Orthopedics

Feature Article 

Classic Markers for Infection Perform Poorly in Predicting Residual Infection Prior to Reimplantation

Amir Herman, PhD, MD; Anthony Albers, MD; Donald S. Garbuz, MD, MPH; Clive P. Duncan, MD, MPH; Bassam A. Masri, MD

Abstract

Two-stage exchange arthroplasty remains the treatment of choice for chronic periprosthetic joint infections. This retrospective study conducted between 2009 and 2015 examined the diagnostic value of biomarkers for residual infection between stages. The biomarkers evaluated included C-reactive protein prior to reimplantation, preimplantation synovial fluid white blood cell count and percent neutrophils, and the intraoperative histologic synovial white blood cell count per high-power field (×400) on permanent sections. Residual infection was defined as either positive cultures (more than 1) at second stage, any further surgery (eg, amputation, arthrodesis, or another 2-stage revision), or the need for infection suppression with antibiotics. Sensitivity, specificity, positive and negative predictive values, and likelihood ratios were calculated accordingly. A total of 182 two-stage exchange operations that included 109 (59.9%) prosthetic hips and 73 (40.1%) prosthetic knees met the inclusion criteria. Residual infection was present in 38 (20.9%) of the procedures. The area under the curve–receiver operating characteristic values were 0.677 for C-reactive protein (P=.002), 0.506 for aspiration white blood cell count (P=.944), 0.623 for aspiration percent neutrophils (P=.200), and 0.524 for white blood cell count per high-power field (P=.801). Positive and negative predictive values were poor and ranged between 26% and 57% and 78% and 85%, respectively. Analyses using specific combinations of biomarkers did not significantly improve predictive values. This study showed that classic markers perform poorly in identifying residual infection prior to second-stage revision. Further research is necessary to evaluate the diagnostic utility of other, more recently introduced biomarkers to determine whether infection has been eradicated between stages. [Orthopedics. 2019; 42(1):34–40.]

Abstract

Two-stage exchange arthroplasty remains the treatment of choice for chronic periprosthetic joint infections. This retrospective study conducted between 2009 and 2015 examined the diagnostic value of biomarkers for residual infection between stages. The biomarkers evaluated included C-reactive protein prior to reimplantation, preimplantation synovial fluid white blood cell count and percent neutrophils, and the intraoperative histologic synovial white blood cell count per high-power field (×400) on permanent sections. Residual infection was defined as either positive cultures (more than 1) at second stage, any further surgery (eg, amputation, arthrodesis, or another 2-stage revision), or the need for infection suppression with antibiotics. Sensitivity, specificity, positive and negative predictive values, and likelihood ratios were calculated accordingly. A total of 182 two-stage exchange operations that included 109 (59.9%) prosthetic hips and 73 (40.1%) prosthetic knees met the inclusion criteria. Residual infection was present in 38 (20.9%) of the procedures. The area under the curve–receiver operating characteristic values were 0.677 for C-reactive protein (P=.002), 0.506 for aspiration white blood cell count (P=.944), 0.623 for aspiration percent neutrophils (P=.200), and 0.524 for white blood cell count per high-power field (P=.801). Positive and negative predictive values were poor and ranged between 26% and 57% and 78% and 85%, respectively. Analyses using specific combinations of biomarkers did not significantly improve predictive values. This study showed that classic markers perform poorly in identifying residual infection prior to second-stage revision. Further research is necessary to evaluate the diagnostic utility of other, more recently introduced biomarkers to determine whether infection has been eradicated between stages. [Orthopedics. 2019; 42(1):34–40.]

Infection after joint arthroplasty is a devastating complication, having an incidence of 0.5% to 1% after primary joint replacement.1,2 That incidence is expected to increase in the next several years due to the high prevalence of obesity and diabetes making the population more prone to infection.3 The infection burden, with all of its potential negative consequences, is expected to rise because of the increasing numbers of joint replacements.

