Haemophilus influenzae, particularly type b (Hib), was once a dominant cause of invasive infection in children. Incidence of disease declined dramatically following the introduction of the H. influenzae type b conjugate vaccine in the United States in 1990, with a >90% reduction reported by the US Centers for Disease Control and Prevention (CDC).1
However, population-based surveillance studies have noted an increase in the incidence of invasive non-H. influenzae type b disease in the United States in recent years,2,3 particularly as a result of non-typeable H. influenzae (NTHi) and H. influenzae type a (Hia). In a study conducted by the CDC Active Bacterial Core surveillance (ABCs) of invasive H. influenzae, disease incidence of Hia increased by approximately 13% annually and NTHi by 3% annually from 2002 to 2015.2 These investigators noted that the highest incidence of invasive Hia disease was in children <4 years old and NTHi was most frequent in infants <1 year of age. This is clinically relevant because non-Hib infections may be associated with significant morbidity and even mortality in children.4 We recently reported an increase in the rate of pediatric hospitalization for invasive H. influenzae at Texas Children’s Hospital (TCH) in Houston, TX.5 The increased frequency was largely driven by cases of NTHi and Hia. Limited data are available regarding the contemporary molecular epidemiology of these organisms in US children; such knowledge is of importance in potentially identifying emerging strains. Additionally, there is some evidence that select sequence types (STs) may be associated with specific infectious diagnoses.6
We performed multilocus sequence typing (MLST)7 on invasive H. influenzae isolates recovered from patients at TCH from January 2011 to December 2018; the clinical course and management of these infections have been presented in a prior publication.5 2011 was chosen as the study start date as this year corresponded to the full initiation of the electronic medical record in our institution. Isolates included were those identified in blood, cerebrospinal fluid, pleural fluid or synovial fluid cultures in patients <18 years of age. Typing was performed routinely by the Texas Department of State Health Service Laboratory (Austin, TX) on blood and CSF isolates; MLST was performed on all viable isolates in the Edward O Mason Infectious Diseases Research Laboratory at TCH using described methods and primers.7 STs were assigned based on a publicly available database (https://pubmlst.org/organisms/haemophilus-influenzae/). Minimum spanning trees were constructed using Phyloviz (www.phyloviz.net).
STs were compared in terms of patient clinical characteristics and diagnoses; categorical variables were compared using Fisher’s exact test and continuous variables with Wilcoxon rank-sum or Kruskal-Wallis tests as appropriate. The number of hospital admissions was used to estimate the frequency of disease for select STs and expressed in terms of cases/10,000 hospital admissions as a surrogate for disease incidence. Incidence trends were examined using Poisson regression (STATA ver. 15).
Among 61 culture-confirmed invasive H. influenzae infections identified by the clinical microbiology laboratory in the course of care, 38 isolates were viable for further study. Eighteen isolates were encapsulated (type a: n = 9; type e: n = 2; type f: n = 7), 14 were NTHi and typing was not available for 6 isolates; no cases of Hib occurred. The median patient age was 29 months (interquartile range [IQR]: 11–60 months). The diagnoses included bacteremia (n = 19), meningitis (n = 8), pneumonia (n = 6) and septic arthritis (n = 5). Six isolates were β-lactamase positive (15.7%); no β-lactamase negative ampicillin-resistant strains were noted.
The 38 isolates encompassed 17 STs (Fig. 1A and B). The most common single STs were ST56 (n = 10) and ST124 (n = 10). Although NTHi were overall the most common isolates, they were genetically highly diverse encompassing 11 unique STs. There were significant correlations between capsular type and ST (P < 0.001). The majority of Hia were ST56 (7/9, 77.8%); similarly, ST124 predominated among type f isolates (H. influenzae type f [Hif], 6/7, 85.7%). Two isolates with novel STs were noted, including infections due to Hia (ST2235, meningitis) and Hif (ST2234, bacteremia), respectively. In the final 2 study years, all NTHi were either ST142 or ST388 (n = 2 each). The 6 β-lactamase positive isolates comprised 5 different STs (one each of ST14, ST56, ST107 and ST127, and 2 ST388 isolates). All ST124 and 9/10 ST56 were β-lactamase negative compared with 13/18 isolates of any other ST (72.2%, P = 0.13).
We had previously reported an increase in all invasive H. influenzae disease during this time period, largely associated with increased incidence of NTHi and Hia disease; a significant increase in other capsular types was not observed.5 No cases of infection due to either ST56 or ST124 occurred from 2011 to 2013. Starting in year 2014, a numeric increase in hospitalizations due to ST56 (2014, 0.95 cases/10,000 hospitalizations compared with 2018, 1.46/10,000 hospitalizations, P = 0.07) and ST124 (2014, 0.47 cases/10,000 hospitalization compared with 2018, 1.16/10,000 hospitalizations, P = 0.06) was observed (Fig. 1C).
The median age was similar among patients with ST56 (29 months, IQR: 3–60) and ST124 (23 months, IQR: 17–49) infection compared with other STs (32 months, IQR: 11–32, P = 0.9). STs were similar with respect to patient demographics and comorbidities (data not shown). Although there was no statistically significant association with any single ST and site of infection, ST56 and ST124 together accounted for 5 of 8 cases of meningitis (62.5%) and 4 of 5 cases of septic arthritis (80%, Fig. 1D). Six children with meningitis experienced sensorineural hearing loss of which 3 cases (50%) were due to ST124.
