High pathogenicity of human adenovirus type 55 in adult males: a meta-analysis and phylogenetic study of HAdV-55 epidemics worldwide over the past six decades

ABSTRACT

Human adenovirus type 55 (HAdV-55), formerly designated as HAdV-11a, is a notable recombinant pathogen linked to severe respiratory infections. Since its reclassification as HAdV-55 in 2010, numerous outbreaks of pneumonia associated with it have been documented over the past decade. We conducted a systematic review of 111 peer-reviewed articles from PubMed, Web of Science, and China National Knowledge Infrastructure (CNKI) spanning the 60 years (1965–2024). Furthermore, we performed phylogenetic analysis on 87 complete HAdV-55 genomes available in GenBank. Bayesian methods were employed to estimate the temporal evolutionary dynamics of HAdV-55. A total of 15,245 HAdV-55 cases were identified, including 3,790 pneumonia cases (24.86%) and 33 fatalities (case fatality rate: 0.22%), of which 21 (63.63%) were male and the remainder unspecified. Military recruits comprised 79.06% of the cases (12,039/15,245), with adults (95.39%) and males (97.61%) predominantly affected. Asia accounted for the majority of cases (15,060, 98.79%), primarily in China (89.90%), followed by North America (0.98%). The infections caused by HAdV-55 are associated with respiratory (87.27%), gastrointestinal (2.71%), ocular (0.30%), and mental (0.07%) disorders. Phylogenetic analysis identified four variants: HAdV-55a through HAdV-55d, with HAdV-55d emerging as the dominant circulating strain since 2006. Additionally, the E3 region harbours distinct amino acid substitutions correlated with specific countries, suggesting evolution for immune evasion. Bayesian evolutionary analysis estimated the most recent common ancestor (tMRCA) originated in 1929 (95% HPD: 1890–1967). This study underscores the high prevalence of HAdV-55 among adults and males, particularly in China, highlighting the urgent necessity for antiviral therapeutics and vaccines.

Introduction

Human adenovirus (HAdV) is a non-enveloped, icosahedral, double-stranded DNA virus with a genome size of approximately 36 kb, classified within the genus Mastadenovirus of the family Adenoviridae [1]. The major capsid structural proteins of adenovirus include the penton base, hexon, and fibre [2]. Historical identification and classification of novel human adenoviruses relied on serological methods, which categorized them into seven species (HAdV-A to -G) based on their biological properties. However, the increasing emergence of recombinant adenoviruses has shifted the focus towards bioinformatics and genomic analysis as primary methods for typing [3]. Genomic analysis enables the systematic identification of novel adenovirus genotypes, which may arise from recombination events in coding gene regions of critical structural proteins (such as the penton base, hexon, and fibre) among existing types. Currently, 117 HAdV genotypes are archived in GenBank (http://hadvwg.gmu.edu) [4].

Serological methods used for human adenovirus identification and classification led to subsequent misidentification of HAdV-55 strains as HAdV-11a, based on the hexon epitopes. The earliest known HAdV-11a strain can be traced to BC34, identified in China in 1965 [5]. The first global detection of HAdV-11a infection occurred in 1969 at a Spanish military training base [6], followed by sporadic outbreaks in Turkey [7], Singapore [8], North America [9], South America [10], and China [11]. A pivotal reclassification occurred when Walsh et al. conducted a whole-genome reanalysis of the QS-DLL strain isolated from an outbreak in Shaanxi, China, revealing HAdV-55 as a recombinant of HAdV-11 and HAdV-14 [12–14]. Consequently, this strain was reclassified as a P14H11F14 recombinant type and renamed HAdV-55 [12].

HAdV-11, HAdV-14, and HAdV-55 are categorized as species B adenoviruses, with differing pathogenic attributes: HAdV-11 is a renal pathogen linked to urinary tract infections (the misidentified HAdV-11a variant was associated with conjunctivitis and pneumonia), while HAdV-14 and HAdV-55 are respiratory tract pathogens associated with respiratory disease [15,16] that can range from mild upper respiratory symptoms (e.g. fever, cough) to severe pneumonia and death. A clinical investigation involving patients aged ≥7 years (with HAdV-55 cases aged 14–80) has demonstrated that HAdV-55 and HAdV-7 infections are generally more severe than the other HAdV types (HAdV-3, HAdV-14, HAdV-50, and HAdV-2), with longer durations of fever, higher rates of oxygen therapy needs, and a greater need for mechanical ventilation [17]. These cases frequently lead to household clusters and community-acquired pneumonia outbreaks [18,19]. Importantly, HAdV-55 was not reported in Europe and North America following the initial, sporadic outbreaks; however, it persists endemically in China and South Korea, reflecting a shift in its global transmission dynamics [20]. The continued relevance of HAdV-55 as a lethal pathogen has been underscored by a recent case study reporting a fatal neonatal pneumonia caused by this recombinant strain in the United States [21], highlighting the need for ongoing global surveillance. Comprehensive global data on the prevalence and clinical impacts of HAdV-55 remain limited.

Since the first identification of HAdV-55 in 1965, there have been no systematic epidemiological investigations. To address this gap, we conducted a systematic review of the research literature on HAdV-55, including under its previous name as HAdV-11a, from 1965 to 2024 with the objective of analyzing its global distribution from an epidemiological perspective. Concurrently, we performed a phylogenetic analysis of all publicly available whole-genome sequences of HAdV-55 to investigate its transmission patterns. These analyses are designed to provide a theoretical foundation for the development of antiviral drugs and vaccines, as well as to inform the development of disease prevention and control policies.