Tsukayama et al4 classified periprosthetic joint infection as chronic, acute, hematogenous, or indolent (positive intraoperative cultures). The surgical and medical treatment of periprosthetic joint infection is determined by the type of infection.5 Early acute postoperative and acute hematogenous infections may be treated with debridement, irrigation, and implant retention with variable degrees of success, whereas chronic infections require implant removal due to the development of a biofilm, which prevents eradication of bacterial infections.5

Recent work was also done to establish consensus criteria for periprosthetic joint infections.6–10 These criteria are based on clinical and laboratory markers for the diagnosis of infection. They require either a sinus tract or 2 positive cultures as major criteria or 3 of 5 minor criteria. The minor criteria are based on infection markers such as C-reactive protein (CRP) or erythrocyte sedimentation rate (ESR), white blood cell (WBC) count and percent neutrophils in joint fluid aspiration, and WBC count per high-power field microscopy (×400) on frozen sections of synovial tissue. The threshold values depend on the type of infection—acute or chronic—and are well established in the literature.11–14

Chronic periprosthetic joint infection can be treated as either a single-stage or, more commonly, a 2-stage revision. Two-stage revision success rates have been reported to range from 72% to 96%, depending on both patient and bacterial characteristics and on duration of follow-up.15–23 To minimize the risk of ongoing or recurrent infection, verifying eradication of infection after implant removal (first stage) and antibiotic-cement spacer implantation is of paramount importance prior to proceeding with the second stage. If infection between stages is deemed to persist, another irrigation and debridement (with another antibiotic-cement spacer implantation) is required. If the patient is deemed to be infection free between stages, the surgeon may proceed to reimplantation of a definitive prosthesis.

Several authors have tried to determine the values of biomarkers for infection prior to reimplantation.23–33 However, these studies were limited by small samples or by focusing on only a subset of the commonly used markers.

The goal of this study was to evaluate the ability of CRP before reimplantation, synovial WBC count and percent neutrophils on preoperative joint aspiration, and WBC count in synovial tissue at the time of second stage to predict infection eradication and ultimately long-term success in infection eradication in 2-stage exchange arthroplasty.

Materials and Methods

After institutional review board approval was received, the registry of the adult reconstruction unit at the authors' center was used to identify patients who had 2-stage revision for infection between January 1, 2009, and December 31, 2015. The authors then retrospectively reviewed the patients' electronic medical records, including outpatient clinic follow-up, blood tests, surgery reports, emergency department visits, and subsequent hospitalizations. To maximize follow-up, the authors used a provincial database to determine if treatment was provided at another center in their province. This was done to maximize the ability to detect complications and to minimize bias due to inadequate follow-up. This database contains data on any blood work, admission, or surgical intervention at hospitals within British Columbia.

Patients older than 18 years who had a 2-stage revision for an infected hip or knee replacement were included. Primary infection was defined by the joint periprosthetic infection committee criteria.6–9 These include 1 of 2 major criteria (sinus tract communicating with the prosthesis or 2 separate positive cultures) and 3 of 5 minor criteria (elevated ESR or CRP, elevated synovial WBC count, elevated synovial percent neutrophils, more than 5 WBCs per high-power field in histology, or a single positive periprosthetic culture).

Patients who did not have a second-stage reimplantation and patients who had arthroplasty due to an oncologic diagnosis were excluded. For patients who had several 2-stage revisions, only the first one was included. Persistent or recurrent infection after the 2-stage procedures was determined according to the joint periprosthetic infection committee criteria and was established by the senior surgeons (D.S.G., C.P.D., B.A.M.) after reviewing the patients' electronic and hardcopy records. A patient who required suppressive antibiotics was also considered chronically infected, even if the infection was successfully suppressed and the patient was asymptomatic, as this was a failure of the initial 2-stage protocol. A total of 212 reimplantation procedures were initially retrieved; 30 of these patients were excluded, leaving 182 patients in the study.

At the authors' institution, after removal of the implants (first stage), an articulating cement spacer is implanted and patients receive 6 weeks of antibiotic treatment under the direction of the infectious disease service. Reimplantation (second stage) is usually performed at 3 to 5 months after the first stage to allow for an antibiotic-free interval and for monitoring with ESR and CRP. Five cultures were obtained at reimplantation, after antibiotics were withheld. One positive culture was considered a contamination, and 2 positive cultures were considered true infection. At the authors' institution, ESR is not covered by health insurance if a CRP is ordered at the same time. Therefore, CRP values were more predictably present in the medical records and were reviewed instead of ESR.

Data collected included demographic data: age at surgery, sex, site (hip or knee), body mass index, and American Society of Anesthesiologists score. C-reactive protein, synovial WBC count and percent neutrophils on joint aspiration, and the histologic count of WBCs per high-power field microscopy (×400) in the synovial biopsies were checked prior to reimplantation. An experienced musculoskeletal pathologist reviewed the histopathology in all cases. Further, CRP values were collected at 1 month (during antibiotic treatment) and 2, 3, and 4 months after the first stage.

