Ceftriaxone-resistant Neisseria gonorrhoeae detected in England, 2015–24: an observational analysis

Introduction

Gonorrhoea is the second most common sexually transmitted infection (STI) in England, with over 85 000 diagnoses in 2023.¹ Untreated gonorrhoea can lead to pelvic inflammatory disease, infertility and ectopic pregnancy, and can increase the risk of HIV transmission. Neisseria gonorrhoeae, the causative pathogen of gonorrhoea, has developed resistance to successive classes of antibiotics, and few antimicrobials remain effective in its treatment.^2,3 Extended-spectrum cephalosporins, such as ceftriaxone, are the last-line option for empirical monotherapy.² The first ceftriaxone-resistant gonococcal strain, H041, was detected in Japan in 2009⁴ but failed to disseminate further. Four more ceftriaxone-resistant N. gonorrhoeae presented a similar picture from 2010 to 2014;^5–8 strains were detected but failed to disseminate. The exception was strain F89, which was first isolated in France during 2010⁵ and was subsequently isolated in Spain in 2011,⁶ albeit with undetected cases along transmission chains. Unfortunately, the sporadic nature of ceftriaxone-resistant N. gonorrhoeae was relatively short-lived, with the emergence of the FC428 strain in Japan during 2015.⁹ The FC428 strain is associated with ceftriaxone resistance determined by the mosaic penA-60.001 allele that encodes the gonococcal PBP 2. Strains that cluster within the FC428 clone have been detected in numerous countries in the Asia-Pacific region including China, Cambodia and Vietnam, as well as in Canada and across Europe,^10,11 with the first case detected in the UK in 2018.¹² The penA-60.001 allele has also been detected in different genomic backbones,^10,11,13 suggesting independent recombination events with a penA-60.001 donor, such as commensal Neisseria subflava,¹⁴ or recombination with other FC428-like isolates. The first ceftriaxone-resistant N. gonorrhoeae strain with high-level resistance to azithromycin (MIC ≥ 256 mg/L) was isolated during 2018 in both the UK and Australia. This phenotype is considered to be XDR, although there is no internationally agreed definition of XDR N. gonorrhoeae. These isolates harboured the penA-60.001 allele, but were unrelated to the FC428 clone.^13,15 There were no identified links between the individuals, thus highlighting the extent of antimicrobial resistance (AMR) surveillance gaps.¹⁵

Previously we reported a concerning increase in ceftriaxone-resistant N. gonorrhoeae cases in the UK, with 10 cases detected in the 6 month period between December 2021 and June 2022, compared with 9 cases detected between December 2015 and September 2021.¹⁶ Most of these cases were associated with travel to or from the Asia-Pacific region, and all were heterosexual individuals. A further 15 cases were detected in England between June 2022 and May 2024, including 5 cases that were XDR, marking the highest number of XDR cases reported in any European country to date. Here we describe the demographic and clinical details of these most recent cases, and phenotypic and molecular characteristics of the isolates. For a comprehensive overview, we have also included details for 16 ceftriaxone-resistant cases previously identified in England, 12 of which have been published,^12,13,16,17 and performed a global genomic comparison of all publicly available ceftriaxone-resistant N. gonorrhoeae strains with mosaic penA alleles.

Methods

Results

Discussion

To date, all individuals with ceftriaxone-resistant gonorrhoea detected in England have been heterosexual, mostly in their 20s and have usually had travel links with the Asia-Pacific region. Although not all partners could be traced, fortunately there appears to have been limited onward transmission within England. This is possibly because most cases are travel-associated and are not part of dense sexual networks, with termination of the transmission chain. However, the frequency of case detection has increased rapidly since 2021, without any changes in clinical practice or surveillance in England, and may possibly reflect the increase in antimicrobial-resistant N. gonorrhoeae in the Asia-Pacific region.²¹ Most notably, there has been an increase of XDR cases, with an unprecedented figure of five cases detected in a recent 6 month period (November 2023–May 2024), compared with the total of seven XDR cases detected in England since 2018.

