Over the past few decades, acute hepatitis and liver failure in children have become major public health issues worldwide (Daniels et al., 2009). Infectious, toxic, autoimmune, and metabolic mechanisms are the most common causes (Figure 1) (Devictor et al., 2011).

Most cases can be attributed to viral infections, with hepatotropic viruses being the primary culprit (Daniels et al., 2009; Zeng et al., 2021). Globally, acute hepatitis A virus (HAV) infection has long been considered the most prevalent cause of acute hepatitis in children (Rahman et al., 2022). However, its prevalence varies by geographical area and vaccination rates (Daniels et al., 2009; Zeng et al., 2021). Studies in the United Kingdom and India indicated that HAV infection accounts for 39% and 48.6% of acute hepatitis in hospitalized children, respectively (Braccio et al., 2017; Sood et al., 2019). Since the beginning of the century, routine childhood immunization has been promoted continuously in numerous nations, including the United States and Europe. In this regard, from 1995 to 2007, the incidence of acute hepatitis A in the United States population decreased by 92%, especially for children under 15 years old (Daniels et al., 2009). In addition to HAV, sporadic and epidemic infections of the Hepatitis E virus (HEV) represent a leading cause of acute viral hepatitis in children, primarily in developing nations (Verghese and Robinson, 2014; Fischler et al., 2016). HEV seroprevalence in children increases with age (Hodges et al., 1998) and is related to residence and socioeconomic status (Arankalle et al., 2001). Two early studies with small sample sizes from Egypt showed that acute sporadic hepatitis E accounted for 22.2% (under 11 years old) (el-Zimaity et al., 1993) and 12% (under 13 years old) (Hyams et al., 1992) of children admitted to hospital for acute hepatitis.

Besides hepatotropic pathogens, other viruses such as Epstein-Barr virus (EBV), cytomegalovirus (CMV), herpes simplex virus, adenovirus, and coxsackievirus can also cause acute liver failure in children, especially in those with immunodeficiency disorders (Gallegos-Orozco and Rakela-Brödner, 2010; Tsunoda et al., 2017). Real-time PCR and serological testing can assist in the detection of non-hepatitis virus-related acute hepatitis in children. A recent study investigated the incidence of CMV and EBV in children hospitalized with acute liver failure. Of the children excluded from hepatitis virus A, B, and C infection, 18% and 9% were positive for both DNA and IgM assay of CMV and EBV, respectively (Tsunoda et al., 2017). Another prospective study proposed that human herpesvirus-6 could cause acute liver failure in children (Chevret et al., 2008).

Other non-viral infections can potentially cause acute hepatitis in children, including amebiasis (Gupta et al., 2022), Brucella (Hassouneh et al., 2019), histoplasmosis, leptospirosis (Aygün et al., 2016), etc. A new study suggested that hepatitis was the most common laboratory abnormality in children with brucellosis (nearly three-quarters) (Hassouneh et al., 2019). In recent years, pediatric autoimmune hepatitis (AIH) has been increasingly reported (Mack et al., 2020). In the United States, the prevalence of AIH is 3 per 100,000 children, and the estimated annual incidence per 100,000 children is 0.4, comparable to Canada (0.23) (Deneau et al., 2013; Jiménez-Rivera et al., 2015).

In addition, drugs represent another potential cause of acute hepatitis. In developed countries, acetaminophen is the leading cause of drug-induced liver injury in children, followed by medications such as anti-tuberculosis and anti-epileptic agents (Squires et al., 2006). As for newborns and infants, acute liver failure is most commonly caused by inherited metabolic diseases (Durand et al., 2001), including Wilson disease, alpha-1 antitrypsin deficiency, tyrosinemia, glycogen storage disease, etc.