Statistical analysis was performed by an experienced biostatistician (A.H.) using SPSS version 23.0 software for Windows (IBM Corp, Armonk, New York). Categorical data were compared using the chi-square test. Continuous data were compared using the Wilcoxon–Mann– Whitney rank sum test.

For each biomarker, the arthroplasty site (hip or knee) was examined for difference in values both in infected and in noninfected cases. That was done by fitting an analysis of variance model for each marker (as dependent variable) with site and infection outcome as primary covariates and interaction term as independent covariates.

For each marker for infection, the receiver operating characteristic (ROC) curve was fitted and the area under the curve (AUC) was calculated for each curve. The Wilcoxon–Mann–Whitney rank sum test was used to test the hypothesis that the AUC=1/2.

The AUC of the ROC can be interpreted as the probability of any single observation in one group being higher than any observation in another group. Formally, this is written as Prob(X>Y), where, in this case, X is a marker in infected patients and Y is the same marker in noninfected patients. Then the null hypothesis tested is that the probability of one being higher than the other is equal to a flip of a coin; Prob(X>Y)=1/2. This is tested by the standard Wilcoxon– Mann–Whitney rank sum test.

Youden's index for each marker is defined as sensitivity+specificity-1 for each of the marker's values. The value that maximizes Youden's index was used to define an optimal cutoff value for each marker. In simple terms, this indicates the value at the top left corner of the ROC's curve.

After establishing the optimal cutoff value for each marker, sensitivity, specificity, positive and negative predictive values, and positive and negative likelihood ratios were calculated for each marker. Any 2-marker combination was also tested for improving the predictive value for residual infection.

Post hoc power analysis showed 86% power for comparing 144 and 38 cases to detect a statistically significant difference between 10 and 27 mg/L (CRP values for noninfected and infected cases, respectively), with standard deviation of 30 and type 1 error of 5%.

Results

The study included 182 patients who were eligible according to the inclusion criteria. Of these, 107 (58.8%) were male and 75 (41.2%) were female. Mean age was 66.6 years (SD, 11.1 years; range, 31–90 years). Mean body mass index was 30.25 kg/m2 (SD, 6.99 kg/m2; range, 18.3–46.6 kg/m2). The study population included 73 (40.1%) knee and 109 (59.9%) hip replacements revised for infection (Table 1).

Demographic and Clinical Data

Table 1:

Demographic and Clinical Data

There were 144 (79.1%) patients in whom 2-stage revision due to infection resulted in infection-free arthroplasty. In 38 (20.9%) patients, persistent infection was diagnosed after the second-stage operation. Of these patients, 8 (21.1%) had additional irrigation and debridement, 10 (26.3%) were receiving antibiotic suppression, 4 (10.5%) had an amputation, 11 (28.9%) had a new 2-stage revision, 3 (8.0%) had a resection arthroplasty, 1 (2.6%) had an arthrodesis, and 1 (2.6%) had multiple plastic surgery procedures for wound closure.

Markers prior to reimplantation included CRP, WBC count, and percent neutrophils in aspiration. During reimplantation, histologic count of WBCs per high-power field microscopy (×400) in the final (not frozen) sections was also included. There was no statistically significant difference in the markers' values in the infected and the noninfected patients between arthroplasty sites (hip or knee). Of all of the markers examined, only the CRP had a statistically significant higher value for residual infection prior to reimplantation (P=.002). However, for CRP, the AUC of the ROC curve was 0.677, indicating that this marker had poor accuracy in diagnosing residual infection prior to reimplantation (Table 2).

Values of Diagnostic Parameters

Table 2:

Values of Diagnostic Parameters

Cultures during the second stage were also highly unreliable. Of 37 patients who were infected after the second stage and had intraoperative culture results, 26 (70.3%) patients had negative cultures. Only 11 patients (29.7%) of the aforementioned 37 had positive intraoperative cultures at the time of reimplantation (Table 2). Although negative cultures predicted lack of infection (P=.001), the AUC was 0.634, indicating poor performance. This is highlighted by the fact that 70.3% of ultimate failures had negative cultures at the time of the second-stage procedure.

