key: cord-0993263-mklf5pnx
authors: Drews, Ashley L.; Atmar, Robert L.; Glezen, W. Paul; Baxter, Barbara D.; Piedra, Pedro A.; Greenberg, Stephen B.
title: Dual Respiratory Virus Infections
date: 1997-12-03
journal: Clin Infect Dis
DOI: 10.1086/516137
sha: 7f99feefc6f7b6cc21d12f88eab60f79114e1089
doc_id: 993263
cord_uid: mklf5pnx

We retrospectively reviewed eight prospective epidemiological studies conducted between 1991 and 1995 for dual respiratory virus infection (DRVI) to determine the frequency, associated comorbid conditions, clinical presentations, and morbidity related to DRVI among immunocompetent persons. Two viruses were identified as the cause of 67 (5.0%) of 1,341 acute respiratory virus infections. DRVI was detected in patients from <1 year to 79 years of age, in both sexes, and in many races. Forty-two percent of patients with DRVI were ⩽4 years old. Fifty-eight percent of patients with DRVI had underlying chronic lung disease. DRVI was associated with upper respiratory tract illness; lower respiratory tract illness, including pneumonia; systemic influenza-like illnesses; and exacerbations of asthma or chronic obstructive pulmonary disease. All of the common acute respiratory viruses were identified; picornaviruses and influenzavirus A were the most common. The rate of DRVI (11.6%) was highest in the epidemiological studies in which cell culture, serology, and polymerase chain reaction were used together. Patients with DRVI were hospitalized significantly more often than those with respiratory infection due to a single virus (46.3% vs. 21.7%; P < .01). The percentage of DRVIs increased proportionally with the number of diagnostic methods used.

Materials and Methods has been reported infrequently. The majority of the studies have Study design. We retrospectively reviewed the charts inbeen published in the pediatric literature, with most infections cluded in epidemiological studies of community-acquired resreported to be due to respiratory syncytial virus (RSV) and a piratory virus infections conducted at our institution between second respiratory virus [1 -8] . The frequency of dual respira-1991 and 1995. A DRVI was defined as an acute respiratory tory virus infection (DRVI) varies widely in the literature, and virus infection and any combination of culture(s), serological the clinical relevance of DRVI is unresolved. Some authors test(s), or PCR(s) positive for two different viruses. An SRVI have found that the morbidity associated with DRVI is higher was defined as an acute respiratory virus infection caused by than that associated with single respiratory virus infection a single virus detected by either culture, positive serology, or (SRVI) [3] , while other authors have not found that DRVIs PCR. The charts of all patients with DRVI in these epidemioare more severe than SRVIs [2, 5, 8 -12] . We performed a logical studies were reviewed by one of us (A.L.D.), and data retrospective review of prospective epidemiological studies of on demographics, comorbid conditions, date of onset of the respiratory virus infection carried out by the Acute Viral Respiacute respiratory illness, results of viral diagnostic tests, and ratory Disease Unit at Baylor College of Medicine between clinical presentation were recorded. A computerized database 1991 and 1995. We examined the incidence, demographics, was used to obtain the results of viral diagnostic tests performed and clinical presentations of DRVI. A review of the literature for the other patients in these studies. Information on demoon DRVI was performed, and our results were compared with graphics and hospitalization status of those with SRVI was also those previously reported.