Methods

Electronic document retrieval and literature screening

A systematic search for peer-reviewed articles was conducted across PubMed, Web of Science, and the China National Knowledge Infrastructure (CNKI) databases, following the search strategy ((Adenovirus) OR (Adenoviruses)) OR (adenoviral) OR (mastadenovirus) OR (AdV) OR (HAdV) AND (type 55 [Text Word]) OR (subtype 55 [Text Word]) OR (genotype 55 [Text Word]) OR (HAdV-B55 [Text Word]) OR (HAdV-55 [Text Word]) OR (HAdV55 [Text Word]) OR (Ad55 [Text Word]) OR (AdV55 [Text Word]) OR (AdV-55 [Text Word]) OR (HAdV-B [Text Word]) OR (Species B [Text Word]) OR (type 11a [Text Word]) OR (subtype 11a [Text Word]) OR (serotype 11a [Text Word]) OR (genotype 11a [Text Word]) OR (HAdV-B11a [Text Word]) OR (HAdV-11a [Text Word]) OR (HAdV11a [Text Word]) OR (Ad11a [Text Word]) OR (AdV11a [Text Word]) OR (AdV-11a [Text Word]) AND (“1965/01/01” [Date – Publication]: “2024/12/31” [Date – Publication]) NOT Fowl NOT echovirus NOT adenovirus vector (Supplementary Table 1). Eligible studies were selected for inclusion in statistical analysis. Independent assessments were performed by two reviewers who screened titles, abstracts, and full texts in accordance with the following eligibility criteria: (1) abstracts or keywords mentioning adenovirus “type/serotype/subtype/genotype” (including abbreviations such as HAdV55, HAdV-55, Ad55, AdV55, HAdV11a, HAdV-11a, Ad11a, AdV11a, etc.); (2) abstracts or main texts describing surveillance/epidemiological studies, case reports, or laboratory/bioinformatics research reporting original/primary evidence of HAdV-55 detection in human clinical samples. Studies were excluded if they met any of the following criteria: (1) no mention of adenovirus type/serotype/subtype/genotype or relevant abbreviations in the abstract or keywords; (2) publishedafter December 2024; (3) original/unreported/insufficient HAdV-55 detection evidence (e.g. identification of HAdV-55 in clinical samples using non-specific serotyping or genotyping methods); (4) duplicate case reports, epidemiological studies, or genomic analyses; (5) focus on non-human subjects; (6) primarily investigated adenovirus antibody prevalence rather than disease manifestation. All included studies are catalogued in Supplementary Table 2.

Phylogenetic analysis of HAdV-55 complete genomes; penton base, hexon, and fibre genes; and E3 region

A search was conducted in the NCBI database using the keywords “human adenovirus 55,” “human adenovirus 11a,” and “human adenovirus P14H11F14” with sequence length filters set between 30,000 and 40,000 bp. A total of 87 complete HAdV-55 genome sequences were retrieved. Multiple sequence alignment was performed using MAFFT (Multiple Alignment Fast Fourier Transform) [22]. A Neighbor-Joining (NJ) phylogenetic tree was constructed utilizing MEGA 12.0.9 [23]. The Maximum Likelihood Method was employed with a Bootstrap value set to 1,000, while all other parameters were maintained at their default settings.

Temporal evolution analysis of HAdV-55 complete genome sequence

We compiled all available complete genome sequences of HAdV-55 from GenBank that included collection dates and locations, excluding any data lacking collection dates (7 genome sequences excluded). This resulted in a dataset of 80 HAdV-55 genome sequences for the analysis of temporal evolutionary dynamics. Initially, ModelTest-ng v0.1.7 was utilized to determine the optimal substitution model for the HAdV-55 complete genome nucleotide sequence matrix, identifying the HKY model based on the Akaike Information Criterion (AIC) [24]. Subsequently, BEAST v1.10.4 was employed to analyze the complete genome, three major capsid protein genes and E3 region of HAdV-55 and to calculate nucleotide evolutionary rates as well as the time to tMRCA [25]. An uncorrelated relaxed log-normal molecular clock model in conjunction with a population exponential growth model was implemented, with a total of 5 × 10⁸ iterations and sampling conducted every 1,000 generations. Tracer v1.7.2 was used to assess the output data from BEAST, ensuring that all model parameters achieved effective sample sizes (ESS) exceeding 200, along with a burn-in rate of 10%. Additionally, a maximum clade credibility (MCC) tree was constructed using posterior probability methods and analyzed with FigTree v1.4.4.

Data extraction and statistical analysis

Relevant data (e.g. HAdV-55 infection, time, geographic location, disease characteristics, and patient demographics) were extracted from full-text articles using specially designed data collection forms. The data were imported into Graphpad Prism 6.0 for correlation analysis.

Results

Literature screening and geographical distribution of HAdV-55 reports

The literature search for abstracts and keywords was conducted across three databases. Initial screening yielded 386 articles from PubMed, 342 from Web of Science, and 117 from CNKI. After deduplication, 611 articles were retained for inclusion. Further analysis of the abstracts and full texts of these 611 articles resulted in the selection of 100 articles plus 11 articles obtained by manual search that met the eligibility criteria (Figure 1, Supplementary Table 1). Among these, 107 articles reported detailed case numbers of HAdV-55 infections. Including multi-regional studies, the geographic distribution of publications was as follows: Asia (101, 94.39%), North America (3, 2.80%), Europe (2, 1.87%), South America (2, 1.87%), and Africa (1, 0.93%). Notably, China (77.47%) and South Korea (11.71%) accounted for the highest shares of publication. The pooled data documented a total of 15,245 HAdV-55 infections, including 3,790 pneumonia cases (24.86% of total infections), with 33 reported fatalities (case fatality rate, CFR: 0.22%). Reported infection clusters ranged from 1 to 2,109 cases per study, with the majority of cases originating from military populations.

Figure 1.

PRISMA flow diagram of the identification, screening, and review of the medical and scientific literature on HAdV-55 infections published from 1965 to 2024.