Prior studies encompassing isolates from a global collection suggest that the dominant ST among Hia is ST23.7,8 ST23 isolates were notably absent in our study with the majority of Hia belonging to ST56, a single locus variant of ST23.9 Interestingly, in a recent pediatric study of Hia disease conducted in Utah, ST62 dominated, accounting for 75% of isolates, with only 10% belonging to ST56.10 Similar to previous studies, encapsulated strains at TCH were rather homogeneous compared with NTHi isolates, which were very diverse.8,9,11 Given the clonal nature of encapsulated strains, our observed increase in pediatric ST56 disease is consistent with both the local and national increase in invasive Hia.2,5 Other investigators have reported a higher incidence of Hia disease in certain minority populations in the United States, particularly Native Americans/Alaskan Natives.3 We were unable to demonstrate an association with ST56 (or any particular ST) and patient demographic characteristics albeit our sample size was limited. We also noted a small temporal increase in hospitalization with ST124 H. influenzae, the majority of which were Hif. Previous work by the ABCs from 2009-2015, found a relatively stable incidence of Hif disease, though noting that the incidence was highest in infants <1 year and adults >65 years old followed by children 1–4 years of age.2 Notably, while most ST124 were Hif, 2 isolates were NTHi and typing was not available for another two. These findings suggest that the increase in ST124 may not be solely attributable to the expansion of Hif alone but perhaps hint at an advantageous genetic background.
The reasons for the observed local and national increase in invasive H. influenzae disease, particularly Hia and NTHi, are unclear. It is conceivable that these findings may be an indirect consequence of widespread use of the pneumococcal and Hib conjugate vaccines; alterations in both the nasopharyngeal flora and the causative agents of otitis media in young children have been described since the introduction of these conjugate vaccines.12,13 Reductions in colonization with pneumococci and Hib may have opened ecologic niches in the nasopharynx of children, allowing colonization by non-H. influenzae type b strains; one can speculate that this may in turn predispose to invasive disease. Additionally, the emergence of highly fit H. influenzae clones may contribute to increased disease incidence, as suggested above. Further large studies are needed to better define the current molecular epidemiology of invasive H. influenzae to guide any future vaccine efforts.
There are limitations to our study which should be acknowledged. As not all invasive H. influenzae cases identified during the study period had isolates viable for MLST, there is the potential for sample bias and our findings should be interpreted cautiously. Furthermore, our data are from a single academic center and may not be representative of H. influenzae epidemiology in the nation as a whole.
Although our sample size is limited and the results may not be generalizable, these data shed light on the contemporary molecular epidemiology of pediatric invasive H. influenzae in our region and suggest subtle differences in Hia and Hif epidemiology in the Houston area compared with other reports. These data are relevant given increasing incidence of non-H. influenzae type b disease and may be informative to future surveillance efforts.
1. MacNeil JR, Cohn AC, Farley M, et al. Current epidemiology and trends in invasive Haemophilus influenzae
disease – United States, 1989-2008. Clin Infect Dis. 2011;53:1230–1236.
2. Soeters HM, Blain A, Pondo T, et al. Current epidemiology and trends in invasive Haemophilus influenzae
disease – United States, 2009-2015. Clin Infect Dis. 2018;67:881–889.
3. Soeters HM, Oliver SE, Plumb ID, et al. Epidemiology of invasive Haemophilus influenzae
serotype a disease – United States, 2008-2017 [published online ahead of print June 26, 2020]. Clin Infect Dis. doi: 10.1093/cid/ciaa875.
4. Adderson EE, Byington CL, Spencer L, et al. Invasive serotype a Haemophilus influenzae
infections with a virulence genotype resembling Haemophilus influenzae
type b: emerging pathogen in the vaccine era? Pediatrics. 2001;108:E18.
5. Vallejo JG, McNeil JC, Hultén KG, et al. Invasive Haemophilus influenzae
disease at Texas Children’s Hospital, 2011 to 2018. Pediatr Infect Dis J. 2019;38:900–905.
6. Collins LF, Havers FP, Tunali A, et al. Invasive nontypeable Haemophilus influenzae
infection among adults with HIV in Metropolitan Atlanta, Georgia, 2008-2018. JAMA. 2019;322:2399–2410.
7. Meats E, Feil EJ, Stringer S, et al. Characterization of encapsulated and noncapsulated Haemophilus influenzae
and determination of phylogenetic relationships by multilocus sequence typing. J Clin Microbiol. 2003;41:1623–1636.
8. Tsang RSW, Ulanova M. The changing epidemiology of invasive Haemophilus influenzae
disease: emergence and global presence of serotype a strains that may require a new vaccine for control. Vaccine. 2017;35:4270–4275.
9. Potts CC, Topaz N, Rodriguez-Rivera LD, et al. Genomic characterization of Haemophilus influenzae
: a focus on the capsule locus. BMC Genomics. 2019;20:733.
10. Crandall H, Christiansen J, Varghese AA, et al. Clinical and molecular epidemiology of invasive Haemophilus influenzae
serotype a infections in Utah children. J Pediatric Infect Dis Soc. 2020;9:650–655.
11. Erwin AL, Sandstedt SA, Bonthuis PJ, et al. Analysis of genetic relatedness of Haemophilus influenzae
isolates by multilocus sequence typing. J Bacteriol. 2008;190:1473–1483.
12. Kaur R, Morris M, Pichichero ME. Epidemiology of acute otitis media in the postpneumococcal conjugate vaccine era. Pediatrics. 2017;140:e20170181.
13. Biesbroek G, Wang X, Keijser BJ, et al. Seven-valent pneumococcal conjugate vaccine and nasopharyngeal microbiota in healthy children. Emerg Infect Dis. 2014;20:201–210.