abstracted. Patient populations. A total of 4,336 patients were enrolled in eight different prospective epidemiological studies of acute respiratory viral disease conducted between 1991 and 1995. with chronic obstructive pulmonary disease (COPD) and con-ered an acute respiratory virus infection on the basis of serology of acute and convalescent sera, a significant increase in titers trol patients ú50 years old were observed longitudinally for the development of acute respiratory illness; in another study, of antibody had to be detected within 6 weeks of the onset of the illness. young adults with asthma were observed longitudinally for the development of acute respiratory disease; and in the third study, A significant rise in antibody titer was defined for influenzaviruses A and B as follows: at least a sixfold increase by adults with exacerbations of acute asthma who presented to a county hospital emergency center (EC) were evaluated. In an-microneutralization assay or at least an eightfold increase by hemagglutination inhibition assay or at least a fourfold increase other smaller study, pairs of mothers and infants from birth to 3 years of age were observed for the development of acute in antibody to the same antigen in two different assays. We excluded titer increases due to influenza vaccination by assum-respiratory disease. Cell culture alone was used for diagnosis in the two outpatient clinic studies. In all of the other studies, ing that all simultaneous increases in antibody to influenzaviruses A H1 and H3 antigens and influenzavirus B antigen were cell culture and serology were used, and in the two studies of asthmatics, PCR was used as well.

due to vaccination. Furthermore, the charts of all patients with potential influenzavirus infection, as determined by serology Viral diagnostic methods. Cell cultures for influenzavirus types A and B, RSV, parainfluenzavirus types 1, 2, and 3, alone, were reviewed for influenza vaccination status. This information had been collected prospectively in all of the epide-adenoviruses, and picornaviruses were performed by using the following cell lines: human embryonic lung fibroblast, human miological studies reviewed. If there was documentation of influenza vaccination within 6 weeks of the acute respiratory epidermoid laryngeal carcinoma, African green monkey kidney, and Madin-Darby canine kidney or primary rhesus mon-illness, these patients were not considered to have influenzavirus infection; however, they could be included as a case of key kidney. Standard detection and identification methods of viruses were used [13] . Cultures positive for herpes simplex SRVI or DRVI if they had infection(s) with any of the other respiratory viruses. virus (HSV) or cytomegalovirus (CMV) were excluded from the analysis.

For parainfluenzaviruses and RSV, a significant rise in antibody titer was defined as at least a sixfold rise by microneutrali-Serology was performed by microneutralization tests for detection of influenzavirus types A and B; RSV; and parainfluen-zation assay. Given the cross-reactivity of antibody to parainfluenzavirus types 1, 2, and 3, a positive serology for more zavirus types 1, 2, and 3 [14, 15] . An ELISA to detect antibody to coronavirus OC43 was performed as described previously than one type of parainfluenzavirus was not included as a DRVI; instead, it was considered a single parainfluenzavirus [16] . Serology for influenzaviruses A and B was also performed by hemagglutination inhibition by using previously described infection, and the type assigned was the prevalent type circulating in the community at that time. techniques [17] . For an acute respiratory illness to be consid-/ 9c40$$de18 11-04-97 18:01:16 cidas UC: CID with DRVI. Thirty-one (58%) of these 53 had underlying chronic lung disease (COPD or asthma). Five patients had hypertension. Two patients had congestive heart failure, two For coronavirus OC43, a significant rise in antibody titer had diabetes mellitus, and two had peptic ulcer disease. was defined as either a single fourfold or greater rise or a §2.5-Clinical presentations. Information was available regardfold rise that was reproducible upon repeated testing.

ing the clinical diagnosis of the acute respiratory illness for 60 Reverse transcriptase-PCR was performed to detect coroof the 67 patients. Upper respiratory tract infection was the naviruses and picornaviruses in two epidemiological studies most common diagnosis (55% of patients), but 40% of the and to detect influenzavirus A in one of these studies. Viral patients had lower respiratory tract infections. An exacerbation nucleic acids were extracted from respiratory secretions, and of underlying lung disease (COPD or asthma) was associated previously described primers for amplification of coronavirus with DRVI in 42% of the patients. Eighteen percent of these OC43 [18] , influenzavirus A [19] , and picornavirus [20] were patients presented with a systemic or influenza-like illness, and used. Complementary DNA synthesis and PCR amplification a small number (5%) had associated congestive heart failure. were performed by using a PTC-100 thermal cycler [20, 21] .