Geographical distribution of HAdV-55 infections

In the reported 15,245 cases of HAdV-55 infection, 15,060 cases originated from Asia (98.79%), followed by 149 cases from North America (0.98%), 15 cases from Europe (0.10%), 11 cases from Africa (0.07%), and 10 cases from South America (0.07%) (Figure 2a).

Figure 2.

(a) Cases of HAdV-55 infections reported in the medical and scientific literature, by geographic region of the world, 1965–2024. Military infections: Asia (12,039/99.89%), North America (0/0.00%), Europe (13/0.11%), South America (0/0.00%), Africa (0/0.00%). Civilian infections: Asia (3,021/94.61%), North America (149/4.67%), Europe (2/0.06%), South America (11/0.34%), Africa (10/0.31%). (b) Cases of military and civilian HAdV-55 infections reported in the medical and scientific literature, 1965–2024. Military infections: 1965–1994 (13/0.11%), 1995–2004 (0/0.00%). 2005–2014 (1,868/15.50%); 2015–2024 (10,171/84.39%). Civilian infections: 1965–1994 (16/0.50%), 1995–2004 (150/4.70%). 2005–2014 (703/22.07%); 2015–2024 (2,324/72.78%). (c) Percents of military and civilian HAdV-55 infections reported in the medical and scientific literature, 1965–2024. (d) Number of peer-reviewed publications reporting HAdV-55 infections in the medical and scientific literature each decade from 1965 to 2024. Military infections: 1965–1994 (1/2.86%), 1995–2004 (0/0.00%). 2005–2014 (4/11.43%); 2015–2024 (30/85.71%). Civilian infections: 1965–1994 (1/1.28%), 1995–2004 (3/3.85%). 2005–2014 (15/19.23%); 2015–2024 (59/75.64%).

Among the reported 33 fatalities, 30 (90.91%) were from China, two from South Korea (6.06%), and one from Turkey (3.03%). A detailed geographic distribution of HAdV-55 cases can be found in Table 1 and Figure 2a. Of the 15,060 infection cases reported in Asia, 89.90% were from China (13,539 cases), 9.41% from South Korea (1,417 cases), 0.28% from Iran (42 cases), 0.17% from Singapore (26 cases), 0.13% from Japan (19 cases), 0.06% from Israel (9 cases), 0.03% from Taiwan, China (5 cases), and 0.02% from Turkey (3 cases). The 149 cases reported in North America were entirely from the United States (100%). Among the European cases, 13 were from Spain (86.67%) and 2 from France (13.33%). All 10 cases in South America originated from the Southern Cone region (100%). The 11 cases in Africa were entirely from Egypt (100%). In the report on the 15,060 HAdV-55 infection cases, military recruits/personnel (12,052 cases) represented 79.06% of infections, while the remaining 20.94% were civilian cases (3,193 cases) (Figure 2b-c). Both military and civilian infections predominantly occurred in Asia.

Table 1.

Geographical distribution of human adenovirus type 55 (HAdV-55) infections, 1965–2024. The bold text should be set to regular weight.

*The data from the Southern Cone lacks detailed source classification.

** The 21 recorded fatalities, which have clear gender classification, were male; the remaining 12 did not describe gender.

Children: age < 18 years old; Adults: age ≥ 18 years old. The percentages for age and gender were derived using the counts of patients with non-missing data for each characteristic as the denominator.

Time distribution of HAdV-55

Prior to 2015, there were 23 reported studies on HAdV-55 infections. Over the past decade, the number of publications documenting HAdV-55 has increased to 88 articles, representing a 3.8-fold rise compared to the pre-2015 period, accounting for 79.28% of all publications (Figure 2d). In the past decade, the reported cases of HAdV-55 have risen to 12,495, reflecting a 3.4-fold increase, including 2,324 civilian infection cases, reflecting approximately a 2.7-fold increase compared to the previous ten years (Figure 2b). Similarly, military infections among recruits/personnel reported in the last decade reached 10,171 cases, making up 84.39% of all military-related infections (Figure 2b). These trends underscore the rapid escalation of HAdV-55 infections in both civilian and military populations. It may be attributed to HAdV-55 being a recombinant strain derived from HAdV-11 and HAdV-14, which was historically classified as a subtype of HAdV-11 (HAdV-11a). Such classification may have led to misdiagnoses of related cases as HAdV-11 infections. Over the last decade, the annual number of publications on HAdV-55 has demonstrated a consistent upward trend, with HAdV-55 emerging and spreading widely in Asia, particularly in China and South Korea. The articles focusing on civilian populations comprised 67.57%, while those involving military populations constituted 32.43% (Figure 2d). Both civilian and military cases of HAdV-55 infection have increased over time. Notably, while studies reporting civilian infections have outnumbered those focused on military populations in the past decade (Figure 2d), the actual case count in civilian populations has been significantly lower than in military settings (Figure 2b). Since the publication years may not align with the actual years of HAdV-55 infection occurrence in the reports, Supplementary Table 2 provides data on the specific years covered in each study.

Clinical features of HAdV-55

In the literature examining the clinical manifestations of HAdV-55 infections, the predominant symptoms reported were respiratory diseases, followed by gastrointestinal and ocular conditions, with very few instances of neurological and psychiatric disorders (Figure 3a). Initial disease presentations typically manifested as self-limiting upper respiratory tract infections, which could progress to pneumonia, posing significant fatal risks. Among 95 articles detailing disease types, 92 articles (96.84%) documented respiratory conditions (13,305 cases, 87.27%), including influenza-like illness (fever, chills, fatigue, headache, dizziness, myalgia, etc.), acute upper respiratory infections, acute respiratory distress syndrome (ARDS), respiratory failure, bronchitis, pharyngitis, and pneumonia (3,790 cases, 24.86%). Gastrointestinal diseases (413 cases, 2.71%) were noted in 30 articles (31.58%), presenting as diarrhoea, abdominal pain, nausea, vomiting, and gastrointestinal bleeding. Ocular conditions (45 cases, 0.30%) were described in 12 articles (12.63%), primarily as conjunctival congestion and conjunctivitis. Only 6 articles (6.32%) reported psychiatric symptoms (11 cases, 0.07%), including insomnia, mania, paranoia, and consciousness disorders; however, no cases of encephalitis were documented. Detailed number of cases and source region are available in Figure 3b and Table 1.