Although in-depth and comprehensive research has been conducted on the etiology of pediatric acute hepatitis and acute liver failure along with the continuous promotion of unprecedented metabolic and viral screening, a definite cause cannot be found in a substantial portion of children (Aydoğdu et al., 2003; Squires et al., 2006; Black, 2009; Narkewicz et al., 2009). When a specific diagnosis is not established in PALF, these children are categorized as indeterminate (iPALF) (Alonso et al., 2017). The Pediatric Acute Liver Failure study group, composed of centers from the United States, Canada, and the United Kingdom, released the initial data of 348 cases in 2006 (Squires et al., 2006). The outcomes of this prospective multicenter research revealed that for nearly half of children with pediatric acute liver failure, a definitive diagnosis could not be determined, consistent with the subsequent results published by this study group in 2008 and the conclusions of studies from Turkey and Germany (Aydoğdu et al., 2003; Squires et al., 2006; Narkewicz et al., 2009; Kathemann et al., 2015). Besides, children with ambiguous diagnoses were younger and more likely to receive liver transplantation after admission than children with a definitive diagnosis (Liu et al., 2006; Narkewicz et al., 2009). In subsequent studies, Narkewicz et al. evaluated the diagnostic mode of 329 children who were eventually diagnosed with indeterminate acute liver failure between 1999 and 2008. The results showed inadequate screening for metabolic and autoimmune acute liver failure causes. Furthermore, an efficient diagnostic evaluation pattern for pediatric acute liver failure based on age and clinical manifestations should be established to help pediatricians formulate timely and effective treatment strategies (Narkewicz et al., 2009). Mechanisms driving injury of iPALF have been postulated to be secondary to overwhelming inflammatory responses and immune dysregulation (Alonso et al., 2017). Recent efforts to further characterize iPALF have identified a subset of patients with unusual liver histology, including a dense CD103+CD8⁺ T-cell infiltrate, further suggesting liver injury is immune-mediated (Squires et al., 2022).

By far, adenovirus is most frequently identified in test samples. As of 28 July 2022, adenovirus has been found in 65.9% (170/258) of cases investigated in the United Kingdom (UK Health Security Agency, 2022c). Type 41 represented 95.2% (59/62) of these adenovirus-positive cases and was found in several cases reported in the United States (Baker et al., 2022; UK Health Security Agency, 2022c).

Recently, the UK Health Security Agency has documented an increase in adenovirus-positive stool samples from children aged 1–4 years compared to pre-pandemic levels, as well as a rise in the occurrence of hepatitis among children aged 1 to 4 (UK Health Security Agency, 2022b). However, on June 14, the CDC released a preliminary survey highlighting that current data do not indicate that pediatric hepatitis or adenovirus types 40/41 are above pre-pandemic baseline levels of COVID-19 (Kambhampati et al., 2022).

Although adenovirus infection is currently the most widely accepted hypothesis, it cannot fully explain the severity of clinical symptoms as the underlying cause. Biopsies of liver specimens from the United Kingdom, Israel, and Alabama, United States, did not reveal adenovirus inclusions, and the adenovirus load of affected children was generally low (Baker et al., 2022; UK Health Security Agency, 2022b; Marsh et al., 2022). Much controversy surrounds the ability of the “adenovirus hypothesis” to explain the pathogenesis of acute hepatitis of unknown origin in children.

It is well-established that adenovirus infections are prevalent in children. Indeed, children’s tonsils and adenoids are common host organs for adenoviruses, with a reported positivity rate of 76% for adenovirus PCR (Jaggi et al., 2013). About 30% of immunocompetent children develop persistent or resident infections after the acute phase of adenovirus infection. Such persistent infections are common in mucosal lymphocytes and are more common in young children, fading away with age (Garnett et al., 2002; Kosulin, 2019).

Then, during systemic immunocompromise, mucosal-site adenoviruses (as well as various other viruses) have the potential to enter the bloodstream and reactivate, eventually presenting as disseminated infection. Current evidence suggests that acute liver injury and liver failure, regardless of the cause, represent a state of immunodeficiency. However, whether liver damage or adenovirus viremia is the cause remains unclear.

Moreover, almost all previously reported cases of adenoviral hepatitis involved immunodeficiency. In this respect, adenoviral inclusion bodies were detected in all liver biopsies of adenoviral hepatitis. Few cases of acute severe hepatitis or liver failure due to adenovirus infection have been reported in immunocompetent children. However, the serotype human adenovirus (HAdV) 41 has not been linked to hepatitis or liver failure (Rocholl et al., 2004; Ozbay Hoşnut et al., 2008; Lynch and Kajon, 2016; Kang, 2017).

Although adenovirus is infrequently related to fulminant liver failure in immunocompetent children, other factors may enhance susceptibility; hence, adenovirus remains a potential contributing factor in addition to other factors such as abnormal susceptibility or host response.