Using optimal cutoff values as determined by Youden's index, the positive predictive values for residual infection ranged between 26.3% and 57.1%, and the negative predictive values ranged between 78.3% and 85.2% (Table 3).

Diagnostic Properties of Infection Markers

Table 3:

Diagnostic Properties of Infection Markers

Combining the results of the infection markers yielded positive predictive values between 50.0% and 80.0% and negative predictive values between 77.3% and 90.3%. The best combination of markers was CRP greater than 15 mg/L and percent neutrophils greater than 80% in aspiration. This combination yielded a positive predictive value of 80.0% and a negative predictive value of 85.4% (Table 4).

Diagnostic Properties When Combining Infection Markers

Table 4:

Diagnostic Properties When Combining Infection Markers

Examinations of monthly CRP values at 1 to 4 months after the first stage had AUC of ROC curves of 0.581, 0.647, 0.591, and 0.603, respectively (Figure 1). Although CRP at 4 months after the first stage was higher in patients with residual infection, its diagnostic value was limited because approximately half of the infected cases had CRP values below the cutoff of 15 mg/L.

Boxplots of C-reactive protein (CRP) values 1, 2, 3, and 4 months after resection arthroplasty prior to reimplantation. The boxes represent first and third quartiles of the values. The middle line represents median value. The whiskers represent highest value up to 1.5 of the interquartile range.

Figure 1:

Boxplots of C-reactive protein (CRP) values 1, 2, 3, and 4 months after resection arthroplasty prior to reimplantation. The boxes represent first and third quartiles of the values. The middle line represents median value. The whiskers represent highest value up to 1.5 of the interquartile range.

Discussion

In this study, the authors showed that classic markers for infection performed poorly for predicting recurrent infection before reimplantation. The markers in this study included CRP, aspiration WBC count and percent neutrophils, and histologic WBC count per high-power field microscopy (×400). For all of the markers examined, the AUC of the ROC curves were between 0.5 and 0.65. Further, using optimal cut-point for continuous markers did not provide tests with high positive and negative predictive values for anticipating residual infection prior to reimplantation. This result is especially important because, prior to reimplantation, the surgeon and the patient must make a decision to either proceed with reimplantation or repeat the first stage, based on the likelihood of persistent infection.

Kusuma et al27 studied 76 patients with infected knee arthroplasty, of whom 7 had persistent infection. They concluded that ESR, CRP, and aspiration WBC count or percent neutrophils performed poorly in predicting residual infection prior to reimplantation. They reported an AUC that ranged between 0.39 and 0.71. Other authors reported similar results.16

Shukla et al26 reported on 87 hips revised for periprosthetic joint infection. They reported 7 hips with persistent infection. The AUC of the markers they studied ranged between 0.55 and 0.91, with the lowest AUC value attributed to the CRP and the highest attributed to aspirated joint fluid WBC counts. The current study did not confirm their finding that WBC counts in the synovial fluid could be reliably used to diagnose residual infection prior to reimplantation.

Janz et al29 studied 69 patients after resection hip arthroplasty prior to reimplantation. They concluded that aspiration WBC counts and CRP had a positive predictive value of 75% and 36%, respectively, and a negative predictive value of 70% and 90%, respectively. They did not report the AUC of the ROC curves. These values do not provide any diagnostic value.

Bori et al24 studied 21 patients prior to reimplantation. They examined WBC count per high-power field microscopy during reimplantation. They concluded that using a WBC count of greater than 5 per high-power field yielded positive and negative predictive values of 100% and 73.6%, respectively. However, their sample was small. They did not report the AUC of the ROC curves of WBC count, and they did not use any statistical means to choose optimal cutoff values.

Some authors considered the use of sonication of antibiotic spacers to identify residual infection prior to reimplantation.31,34 However, sonication of the antibiotic spacers can only be used for cultures during reimplantation and not before stage 2 for the diagnosis of residual infection. Another option gaining popularity is the use of alpha-defensin-1 to diagnose residual infection.35 It has been shown to be a good diagnostic tool for a wide range of bacteria and even in patients already treated with antibiotics.36,37 Alpha-defensin-1 has shown promising results with close to 100% sensitivity and specificity for identifying periprosthetic joint infection, outperforming all available markers.35 However, only 1 study included 6 patients (of 61) who had alpha-defensin-1 tested before reimplantation.38 In that study, the authors did not analyze these patients separately. Other authors considered serum D-dimer as a marker for residual infection to guide the timing of reimplantation.39

The limitations of the current study were those inherent to every retrospective study—incomplete data due to reliance on available records. No standardized a priori protocol was used for determining infection. It was left to the surgeon to determine whether the joint was infected. Further, the focus of this study was symptomatic infection cases. However, because the surgeon's decision regarding infection was influenced by the markers' values, the bias would be to support the markers' diagnostic value, which is opposite of this article's conclusion.

