[Comprehensive diagnosis of infection in revision total replacements of large joints].

Koudela, K; Geigerová, L; Hes, O; Koudela, K

ABSTRACT

PURPOSE OF THE STUDY To make comprehensive diagnoses of the infections associated with revision total knee and hip arthroplasties in our group of patients MATERIAL AND

METHODS:

From September 2002 till November 2004, a group of 69 patients undergoing revision total joint replacement (65 hips and four knees) were evaluated. The period between primary and revision surgery ranged from 6 months to 25 years. The patients were examined for CRP, erythrocyte sedimentation rate (ESR) and white blood cell (WBC) counts. The samples of their periprosthetic tissue were assessed for biopsy and microbial findings. The removed prosthetic components were sonicated.The samples were cultured in both aerobic and anaerobic conditions for 16 days. A finding of more than 10 neutrophils per viewing field was taken as a positive biopsy result. The definition of an infection was based on the detection of a microorganism with the identical phenotype in two or more samples.

RESULTS:

Before surgery, 13 patients had a suspected infection which was subsequently diagnosed. A positive culture result in at least one of the collected samples was found in 48 patients; of these, a positive culture finding of a phenotypically identical microorganism in two or three samples was in 32 patients, who thus met the definition of infection. The average values for the whole group were CRP, 16 mg/L (1-109); ESR, 25 mm/h (3-110); peripheral WBC count, 6.2 x 109/L (3.6-11.6). The microorganisms most frequently growing in culture were coagulase-negative staphylococci and propionibacteria accounting for 41% and 29% of the total isolates obtained, respectively. From the total number of samples, positive culture results were obtained in 36% of sonicate femoral components; 40% of sonicate acetabular cups, 51% of periprosthetic tissues and 48% of swabs. In these positive microbial cultures strictly anaerobic microorganisms were found in 41% of femoral component, 49% acetabular component and 42% periprosthetic tissue samples and in 27% of swabs taken at arthrotomy. Prolonged cultivation of the 151 isolates initially obtained yielded 81 (54%) isolates which would have failed to be detected by primary culture. The results of laboratory tests in the patients with negative culture findings, in those with a phenotypically identical micro- organism found in one sample, and in those with positive findings in two or more samples were CRP, 4.3 mg/L; 9.8 mg/L; and 21.7 mg/L, respectively; ERS, 13.5mm/h; 20.1 mm/h; and 33.0 mm/h, respectively; and WBC counts, 6.27 x 109/L; 6.25 x 109/L; and 6.16 x 109/L, respectively. The t-test was used for the statistical analysis of CRP, ESR and WBC count values, and it revealed a significant differences between the patients with negative microbial findings and those with positive microbial findings in two and more samples in all three values, i.e., CRP (p = 0.01), ESR (p = 0.01) and WBCs (p = 0.96). Biopsy findings showed a sensitivity of 62.5 % and a specificity of 91% in relation to the microbial findings.

DISCUSSION:

Our results as well as relevant literature data suggest that microorganisms may survive on implant surfaces even in the cases regarded as aseptic. They often grow slowly and, theoretically, can have an adverse effect on the longevity of revision arthroplasty. However, because of current endoprosthetic practices and the ubiquitous presence of microorganisms, contamination of some samples cannot be excluded.

CONCLUSIONS:

In our group of patients, the CRP and ESR values proved to be useful in making the diagnosis of infection. For this pur- pose, WBC counts in blood samples were not sensitive enough. Biopsy findings had low sensitivity, but appeared to be a specific marker of infection. Prolonged cultivations of samples and cultivation under anaerobic conditions resulted in a marked increase in isolates obtained, as compared with the routine cultivation technique.