Figure 3.

(a) Number of peer-reviewed publications reporting different HAdV-55 symptoms in the medical and scientific literature from 1965 to 2024. Among 95 articles detailing disease types, respiratory (92/96.84%), gastrointestinal (30/31.58%), ocular (12/12.63%), and mental (6/6.32%). (b) Cases of different HAdV-55 symptoms in the medical and scientific literature from 1965 to 2024. Respiratory (92/96.84%). Gastrointestinal (30/31.58%), Ocular (12/12.63%); Mental (6/6.32%).

Among the 33 documented fatalities, two were from the South Korean military and one from the Turkish military, with causes of death not specified in detail. Three cases were from the Chinese military, whose deaths were attributed to multiple organ dysfunction syndrome (MODS) induced by HAdV-55 infection. The remaining 27 cases were all Chinese civilians. Among these, one case died of HAdV-55 infection accompanied by megaloblastic anaemia, 11 cases succumbed to acute respiratory distress syndrome (ARDS) or MODS resulting from severe pneumonia, and 4 cases died of complications from bacterial or fungal co-infections, including catheter-related bloodstream infection (CRBSI), invasive pulmonary aspergillosis (IPA), and septic shock. The causes of death for the remaining 11 cases were not clearly elucidated. Notably, while military infection numbers in the research literature significantly exceeded civilian cases, military fatalities remained disproportionately low (6 cases). This discrepancy may be related to the military's non-disclosure of relevant data. Detailed mortality information and source articles are available in Table 1 and Supplementary Table 2.

High prevalence of HAdV-55 in adults and males

Among the 111 articles on HAdV-55 infection, gender data were available for 4,638 cases. Male cases significantly outnumbered female cases (4,527, 97.61% vs. 111, 2.39%), suggesting higher susceptibility to HAdV-55 among males (Figure 4a–b). Regarding fatal outcomes, the gender disparity persisted: of the 33 deaths, 21 (63.63%) were male, none were female, and the gender of the remaining 12 (36.36%) cases was undetermined. Moreover, analysis of age distribution among the 10,683 patients revealed that the majority of infections occurred in adults (≥18 years old), accounting for 10,190 cases (95.39%), compared with only 493 cases (4.61%) in children (<18 years old) (Figure 4c–d). Of the adult cases, 7,721 (72.27%) were military recruits/personnel highlighting a heightened susceptibility to clustered HAdV-55 outbreaks in military settings. After excluding military recruits/personnel, 2,469 (23.11%) adult cases remained, suggesting that adults constitute the main susceptible population in the general community.

Figure 4.

(a) Cases of different gender HAdV-55 infections reported in the medical and scientific literature, 1965–2024. Male infections: 1965–1994 (15/0.33%); 1995–2004 (124/2.74%); 2005–2014 (436/9.63%); 2015–2024 (3,952/87.30%). Female: 1965–1994 (1/0.90%), 1995–2004 (22/19.82%); 2005–2014 (14/12.61%); 2015–2024 (74/66.67%). (b) Percents of different gender HAdV-55 infections reported in the medical and scientific literature, 1965–2024. (c) Cases of different age HAdV-55 infections reported in the medical and scientific literature, 1965–2024. Adult infections: 1965–1994 (13/0.13%); 1995–2004 (150/1.47%); 2005–2014 (607/5.96%); 2015–2024 (9,420/92.44%). Children and Teenagers: 1965–1994 (16/3.25%), 1995–2004 (0/0.00%); 2005–2014 (286/58.01%); 2015–2024 (191/38.74%). (d) Percents of different age HAdV-55 infections reported in the medical and scientific literature, 1965–2024.

Phylogenetic analysis of HAdV-55 complete genomes; penton base, hexon, fibre genes; and E3 region

We conducted a phylogenetic analysis of 87 complete genome sequences of HAdV-55 sourced from the GenBank database, originating from various countries across North America, Europe, and Asia, specifically including the United States, Spain, Japan, Egypt, Argentina, China, and South Korea (Figure 5a). The phylogenetic tree constructed from these sequences revealed five distinct lineages. Lineage 1 (Red) included the prototype HAdV-11/14 strain 273 (MN654395.1, 1969, Spain), 273/P14H11F14 (MG905110.1, 1969, Spain), and CADOH VRDL 76-0669 (MN654394.1, 1976, U.S.), which we designated as the HAdV-55p prototype. Lineage 2 (Green) comprised two Chinese strains and 13 South Korean strains, designated as HAdV-55a variant. Lineage 3 (Blue) included three strains: CQ-814 (JX123027.1, 2010, China), ak36_AdV11a (JX121, 2005, Argentina), and CHOP6887 (MT513753.1, 2006, U.S.), designated as HAdV-55b variant. Lineage 4 (Purple) consisted of 16 strains from various countries, with Egyptian strains accounting for the largest proportion, followed by those from Singapore, the United States, and Japan, designated as HAdV-55c variant. The remaining 50 strains formed Lineage5 (Yellow), comprising strains that have been continuously circulating in China in recent years, designated as HAdV-55d variant. These findings indicate the persistent regional transmission of HAdV-55 in specific areas. Notably, genomic analysis showed high whole-genome conservation (99.80% identity), with intra-lineage identity exceeding 99.94%, indicating that all lineages of HAdV-55 exhibit a high degree of genetic conservation. However, the prototype strain 55p exhibited greater nucleotide divergence (0.056%−0.172%) than other lineages, indicating that prolonged circulation and mutation led descendant lineages (HAdV-55a-55d) to accumulate significant genetic variations and differentiate into distinct variants.

