Clinical Characteristics and Molecular Epidemiology of Invasive Nontypeable Haemophilus influenzae at a Tertiary Medical Center in Washington, DC Authors: Jose Lucar, Stephen Stone, Matthew Spence, Marc O Siegel, Afsoon Roberts, Tara N Palmore, Maria Elena Ruiz, Huma Saleem, Sarah Lohsen, Charlotte J Royer, Lauren F Collins, Rebecca Yee, Sarah W Satola Published: 2026 ABSTRACT AbstractWe observed a sharp rise in invasive nontypeable Haemophilus influenzae disease at a tertiary medical center in Washington, DC. Nine adults presented over 9 months with severe illness, with both clonal and nonclonal strains identified by genome sequencing. These findings suggest emerging transmission patterns and the need for expanded surveillance.Keywords: invasive bacterial infections, molecular epidemiology, nontypeable haemophilus influenzae, people with HIV, whole-genome sequencing Nontypeable Haemophilus influenzae (NTHi) has become the predominant cause of invasive H. influenzae disease in the United States following widespread type b vaccination, which has dramatically reduced disease caused by encapsulated strains [1–4]. U.S. surveillance data have reported increased NTHi incidence across multiple age groups, and international studies similarly highlight a predominance of NTHi among invasive isolates [1, 2, 4–6]. Recent population-based studies have also identified distinctive demographic patterns among adults with invasive NTHi. In metropolitan Atlanta, clusters disproportionately involved Black men, people with HIV (PWH), and men who have sex with men (MSM), with whole-genome sequencing (WGS) confirming repeated involvement of 2 clonal lineages termed cluster 1 (C1) and cluster 2 (C2) [7, 8]. These strains were associated with musculoskeletal infections and geographic proximity, suggesting that interconnected social networks may influence transmission. Beginning in late 2022, clinicians at our tertiary medical center in Washington, DC noted an unexpected increase in adult invasive NTHi cases, involving demographic groups similar to those previously reported in Atlanta [7, 8]. In response, we conducted a retrospective assessment to describe the clinical presentations, outcomes, and molecular characteristics of invasive NTHi identified at our institution. METHODS We conducted a retrospective review of adults hospitalized with invasive NTHi at the George Washington University Hospital in Washington, DC, between December 2022 and August 2023. Cases were identified through microbiology laboratory records. Invasive disease was defined as isolation of NTHi from a normally sterile site or clinical presentations indicating infection extending beyond the respiratory tract or skin. Haemophilus influenzae was isolated from patient specimens using standard culture methods on chocolate agar (Thermo Scientific Remel, Waltham, MA, USA) incubated at 37°C with 5% CO2, and species identification was performed using the Bruker MALDI Biotyper MBT Smart MS instrument (Bruker Daltonics, Bremen, Germany) with the FDA-approved database. β-Lactamase production was assessed using BD BBL Cefinase β-Lactamase Detection Discs (BD Diagnostics, Sparks, MD, USA) in accordance with the manufacturer's instructions. Data abstracted from the electronic medical record included demographics, comorbidities, HIV status, antiretroviral therapy use, CD4 cell count, HIV viral load, and documented sexual behavior when available. Clinical manifestations, microbiologic findings, antimicrobial therapy, length of hospital stay, intensive care unit (ICU) admission, and all-cause in-hospital mortality were recorded. Frozen stocks of patient isolates were sent to the Georgia Emerging Infections Program laboratory (Atlanta, Georgia) for whole-genome sequencing. Upon receipt, frozen stocks were plated onto BD BBL Chocolate II Agar (BD Diagnostics, Sparks, MD, USA) and incubated overnight at 37°C with 5% CO2. Cluster designation (C1, C2, or noncluster) was assigned according to previously established genomic criteria [7, 8]. Laboratory personnel were blinded to all clinical information. Continuous variables were summarized as medians or means with ranges, and categorical variables as