Acute Kidney Injury When Treating Periprosthetic Joint Infections After Total Knee Arthroplasties with Antibiotic-Loaded Spacers Incidence, Risks, and Outcomes

摘要

Background: Two-stage exchange arthroplasty with a high-dose antibiotic-loaded bone cement (ALBC) spacer and intravenous or oral antibiotics is the most common method of managing a periprosthetic joint infection (PJI) after a total knee arthroplasty (TKA). However, little is known about the contemporary incidence, the risk factors, and the outcomes of acute kidney injuries (AKIs) in this cohort. Methods: We identified 424 patients who had been treated with 455 ALBC spacers after resection of a PJI following a primary TKA from 2000 to 2017. The mean age at resection was 67 years, the mean body mass index (BMI) was 33 kg/m(2), 47% of the patients were women, and 15% had preexisting chronic kidney disease (CKD). The spacers (87% nonarticulating) contained a mean of 8 g of vancomycin and 9 g of an aminoglycoside per construct (in situ for a mean of 11 weeks). Eighty-six spacers also had amphotericin B (mean, 412 mg). All of the patients were concomitantly treated with systemic antibiotics for a mean of 6 weeks. An AKI was defined as a creatinine level of >= 1.5 times the baseline or an increase of >= 0.3 mg/dL within any 48-hour period. The mean follow-up was 6 years (range, 2 to 17 years). Results: Fifty-four AKIs occurred in 52 (14%) of the 359 patients without preexisting CKD versus 32 AKIs in 29 (45%) of the 65 patients with CKD (odds ratio [OR], 5; p = 0.0001); none required acute dialysis. Overall, when the vancomycin concentration or aminoglycoside concentration was >3.6 g/batch of cement, the risk of AKI increased (OR, 1.9 and 1.8, respectively; p = 0.02 for both). Hypertension (beta = 0.17; p = 0.002), perioperative hypovolemia (beta = 0.28; p = 0.0001), and acute atrial fibrillation (beta = 0.13; p = 0.009) were independent predictors for AKI in patients without preexisting CKD. At the last follow-up, 8 patients who had sustained an AKI had progressed to CKD, 4 of whom received dialysis. Conclusions: In our study, the largest series to date that we are aware of regarding this issue, AKI occurred in 14% of patients with normal renal function at baseline, and 2% developed CKD after undergoing a 2-stage exchange arthroplasty for a PJI after TKA. However, the risk of AKI was fivefold greater in those with preexisting CKD. The causes of acute renal blood flow impairment were independent predictors for AKI.