Diagnostic Accuracy and Cost-effectiveness Analysis of Serum and Synovial Fluid Markers for the Diagnosis of Periprosthetic Hip and Knee Joint Infections

Total joint arthroplasties (TJA) are constantly increasing in older and multi-morbid patients; knee and hip joints are the most frequently treated due to high incidence of osteoarthritis in these joints. The indication to TJA is now extended to younger and active patients with high functional demands and longer life expectancy. Because of these, an increase in revision surgeries and a PJI risk is also expected. The type of revision surgery (RS) approach and of antimicrobial treatments is based on the presence or absence of infection: one-stage revision with maintenance/replacement of prosthesis in combination with debridement and irrigation (aseptic RS); and two-stage revision with prosthesis removal, debridement and irrigation, the possible use of a temporary spacer, and the potential reimplantation after weeks or months (septic RS).

Today, PJI remains an important dangerous and devastating complication; it has been reported for hip and knee joints an incidence of 0.2-1.5%, prosthesis failure of 15-25%, revision success rate of 80-95% and annual cost per infection of $20,000-$40,000. The timely and most accurate diagnosis is essential for the correct treatment of PJI as patients undergoing revision need extensive physical, psychological and economic supports. Following the definition and standardization of diagnostic major and minor criteria for PJI by the Musculoskeletal Infection Society (MSIS) an improvement in the diagnosis and treatment of PJI was observed. Recently, the International Consensus Group on PJI revised MSIS criteria defining that patients should be considered to have PJI in the presence of one of the major criteria or three of five minor criteria. Finally, an evidence-based and validated criteria for hip and knee PJI definition to diagnose patients within the preoperative period was published in 2018. Patients with an aggregate score of more than or equal to 6 are considered infected, while a 2-5 score requires the inclusion of intraoperative findings (positive histology, purulence and single positive culture) for confirming or refuting the diagnosis. A 4-5 score is inconclusive, and a score of less than or equal to 3 is not infected.

During the last International Consensus Meeting on Musculoskeletal Infection, various questions on PJI were discussed and among them the working groups (WG) provided recommendations based on limited (4 - due to small numbers) and moderate (12 and 19) levels of evidence: WG4, on which patient-specific factors influence the thresholds for serum and synovial markers in acute and chronic PJI; WG12, on which serum tests have the best diagnostic accuracy for PJI and if the combination of any number of tests increase the diagnostic accuracy; WG19, if the profile of organisms causing surgical site infection or PJI following orthopaedic procedures changed over recent years. The WG4 highlighted that no inflammatory arthritis (IA)-specific factors are known to influence the thresholds for serum and synovial markers in PJIs. It appears that in IA patients there are overlaps in values of synovial markers such as alphadefensin, white blood cell count (sWBC) and C-reactive protein (CRP), limiting their usefulness in differentiating septic from aseptic patients. The WG12 stated that it is evident that diagnosis of PJI cannot be based solely on serological tests and that CRP and erythrocyte sedimentation rate (ESR) are well-researched screening tests and have high sensitivity when used alone. However, the combination of serological tests has shown to improve diagnostic accuracy, but the identification of the optimal combination needs further investigations. Finally, the WG19 recommended to perform further investigations regarding the profile of microorganism with next generation sequencing (NGS), as it may confer significant antibiotic selection implications.

Among pre-operative test on synovial fluid, alpha-defensin showed the highest degree of accuracy in the diagnosis of PJI when measured by laboratory-based immunoassay rather than lateral flow test, suggesting that care must be taken with interpretation of the lateral flow test when relying on its results for the intra-operative diagnosis. Synovial fluid could be tested for other newly markers such calprotectin and cathelicidin LL-37, whose sensitivity and specifity are to be defined or for microbiological tests.

Synovial fluid analysis for WBC and %PMN has high sensitivity and specificity in diagnosing PJI, and it represents an easy and widespread available method. Our meta-analysis evaluated: 1) accuracy in terms of sensitivity and specificity of synovial fluid WBC and %PMN tests in the diagnosis of PJI after total knee (TKA) and hip (THA) arthroplasty; 2) which test yielded superior test performance; and 3) the influence of study characteristics such as anatomic site and threshold value on the diagnostic accuracy of these tests. The meta-analysis results revealed that both synovial fluid WBC and %PMN have a high specificity and sensitivity in detecting PJI, and no clear superiority of one test over the other exists. In addition, an anatomic site effect was demonstrated on sensitivity of synovial fluid %PMN, suggesting that cutoff thresholds should be identified separately for TKA and THA. The differences in bearing surfaces and the possibility of fretting corrosion in dual taper stem designs might substantially alter automated synovial fluid analysis in THA compared to TKA.