Interestingly, it has been reported that the outbreak control during the 2019 coronavirus epidemic may have reduced young children’s exposure to common pathogens, leading to the development of immune systems different from their peers before the 2019 coronavirus epidemic. When young children are subsequently infected with common pathogens such as adenovirus, the immune response or disease manifestations differ from their peers. According to the UKHSA, adenovirus, enterovirus, human metapneumovirus, rhinovirus, and norovirus have been significantly overrepresented among children under 10 years old since the end of 2021. Respiratory syncytial viruses have been reported since late summer 2021, possibly due to behavioral alterations and population vulnerability following the pandemic’s low incidence (UK Health Security Agency, 2022b).

Up to now, more than 10% of positive SARS-CoV-2 cases have been detected by PCR, while the positive rate of SARS-CoV-2 serological tests is about 70%. Serological tests are underway to explore previous infections further. (World Health Organization, 2022a; European Centre for Disease Prevention and Control, 2022). Most patients were only tested for SARS-CoV-2 PCR but not for SARS-CoV-2 serology. Notably, SARS-CoV-2 PCR can only determine if the child was infected with SARS-CoV-2 then, but serological testing is indicated to assess whether the child has been infected with SARS-CoV-2 in the past.

SARS-CoV-2 can result in liver function impairment in COVID-19 patients. Recent research by Ellen K. Kendall et al. revealed that children infected with COVID-19 had a two-fold higher risk of liver injury than children infected with other respiratory infections. This finding implies that COVID-19 causes acute and chronic hepatic complications in pediatric patients (Kendall et al., 2022).

Recently, Shiri Cooper et al. analyzed the clinical presentation and laboratory test results of five hospitalized children who recovered from COVID-19 and subsequently exhibited liver injury. It has been suggested that SARS-CoV-2 infection is more likely to cause acute hepatitis of unknown etiology in children. The two patterns of long-term hepatic manifestations of COVID-19 in children include acute liver failure and acute hepatitis with cholestasis (Cooper et al., 2022).

In 2020, Yijin Wang et al. studied 156 COVID-19 patients from two designated centers in China and compared the clinical characteristics of patients with and without increased aminotransferases. The findings indicate that SARS-CoV-2 infection of the liver is a major contributor to the impairment of liver function in COVID-19 patients. However, the median levels of ALT were 50 U/L and 19 U/L, and AST levels were 45.5 U/L and 19 U/L in the abnormal and normal transaminase groups, respectively (Wang et al., 2020). Even the transaminase values in the abnormal transaminase group were much lower than those in the case definition.

However, two independent research groups in Glasgow and London, UK, have recently found no direct association between acute hepatitis of unknown origin in children and SARS-CoV-2 infection. The prevalence of SARS-CoV-2 antibodies did not differ between cases and age-matched controls. Nevertheless, they could not completely rule out an immune-mediated post-COVID-19 phenomenon in susceptible children (Ho et al., 2022; Morfopoulou et al., 2022).

In July, two non-peer-reviewed studies from the United Kingdom provided new insights into the possible etiology of the disease: children may be infected with both viruses and harbor genes that lead to an overactive immune response. In addition to the many studies that have reported the detection of adenovirus 41 in children with acute hepatitis of unknown origin, these two studies detected another virus, adeno-associated virus 2 (AAV2), in most children and carried a specific HLA genotype (DRB1*04:01).

AAV2 is a human parvovirus that depends on coinfection with helper viruses such as adenovirus or herpesvirus for efficient reproduction. Many people are infected with AAV2 by age 10, although the virus can remain latent in cells until a helper virus activates it.

A study conducted in Glasgow, Scotland, reported nine early cases and 58 controls. NGS and PCR detected AAV2 in plasma samples from nine patients and liver samples from four patients but not in matched control children. In addition, they found a higher prevalence of the class II HLA-DRB1*04:01 allele in eight of nine children (89%) compared to this HLA gene among Scottish blood donors (15.6%). HAdV (species C and F) and human herpesvirus 6B (HHV6B) were detected in 6/9 and 3/9 affected cases, including 3/4 and 2/4 liver biopsies, respectively (Ho et al., 2022).

Another study by a team in London investigated 28 cases of acute hepatitis in children of unknown origin, including liver samples from five children who required transplantation and blood samples from the remaining children who did not. 16 of the 17 samples tested for AAV2 were positive during RNA sequencing of liver samples (Morfopoulou et al., 2022).

These two studies propose a new etiological hypothesis that acute hepatitis in children of unknown origin may result from the interplay between AAV2, helper virus (e.g., adenovirus), and genetic susceptibility (a specific HLA genotype).