counts with percentages. No inferential statistics were performed due to small sample size. The George Washington University Institutional Review Board determined that the study met criteria for exemption. Bioinformatics Methods Primary analytics, including read preparation, genome assembly, and MLST, were conducted using Bactopia v3.1.0 as described previously [8, 9]. Patient H. influenzae assemblies generated by Bactopia were aligned with Parsnp v2.1.5 [10]. This alignment included the complete C1 and C2 genomes published in 2025 (GA-C1 and GA-C2, respectively) for comparison, with GA-C2 used as the reference genome for alignment [7, 8]. ClonalFrameML v1.13 [11] was used to generate a core genome–masked alignment, and the resulting phylogeny was visualized using FigTree v1.4.4 [12]. The tool snp-dists v1.2.0 [13] was used to produce an all-versus-all matrix of single nucleotide polymorphisms (SNPs), which was visualized using Python 3.12 with the packages pandas [14], Seaborn [15], and matplotlib [16]. RESULTS Patient Characteristics, Clinical Manifestations, and Outcomes Nine adults with invasive NTHi were identified. The median age was 55 years (range 32–76). All individuals were male, and 8 (89%) were Black. Six (67%) identified as MSM. Five patients (56%) were PWH receiving antiretroviral therapy; the median CD4 count among PWH was 348 cells/mm3 (range 32–984), and 4 were virologically suppressed. Table 1 summarizes demographic and clinical characteristics. Table 1.Characteristics of Patients With Invasive Nontypeable Haemophilus Influenzae Infection.PatientAgeSexRaceHIV infectionCD4 count (cells/mm3)MSMClinical presentationPositive culture source(s)BacteremiaMSK manifestationsClonal clusterAntibiotic agents used during treatmentLOS, DaysOutcomePT-163MaleBlackYes566-Acute cough, found down at home with hypothermia, bradycardia, septic shock, diagnosed with bilateral pneumoniaBloodYesNoNCPiperacillin/tazobactam, ceftriaxone17DischargePT-235MaleBlackYes984YesAcute left shoulder pain and feverBlood, synovial fluidYesMonoarthritisC1Ceftriaxone, ampicillin, cefdinir11DischargePT-355MaleBlackYes129YesAcute polyarthralgia (bilateral hands and wrists, bilateral knees, bilateral elbows), feverBloodYesPolyarthritis, rhabdomyolysisNCCeftriaxone, amoxicillin/clavulanate28DischargePT-432MaleBlackNo-YesAcute right knee and right shoulder painSynovial fluidNoPolyarthritisN/ACeftriaxone, amoxicillin11DischargePT-541MaleBlackYes110YesAcute left elbow and right shoulder painBloodYesTenosynovitis, monoarthritisC2Piperacillin/tazobactam, ceftriaxone, amoxicillin5DischargePT-660MaleBlackNo--Found down with acute encephalopathy, found to be in septic shock, multiorgan failure, bowel ischemiaBloodYesNoC2Cefepime, piperacillin/tazobactam3DeathPT-742MaleBlackYes32YesBilateral knee pain and feverSynovial fluidNoPolyarthritisNCCeftriaxone, ampicillin12DischargePT-856MaleBlackYes803YesAltered mental status, septic shockBloodYesNoC2Piperacillin/tazobactam1DeathPT-976MaleWhiteNo--Septic shock, aortic valve endocarditisBloodYesNoN/APiperacillin/tazobactam, ceftriaxone, meropenem17DeathAbbreviations: C1, clonal cluster 1; C2 clonal cluster 2; HIV, human immunodeficiency virus; LOS, length of hospital stay; MSK, musculoskeletal; MSM, men who have sex with men; N/A, isolate not available for testing; NC, noncluster. Characteristics of Patients With Invasive Nontypeable Haemophilus Influenzae Infection. Abbreviations: C1, clonal cluster 1; C2 clonal cluster 2; HIV, human immunodeficiency virus; LOS, length of hospital stay; MSK, musculoskeletal; MSM, men who have sex with men; N/A, isolate not available for testing; NC, noncluster. Musculoskeletal disease was common: 5 patients presented with septic arthritis and/or osteomyelitis. Other presentations included bacteremia without a clear focal source and mixed invasive syndromes. Empiric therapy consisted primarily of ceftriaxone-based regimens tailored to final culture results. All 9 isolates tested negative for beta-lactamase production. The mean hospital stay was 11 days (range 3–28). Three patients required ICU admission, all of whom died during hospitalization. Genomic Characterization Seven of the 