In addition, because most of the cultures during reimplantation were negative (even when infected), it was difficult to determine whether an infection was a new event or a persistence of the primary infection. This is especially true because some studies of infection recurrence after 2-stage exchange have shown a preponderance of infections with new organisms.20,21 In that situation, it is not clear whether this new infection is a second, totally unrelated infection or if it is an undiagnosed initial organism that was not treated with appropriate antibiotics because the cultures were negative for that organism at the time of the initial operation.

Conclusion

There is an obvious need for better decision making to determine eradication of infection prior to reimplantation. Further research is needed to determine whether new markers such as alpha-defensin-1 perform better in diagnosing residual infection prior to reimplantation. In the meantime, clinical judgment with judicious but cautious use of biomarkers will remain the mainstay of management.

References

  1. Cochrane AR, Ong KL, Lau E, Mont MA, Malkani AL. Risk of reinfection after treatment of infected total knee arthroplasty. J Arthroplasty. 2016;31(9)(suppl):156–161. doi:10.1016/j.arth.2016.03.028 [CrossRef]
  2. Jämsen E, Varonen M, Huhtala H, et al. Incidence of prosthetic joint infections after primary knee arthroplasty. J Arthroplasty. 2010;25(1):87–92. doi:10.1016/j.arth.2008.10.013 [CrossRef]
  3. O'Toole P, Maltenfort MG, Chen AF, Parvizi J. Projected increase in periprosthetic joint infections secondary to rise in diabetes and obesity. J Arthroplasty. 2016;31(1):7–10. doi:10.1016/j.arth.2015.07.034 [CrossRef]
  4. 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(4):512–523. doi:10.2106/00004623-199604000-00005 [CrossRef]
  5. Tsukayama DT, Goldberg VM, Kyle R. Diagnosis and management of infection after total knee arthroplasty. J Bone Joint Surg Am. 2003;85(suppl 1):S75–S80. doi:10.2106/00004623-200300001-00014 [CrossRef]
  6. Parvizi J, Zmistowski B, Berbari EF, et al. New definition for periprosthetic joint infection: from the Workgroup of the Musculoskeletal Infection Society. Clin Orthop Relat Res. 2011;469(11):2992–2994. doi:10.1007/s11999-011-2102-9 [CrossRef]
  7. Workgroup Convened by the Musculoskeletal Infection Society. New definition for periprosthetic joint infection. J Arthroplasty. 2011;26(8):1136–1138. doi:10.1016/j.arth.2011.09.026 [CrossRef]
  8. Springer BD. The diagnosis of periprosthetic joint infection. J Arthroplasty. 2015;30(6):908–911. doi:10.1016/j.arth.2015.03.042 [CrossRef]
  9. Enayatollahi MA, Parvizi J. Diagnosis of infected total hip arthroplasty. Hip Int. 2015;25(4):294–300. doi:10.5301/hipint.5000266 [CrossRef]
  10. Parvizi J, Tan TL, Goswami K, et al. The 2018 definition of periprosthetic hip and knee infection: an evidence-based and validated criteria. J Arthroplasty. 2018;33(5):1309–1314. doi:10.1016/j.arth.2018.02.078 [CrossRef]
  11. Bedair H, Ting N, Jacovides C, et al. The Mark Coventry Award. Diagnosis of early postoperative TKA infection using synovial fluid analysis. Clin Orthop Relat Res. 2011;469(1):34–40. doi:10.1007/s11999-010-1433-2 [CrossRef]
  12. Greidanus NV, Masri BA, Garbuz DS, et al. Use of erythrocyte sedimentation rate and C-reactive protein level to diagnose infection before revision total knee arthroplasty: a prospective evaluation. J Bone Joint Surg Am. 2007;89(7):1409–1416.
