Classifications in Brief: The McPherson Classification of Periprosthetic Infection

Abstract

History Periprosthetic joint infections (PJIs) after arthroplasty cause morbidity and contribute to mortality; they also increase the costs of treatment, both to the patient and the healthcare system [27]. The risk of THA and TKA infection is often quoted in the literature as less than 2% [23]. Most patients who develop a PJI can be treated, however, as many as 30% [11] of these patients will have persistent infection despite appropriate treatment. Decision making about the best treatment approach for these patients remains controversial. Widely accepted management of PJI is a two-stage revision with a targeted antibiotic regimen [27]. Other surgical interventions include débridement and implant retention (DAIR), with or without modular component exchange, and single-stage direct-exchange revision arthroplasty. Excisional arthroplasty and amputation as an index operation typically are reserved for unusual circumstances. McPherson et al. [20] were the first to propose a clinical staging system for knee PJI. This system has three components: infection type, systemic host grade, and local extremity grade (Table 1). The premise behind this classification was the authors’ subjective experience with patients with systematic comorbidities that predispose a patient to infection (or that can potentiate its persistence) were more difficult to treat, as were those with local soft-tissue compromise [20]. Their initial paper retrospectively applied the staging system to 70 infected TKAs, and they reviewed outcome measures including the Knee Society Scores, complications, amputations and permanent resections after treatment. Subsequently, McPherson et al. [21] applied the same system to infected THAs treated with a two-stage revision, with a similar observation; patients who were healthier at baseline and who had fewer local findings suggestive of severe infection or poor tissue healing ability (like soft-tissue abscess or sinus formation) were more likely to have successful reimplantation after PJI.Table 1.: McPherson Staging system for prosthetic joint infectionBefore the publication of McPherson et al.’s classification [20], no systems dedicated to PJI had been described. Grouping orthopaedic-related infections based on duration, the host’s systemic condition, and the host’s local condition was not unique [6, 14, 17, 30], McPherson and his team though, in 1999, were the first to apply this to arthroplasty. However, similarities can be drawn between McPherson’s classification [20] and the Cierny and DiPasquale [5], and the Cierny and Mader [6] classification system of osteomyelitis in the long bones, which was later modified and adapted to grade PJI. Both systems stratify patients based on their current morbid state, related to both systemic and local health. Purpose Treatment of PJI is challenging and can lead to functional impairment, pain, and even death. The development and application of an objective and clinically relevant system, like McPherson’s classification, provides important information to facilitate communication, research, prognosis and treatment decisions. Currently, the McPherson classification system is recommended by the Musculoskeletal Infection Society for grading PJI, and there is no internationally accepted equivalent [1]. Using the McPherson classification to identify those patients who might achieve eradication of PJI with a single-stage procedure could provide a cost benefit to health systems, as well as an easier course of treatment for patients to endure [12, 27]. This classification system may help clinicians to identify patients who are at a higher risk for failed surgical management, which may result in outcomes including arthrodesis and amputation. Description The McPherson staging system for PJI incorporates three distinct parameters: infection type (temporal phase), systemic host grade, and local extremity grade (Table 1); these are then used together to assign a status; for example, II, B, and 1 (compromised host, with acute hematogenous infection (< 4 weeks) of the joint, and uncompromised skin at the surgical site). Stratification of patients by the McPherson system starts with the duration of infection, using 4 weeks as a division point for acute and chronic infection. Acute infections are further divided into early postoperative and hematogenous infection. The number and type of medical comorbidities are used to segregate patients into grades (A) uncompromised, (B) compromised, and (C) substantially compromised host status groups. Compromising systemic factors either impair the immune system or the body’s synthetic and anabolic capacity or drive a continuous inflammatory state (Table 2).Table 2.: Systemic and Local Wound factors in patients at risk for prosthetic joint infectionThe assessment of the local and extremity status again highlights issues of soft-tissue coverage, delivery of essentials for metabolism and immune function, and skin integrity. Based on these parameters, the extremity is graded as 1, 2, or 3 (uncompromised, compromised, and severely compromised, respectively), depending on the number of comorbidities associated with the limb. Validation Although widely used, the McPherson staging system for periprosthetic infection has not been validated with intra- and interobserver studies in the current evidence. The objective nature of the system would suggest that little variability would be encountered. Studies