4. Discussion
The overall HAdV detection rate among AGE cases was 5.2% in four cities of Western India and ranged between 4.3% (Mumbai) and 5.7% (Ahmadabad). Previously, 4.9 - 11.8% detection rate has been reported from Western, Northern, and Eastern parts of India (Verma et al., 2009; Borkakoty et al., 2016;  Akhil et al., 2016, Gupta et al., 2017; Banerjee et al., 2017; Gopalkrishna et.al, 2021). Globally, the prevalence of HAdV among AGE patients was reported to be between 3.3 - 16.2% (Tran et al., 2010; Sanaei Dashti et al., 2016; Mukhtar et al., 2016; Colak et al., 2017; Afrad et al., 2018). Exceptionally, high HAdV detection rates (26.6- 42.5%) were reported in fecal specimens of diarrheal children <14 years of age with known or unknown HIV status from rural and urban settings in Kenya and Africa (Magwalivha et.al 2010).
Similar to earlier reports, majority of the (85.2%) HAdV infections were detected in children below two years of age (Verma et.al, 2009; Gopalkrishna et.al, 2021) and significantly high in children aged between 7 and 18 months (Table 1). The patients with HAdV infection were seen throughout the year with a maximum number of cases during warmer months in line with earlier reports (Borkakoty  et.al 2016; Afrad et.al 2018; Gopalkrishna et.al, 2021). The severity of the clinical disease among patients with sole HAdV infection was not significantly different as compared to those with sole RVA infection or mixed infection.
The HAdV-F, the known causative agent of gastroenteritis, is reported to be responsible for 25 to 52% of gastroenteritis cases (Filho et.al 2007, Verma et.al 2009, Qiu et.al 2018; Kumthip et.al 2019; Gopalkrishna et.al, 2021). The study strains showed the dominance of the HAdV-F (52.5%) followed by an occurrence of non-enteric adenoviruses of the HAdV-A (17.4%), HAdV-C (11.4%), HAdV-B (8.2%) and HAdV-D (3.2%). The HAdV-F strains were detected in all four cities with their predominance in Ahmadabad (78.5%), Mumbai (61.5%), and Surat (57.1%) cities consecutively for 3-4 years along with second most dominance of HAdV-A strains. The role of HAdV-18 and HAdV-31 strains in sporadic and HAdV-12 in sporadic and outbreak cases of AGE has been reported (Verma et.al 2009; Portes et.al 2016; Afrad et al., 2018; Qiu et.al 2018; Kumthip et.al 2019 ; Gopalkrishna et.al, 2021). Identification of the common epitope of the HAdV-A and F strains on protein VI was suggested as a sign of potential determinants of the tropism and their close evolutionary relationship. However, it is not shared by the HAdV B, C, D, or E strains (Brown et.al 1996). It is noteworthy that during the phylogenetic analysis of the study strains using the penton region, clustering of HAdV-A and F strains together with high bootstrap support was observed (Figure 2).
Multiple studies reported considerable change in the relative incidence of HAdV-40 and HAdV-41 infections over the period in specimens collected from the same geographical region (de Jong et.al 1993; Dey et.al 2011; Banerjee et.al 2017). Among HAdV-F (n=32), 56.2% of strains belonged to the HAdV-41 genotype, and of which only 16.6% were of Pune city. In AGE patients admitted during 2005 - 2007 and 2017 - 2019 in Pune city, 80% and 84.6% of the HAdV-F strains were reported to be HAdV-41 (Verma et.al 2009, Gopalkrishna et.al, 2021). The dominance of the HAdV-B (n=5) and C (n=5) (10/16, 62.5%) strains in Pune city followed by HAdV-F (29.4%) was observed with the absence of HAdV-A strains. The absence of HAdV-A strains was in line with earlier studies on the samples collected between 2005 and 2007 (Verma et.al, 2009) and 2017-2019 (Gopalkrishna et.al, 2021) reporting a very rare occurrence of HAdVA-31 in Pune city. The immune selection pressure in favor of one genotype over the other is known for the shift in dominant genotypes over the period and thus, emphasizes the need for long-term monitoring.
The clinical presentation of the patients infected with only HAdV-F and non-HAdV-F strains were indistinguishable from each other with exception of the less severe patients in the HAdV-B group. No difference in viral load of the AGE patients and control groups or between HAdV41 positive patients and healthy controls was reported suspecting the diarrheal etiological role of HAdV-41 (Qiu et.al 2018). However, a high viral load was observed in HAdV-F-positive study specimens as compared to non-HAdV-F. The absence of correlation of the Ct values with the severity of the disease observed in the study is in line with an earlier report (Bergallo et.al, 2019).