  13. Schinsky MF, Della Valle CJ, Sporer SM, Paprosky WG. Perioperative testing for joint infection in patients undergoing revision total hip arthroplasty. J Bone Joint Surg Am. 2008;90(9):1869–1875. doi:10.2106/JBJS.G.01255 [CrossRef]
  14. Spangehl MJ, Masri BA, O'Connell JX, Duncan CP. Prospective analysis of preoperative and intraoperative investigations for the diagnosis of infection at the sites of two hundred and two revision total hip arthroplasties. J Bone Joint Surg Am. 1999;81(5):672–683. doi:10.2106/00004623-199905000-00008 [CrossRef]
  15. Brimmo O, Ramanathan D, Schiltz NK, Pillai AL, Klika AK, Barsoum WK. Irrigation and debridement before a 2-stage revision total knee arthroplasty does not increase risk of failure. J Arthroplasty. 2016;31(2):461–464. doi:10.1016/j.arth.2015.08.044 [CrossRef]
  16. Frangiamore SJ, Siqueira MB, Saleh A, Daly T, Higuera CA, Barsoum WK. Synovial cytokines and the MSIS criteria are not useful for determining infection resolution after periprosthetic joint infection explantation. Clin Orthop Relat Res. 2016;474(7):1630–1639. doi:10.1007/s11999-016-4710-x [CrossRef]
  17. Masri BA, Panagiotopoulos KP, Greidanus NV, Garbuz DS, Duncan CP. Cementless two-stage exchange arthroplasty for infection after total hip arthroplasty. J Arthroplasty. 2007;22(1):72–78. doi:10.1016/j.arth.2006.02.156 [CrossRef]
  18. Haddad FS, Muirhead-Allwood SK, Manktelow AR, Bacarese-Hamilton I. Two-stage uncemented revision hip arthroplasty for infection. J Bone Joint Surg Br. 2000;82(5):689–694. doi:10.1302/0301-620X.82B5.9668 [CrossRef]
  19. Ibrahim MS, Raja S, Khan MA, Haddad FS. A multidisciplinary team approach to two-stage revision for the infected hip replacement: a minimum five-year follow-up study. Bone Joint J. 2014;96-B(10):1312–1318. doi:10.1302/0301-620X.96B10.32875 [CrossRef]
  20. Gooding CR, Masri BA, Duncan CP, Greidanus NV, Garbuz DS. Durable infection control and function with the PROSTALAC spacer in two-stage revision for infected knee arthroplasty. Clin Orthop Relat Res. 2011;469(4):985–993. doi:10.1007/s11999-010-1579-y [CrossRef]
  21. Leung F, Richards CJ, Garbuz DS, Masri BA, Duncan CP. Two-stage total hip arthroplasty: how often does it control methicillin-resistant infection?Clin Orthop Relat Res.2011;469(4):1009–1015. doi:10.1007/s11999-010-1725-6 [CrossRef]
  22. Gomez MM, Tan TL, Manrique J, Deimengian GK, Parvizi J. The fate of spacers in the treatment of periprosthetic joint infection. J Bone Joint Surg Am. 2015;97(18):1495–1502. doi:10.2106/JBJS.N.00958 [CrossRef]
  23. Biring GS, Kostamo T, Garbuz DS, Masri BA, Duncan CP. Two-stage revision arthroplasty of the hip for infection using an interim articulating Prostalac hip spacer: a 10- to 15-year follow-up study. J Bone Joint Surg Br. 2009;91(11):1431–1437. doi:10.1302/0301-620X.91B11.22026 [CrossRef]
  24. Bori G, Soriano A, García S, Mallofré C, Riba J, Mensa J. Usefulness of histological analysis for predicting the presence of microorganisms at the time of reimplantation after hip resection arthroplasty for the treatment of infection. J Bone Joint Surg Am. 2007;89(6):1232–1237. doi:10.2106/00004623-200706000-00011 [CrossRef]
  25. George J, Kwiecien G, Klika AK, et al. Are frozen sections and MSIS criteria reliable at the time of reimplantation of two-stage revision arthroplasty?Clin Orthop Relat Res.2016;474(7):1619–1626. doi:10.1007/s11999-015-4673-3 [CrossRef]
  26. Shukla SK, Ward JP, Jacofsky MC, Sporer SM, Paprosky WG, Della Valle CJ. Perioperative testing for persistent sepsis following resection arthroplasty of the hip for periprosthetic infection. J Arthroplasty. 2010;25(6)(suppl):87–91. doi:10.1016/j.arth.2010.05.006 [CrossRef]
