C-reactive protein (CRP), and 39.5 mm/hr for serum erythrocyte sedimentation rate (ESR). No significant differences in serum values were found when reducing the analysis time from within 90 days to 30 days. However, the cell count threshold increased to >10,000 cells/µL using the definition for 30 days rather than that for 90 days40. Serum Biomarkers Although CRP and ESR are the main screening tests, their thresholds may vary with patient demographic characteristics. ESR levels were higher in older, female, and African-American patients, resulting in more false-positive PJI cases. No differences were found in CRP, which seems more accurate in these subgroups41. D-dimer in revision cases had excellent sensitivity (96%) but low specificity (32%) and poor accuracy (61%) to diagnose PJI42. Qin et al. advocated for CRP and ESR as the best screening tests; however, the combination of D-dimer and CRP had a sensitivity of 98% and a negative predictive value of 96%, suggesting that such combinations might improve PJI diagnostic accuracy43. Bin et al. studied fibrinogen for PJI diagnosis in a small patient cohort and found that it had high specificity (94.6%) with acceptable sensitivity, showing the potential to rule out PJI and confirm infection control before reimplantation44. Moreover, comparisons of fibrinolytic markers (D-dimer, fibrin-degradation product, and fibrinogen) with CRP and ESR found only fibrinogen as a promising marker45-48. Procalcitonin showed poor specificity and sensitivity in a small study, and Busch et al. recommended against its use to diagnose PJI49. Synovial Fluid Biomarkers Alpha-defensin was shown by Deirmengian et al. as the most accurate diagnostic biomarker currently available (90% sensitivity and 95% specificity). In a multicenter, prospective study (n = 305) used to validate a lateral-flow test for the U.S. Food and Drug Administration (FDA)50, the lateral-flow test had no significant differences compared with the enzyme-linked immunosorbent assay (ELISA). Calprotectin, a biomarker used in chronic inflammatory diseases, has shown a sensitivity of 100% and a specificity of 95% with an AUC of 0.97, when compared with 2018 ICM criteria as the standard51. Saline solution lavage may help when dry taps occur during aspiration if the percentage of polymorphonuclear cells (PMNs) remains similar to the 80% cutoff (sensitivity, 75%)52. Cultures Rockov et al. showed that aspiration cultures are more likely to correlate with intraoperative cultures with higher aspiration volumes, with a cutoff of 3.5 mL for typical organisms and 12.5 mL for slow-growing organisms53. Microcalorimetry may be equivalent to culture results as shown by Morgenstern et al.54; however, this test alone still has poor sensitivity (39%) when compared with institutional criteria. Treatment Irrigation and Debridement Machine learning has recently been used to predict outcomes following DAIR for PJI. The most important variables associated with failed treatment were elevated serum CRP levels and the presence of positive blood cultures55. A second DAIR procedure continues to show superior results to single DAIR procedures, with a 74% success rate in a study of 144 cases56. The timing of DAIR for TKA remains important to the success rate. The best results were found in patients who underwent DAIR within 2 weeks of the index arthroplasty, with a 223% increase in reinfection if these procedures were performed after 2 weeks57. Despite these results, the incidence of DAIR has increased over a 10-year period in the United States, with the most substantial increase seen in the first 90 days58. One series of 263 patients who underwent DAIR and had a mean follow-up of 8.3 years showed a lower rate of reinfection with a dual surgical setup (25%) compared with a single setup (62.5%)59. The reported success rates for 2-stage exchange of the hip following a failed irrigation and debridement procedure remain variable. In a series of 114 patients with periprosthetic hip infection who had a mean follow-up of 6.2 years, the failure rate was 43% (21 of 49) in the irrigation and debridement group and 12% (8 of 65) in the direct staged revision group60. Two-Stage Exchange Two-stage exchange is regarded as the gold standard; however, recent studies have called this into question61,62. The exclusion of those patients who never underwent a reimplantation procedure provided a 9% overestimation of success in patients treated with 2-stage exchange61. Another study of 89 patients found a very high attrition rate of those patients undergoing 2-stage exchange for PJI due to a high risk of death (24%) and inability to complete the 2-stage protocol (32%); patients who completed the 2-stage protocol had a reinfection rate