Antibiotic resistance among invasive Neisseria meningitidis isolates in England, Wales and Northern Ireland (2010/11 to 2018/19)

Invasive meningococcal disease (IMD), caused by Neisseria meningitidis, can have a fatality rate as high as 10%, even with appropriate treatment. In the UK, penicillin is administered to patients in primary care whilst third generation cephalosporins, cefotaxime and ceftriaxone, are administered in secondary care. The first-choice antibiotic for chemoprophylaxis of close contacts is ciprofloxacin, followed by rifampicin. Immunocompromised individuals are often recommended antibiotic chemoprophylaxis and vaccination due to a greater risk of IMD. Resistance to antibiotics among meningococci is relatively rare, however reduced susceptibility and resistance to penicillin are increasing globally. Resistance to third generation cephalosporins is seldom reported, however reduced susceptibility to both cefotaxime and ceftriaxone has been observed. Rifampicin resistance has been reported among meningococci, mainly following prophylaxis, and ciprofloxacin resistance, whilst uncommon, has also been reported across the globe. The Public Health England Meningococcal Reference Unit receives and characterises the majority of isolates from IMD cases in England, Wales and Northern Ireland. This study assessed the distribution of antibiotic resistance to penicillin, rifampicin, ciprofloxacin and cefotaxime among IMD isolates received at the MRU from 2010/11 to 2018/19 (n = 4,122). Out of the 4,122 IMD isolates, 113 were penicillin-resistant, five were ciprofloxacin-resistant, two were rifampicin-resistant, and one was cefotaxime-resistant. Penicillin resistance was due to altered penA alleles whilst rifampicin and ciprofloxacin resistance was due to altered rpoB and gyrA alleles, respectively. Cefotaxime resistance was observed in one isolate which had an altered penA allele containing additional mutations to those harboured by the penicillin-resistant isolates. This study identified several isolates with resistance to antibiotics used for current treatment and prophylaxis of IMD and highlights the need for continued surveillance of resistance among meningococci to ensure continued effective use.


Introduction
Invasive meningococcal disease (IMD), including meningitis and/or septicaemia, is a severe, life-threatening illness caused by the bacterium Neisseria meningitidis (the meningococcus). Six meningococcal serogroups (A, B, C, W, X and Y) are responsible for the majority of IMD cases worldwide [1][2][3]. Meningococci can be further differentiated into sequence types (STs), with related STs grouped into lineages termed clonal complexes (CCs) [4]. Serogroup B meningococcal disease is the most prevalent in the UK and is the most diverse in terms of clonal complex distribution [5].
IMD is fatal in 5-10% of cases in developed countries, despite prompt commencement of antibiotic treatment [6]. In the UK, patients in primary care with suspected meningococcal disease are administered penicillin, whilst those in secondary care are administered third generation cephalosporins (3GCs), ceftriaxone or cefotaxime [7]. Ciprofloxacin is the chemoprophylactic agent of choice for close contacts of cases and rifampicin is a suitable alternative [8]. Chemoprophylaxis is recommended for close contact of IMD cases and for individuals with deficiencies of the terminal complement pathway, where the risk for IMD development is up to 10,000 fold higher than in the general population [9].
N. meningitidis has remained largely susceptible to the antibiotics used in the treatment and prevention of IMD. Resistance to ciprofloxacin, due to alterations in the gyrA gene (encoding DNA gyrase subunit A), is relatively uncommon, except for in China, where the prevalence of resistance among isolates, in particular of the ST-4821 CC, has led to the withdrawal of its use as a first line prophylactic agent [10,11]. More recently, ciprofloxacin resistance has been reported in invasive ST-23 CC isolates in the USA [12] and in invasive ST-175 CC isolates in immunocompromised individuals in the UK and Europe [13]. Mutations in the parC gene (encoding DNA topoisomerase IV, subunit A), when accompanied by mutations in the gyrA gene, cause enhanced levels of ciprofloxacin resistance [14].
Primary cases of rifampicin-resistant meningococci due to alterations in the rpoB gene (encoding the beta subunit of RNA polymerase), is uncommon and not associated with particular lineages. It is usually observed following rifampicin prophylaxis [15][16][17].
Meningococcal resistance to 3GCs is rare. A report of cefotaxime and ceftriaxone resistance among serogroup A isolates in India [24] has been met with some scepticism [25]. Reduced susceptibility to cefotaxime has been observed among several invasive isolates in France [26]. The penA gene responsible, penA327, was identical to that causing reduced susceptibility to 3GCs in N. gonorrhoea [27]. The allele, containing only four out of the five aforementioned AASs, also causes penicillin resistance and reduced susceptibly among meningococci [20,26]. Amino acid substitution A501P in penA has been shown to enhance levels of 3GC resistance among N. gonorrhoea, and there are some concerns regarding the potential selection of this same mutation among meningococci [28]. Surveillance of antibiotic resistance among this species therefore remains of paramount importance.
The Public Health England (PHE) Meningococcal Reference Unit (MRU) offers a complimentary national reference service for confirmation and characterisation of meningococcal isolates in England, Wales and Northern Ireland (E, W and NI). The aim of this study was to characterise isolates from culture-confirmed cases of IMD received at the MRU from July 2010 to June 2019 inclusive (epidemiological years 2010/11 to 2018/19), and to determine the distribution of antibiotic susceptibility.

