1 INTRODUCTION

Respiratory syncytial virus (RSV) is a common pathogen responsible for mild cold-like symptoms to lower respiratory tract infection (LRTI). Generally people recover within 2 weeks, but it can be serious in infants, children <5 years, older adults (especially ≥60 years), immunocompromised, or people with underlying cardio-pulmonary conditions.1 Globally, in 2015, approximately 33.1 million children <5 years reported RSV-associated acute LRTI, of which 3.2 million were hospitalized and 59,600 died, while 336,000 older adults were hospitalized due to RSV-associated acute respiratory infection, of which 14,000 died.2, 3 Annually, over 177,000 older adults are hospitalized and 14,000 die in the United States (US) with RSV infection.1

Older adults have higher risk of severe complications from RSV due to immunosenescence.1 A 4-year prospective study in the US during 1999 to 2003 showed that 3%–7% of healthy older adults, 4%–10% of high-risk patients, and 8%–13% of hospitalized patients had RSV infection.4 Another prospective observational cohort study in Europe (EU) during 2018 to 2019 showed that RSV infection is prevalent in community-dwelling older adults (≥60 years) with an annual incidence of 2.1%–4.9%; however, it rarely causes severe disease.5 In long-term care facilities, RSV is estimated to infect 5%–10% of residents per year with significant rates of pneumonia (10%–20%) and death (2%–5%).6

In the EU and the US, RSV is increasingly recognized, including impact on older adults; however, little is known about RSV infection in Japan.7 Several prophylactic vaccines against RSV disease in older adult populations are currently under development. As Japan has a significant population over 60 years of age (~35.6 million),8 it is important to understand the burden of RSV disease in this population.

The objectives of this study were to estimate the occurrence rate of RSV infection, and to characterize the RSV disease outcome in older adults in Japan.

2 METHODS 2.1 Study population

Ambulatory adults (≥65 years) living in Japan in community or assisted-living or long-term care residential facilities, and who were, healthy or with clinically stable medical conditions, were enrolled. Participants with acute respiratory disease (ARD) at consent and Week 0 (baseline), or who had received experimental antiviral drugs or vaccines for RSV within 6 months prior to the study, or had a life expectancy <1 year, were excluded.

2.2 Study design

The study was conducted at 10 sites in Japan from April 2019 to July 2020. Participants were followed up bi-weekly (weekly during RSV season, predefined from 01 June 2019 to 30 November 2019)9 through telephone to monitor ARD. Each participant was observed for 52 weeks from the day of registration or until withdrawal (Figure 1).

Schematic diagram of the study. Greyed cells in the above figure indicate scheduled activities to be performed at home. White cells in the above figure indicate scheduled activities to be performed at the study site. *Of which >30% aged 65 to 74 years and >30% aged ≥75 years and older (including >5% of 85 years and older as possible). **In the case that participants recovered and symptoms recurred, ARD was defined as a single episode or a separate episode of respiratory illness by the clinical judgment of the investigator; however, the episodes of respiratory illness that occurred >28 days of each other as a separate episode of illness were to be counted.10 Abbreviations: ARD = acute respiratory disease, ICF = informed consent form, RiiQ™ = Respiratory Infection Intensity Questionnaire, RSV = respiratory syncytial virus

The study was conducted in accordance with the principles defined in the Declaration of Helsinki, Ethical Guidelines for Medical and Health Research Involving Human Subjects, Japanese regulations, and sponsor policy.