of 15% at 4.5 years62. Another study evaluating 2-stage treatment for periprosthetic hip infection found a success rate of 95.7% at a mean follow-up of 67 months63. Although overall success rates remain high with a 2-stage protocol, a study examining all complication rates (medical and surgical) during treatment found that 52% of patients incurred a complication at some point during the staged treatment of the PJI, with an 18% mortality rate at a mean follow-up of 3.7 years64. Three months of oral antibiotics after reimplantation reduced the recurrence rate in patients undergoing a 2-stage exchange. In 185 patients, treatment success was achieved in 87.5% of patients who received 3 months of antibiotics compared with 71.4% in those who did not65. Articulating antibiotic spacers remain the standard for interim treatment of PJI. Articulating spacers using new primary knee components are becoming increasingly popular, with a success rate of 79% following reimplantation at a follow-up of 3.7 years66. A randomized trial of static spacers (n = 32) and articulating spacers (n = 36) for PJI showed significantly higher range of motion and Knee Society score following reimplantation in the articulating spacer group at a mean follow-up of 3.5 years67. One-Stage Exchange Although a prospective, multicenter, randomized study comparing 1-stage exchange with 2-stage exchange is ongoing in the United States, results are not yet available and therefore have not been published. However, 1-stage exchange continues to gain enthusiasm as treatment for PJI. A series of 72 patients undergoing a 1-stage knee exchange for PJI with a hinged arthroplasty system and tantalum cones were found to have an infection-free survival of 89% and survival free of any revision of 83% at a mean follow-up of 4 years68. Antimicrobial Treatment Vancomycin Powder Intrawound vancomycin powder did not reduce the risk of surgical site infection in patients with acetabular fractures who underwent open reduction and internal fixation (ORIF)69. A retrospective study that evaluated patients who underwent TKA demonstrated a significant decrease in PJI in patients who had powdered vancomycin applied prior to wound closure70. Single-center retrospective studies reviewed patients who had undergone TJA and found that those who received topical vancomycin were less likely to develop PJI, although the difference in patients who underwent THA was not significant in 1 of the studies and baseline PJI rates were low71. A systematic review found that all published studies on intrasite antibiotic powder to prevent surgical site infection were retrospective; some small studies demonstrated a clinical benefit, but no recommendations could be made because of a lack of high-quality evidence72. Oral Antibiotics in Revision Arthroplasty A multicenter, randomized controlled trial demonstrated that a 3-month course of oral antibiotics following 2-stage revision reduced the rate of failure due to infection65. The antibiotic regimens were heterogeneous, antibiotic-related adverse events occurred, and it was noted that the majority of infections in both the treatment and control groups were caused by microorganisms that differed from the original infection. Although it is still undetermined which antibiotics should be utilized and if a duration of 3 months of antibiotics is necessary, this trial suggested that there is a potential benefit of oral antibiotics following a 2-stage revision surgical procedure. Receipt of an extended course of oral antibiotics after DAIR independently predicted treatment success in 1 study; however, extending oral antibiotic therapy beyond 1 year did not improve outcomes, suggesting that the benefit of chronic suppressive antibiotic therapy may plateau73. A retrospective cohort study evaluating oral antibiotics as a chronic suppressive therapy demonstrated the development of resistance (23.1%) and adverse events necessitating suspension of antibiotics (5.6%). Failure of chronic suppressive therapy was associated with patient age of <70 years, non-gram-positive cocci etiology, and an upper-limb prosthesis74. A retrospective study with a 3.2-year follow-up concluded that chronic suppressive therapy with oral antibiotics is a viable option in patients in whom a further surgical procedure is contraindicated or declined75. The utility of adjunct rifampin in staphylococcal PJI was evaluated in a randomized controlled trial, and no significant advantage of adding rifampin to standard antibiotic therapy was observed, calling into question the need to add rifampin to treat implant-related infections76. Antibiotic Stewardship A systemic review of the available literature found no direct evidence that prophylactic antibiotics