The fecal specimens of HAdVA-12 positive AGE patients < 25 months of age were demonstrated with high viral loads (mean 1.9 X 107 DNA copies/g stool) using qRT-PCR assay (Portel et.al 2016). Among HAdVA-12 positive fecal specimens (n=4) of the study, three showed viral load between 2.4 X 103 and 1.4 X 105. The viral load in the HAdV-18 (n=2) positive specimen with dual infection of Norovirus was 8.4 X 102, while in the remaining sample it was 7.8 X 107. Among five HAdV-31 positive specimens, three showed viral load between 1.2 X 103 and 7.1 X 106. The remaining two specimens with dual infection with RVA and Norovirus showed a viral load of 7.2 X 101 and 1.2 X 106 respectively. Overall, eight out of 11 subgroup A strains showed viral load between 103 to 107 which suggests the need for viral load studies for further confirmation.
All of the HAdV-B, majority of the HAdV-C, and single HAdV-D subgroup strains identified in the study belong to Pune city. To date, several studies reported the presence of non-enteric HAdV strains among AGE patients along with a recent study documenting HAdV-C as the leading agent followed by HAdV-F and B (Hierholzer 1992; Li et.al 2005; Filho et.al 2007; Banyai, et al. 2009; Lee et.al 2012; Knipe et.al 2013; Sriwanna et.al 2013; Afrad et al., 2018; Kumthip et.al 2019). Failure to isolate the majority of the HAdV-C strains from fecal specimens of the patient and detection of the HAdV-F along with other serotypes in healthy individuals was speculated to be due to viral persistence (Fox et al., 1977; Garnett et al., 2002; Roy et.al 2011; Qiu et.al 2018). Therefore, HAdV strains responsible for non-AGE diseases, past infections, or as a part of normal human virome in the gut due to persistent infection needs to be differentiated from the causative HAdV strains of AGE. To rule out the possibility of non-AGE diseases clinical history of the HAdV-positive patients of the study was analyzed. Two patients with upper respiratory tract infection were observed, of which one was positive for HAdV-41 (Ct value- 28.5), and the other for HAdV-18 (Ct. value – 19.3) genotype.
Further, it should be noted that all HAdV-B positive specimens (viral load 327 - 6.2X107copies/run) were collected in winter (October-December months), and the majority of HAdV-C strains (viral load 8.4X102 – 4.5 X106) in summer season (March-Aug months). Taking into consideration the summer season pattern of the HAdV AGE cases, further studies in context to seasonality, subgroup/genotype, and viral load are highly essential to ascertain the etiology. High detection rates and viral load were reported in the summer season (2.19×103 to 6.72×105 gc/g of feces) among asymptomatic adult humans (Vetter et.al, 2015). Hence, further studies, on fecal viral load between symptomatic and asymptomatic children might be useful to identify the role of enteric and non-enteric HAdV strains identified in diarrheal patients.
The difference in the genome-wide composition or the susceptibility among diarrhea and healthy populations suspected to be the reason behind the selective pathogenicity of the HAdV41 strains (Qiu et.al 2018). Further, full genome studies on HAdV strains will be necessary to address this issue. Recently, a serotonin-dependent cross talk between HAdV-41, and human Enterochromaffin cells and Enteric Glia Cells suspected to be the mechanism behind enteric adenovirus diarrhea (Westerberg et.al 2018). Similar studies on non-enteric HAdV strains will help to know their exact role in diarrheal patients.
In Maharashtra and Gujrat states of Western India, universal immunization program for RVA was introduced in August 2019. Recently, a significant proportion of HAdV infections were documented in rotavirus-vaccinated children (Msanga et al., 2020). On this background, the present study on faecal specimens collected between 2013 and 2016 (pre vaccination era) from AGE patients hospitalized in four cities will be of major importance for future comparative analysis to know the exact impact of RVA vaccination in children in Western India. Overall, the existence and co-circulation of multiple HAdV genotypes in AGE patients in the region further demands the necessity of surveillance studies in different parts of India to understand the etiological role and disease burden of adenoviruses in sporadic and outbreak cases of gastroenteritis.
Competing interests: The authors declared that there is no conflict of interest.