Respiratory viruses causing DRVI. There were 20 different Positive results were identified by slot blot or Southern blot virus combinations isolated from patients with DRVI (table 3) . hybridization with use of digoxigenin-labeled oligonucleotides Picornaviruses (33 patients), most of which were rhinoviruses, [19, 22] . Pre-and post-PCR procedures were performed in and influenzavirus A (28) were the most common viruses idenseparate rooms on different floors, and other standard precautified in DRVIs, and the combinations of a picornavirus with tions were used to prevent carryover contamination during the influenzavirus A (10 patients) or a picornavirus with coronaviperformance of reverse transcriptase-PCR assays [23] . For each rus OC43 (10) Overall, there was no significant difference in the age of hospitalized patients with DRVI vs. SRVI or in the percentage cell culture was the only diagnostic method used had the lowest of those õ1 year old or ú65 years old. One asthmatic patient had two DRVIs and two SRVIs, and she was not hospitalized . The rates of DRVI in these reports were not higher than the overall rate of SRVIs; three were not hospitalized with any of these infections, and one was hospitalized with the DRVI but not with either DRVI in all of the studies reviewed. DRVIs have been reported in both inpatients and outpatients SRVI. One patient had one DRVI and three SRVIs and was not hospitalized with any of these infections. No patient with with diagnoses of upper respiratory tract illness, lower respiratory tract illness, systemic or flu-like illness, and exacerbation both DRVI and SRVI was hospitalized with SRVI and not DRVI.

of asthma. Our review also found DRVI to be associated with these various clinical presentations; however, no predilection for DRVI in a specific clinical syndrome was apparent.

All of the usual respiratory viruses have been reported in cases of DRVI. In addition, many studies include HSV [3, 5, Over the past 45 years, more than 430 cases of DRVI in immunocompetent hosts have been reported ( [24] , reovirus [24] , and mumps virus of DRVI in these reports varies widely. Many published studies of acute respiratory illness report no cases of DRVI [43 -53].

[28] have also been reported in DRVI. Many of the studies in the literature do not specify which of the identified respiratory In the published studies in which DRVIs have been reported, the rate of DRVI ranged from 1.33% to 100% in a single case viruses were specifically involved in DRVI. In the studies that do specify the viruses involved in DRVI, RSV is generally the report. In our study, the overall rate of DRVI was 5%, but the rate of DRVI in the individual epidemiological studies that we most common one, and influenzaviruses A, B, and C, parainfluenzaviruses, adenoviruses, and rhinoviruses also are fre-reviewed ranged from 1.8% to 15.8%. The wide range of DRVI rates in our studies and in the literature is due to multiple factors quently involved. In our patients with DRVI, picornaviruses and influenzavirus A were the most common viruses, followed including differences in patient populations (e.g., differences in age and comorbid conditions), time of study (e.g., winter vs. by coronavirus OC43. Most of the published reports did not attempt to identify summer or during epidemics of respiratory virus infections), and diagnostic methods employed. Both the literature and our coronavirus as an etiologic agent, but Lina et al. [42] reported that 15 of 16 cases of DRVI in their study included a coronavi-analysis show that the number of DRVIs identified increases with the number of viral diagnostic methods used.