should be given prior to dental procedures in patients with TJA, but suggested that unusual situations may call for assessment on a case-by-case basis77. A comprehensive review of the current guidelines and available evidence describes a method to engage both dentists and orthopaedic surgeons in dental stewardship78. A review highlighting antibiotic stewardship strategies in TJA is also available79. New Therapies A bacteriophage-derived lysin showed antimicrobial potential in vitro and in a murine model80. A case report81 and a case series82 detailing the successful use of bacteriophage therapy in patients with Staphylococcus aureus PJI further introduced bacteriophage therapy as a potential salvage treatment option. Other Topics Spine A systematic review and meta-analysis of 27 studies found a surgical site infection incidence of 3.1%, with a higher incidence observed in instrumented surgical procedures and a posterior approach83. Given the associated cost and morbidity of surgical site infections, prevention and identification of at-risk patients are critical. Another systematic review and meta-analysis suggested no additional benefit to prolonged postoperative antimicrobial prophylaxis after spine surgery84. A multicenter study observed that a surgical procedure performed at an academic center and an American Society of Anesthesiologists (ASA) classification of ≥3 were risk factors for surgical site infection after instrumented fusion, possibly reflecting case complexity and cumulative comorbidities85. Urinary tract infection was associated with a higher rate of infection, in the American College of Surgeons National Surgical Quality Improvement Program data set, even after adjustment for confounding factors86. When spinal implant infection is suspected, accurate diagnosis is critical to management. Histopathology was not consistently helpful, with a sensitivity of only 51% for spinal implant infection87. Improved sensitivity was noted for Staphylococcus and other virulent organisms, with poor sensitivity for Cutibacterium acnes and coagulase-negative staphylococci; in contrast, the sonication of removed spinal implants is a sensitive and specific technique for the diagnosis of infection88. A diagnostic threshold of 20 colony-forming units (CFU)/10 mL yielded an optimal sensitivity and specificity. Trauma A multicenter, randomized controlled trial found no difference in deep surgical site infection between negative pressure wound therapy (NPWT) compared with a standard wound dressing following a surgical procedure for a lower-extremity fracture89. However, this study excluded patients with open fractures that could not be closed at the first operative debridement, those most likely to develop infection. Another multicenter randomized trial was performed among patients with Gustilo-Anderson (GA) classification GA-II to III-C open tibial fractures90. In patients who underwent a delayed primary closure, there was no significant difference in the level of disability, deep infection, or nonunion with NPWT. There was a numerically higher rate of deep infection at 30 days in the conventional dressing group (10.1%) compared with the NPWT group (7%); however, the study was not powered to evaluate infection as an outcome. These conflicting findings, in addition to an additional systematic review and meta-analysis that observed a decrease in deep infection when NPWT was used after open fracture91, suggest that the findings must be taken in the context of the overall literature when applied to those individuals with open fractures. The optimal use of antimicrobial prophylaxis to prevent infection after an open fracture remains debated. One retrospective study involving a relatively similar number of GA-I, GA-II, and GA-III fractures demonstrated no benefit of prophylaxis beyond 72 hours92. The choice of specific antimicrobials remains under debate. Historically, a combination of an aminoglycoside and cefazolin has been used for GA-III open fractures. A retrospective, single-center cohort study of patients with GA-III open fractures observed a lower rate of surgical site infection at 30 days in the group receiving piperacillin-tazobactam, compared with the group receiving cefazolin with tobramycin93. Evidence-Based Orthopaedics The editorial staff of JBJS reviewed a large number of recently published studies related to the musculoskeletal system that received a higher Level of Evidence grade. In addition to articles cited already in this update, 6 other articles relevant to musculoskeletal infection surgery are appended to this review after the standard bibliography, with a brief commentary about each article to help guide your further reading, in an evidence-based fashion, in this