Serial Monitoring of Complement Levels Identifies Kidney Transplant Recipients at High Risk of Infection: A Prospective Study

Abstract

Objectives: The complement system plays an important role in mediating both acquired and innate responses to defend against infection. The usefulness of serial monitoring of complement levels to assess the risk of infection in kidney transplant (KT) recipients is not well established. Methods: In this observational cohort study we prospectively included 235 patients (151 males; mean age: 53.9 ± 14.2 years) who underwent KT at our institution from November 2008 to August 2010. Complement levels (C3 and C4) were determined by nephelometry just before transplantation (T0) and at months 1 (T1) and 6 (T6) post-transplant. We defined C3 hypocomplementemia as a decrease in serum levels below the lower normal limit in our laboratory (83 mg/dL). The primary study outcome was the occurrence of at least one episode of post-transplant infection. We divided the post-transplant follow-up period in three intervals: early (first month), intermediate (1-6 months), and late (>6 months). Multivariate adjusted odds ratios (ORs) were calculated using those covariates that were found to be significant at P < 0.10 by univariate analysis. Results: The incidence of C3 hypocomplementemia progressively decreased along the study period: 25.1% at baseline (T0), 18.7% at T1, and 8.9% at T6 (P = 0.004). The median duration of follow-up was 509 days. A total of 150 recipients (63.8%) developed 356 episodes of post-transplant infection (overall incidence rate: 2.75 per 1,000 transplant-days). Patients with C3 hypocomplementemia at T0 did not experience higher rates of infection in the early period. Patients with C3 hypocomplementemia at T1 had increased incidences of any type of infection (P = 0.013), bacterial infection (P = 0.008), acute graft pyelonephritis (P = 0.030), cytomegalovirus infection (P = 0.031), and other viral infection (P = 0.007) during the intermediate period. The incidence rates of overall infection during this period in patients with and without C3 hypocomplementemia at T1 were 9.81 and 4.68 episodes per 1,000 transplant-days, respectively (P < 0.001). Finally, C3 hypocomplementemia at T6 was associated with a higher incidence rate of overall infection in the late period (1.90 vs. 0.82 episodes per 1,000 transplant-days; P = 0.032), and higher cumulative incidence of bacterial infection (P = 0.089). On multivariate regression, C3 hypocomplementemia at T1 was found to be associated with an increased risk for bacterial infection (OR 2.41; 95% confidence interval [CI] 1.08 - 5.34; P = 0.031) and had a nearly significant impact on the risk for any type of infection (OR 2.14; 95% CI 0.96 - 4.77; P = 0.063) in the intermediate period. Conclusions: Our study suggests that the serial measurement of complement levels during the post-transplant period provides an affordable and useful approach to the risk of infection in KT recipients.