Ceftriaxone resistance in England is still mainly conferred by the penA-60.001 allele, although the STs of isolates are diverse, and tend to cluster with isolates identified in countries of the Asia-Pacific region, consistent with the reported travel links. Ceftriaxone-resistant N. gonorrhoeae has not yet been detected in the gay or bisexual population in England. Heightened surveillance and public awareness activities are currently underway to identify cases as soon as they emerge, with the aim to interrupt the transmission of cases that can impose challenging treatment situations.

Most of the globally sequenced ceftriaxone-resistant isolates have belonged to the international FC428 clone. However, the phylogenetic analysis revealed multiple other phylogroups comprising isolates bearing mosaic penA alleles, which supports the evidence that resistance to ceftriaxone has emerged across multiple genetic backgrounds^10,11,13 that might be even more diverse than reported in this analysis. Early ceftriaxone-resistant isolates (identified between 2009 and 2015), including those from England, expressed various mosaic penA variants,^4–8 but these were later dominated by those carrying the penA-60.001 allele. More recently, ST7363 isolates carrying penA-232.001 (which were genetically related to H041⁴ and A8806⁸ isolates reported in Japan and Australia between 2009 and 2013) have been reported in China,²² but none was detected in England. In contrast, ST1901 isolates carrying penA-237, first reported in France in 2022,²³ were detected in two English isolates belonging to ST1901 and its SLV ST1579 (H22-722 and H23-343). Genomic analysis showed that the XDR isolates remained limited to ST16406 and its SLVs, first reported in the UK in 2018 in a heterosexual male after sexual contact in Thailand. These findings suggest that XDR N. gonorrhoeae isolates might be currently linked, and possibly restricted, to the global dissemination of this clade, although a high degree of diversity is still present within the XDR clade IV established from our analysis.

The prevalence of N. gonorrhoeae with ceftriaxone resistance in China was reported to be 8.1% in 2022, with five provinces reporting >10%.²¹ The Enhanced Gonococcal Antimicrobial Surveillance Programme (EGASP) in Cambodia recently reported ceftriaxone MICs of ≥0.125 mg/L in 29 isolates that harboured the penA-60.001 allele distributed across nine different MLST STs detected during 2021–22.¹¹ Three were ST-16406 that also displayed high-level azithromycin resistance, i.e. were XDR.¹¹ Worryingly, the Cambodian situation continues to worsen, with 15.4% ceftriaxone resistance and 6.2% XDR detected during 2022–23.²⁴

In our English case series, all genital infections were successfully treated with ceftriaxone, despite the isolates being categorized as resistant to ceftriaxone, according to EUCAST breakpoints.¹⁹ However, some patients received dual therapy with 1 g azithromycin, and so we cannot know for certain that these infections would have cleared with ceftriaxone alone. Our dataset is currently the most comprehensive available, with full treatment outcome and site of infection data from individuals who have been infected with ceftriaxone-resistant N. gonorrhoeae, and were treated using recommended regimens. There are very few data on the frequency of treatment failures in regions with a high prevalence of ceftriaxone resistance. One study from China of 1686 patients with uncomplicated gonorrhoea found that all patients were cured with ceftriaxone, even though nearly 10% had an isolate with decreased susceptibility to ceftriaxone.²⁵ However, it was noted that 72.7% of patients in the study were treated with a higher-than-standard dosage (>1 g), and it was suggested that this may be because clinicians were concerned that Chinese manufactured ceftriaxone was less potent than drug manufactured elsewhere.²⁵

