Lamivudine therapy was associated with (i) a normalization of alanine aminotransferase levels in four of five patients when these levels were increased at the beginning (n=5); (ii) a rapid disappearance of HBV DNA from the serum (detected by hybridization) in all of the patients; (iii) the negativity of HBV DNA by polymerase chain reaction in four patients; and (iv) no change in renal function and in proteinuria when present (one patient). Finally, no adverse effects were noted. When lamivudine therapy was stopped for four patients after 6 months, it was associated with a biochemical and virological relapse within the weeks that followed. Lamivudine therapy was therefore resumed for these patients.

Hodgkin's disease is commonly associated with EBV latent infection. The incidence of EBV reactivation (active infection or EBV infection with replicative cycle) was evaluated in a series of 30 patients with untreated Hodgkin's disease (except for one case with chronic lymphocytic leukemia) by quantitation of EBV DNA and titration of anti-ZEBRA antibodies in serum samples. DNA was detected in serum (>2.5 x 10(2) genomes/ml) in 15 of 30 patients and was more frequent in Hodgkin's disease with EBV-positive Reed-Sternberg cells (10/12) than in EBV-negative cases (5/18), (P< 0.01). Of interest was the demonstration that viremia correlated well with increased titers of anti-ZEBRA IgG and/or standard serological profiles of EBV reactivation (12/15), (P < 0.05). However the lack of EBV replicative cycle in Reed-Sternberg cells (negative for ZEBRA antigen and early antigen BHLF1) suggests that the viral replication occurs in a nonneoplastic cell compartment rather than in tumor cells. The measurement of EBV DNA loads and the titration of anti-ZEBRA antibodies shed new lights on the link between activation of EBV replication and Hodgkin's disease: these serological markers together with the determination of the EBV status of the tumor suggest that replication of the viral genome occurs with a decreased efficiency of the immune system, thus allowing progression of the tumor.

Because the prevention of influenza infection by vaccination warrants a global strategy to target the different reservoirs, we suggest that the modern policy of vaccinating solid organ patients should be extended both to their relatives and to the healthcare workers of transplant units.

A polymerase chain reaction (PCR) assay for the detection of Epstein-Barr virus (EBV) sequences in various clinical samples, especially peripheral blood leukocytes (PBL) and serum, was carried out and the results obtained were compared with specific EBV serology. One hundred seventy patients were enrolled in the study: 89 healthy blood donors, 22 asymptomatic patients, 36 individuals with primary EBV infection (including 19 patients with infectious mononucleosis [IM]), 22 HIV-infected subjects (including 4 with hairy oral leukoplakia, 3 with central nervous disorders, and 15 with non-Hodgkin's lymphoma). All the serum samples from the healthy blood donors were negative. In patients with IM and in AIDS-non Hodgkin's lymphoma (ARNHL), PCR was strongly positive in leukocytes (> 2,000 genome equivalents/10(4) cells), which was correlated with detectable amounts of EBV DNA in serum. The overall positivity rate of PCR in serum was 58.8%, 68%, and 73% of cases for non-IM primary EBV infections, IM, and ARNHL, respectively. In two cases of EBV primary infection, the viral DNA was detected in serum, respectively 1 month and 2 months before IgM positivity and IgG rise. In one case of ARNHL followed up for several months, PCR (viral load of 2,000 genome equivalents/10(4) cells) became positive concurrently with appearance of lymphoma. In immunocompromised individuals, PCR EBV, if carried out in larger prospective studies, could be considered as a tumor marker, useful for predicting EBV-driven lymphoma and follow-up therapy.

JC virus (JCV) induces progressive multifocal leukoencephalopathy (PML), especially in human immunodeficiency virus (HIV)-infected patients. Although JCV genotypes have primarily been associated with geographic patterns, a distinctive neuropathogenicity was recently attributed to genotype 2. A multicenter study was conducted to describe the distribution of JCV genotypes in France and to investigate correlations between genotypes and PML. Genotypes were determined by sequencing 494 bp in the VP1 capsid gene. Peripheral JCV was studied in 65 urine samples from 43 HIV-infected patients and from 22 control subjects. Genotypes 1, 4, 2, and 3 were detected in 52.3%, 30.8%, 12.3%, and 4.6% of the samples, respectively. In 56 brain or cerebrospinal fluid samples, PML-associated JCV of genotypes 1, 2, 4, and 3 was found in 66%, 19.7%, 8.9%, and 5.4%, respectively. Infection with JCV genotypes 1 or 2 was correlated with PML (odds ratio, 3.29). On the other hand, infection with JCV genotype 4 could represent a lower risk for PML.