A comparative study of intraoperative frozen section and alpha defensin lateral flow test in the diagnosis of periprosthetic joint infection.

Background and purpose - For decision-making (aseptic vs. septic), surgeons rely on intraoperatively available tests when a periprosthetic joint infection (PJI) cannot be confirmed or excluded preoperatively. We compared and evaluated the intraoperative performances of the frozen section and the alpha defensin lateral flow test in the diagnosis of PJI. Patients and methods - In this prospective study, consecutive patients with indicated revision surgery after arthroplasty were included. Patients were classified as having PJI using the MusculoSkeletal Infection criteria. The presence of alpha defensin was determined using the lateral flow test intraoperatively. During revision surgery, tissue samples were harvested for frozen and permanent section. Analysis of diagnostic accuracy was based on receiver-operating characteristics. Results - 101 patients (53 hips, 48 knees) were eligible for inclusion. Postoperatively, 29/101 patients were diagnosed with PJI, of which 8/29 cases were definitely classified as septic preoperatively. Of the remainder 21 septic cases, the intraoperative alpha defensin test and frozen section were positive in 13 and 17 patients, respectively. Sensitivities of the alpha defensin test and frozen section were 69% and 86%, respectively. The area under the curves of both tests showed a statistically significant difference (p = 0.006). Interpretation - The frozen section showed a significantly higher performance compared with the alpha defensin test and a near perfect concordance with the definitive histology, and therefore remains an appropriate intraoperative screening test in diagnosing PJI. Although the sensitivity of the alpha defensin test was lower compared with that of frozen section, this test is highly specific for confirming the diagnosis of PJI.

Effect of simulated metastatic lesions on the biomechanical behavior of the proximal femur.

Pathologic fractures of femora in patients with metastatic cancer are associated with high morbidity and mortality. Prediction of impending fractures is based on unspecific clinical criteria or past clinician's experience, which leads to underestimation or overtreatment. The aim of this study was to investigate the effect of the site of metastatic lesions on biomechanical behavior of the proximal femur. Sixteen pairs of human femora were scanned with quantitative computed tomography (QCT) to asses bone mineral density. One femur of each pair remained intact while a defined lesion was reamed out in either the superolateral or inferomedial portion of the femoral neck of the contralateral femur. All femora were loaded in a mechanical test setup mimicking one-legged stance and stiffness, failure load, and fracture location were determined. In the biomechanical experiments the superolateral lesion and the inferomedial lesion caused a stiffness reduction of 19% and 66%, respectively. The average failure load was 40% and 75% lower for specimens with the superolateral (4.53 ± 1.56 kN) and inferomedial (1.89 ± 1.73 kN) lesions, respectively, compared to intact specimens (7.66 ± 3.34 kN). Lesions in the femoral neck led to reduction in both stiffness and failure load of the proximal femur. Furthermore, the site of the lesion had a large effect on the magnitude of the reduction in biomechanical properties. The presented data emphasize the importance of differentiating between locations of the lesion in pathologic fracture prediction of the metastatic femur and underline the insufficient accuracy of current predictive guidelines. © 2017 Orthopaedic Research Society. Published by Wiley Periodicals, Inc. J Orthop Res 35:2407-2414, 2017.