Materials and methods
The study included 4,122 IMD isolates received by the MRU from July 2010 to June 2019, inclusive.

Serogrouping
Serogrouping was performed using the dot-blot ELISA as previously described [29].

Antibiotic susceptibility testing
Antibiotic susceptibility testing for penicillin, cefotaxime, rifampicin and ciprofloxacin was performed using the Etest (bioMerieux UK Limited, Basingstoke, UK) method on either Mueller Hinton agar plates supplemented with 5% horse blood and 20 mg/L β-nicotinamide adenine dinucleotide (Oxoid Limited, Basingstoke, UK) (isolates received on or after 9 th May 2019) or Iso-sensitest blood agar supplemented with 20 mg/L NAD (isolates received prior to 9 th May 2019).
Isolates with penicillin MICs �0.06 mg/L were categorised as 'susceptible, standard exposure (S)' (where there is a high likelihood of therapeutic success using a standard dosing regimen of the agent) and isolates with penicillin MICs >0.25 mg/L were categorised as 'resistant (R)' (where there is a high likelihood of therapeutic failure even when there is increased exposure). Isolates with intermediate penicillin MICs were categorised as 'susceptible, increased exposure (I)' (where there is a high likelihood of therapeutic success because exposure to the agent is increased by adjusting the dosing regimen or by its concentration at the site of infection) [30].

Beta-lactamase testing
Beta-lactamase testing was performed using Interlactam strips (MAST, UK) following the manufacturer's instructions.
Genotypic data were obtained using the BLAST and/or Export Dataset tools [31,32].

Distribution of antibiotic susceptibility
More than half of the isolates (63%; n = 2,591) were penicillin-susceptible, standard exposure

PLOS ONE
Only one isolate was resistant to more than one antibiotic (penicillin and cefotaxime; 0.5 mg/L and 0.25 mg/L, respectively; PubMLST ID 20267). Details of all isolates that were resistant to at least one antibiotic are listed in S1 Table. Penicillin susceptibility Fifty-five percent of serogroup B isolates were penicillin-susceptible (Fig 1). The proportion of penicillin-susceptible serogroup C, W and Y isolates were 53%, 76% and 73%, respectively. The proportion of serogroup B PenS isolates decreased from 67% in 2010/11 to 38% in 2018/ 19 (S1 Fig). There was an increase in the proportion of serogroup W PenR isolates over the last four years, with the highest proportion observed in 2018/19 (8%; S1 Fig).
The most prevalent clonal complex (all serogroups) was ST-11 CC followed by ST-41/44 CC where 75% and 56% of isolates were penicillin-susceptible, respectively (Fig 1). The clonal complex with the lowest proportion of PenS isolates was ST-461 CC (13%), which was also associated with the highest proportion of PenR isolates (21%).
The increase in PenR serogroup W isolates over the last four years (S1 Fig) was due to a rise in the number of serogroup W ST-11 CC isolates harbouring penA9, as reported previously [22]. The decrease of serogroup B PenS isolates overtime (S1 Fig) was largely due to a decrease in serogroup B PenS ST-41/44 CC isolates with penA1 and ST-269 CC isolates with penA3 and penA27 (Fig 3).
Allele penA419, which was also associated with penicillin resistance, was not found in any other isolates on PubMLST (accessed 10 th June 2021). In addition to the five AASs associated with penicillin resistance, the allele had substitutions A501T and D511V that were not found among penA alleles harboured by cefotaxime-susceptible PenR isolates received at the MRU.

