Endotoxemia is associated with better clinical outcome in incident Chinese peritoneal dialysis patients: a prospective cohort study.

Background: Endotoxemia is common in peritoneal dialysis (PD) patients; circulating lipopolysaccharide (LPS) level is related to the degree of systemic inflammation and atherosclerosis. We examine whether baseline plasma LPS level represents a prognostic marker in new PD patients.

Methods: We studied 158 new Chinese PD patients (80 males). Baseline plasma LPS level at initiation of PD was measured. Patients were stratified into quartiles according to plasma LPS level: quartile I, <0.45 EU/mL; II, 0.45 - <0.70 EU/mL; III, 0.70 - <0.95 EU/mL; and IV, ≥ 0.95 EU/mL. The patients were then prospectively followed for the development of cardiovascular events. All-cause mortality and duration of hospitalization were also recorded.

Results: Average age was 55.6 +/- 14.7 years; average endotoxin concentration was 0.70 +/- 0.30 EU/mL; average follow-up was 55.5 +/- 36.9 months. At 60 months, event-free survival was 41.0%, 52.5%, 65.0%, and 61.5% for LPS level quartiles I, II, III, and IV, respectively (log rank test p = 0.066). By multivariate analysis with the Cox proportional hazard model to adjust for confounders, plasma LPS level had no independent effect. At 60 months, technique survival was 20.5%, 20.0%, 32.5%, and 51.3% for LPS level quartiles I, II, III, and IV, respectively (log rank test p = 0.0009). By Cox proportional hazard model, each higher quartile of LPS conferred 28.6% protection (95% confidence interval 15.6% - 40.3%, p = 0.0002) from developing technique failure. A higher plasma LPS level had a lower all-cause mortality (unadjusted hazard ratio 0.486, p = 0.046) and cardiovascular mortality (unadjusted hazard ratio 0.251, p = 0.025), but the result became insignificant after adjusting for potential confounders.

Conclusion: A higher baseline plasma LPS level is an independent predictor of better technique survival in new Chinese PD patients, with an insignificant trend of fewer cardiovascular events. The observation seems to conform to the phenomenon of reverse epidemiology for other traditional cardiovascular risk factors in dialysis patients but the exact reason for this paradoxical phenomenon requires further investigation.