subspecialty area. Evidence-Based Orthopaedics Amanatullah DF, Cheng RZ, Huddleston III JI, Maloney WJ, Finlay AK, Kappagoda S, Suh GA, Goodman SB. The routine use of synovial alpha-defensin is not necessary. Bone Joint J. 2020 May;102-B(5):593-9. In a review of 158 patients, the routine use of alpha defensin testing did not substantially change the ability of orthopaedists or infectious disease consultants to diagnose or rule out PJI in most cases. However, when the diagnosis was in doubt after traditional testing, alpha defensin was useful. Of 27 undecided cases with traditional testing, 24 (89%) benefited from the addition of testing with alpha defensin. It appears that the traditional PJI workup of ESR, CRP, aspirated cell counts with the percentage of PMNs, and culture remain the mainstay of diagnosis, with alpha defensin used in difficult-to-diagnose cases. Belay ES, Danilkowicz R, Bullock G, Wall K, Garrigues GE. Single-stage versus two-stage revision for shoulder periprosthetic joint infection: a systematic review and meta-analysis. J Shoulder Elbow Surg. 2020 Dec;29(12):2476-86. Epub 2020 Jun 18. Single-stage revision had a recurrence rate of 6.3% compared with a 10.1% recurrence rate for 2-stage revision. However, all of the studies in this review were nonrandomized and retrospective. Selection bias was admitted because there was a higher propensity for treatment of virulent and drug-resistant organisms in the 2-stage group. Such a selection bias could be mitigated by a prospective randomized trial in the future. Deirmengian CA, Kazarian GS, Feeley SP, Sizer SC. False-positive automated synovial fluid white blood cell counting is a concern for both hip and knee arthroplasty aspirates. J Arthroplasty. 2020 Jun;35(6S):S304-7. Epub 2020 Jan 28. The correlation between automated and manual cell counts in a series of >47,000 synovial fluid samples (44,824 from arthroplasties and 2,242 from native knees) found that the false-positive rate of automated counts when compared with manual counts was 4.4% for native knees, 10.9% for TKAs, and 34.3% for THAs. Although the false-positive issue has been recognized in aspirates of metal-on-metal or corrosion cases, this is the first study demonstrating that automated cell counts may overestimate synovial WBC counts due to arthroplasty-related debris. The authors’ recommendation to verify elevated WBC counts in synovial fluid with a manual cell count may have merit. Gutman MJ, Stone MA, Namdari S, Abboud JA. Treatment of elbow periprosthetic joint infection: a systematic review of clinical outcomes. J Shoulder Elbow Surg. 2020 Feb;29(2):411-9. In this review of 309 patients in nonrandomized retrospective studies, staphylococci were the most common infecting organism. The success rate for irrigation and debridement and component retention was 56%, and the success rate for a 2-stage exchange was 81%, similar to the data from PJI of the hip and knee. As only 8 cases were treated in a 1-stage exchange, the authors could not comment on this treatment. Kwon YM, Mahajan J, Tirumala V, Oganesyan R, Yeo I, Klemt C. Sensitivity and specificity of serum and synovial fluid markers in diagnosis of infection in head-neck taper corrosion of metal-on-polyethylene total hip arthroplasty. J Arthroplasty. 2020 Dec;35(12):3737-42. Epub 2020 Jun 23. This study points out the difficulty in diagnosing concomitant adverse local tissue reactions and PJI in cases of head and neck corrosion. In a series of 89 patients with taper corrosion, infected and noninfected cases had similar serum ESR (35 compared with 31 mm/hr), but there were significantly higher CRP values in the infected cases (48 compared with 32 mg/L). Synovial WBC counts and PMN percentages were significantly higher in the infected group: 60,380/µL (89% PMNs) compared with 1,194/µL (32% PMNs). These results must be tempered by the fact that only 11 of 89 patients were infected. Sigmund IK, Dudareva M, Watts D, Morgenstern M, Athanasou NA, McNally MA. Limited diagnostic value of serum inflammatory biomarkers in the diagnosis of fracture-related infections. Bone Joint J. 2020 Jul;102-B(7):904-11. In this study, 106 patients were retrospectively evaluated for suspected septic nonunion of fractures. CRP and the WBC count with differential were analyzed. Cases were classified as septic or aseptic based on the Fracture-Related Infection Consensus document. Forty-six patients (43%) were truly infected. Although CRP performed the best in this study, it is not sufficiently accurate to diagnose a septic nonunion, especially when caused by low-virulence organisms. Histopathology and microbiology were more accurate in this study. Hence, the diagnosis of fracture-related infection remains somewhat elusive.