Etiologies of Acute Respiratory Tract Infections

In this issue of Clinical Infectious Diseases, there are 2 articles that provide us with some insight into the various etiologic agents that can cause acute respiratory tract infection (ARTI) in general practice patients in The Netherlands [1] and into the significance of the human metapneumovirus (hMPV) in patients with community-acquired pneumonia (CAP) and exacerbations of chronic obstructive pulmonary disease (COPD) in Quebec, Canada [2].

If we examine the broad spectrum of infectious and noninfectious illnesses experienced by people of all ages worldwide, it is clear that ARTIs are, without question, the most common maladies encountered, regardless of age or sex [3]. The objectives of the Dutch study were to estimate the incidence of influenza-like illnesses (ILIs) and of other ARTIs in patients visiting their general practitioners (to determine the etiologic agents) and to test the hypothesis that asymptomatic persons with subclinical infection may act as sources of transmission [1]. The objective of the Canadian study was to examine the role of hMPV infection in adults with CAP and adults with exacerbations of COPD [2].

A case-control method was used in the Dutch trial, and ILI was defined as an “acute onset of illness” (duration of prodromal stage, ⩾3–4 days) and the presence of at least 1 of the following symptoms: cough, rhinitis, sore throat, frontal headache, retrosternal pain, or myalgia. ARTI was defined as an acute respiratory illness other than ILI with at least 1 of the following symptoms: cough, rhinitis, or sore throat.

Nose and throat swab specimens were obtained from case patients and control subjects, and viral cultures and PCR tests were performed for detection of adenovirus, coronavirus, enterovirus, hMPV, influenza virus, parainfluenza virus, rhinovirus, and respiratory syncytial virus (RSV), as well as for Mycoplasma pneumoniae, Chlamydophila pneumoniae, and Chlamydophila psittaci. Bacterial cultures of throat swab specimens were also performed to detect all bacterial pathogens that are known to potentially cause “community-acquired” respiratory infection. The findings were then semiquantitatively classified into 5 groups: no colonies, sporadic colonies, few colonies, several colonies, or many colonies.

The spectre of respiratory illnesses and of epidemic respiratory illnesses in particular, such as can be seen with influenza, was heightened recently by the epidemic of severe acute respiratory syndrome (SARS) during 2002–2003 and widespread publicity about concerns regarding a possible epidemic or even pandemic attributable to avian influenza [4, 5]. The more we learn and understand about the epidemiology and etiology of ARTIs, the better position we will be in to prevent and treat them.

Essentially, the article by van Gageldonk-Lafeber et al. [1] shows us a number of things. The overall incidences of consultations for ILIs and other ARTIs are 132 and 413 consultations per 10,000 person-years, respectively. The diagnoses made for case patients with ARTIs other than ILIs were primarily the common cold (36% of patients), acute pharyngitis (30%), and acute tonsillitis (20%). The main symptom for case patients with ILI was fever (90% of patients), and for those with other ARTIs, it was sore throat (76%). For the control subjects, the most common symptoms were joint-muscle complaints (21% of patients) and skin disorders (14%). Approximately 20% of control subjects consulted their family physicians for other reasons, such as to pick up prescriptions, to undergo a routine physical examination, and to accompany relatives.

Not surprisingly, no pathogens were detected in 35% of case patients; in contrast, pathogens were found in 31% of control subjects without airway complaints. In the case patients, the pathogens were viruses (58% of patients), group A β-hemolytic streptococci (11%), and mixed pathogens (virus plus group A β-hemolytic streptococci; 3%).

Case patients were defined as persons with ILI or an ARTI other than ILI who had not used antimicrobials in the previous 2 weeks, and the findings reported above are certainly in keeping with the general supposition that ARTIs are primarily caused by viruses. A breakdown of the most common pathogens in case patients and control subjects revealed that influenza A viruses were the most common pathogens in case patients with ILI (42% of patients), and rhinovirus was the most common pathogen in case patients with ARTI (25%).

In control subjects, RSV (17% of subjects) was the most common pathogen. The main conclusions of van Gageldonk-Lafeber et al. [1] are that most ARTIs are viral, rhinovirus was the most common pathogen in both case patients and control subjects, and patients without symptoms suggestive of ARTIs may harbor pathogens and could be a potential source of transmission of respiratory pathogens.