2.3 Data collection

Clinical data of coexisting medical conditions including conditions conferring increased risk for severe RSV disease: congestive heart failure, coronary artery disease, chronic obstructive pulmonary disease, asthma, or immunocompromised condition due to medications, transplant, asplenia, or innate immunodeficiency, etc., were collected at Week 0. The participants visited study sites after the onset of any ARD symptoms, and a nasopharyngeal swab sample per episode was collected if they were diagnosed with ARD by the investigator. The ARD symptoms to be filled in the survey sheet included LRTI (cough, shortness of breath, sputum production, and wheezing); upper respiratory tract infection (URTI; nasal congestion and sore throat); and systemic symptoms (headache, fatigue, fever [>38°C], feverishness, and myalgia) by the investigator. The Respiratory Panel of BIOFIRE® FILMARRAY® Multiplex polymerase chain reaction (PCR) system was used for the identification of pathogens in the nasopharyngeal swabs. The RSV-positive samples were further analyzed for RSV A or B subtyping, performed using RealStar® RSV reverse transcriptase (RT)-PCR kit 3.0. The Respiratory Infection Intensity and Impact Questionnaire (RiiQ™) was used to assess ARD symptoms by the participants. Its dimensions included LRTI, URTI, and systemic symptoms. The RiiQ™, adapted from Influenza Intensity and Impact Questionnaire (Flu-iiQ™), was used under license from Measured Solutions for Health P/L, Australia. This is a patient-reported outcome instrument designed to assess symptom intensity and impact of respiratory infection. The RiiQ™ scales are scored as an average of each symptom related to each scale. Each symptom was scored from 0 (none) to 3 (severe).11, 12 The RiiQ™ and body temperature were assessed daily in all ARD participants from the onset of ARD symptoms until recovery as confirmed by investigators or until Week 52 (if participants did not recover). Investigator also collected information regarding utilization of medical services (outpatient visits, emergency room [ER] visits, hospitalizations, or intensive care unit [ICU] hospitalizations) during the ARD period and assessed the impact of ARD on basic functioning and independence status using the Barthel Index of Activities of Daily Living (BIADL) Assessment Sheet. The BIADL is a 10-item ordinal scale with scores ranging from 0 to 20 (lower scores indicating increased disability).13 Functional status >16 was defined as independent.14, 15

2.4 Statistical Methods 2.4.1 Sample size determination

Approximately 1000 participants (>30% aged 65 to 74 years and >30% aged ≥75 years [including >5% aged ≥85 years]) were planned. The occurrence rate of RSV infection in older adult population was conservatively estimated to be 2%.4 A total of 1000 participants were expected to allow for precision of estimation of the true attack rate of 2%, as measured by the width of the two-sided 95% confidence interval (CI).

2.4.2 Endpoints Primary

Occurrence of RSV–ARD (defined as new onset or worsening of ≥1 symptom of URTI and/or LRTI with or without systemic symptoms with an RT-PCR confirmed diagnosis of RSV) in older adults in Japan during the study period.

Secondary

Occurrence of RSV-lower respiratory tract disease (LRTD; defined as new onset or worsening of at least 1 symptom of LRTI, assessed by the investigator, with an RT-PCR confirmed diagnosis of RSV), occurrence of ARD/LRTD caused by respiratory pathogens other than RSV; outcome of RSV–ARD and ARD caused by respiratory pathogens other than RSV (recovery conditions, signs and symptoms, types of medical services utilized for ARD); and occurrence rates of RSV–ARD or ARD caused by other respiratory pathogens by subgroups (age groups, gender, presence of underlying comorbidities or immunosuppression, functional status, smoking history, influenza vaccination in 2019/2020, and place of residence).

Of the above secondary endpoints, occurrence of LRTD caused by other respiratory pathogens, occurrence rate of RSV–ARD by smoking history and influenza vaccination in 2019/2020, and occurrence of ARD caused by other respiratory pathogens by demographic data were evaluated as post hoc analyses.

Exploratory

Number of RT-PCR-confirmed RSV–LRTD cases in clinical case definitions, defined as ≥3 LRTI symptoms for case definition #1, ≥2 LRTI symptoms for case definition #2 and ≥2 LRTI symptoms or ≥1 LRTI combined with ≥1 URTI and ≥1 systemic symptom for case definition #3, based on clinical outcome assessed by RiiQ™ and Body Temperature Record Sheet; trend of the seasonality of RT-PCR-confirmed RSV infection and other respiratory pathogens; and change from baseline functioning and independence status assessed by clinician using BIADL questionnaire.

The diagnosis and severity at ARD diagnosis by investigators, and the time from the onset day of ARD symptoms to the day of ARD diagnosis (ARD Visit Day 0) were also evaluated as post hoc analyses.

2.4.3 Statistical analyses

This was an exploratory descriptive study; therefore, no hypothesis testing was conducted. The precision of the estimate of the occurrence rate was determined using the Clopper–Pearson exact 95% CI. All participants having baseline (Week 0) data were included in the analyses. All data were summarized descriptively.

3 RESULTS 3.1 Study population

A total of 1000 participants were enrolled; 913 (91.3%) completed and 87 (8.7%) discontinued the study. Discontinuations were due to consent withdrawal (n = 58 [5.8%]), lost to follow-up (n = 22 [2.2%]), death (n = 6 [0.6%]), and failure to comply with protocol requirements (n = 1 [0.1%]). Overall, 313/1000 (31.3%) participants had 459 ARD visits (Table 1). Overall compliance rate of recording patient reported outcome during each ARD period was 81.8% (data not shown).