Figure 5.

(a) Phylogenetic analysis of the whole genome of HAdV-55 (n = 87). Lineage 1 (HAdV-55p prototype) in Red: The prototype strain and early U.S. isolate (1976); Lineage 2 (HAdV-55a variant) in Green: All South Korea isolates and 2 China isolates; Lineage 3 (HAdV-55b variant) in Blue: China isolate, Argentina isolate, and U.S. isolate; Lineage 4 (HAdV-55c variant) in Purple: All Egypt, Singapore isolates, and closely related isolates; Lineage 5 (HAdV-55d variant) in Yellow: The predominant contemporary variants in China (2006–present). (b) Phylogenetic analysis of the E3 region of HAdV-55 (n = 87). Lineage 1 (HAdV-55p prototype) in Red: The prototype strain and early USA isolate (1976); Lineage 2 (HAdV-55a variant) in Green: All South Korea isolates and 2 China isolates; Lineage 3 (HAdV-55b variant) in Blue: China isolate, Argentina isolate, and USA isolate; Lineage 4 (HAdV-55c variant) in Purple: All Egypt, Singapore isolates, and closely related isolates; Lineage 5 (HAdV-55d variant) in Yellow: The predominant contemporary variants in China (2006–present). (c) Phylogenetic analysis of the E3 15.3 kDa gene of HAdV-55 (n = 87). Lineage 1 (HAdV-55p prototype) in Red: The prototype strain and early USA isolate (1976); Lineage 2 (HAdV-55b variant) in Blue: China isolate, Argentina isolate and USA isolate; Lineage 3 (HAdV-55c variant) in Purple: All Egypt, Singapore isolates and closely related isolates; Lineage 4 (HAdV-55a + d variants) in Yellow: All South Korea isolates and 2 China isolates and the predominant contemporary variants in China (2006–present).

Analysis of the three major capsid genes also revealed high sequence conservation in HAdV-55 strains (Supplementary Figures 3–5). Hexon and penton base genes showed extremely high sequence conservation (>99.975%), with only minor variations: penton base had three mutations in two HAdV-55d strains, while hexon contained just two mutations (Supplementary Table 3). No significant HVR mutations were found, indicating stable antigenicity across all lineages. In contrast, fibre gene exhibited more inter-lineage sequence variation, particularly in the knob domain. While nucleotide identity remained high (99.807%−99.979%), ten amino acid mutations were identified in the fibre, among which 2 mutations occurred in the shaft domain and 8 mutations occurred in the knob domain (Supplementary Table 4). The HAdV-55d variant exhibited the most amino acid mutations. A total of eight strains had amino acid mutation sites in the fibre of HAdV-55d variants, among which the TJ01 strain (KP896484.1, 2013) showed the greatest degree with 6 amino acid mutations. Moreover, compared to the HAdV-55p, all variants shared the H149D knob mutation. These knob domain changes likely enhanced receptor binding and transmission fitness, contributing to the variants’ epidemiological success.

In contrast, the nucleotide similarity of the E3 region was relatively low (99.724%–99.876%). Phylogenetic analysis revealed that the evolutionary tree of the E3 region exhibited a branching structure consistent with that of the whole genome, suggesting variant-specific variations in the E3 region (Figure 5b). This was particularly evident in the E3 15.3 kDa gene, whose phylogenetic clustering closely resembled that of the whole genome (Figure 5c). Further amino acid mutation analysis identified characteristic substitutions in specific variants: a D72N mutation was present in HAdV-55a and HAdV-55d, an A11 V mutation in HAdV-55b, and dual E32H and E113 K mutations in HAdV-55c (Supplementary Table 5). Given the region-specific distribution pattern observed among HAdV-55 variants, we propose that these mutations in the E3 region may have arisen from differential immune pressures in distinct human populations. Such adaptations likely enhance viral fitness in local populations and could facilitate more efficient transmission.

Temporal evolution of HAdV-55 genome sequences

Employing the Bayesian-MCMC approach, we conducted an analysis of the complete genome of HAdV-55 to elucidate its temporal evolutionary dynamics and to estimate its mean nucleotide substitution rate, time to tMRCA, and evolutionary divergence (Figure 6). The mean nucleotide substitution rate for HAdV-55 complete genome was determined to be 1.6872 × 10⁻⁵ substitutions/site/year (95% highest posterior density interval [HPD]: 1.133–2.2733 × 10⁻⁵), the nucleotide substitution rates for the penton base, hexon, and fibre genes were estimated at 8.426 × 10⁻⁶ (95% HPD: 3.9249 × 10⁻⁹ to 2.1766 × 10⁻⁵), 2.174 × 10⁻⁵ (95% HPD: 9.0594 × 10⁻⁶ to 3.6458 × 10⁻⁵), and 2.676 × 10⁻⁵ (95% HPD: 8.2391 × 10⁻⁶ to 4.8467 × 10⁻⁵), respectively. In contrast, the E3 region exhibited the highest substitution rate of 3.898 × 10⁻⁵ (95% HPD: 2.2693 × 10⁻⁵ to 5.5947 × 10⁻⁵). The tMRCA traced back to 1929 (95% HPD: 1717–1967). This suggests that HAdV-55 experienced its initial evolutionary divergence in 1929, resulting in the separation into HAdV-55p and other strains. Subsequent divergences occurred in 1974 (95% HPD: 1963–1986), leading to the emergence of HAdV-55b, and in 1988 (95% HPD: 1977–1999), which resulted in the formation of HAdV-55c. The fourth divergence transpired in 1995 (95% HPD: 1987–2004), culminating in the emergence of HAdV-55a and HAdV-55d, the latter of which have been persistently circulating in China.