rus. Coronavirus was detected in 32.8% of our patients with DRVI, while this virus was detected in only 2.3% of the patients Most reports of DRVI have involved children, but there have been reports of DRVI in immunocompetent adults as well. Our with SRVI. The increased frequency of coronavirus in cases of DRVI may be an artifact of our viral detection methods study shows that a greater percentage of patients with DRVI than those with SRVI are õ1 year old (25% vs. 15%). It is rather than a true difference. Our detection methods included PCR for coronavirus, and ú40% of our coronavirus infections unclear from the literature and our study why younger patients are more often dually infected with respiratory viruses. The were detected by PCR alone. In most of the published studies of DRVI, more than one immature immune system of infants and lack of previous exposure to respiratory viruses could increase susceptibility to si-viral diagnostic technique was used to identify respiratory viruses. The rate of DRVI in the literature is dependent upon the multaneous infection with two or more respiratory viruses. Another possible but unproven explanation for the increased number of viral diagnostic methodologies used. When only one diagnostic method was used, the overall rate of DRVI was rate of DRVI among children is that there is something unique about RSV that facilitates infection with a second respiratory 1.8%, while when two virus detection methods were used, the rate of DRVI was 9.9%, and when three methods were used, virus. It is also possible that prolonged shedding of respiratory viruses occurs more commonly in children. We cannot exclude the rate was 8.4%. Likewise, our study shows that the rate of detection of DRVI increased with the number of viral diagnos-the possibility that our apparent DRVIs were actually separate but closely timed infections. tic methods used. It seems that DRVIs occur more frequently than is currently appreciated, and as viral diagnostic ability Fifty-eight percent of our patients with DRVI had underlying lung disease (asthma or COPD). The importance of this obser-improves, the number of DRVIs detected will increase. In the published reports in which serology was used, a four-vation is unclear because complete information regarding underlying lung disease for all enrolled patients in our studies fold or greater increase in the antibody titer between acute and convalescent sera was accepted as the definition of infection. was not available. This high rate may reflect the prevalence of underlying lung disease in our study populations rather than

We used more stringent criteria for serological diagnosis, which affected the number of DRVIs detected. If a fourfold rise in an increased susceptibility to DRVI in patients with underlying lung disease, since three of our studies specifically included antibody titer between acute and convalescent sera was considered diagnostic in our study as it is in the literature, then the patients with asthma or COPD. Most of the reports of DRVI / 9c40$$de18 11-04-97 18:01:16 cidas UC: CID The finding of a significantly higher hospitalization rate for performance characteristics of the serological tests, we chose to use a higher threshold for serological diagnosis to exclude patients with DRVI than for those with SRVI when identical viral diagnostic methods were used suggests that DRVIs are more potential false-positive increases in antibody titers.

It is not clear from the literature whether infection with two associated with increased morbidity. In addition, a small number of SRVIs were detected in asymptomatic patients, but all or more respiratory viruses causes more-severe clinical illness than infection with a single respiratory virus. Tristram et al.

DRVIs were associated with illness. Whether DRVIs are associated with increased morbidity can only be determined by a [4] reported an increased incidence of respiratory failure among patients dually infected with RSV and adenovirus, suggesting prospective clinical study in which patients with DRVI, SRVI, and respiratory illnesses without a known etiology are followed that DRVI may be more severe than SRVI. On the other hand, several reports state that the severity of DRVI is not greater longitudinally. DRVIs occur in immunocompetent hosts of all ages, particu-than infection with either agent alone, as was stated in the first reported case of DRVI [9] . Similarly, by using duration of larly infants. DRVIs are associated with all syndromes of respiratory illness and are found in both inpatients and outpatients. hospitalization as an indicator of severity, Portnoy et al. [10] did not find any significant difference in terms of severity All of the common acute respiratory viruses are involved in DRVI. All viral diagnostic methods contribute to detection of between DRVI and SRVI.

Mufson et al. [12] found no association between DRVI and DRVI, and the rate of DRVI increases as the number of diagnostic methods used increases. Wider application of PCR a specific disease syndrome. In reporting a simultaneous epidemic of RSV and parainfluenzavirus type 3 in a neonatal should increase the rate of detection of dual infections and allow clearer distinction between DRVI and SRVI. The per-intensive care unit, Meissner et al.