Although genetic material from AAV2 was discovered in the liver cells of patients, neither viral proteins nor actual copies of the virus were detected, emphasizing that AAV2 may trigger an immunological response that is harmful to the organ rather than directly harming liver cells.

There were a few limitations associated with these two studies. First, the number of study subjects was small. Second, current evidence only suggests a correlation with AAV2 with no proof of a causal relationship. It is highly conceivable that AAV2 is only a biomarker for coinfection with ADV or even HHV6B. Many scientists are still cautious about the conclusions of these two studies, and further mechanistic studies and large-sample research studies are warranted to confirm them.

Gao, SH. hypothesized that children with severe acute hepatitis may have an underlying immunodeficiency, given that most cases were in children aged five or younger. In addition to immune insufficiency in children of these ages and immunodeficiency due to lack of pathogen exposure in children as a result of COVID-19 prevention and control measures, primary immunodeficiency should also be investigated (Gao et al., 2022).

In July, the Glasgow Research Group revealed that most children (8/9) carried a specific HLA genotype (DRB1*04:01) (Ho et al., 2022), associated with some autoimmune diseases, such as autoimmune hepatitis (AIH) and rheumatoid arthritis. It is well known that AIH can be classified into two subtypes; type 1 and type 2. Approximately two-thirds of pediatric AIH cases are AIH-1, which usually appears in adolescence, whereas AIH-2 affects younger children, including infants. Up to 67% of patients in children present with an acute onset. Fulminant manifestation is more frequent in AIH-2. AIH-1 susceptibility is conferred by MHC class II HLA DRB1*03 at all ages, whereas DRB1*04 is susceptible to late-onset disease; AIH-2 is associated with possession of DRB1*07 and DRB1*03 (Mieli-Vergani et al., 2018; Di Giorgio et al., 2020; Terziroli Beretta-Piccoli et al., 2022). The class II HLA DRB1*0401 allele may represent a disease susceptibility gene, but the local population structure may overestimate the significance (Ho et al., 2022). Indeed, it has been established that HLA serotypes vary across ethnic groups and regions. AIH-1 is closely related to DRB1*0301 and DRB1*0401 alleles in Caucasian adults; DRB1*0405 in Japan and Argentina (Terziroli Beretta-Piccoli et al., 2022). Moreover, DRB1*04 does not increase children’s susceptibility to AIH and even exerts a protective effect compared to adult patients (Gregorio et al., 1997).

Therefore, genetic testing, major histocompatibility complex molecular phenotyping, and immune function testing, including testing for humoral, cellular, and innate immune responses, should be undertaken on these children.

In addition, acute hepatitis in children of unknown origin is likely caused by the emergence of a novel adenovirus/a new variant of SARS-CoV-2/a new variant of another known pathogen (cytomegalovirus, EBV, respiratory syncytial virus, enterovirus, parainfluenza virus, etc.)/a new pathogen with altered characteristics, with or without the cofactors mentioned above.

Suppose it is a new variant of adenovirus. It is of interest whether it is related to the recombination and mutation of adenovirus under selection pressure that may occur after mass vaccination with adenovirus vector vaccines.

There is a lack of data to support this hypothesis, especially when whole-genome sequencing results from multiple cases have not been established. However, the phenomenon of a novel adenovirus/a novel variant of SARS-CoV-2/a novel variant of other known pathogens/a new pathogen arising from naturally occurring genome recombination is highly possible.

Toxicological investigations have revealed no major findings, with no indication of paracetamol, fluconazole, or aflatoxin B1-induced toxicity (UK Health Security Agency, 2022b).

Environmental exposure that may influence adenovirus infection has not been elucidated. According to the UKHSA report, approximately 70% of children affected whose families were interviewed reported recent contact with pet dogs. In the extended investigations, however, there was no indication that dogs played a role in hepatic syndrome, and no public health issues were associated with dog exposure and hepatic syndrome (UK Health Security Agency, 2022a).

With limited studies based on liver biopsy samples, there is little evidence to support direct viral infection of HAdV-F41, SARS-CoV-2, or AAV2 in explanted livers, and neither immunohistochemistry nor proteomic analysis has been used to detect viral proteins. Instead, studies have found evidence of high HLA class II allele expression in transplant cases, which increases the genetic risk of unknown hepatitis in these children (Ho et al., 2022; Morfopoulou et al., 2022). Nevertheless, we cannot completely rule out an immune-mediated phenomenon in susceptible children with prior viral infections. Therefore, the first four hypotheses mentioned above remain the most likely hypotheses.