Rifampicin susceptibility
Two isolates (ST-11 CC and ST-41/44 CC, respectively) were rifampicin resistant (MICs = 0.5 and >32 mg/L; S1 Table). These harboured rpoB alleles; rpoB238 and rpoB84, respectively. RpoB238 had AAS D542E and RpoB84 had AAS H552N. There was no evidence that suggested either of the two isolates were from close contacts of index cases.

Discussion
From 2010/11 to 2018/19, antibiotic resistance was uncommon among IMD isolates in E, W and NI. However, resistance to all four antibiotics (penicillin, rifampicin, cefotaxime and ciprofloxacin) was observed.
As has been reported in several other countries [20,23,34], the proportion of PenR isolates in E, W & NI has increased over time, thus, highlighting the need for surveillance of penicillin susceptibility in order to ensure its continued effective use as a first-line prophylactic agent.
The recent increase was largely due to PenR serogroup W ST-11 CC isolates that have emerged in a lineage that originated in South America in the early 2000s, and that has since spread to Europe, Australasia and North America [35][36][37]. Since 2016, Australia and the UK have both observed an increase in PenR isolates within the so-called original UK W:cc11 strain or the South American strain sublineage [22,23]. Reports of penicillin-resistant meningococcal strains causing IMD in complement deficient individuals receiving chemoprophylaxis [38,39] make this strain particular cause for concern.   Out of the 113 PenR isolates, three had a penA allele (penA327, penA209 and penA171) that did not contain all five of the AASs commonly associated with penicillin resistance (F504L, A510V, I515V, H541N, I566V).
PenA327, (I566V absent), has been previously reported in penicillin-resistant meningococcal isolates and also those with reduced susceptibility to 3GCs [26]. All IMD isolates with penA327 received at the MRU displayed reduced susceptibility to cefotaxime (MICs 0.047-0.125 mg/L).
PenA209 had only three of the five AASs (F504L, A510V, I515V) alongside V447L and A516G. PenA alleles identified on PubMLST that contained F504L, A510V and I515V without V447L and A516 had lower penicillin MICs (0.047-0.094 mg/L) compared to those that also had V447L and A516 (0.12-0.5 mg/L). This may suggest that penA alleles harbouring the three AASs may not be adequate to confer penicillin resistance alone, however, do when in combination with V447 and A516. Future work is needed to confirm the role that these AASs have in conferring resistance to penicillin among meningococci.
PenA171 contained none of the five AASs commonly associated with conferring reduced susceptibility and resistance to penicillin. It did contain P551S, a mutation which has been observed among gonococcal isolates resistant to penicillin [40], however, other meningococcal isolates with penA171 on the PubMLST database were penicillin-susceptible, suggesting penA171, and the P551S AAS alone, are not responsible for causing penicillin resistance among meningococci. Further work is required to determine the cause of penicillin resistance in this isolate which was beta-lactamase negative, as reflected in the relatively low MIC (0.38 mg/L). Other mechanisms of penicillin resistance have been determined among N. gonorrhoeae, which include mutations in the transcriptional repressor of the efflux pump (mtrCDE) operon, which is encoded by the mtrR gene [41]. The operation of the mtrCDE efflux pump in meningococci appears to differ than in that of N. gonorrhoeae [42], with no evidence to suggest that mutations in the mtrR gene of meningococci can cause penicillin resistance.
Only seven isolates had cefotaxime MICs � 0.047 mg/L, one of which was cefotaxime-resistant (MIC = 0.25 mg/L). Five of the seven isolates harboured penA327, which has previously been associated with reduced susceptibility to cefotaxime and penicillin in meningococci. Four of these were serogroup C isolates belonging to ST-11 CC and one was a serogroup B strain belonging to ST-41/44 CC. The allele has been previously identified among ST-11 CC isolates of serogroups B and C, suggesting successful clonal expansion of this particular strain PLOS ONE [26]. The allele is identical to that of an allele identified among gonococcal isolates displaying resistance to ceftriaxone in this species [27]. Apart from penA327, no other penA allele has been associated with reduced susceptibility or resistance to 3GCs among meningococci. High levels of 3GC resistance due to A501P have been observed among gonococcal isolates [43]. Other AASs at this position close to the core of the active site motif of PBP2, A501T and A501V, have been described among gonococcal isolates displaying reduced susceptibility to cefotaxime [44][45][46]. It is thought that the substitution of methyl side chain alanine (A501) interferes with 3GC binding, or leads to structural modifications causing higher levels of 3GC resistance [43].