Figure 6.

Bayesian estimation of the time to the most recent common ancestor for HAdV-55. The red dots represent the major evolutionary divergences for the five subtypes, and the 95% highest posterior density (HPD) range for the time to tMRCA is indicated. The average nucleotide substitution rate of HAdV-55 was calculated at 1.6872 × 10⁻⁵ mutations/site/year (95% HPD: 1.133–2.2733 × 10⁻⁵). The most recent common ancestor is estimated to have appeared in 1929 (95% HPD: 1890–1967).

Discussion

Since the first reported case of HAdV-55 infection in Spain, this virus has gradually garnered global attention. We conducted a systematic review of articles related to HAdV-55 from 1965 to 2024. Over the years, both the number of publications and reported cases associated with HAdV-55 infections have demonstrated a significant increase (Figure 2b,d). With the exception of Oceania and Antarctica, HAdV-55 cases have been documented across all continents. The rapid increase in the number of HAdV-55 cases can be attributed to two primary factors. First, advancements in adenovirus genotyping methods have supplanted traditional serological typing with sequence-based genotyping, thereby enhancing the accuracy of adenovirus genotype identification. Second, HAdV-55 infection causes severe clinical symptoms, often leading to acute respiratory distress [17,18], and this would lead clinicians to search for a cause. Similar to HAdV-7 infections, HAdV-55 can provoke a severe cytokine storm, characterized also by high viral loads and prolonged shedding durations of up to 55 days [26–28]. Research conducted by Zhao et al. using airway and alveolar organoid infection models demonstrated that HAdV-55 exhibits higher replication efficiency and more specific cell tropism compared to HAdV-3, potentially leading to impaired organ repair and loss of lung cell differentiation, thereby enhancing its pathogenicity [29]. Most HAdV-55-related fatalities resulted from multiple organ dysfunction syndrome or ARDS secondary to severe pneumonia.

We conducted a systematic review of global data on HAdV-55 infections, providing detailed statistics on gender, age, and mortality, limited to studies that explicitly reported these parameters. Our analysis indicates that HAdV-55 infections predominantly affect adults and males, with a CFR of 0.22%. The CFR for HAdV-55 (0.22%) was higher than that previously reported for HAdV-4 (0.11%) [30]. The data compiled in this study show that adult cases account for 95.39% of all reported infections. In contrast, the meta-analysis of HAdV-7 shows that its infections affect both children and adults, with a higher proportion observed in children (56.75%), and children account for 87.63% of all fatal cases [31]. Importantly, HAdV-7 demonstrates a significantly higher overall CFR [31], suggesting greater pathogenic severity in children, particularly associated with the recently prevalent 7E2 variant. In contrast, the high pathogenicity of HAdV-7 in children may be linked to the immature host immune system, enhanced virulence due to viral mutations, and excessive inflammatory activation [15]. However, the factors that facilitate the ease of transmission and also severe outcomes of HAdV-55 infection may relate to dense living conditions and to increased physical and/or psychological stress [32].

Regarding gender distribution, numerous adenovirus genotypes exhibit a male-biased infection trend, including HAdV-B, HAdV-C, and HAdV-E4 [33]. For example, a male predilection has been observed for HAdV-14 infection by multiple reports [34–36]. HAdV-7 infection also appears to exhibit a significant male predilection, with cases in males being 3.21 times more frequent than in females (4,484 vs. 1,396) [31]. The reasons for a male predominance in HAdV infections are not clear, but may be related to factors such as behavioural exposure, immune responses, or receptor expression. Male cases significantly outnumbered female cases in both civilian and military populations, aligning with previous studies that highlight male susceptibility to HAdV-55 [37–39]. Notably, a study on CD46 gene polymorphisms in males infected with HAdV-55 suggested that the TT genotype at rs2724385 on CD46 might confer protective effects [39]. Although CD46 is recognized as a primary receptor for species B adenoviruses [41], recent studies indicate that human desmoglein-2 (DSG2) serves as the main cellular receptor for HAdV-55, with CD46 playing a secondary role in mediating infection [42,43]. Therefore, further investigation into the potential role of DSG2 is warranted. Additionally, Esposito et al. found that most patients infected with HAdV-14 were either smokers or shared smoking materials, indicating that close contact during smoking may facilitate transmission [44], although its impact on HAdV-55 remains unclear. In animal studies, Chen et al. injected HAdV-5 into BALB/c mice via the tail vein and found that 60% of male mice died, whereas all female mice survived. Moreover, sex-specific differences were evident in the down-regulated expression of the H2-Aa, H2-Ea-ps, CD74, and H2-Eb1 genes in male mice and the AHR gene in female mice, providing provocative insights into potential mechanisms underlying male susceptibility to HAdV infections [45]. Overall, the exact mechanisms underlying the sex predilection of HAdV-55 for adult males remain unclear and warrant further in-depth investigation.

To elucidate the evolutionary relationships of HAdV-55, we downloaded 87 complete HAdV-55 genome sequences isolated between 1969 and 2021 from the GenBank database and performed phylogenetic analysis. The results revealed that the phylogenetic tree constructed from these genomes could be divided into five lineages, designated as HAdV-55p, HAdV-55a, HAdV-55b, HAdV-55c, and HAdV-55d. While the findings of Hang et al. align with our conclusion [20], the phylogenetic tree constructed in this study incorporates complete genome sequences spanning 1965–2024, offering a larger sample size, broader temporal coverage, and more diverse genetic variation data. This provides a detailed evolutionary history of viral strains, enhances our understanding of HAdV-55's global transmission patterns and regional genetic divergence, and strengthens insights into its molecular evolution. These findings serve as a critical scientific foundation for developing future prevention and control strategies.

