Prospective surveillance of human adenovirus in acute respiratory infections reveals epidemiological features and the disappearance of species B during the COVID-19 pandemic in Beijing, China

Dear editor,

Human adenovirus (HAdV) is an important pathogen in acute respiratory tract infections(ARTIs).¹ In this Journal, both Poole² and Lumley³ reported significant variation in HAdV prevalence during the COVID-19 pandemic. However, the epidemiology of HAdV remains incompletely understood. We conducted a multiple-center prospective surveillance, including ARTI patients of all age categories admitted to 35 sentinel hospitals situated across all 16 districts of Beijing, with the aim of understanding contemporary epidemiological characteristics of HAdV infections in Beijing, China.

A total of 47,338 ARTIs cases were enrolled from 2015 to 2021, including 13,927 children (<18 years old), 32,226 adults (≥18 years old), 1,185 without age information, the median age was 38 years (IQR, 9–68). HAdV was detected in 581 cases, including 337 children, 240 adults, 4 individuals without age information, with a median age of 9 years (IQR, 4–30).

The overall HAdV detection rate was 1.23% (581/47,338). The detection rates varied yearly as follows: 1.40% (93/6,643) in 2015, 1.48% (111/7,522) in 2016, 1.22% (93/7,650) in 2017, 1.49% (117/7,839) in 2018, 1.73% (135/7,806) in 2019, 0.56% (22/3,961) in 2020 and 0.17% (10/5,917) in 2021. The entire HAdV epidemic period could be divided into three stages: epidemic period as period 1 (January 1, 2015–December 31, 2017), high epidemic period as period 2 (January 1, 2018–January 31, 2020), and COVID-19 pandemic period as period 3 (February 1, 2020–December 31, 2021) (Fig. 1 ). The HAdV-positive rates of period 1, period 2, and period 3 were 1.36% (297/21,815), 1.65% (269/16,333), and 0.16% (15/9,190), respectively, with significantly increasing in period 2 (p < 0.05) and dropping in period 3 (p < 0.001) (Fig. 1a). It was consistent with the reports that public health measures against COVID-19 had a great impact on the prevention of HAdV and other respiratory pathogens.^{2 , 4 , 5}

Fig. 1.

The prevalence and distribution of HAdV in population, Beijing, 2015–2021. a, the number of HAdV-positive cases and detection rate among patients with ARTIs by month and period, Beijing, 2015–2021. Note: The first COVID-19 case was identified in Beijing on January 19, 2020, and a Level I Public Health Emergency Response (PHER) was implemented on January 24, 2020. Therefore, February 1, 2020, was used as the cutoff point to differentiate the COVID-19 epidemic period. b, the distribution of HAdV infection by age in Beijing, 2015–2021.

HAdV-positive cases were detected in all age groups. The highest detection rate was observed in the age group of 1–4 years old (2.79%, 177/6,354), followed by the age group of 15–19 years old (2.70%, 36/1,331), 5–9 years old (2.46%, 97/3,941), 10–14 years old (2.17%, 31/1,427), and 20–24 years old (2.02%, 36/1,779). The detection rates decreased dramatically to a low level among the population aged 30 years or older (Fig. 1b). During the COVID-19 pandemic, a total of 15 HAdV-positive cases were detected, including 10 cases aged 1–4 years old. The remaining age groups were almost free of infection. This suggests that HAdV infection in children under 5 years old, except for RSV, is a crucial public health problem that should be taken seriously.

Eleven HAdV types were successfully identified by sequencing the hexon and fiber gene⁶ including HAdV-3 (49.74%, 289/581), HAdV-7 (14.29%, 83/581), HAdV-2 (8.78%, 51/581), HAdV-1 (6.71%, 39/581), HAdV-4 (6.54%, 38/581), HAdV-55 (5.51%, 32/581), HAdV-5 (2.58%, 15/581), HAdV-57 (1.03%, 6/581), HAdV-14 (0.86%, 5/581), HAdV-21 (0.69%, 4/581), and HAdV-41 (0.17%, 1/581).

