Discussion

Over the past 20 EYs, there has been considerable variation in incidence rates of laboratory-confirmed IMD, in the ROI. The highest rate recorded was 11.6/100 000 in EY1999/2000, which has since gradually declined to the lowest ever recorded rate of 1.50/100 000 in EY2015/2016. The decline, largely attributed to the success of the MCC vaccine introduced in October 2000 in virtually eliminating menC disease, is also due to a spontaneous decrease in menB disease in the absence of any obvious intervention. Similar significant decreasing trends in menC IMD were observed in other countries that also implemented the MCC vaccine into their routine schedule compared with rates in countries that had no MCC vaccination policy [Reference Whittaker4]. This suggests the high uptake rate of the vaccine and the development of herd immunity among the unvaccinated cohorts [Reference Fitzgerald5] similar to the experience in the UK [Reference Maiden6, Reference Trotter and Maiden7].

Nonetheless, with the reduction in menC, disease due to menB predominated and accounted for the vast majority of cases in each EY after the introduction of the MCC in the ROI, as was the case in many EU countries [Reference Whittaker4], in particular those with MCC vaccination policies. MenB cases have in ROI however, also reduced significantly over the period with, in EY2015/2016 a sixfold drop in actual numbers since peak case figures in EY1999/2000 representing an eightfold drop in incidence rate. No specific anti-menB interventions were introduced to trigger this, although perhaps lifestyle changes including the introduction of the smoking ban from public places may have contributed to this decline [Reference Kabir8]. Similar spontaneous gradual declines in menB IMD cases have also been observed in other European countries, also in the absence of any specific interventions [Reference Stoof9–Reference Bijlsma11]. Furthermore, despite early concerns of possible vaccine-mediated capsule switching or serogroup replacement, only 4.3% of all IMD diagnosed since MCC introduction was due to non-menB or menC. In addition, as is the case in most EU countries [Reference Halperin1, Reference Whittaker4], IMD due to menW and menY was rare in ROI throughout the 20 EYs. Unlike experiences in the UK and Sweden, where significant recent rises in rates of disease due to menW and menY have been observed [Reference Hedberg12–Reference Ladhani16], constant but low rates of menW and Y disease were observed here. Nevertheless in 2014, ROI still had one of the highest rates of IMD in Europe, and the highest rate among countries that have a routine MCC vaccination programme at 1.3/100 000 population [Reference Whittaker4, 17].

The slight male predominance among IMD cases in ROI has also been documented by several other studies [Reference Hellenbrand10, Reference Bijlsma11, Reference Edge18–Reference Guerra-Silveira and Abad-Franch20] and which has been partly attributed to gender-related physiology and gender-specific behaviour [Reference Guerra-Silveira and Abad-Franch20–Reference Muenchhoff and Goulder23]. Also, the distribution of case age profiles was similar to the age profiles found with IMD in other countries [Reference Halperin1, Reference Hellenbrand10, Reference Bijlsma11], with infants and young children experiencing the greatest burden of disease and a smaller secondary disease peak observed in older teenagers. However, an age-shift was discerned with an increase in median case age observed between EY1996/1997 and EY2015/2016. A similar increase in case age was also documented in The Netherlands [Reference Bijlsma11] and those authors speculated that this was due to MCC vaccination and an ageing population overall. These explanations could also be appropriate for ROI. The MCC vaccine was rolled out as part of the infant immunisation schedule in 2000 with a catch-up campaign to include those up to 23 years of age. Overall levels of coverage were estimated to be 70%; highest in the <18 year olds at more than 80% [Reference Fitzgerald5]. The implementation of this vaccine resulted in the significant decline in menC IMD as a consequence of population herd immunity with the eradication of menC carriage among the principal carriers (older teenagers/young adults), thereby interrupting the transmission of menC [Reference Maiden6, Reference Trotter and Maiden7]. Furthermore, it is in this age group (<25 years) where the decline in IMD (menC, but largely menB) was most pronounced compared with the older age groups. The increase in the annual median case age values reflects this shift (the median age value is less influenced by the size of <25 age group and as a result the median age each year increases or the proportion of cases attributable to those aged <25 has declined much more sharply over time (compared with older age groups) and therefore, has resulted in a change/increase in the median case age). However, there were no discernible impacts due to menY or menW cases, which are traditionally associated with causing IMD in individuals aged over 65 years [Reference Stoof9, Reference Gunaratnam24], reflecting their long-standing low incidence rates in ROI.

Significant declines in menB disease rates were evident in all age groups except in ⩾65 year olds. Disease incidence due to menB (and other meningococcal serogroups) in this older population has, however, remained low. The greatest decreases were observed in children <5 years old and also in the 10–14 year olds, although these age groups still accounted for just over a quarter of all menB cases reported in EY2015/2016. The <1 year olds still had the highest menB incidence rate of any age group at 16.3/100 000 in EY2015/2016 supporting the introduction of the multi-component menB vaccine, 4CMenB, into the national infant immunisation programme in December 2016. However, it may be difficult to assess the impact of this vaccine in this age group due to low case numbers; the actual numbers of menB diagnosed in EY2015/2016 in <1 year olds (n = 11) was over 40% lower compared with the average number of menB cases diagnosed in this age group in the previous 4 years (18.7 cases). Furthermore, during the last four EYs, disease due to menY, menC and menW has increased in this age group (possibly due to spread from older populations), and accounted for 12.3% of all IMD in this age group in that period. 4CMenB is registered to prevent MenB disease [Reference O'Ryan25] and its efficacy against meningococcal disease caused by other serogroups is unclear [Reference Serruto26], although there may be some protection against menW [Reference Medini, Stella and Wassil27, Reference Ladhani28].

