Low Prevalence of Conjunctival Infection with Chlamydia trachomatis in a Treatment-Naïve Trachoma-Endemic Region of the Solomon Islands

Abstract

Background: Trachoma is endemic in several Pacific Island states. Recent surveys across the Solomon Islands indicated that whilst trachomatous inflammation-follicular (TF) was present at levels warranting intervention, the prevalence of trachomatous trichiasis (TT) was low. We set out to determine the relationship between chlamydial infection and trachoma in this population.

Methods: We conducted a population-based trachoma prevalence survey of 3674 individuals from two Solomon Islands provinces. Participants were examined for clinical signs of trachoma. Conjunctival swabs were collected from all children aged 1-9 years. We tested swabs for Chlamydia trachomatis (Ct) DNA using droplet digital PCR. Chlamydial DNA from positive swabs was enriched and sequenced for use in phylogenetic analysis.

Results: We observed a moderate prevalence of TF in children aged 1-9 years (n = 296/1135, 26.1%) but low prevalence of trachomatous inflammation-intense (TI) (n = 2/1135, 0.2%) and current Ct infection (n = 13/1002, 1.3%) in children aged 1-9 years, and TT in those aged 15+ years (n = 2/2061, 0.1%). Ten of 13 (76.9%) cases of infection were in persons with TF or TI (p = 0.0005). Sequence analysis of the Ct-positive samples yielded 5/13 (38%) complete (>95% coverage of reference) genome sequences, and 8/13 complete plasmid sequences. Complete sequences all aligned most closely to ocular serovar reference strains.

Discussion: The low prevalence of TT, TI and Ct infection that we observed are incongruent with the high proportion of children exhibiting signs of TF. TF is present at levels that apparently warrant intervention, but the scarcity of other signs of trachoma indicates the phenotype is mild and may not pose a significant public health threat. Our data suggest that, whilst conjunctival Ct infection appears to be present in the region, it is present at levels that are unlikely to be the dominant driving force for TF in the population. This could be one reason for the low prevalence of TT observed during the study.

Fig 1

Photographs of conjunctivae showing (A) no evidence of active trachoma, (B) mild trachomatous inflammation-follicular (TF) and (C) more severe TF. All three photographs were taken of conjunctival C. trachomatis infection-negative children aged 1–9 resident in Temotu, Rennell or Bellona, Solomon Islands, October-November 2013, in whom the photo grade and field grade matched.

Fig 2. Age-specific prevalence (grey bars) and 95% confidence interval (arrows) of trachomatous inflammation—follicular (TF) in individuals aged 1–15 years, recorded during a trachoma survey of Temotu, Rennell and Bellona, Solomon Islands, October-November 2013.

Fig 3

Maximum likelihood phylogram of (A) genome and (B) plasmid sequences from clinical specimens collected in the Solomon Islands in October and November 2013, assembled to C. trachomatis A/HAR-13 and B/Jali20/OT reference, respectively. All branches had bootstrap values over 85/100.

Fig 4. Comparison of unadjusted prevalence of trachomatous trichiasis (TT) in 15+ year olds, and trachomatous inflammation—follicular (TF) in 1–9 year-olds, in treatment-naïve trachoma-endemic EUs for which data have previously been published (n = 58), and in which the TF prevalence in 1–9 year-olds was 10–40%.

Correlation coefficient (R) is 0.40 for this subset, but 0.77 if including studies from areas of any prevalence.

Fig 5. Published age-specific trachomatous inflammation—follicular (TF) prevalence data in studies undertaken in districts with >10% TF in overall child population (dotted and dashed lines), compared to the same in Temotu, Rennell and Bellona, Solomon Islands, October-November 2013 (grey columns).

Fig 6. Relationship between the prevalence of conjunctival C. trachomatis infection (diagnosed by nucleic acid amplification test) and trachomatous inflammation—follicular (TF) in the total 0–9 year-old population or a subset of that group at the district level (n = 35 districts).

Correlation coefficient (R) is 0.84.