The prevalence of HAdV types varied during three periods. In period 1 and period 2, HAdV-B (3, 7, 14, 21, and 55), HAdV-C (1, 2, 5, and 57), and HAdV-E (4) were all detected. Among them, HAdV-B was the dominant species. In period 3, HAdV-B was undetected, while HAdV-C and HAdV-E remained at a low prevalence (Fig. 2a ). This is the first to describe the disappearance of HAdV-B during the COVID-19 pandemic, which had important public health implications. Given that HAdV-B was the most common type of ARTIs in Asia, which had caused outbreaks and severe pneumonia,⁷ it was necessary to maintain continuous surveillance in order to keep on alert for the reemergence of HAdV-B in the future.

Fig. 2.

The prevalence of different HAdV types by month and period, Beijing, 2015–2021. a, The prevalence of HAdV type. b, The prevalence of HAdV-B. c, The prevalence of HAdV-C. d, The prevalence of HAdV-E.

The prevalence of HAdV-B varied significantly in period 1 and period 2 (0.99% [216/21,815] vs. 1.21% [197/16,333]; p < 0.05) (Fig. 2b). HAdV-3 and HAdV-7 were the dominant type. Compared to period 1, HAdV-3 still remained the most dominant type, with no significant variance in period 2 (0.73% [160/21,815] vs. 0.79% [129/16,333]; p = 0.551). However, the detection rate of HAdV-7 increased significantly in period 2 (0.14% [31/21,815] vs. 0.32% [52/16,333]; p < 0.001). The detection rates of other types of HAdV-B in period 2 remained at a similar level.

The prevalence of HAdV-C did not vary significantly among the three periods (0.26% [56/21,815] vs. 0.25% [41/16,333] vs. 0.15% [14/9,190]; p = 0.192) (Fig. 2c). HAdV-C could cause an asymptomatic persistent infection after initial infection.⁸ In period 3, the continued prevalence of HAdV-C may be associated with the persistent infection in children. Species E (HAdV-4) increased significantly in period 2(0.07% [15/21,815] vs. 1.21% [197/16,333]; p < 0.05) (Fig. 2d). Only one sample of HAdV-4 was detected in period 3. These observations suggested that the rise of HAdV-7 and HAdV-4 led to the high prevalence of HAdV in period 2. HAdV-7 was associated with more severe disease than other HAdV types, and the emergence of HAdV-7 often led to HAdV outbreaks.⁹ HAdV-4 has been known to circulate sporadically, while increasing HAdV-4 infections have been found in many countries.⁶ The prevalence changes of these HAdV types emphasize the need to closely monitor type in HAdV for early warning and intervention.

The most common clinical manifestation of the HAdV-positive patients was moderate or high fever (≥38 °C, 96.68%), followed by cough (72.93%), sore throat (48.25%), sputum production (44.75%), and rhinorrhoea (24.86%). These clinical characteristics were helpful to differential diagnosis with infections of other respiratory pathogens. We also found HAdV-7 and HAdV-4 had the highest prevalence in severe community-acquired pneumonia (SCAP) in 18–24-year-old patients. Notably, SCAP cases were most commonly observed at children (<1 years old) and the elderly (≥60 years old), which was similar to respiratory syncytical virus.¹⁰ SCAP cases more often experienced respiratory failure, liver damage and kidney failure (p < 0.05).

In summary, children were the main population for HAdV infections in Beijing. The elevated prevalence of HAdV in 2018 and 2019 was probably linked to the rise of HAdV-7 and HAdV-4. During the COVID-19 epidemic, the prevalence of HAdV significantly decreased and the long-term dominant HAdV-B disappeared. Continuous, population-wide molecular epidemiological surveillance is essential for the prevention and control of HAdV.

Declaration of Competing Interest

The authors report no conflict of interest.