IntroductionMeasles and congenital rubella syndrome remain significant causes of morbidity and mortality among children, particularly in sub-Saharan Africa.1 While safe and effective vaccines are available, high levels of vaccine coverage are required to interrupt measles virus transmission and few sub-Saharan African countries have introduced rubella vaccine into their routine immunization schedule as of 2015.2 Plans for rubella vaccine introduction in 49 countries by the end of 2020 are underway.3
Large outbreaks of measles have occurred recently,4–6 and in sub-Saharan Africa they increasingly involve older children and adolescents.7 Many factors contribute to these outbreaks, including HIV infection. HIV-infected children, the majority of whom live in sub-Saharan Africa,8 can have poorer primary measles vaccine responses and a higher likelihood of waning immunity than HIV-uninfected children in the absence of treatment,9–12 and therefore can remain susceptible to measles despite vaccination. Until recently, the impact of HIV infection on the buildup of measles susceptibles was limited by the high mortality rate of HIV-infected infants.13,14 However, with increased access to life-prolonging therapies,15 more HIV-infected children survive into adolescence and adulthood, thus creating pockets of susceptible individuals that may sustain measles virus transmission and jeopardize elimination efforts.14
While antiretroviral treatment (ART) reverses some of the immunologic damage due to HIV infection,16,17 immunity to vaccine-preventable diseases does not appear to be restored,18,19 as immune reconstitution in children primarily occurs with naïve CD4+ T cells20–22 and abnormalities in B cell function persist.23,24 Consequently, these children may remain susceptible to measles and need revaccination to restore protective immunity. Recently, the World Health Organization’s Strategic Advisory Group of Experts (SAGE) recommended measles revaccination for HIV-infected children receiving highly active antiretroviral therapy following immune reconstitution.25
Limited data are available on immunity to vaccine-preventable diseases among older HIV-infected children and adolescents with and without ART,26–29 particularly in sub-Saharan Africa.30 These data will be needed to guide vaccine policy on revaccination against measles and catch-up rubella vaccination. Older children are not traditionally targeted by immunization programs but are an increasingly important age group, particularly as countries consider wider age ranges for supplemental immunization activities with measles and measles-rubella containing vaccines. For rubella, failure to target older girls in susceptible age groups could lead to an increase in congenital rubella cases.31
The objectives of this study were to estimate and compare the prevalence of immunity to measles and rubella viruses among HIV-infected treatment-naïve youth, HIV-infected treatment-experienced youth, and HIV-uninfected youth in rural Zambia.
Material and MethodsStudy Setting and PopulationThe study was conducted in the catchment area of Macha Hospital in Choma District, Southern Province, Zambia. This area is populated by traditional villagers living in scattered homesteads, characteristic of much of rural sub-Saharan Africa. The prevalence of HIV infection in Choma District was estimated to be 15.7% in 2010.32 Zambia had a large measles outbreak in 2010–2011 but has since reported few cases after a nationwide supplementary measles immunization campaign in 2012 targeting children from 9 months to 14 years of age. Reported vaccine coverage in Southern Province was 86% for measles and 69% were estimated to be fully vaccinated (BCG, measles, and three doses each of DPT-HepB-Hib and polio vaccine) in 2013.33 Zambia plans to introduce rubella vaccine in 2016.
Patients and SamplesSamples for testing antibodies to measles and rubella viruses were selected from participants within two ongoing studies conducted in Macha from 2009–2013. Three groups of youth 5–15 years of age were selected for comparison: 1) HIV-uninfected youth; 2) HIV-infected treatment-naïve youth; and 3) HIV-infected youth receiving ART. Measles vaccination status of the study children was not known.
All studies were approved by the Institutional Review Board at the Johns Hopkins Bloomberg School of Public Health, a local research ethics committee and the Government of Zambia Ministry of Health.
Source of HIV-uninfected youthIn 2007, community-based serial cross-sectional surveys were initiated around Macha Hospital to determine changes in the prevalence of malaria parasitemia.34,35 Satellite images were used to construct a sampling frame and approximately 24 households were randomly selected to be enrolled in each cross-sectional survey conducted every other month. All individuals residing within a household were eligible to participate. Permission was obtained from the head of household and written informed consent was obtained from the individual or parent. For each participant, a questionnaire was administered to collect information on demographics, health-seeking behavior and use of medications, and a blood sample was collected by finger prick on filter paper (Whatman, Protein Saver card 903, Piscateway, New Jersey), dried overnight and stored individually with desiccant in a sealed plastic bag at −20°C.
