General results of respiratory virus screening using RVP FAST assay
During Nov 2016 to Feb 2019, 2090 respiratory specimens were collected for respiratory virus screening using RVP FAST assay. For there were more specimens belonging to one patient in one time hospitalization, we only kept the first or the EVs/HRVs positive one. Then 1835 specimens were left for statistical analysis. Among these specimens, there were 402 (21.9%, 402/1835) for RSV, including 300 for RSV A and 114 for RSV B; 363 (19.8%, 363/1835) positive for EVs/HRVs; 196 (10.7%, 196/1835) for PIVs, including 27 for PIV1, 8 for PIV2, 136 for PIV3, and 30 for PIV4; 173 (9.4%, 173/1835) for HBoV; 154 (8.4%, 154/1835) for ADV; 140 (7.6%, 140/1835) for Flu, including 121 for Flu A (H1 91, H3 24), and 23 for Flu B; 80 (4.4%, 80/1835) for HMPV; 60 for Coronavirus (3.3%, 60/1835) , including 29 for NL63, 6 for HKU1, 5 for 229E, and 21 for OC43.
Identification and genotyping of HRVs or EVs
For HRVs and EVs cannot be distinguished from each other in RVP FAST assay, RT-PCR and sequence analysis were used. Among 363 specimens positive for EVs/HRVs, 314 (86.5%, 314/363) were determined as HRVs positive, including 177 (56.4%, 177/314) positive for HRV A, 29 (9.2%, 29/314) for HRV B, and 108 (34.4%, 108/314) for HRV C, 19 (5.2%, 19/363) were determined as EVs positive, and 30 were undetermined for low amplification products (Fig 1).
The epidemiology of HRVs
The monthly distribution of HRVs positive specimens during Nov 2016 to Feb 2019 was shown in Fig 2. HRV infections were shown all year round. More HRVs were detected through all Autumn months followed by that in Winter months. In August and September, HRV A and B are the dominant species, then HRV C became the dominant species in October and December.
By MEGA version 6.0 software, the phylogenetic tree shown in Fig 3 was constructed to identify the sero- or geno-types of HRV positive specimens. Among 177 specimens positive for HRV A, 53 types were confirmed, including 15 positive for A49, 12 for A24, 10 for A12, and 9 for A101; Among 29 specimens positive for HRV B, 11 types were confirmed, including 8 for B79; Among 108 specimens positive for HRV C, 32 types were confirmed, including 14 for C2, 8 for C5, and 6 for C15.
In Fig 4, no obvious circulation pattern was shown for each sero- or geno- type, therefore, more data should be accumulated to evaluate the epidemiology characters of HRV sero- or geno- types. For the 96 HRV types detected in the study, only 29 types were detected in consecutive years. The most prevalent strain of HRV A, A49 was found only between March 2017 and October 2017, which was followed by A24 more detected in the Autumn of 2018. The most prevalent strain of HRV B, B79 was detected in the winter of 2018. The most prevalent type of HRV C, C2 was detected more often in the late of 2018.
Clinical characters of different HRV species in severe or mild acute lower respiratory tract infection
As shown in Table 1, there were 925 patients discharged from Intensive Care Unit (ICU) with severe ALRI (ICU group) and 508 patients discharged from the Department of Respiratory with mild ALRI (Respiratory group) compared to those from ICU. The HRVs positive rates are similar and HRV A and HRV C are the major pathogens in two groups. However, the positive rates of HRVs species were significant different between the two groups (p=0.011), which may be explained by more HRV A in ICU group (63.0%: 45.2%, p=0.006), and more HRV B (15.1%:6.5%, p=0.028) in Respiratory group. However, no significant difference was shown on HRV C (39.8%: 30.5%, p=0.136) between the two groups. Compared to patients positive for HRVs from Respiratory group, patients who were positive for HRVs from ICU group were significantly younger (0.39 years: 1.80 years, p=0.000).
In the ICU group, about 44.8% (n=69) were viral co-infection, with HBoV (n=27) as the major virus followed by PIV (n=19) and RSV (n=18). In the Respiratory group, about 45.2% (n=42) were viral co-infection, with RSV as the most common viral pathogen (n=12) followed by PIV (n=11) and ADV (n=9). No significant difference was shown in the whole viral co-infection (p=0.957) and viral co-infection in different HRV species (p=0.239) between ICU group and the Respiratory group.
Then a logistic model was constructed to test and verify the variables associated with severe ALRI (Table 2). By using HRV C infection as the reference group among HRV species when gender, age and co-infections were considered as factors, there is a significant difference as the odds of HRV A related to severe ALRI was OR=1.983 (95%CI=1.091-3.605) (p=0.025), while no significant difference was shown (OR=0.707, 95%CI=0.260-1.920) (p=0.496) when the odds of HRV B related to severe ALRI was evaluated. In the logistic model, gender was excluded from variables associated with severe ALRI (OR=0.710, 95% CI=0.389-1.295, p=0.264), and age (OR=0.703, 95% CI=0.607-0.813, p=0.000) was confirmed as a protective factor (OR<1, the upper limit of 95% CI<1) for severe ALRI.
When the correlation of HRV sero- or geno- types with severe or mild ALRI was evaluated (Fig 5), 39 HRV types were shown in both groups. In ICU groups, HRV A24 followed by A49, HRV B79 followed by B6, and HRV C2 were the most popular serotypes of different species, while HRV A49 followed by A18, HRV B14, B79 and B92, and HRV C5 followed by C2 were the most popular serotypes of different species in Respiratory group. However, as a result of restrictions on the numbers of specimens belonging to each sero- or geno-type, no skewing towards one type in either group was observed.