[2] stated, ''The infants infected by two viruses simultaneously could not be distin-centage of patients with DRVI who are hospitalized is greater than that of patients with SRVI, suggesting possible increased guished clinically from infants infected by a single agent.'' In a case-control study of children infected with RSV alone vs. morbidity associated with DRVI; however, only a prospective study with use of identical viral diagnostic methodologies for RSV and another respiratory virus, Subbarao et al. [5] found no difference between the two groups with use of a scoring all patients could resolve this question. Determination of the clinical significance of DRVI should become more important system for severity of clinical illness. In a study comparing infants who had lower respiratory tract infections caused by as more DRVIs are detected by newer, more sensitive viral diagnostic techniques. RSV alone vs. RSV and a second viral agent, Ray et al. [8] found that infants õ6 months old were more likely to have a second viral agent identified in addition to RSV than were 1. Jacobs JW, Peacock DB, Corner BD, Caul EO, Clarke SKR. Respiratory syncytial and other viruses associated with respiratory disease in infants.

of hospitalization among patients with DRVI than among those 

acute viral infections of the respiratory tract in Thai children, with emphasis on laboratory diagnosis

Respiratory syncytial virus

a general hospital: etiology and host risk factors

Simultaneous recovery of type A and type C 1972

The etiology of respira

Multiple respiratory virus tory tract infections in a tropical country

Bacterial-viral interrelations in respiratory infec-31

pneumoniae and bacteria associated with lower respiratory tract disease 32

Concurrent respiratory syncytial virus and comparison of identification methods for influenza and other respiraand influenza A infections in the institutionalized elderly and chronically tory viruses of humans

Association of viral enza A and B and parainfluenza 1 and 2 viruses that uses continuous and Mycoplasma pneumoniae infections with acute respiratory illness cell lines and fresh serum enhancement

Detection of respiratory ogy associated with the use of formalin-inactivated respiratory syncytial syncytial, parainfluenza type 2, and adenovirus antigens by radioimmuvirus vaccine in cotton rats is not a unique viral phenomenon

Development and applica-258 -65. tion of an enzyme immunoassay for coronavirus OC43 antibody in acute 36. Tùmová B, Kasová V, Stumpa A. Mixed infections with influenza A and respiratory illness

Influenza viruses

Diagnostic procedures for viral, rickettsial and chla-Pathogens associated with acute lower respiratory tract infection in mydial infections

Isolation of viruses

Corofrom specimens submitted for direct immunofluorescence test for respinavirus infects and causes demyelination in primate central nervous ratory syncytial virus

Comparison of reverse pediatric outpatients: cause and clinical manifestations

transcription-PCR with tissue culture and other rapid diagnostic assays 111:194 -200. for detection of type A influenza virus

Advantages of multiple cell culture systems for 1993; 59:631 -5. detection of mixed-virus infections

Surveillance of community-acquired isolates as enteroviruses or rhinoviruses by using reverse transcriptionviral infections due to respiratory viruses in Rhone-Alpes (France) durpolymerase chain reaction

Diagnoses of acute lower respirahepatitis A virus in shellfish tissues with the PCR. Appl Environ Microtory tract infections in children in Rawalpindi and Islamabad

Avoiding false positives with PCR

Studies on the role of viruses, bacteria, and M. pneumoniae as causes

72: among children and adults: field study of contemporary importance of 161 -76. several viruses and appraisal of the literature

Patterns of virus infections in families with acute respiratory illnesses

A comparative virologiinfections in the community

cal study of children in hospital with respiratory and diarrhoeal illnesses

Aetiology of community-acquired

Clinical virology and viral surveillance in a pediatric group practice: the use of double-seeded hospital

/ 9c40$$de18

Viral respiratory tract infection and exacerbations of asthma in adult tion of respiratory viruses to severe exacerbations of asthma in adults

Relationship of bacterial and viral infections to exacerbations of asthma

prospective study of respiratory infection in adult asthmatics and their normal spouses

older infants (P Å .08), but there were no differences in type of lower respiratory tract illness (bronchitis, bronchiolitis, The authors thank Robert B. Couch, M.D., for his review of the croup, or pneumonia) or duration of illness between the two manuscript and his helpful suggestions.

When hospitalization is used as a marker for severity of References disease, our findings suggest that DRVIs are associated with increased morbidity. We found a significantly increased rate