MARKERS OF HEPATITIS-C INFECTION AMONG HEMODIALYSIS-PATIENTS WITH ACUTE AND CHRONIC INFECTION - IMPLICATIONS FOR INFECTION-CONTROL STRATEGIES IN HEMODIALYSIS UNITS

To evaluate strategies for screening patients on hemodialysis (HD) for markers of acute and chronic hepatitis C virus (HCV) infection, we st udied sixty-nine patients at a single center over a 36-month period. S erum samples were tested for alanine aminotransferase (ALT) levels, an ti-HCV and HCV RNA at 3-4 month intervals. Anti-HCV was tested for by EIA1, EIA2, and RIBA2 HCV RNA was detected by polymerase chain reactio n (PCR). In addition, IgM antibody to the c33 antigen of HCV was defec ted by an experimental EIA. Of the 43 HD patients at the start of the study, anti-HCV was detected by EIA1 in 13 (30%). All EIA1 positive pa tients and 14 (47%) of the 30 EIA1 negative patients tested positive b y EIA2. Thus, at the start of the study 27 (63%) of 43 patients tested positive for anti-HCV by EIA2. The presence of anti-HCV among EIA2 po sitive patients was confirmed by RIBA2 in all patients. Based on the P CR results, She sensitivity, specificity, positive predictive value an d negative predictive value for EIA1 were 48%, 100%, 53% and 100%, res pectively, and for EIA2 were 100%, 100%, 100% and 100%, respectively. During follow-up, 26 EIA2 negative patients began HD in the unit. Of t he 42 EIA2 negative patients, five (12%) seroconverted for anti-HCV du ring follow-up. All five patients with new HCV infection tested positi ve for HCV RNA three months prior to the defection of anti-HCV by EIA2 . In addition, IgM to c33 was detected three months prior to seroconve rsion for anti-HCV by conventional tests in the two patients in whom t esting was performed. All patients in whom HCV RNA was detected remain ed HCV RNA positive through the course of the study. The HCV subtype t esting revealed that 25 (86%) of 29 patients had HCV type II infection and the remainder had mixed infection with both type I and II. The re sults of this study demonstrate that EIA2 is a reliable predictor of H CV infection among HD patients, and that RIBA2 or PCR testing is not n ecessary in routine clinical practice. Further, clearance of viremia a nd recovery from HCV infection is unusual, and hence regular testing f or HCV should be limited to patients who are EIA2 negative. Serum ALT levels and conventional anti-HCV tests are poor predictors of early HC V infection. The anti-c33 IgM test may have a role in the identificati on of individuals with early HCV infection.