Respiratory viral coinfection in a birth cohort of infants in rural Nepal

Abstract Background Acute respiratory illnesses are a leading cause of global morbidity and mortality in children. Coinfection with multiple respiratory viruses is common. Although the effects of each virus have been studied individually, the impacts of coinfection on disease severity are less understood. Methods A secondary analysis was performed of a maternal influenza vaccine trial conducted between 2011 and 2014 in Nepal. Prospective weekly household‐based active surveillance of infants was conducted from birth to 180 days of age. Mid‐nasal swabs were collected and tested for respiratory syncytial virus (RSV), rhinovirus, influenza, human metapneumovirus (HMPV), coronavirus, parainfluenza (HPIV), and bocavirus by RT‐PCR. Coinfection was defined as the presence of two or more respiratory viruses detected as part of the same illness episode. Results Of 1730 infants with a respiratory illness, 327 (19%) had at least two respiratory viruses detected in their primary illness episode. Of 113 infants with influenza, 23 (20%) had coinfection. Of 214 infants with RSV, 87 (41%) had coinfection. The cohort of infants with coinfection had increased occurrence of fever lasting ≥ 4 days (OR 1.4, 95% CI: 1.1, 2.0), and so did the subset of coinfected infants with influenza (OR 5.8, 95% CI: 1.8, 18.7). Coinfection was not associated with seeking further care (OR 1.1, 95% CI: 0.8, 1.5) or pneumonia (OR 1.2, 95% CI: 0.96, 1.6). Conclusion A high proportion of infants had multiple viruses detected. Coinfection was associated with greater odds of fever lasting for four or more days, but not with increased illness severity by other measures.


| INTRODUC TI ON
Acute lower respiratory tract infections (ALRI), including pneumonia and bronchiolitis, were responsible for an estimated 650 000 deaths of children under five years old in 2016 and continue to be a major cause of infant morbidity and mortality worldwide. 1 Multiple viruses cause respiratory illness in children, including influenza, respiratory syncytial virus (RSV), rhinovirus, enterovirus, and adenovirus. 2 Influenza is among the most common etiologies of ALRI episodes and is responsible for an estimated 39 million cases each year among both children and adults. An estimated 30 000-100 000 children under 5 years old die from influenza each year, with 99% of these occurring in developing countries. 3 RSV is responsible for over 33 million new cases of childhood ALRI each year, and an estimated 55 000 to 190 000 deaths of children under 5 years old can be attributed to ALRI from RSV alone. 4 Multiple viruses are often simultaneously detected, though the impact of coinfection is not well-defined. Further, little research has been done to examine viral coinfection in low resource settings.
Prevention of morbidity and mortality due to respiratory viral disease remains a major goal worldwide. Currently, seasonal influenza vaccines are widely available, and developing an RSV vaccine is among the primary 2020-2025 goals for GAVI (Global Alliance for Vaccines and Immunization) and the WHO. 5,6 Understanding the role of coinfections will help inform the relative burden of disease of each respiratory virus and add context to future vaccination intervention and development efforts.
Polymerase chain reaction (PCR) technology has allowed for detection of multiple simultaneous viruses in children with acute respiratory illness. 7 Before widespread use of molecular diagnostics, concurrent infection by two or more viruses in children under five was considered rare. However, recent studies estimate viral coinfections in 10%-30% of pediatric patients with ALRI. 8 A number of previous studies have investigated the relationship between coinfection with multiple viruses and disease severity. These studies have mostly occurred in hospitalized settings in industrialized countries, and findings are discordant on the effect of coinfection on symptom severity and duration. 2,8 The hypothesis that multiple viral infections can have an additive effect on illness severity is intuitive. A prospective study of hospitalized infants in 2005 found that coinfections of any type were associated with more severe ALRI than single infections in infants under 6 months old. 2 Other studies confirmed that coinfection resulted in worse outcomes, including higher likelihood of fever, 9 longer duration of symptoms, 10 higher rates of admission to the pediatric intensive care unit, 11 and more severe bronchiolitis. 12 However, recent viral modeling has indicated that one virus can occupy all of the patient's available cellular mechanisms, inhibiting progress of the other infections. According to this model, competition between two different infections can attenuate illness severity. 13 Additionally, one infection can trigger an immune response that is protective against the subsequent infections, making coinfections less severe than one virus alone. 14 By contrast, other reports have found that coinfection did not lead to more severe disease and demonstrate the need for further investigation of the impacts of concurrent infections in infants. 10,15 As most studies of coinfection have been conducted in hospitalized settings in industrialized countries, little is known about the impact of coinfection on the severity of respiratory viral infection in rural settings. The presence of multiple viral infections may play a role in the severity of these illnesses. The objective of this study is to characterize the frequency and patterns of coinfection in infants in rural Nepal and compare symptom severity in coinfection and singular infection.

