Identification of individuals with gonorrhoea within sexual networks: a population-based study

Summary

Background

Molecular typing of Neisseria gonorrhoeae and contact tracing provide a combined approach for analysis of sexual networks in metropolitan areas, although there are some difficulties in application. Our aim was to examine the application of high-throughput molecular approaches that can identify individuals in linked sexual networks.

Methods

We characterised 2045 isolates of N gonorrhoeae from patients presenting at 13 major sexually transmitted infection clinics in London, UK, between June 1 and Nov 30, 2004. All isolates were assigned a sequence type (strain) on the basis of the sequences of internal fragments of two highly polymorphic loci, por and tbpB. These types were matched to demographic and behavioural data obtained at the clinic for each patient. We assessed the congruence in the demographic and behavioural characteristics of individuals infected with the same strain.

Findings

We identified 21 prevalent strains in this diverse gonococcal population, each infecting between 20 and 124 individuals. Seven of these strains were predominantly from men who have sex with men; the remaining 14 were predominantly from heterosexual people. No differences were recorded between the strains associated with men who have sex with men in the demographic or behavioural characteristics of infected individuals. By contrast, significant differences in age (p<0·0001), ethnicity (p=0·001), proportion of women (p=0·01), and HIV status (p=0·03) were noted between the 14 prevalent heterosexual-associated strains. Heterosexuals with strains not shared by others in the sample were significantly older (p=0·0005) and more likely to have had sex outside the UK (p<0·0001) than those sharing a strain with at least one other.

Interpretation

The discriminatory high throughput strain characterisation method applied here identified localised transmission networks and suggests little bridging between networks of men who have sex with men and heterosexual networks.

Introduction

The incidence of sexually transmitted infections (STIs) has risen substantially in the UK over the past decade, increasing the pressure on genitourinary services and resulting in increased morbidity.¹ This increase has been attributed to changes in sexual behaviour and risk-taking, together with increasing delays in access to treatment due to a crisis in sexual health service provision.1, 2 The prompt identification of cases of bacterial STIs and the notification and treatment of their sexual contacts (either through provider-led contact tracing or patient-led partner notification) has been the main public-health measure implemented to reduce infection.³ Identification and understanding of the nature of local sexual networks can help to target interventions towards places and people who are most at risk,3, 4 but is difficult, especially in major cities.5, 6

In London, UK, there are 32 genitourinary medicine clinics that are open access, require no referral, and provide free treatment for STIs. These clinics should provide fast and effective diagnosis and treatment to break transmission. However, the incidence of gonococcal infection remains high, and London has a disproportionate burden (45–50%) of all cases diagnosed in England and Wales.¹ Gonorrhoea incidence in London rose sharply from the mid-1990s onwards, and only started to level off from 2003. Overall, gonorrhoea diagnoses fell by 11% from 2003 to 2004, with the most pronounced decrease occurring in heterosexual men and women. By contrast, diagnoses in men who have sex with men increased by 7% over the same period.¹ Interventions to control STIs in London and many other major metropolitan areas are challenging; exhaustive contact tracing is unrealistic because patients may present to many different clinics, patients are unwilling or unable to identify many sexual partners, some of whom are anonymous, and partner notification in travellers and migrants to identify overseas partners might be difficult. Partner notification data therefore provide incomplete information on the nature and size of the major sexual networks, which limits attempts to control endemic infection.

Molecular epidemiology has been widely used to analyse outbreaks of disease in hospitals and the community⁷ and, for STIs, has the potential to identify individuals within the same sexual network.⁸ Provided that the molecular typing procedure is highly discriminatory, individuals who are not epidemiologically linked should be shown to be infected with different strains, whereas sexual contacts, and those within the same sexual network, should be shown to be infected with the same strain.⁸ For gonorrhoea, valuable information about the nature and size of sexual networks has been obtained by combining molecular typing with epidemiological data on sexual networks obtained through contact tracing, and this approach has elucidated sexual networks that would otherwise remain hidden. This approach has also identified larger networks than indicated by epidemiological data alone.9, 10, 11 However, so far the utility of molecular typing data to inform control has not been tested, mainly because data collection and analysis have been done retrospectively, and has not been attempted in a large metropolitan area with high rates of gonorrhoea.

We report here the use of a highly discriminatory typing method, Neisseria gonorrhoeae multi-antigen sequence typing (NG-MAST), which has been shown to identify epidemiologically linked individuals, but to distinguish between patients with no links.¹² The aim of this study was to test the feasibility of characterising isolates from a large proportion of cases of gonorrhoea in a metropolitan area and to identify clusters of linked patients in the absence of contact tracing data.