investigating the reliability of the system to predict patient outcome have been typically small, retrospective, and variable in their findings. Despite these shortcomings, the International Consensus Group on Orthopaedic Infections advocates the use and further development of this system [1, 10]. Since McPherson’s first description of this system [20] and their series of TKA and THA [20, 21], multiple studies have used this system to report success or failure of treatment. To date, there are no well-designed prospective series applying this system; most are retrospective case series, and many focus on the McPherson host grade alone in patients either with acute or chronic infections, with variable local extremity grading. Two-stage exchange arthroplasty for infection is a commonly recommended procedure to cure infections. In a review of 245 patients with infected TKAs treated with two-stage revision arthroplasty, the application of the McPherson classification identified three factors that were associated with an increased risk of re-infection: McPherson host grade C, previous revision surgery, and a BMI greater than 30 kg/m2 [24]. The latter two parameters suggest a worse McPherson extremity grade, but in their direct comparison of Grade 2 and 3 limbs, no difference was found (no Grade 1 limbs were treated). In a series that replicated the original McPherson THA paper [21], 60 Type III infections were treated with a two-stage hip revision [2]. The authors found that worsening of the host’s status was related to a higher risk of a postoperative draining sinus, and the local extremity grade was positively associated with wound healing complications [2]. Interestingly, host grade was not associated with an increase in one or more complications, or a difference in amputation and reimplantation rates nor death during a 24-month follow-up period. This may be related to the low sample size, a difference in patient group heterogeneity, or different treatment protocols (exclusive use of spacers [2] versus resection and spacers [21]). A study on the outcomes of 92 patients with infected primary THAs, with a rigorous application of the classification system and a minimum of 2 years of follow-up concluded that only patients with McPherson Grade I/A/1 could or should be considered for single-stage revision [32]. This study is the largest published THA series on this subject, and in particular, the largest using all three facets of the scoring system. It provides support for the use of the system when deciding between a single-stage or two-stage THA revision. The use of débridement and implant retention (DAIR) to treat acute PJI remains controversial. In a review of 90 patients with acute PJI after THA at a single center who underwent DAIR, 17% (15 of 90) of THA revisions failed due to acute reinfection and underwent either resection arthroplasty or revision DAIR; failure proportions increased with the host’s grade (Host grade A 8%, Host grade B 16% [p = 0.04], Host grade C 44% [p = 0.006]) [4]. They were unable to identify a positive relationship between the temporal infection type (that is, McPherson Type I or II) or the local extremity grade regarding failure to eradicate infection (component removal or repeat DAIR due to infection recurrence). Affirming that for THA host grade should be considered when offering DAIR procedures; local extremity grade may be of less importance around the hip due to good soft tissue coverage. Conversely, in the treatment of acute infection (McPherson Types I and II) using the DAIR technique in a series of 134 patients with infected primary TKAs, McPherson’s host grade was not shown to be associated with an increased risk of subsequent infection or component removal [31]. The protocol used in this study included antibiotics for the life of the implant. Mortality of these patients was 15% by 2-year follow-up, of which grade C hosts had an increased risk of death (HR 18.4 (95% confidence interval 3.9 to 87.2); p < 0.001) compared with grade A hosts, as did grade B hosts (HR 7.4 (95% CI 1.6 to 33.3); p = 0.01) [31]. The notion that single-stage and DAIR-type procedures carry a high risk of recurrent infection in compromised hosts was supported in a review of chronically infected (Type III) lower-limb arthroplasties in immunocompromised patients [15]. Thirteen of the 24 Grade B hosts had a recurrence of infection, and three of five Grade C hosts had a recurrence. No Grade A hosts were included as controls. Two studies from the Mayo Clinic reviewed patients undergoing a second two-stage exchange arthroplasty for PJI after TKA and THA [3, 9]. Although both studies had small groups (45 TKA patients and 19 THA patients), the repeat two-stage TKA results showed that the chance of success (prosthesis retention without subsequent procedures) in Grade A hosts was only 70% (7 of 10), falling to 50% (10 of 20) in Grade B hosts and 0% in Grade C hosts (0 of 2) [9]. In contrast, within the repeat two-stage THA group, there was no association between the host’s grade and reinfection risk after THA [3]. This was likely influenced by the small cohort. Other studies on the treatment and salvage of infected arthroplasty support that local limb conditions and systemic factors should be considered [19, 24, 29]. A review of 38 two-stage TKA revisions identified that local limb conditions affect infection clearance, with increased infection recurrence in extensively scarred limbs, and those with chronic venous insufficiency [29]. Additionally, systemic