Herein, we describe the prevalence and clinical manifestations of HAdVs in patients hospitalized due to acute diarrhea. The strengths of our study were that an unbiased systematic sampling were followed irrespective of age, sex, nutrition status, disease severity or socioeconomic context for a long period of time over four consecutive years. The overall detection rate was found to be 10.7% which is a three-fold increase compared to the earlier studies (1.9-3.9%) conducted in Bangladesh 21.
This concordance can be explained by the enrollment of only hospitalized patients in the current study. However, our observation is comparable to many studies conducted worldwide, such as India, Gabon, Japan, Ireland, and China 32-36.The age distribution of patients with HAdV infections confirmed that HAdV infections occurred preferentially in children under 5 years of age (54.8%), demonstrating HAdV as an important pediatric pathogen and HAdV specific maternal antibodies are not protecting neonates against infection.
Similar observation has been demonstrated in other reports conducted on enteric adenoviruses worldwide 9, 11, 21, 36. During the study period, HAdVs were detected throughout the year and we did not observe any seasonality pattern but several high peaks in summer and rainy seasons and a low incidence in winter. Our results, together with the reports from Lanzhou, Brazil, and Australia provide evidence that, HAdV infection has no consistent seasonal pattern in temperate regions 37-40.The most common sign and symptoms infected with diarrheal pathogens are dehydration and vomiting. We observed more patients with some or severe dehydration (61.5%) and vomiting (78.8%) than with abdominal pain (40.
4%) and fever (40.4%). Number of watery stool per day was extended with most of the individuals having at least 6 times (80.7%) stool per day.In this study, we used broad-range primers for PCR amplification that is able to detect all known HAdV serotypes. A variety of HAdV sub-genera (five) and serotypes (seventeen) were identified to be circulating during the study period. Sub-genera F (HAdV-40 and -41) was the most prevalent, 54.
8% (51/93) of the HAdV positive samples. HAdV-41 was the most predominant (30/93, 32.3%), followed by HAdV-40 (21/93, 22.6%). Earlier molecular study conducted in Bangladesh 21 documented that HAdV-40 was the only serotype found in Bangladesh, and they did not observe presence of any HAdV-41 during their one-year study period (2004-2005).
However, that study was limited to only one-year time span. Our observation is in accordance with the other study published (1993) from Bangladesh by Khan et al. 20 and with many reports where predominance of HAdV-41 over HAdV-40 has been documented 11, 37, 38, 41. It is worthy to note that, we detected nine HAdV-70, which is similar (at the nucleotide level) to the recombinant HAdV-70 P70H70F29 isolated from diarrheal feces of a hematopoietic stem cell transplantation recipient (2015) 4. Notably, six of the nine HAdV-70 detected cases, the virus was detected as the only pathogen which may suggests that HAdV-70 be another serotype leads to diarrhea. More than half of the HAdV positive cases were also co-infected with other viral or bacterial pathogens. V.
cholerae, rotavirus, and norovirus were the major co-infecting pathogens identified (Table 2). Similar evidence of mixed infections was evident in other HAdV epidemiological studies conducted in other parts of the globe 42-44. Many epidemiology studies demonstrated that many HAdVs types can be found in stool due to prolonged shedding in the feces after a previous infection28, 38, 45, but only enteric HAdVs are the causative agents of gastrointestinal disease. However, although enteric AdVs (genotypes 40 and 41) were the most common cause of gastroenteritis in our study, other non-enteric genotypes were also associated with enteric symptoms. Thus, we cannot rule out the role of species A/B/C/D HAdVs in diarrhea etiology, since 20/41 (49%) of non-enteric HAdVs (HAdV-1, -2, -6, -7, -10, -12, -16, -21, -37, -61, -64, and -70) were detected as sole viral pathogen in this study. Moreover, some sporadic reports have described an association of non-enteric HAdVs with acute gastroenteritis, particularly HAdV-12, -18, and -31 (species A), HAdV-3 and -7 (species B), and HAdV-1, -2, -5 (species C) 38, and HAdV-52 (species G) 17. Therefore, detection methods should not be limited to target only established enteric genotypes 40 and 40 to determine true burden of gastroenteritis caused by HAdVs.
There were a number of limitations. First, the study was conducted in a hospital specialized on diarrhea disease, where all patients presented diarrhea and investigation on other diseases such as respiratory infections was not appreciated. Therefore, it was not possible to assess the importance of identification of so many non-enteric HAdVs in diarrheal patients. Second, all subject in our study where hospitalized patients which may not adequately represent the actual burden of HAdV infections associated with diarrhea in the larger community. Third, the clinical relevance of pathogens with gastroenteritis was difficult especially in co-infected patients. Fourth, only a 301 bp region of hexon gene was sequenced for genetic analysis.
Full-length genome could help to understand the extent of genetic variation and any potential recombination in these strains.In conclusion, this 4-year study documented the prevalence, age distribution, seasonality and molecular epidemiology of HAdV infections associated with diarrhea in a low-income country setting. HAdVs showed a huge genetic diversity including enteric- and non-enteric HAdVs and co-infections with other diarrheal pathogens. Taken together, our results provide a foundation for further clarification of the role of non-enteric HAdVs and define the clinical and public health significance of HAdV infections.