| MATERIAL S AND ME THODS
This is a secondary analysis of data from a prospective, randomized, placebo-controlled study of the effects of influenza immunization among pregnant women in the Sarlahi District of rural Nepal.
Methods and results of this clinical trial have been published and registered with ClinicalTrials.gov (Trial no. NCT01034254). 16 Between 2011 and 2014, 3693 women were enrolled during pregnancy into a study of maternal influenza immunization. Both women and their infants were monitored via weekly household visits for respiratory illness. Infants enrolled in this study were enrolled in a home visit with an infant birth assessment within 72 hours of birth and followed by weekly visits until they reached 180 days of age. 17 Mothers were asked about respiratory symptoms in infants for each day in the past week. Nasal swabs were collected from the infants if they experienced one or more symptoms of respiratory illness in the past week, defined as subjective fever, cough, wheeze, difficult or rapid breathing, or a draining ear. Samples were tested for respiratory viruses by real-time reverse transcription PCR. 18 This secondary study analyzed the subset of infants experiencing their first case of respiratory viral illness. First viral illness was defined as the first symptomatic respiratory illness with a virus de- We evaluated the associations between coinfection compared to monoinfection and three pre-specified outcomes: subjective fever lasting four days or longer, pneumonia, and seeking further medical care. Fever duration was analyzed as a binary outcome to maintain consistency in analysis with other measures of severity. Pneumonia was defined as cough and/or difficulty breathing, combined with wheeze and/or age-specific tachypnea. 17 Seeking further medical care was defined as a visit to a hospital, a doctor, or a non-doctor provider for the primary illness episode in the week preceding the home visit. Non-doctor providers included visits to local healers and medicine shops. These measures for severe disease were chosen based off of the symptoms and behaviors attributable to respiratory illness observed in this population.
Sociodemographic characteristics and symptoms were compared between infants with coinfection and monoinfection. Continuous variables were described using mean (standard deviation), and binary variables were described using counts and percent of coinfection or monoinfection cases. Two-sided t tests and chi-squared tests were used to compare differential distributions of demographic characteristics between infants with coinfection and monoinfection.
A variable was determined to differ significantly if the P-value was <.05.

| RE SULTS
Altogether, 1730 (47%) of the 3646 enrolled infants had at least one episode of viral respiratory illness over the course of this study.
Among those experiencing their first respiratory viral infection, 1403 (81%) were infected with a single virus and 327 (19%) had at least two respiratory viruses detected simultaneously. Of those with coinfection, 32 (9.8%) had three different viruses detected, and two (0.6%) had four different viruses. Subjects had a mean age of 11 weeks, 54% were male (n = 936), 13% were premature (n = 225), and 20% were born low birth weight (n = 349).
Coinfection status did not differ substantially by sex, whether the infant was premature or low birth weight, the number of rooms in the home, the number of children under 5 years old in the household, or whether the mother smoked (Table 1). Mother's age at delivery was older among mothers of monoinfected infants than among mothers of coinfected infants. Infants with a single infection were more likely to have a latrine in the home (45.2%) than infants with coinfection (38.2%). Among mothers of infants with coinfections, 45.5% had received the flu vaccine compared to 49.5% of monoinfected infants' mothers receiving the vaccine. Infants with coinfection had a mean age of 12.1 weeks and were an average of 1.3 weeks older than infants with monoinfection. Fever was detected among 77% of infants who tested positive for influenza, while the rate of fever among infants with other viral infections was between 46% for infants with rhinovirus alone and 66% for infants with RSV alone (see Appendix S1).    was found between coinfection and severe disease (see Appendix S1).