The majority of respiratory tract infections in general and of viral infections in particular are caused by rhinoviruses and coronaviruses [3]. In the category of respiratory infections, influenza was reported more frequently than the common cold, despite the fact that colds occur more frequently [3]. This is further supported by data on viruses associated with the common cold in which influenza virus was reported to occur in 25%–30% of cases [6, 7]. In another community-based study, the Tecumseh study [8], the viruses that caused all respiratory infections were reported. Here too, as in the Dutch study, rhinoviruses were the most frequently found pathogens.

The pathogens capable of causing viral respiratory disease vary in their ability to both initiate and transmit infection. If one examines the relative role or impact of different respiratory viruses in producing ARTIs, the top 3 etiologic agents (expressed as estimated percentages of all respiratory illnesses) are as follows: rhinovirus, 34%; coronaviruses, 14%; and influenza viruses, 9% [3].

The viral pathogens also vary in their ability to transmit infection. Coronaviruses are generally transmitted by large droplet spread, whereas influenza viruses are spread by airborne methods, including both aerosol and droplet spread [9]. The exact mode of transmission of rhinoviruses, however, is not clear, and it is still debated whether rhinovirus is transmitted primarily by direct contact (e.g., droplet nuclei) or by indirect contact. For example, it has been shown that these viruses can survive on surfaces and can be spread by inoculation of nose or eyes with one's fingers, and others have shown that spread by droplet transmission can take place as well [10].

One of the most recent outbreaks that drew worldwide attention to the potential of a pandemic was the outbreak of the SARS. This began in the Guandong province of China in November 2002 and subsequently spread to Hong Kong and, ultimately, to far reaches of the world. The etiologic pathogen was determined to be a new coronavirus (SARS-CoV) that presumably spread from animals that are sometimes eaten as part of exotic banquets by humans. As happens with other coronaviruses, the SARS-CoV probably first mutated and then spread to new and susceptible species (e.g., humans). It is likely that spread among humans happened primarily by close contact that probably included droplet spread, direct contact, or fomite transmission. There were also several instances that may have involved small particle airborne or fecal/oral transmission. Usually, 1 case would only result in spread of the virus to a few other individuals, but there were several well-documented cases of “super transmitters,” in which 1 person could infect ⩾10 other persons [11, 12].

One of the interesting findings of the Dutch study was the fact that the researchers found pathogens in 30% of control subjects. This was higher for the youngest age groups and might act as a potential source of infection of others. They drew attention to the fact that only 2 other studies have addressed this issue, one that involved children and adults and the other that examined only adults [13, 14]. The former showed that results of virologic assessments of asymptomatic subjects were positive for 12% of children and 4% of adults for rhinovirus and enterovirus, respectively [13]. The latter study showed that only 4% of asymptomatic subjects aged ⩾60 years tested positive for rhinovirus or coronavirus [14]. These figures are significantly lower than those reported by the Dutch study [1]. The frequencies of subclinical infection and of asymptomatic subjects, by age group, were as follows: children (age, 0–15 years), 68%; adults, 55%; and persons aged ⩾65 years, 51%. A number of possible explanations are offered, but the important point still remains that asymptomatic persons, whether young or old, may act as unsuspected sources of infection.

The article by Hamelin et al. [2] is an attempt to assess the role that hMPV plays in adults with CAP and exacerbations of COPD. hMPV is a newly discovered viral respiratory pathogen for which, thus far, there are limited epidemiologic data to describe it. It was first isolated in The Netherlands from nasopharyngeal secretions of 28 children collected over a 20-year period. The virus itself is a single-stranded negative-sense RNA virus, and sequence studies of isolates have identified 2 main lineages. It belongs to the order Mononegavirales, family Paramyxoviridae, and genus Metapneumovirus [15, 16].