factors such as chronic pulmonary disease, low preoperative lymphocyte count, and diabetes mellitus increase risk of unsuccessful infection eradication and wound complications [8, 18]. Limitations A primary limitation of the McPherson classification is the paucity of studies confirming its inter- and intra-observer reliability. Despite this shortcoming, this system does have international acceptance and continues to be reported in current scientific literature. The McPherson classification of PJI is comprehensive but lacks some defining criteria that may ultimately affect the outcome [13]. The elements of the “ideal” staging system are proposed to include clinical presentation, etiopathogenesis, anatomo-pathologic findings, host type, micro-organisms, bone defects, and soft tissues [25]. Similar conclusions have been made regarding osteomyelitis of the long bones [14]. McPherson’s system addresses three of these: duration (clinical presentation), host status, and soft tissues. The McPherson classification does not account for the infecting organisms and resistance profiles, which are important to consider when treating infected arthroplasty. Two-stage [28] and DAIR [26] procedures have a high failure rate in those with infections with multi-resistant organisms. A counterargument was presented in a study of 37 TKAs infected with methicillin-resistant Staphylococcus aureus or Staphylococcus epidermidis, among which only 11% of patients had reinfection with the same organism after two-stage revision; the overall reinfection rate was 24% [22]. In addition to factors affecting host status (Table 2), other authors have suggested there may be more affecting variables; however, to what extent each host comorbidity drives the host’s inability to overcome infections and operative complications is unknown. These include coronary artery disease, anemia, psychiatric disease, and sleep apnea [8]. Few studies have examined the management of massive bone loss in patients with PJI. Minor bone loss during reimplantation can be addressed with cement, sleeves, or distal-fit prostheses, whereas patients with massive bone loss may need a megaprosthesis. Successful treatment of infected THA and TKA with megaprostheses has been reported in a small series at 83% (24 of 29) [7]. Bone loss around the infected implant may be less of a concern than it is in the management of osteomyelitis of the long bones and may not need to be included in hypothesized classification systems of PJI. A logical assumption is that older patients are likely to do worse than younger patients, but some authors suggest that patients younger than 60 years may also be at risk [28, 31]. A review of DAIR in patients with chronically infected TKAs concluded that patients younger than 60 years had both an increased risk of subsequent infections and a need for component removal [31]. In another analysis of infected TKAs, the results showed that the mean age of patients who successfully underwent two-stage revision was higher than that of patients with reinfection (mean success 74 years, mean failure 65; p = 0.01) [28]. Because both of these studies were retrospective, the underlying nature of the poorer outcome in younger patients may be due to selection bias and prevention of “sicker, older” patients from completing their surgical protocols. The distinction between a Type II and III infection is the presence of infection symptoms for more than 4 weeks; some suggest this temporal period must be reduced. Symptoms for less than 48 hours was identified to improve infection clearance in patients undergoing a DAIR procedure, in a study of 99 patients [16]. Conclusions The McPherson classification system provides a way to group patients with infected THAs and TKAs. It provides a tool to communicate host and infection characteristics in what is a very complex and specialized field. Of significant concern is the seemingly widespread, international use of this classification despite an absence of validation studies. Two-stage revision for infected arthroplasty is commonly accepted as necessary for the treatment of periprosthetic infection, but it is laden with risk and burden to the patient. The use of the McPherson staging system has led to existing literature which suggests that DAIR and single-stage revision arthroplasty for infected THA and TKA can be favorably considered in McPherson Type I/A/1 patients. Caution and consideration of treatment goals should be strongly considered if providing DAIR or single-stage revision to patients with host Grade B or C, or those with Type III infections. Demographic changes among future arthroplasty patients means the development of a robust, validated system to stratify failed septic arthroplasties for determining a treatment regimen will continue to be important. Within the literature to date, robust studies of periprosthetic classification systems are absent. This provides opportunity for future research, although this will remain difficult due to low numbers of patients with infected arthroplasties. Many treatment protocols are based on innominate regimes proposed through expert opinions and available studies. The McPherson classification system attempts to address this issue, although it remains an unvalidated system, supported by only small case series. Although the McPherson classification system is accepted by a number of relevant large international groups, we caution the use of this system until research validating its use exists.

Topics & Concepts