When is the best time for measles vaccination in infants born to HIV-infected mothers?

Children born to HIV-infected Zambian mothers had lower levels of passively acquired maternal antibodies against the measles virus (MV) and an increased risk of measles before routine childhood vaccination, according to the findings of an observational study published in the December 1st edition of Clinical Infectious Diseases. The authors and an accompanying editorial recommend administering the first dose of measles vaccine as early as six months of age in these children.

Measles is still an important cause of childhood deaths in sub-Saharan Africa before they are vaccinated. Passively acquired maternal antibodies against MV provide protection to young infants. However, these antibodies might also neutralise the vaccine virus and interfere with vaccine-induced immunity.

The timing of measles vaccination is therefore crucial. In many African countries, the measles vaccine is delivered during routine childhood immunisation at nine months of age when 85 % of infants born to mothers with naturally acquired immunity respond to the vaccine. The usefulness of this policy in infants of HIV-negative mothers remains unknown.

There is currently no information about the time at which maternally acquired anti-MV antibodies disappear in children born to HIV-positive mothers. This information is critical for the timing of measles vaccination. In order to address this issue, a team of British, US, and Zambian investigators compared anti-MV antibodies in infants aged six weeks to nine months and their HIV-infection or exposure status.

The study took place between May 2000 and November 2002 at a public clinic for routine childhood immunisation in Chawama Township, Lusaka, Zambia. Study subjects were HIV-uninfected and HIV-positive children who were visiting the clinic for immunization but had not been vaccinated against measles. The Edmonton-Zagreb measles vaccine was administered at nine months of age to both groups of children.

Blood samples were obtained from each child at enrolment and a standard questionnaire was completed about the clinical history. After vaccination, a second blood sample was obtained and another questionnaire completed.

Laboratory assays consisted of a rapid HIV immunoassay, HIV RNA determination using reverse-transcription polymerase chain reaction, and measurement of anti-MV neutralising antibodies by a viral neutralisation assay in infants aged between six weeks and nine months. Anthropometric scores for children’s growth were determined using standard methods.

Antibody levels of 120 mIU/ml or more were considered to be protective against wild-type MV infection. Antibody levels of 50-120 mIU/ml were considered sufficient to interfere with immune responses against the measles vaccine but insufficient to protect against MV infection. Antibody levels of less than 50 mIU/ml were considered non-protective and insufficient to interfere with immune responses. Potential predictors of anti-MV antibody levels such as maternal age and infant variables like sex, wasting, stunting, underweight, and illness at the time of sampling, were assessed using a statistical model.

Neutralising antibodies to MV were measured in 652 plasma samples collected from 448 infants, of whom 13.6% were HIV-1 infected, 53.4% were HIV-seropositive but uninfected, and 33% were HIV-seronegative. There were no major differences among infants according to HIV exposure or infection status. However, HIV-infected infants were more likely to be underweight, stunted, undergoing wasting, or ill by comparison with HIV-seropositive but uninfected infants and HIV-seronegative infants.

Neutralising anti-MV antibodies were lowest in HIV-1-infected infants and highest in HIV-seronegative infants. This was confirmed by a statistical model which suggested that HIV-1–infected infants had lower levels of passively acquired antibodies to MV at birth than did HIV-seronegative infants. No factors other than HIV status was associated with antibody levels.

By six months of age, 91% of HIV-1–infected infants, 83% of HIV-seropositive but uninfected infants, and 58% of HIV-seronegative infants were estimated to have maternal antibody levels that were unlikely to interfere with measles vaccination. By nine months of age, 99% of all infants had non-protective antibody levels.

The findings of Scott et al demonstrate that HIV-infected infants born to HIV-seropositive Zambian women do not have maternally-acquired passive protection against MV and therefore run an increased risk of measles before routine measles vaccination at 9 months. The policy implication is that these children should be immunised by 6 months old, say the authors and the editorial.