TABLE 1. Summary of participant disposition, demographic variables, and baseline characteristics Variables Value Participants enrolled 1000 Completed study n (%) 913 (91.3) Discontinued from the study n (%) 87 (8.7) Reason for discontinuation Lost to follow-up 22 (2.2) Consent withdrawal 58 (5.8) Death 6 (0.6) Repeated failure to comply with protocol requirements 1 (0.1) Time of discontinuation n (%) Before start of RSV season 32 (3.2) During RSV season 16 (1.6) After RSV season 39 (3.9) Number of ARD visits 459 Participants with ARD visit n (%) 313 (31.3) Sex n (%) Female 551 (55.1) Male 449 (44.9) Age (years) Mean (SD) 74.3 (6.23) Age group (years) n (%) 65–74 543 (54.3) 75–84 383 (38.3) ≥85 74 (7.4%) Resident information n (%) Home 996 (99.6%) Assisted-living or long-term care residential facility 4 (0.4%) Coexisting medical condition n (%) No 42 (4.2%) Yes 958 (95.8%) Chronic heart disease Congestive heart failure 18 (1.8%) Coronary artery disease 66 (6.6%) Chronic lung disease COPD 31 (3.1%) Asthma 104 (10.4%) Immunocompromised 7 (0.7%) Other 923 (92.3%) Abbreviation: ARD = acute respiratory disease, COPD = chronic obstructive pulmonary disease, n = number of participants, RSV = respiratory syncytial virus, SD = standard deviation.

The proportion of females (55.1%) was higher than males (44.9%); the mean (standard deviation [SD]) age was 74.3 (6.23) years; more than half of participants (54.3%) were 65 to 74 years. Most participants were home residents (99.6%), living with family/cohabiter (80.9%), nonsmokers (57.6%), and were vaccinated with influenza vaccine in 2019/2020 (64.5%). At baseline, most participants (998 [99.8%]) had a functional status of >16. The most common (reported in >10% of participants) coexisting medical condition was asthma (n = 104 [10.4%]) among conditions with increased risk for severe RSV (Tables 1 and 2).

TABLE 2. Summary of occurrence rate of RSV–ARD or ARD caused by influenza A/H1 by demographic and baseline characteristics subgroups during the whole study period Subgroups N = 1000 RSV–ARD ARD by influenza A/H1 n (%) 95% CI n (%) 95% CI 24 (2.4) 1.54, 3.55 11 (1.1) 0.55, 1.96 Age (years) 65–74 543 14 (2.6) 1.42, 4.29 8 (1.5) 0.64, 2.88 75–84 383 8 (2.1) 0.91, 4.07 3 (0.8) 0.16, 2.27 ≥ 85 74 2 (2.7) 0.33, 9.42 0 — Sex Female 551 19 (3.4) 2.09, 5.33 8 (1.5) 0.63, 2.84 Male 449 5 (1.1) 0.36, 2.58 3 (0.7) 0.14, 1.94 Presence of underlying comorbidities or immunosuppression Yes 958 23 (2.4) 1.53, 3.58 11 (1.1) 0.57, 2.05 Chronic heart disease (CHF + CAD) 78 1 (1.3) 0.03, 6.94 1 (1.3) 0.03, 6.94 Chronic lung disease (COPD+Asthma) 127 4 (3.1) 0.86, 7.87 2 (1.6) 0.19, 5.57 Immunocompromised 7 0 — 0 — Other 923 23 (2.5) 1.59, 3.72 11 (1.2) 0.60, 2.12 Diabetes mellitus (SMQ) 424 9 (2.1) 0.98, 3.99 4 (0.9) 0.26, 2.40 Hematological malignant tumors (SMQ) 2 0 — 0 — No 42 1 (2.4) 0.06, 12.57 0 — Functional statusa ≤16 2 0 — 0 — >16 998 24 (2.4) 1.55, 3.56 11 (1.1) 0.55, 1.96 Place of residence Home 996 24 (2.4) 1.55, 3.56 11 (1.1) 0.55, 1.97 Assisted-living or long-term care residential facility 4 0 — 0 — Structure of family living together/cohabiters Alone 191 5 (2.6) 0.86, 6.00 0 — Living with family/cohabiter 809 19 (2.3) 1.42, 3.64 11 (1.4) 0.68, 2.42 Smoking history Smoker 85 1 (1.2) 0.03, 6.38 0 — Former smoker 339 8 (2.4) 1.02, 4.60 2 (0.6) 0.07, 2.11 Nonsmoker 576 15 (2.6) 1.46, 4.26 9 (1.6) 0.72, 2.95 Seasonal influenza vaccination Yes 645 15 (2.3) 1.31, 3.81 8 (1.2) 0.54, 2.43 No 330 8 (2.4) 1.05, 4.72 2 (0.6) 0.07, 2.17 Unknown or missing 25 1 (4.0) 0.10, 20.35 1 (4.0) 0.10, 20.35 Abbreviations: ARD = acute respiratory disease, CAD = coronary artery disease, CHF = congestive heart failure, CI = confidence interval, COPD = chronic obstructive pulmonary disease, RSV = respiratory syncytial virus, SMQ = Standardised MedDRA Queries. a The score is from the Barthel Index of ADL. 3.2 Primary outcomes