9 patients' isolates were available for whole-genome sequencing. One isolate belonged to C1 (PT-2), 3 to C2 (PT-5, PT-6, PT-8), and 3 were noncluster (PT-1, PT-3, PT-7), demonstrating cocirculation of both known clonal lineages and unrelated strains during the study period (Figure 1). Characterization revealed C1 and C2 samples were closely related, forming distinct lineages (Figure 1A), with comparatively few SNPs between lineage members (Figure 1B). C1 (PT-2) and C2 (PT-5, PT-6, PT-8) isolates belong to ST164 and ST1714, respectively. The noncluster isolates from PT-1 and PT-3 belong to ST436 and ST44 while the isolate from PT-7 did not match any known ST. These findings are consistent with previously described characteristics of the C1 and C2 lineages [7, 8]. Phylogenetic and SNP analyses of nontypeable Haemophilus influenzae patient isolates. A, Phylogenetic tree of patient nontypeable Haemophilus influenzae (NTHi) isolates. Patients were assigned study identifiers in the format PT + number (eg, PT-1, PT-2). GA-C1 and GA-C2 refer to complete reference genomes of C1 (ST164) and C2 (ST1714), published as part of the Georgia Emerging Infections Program NTHi surveillance project [7, 8]. H. influenzae assemblies were aligned using Parsnp [10], with C2 selected as the reference genome for alignment. A masked core-genome alignment was generated with ClonalFrameML [11], and the resulting phylogenetic tree was produced using FigTree [12]. B, Cluster map of single-nucleotide polymorphisms (SNPs) distances between patient samples and the 2 reference strains. The SNP matrix was generated using snp-dists [13] and visualized in Python [16]. DISCUSSION This case series describes a notable increase in invasive nontypeable Haemophilus influenzae (NTHi) among adults hospitalized at a tertiary medical center in Washington, DC over a 9-month period. The cohort was composed predominantly of Black men, many of whom were PWH and/or MSM. Musculoskeletal infections, including septic arthritis and osteomyelitis, were frequent, and clinical severity was substantial, with all intensive care unit admissions resulting in death. Whole-genome sequencing (WGS) demonstrated both recognized clonal clusters (C1/C2) and unrelated lineages, indicating that multiple transmission patterns were likely operating simultaneously. These findings align with U.S. epidemiologic trends demonstrating an increase in invasive NTHi across all age groups following widespread type b vaccination [1–4]. The overrepresentation of Black men, PWH, and MSM closely parallels observations from metropolitan Atlanta, where clusters of invasive NTHi were linked to C1 and C2 lineages and associated with both social networks and musculoskeletal disease [7]. The detection of a mix of clonal and nonclonal isolates in our cohort also mirrors broader U.S. molecular surveillance data, which report heterogeneous strain populations that include both sporadic and cluster-associated infections [1]. International surveillance reinforces the predominance of NTHi among invasive H. influenzae isolates and highlights increasing antimicrobial resistance, an area of emerging concern [5, 6]. The strengths of this study include the use of detailed clinical data, integration of molecular characterization through WGS, and the identification of a demographic signal consistent with previously described risk networks. The ability to compare clinical and genomic patterns strengthens the relevance of the findings to both clinical practice and public health response. However, several limitations warrant consideration. The sample size was small, limiting generalizability and precluding statistical inference. Only 7 of 9 isolates were sequenced due to nonviability of the other 2. Behavioral information may have been under-documented in clinical notes, resulting in potential misclassification. As a single-center, retrospective analysis, the study may not reflect broader regional trends. In conclusion, we identified a short-interval rise in invasive NTHi among adults in Washington, DC, marked by a demographic pattern similar to prior U.S. cluster-associated outbreaks and characterized by severe presentations, frequent