Unraveling the Pneumonia Burden Associated With Human Metapneumovirus Infection

(See the Major Article by Howard et al on pages 108–17.) Community-acquired pneumonia (CAP) is an important cause of morbidity and mortality worldwide, disproportionately affecting children younger than 5 years of age and adults older than 65 years or those with chronic pulmonary or cardiac disease [1, 2]. Although most cases in the United States are treated as outpatients, CAP is a leading reason for hospitalization among adults and children, with associated medical costs exceeding $10 billion annually [3]. Despite the large burden of disease caused by CAP, important knowledge gaps remain regarding the microbial etiology of severe infection requiring hospitalization. Implementation of conjugate vaccine immunization programs has resulted in important reductions in the incidences of pneumococcal and Haemophilus influenzae type b CAP over the last 3 decades. In addition, the development of molecular diagnostics has increased the detection of respiratory viruses in patients hospitalized for pneumonia. As a result of these advances, viruses have become the most frequently identified pathogens in pediatric and adult CAP. Among viral pathogens, human metapneumovirus (HMPV) figures prominently as a cause of upper and lower respiratory infection. Human metapneumovirus was first identified by Dutch researchers in 2001, who described a previously unknown, nonsegmented, negative-strand, enveloped RNA virus isolated from the respiratory secretions of 28 children with acute respiratory infection [4]. Since its initial report, broader use of molecular diagnostic testing has increased identification and awareness of HMPV disease across all age groups, with younger children and the elderly found to be at greater risk of severe illness requiring hospitalization [5–7]. The clinical manifestations of HMPV resemble those of respiratory syncytial virus (RSV), the human pathogen most closely related to HMPV [8, 9]. With regard to seasonality, HMPV tends to circulate from midwinter to early spring in North America and Europe, with peak activity slightly later than that of RSV and influenza [10]. In this issue of Clinical Infectious Diseases, Howard et al [11] report on a secondary analysis of the US Centers for Disease Control and Prevention Etiology of Pneumonia in the Community (EPIC) database, a prospective, multicenter, population-based active surveillance study conducted from January 2010 through June 2012, that aimed to determine the etiology of CAP in hospitalized children and adults [12, 13]. Strengths of the EPIC study include its systematic enrollment within defined catchment areas that allows for the calculation of population-based incidence rates and the comprehensive diagnostic testing performed with multiple modalities to identify causative pathogens. Howard et al analyzed the EPIC database to determine the clinical features of HMPV pneumonia in children and adults compared with pneumonia caused by RSV, other viruses, atypical bacteria (Mycoplasma pneumoniae, Chlamydophila pneumoniae, and Legionella pneumophila), and pyogenic bacterial pathogens. The authors included 2358 pediatric patients and 2320 adult patients hospitalized for radiographically confirmed CAP, making this one of the largest HMPV-related studies published to date. Human metapneumovirus was detected in 12.6% and 3.8% of children and adults, respectively, resulting in annual HMPV CAP incidences of 1.9 hospitalizations per 10 000 children and 0.9 hospitalizations per 10 000 adults. Co-detection of other viral, atypical, and bacterial pathogens among HMPV-positive patients was common and seen in 42% of children and in 24% of adults. For the purposes of comparing clinical presentation and disease severity across etiologies, Howard et al limited their analyses to subjects with samples tested for both viruses and bacteria and excluded subjects with co-detection of other pathogen types. The main finding was that no clinical features at presentation could reliably distinguish HMPV-associated CAP from other causes of pneumonia. This highlights a major issue in the management of CAP requiring hospitalization. Considering that a bacterial pathogen was detected in only 15% of children and 14% of adults in the EPIC study, how can clinicians identify the minority of patients who have a bacterial infection needing antibiotic therapy from the majority who will not benefit from it? Finding a virus in the upper respiratory tract does not exclude the possibility of bacterial coinfection. In fact, in vitro data, animal models, and epidemiological studies suggest that viral infection, including with HMPV, facilitates secondary bacterial lung infection and predisposes to severe pneumococcal disease [14, 15]. Moreover, establishing a bacterial etiology in CAP remains extremely difficult because current standard bacteriology testing is insensitive (most cases cannot be confirmed by sputum or blood culture, especially if the patient already began antibiotic treatment) and has poor specificity (sputum cultures and urine antigen tests may represent nasopharyngeal colonization) [16, 17]. Therefore, it is not surprising that the identification of a virus does not consistently reduce physician antibiotic prescribing among children [18] or adults [17] admitted for acute respiratory infection. Moreover, given the difficulty in determining if CAP is due solely to a virus at the time of presentation, current CAP clinical practice guidelines recommend initially treating all hospitalized patients empirically for bacterial infection or coinfection [19, 20]. An epidemiologic approach used in other studies to address the imperfect specificity of the identification of a viral pathogen in the nasopharynx when trying to determine pulmonary disease etiology is the use of a case-control design. By comparing CAP cases with healthy controls, researchers can calculate the virus-specific attributable fraction, which estimates the proportion of patients with CAP who are positive for a virus and whose illness is actually due to that virus [21]. In another EPIC substudy, Self et al [21] found that detection of HMPV was rare in asymptomatic controls compared with detections in patients with CAP in the same communities and during the same time period. Observed HMPV-specific attributable fractions were 0.90 and 0.93 in children and adults, respectively. Moreover, when HMPV was detected in controls, polymerase chain reaction cycle threshold values were higher, suggesting lower viral loads, than in patients with CAP. The Pneumonia Etiology Research for Child Health (PERCH) study—a case-control study of radiographically confirmed CAP requiring hospitalization in 4232 children without HIV infection between 1 month and 5 years of age who were hospitalized with severe or very severe pneumonia in 7 African and Asian countries during 2011 to 2013—developed a partial latent-class Bayesian analytic approach to estimate pneumonia etiology at the population and individual levels [22]. Again, identification of HMPV was strongly associated with CAP (odds ratio, 6.3; 95% confidence interval [CI], 4.78–8.20%). PERCH investigators determined that HMPV thus accounted for 7.5% (95% CI, 5.9–9.5%) of hospitalized CAP in this population. Howard et al also found that, compared with HMPV pneumonia, children and adults with pyogenic bacterial pneumonia were approximately 3 times as likely to have more severe disease. However, given currently available diagnostic methods, bacteria are more readily identifiable in CAP cases with complicated disease (eg, patients presenting with bacteremia or empyema requiring drainage) or in patients with disease severity that warrants more invasive procedures (eg, bronchoalveolar lavage or endotracheal aspirate). Thus, detection of laboratory-confirmed bacterial pneumonia is inherently biased toward more severe cases. This notion is supported by the results of a 2007–2017 study in which one-quarter of patients admitted to a Spanish intensive care unit with hMPV-associated CAP had a documented Streptococcus pneumoniae coinfection. Further progress in our understanding of CAP etiology will be important to help guide policy and future treatment recommendations. The results of the current study by Howard et al highlight the need for HMPV-specific preventive and therapeutic strategies to reduce the considerable HMPV disease burden in children and adults. Financial support. J. P. is supported by a Chercheur-boursier clinicien career award from the Fonds de Recherche Québec Santé. Potential conflicts of interest. J. P. received consulting/speaker fees/honoraria from AbbVie, Seegene, BD Diagnostics, and Cepheid and research grant funding outside of the current work from Janssen Pharmaceuticals, AbbVie, BD Diagnostics, MedImmune, Seegene, and Sanofi Pasteur.