Display options
Share it on

Clin Diagn Virol. 1995 Aug;4(2):121-34. doi: 10.1016/0928-0197(95)00002-p.

Polymerase chain reaction for the early diagnosis of cytomegalovirus hepatitis in liver transplant patients.

Clinical and diagnostic virology

L M Barkholt, B Johansson, B Veress, J P Andersson, A Ehrnst

Affiliations

  1. Department of Transplantation Surgery, Huddinge Hospital, Karolinska Institute, S-14186 Huddinge, Sweden.

PMID: 15566834 DOI: 10.1016/0928-0197(95)00002-p

Abstract

BACKGROUND: In liver transplant (LTX) patients, cytomegalovirus (CMV) hepatitis as a cause of graft dysfunction occurs in 15-25% of the patients. Polymerase chain reaction (PCR), applied to liver biopsy specimens, may increase the ability to detect CMV DNA at a local site. In this study, PCR was used to compare its relation to the development of clinical CMV hepatitis.

STUDY DESIGN: Nested polymerase chain reaction (nPCR), derived from a conserved region of the CMV major immediate-early gene, was used to examine 141 frozen liver biopsies from 61 LTX patients for the presence of CMV DNA. 134 biopsies were obtained from 54 patients with pathological liver function tests within four months after transplantation. The remaining seven patient biopsies were derived from the one-year investigation after LTX and served as controls. The results were compared to virus isolation, antigen detection by immunohistology and in situ hybridization for CMV DNA of the biopsy specimens. Histological examination was performed to verify a diagnosis of viral hepatitis.

RESULTS: CMV DNA was amplified in 11% (15/134) of the biopsies, corresponding to 20% (11/54) of the patients. Virus isolation revealed CMV in 5% (7/134) of the samples. None of the nPCR-negative biopsies was virus culture positive. CMV genomes were detected by nPCR more frequently than CMV hepatitis was diagnosed by using the combination of CMV-specific histopathology and/or immunohistology and/or CMV-positive virus isolation (p < 0.01). However, when this comparison was performed within individual patients, the difference was not significant (p > 0.05). If the results of in situ hybridization were included in the diagnostic criteria of CMV hepatitis, the nPCR was comparable to these, both at the biopsy and the patient levels (p > 0.1 and p > 0.05, respectively). For the diagnosis of CMV hepatitis the negative predictive value of CMV-nPCR was 1.0. The positive predictive value ranged from 0.55 to 0.82 depending on the criteria of CMV hepatitis. The nPCR also detected signs of CMV infection in the liver graft earlier than virus isolation, 11 versus 21 days, respectively, after transplantation.

CONCLUSION: The frequency of CMV DNA positivity, measured by nPCR, was similar to that revealed by other combined methods. We suggest that the combined findings of histological cholangitis and/or lobulitis together with nPCR for CMV DNA can be used as a diagnostic criterion for initiation of antiviral treatment against CMV hepatitis.

Publication Types