The HAdV-55p comprises three strains: the prototype strain HAdV-11/14 273 (MN654395.1, 1969, Spain), 273/P14H11F14 (MG905110.1, 1969, Spain), and CADOH VRDL76-0669 (MN654394.1, 1976, U.S.). The HAdV-55a consists of two Chinese strains and 13 South Korean strains, clustering closely with the HAdV-55p. This suggests that the transmission chain of HAdV-55 in South Korea remained relatively singular, resulting in limited viral diversification and closer genetic proximity to the HAdV-55p. In contrast, HAdV-55d includes 50 Chinese strains, accounting for 57.4% of all HAdV-55 strains, and forms multiple sublineages in the phylogenetic tree. Additionally, HAdV-55a and -55b lineages were also found in Chinese isolates. These indicate that HAdV-55 likely underwent extended localized transmission and rapid evolution within China. The country's large population and vast geographical area provided favourable conditions for the spread of HAdV-55d, leading to significant evolutionary divergence from strains in other regions. As a major hotspot for HAdV-55, China may harbour large-scale transmission activities with multiple independent transmission chains. To mitigate future risks, it is imperative to strengthen prevention and control measures, implement robust surveillance systems, and rapidly identify and characterize emerging variants to prevent widespread viral proliferation.

The HAdV-55b variant comprises three strains: CQ-814 (JX123027.1, 2010, China), ak36_AdV11a (JX423384.1, 2005, Argentina), and CHOP6887 (MT513753.1, 2006, USA). The CHOP6887 strain, isolated in 2006 from a paediatric patient in Pennsylvania, exhibits significant evolutionary divergence from other U.S. strains but shares nearly identical sequence homology with the Argentine ak36_AdV11a strain, differing only by a 13-nucleotide deletion in an untranslated region [46]. Chen et al. observed that CQ-814 and ak36_AdV11a had only 17 single-nucleotide substitutions across their complete genomes [47]. These results demonstrate high sequence similarity among HAdV-55b strains, indicating possible cross-country spread among USA, Argentina, and China.

The HAdV-55c variant consists of 16 strains distributed across multiple countries and regions, including North America, South America, and Africa. Egyptian strains dominate this lineage, followed by those from Singapore, suggesting that Egypt and Singapore may represent key endemic regions for HAdV-55c, with sustained viral transmission. Studies indicate that strains isolated in Singapore in 2005 closely resemble Egyptian strains, implying potential cross-border transmission or importation between these regions [20]. These findings highlight HAdV-55's capacity for international spread, triggering localized outbreaks or prolonged endemicity in diverse geographic areas.

Genomic and capsid protein analyses revealed extremely high genetic conservation among all HAdV-55 lineages (HAdV-55p and HAdV-55a-55d variants), with the whole-genome nucleotide identity exceeding 99.80%. The high conservation of the whole genome indicates that HAdV-55 remains in an evolutionarily stable stage during human-to-human transmission. Notably, the large-scale epidemic of HAdV-55d variant in China is strongly associated with its significant intra-lineage genetic diversity. The core structural/functional proteins, hexon and penton, exhibit extremely high sequence conservation (with identities >99.80% and >99.99%, respectively), and there are no significant variations in amino acids within HVRs. This stable antigenicity and conserved viral invasion mechanism provide reliable targets for vaccine and antiviral drug development. In contrast, the receptor-binding protein fibre exhibits more significant inter-lineage variations, which are mainly concentrated in the functionally critical knob domain. Among them, the fibre protein of HAdV-55d variant has the most diverse amino acid mutations, which may be due to its wide spread in China, the largest number of reported cases, and the existence of multiple transmission chains, leading to the most significant intra-lineage variations, increased replication cycles, enhanced host exposure, and accumulation of random mutations. The continuous variations in fibre knob domain require vigilance. It is recommended to establish a molecular surveillance network based on fibre knob to warn the emergence of potential variants with enhanced transmissibility.

Notably, while its major capsid genes of HAdV-55 are conserved, the E3 region exhibits phylogeographic clustering congruent with the whole genome. The E3 region exhibits lineage-specific mutations: D72N in HAdV-55a/d (predominant in East Asia), A11 V in HAdV-55b (transcontinental distribution), and two mutations (E32H + E113 K) in HAdV-55c (mainly in Egypt with global spread). These mutations show strong geographic clustering, suggesting adaptive evolution of the E3 region in response to regional immune pressures, possibly providing a local transmission advantage. Specifically, the last open reading frame in the E3 region, 15.3 K, shares similar evolutionary characteristics with the whole genomes, suggesting that E3 region may be under specific selective pressures related to interactions with the host immune system. Seroprevalence data support this hypothesis: in East Asia, where HAdV-55a/d variant (D72N) predominates, high seropositivity rates in Chinese adults (22.4–34.0%) and Korean military personnel (56.0%) indicate widespread transmission and substantial immune pressure driving selection of this lineage [48–50]. Notably, individuals aged 1–18 years typically show no detectable antibodies [50], which aligns with our previous epidemiological analysis that HAdV-55 infections predominantly occur in adults. Regional variations in population immunity likely contribute to the geographic clustering of lineage-specific E3 mutations. Supporting this, studies have shown significant differences in immune cell responses to pathogens across human populations. For instance, monocytes and macrophages from individuals of European ancestry mount a weaker inflammatory response to TLR1 agonists compared to those of African ancestry, a difference partly attributed to population-specific eQTLs in the TLR1 gene [51,52]. The E3 region encodes immunomodulatory proteins, such as the 14.7 K protein – a potent antagonist of TNF-mediated cytolysis that plays a critical role in suppressing host inflammatory responses [53]. Its mechanism involves direct binding to the NF-κB p50 subunit and inhibiting its DNA-binding capacity, thereby blunting the innate immune response [54]. Furthermore, the CR1-γ gene within the E3 region shows a highly variable luminal domain across different HAdV-D types, implying host immune pressure-driven evolution [55]. Collectively, these findings indicate that the geographic clustering of the HAdV-55 E3 region reflects local viral adaptation to the distinct immune landscapes of different human populations – shaped by natural selection, particularly in pathways like TNF-NF-κB signalling. This underscores a co-evolutionary arms race between the virus and its human host. Consequently, our findings suggest that effective viral surveillance must look beyond three major capsid proteins and prioritize monitoring variations in immunomodulatory regions like E3, as these mutations serve as crucial “early warning signals” for predicting viral immune evasion potential and region-specific transmission risks.