| D ISCUSS I ON
In this study utilizing prospective community-based home surveillance for respiratory illness of a birth cohort of infants in rural Nepal, we examined the correlation between respiratory viral coinfection on clinical outcomes. A substantial percentage (19%) of infants who tested positive for a viral respiratory infection had two or more concurrent viruses at their first illness, and infants with coinfection were more likely to experience extended duration of febrile illness than infants with monoinfection.
Our findings confirmed recent reports that coinfection is more prevalent than was believed before routine use of molecular viral diagnostics and that multiple viruses are often present during a respiratory illness episode. Infants with multiple infections tended to be one week older than infants with single infections, but this difference was not statistically significant. In a hospital-based study in Austria of respiratory viral coinfection among infants, the frequency of viral coinfection compared to monoinfection did not differ by month of age, which aligns with these findings. 2 This observed age difference may have been due to the effects of certain viruses, such as bocavirus or rhinovirus which are associated with prolonged periods of asymptomatic shedding. 19 The seven viruses in this analysis differed by their proportion of infections that were monoinfections and those that were coinfections. Bocavirus had a higher ratio of coinfection to monoinfection cases than any other virus in this analysis. Bocavirus has been found to have an extended period of viral shedding after infection, often remaining detectable in young children for over a month after their primary illness event. 19 Bocavirus is known to often be asymptomatic, so the frequency of multiple infections among infants with bocavirus may be due to infants experiencing symptoms from a subsequent infection and testing positive for a shedding bocavirus.  14 The mechanism of infection of the influenza virus may also play a role in its relative adversity to coinfection: The virus has been shown to block the progression of other viruses, and models have illustrated that the influenza virus can rapidly proliferate and generate a robust host inflammatory response, leaving little room for a subsequent viral infection. 14  Rhinovirus is usually associated with the common cold, but can lead to a range of symptom severities; rhinovirus can cause pneumonia in infants under 6 months old, but an estimated 15%-30% of cases are asymptomatic. 22 The range of clinical presentations associated with rhinovirus monoinfection makes it difficult to determine the source of symptoms in the 276 infants with rhinovirus coinfection.
Disease severity outcomes of pneumonia, seeking further medical care, and fever lasting longer than three days were used because they were measurable attributes of respiratory illness in this population. Fever lasting longer than three days may capture illness episodes that would have resulted in a hospital visit: A United States study found the mean duration of fever for children who were hospitalized to be four days. 23 Due to the inaccessibility of hospitals in this region, common outcomes of extended hospital stay, admission to an intensive care unit, or requiring supplemental oxygen were unable to be measured. Seeking medical care included visits to local healers and other providers, which is a more accessible form of care for this cohort. This limits the ability to compare these results with hospital-based studies, but the outcomes used in this analysis better capture the severity of illnesses among infants in this population.
Odds of pneumonia and odds of seeking further medical care did not differ between infants with any monoinfection and those with any coinfection. The subsets of infants with influenza and RSV, respectively, also did not have associations between number of infections and odds of pneumonia or seeking further medical care. This is consistent with findings from hospital-based studies in South Africa, Japan, Brazil, and Canada, where infants and young children with coinfection did not have increased illness severity compared to those with monoinfection. 7,8,10,24,25 A 2016 hospital-based study found RSV coinfection and monoinfection to not differ in disease severity. Those hospitalized with coinfections of RSV and influenza had greater odds of a prolonged stay than RSV alone. 10 While these reports support the results of this analysis, their hospitalized settings pose a challenge to comparing the data. A 2013 study among childcare attendees reported a similar lack of association between coinfection and illness severity. 26 Despite not being significantly associated with pneumonia or seeking further care, coinfection was associated with 1.4 times greater odds of having subjective fever for four or more days com- This analysis of infants experiencing their first respiratory illness demonstrated that coinfection is associated with increased duration of febrile illness, especially among infants with influenza. While seeking care and having symptoms of pneumonia did not notably differ between infants with monoinfections and infants with coinfections, extended fever duration is a serious concern for rural communities where respiratory illness is a leading cause of infant mortality. This analysis demonstrates that coinfection affects clinical outcomes in a cohort of infants in a rural setting. On an individual level, these findings demonstrate the importance of diagnostic methods that can accurately identify coinfection, which may play an important role in clinical outcomes. On a broader scale, vaccines for influenza, RSV, and other respiratory illnesses are being considered internationally, and the primary trial results showed maternal influenza immunization to effectively protect infants from respiratory illness. 14 As vaccines for respiratory viruses continue to be developed and tested around the world, coinfection remains an important clinical factor that needs to be considered when designing and measuring the effects of these interventions.