Dialysate bacterial endotoxin as a prognostic indicator of peritoneal dialysis related peritonitis

• Published in: Nephrology (Carlton, Vic.) Vic.), 2016-12, Vol.21 (12), p.1069-1072
• Main Authors: Szeto, Cheuk-Chun, Lai, Ka-Bik, Chow, Kai-Ming, Kwan, Bonnie Ching-Ha, Law, Man-Ching, Pang, Wing-Fai, Ma, Terry King-Wing, Leung, Chi-Bon, Li, Philip Kam-Tao
• Format: Article
• Language: English
• Subjects: Anti-Bacterial Agents - therapeutic use; Bacteria; Bacteria - drug effects; Bacteria - metabolism; Biomarkers - metabolism; Dialysis Solutions - metabolism; Endotoxins - metabolism; Humans; infection; Leukocyte Count; Peritoneal Dialysis - adverse effects; Peritonitis - diagnosis; Peritonitis - drug therapy; Peritonitis - microbiology; Predictive Value of Tests; Recurrence; renal failure; Reproducibility of Results; survival; Time Factors; Treatment Failure
• ISSN: 1320-5358; 1440-1797
• DOI: 10.1111/nep.12828

Peritonitis is the major complication of peritoneal dialysis (PD). The aim of our present study is to explore the prognostic value of endotoxin level in PD effluent for the prediction of treatment failure in PD‐related peritonitis. We studied 325 peritonitis episodes in 223 patients. PD effluent (PDE) was collected every 5 days for endotoxin level and leukocyte count. Patients were followed for relapsing or recurrent peritonitis. We found 20 episodes (6.2%) had primary treatment failure; 41 (12.6%) developed relapsing, 19 (5.8%) had recurrent, and 22 (6.8%) had repeat episodes. Endotoxin was detectable in the PDE of 19 episodes (24.4%) caused by Gram negative organisms, 4 episodes (6.8%) of mixed bacterial growth, and none of the culture negative episodes or those by Gram positive organisms. For episodes caused by Gram negative bacteria, a detectable endotoxin level in PDE on day 5 had a sensitivity and specificity of 66.7% and 83.3%, respectively, for predicting primary treatment failure. In contrast, PDE leukocyte count > 1000 per mm3 on day 5 had a sensitivity and specificity of 88.9% and 89.1%, respectively; the addition of PDE endotoxin assay did not improve the sensitivity or specificity. We conclude that detectable endotoxin in PDE 5 days after antibiotic therapy might predict primary treatment failure in peritonitis episodes caused by Gram negative organisms. However, the sensitivity and specificity of PDE endotoxin assay was inferior to PDE leukocyte count.