Epidemiology and pathogenesis of HHV-6

Abstract

Human herpesvirus-6 (HHV-6), a beta-herpesvirus, was discovered in 1986. It is divided into two variants, A and B. HHV-6 variant B is the cause of exanthema subitum, while variant A has not yet been associated to any disease. HHV-6 is ubiquitous and seroepidemiological studies have shown that more than 90% of children older than two years of age are seropositive. HHV-6 infections in the immune competent individual are usually benign, but complications like encephalitis, hepatitis and disseminated disease have been reported.

We developed an ELISA for detection of HHV-6 IgG antibodies for seroepidemiological studies and compared it to an immunofluorescence assay (IFA). Both methods revealed a seropositivity of > 90%, and the ELISA was a reliable alternative for measurement of HHV-6 IgG antibodies. The prevalence of variants A and B is not known, since available serological assays do not discriminate between the two variants. Polymerase chain reaction (PCR) and T-lymphocyte proliferation assays were used to identify the distribution of variants A and B in a group of healthy adults. HHV-6 DNA was detected in 22% of the leukocyte samples. All samples that could be typed were variant B. Furthermore, 58% of healthy adults responded to variant B in the lymphocyte proliferation assay and 25% to variant A, but the responses did not correlate to the PCR assays.

In order to investigate possible congenital HHV-6 infection we analysed the presence of HHV-6 DNA in matemal blood and in cord blood from 107 mother-infant pairs. In a second group, sequential samples were taken from women at months 3, 5 and 6-8 during pregnancy, and from mother and infant at delivery. Cord blood samples from 104 mother-infant pairs were also analysed. HHV-6 DNA was detected in around 40% of samples during month 3-8 of pregnancy, compared to 25% in non-pregnant age matched controls. HHV-6 IgG antibodies were found in 96%, and the titres were significantly higher in women with HHV-6 DNA in their leukocytes. In 2/211 mother /cord blood pairs we detected HHV-6 variant B in leukocytes and plasma indicating congenital infection. There was no indication that HHV-6 caused fetal damage, since all children were apparently healthy at birth. Cytomegalovirus (CMV), another beta-herpesvirus, infects the fetus in 0.5-2.5% of normal pregnancies, and CMV could be transferred by infection of placental trophoblasts. The possibility to infect trophoblasts in vitro with HHV-6 in vitro was investigated by use of four different strains of HHV-6. Trophoblasts could not be infected with HHV-6, despite positive CMV isolation results.

Allogeneic stem cell transplantation (allo-SCT) can cure patients with some severe diseases, but treatment-related herpesvirus infections remain a problem. HHV-6 infection has been suggested to be associated various symptoms. By HHV-6 DNA analysis we confirmed that HHV-6 variant B was associated with delayed granulocyte and platelet engraftment, but no relation to acute graft versus host disease was observed. HHV-6 DNA can be found without obvious relation to disease, but high copy numbers has been correlated to symptomatic disease. A real-time PCR was developed to facilitate quantification and different DNA extraction methods were evaluated. HHV-6 DNA was found in blood from 30%, and in mouth washings from 80% of the patients. The quantitative method will be used to verify symptomatic infections and monitor the effect of antiviral treatment. HHV-6 is an important pathogen in the population of immune compromised patients and proper diagnostic tools are of great value.