For this study, samples were randomly selected from participants 5–15 years of age who enrolled from 2009–2013, had at least three filter paper spots available, and did not report use of ART. The target sample size was four HIV-uninfected youth per HIV-infected youth. HIV testing was not conducted as part of the study and all participants not reporting ART were presumed HIV-uninfected given the low prevalence of HIV infection expected in a population-based sample of youth.
Source of HIV-infected youthIn 2007, a cohort study was initiated and all HIV-infected children younger than 16 years attending the Macha HIV clinic were eligible for enrollment.36,37 After obtaining written informed consent from a parent or guardian and assent from children 8–15 years of age, children were seen at study visits every three months, at which time a questionnaire was administered, the child was examined and a blood sample was collected. Beginning in 2009, blood spots were collected on filter paper (Whatman, Protein Saver card 903) at enrollment and after ART initiation. The blood spot cards were dried overnight and stored individually with desiccant in a sealed plastic bag at −50°C.
For the group of HIV-infected treatment-naïve youth, all available samples were selected from the 163 treatment-naïve youth 5–15 years of age at enrollment from 2009–2013. For the group of HIV-infected youth receiving ART, the first available sample was selected from the 186 youth 5–15 years of age who had been receiving ART for 9 to 24 months from 2009–2013. Thirty-five youth contributed samples to both groups.
To evaluate the impact of ART on measles and rubella immunity, pairs of samples were obtained for children contributing to both the treatment-naïve and ART groups. Additional samples were selected for HIV-infected youth receiving ART who had a pre-ART sample available at study enrollment prior to 5 years of age.
Laboratory MethodsSerum eluted from dried blood spots (DBS) was tested by enzyme immunoassay (EIA) for IgG antibodies to measles and rubella virus (Enzygnost, Siemens, Germany) at the Clinical Research Laboratory at Macha Research Trust. Two dried blood spots per participant were used. Samples from each of the three groups were randomly selected for testing on each plate. Serum was eluted from DBS using 250 µL of elution buffer according to protocol 6 outlined by Mercader et al.38 50 µL of eluted sample was transferred to labeled 96-well plates and sealed. Samples not tested immediately were stored at 4°C overnight for testing the following day. The EIA was performed according to the manufacturer’s protocol. The optical density (OD) for each sample was read at 450 nM using a BioTek ELx800 microplate reader. Corrected OD differences (cOD) were calculated using the OD differences (antigen-control well), the mean of the positive and negative references and the nominal value as specified by the manufacturer.
Volume adjustmentDBS from a known measles and rubella positive individual were eluted and tested for IgG antibodies to measles and rubella virus using serum elution volumes ranging from 5µL to 50µL. Linear regression was used to determine the effect of volume on OD value (estimated mean increase of 0.03 in OD value per 1µL increase in volume for the antigen well). The results were the same for measles and rubella tests. The regression coefficient was used to calculate adjusted OD values for the antigen well.
Statistical AnalysisTen percent of measles samples and 3% of rubella samples had less than the required 50 µL of sample and adjusted cOD values were calculated (results based on unadjusted cOD values are presented in Supplemental Digital Content 1). For rubella, samples with cOD values ≥2.5 (12% of samples) were assigned a value of 2.5. For all samples with cOD values >0.1, measles (mIU/mL) and rubella antibody concentrations (IU/mL) were calculated using the alpha method, as specified by the manufacturer. cOD values between 0.1 and 0.2 were considered equivocal and cOD values >0.2 were considered positive for antibodies to measles and rubella viruses. A second definition was also used in which cOD values ≥0.1 were considered positive.
Measles and rubella seroprevalence was compared between the three groups of HIV-infected and uninfected youth using chi-square tests and log-binomial regression to adjust for age. Antibody concentrations were compared using Wilcoxon rank sum tests. Risk factors for susceptibility (cOD≤0.2) were assessed separately for HIV-infected treatment-naïve youth and youth receiving ART using log-binomial regression.
DiscussionIn this cross-sectional study, HIV-infected youth in rural Zambia were less likely to be seropositive for measles and rubella viruses than HIV-uninfected youth. Based on these results, approximately 25% and 50% of HIV-infected youth in this area were susceptible to measles and rubella viruses, respectively. The more substantial differences in measles seroprevalence likely reflect higher rates of primary or secondary measles vaccine failure, whereas the smaller differences in rubella seroprevalence may reflect waning immunity following wild-type virus infection.