RSV–ARD was detected in 24/1000 (2.4%) participants (95% CI: 1.54, 3.55) by RT-PCR; of which, 13/24 (54.2%) and 11/24 (45.8%) had Subtypes A and B RSV–ARD, respectively (Table 3).

TABLE 3. Summary of occurrence rate of RSV–ARD, RSV–LRTD, and ARD and LRTD by other respiratory pathogens during the whole study period RSV–ARD (N = 1000) RSV–LRTD (N = 1000) Total Subtype A Subtype B Total Subtype A Subtype B n (%) 24 (2.4%) 13 (1.3%) 11 (1.1%) 8 (0.8%) 7 (0.7%) 1 (0.1%) 95% CI 1.54, 3.55 0.69, 2.21 0.55, 1.96 0.35, 1.57 0.28, 1.44 0.00, 0.56 Pathogens ARD by other pathogens LRTD by other pathogens n (%) 95% CI n (%) 95% CI Coronavirus 229E 3 (0.3) 0.06, 0.87 — — Coronavirus HKU1 1 (0.1) 0.00, 0.56 — — Coronavirus OC43 12 (1.2) 0.62, 2.09 2 (0.2) 0.02, 0.72 Coronavirus NL63 4 (0.4) 0.11, 1.02 — — Human Metapneumovirus 12 (1.2) 0.62, 2.09 6 (0.6) 0.22, 0.72 Human Rhinovirus/Enterovirus 100 (10.0) 8.21, 12.03 21 (2.1) 1.30, 3.19 Influenza A/H1 11 (1.1) 0.55, 1.96 2 (0.2) 0.02, 0.72 Influenza B 1 (0.1) 0.00, 0.56 — — Parainfluenza 1 2 (0.2) 0.02, 0.72 — — Parainfluenza 2 3 (0.3) 0.06, 0.87 1 (0.1) 0.00, 0.56 Parainfluenza 3 20 (2.0) 1.23, 3.07 3 (0.3) 0.06, 0.87 Parainfluenza 4 1 (0.1) 0.00, 0.56 — — Bordetella pertussis 1 (0.1) 0.00, 0.56 — — Mycoplasma pneumoniae 1 (0.1) 0.00, 0.56 — — Abbreviations: ARD = acute respiratory disease, CI = confidence interval, LRTD = lower respiratory tract disease, RSV = respiratory syncytial virus. Note: LRTD cases are shown in this table as ARD. 3.2.1 Secondary outcomes

RSV–LRTD, as assessed by the investigator was detected in 8/1000 (0.8%) participants (95% CI: 0.35, 1.57); 7 (0.7%) and 1 (0.1%) had Subtypes A and B RSV–LRTD, respectively (Table 3).

The median (range) duration of RSV–ARD was 18.0 (10 to 33) days (Table S1). The most frequent symptoms recorded by the participants (mean change of >1.0 in symptom score from baseline) for RSV–ARD were cough, sore throat, nasal congestion, and expectoration. For most participants (n = 15), symptoms improved by Day 14; however, for few participants (n = 7), the symptoms lasted for 21 to 28 days after onset (data not shown). Respiratory tract symptoms tended to take longer to return to baseline for RSV and human metapneumovirus than for influenza A/H1 (Figure S1). The body temperature of participants with the RSV–ARD was stable over time (Days 0 to 28; 35.0°C to 38.4°C) and was similar in Subtypes A and B. Only 1 participant with Subtype B had fever (38.4°C) on Day 4 from onset of ARD. All 24 participants detected with RSV had utilized medical services of outpatient visits and 1 (4.2%) was hospitalized for 10 days for RSV–ARD. No participant died, was admitted to the ICU, or visited an ER due to RSV–ARD.