Based on Table 1 “Geographical Distribution of Human Adenovirus Type 55 (HAdV-55) Infections,” it can be observed that both HAdV-55a variant (South Korea) and HAdV-55d variant (China) predominantly manifest respiratory symptoms. Gastrointestinal symptoms are significantly more frequent in the Korean strain (9.8% vs. 1.7% in the Chinese strain), indicating a greater propensity for severe respiratory complications (e.g. pneumonia) and gastrointestinal involvement. HAdV-55d variant (China) has reported 11 cases with psychiatric symptoms, including insomnia, irritability, anger, and epileptic seizures [56–59]. Therefore, further animal experiments and clinical studies are required to confirm whether the HAdV-55d Chinese strain has neuroinvasiveness. Strains from other regions also present characteristic disease symptoms.

The evolutionary rate of HAdV-55 revealed a mean nucleotide substitution rate of 1.6872 × 10⁻⁵ substitutions/site/year (95% HPD: 1.133–2.2733 × 10⁻⁵) across the complete genome, approximately consistent with the rate of 1.7 × 10⁻⁵ substitutions/site/year calculated by Lu et al. [56], which is of the same order of magnitude as the evolutionary rates calculated by Gonzalez [60] and Firth [61] (HAdV-E4: 3.24-3.69 × 10⁻⁵, HAdV-C: 3.46 × 10⁻⁵ and HAdV-B: 7.20 × 10⁻⁵). Based on the MCC tree constructed from HAdV-55 complete genomes, the tMRCA of HAdV-55 was estimated to be in 1929 (95% HPD: 1717–1967). This differs from the tMRCA of 1963 (95% HPD: 1915–2062) reported by Mao et al., likely due to their use of hexon sequences from only 40 strains isolated between 1969 and 2016 [62]. Since tMRCA calculations for a single capsid protein cannot represent the origin time of the entire genome, the tMRCA inferred from our MCC tree is more representative of the actual evolutionary timeline. It indicates that HAdV-55 underwent four major evolutionary divergences. Following the fourth divergence in 1995 (95% HPD: 1987–2004), HAdV-55d variant emerged as the predominant circulating strain, exclusively distributed in China. This suggests that this variant has achieved widespread dissemination within the country.

This study presents several limitations: First, the clinical presentation of HAdV-55 infection is often non-specific, primarily manifesting with influenza-like symptoms. Moreover, as a recombinant strain of HAdV-11 and HAdV-14, diagnostic assays targeting single structural protein regions may lead to type misidentification, complicating clinical typing in endemic areas. Second, incomplete data disclosure by military organizations regarding infectious disease surveillance, coupled with inconsistent reporting of demographic variables (e.g. age and sex) in the original studies, resulted in substantial missing data (Table 1). Third, military outbreaks often involve large clusters, while civilian cases are mostly sporadic and limited in numbers. This reporting imbalance may overrepresent military-associated findings in our analysis. Finally, while we conducted systematic searches through PubMed, Web of Science, and CNKI databases, the potential omission of literature from other sources may lead to underreporting of infections and fatalities, possibly inflating the apparent proportion of cases in Asia. Notwithstanding these limitations, we have comprehensively characterized the epidemiological profile of HAdV-55 and compared regional and population-specific transmission patterns.

Notably, in recent years, the detection rate of HAdV-55 and the number of related clinical reports have demonstrated a marked upward trend, paralleling the strengthening of global public health surveillance systems and the widespread application of molecular diagnostic technologies such as PCR and metagenomic sequencing. This increase may not fully reflect a true rise in the actual incidence of the virus, but rather underscores substantial improvements in our diagnostic sensitivity and understanding of this pathogen. Therefore, continuing systematic pathogen surveillance and molecular epidemiological studies is essential to accurately assess the true disease burden and evolutionary dynamics of HAdV-55. We aim to draw attention from global public health authorities to enhance HAdV-55 surveillance and control measures, particularly emphasizing its emerging prevalence among adults, the potential for severe clinical outcomes, and the necessity for the development of antiviral therapeutics and vaccines.

Funding Statement

This work was supported by grants from the Prevention and Control of Emerging and Major Infectious Diseases-National Science and Technology Major Project (2025ZD01903704), the National Natural Science Foundation of China (32570158), the R&D Program of Guangzhou Laboratory (No. SRPG22-006), and the Open Research Project of the Key Laboratory of Viral Pathogenesis & Infection Prevention and Control (Jinan University), Ministry of Education (2025VPPC-R07, 2025VPPC-S04, 2025VPPC-S05, 2025VPPC-S07, and 2025VPPC-S10). This work was also supported by grants from the National Institutes of Health EY021558 and EY013124 (JC).

Disclosure statement

No potential conflict of interest was reported by the author(s).

Data sharing

The data used for this meta-analysis are available in the included studies and the appendix. Requests for the meta-analysis data can be made by contacting the corresponding author. All the sequence data are available from GenBank.

Supplemental Material

Supplemental data for this article can be accessed online at https://doi.org/10.1080/22221751.2026.2648885.