  27. Kusuma SK, Ward J, Jacofsky M, Sporer SM, Della Valle CJ. What is the role of serological testing between stages of two stage reconstruction of the infected prosthetic knee?Clin Orthop Relat Res.2011;469(4):1002–1008. doi:10.1007/s11999-010-1619-7 [CrossRef]
  28. Della Valle CJ, Bogner E, Desai P, et al. Analysis of frozen sections of intraoperative specimens obtained at the time of reoperation after hip or knee resection arthroplasty for the treatment of infection. J Bone Joint Surg Am. 1999;81(5):684–689. doi:10.2106/00004623-199905000-00009 [CrossRef]
  29. Janz V, Bartek B, Wassilew GI, Stuhlert M, Perka CF, Winkler T. Validation of synovial aspiration in Girdlestone hips for detection of infection persistence in patients undergoing 2-stage revision total hip arthroplasty. J Arthroplasty. 2016;31(3):684–687. doi:10.1016/j.arth.2015.09.053 [CrossRef]
  30. Munemoto M, Inagaki Y, Tanaka Y, Grammatopoulos G, Athanasou NA. Quantification of neutrophil polymorphs in infected and non-infected second-stage revision hip arthroplasties. Hip Int. 2016;26(4):327–330. doi:10.5301/hipint.5000365 [CrossRef]
  31. Mariconda M, Ascione T, Balato G, et al. Sonication of antibiotic-loaded cement spacers in a two-stage revision protocol for infected joint arthroplasty. BMC Musculoskelet Disord. 2013;14:193. doi:10.1186/1471-2474-14-193 [CrossRef]
  32. Lindsay CP, Olcott CW, Del Gaizo DJ. ESR and CRP are useful between stages of 2-stage revision for periprosthetic joint infection. Arthroplast Today. 2017;3(3):183–186. doi:10.1016/j.artd.2016.08.002 [CrossRef]
  33. Stambough JB, Curtin B, Odum SM, Cross MB, Martin JR, Fehring TK. Does change in ESR and CRP guide the timing of two-stage arthroplasty reimplantation? [published online ahead of print April 23, 2018] Clin Orthop Relat Res. doi:10.1097/01.blo.0000533618.31937.45 [CrossRef].
  34. Nelson CL, Jones RB, Wingert NC, Foltzer M, Bowen TR. Sonication of antibiotic spacers predicts failure during two-stage revision for prosthetic knee and hip infections. Clin Orthop Relat Res. 2014;472(7):2208–2214. doi:10.1007/s11999-014-3571-4 [CrossRef]
  35. Wyatt MC, Beswick AD, Kunutsor SK, Wilson MJ, Whitehouse MR, Blom AW. The alpha-defensin immunoassay and leukocyte esterase colorimetric strip test for the diagnosis of periprosthetic infection: a systematic review and meta-analysis. J Bone Joint Surg Am. 2016;98(12):992–1000. doi:10.2106/JBJS.15.01142 [CrossRef]
  36. Deirmengian C, Kardos K, Kilmartin P, Gulati S, Citrano P, Booth RE. The alpha-defensin test for periprosthetic joint infection responds to a wide spectrum of organisms. Clin Orthop Relat Res. 2015;473(7):2229–2235. doi:10.1007/s11999-015-4152-x [CrossRef]
  37. Shahi A, Parvizi J, Kazarian GS, et al. The alpha-defensin test for periprosthetic joint infections is not affected by prior antibiotic administration. Clin Orthop Relat Res. 2016;474(7):1610–1615. doi:10.1007/s11999-016-4726-2 [CrossRef]
  38. Bingham J, Clarke H, Spangehl M, Schwartz A, Beauchamp C, Goldberg B. The alpha defensin-1 biomarker assay can be used to evaluate the potentially infected total joint arthroplasty. Clin Orthop Relat Res. 2014;472(12):4006–4009. doi:10.1007/s11999-014-3900-7 [CrossRef]
  39. Shahi A, Kheir MM, Tarabichi M, Hosseinzadeh HRS, Tan TL, Parvizi J. Serum D-dimer test is promising for the diagnosis of periprosthetic joint infection and timing of reimplantation. J Bone Joint Surg Am. 2017;99(17):1419–1427. doi:10.2106/JBJS.16.01395 [CrossRef]