It is concerning that in our analysis, three of five pharyngeal infections and the only rectal infection failed treatment. It may be appropriate to reconsider clinical breakpoints according to the site of infection, with a higher MIC breakpoint introduced for genital infections. There is no clinical breakpoint or ECOFF for ertapenem, but it has been shown that for isolates with raised ceftriaxone MICs, the ertapenem MIC is lower,²⁶ as was the case for all 29 isolates described in this paper. However, this is not universal, particularly in the presence of a penA mosaic allele.²⁷ In a recent clinical trial, ertapenem 1 g IM was shown to be non-inferior to ceftriaxone 1 g IM, but the isolates in this trial were susceptible to ceftriaxone.²⁸ To note, in the UK, ertapenem is licensed for IV but not IM use. In 2018, ertapenem 1 g IV for 3 days was used to treat two cases of ceftriaxone treatment failure;¹³ however, the most recent treatment failure in 2024 within this current analysis was successfully treated with a single dose of ertapenem 1 g IV. Single-dose treatment is preferable for both provider and patient. Although all isolates were susceptible to spectinomycin and had low gentamicin MICs, neither antibiotic is effective for treating pharyngeal infections.¹⁸ Additionally, spectinomycin is no longer available in many countries, including the UK.

The UK gonorrhoea management guideline recommends universal TOC at 2–3 weeks after treatment;¹⁸ settings where TOC is not performed will not detect asymptomatic treatment failures. Pharyngeal infection is recognized as being harder to treat, possibly because tissue penetration of antimicrobials is limited;² the UK guideline recommends that all individuals with epidemiological links to the Asia-Pacific region, or anyone with urogenital ceftriaxone-resistant N. gonorrhoeae, should have a pharyngeal NAAT and culture taken prior to treatment.¹⁸ As described in this analysis, ceftriaxone 1 g IM can be expected to clear urogenital infection, but pharyngeal infection may persist and may be a potential source for onward transmission. Without testing the pharynx, these asymptomatic infections will remain undetected. National guidelines of most countries do not recommend routine testing of the pharynx in heterosexual individuals; the prevalence of pharyngeal gonorrhoea in heterosexuals is largely unknown, and routes of transmission remain ill-defined. It remains unclear whether kissing is a factor in gonorrhoea transmission.²⁹ In this analysis, the four heterosexual men with pharyngeal infection denied having oral sex. Studies have reported a high prevalence of pharyngeal gonorrhoea in heterosexual men and women who were contacts of gonorrhoea cases, and that oral sex is not associated with pharyngeal infection in heterosexual men.³⁰

It is likely that the detection of these cases in England represents just the tip of the iceberg of gonococcal AMR, given the large surveillance gaps globally. England has a comprehensive surveillance system including both sentinel and real-time detection of gonococcal AMR, enabled by the capacity to culture all gonococcal infections and perform universal TOC. Given that nearly all our cases had epidemiological links to the Asia-Pacific region, and that countries in the region are reporting worryingly high levels of ceftriaxone resistance,^21,24 it is essential that action is taken at a global level. This includes access to rapid testing, quality-assured antimicrobial susceptibility testing, robust surveillance, and monitoring of treatment failures to maintain the viability of the current treatment options. Ultimately, new treatment options along with robust antimicrobial stewardship and innovative control strategies such as resistance-guided therapy are required to ensure gonorrhoea remains a treatable infection.

Supplementary Material

Funding

This work was supported by the United Kingdom Health Security Agency.

Transparency declarations

All authors declare no competing interests.

Author contributions

H.F., M.Do. and M.C. prepared the manuscript, and all authors contributed to the review of the final manuscript. M.C., M.Da., R.P.-K., A.V., J.M., S.A. and N.W. managed and performed the laboratory work. M.Do. developed and performed the bioinformatic analysis. S.Su., P.N., E.C., K.F.B., J.E.C.J. and J.S. performed data collection and analysis for epidemiological information. L.R., L.M., M.W., S.Si., G.J.S., M.R., J.E.-J., A.B., O.D., K.P. and A.N. managed the clinical cases. H.F., H.M. and K.S. coordinated the incident response teams.

Supplementary data

Tables S1 and S2 are available as Supplementary data at JAC Online.