As mentioned in the study by van Gageldonk-Lafeber et al. [1], a significant percentage of patients with ARTIs never have an etiologic pathogen identified. This is true not only for ILIs and other ARTIs but also in serious cases of CAP that require admission to the hospital and in nosocomial and ventilator-associated pneumonia that is managed in medical or surgical wards and the intensive care units in tertiary care university hospitals, despite the availability of the best equipment and access to excellent diagnostic laboratories. The article by Hamelin et al. [2] assesses 2 patient populations: patients with exacerbations of COPD aged ⩾40 years (with or without CAP) who sought help at one of the emergency departments at study hospitals, and patients aged ⩾18 years without COPD who had CAP that required admission to the hospital. Three hospitals were involved for the periods of 17 January 2002 through 6 May 2002 and 6 January 2003 through 6 May 2003.

Statistical analysis involving regression modeling showed an association between ARTIs and excess winter deaths, particularly for influenza and RSV, but it was also implied that other agents may have been involved as well [17]. The large number of negative results of culture, PCR, and serologic testing to identify a causative agent certainly bears this out as well.

Identification of the different pathogens that contribute to the various respiratory illnesses is very important not only from an epidemiological point of view but also from a treatment and prevention point of view, because such information may also help in the development of specific treatments aimed at these pathogens, as well as the development of preventive measures through the use of chemoprophylaxis or immunoprophylaxis (i.e., vaccines).

Hamelin et al. [2] studied nasopharyngeal aspirate specimens for the presence of influenza viruses A and B, RSV, and hMPV by PCR, and they studied paired blood samples (obtained at a 3–4-week interval) for antibody to hMPV (by ELISA) and antibody to RSV and influenza (both by complement fixation), using serologic testing. All of the patients with CAP and 81% of the patients with exacerbations of COPD were admitted to the hospital. Among the 145 patients from the study period, the incidence of hMPV infection was 4.1%, the incidence of influenza A was 6.2%, the incidence of influenza B was 0%, and the incidence of RSV infection was 9%, as determined by PCR and/or serologic testing.

Of the 6 patients in whom hMPV was identified, 4 had CAP (2 of whom also had COPD), and 2 had acute exacerbations of COPD (but not CAP). Other features of note are that all 6 patients presented to the emergency department with cough and sore throat, 5 had a mean elevated oral temperature of 38.3°C, 5 of 6 were dyspneic, 4 of 6 had nasal congestion and/or rhinorrhea, 3 of 6 had purulent sputum, 3 of 6 had wheezing, and 2 of 6 had myalgias.

One of the largest studies of patients with hMPV infection is that by Williams et al. [18] published in the New England Journal of Medicine. Nasal wash specimens were obtained over a 25-year period from otherwise healthy children who presented to the pediatric division of infectious diseases at Vanderbilt University Medical Center (Nashville, TN) with acute respiratory tract illness. The samples were studied for the presence of hMPV, which was found in 49 of 248 specimens.

It is difficult to compare the Vanderbilt cohort and the Quebec cohort, because the former consisted exclusively of otherwise healthy infants and children, whereas the latter included adults only, many of whom had comorbid illnesses. Also, the sample sizes were significantly different: the Quebec study had a total of only 145 subjects, 6 (4.2%) of whom were hMPV positive. In the Vanderbilt study, the authors concluded that 12% of all lower respiratory tract infections were most likely caused by hMPV, whereas the figure is only 4.1% in the Quebec study. The symptoms of diarrhea and vomiting were seen in 17% and 10% of the children in the Vanderbilt study, respectively, whereas these symptoms were not even mentioned in the Canadian study, and although fever occurred in 5 (83.3%) of the 6 subjects in the Quebec study, it was only found in 52% of children in the Vanderbilt study.

Hamelin et al. [2] claimed that hMPV is associated with a significant number of cases of CAP and exacerbations of COPD in adults during the early spring. I would take issue with this claim, unless we interpret the statement literally (i.e., if we consider the etiology to be specifically in the early spring [all 6 cases occurred in April]). Otherwise, it is hard to imagine that 4% is a particularly significant figure. The study, however, is an important one, and it is definitely a step in the right direction. To date, we have few data on hMPV, and anything that sheds light on its role as a pathogen is welcome. What is very important is the fact that the Quebec data show that, in the elderly population—particularly for persons with underlying lung disease—hMPV infection can lead to severe infection that requires hospitalization (mean duration, 10 days). Hamelin and colleagues' finding that ∼85% of the adult population in their study had preexisting hMPV antibody is consistent with the premise that hMPV infection is almost universal by the age of 5 years.