N. gonorrhoeae has an extensive history of antimicrobial resistance and has demonstrated that it is capable of utilizing a variety of mechanisms to escape antimicrobial pressure. Extended-spectrum cephalosporins have been the CDC recommended treatment only since 2007, but it appears that their usefulness is already on the decline. In 2010, the CDC issued treatment guidelines recommending dual therapy with CFM plus AZM or doxycyline due to concerns about increasing antimicrobial resistance. In 2012, the CDC again updated its recommendations and abandoned oral extended-spectrum cephalosporins in favor of dual therapy with CRO plus AZM or doxycycline. The data here support this recommendation since no isolates were found with alert value MICs to both AZM and an extended-spectrum cephalosporin. Two isolates with alert value MICs to AZM were detected in San Francisco and San Diego, however, so the emergence of dual-resistant strains in California remains a possibility.
Three mosaic penA alleles were detected in this study and a novel one was identified. The mosaic XXXIV and XXXVIII penA alleles were both previously detected in San Francisco in 2009 , while the LA-A penA allele is novel. The mosaic XXXIV penA allele has all three amino acid changes previously associated with reduced susceptibility to oral extended-spectrum cephalosporins . Homology modeling studies have shown that these amino acid changes result in conformational alterations of the β-lactam-binding pocket and affect the ability of extended-spectrum cephalosporins to bind, particularly those with large R groups . More recently, it has been shown that these amino acid changes are epistatic and only result in higher MICs in the context of other mutational changes in penA. Although the mosaic XXXVIII and LA-A penA alleles both have two of these amino acid changes, the distributions of extended-spectrum cephalosporin MICs for isolates with these penA alleles were similar to those with non-reactive penA alleles. These data support the hypothesis that the I312M and V316T amino acid changes act synergistically with the G545S amino acid change and that they are necessary, but not sufficient, to produce elevated oral extended-spectrum cephalosporin MICs.
The mosaic XXXIV penA allele detected in this study is concerning because it was found in isolates with extended-spectrum cephalosporin MICs that have been reported in isolates associated with treatment failures. Since first being published in 2009, the mosaic XXXIV penA allele has been found worldwide and is now associated with ST1407 strains and reduced susceptibility to extended-spectrum cephalosporins. A pharyngeal ST1407 isolate with the mosaic XXXIV penA allele recently caused a CRO treatment failure in Slovenia  and strains with closely related mosaic penA alleles have caused treatment failures in France (mosaic CI) , Austria (mosaic XXXIV with T534A amino acid alteration)  and Hong Kong (mosaic X) . The mosaic XXXIV penA allele is only a single amino acid change (A501P) different from the mosaic CI penA allele described in extensively-drug resistant isolates from France and Spain that exhibited high-level resistance to both oral and injectable extended-spectrum cephalosporins [23, 30]. This additional mutation might be capable of increasing CRO MICs from the elevated levels seen in this study (CRO MIC = 0.094-0.25 μg/mL) to levels at which CRO treatment failures have occurred even with much higher 1 g doses (CRO MIC = 1.0-2.0 μg/mL) . The extensively-drug resistant isolates described in France and Spain were also ST1407, indicating a genetically similar background. It is possible, however, that determinants of resistance other than penA also affect the extended-spectrum cephalosporin MICs observed.
In 2009, the mosaic XXXIV and XXXVIII penA alleles were described only in ST1407 isolates and a very closely related isolate (ST1513, tbpB 110, por 971, 99.8% similar to por 908) in San Francisco and appeared to be spreading in a clonal fashion. The mosaic XXXVIII penA allele has remained clonal within ST1407 in California, while the mosaic XXXIV penA allele was found in 10 STs other than ST1407 in 2011. In isolates with the mosaic XXXIV penA allele and STs other than ST1407, the tbpB allele was the same as ST1407 (tbpB allele 110) or was a closely related tbpB allele (tbpB allele 1431, 99.5% similar to tbpB 110). ST1407 remains the dominant ST associated with reduced extended-spectrum cephalosporin susceptibilities in California, but strains with unrelated STs now carry the mosaic XXXIV penA allele. Genomic mapping of ST1407 and ST5895 isolates from San Francisco recently detailed the transfer of the mosaic XXXIV penA allele into a completely different genetic background . In this study, four ST5895 isolates with the mosaic XXXIV penA allele were found in Los Angeles, Orange County and San Diego, indicating that this strain is likely circulating throughout California. Unfortunately, the horizontal spread of the mosaic XXXIV penA allele into new STs makes tracking the spread of extended-spectrum cephalosporin resistance with NG-MAST problematic.
Of note is the fact that the mosaic XXXIV penA allele was only found in isolates with tbpB allele 110 or a very closely related allele (tbpB 1431, 99.6% similar to tbpB allele 110). In the N. gonorrhoeae genome, penA is located ~50 kb downstream of tbpB. The linkage between these two loci likely explains the high level of clonality of tbpB alleles in isolates with the mosaic XXXIV penA allele. The linkage between penA and tbpB was also evident in isolates with the mosaic LA-A penA allele where 13/16 isolates had tbpB 18.
Detecting strains with reduced susceptibility to extended-spectrum cephalosporins is a challenge in the U.S. The vast majority of N. gonorrhoeae infections are diagnosed using nucleic acid-based methods where no live organism is present. Without a culture and antimicrobial susceptibility testing, molecular assays are currently the only option available for detecting isolates with reduced susceptibility to extended-spectrum cephalosporins. The modified RTPCR presented here is a simple and rapid tool capable of specifically detecting isolates or urine specimens that may have alert value extended-spectrum cephalosporin MICs. The assay may find utility in the surveillance and epidemiological investigation of strains with reduced susceptibility to extended-spectrum cephalosporins.
This assay, however, has several limitations. The mosaic penA alleles which the assay was designed to discriminate between were detected in California in 2011. It is possible that other mosaic penA alleles are circulating elsewhere that produce elevated extended-spectrum cephalosporin MICs, but do not share the region of sequence identity that the modified RTPCR detects. Also, the modified RTPCR should be used with caution on pharyngeal specimens because of the potential for cross-reactivity with penA alleles from other Neisseria spp. The assay performed well on N. gonorrhoeae positive urine specimens matched to isolates with the mosaic XXXIV penA allele, but only a small number of specimens were available for testing.
The main limitation of this study is the sampling strategy used to collect N. gonorrhoeae isolates. The samples were collected in parallel with the CDC GISP which collects isolates only from men. Additionally, the samples were collected from public health clinics whose populations may not be representative of the general population and are known to oversample MSM . Since MSM populations have historically been associated with higher rates of antimicrobial resistance, it is possible that these data overestimate the prevalence of alert value isolates in the general population . Another limitation was the use of RTPCR to screen isolates for the presence of a mosaic penA allele. The screening RTPCR detected a region of shared sequence identity found in many mosaic penA alleles, but it is possible that other mosaic alleles were present that are non-reactive with this assay.