Demographic and Clinical Data

CharacteristicTotal (N=182)No Residual Infection (N=144, 79.1%)Residual Infection (N=38, 20.9%)P
Sex, No.
  Male107 (58.8%)84 (58.3%)23 (60.5%)
  Female75 (41.2%)60 (41.7%)15 (39.5%).807
Age, mean±SD, y66.6±11.167.5±10.563.4±12.8.106
Site, No.
  Knee73 (40.1%)52 (36.1%)21 (55.3%)
  Hip109 (59.9%)92 (63.9%)17 (44.7%).032
Side, No.
  Left90 (49.5%)69 (47.9%)21 (55.3%)
  Right92 (50.5%)75 (52.1%)17 (44.7%).533
Body mass index, mean±SD, kg/m230.25±6.9929.62±6.0932.91±9.66.154
American Society of Anesthesiologists score, mean±SD2.45±0.582.39±0.562.66±0.58.018

Values of Diagnostic Parameters

ParameterTotalNo Residual InfectionResidual InfectionAUC of ROCP
CRP prior to reimplantation, mean±SD, mg/L14.07±32.0110.51±22.1527.79±53.950.677.002
WBC count in aspiration, mean±SD, cells/L3572.00±14,381.52950.53±13,768.25303.10±16,395.60.506.944
Percent neutrophils in aspiration, mean±SD48.33%±31.46%45.76%±31.23%56.91%±32.04%0.623.200
Percent neutrophils in histology high-power field at reimplantation, No.
  <5136 (87.7%)110 (80.9%)26 (19.1%)
  5–107 (4.5%)5 (71.4%)2 (28.6%)
  11–3011 (7.1%)8 (72.7%)3 (27.3%)
  >301 (0.6%)1 (100%)0 (0%)0.524.801
Cultures at reimplantation, No.
  Negative160 (91.4%)134 (83.8%)26 (16.3%)
  Positive15 (8.6%)4 (26.7%)11 (73.3%)0.634.001

Diagnostic Properties of Infection Markers

MarkerSensitivitySpecificityPositive Predictive ValueNegative Predictive ValuePositive Likelihood RatioNegative Likelihood Ratio
C-reactive protein >15 mg/L43.8%84.6%42.4%85.2%2.830.67
Aspiration white blood cells >1100 cells/L28.6%92.3%57.1%78.3%3.710.77
Aspiration percent neutrophils >80%41.7%85.0%45.5%82.9%2.780.69
White blood cells per high-power field ≥516.1%88.7%26.3%80.9%1.430.95

Diagnostic Properties When Combining Infection Markers

MarkerSensitivitySpecificityPositive Predictive ValueNegative Predictive ValuePositive Likelihood RatioNegative Likelihood Ratio
CRP >15 mg/L and aspiration WBC count >1100 cells/L16.7%97.1%66.7%77.3%5.830.86
CRP >15 mg/L or aspiration WBC count >1100 cells/L66.7%82.9%57.1%87.9%3.890.40
CRP >15 mg/L and aspiration percent neutrophils >80%40.0%97.2%80.0%85.4%14.400.62
CRP >15 mg/L or aspiration percent neutrophils >80%70.0%77.8%46.7%90.3%3.150.39
Aspiration WBC count >1100 cells/L and aspiration percent neutrophils >80%33.3%92.3%50.0%85.7%4.330.72
Aspiration WBC count >1100 cells/L or aspiration percent neutrophils >80%50.0%84.6%50.0%84.6%3.250.59
CRP >15 mg/L and aspiration WBC count >1100 cells/L and aspiration percent neutrophils >80%20.0%97.1%66.7%81.0%7.000.82
Authors

The authors are from the Assuta Ashdod Medical Center (AH), Ashdod, Israel; and McGill University (AA), Montreal, Quebec, and the University of British Columbia (DSG, CPD, BAM), Vancouver, British Columbia, Canada.

Drs Herman, Albers, Duncan, and Masri have no relevant financial relationships to disclose. Dr Garbuz is a paid consultant for Stryker.

Correspondence should be addressed to: Amir Herman, PhD, MD, Assuta Ashdod Medical Center, 7 Harefua St, Ashdod, Israel 7747629 ( amirherm@gmail.com).

Received: July 10, 2018
Accepted: December 18, 2018

10.3928/01477447-20190103-03

Sign up to receive

Journal E-contents