Does cytomegalovirus infection increase the risk of tuberculosis in UK children?

There is a hypothesised association between pre-existing cytomegalovirus (CMV) infection and risk of acquiring tuberculosis (TB). We aimed to explore if CMV seroprevalence and CMV IgG levels in children were associated with TB disease or Mycobacterium tuberculosis (Mtb) infection compared to children who were exposed to TB but remained well.  In this cross-sectional analysis from an observational cohort study of children exposed to TB in their household in the United Kingdom, we examined samples from seventy-five participants, of whom 40 (53%) were male. Median age of the cohort was 6 years (interquartile range: 3-11 years). Twenty-one (28%) children had TB disease, 27 (36%) had Mtb infection, and 27 (36%) had TB exposure only. There was no increased risk of TB in children who were CMV-seropositive (OR 2.18 (0.75-6.48)), and there were no differences in CMV IgG quantification by TB category. There was no detectable CMV viraemia in any of the children in our study. We found higher levels of CMV seroprevalence (49%) than previously described in the United Kingdom. In this small study of children exposed to TB, in a low TB burden setting, we found no association between CMV serostatus or CMV IgG titre and TB status.

Methods
Participants
Participants aged 0-16 years were recruited between January 2011 and December 2014 from 11 paediatric TB clinics in the United Kingdom, as part of the NIKS Cohort Study. In 2006, the National Institute for Health and Care Excellence (NICE) in the United Kingdom recommended that, following household TB exposure, children should be evaluated using tuberculin skin testing (TST). Those with positive TST results should undergo interferon-gamma release assays (IGRA) evaluation, and only those with positive results would be eligible for TB preventive treatment. The NIKS Cohort Study sought to evaluate the negative predictive performance of interferon-gamma release assays in TST-positive children. Serial serum samples were taken at baseline and over the period of follow-up (19). Children who had previously had a positive test for TB were excluded. Of the 392 participants of the NIKS Cohort Study, samples were only available from a subset of 75 children from six of the study sites (The Royal London, Birmingham, Bristol, St Mary’s, Northwick Park, and Southampton hospitals). Data available included age, sex, TB symptoms at first clinic appointment, TB diagnosis, TB treatment, Bacillus Calmette-Guérin (BCG) status, TST result, and IGRA results.

Study design

In this cross-sectional NIKS sub-study, samples were obtained from three clinical groups: children with TB disease, those with Mtb infection, and those with TB exposure (household TB contact, but no evidence of Mtb sensitisation). Evaluation for TB disease in the original NIKS Cohort Study for TB disease included history, examination, chest radiography, TST and IGRA tests, and microbiology if indicated. TB disease was diagnosed based on NICE guidelines for children above the age of 2 years, in conjunction with the clinical context, by paediatricians(19).

Mtb infection included children with a positive TST and/or a positive IGRA. A positive TST was defined as a transverse diameter of the induration was ≥6 mm in BCG-unvaccinated children and ≥15 mm in vaccinated children, in line with the 2006 NICE recommendations. The child was classified as IGRA positive if either the baseline or the 2-month IGRA was positive. Further details are available in the NIKS Cohort Study(19). Children who were a household TB contact and had no evidence of Mtb sensitisation had a negative TST and IGRA.

Samples

Serum samples were obtained from the participants’ first clinic visit. Samples had been stored at -80 degrees Celsius. Samples with <250 microlitres were excluded due to insufficient sample remaining for both polymerase chain reaction (PCR) and enzyme-linked immunosorbent assay (ELISA) to be carried out.

Laboratory analyses

CMV-specific IgG was measured using GenWay Biotech Cytomegalovirus IgG ELISA as per the manufacturer’s instructions, and optical density was used to estimate the quantity of CMV-specific IgG using the VersaMax ELISA Microplate reader. Seropositivity was defined as a CMV IgG detected at >1.1 IU/ml. QIAGEN QIAamp® Blood Kit was used to extract DNA from serum, and CMV DNA was quantified by real-time PCR using the BioMérieux ARGENE CMV R-GENE® kit. Samples were run in triplicate on the Applied Biosystems StepOnePlus Real-Time PCR system. The maximum sensitivity of the assay was 446 copies per ml. Presence of CMV IgG in serum was taken to represent any time exposure to CMV, and presence of CMV DNA in serum was defined as acute infection.

Statistical analysis

StataTM 16.1 was used for all analyses. A Kruskal-Wallis equality of proportions rank test was carried out for CMV IgG level and evaluated children with TB disease, Mtb infection, and TB exposure as three independent groups. An odds ratio was performed to determine the association between CMV seropositivity and TB clinical state.