Pathogens were detected in 199/459 samples. Coinfection was confirmed in 2 samples, and others detected a single pathogen. The most common viruses other than RSV that caused ARD (detected in >10 participants) were human rhinovirus/enterovirus (100 [10.0%]), parainfluenza 3 (20 [2.0%]), coronavirus OC43 and human metapneumovirus (12 [1.2%]), and influenza A/H1 (11 [1.1%]). ARD due to influenza B virus was reported in 1 participant. Only human rhinovirus/enterovirus-infected participants reported with >1 episode of ARD. The most common viruses other than RSV that caused LRTD (detected in >10 participants) was human rhinovirus/enterovirus (21/1000 [2.1%]) (Table 3).

No participant died or was admitted to the ICU due to ARD caused by pathogens other than RSV. The median (range) duration of ARD caused by those viruses or bacteria (detected in >10 participants) was 20.5 (10–36) days, 15.5 (9–31) days, 15.0 (5–55) days, 15.0 (5–38) days, and 12.5 (6–35) days for human metapneumovirus, parainfluenza 3, human rhinovirus/enterovirus, influenza A/H1, and coronavirus OC43, respectively (Table S1). The most frequent symptoms recorded by participants were cough, sore throat, expectoration, and loss of appetite with parainfluenza 3; cough, feeling feverish, and fatigue with influenza A/H1; cough and sore throat with human rhinovirus/enterovirus; cough and expectoration with human metopneumovirus; and nasal congestion with coronavirus OC43. Fever was reported with human metapneumovirus, human rhinovirus/enterovirus, influenza A/H1, parainfluenza 2, parainfluenza 3, and Bordetella pertussis (data not shown). Two participants with ARD were hospitalized, 1 due to human metapneumovirus for 6 days and 1 due to influenza A/H1 for 9 days.

Of the 24 participants who had RSV–ARD; 14 (58.3%) were 65- to 74-years-old, 8 (33.3%) were 75- to 84-years-old, and 2 (8.3%) were ≥85-years-old. The occurrence rate of RSV–ARD was higher in females (19/551; 3.4%) than males (5/449; 1.1%). Four of 127 (3.1%), 9/424 (2.1%), and 1/78 (1.3%) participants with underlying comorbidities of chronic lung disease, diabetes mellitus, and chronic heart disease, respectively, had RSV–ARD. RSV–ARD was detected in 15/576 (2.6%) nonsmokers, 8/339 (2.4%) former smokers and 1/85 (1.2%) smokers with a history of smoking; and in 15/645 (2.3%) vaccinated and 8/330 (2.4%) not vaccinated among seasonal influenza vaccination (Table 2).

Of the other respiratory pathogens, only influenza A was analyzed for subgroups, given influenza A is a well-recognized respiratory disease caused ARD. All 11 participants with influenza A detected influenza A/H1. Of the 11 participants; 8 (72.7%) were 65- to 74-years-old and 3 (27.3%) were 75- to 84-years-old. The occurrence rate of ARD by influenza A/H1 was higher in females (8/551; 1.5%) than males (3/449; 0.7%). Two of 127 (1.6%), 1/78 (1.3%), and 4/424 (0.9%) participants had ARD by influenza A/H1 with underlying comorbidities of chronic lung disease, chronic heart disease, and diabetes mellitus, respectively. The ARD by influenza A/H1 was detected in 9/576 (1.6%) nonsmokers and 2/339 (0.6%) former smokers among smoking history; and in 8/645 (1.2%) participants vaccinated and 2/330 (0.6%) not vaccinated among seasonal influenza vaccination (Table 2).

3.2.2 Exploratory outcome

The three case definitions ranged from the most stringent case definition #1 to the least stringent case definition #3. Of the 1000 participants, 17 (1.7%), 22 (2.2%), and 23 (2.3%) participants had RT-PCR confirmed RSV–LRTD based on case definitions #1, #2, and #3, respectively. One participant was not classified in any case definition (Table 4).

TABLE 4. Summary of RSV participants with clinical case definitions #1 to #3 during the whole study period Clinical case definitions Total Subtype A Subtype B n (%) 95% CI n (%) 95% CI n (%) 95% CI #1a 17 (1.7) 0.99, 2.71 10 (1.0) 0.48, 1.83 7 (0.7) 0.28, 1.44 #2