Chlamydia trachomatis serology as a means of monitoring intervention activities to eliminate trachoma as a public health problem

Abstract

Trachoma causes blindness as a result of repeated ocular infection with Chlamydia trachomatis (Ct). International efforts are focused on eliminating trachoma as a public health problem by 2020; the World Health Organisation’s (WHO) and the Global Alliance for the Elimination of Trachoma by 2020 (GET2020) (1) recommend the SAFE strategy (Surgery to treat trichiasis, Antibiotics to treat infection with Ct, Facial cleanliness and Environmental change to reduce transmission) (2). Current implementation guidelines for the A, F and E aspects of the strategy are based on the prevalence of trachomatous inflammation-follicular (TF) in children aged 1-9 years. Antibiotics are administered as azithromycin mass drug administration (MDA) to all residents of districts wherein the prevalence of TF in 1-9 year old children is greater than 10% (3). As we approach the global elimination of trachoma, the prevalence of TF in children will decline towards the elimination threshold of 5%, and with it, the positive predictive value of TF. This may result in inappropriate, continued administration of antibiotics, which is a mis-use of valuable financial and person resources, as well as raising concerns about antimicrobial stewardship. Once the prevalence of TF in children aged 1-9 years is below 5%, re-emergence must be monitored. The WHO recommends a ‘pre-validation’ survey to determine if re-emergence has occurred (4). The 2014 Technical Consultation on Trachoma Surveillance (4) recommended exploring the district-level prevalence of TF, the district-level prevalence of conjunctival infection with Ct and the district-level prevalence of antibodies against Ct, to determine an appropriate measure, or combination of measures, for deciding when to stop MDA. Numerous methods exist for detecting antibodies against Ct antigens (5–9). Specimens may be collected by venepuncture or from fingerprick, stored as whole blood, serum or dried blood spots on filter paper, and assayed using enzyme-linked immunosorbent assays (ELISA), lateral flow assays (LFA), or multiplex bead assay (MBA); of theses, MBA allows for the testing for antibodies against numerous antigens from endemic infections, but the required instrumentation and reagents are more cost-intensive. Historically, serology for Ct fell out of favour due to the high cross-reaction with antigens from C pneumoniae, however, with the advent of new proteomic tools and assays with greater reported sensitivity and specificity, serology has re-emerged as a potential tool for monitoring the prevalence of antibody-inducing chlamydial infection (4,10,11). Both ELISA and MBA assays present output as numerical data, which must be translated to a population seroprevalence value. Therefore, a method must be determined for dichotomising numerical data and setting a threshold between seropositive and seronegative samples. Receiver 5 Operating Characteristic (ROC) curves have previously been used, but these rely on appropriate reference standards. Alternate methods that rely solely on the data generated within a study, such as finite mixture modelling, may be more appropriate (12–15). Once seropositivity has been estimated, it is of interest to detect changes in the age-specific seroprevalence of a population. Catalytic models have previously been used to monitor changes in the prevalence of malaria (13,16–22) and this methodology can be applied to detect changes in transmission as a proxy for the force of infection (FOI) as well as to estimate seroconversion rates, and to a lesser extent, seroreversion rates (13,19). The simplicity of dichotomous seroprevalence estimates may provide lower resolution information than the quantitative antibody levels. Recent work uses quantitative antibody levels to measure changes in transmission (23). Ensemble machine learning can be used to produce characteristic agespecific antibody curves that may reveal changes in population mean antibody levels that would otherwise be masked because changes occurred above or below the seropositivity threshold. There remain several areas to be elucidated. While antigens used in modern serological studies are considered specific for Ct, there is yet no way to distinguish between antibodies due to ocular infection or genital infection, making serological studies in anyone over the age of sexual debut a challenge. There is no standard reference for antibody levels, making a comparison between different studies challenging. The two most commonly used assays produce different output data: ELISAs measures optical density- the amount of light absorbed by the specimen- while MBAs detect fluorescence in an assayed specimen. An internal standard would allow for comparison between the two assays. This PhD research addresses several key questions about the use of serology and Ct-specific antibodies for monitoring the prevalence of trachoma. As more countries progress to eliminating trachoma as a public health problem by 2020, efforts will need to be increased to monitor and evaluate elimination efforts and to prevent re-emergence of the disease. Serological techniques may be ideal for such activities.