The use of PCR for the national non-culture diagnosis of IMD with determination of meningococcal serogroup has undoubtedly had a significant impact on the speed and accuracy of case confirmation. It has enabled not only improved individual case management, but also better public-health management through accurate evaluation of the impact of MCC vaccination and our understanding of meningococcal disease epidemiology. Over the 20 EY period, without PCR diagnosis in ROI, there would have been a large gap in the confirmation of disease and associated serogroup for almost 60% of cases from whom the causative meningococcal isolate was not cultured. PCR diagnosis of IMD is now an important fundamental tool in many reference laboratories and has made an invaluable contribution to the diagnosis of similar proportions of cases in many countries since its introduction [Reference Edge18, Reference Clarke29–Reference Tzanakaki31]. However, despite the advantages of case confirmation by PCR, there are also several limitations, not least in its inability to accurately monitor antimicrobial susceptibility and perform detailed meningococcal strain characterisation including examining antigen expression in the absence of an isolate. Although, whole genome sequencing (WGS) of pathogens directly from clinical material without culture is possible, process improvements are necessary to ensure success with all pathogen-positive material [Reference Anson32, Reference Rossen, Friedrich and Moran-Gilad33]. Meanwhile, the lack of isolates recovered from cases within the age groups associated with the highest incidence of IMD in ROI is of grave concern as currently only limited epidemiological information can be obtained concerning the causative meningococcal strain from cases confirmed by PCR only. Also, given that meningococcal isolates were more likely to be recovered from cases aged ⩾20 years than from those aged ⩽19 years, our future understanding of meningococcal strain epidemiology could be biased.

Owing to these issues, it is imperative that every effort is made in an attempt to recover the causative organism associated with IMD, including correct sampling and rapid culturing of throat swabs. A recent study examining throat–blood meningococcal isolate pairs recovered from individuals with IMD using detailed characterisation by WGS found that they differed by only a very small number of genetic changes [Reference Klughammer34]. The importance of an isolate is even more crucial in view of the recent introduction of 4CMenB. 4CMenB is a multi-component protein-based vaccine with predicted incomplete strain coverage [Reference Medini, Stella and Wassil27], and unlike traditional capsular-based meningococcal vaccines (e.g. MCC), its efficacy cannot be measured by examining meningococcal serogroup, alone. It is not a menB vaccine per se as other meningococcal serogroups possess many (and variants) of the components contained within the vaccine, as in fact do many of the non-pathogenic commensal Neisseria species [Reference Serruto26, Reference Medini, Stella and Wassil27]. Indeed, following vaccination and exertion of immunologic pressure, it is reasonable to assume that strains not covered by 4CMenB or with incomplete coverage may emerge with variation occurring in strains once susceptible (those that harboured all or some of the exact vaccine components acquiring unmatched components from either other meningococci or from other Neisseria spp.) leading to the emergence of natural or acquired vaccine escape strains and ‘vaccine failure’ [Reference Lucidarme35, Reference Brehony36]. Definitive and accurate evaluation of the impact of 4CMenB vaccination and monitoring of meningococcal epidemiology will only be achievable through detailed strain characterisation using WGS [Reference Brehony36] and by determining levels of antigen expression, the latter is only possible with a viable meningococcal isolate [Reference Serruto26].

In conclusion, important changes in the epidemiology of IMD have been observed in ROI during the last two decades. MenC IMD virtually disappeared following the implementation of a universal childhood immunisation programme effecting that most cases since then have been due to menB. Overall IMD incidence has decreased significantly and now remains relatively low, although the most affected cohorts are still infants and the very young with teenagers and adolescents also at an increased risk. Furthermore, given historical dynamics of IMD epidemiology (even over the past 20 EYs), it is unpredictable whether low disease incidence will persist and what potential regional or global impact emerging strains may have. In December 2016, 4CMenB (Bexsero®, GlaxoSmithKline, https://www.gsk.com) was introduced into the infant immunisation programme in Ireland to help mitigate this, but data regarding its overall efficacy are scant, especially in relation to the level and duration of protection offered, specific targeted strains and effect on carriage/herd immunity. Now more than ever, detailed surveillance is essential to monitor both the disease-associated and carried meningococcal populations circulating in ROI. WGS of meningococci allows the amassing of unprecedented levels of comprehensive and detailed information on a strain's genetic make-up, making it pivotal to the interpretation of meningococcal population biology dynamics and aid in the understanding of how the meningococcus can cause life-threatening invasive disease in some individuals and be just a harmless commensal in most others. From an epidemiological aspect, it is an invaluable tool that allows the exploration of capsule switching and, more specifically, the early detection of the emergence or re-emergence of hyperinvasive strains.