Descriptive epidemiology of rubella disease and associated virus strains in Uganda

Abstract Rubella virus causes a mild disease; however, infection during the first trimester of pregnancy may lead to congenital rubella syndrome (CRS) in over 80% of affected pregnancies. Vaccination is recommended and has been shown to effectively reduce CRS incidence. Uganda plans to introduce routine rubella vaccination in 2019. The World Health Organization recommends assessing the disease burden and obtaining the baseline molecular virological data before vaccine introduction. Sera collected during case‐based measles surveillance from January 2005 to July 2018 were tested for rubella immunoglobulin M (IgM) antibodies. Sera from confirmed rubella outbreaks from January 2012 to August 2017 were screened using real‐time reverse‐transcription polymerase chain reaction (RT‐PCR); for positive samples, a region within the E1 glycoprotein coding region was amplified and sequenced. Of the 23 196 suspected measles cases serologically tested in parallel for measles and rubella, 5334 (23%) were rubella IgM‐positive of which 2710 (50.8%) cases were females with 2609 (96.3%) below 15 years of age. Rubella IgM‐positive cases were distributed throughout the country and the highest number was detected in April, August, and November. Eighteen (18%) of the 100 sera screened were real‐time RT‐PCR‐positive of which eight (44.4%) were successfully sequenced and genotypes 1G and 2B were identified. This study reports on the seroprevalence and molecular epidemiology of rubella. Increased knowledge of former and current rubella viruses circulating in Uganda will enhance efforts to monitor the impact of vaccination as Uganda moves toward control and elimination of rubella and CRS.

These requirements are often hard to achieve through the routine surveillance for measles and rubella. Alternative samples such as serum have been shown to provide molecular virological data retrospectively 10,11 and are a potential source of baseline data on the genetic diversity of the rubella viruses in the prevaccine era. Rubella virus RNA may be detected in acute phase serum specimens (0-3 days from rash onset) that are rubella IgM-positive by using the polymerase chain reaction and positive samples can be genotyped. 10,11 We report on the rubella disease burden from January 2005 to July 2018 using case-based measles surveillance data as well as the molecular virological data using archived serum samples from confirmed rubella outbreaks from January 2012 to August 2017. This documentation will provide baseline information on the epidemiology of rubella before rubella vaccine introduction in Uganda.

| MATERIALS AND METHODS
The Regional Measles Reference laboratory in Uganda has been a member of the Global WHO Measles and Rubella Laboratory Network 12 since 2000 and the rubella cases in this study were identified through measles case-based surveillance following guidelines provided by the WHO Regional Office for Africa. 13 Based on these guidelines, a suspected measles case is defined as "any person with generalized maculopapular rash and fever plus one of the following, cough or coryza (runny nose) or conjunctivitis (red eyes) or any person in whom a clinician suspects measles." Individuals fitting these criteria and who presented at health facilities within 30 days of rash onset were investigated using a standard case investigation form for demographic and clinical history data and a serum specimen collected for laboratory analysis. When five or more cases were investigated in a health facility or district in a month, this was classified as a suspected outbreak. Specimens were transported to the laboratory under cold chain.

| Enzyme-linked immunosorbent assay
Once in the laboratory, sera were tested by both the Enzygnost ® Serum samples from confirmed rubella outbreaks collected between January 2012 and August 2017 that tested rubella IgMpositive and had been collected 0 to 2 days from rash onset were selected for this study. These samples that were being stored at −20°C were retrieved from storage, thawed, and 400 µl aliquots

| Genotyping and sequencing assays
RNA from all real-time RT-PCR positives were added to a nested RT-PCR assay with Superscript III One-Step Platinum Taq kit (Invitrogen) as described by Pukuta et al. 11 In brief, 2.5 µl of the extracted RNA was run in a 25 µl reaction containing 12.5 µl of 2× reaction buffer, 0.25 µl of each of the forward and reverse primers (Table 1) For the 2nd round, 1 µl of the 1st round PCR product was used with the primers in Table 1  were used.

| Phylogenetic analyses
Analyses of the sequences were performed using Sequencher

Regional distribution of investigated cases
Central (10 211

| Rubella virus sequence distribution and phylogenetic analysis
Of the 100 sera tested at CDC, 18 were positive by real-time PCR. For eight of these, the 739-nt window of the E1 glycoprotein was successfully amplified and sequenced. Two genotypes were detected, 1G (three sequences) and 2B (five sequences). The majority of these (5 out of 8; 62.5%) were obtained from cases whose serum had been collected within a day of rash onset. Genotype 1G was detected in 2012 and 2016 and genotype 2B in 2014, 2015, and 2017 ( sequences belonged to sublineage 2B-L2c (Figure 3).

| DISCUSSION
The present study documents that rubella in Uganda is a childhood disease affecting mainly those below 15 years of age. This is in agreement with previously published findings. 4 Combined data from the current and a previous report of rubella in Uganda support a 3-to 4-year rubella cycle in Uganda. 8 21.5% RT-PCR positive and 7.5% genotyped) providing additional evidence that this is an effective method for retrospective genotype baseline determination for rubella virus. 10