musculoskeletal involvement, and mixed clonal and nonclonal strains. These findings underscore the need for expanded genomic epidemiologic surveillance to better understand transmission networks and guide prevention strategies for invasive NTHi in adult populations. Notes Acknowledgments. We honor the memory of Dr. Lauren F. Collins, whose incredible kindness, fierce devotion to her patients, and unwavering advocacy for programs that supported them, touched countless lives. She possessed an extraordinary ability to bring people together, fostering community, collaboration, and compassion wherever she worked. We extend our heartfelt condolences to her family, colleagues, friends, and patients for their profound loss. We acknowledge the work of the Georgia Emerging Infections Program (EIP) and staff who collected and maintained these data and/or specimens. Data availability statement. Haemophilus influenzae whole-genome sequencing data generated for this study are available in the NCBI Sequence Read Archive (SRA) under BioProject PRJNA1427304. The complete reference genomes for isolates C1 and C2 were previously published and are accessible under GenBank accession numbers GCA_044096575.1 and GCA_044096445.1, respectively (BioProject PRJNA544724). Financial support . This study was supported in part by the Investigational Clinical Microbiology Core (ICMC), which was supported by the Department of Medicine, Division of Infectious Diseases, Emory University School of Medicine. The Georgia EIP project described was supported by Cooperative Agreement Number NU50CK000645 from the Centers for Disease Control and Prevention (CDC). Its contents are solely the responsibility of the authors and do not necessarily represent the official views of the CDC. Contributor Information Jose Lucar, Department of Medicine, Division of Infectious Diseases, The George Washington University School of Medicine and Health Sciences, Washington, DC, USA. Stephen Stone, Department of Medicine, Division of Infectious Diseases, The George Washington University School of Medicine and Health Sciences, Washington, DC, USA. Matthew Spence, Department of Medicine, Division of Infectious Diseases, The George Washington University School of Medicine and Health Sciences, Washington, DC, USA; Department of Pharmacy Services, George Washington University Hospital, Washington, DC, USA. Marc O Siegel, Department of Medicine, Division of Infectious Diseases, The George Washington University School of Medicine and Health Sciences, Washington, DC, USA. Afsoon Roberts, Department of Medicine, Division of Infectious Diseases, The George Washington University School of Medicine and Health Sciences, Washington, DC, USA. Tara N Palmore, Department of Medicine, Division of Infectious Diseases, The George Washington University School of Medicine and Health Sciences, Washington, DC, USA. Maria Elena Ruiz, Department of Medicine, Division of Infectious Diseases, The George Washington University School of Medicine and Health Sciences, Washington, DC, USA. Huma Saleem, Department of Medicine, Division of Infectious Diseases, The George Washington University School of Medicine and Health Sciences, Washington, DC, USA. Sarah Lohsen, Department of Medicine, Division of Infectious Diseases, Emory University School of Medicine, Atlanta, Georgia, USA; Georgia Emerging Infections Program, Atlanta, Georgia, USA. Charlotte J Royer, Department of Medicine, Division of Infectious Diseases, Emory University School of Medicine, Atlanta, Georgia, USA; Georgia Emerging Infections Program, Atlanta, Georgia, USA. Lauren F Collins, Department of Medicine, Division of Infectious Diseases, Emory University School of Medicine, Atlanta, Georgia, USA; Georgia Emerging Infections Program, Atlanta, Georgia, USA. Rebecca Yee, Department of Medicine, Division of Infectious Diseases, The George Washington University School of Medicine and Health Sciences, Washington, DC, USA. Sarah W Satola, Department of Medicine, Division of Infectious Diseases, Emory University School of Medicine, Atlanta, Georgia, USA; Georgia Emerging Infections Program, Atlanta, Georgia, USA. References Associated Data Data Citations