Increasing numbers of HIV-infected youth could contribute to lower levels of population immunity in areas with a high HIV prevalence.14 Similar to other studies,26,39 HIV-infected youth in this study were less likely to be seropositive for measles virus than HIV-uninfected youth, despite presumed vaccination during childhood, possible revaccination during the supplemental immunization activities, and immune reconstitution with ART. Lower measles vaccine coverage among HIV-infected youth is an alternative explanation, although this is unlikely given that these children were receiving intensive care. In addition, lower antibody concentrations were found in this group, leading to a higher proportion of youth in the equivocal range compared to uninfected youth. It is not possible to determine if this was due to poorer response to primary vaccination or waning immunity over time. It is also unclear whether these children would develop clinical measles or rubella upon re-exposure as they may have an anamnestic response or cellular immunity that would confer some protection.18 These results support existing studies suggesting that ART does not restore immunity to measles virus and that HIV-infected children and adolescents could benefit from revaccination,18,19,40 as recently recommended by the World Health Organization.25 This group could easily be targeted for revaccination through routine care, as they are frequently seen at healthcare facilities, or through inclusion of youth in supplemental immunization activities.
All HIV-infected youth in this study initiated ART after the age at which they would have received measles vaccination. With early infant diagnosis more widely available, infants are now initiating ART before their first dose of measles vaccine. Although data are limited, immune responses among these infants appear to be better and more similar to those observed among healthy HIV-uninfected infants.41,42 Early initiation of ART preserves normal development and maintenance of memory B cells,42 and these improvements appear to be sustained into adolescence.26 As more children initiate ART in infancy, their immune function and responses to vaccines will need to be documented to guide policies.
In this study, approximately half of all adolescents, regardless of HIV or ART status were susceptible to rubella virus. As rubella vaccines have not yet been introduced in Zambia, immunity in this study was acquired through exposure to wild-type virus. As expected, the proportion seropositive increased with age, but was lower than previous reports in the region.43–45 This may be due to differences in urban and rural populations, with lower population densities and risk of exposure in rural areas. Introduction of measles-rubella vaccines may increase the age of infection,31 therefore a large number of susceptible youth is a cause for concern. This group may need to be targeted for catch-up immunization campaigns when introducing measles-rubella vaccines into the national immunization program.
This study had several limitations. First, youth in the control group were not tested for HIV and were presumed uninfected. While we expect the HIV prevalence among youth in the general population in this area to be low, it is possible that some youth were infected with HIV and that the measles and rubella seroprevalence in HIV-infected youth was underestimated. Second, information on vaccination status was not known. Antibodies to measles virus were presumed to be primarily vaccine induced but some children may have been exposed to wild-type virus. In addition, supplemental immunization activities occurred before and during the study period, therefore antibodies levels may be the result of multiple vaccinations. As described above, lower measles vaccine coverage among HIV-infected youth could also explain the differences in seroprevalence. Regardless of the mechanism, these children would likely benefit from revaccination. Third, all HIV-infected youth survived to enter the study and initiate ART at older ages. Their experiences may be different from children with more rapid disease progression and children initiating ART in infancy. Fourth, measles antibody levels were measured by EIA rather than the gold standard plaque reduction neutralization assay. While these two assays have good agreement, EIAs are less sensitive at lower antibody levels,46 which may have led us to underestimate measles seroprevalence among all groups. Fifth, DBS were used that had been stored for up to four years. Several studies have documented the concordance of results for antibody levels measured from DBS and serum,47–49 even after being stored at −20°C for up to two years.50,51 It is possible that longer term storage resulted in some sample degradation, which may also have led to underestimation of measles and rubella seroprevalence among all groups. Lastly, the number of HIV-infected youth in each treatment group was small, which limited our ability to evaluate risk factors for susceptibility.
Global efforts have succeeded in decreasing morbidity and mortality related to measles and rubella viruses. Large outbreaks in the past few years, however, have demonstrated that high levels of population immunity are required and pockets of susceptible individuals can jeopardize control efforts. HIV-infected youth represent an important group that needs to be considered in these control efforts. These results support recommendations to revaccinate HIV-infected children against measles after initiation of ART and for inclusion of youth in supplemental immunization activities and catch-up campaigns for both measles and rubella.