The cefotaxime-resistant meningococcal isolate identified in this study harboured penA419 with two unique AASs, A501T and D511V, when compared to cefotaxime-susceptible penicillin-resistant isolates. Among other isolates on PubMLST with penA AASs at positions 501 and 511, isolates harbouring A501V had lower MICs than those with A501T. This suggests that AAS A501V does not confer reduced susceptibility to cefotaxime in meningococci, unlike in gonococci [45], whilst A501T may. When A501T was present in combination with an AAS at 511 (D511V or D511G), cefotaxime MICs among the meningococcal isolates were further enhanced. Amino acid substitutions at D511 are extremely rare. Further work is therefore required to confirm the role that this AAS plays in susceptibility to cefotaxime among meningococci. No meningococcal isolates with AAS A501P were identified, however, it has already been suggested that this mutation may eventually be selected in meningococci [28].
Five isolates were ciprofloxacin-resistant (MICs = 0.06-0.5 mg/L). Each of these harboured AASs in the gyrA gene (D95N or T91I) which have previously been associated with conferring ciprofloxacin resistance among meningococcal isolates worldwide [47][48][49]. The ciprofloxacinresistant isolate with the highest MIC (MIC = 0.5 mg/L) also harboured an AAS in its parC allele (D86N). Mutations in parC among meningococcal isolates which also harbour mutations in gyrA have been shown to cause enhanced levels of resistance to ciprofloxacin (MICs of 0.5-1 mg/L).
The five ciprofloxacin-resistant isolates received at the MRU belonged to five different clonal complexes. In China, where ciprofloxacin resistance is prevalent, resistance is commonly seen among isolates belonging to ST-4821 CC, which has also been observed among ciprofloxacin-resistant isolates in other countries [10,48,50]. In the USA, a recent study identified ciprofloxacin resistance among several meningococcal isolates belonging to ST-23 CC [12]. Since the ciprofloxacin-resistant isolates in the present study do not seem to be prevalent among a certain clonal complex, it may suggest that these particular strains in the UK have not successfully expanded with the altered gyrA alleles and indicates that levels of ciprofloxacin resistance may continue to remain low within the population. Recently however, ciprofloxacin resistance has been observed among a particular strain of NG meningococci of the ST-175 CC in the UK and Europe, causing IMD in immunocompromised individuals. Occasionally, ciprofloxacin may be recommended to immunocompromised individuals or prescribed as a rescue therapy for when symptoms of IMD occur despite other prophylactic regimes [51][52][53] and so the presence of this non-groupable ST-175 CC strains in the UK remains a concern.
Only two isolates were resistant to rifampicin which is rare among meningococci considering its widespread use as a prophylactic agent. The majority of meningococcal rifampicin resistance is found among isolates from cases of close contacts following rifampicin prophylaxis [15,17]. However, rifampicin resistance has also been observed among index cases [54]. There was no evidence to suggest that the two rifampicin-resistant isolates received at the MRU were from close contact cases.
One rifampicin-resistant isolate harboured rpoB238 (MIC = 0.5 mg/L), which had AAS 542E. The other rifampicin-resistant isolate rpoB84 (MIC = >32 mg/L), which had AAS H552N. Both AASs have been previously reported among rifampicin-resistant meningococcal isolates, with mutations at AA position 552 associated with higher levels of resistance than those with mutations at AA position 542 [55]. Studies have also identified the same AASs among multiple isolates with varying levels of rifampicin resistance [54,56]. This, along with the observed difference in rifampicin MIC values displayed by the two resistant isolates in this study, may indicate that additional unknown mechanisms can cause increased levels of rifampicin resistance among meningococci.
This study identified several IMD isolates with resistance to antibiotics used for current treatment and prophylaxis of IMD. Future genetic engineering work is needed to confirm the role that certain penA mutations have in conferring resistance to penicillin and cefotaxime among meningococci. Further molecular work is also required to determine possible alternative penicillin resistance mechanisms of meningococci. This study also identified emerging threats and the progression of antibiotic resistance among strains over time. Consideration should be given before the mass use of antibiotics in single-dose prophylaxis strategies to control outbreaks of meningococcal disease, particularly where outbreaks frequently occur such as in the African meningitis belt [57]. Sustained surveillance of antibiotic resistance among the meningococcal population is essential to maintain successful treatment regimens for IMD and chemoprophylaxis regimes.