Abstract

Human metapneumovirus virus (hMPV) is a newly discovered respiratory pathogen with limited epidemiological data available. Cohorts of young and older adults were prospectively evaluated for hMPV infection during 2 winter seasons. Patients hospitalized for cardiopulmonary conditions during that period were also studied. Overall, 44 (4.5%) of 984 illnesses were associated with hMPV infection, and 9 (4.1%) of 217 asymptomatic subjects were infected. There was a significant difference in rates of hMPV illnesses between years 1 and 2 (7/452 [1.5%] vs. 37/532 [7.0%]; P<.0001). In the second year, 11% of hospitalized patients had evidence of hMPV infection. Infections occurred in all age groups but were most common among young adults. Frail elderly people with hMPV infection frequently sought medical attention. In conclusion, hMPV infection occurs in adults of all ages and may account for a significant portion of persons hospitalized with respiratory infections during some years

Human metapneumovirus virus (hMPV) is a newly discovered human respiratory pathogen [1]. Genetically, hMPV is similar to avian pneumovirus and has been classified in the subfamily Pneumovirinae, genus Metapneumovirus [2]. The virus was first isolated in The Netherlands from 28 young children with illnesses similar to human respiratory syncytial virus (hRSV) infections. Isolation of the virus in cell culture requires prolonged incubation, and identification is difficult when traditional techniques are used. Sequence studies of isolates have identified 2 main lineages of hMPV, which may cocirculate during the same season [3]. Recent studies from Canada and England have described patients with respiratory illnesses between ages 2 months and 87 years from whom hMPV was identified by culture and/or by reverse transcription–polymerase chain reaction (RT-PCR), which indicates that hMPV infections can occur over a wide age range [3, 4]. Given the paucity of epidemiological and clinical data available on hMPV infection in adults, we evaluated young and elderly adults for evidence of hMPV infection using RT-PCR and serological testing during 2 winter seasons

Subjects and Methods

Subjects

As part of an ongoing study of hRSV and influenza virus infection, 4 cohorts of adults were evaluated for hMPV infection in Rochester, New York, during the 1999–2001 winter seasons. Subjects included 305 healthy adults ⩾65 years old, 304 high-risk adults with underlying cardiopulmonary diseases, 195 healthy adults <40 years old, and 134 residents of a long-term-care facility (LTCF). After baseline evaluations, volunteers were evaluated during home visits for respiratory symptoms between 15 November and 15 April. Evaluations consisted of a history, physical exam, and functional evaluation and the collection of acute and convalescent serum and nasal samples for viral culture and RT-PCR. During the same period, all patients hospitalized at Rochester General Hospital for acute cardiopulmonary conditions were recruited and evaluated within 48 h of admission

hMPV IgG Enzyme Immunoassay

Antigen preparationRepresentative strains of the 2 major genotypes of hMPV (CAN97-83 and CAN98-75; kindly provided by Dr. Guy Boivin, Laval University, Quèbec, Canada) were grown on semiconfluent monolayers of LLC-MK2 cells until the cytopathic effect (CPE) was extensive (typically 7–10 days). Cultures were frozen and thawed 3 times, and the lysate was centrifuged for 15 min at 500 g to remove cellular debris. Clarified supernatant was frozen at −70°C until use

EIATo prepare EIA plates, an equal volume pool of CAN97-83 and CAN98-75 antigen and similarly prepared uninfected cells were diluted 1:20 in carbonate-bicarbonate buffer (pH 9.3), coated on Immulon II microtiter plates (Dynex Technologies), and incubated overnight at 4°C. After washing plates with PBS and 0.05% Tween 20, serum specimens diluted 1:200 in PBS, 0.5% gelatin, and 0.15% Tween 20 (PBS/G/T) supplemented with 4% normal goat serum and 1% uninfected cell lysate were added in duplicate to positive and negative antigen wells. Plates were incubated for 1.5 h at 37°C and then washed. Anti–human IgG peroxidase (Sigma) diluted 1:20,000 in PBS/G/T was added and incubated for 1 h at 37°C. Plates were washed, and tetramethyl benzidine substrate was added and incubated for 15 min at room temperature. The reaction was stopped by the addition of 2 M H3PO4, and absorbance was measured at 450 nm

The difference ( P-N ) in the mean absorbance values of 2 antigen-positive (P) and 2 antigen-negative (N) control wells was calculated for each serum specimen. All paired specimens with ratios of convalescent:acute-phase P-N>1.5 were retested by serial 2-fold dilution. A ⩾4-fold increase in IgG antibody titer was considered to be positive for recent infection. A positive serum control specimen was included in each assay, to ensure the reproducibility of the results

hMPV RT-PCR

RT-PCR for hMPV was performed on all specimens from patients with illnesses showing an antibody response and an equal number of seronegative illnesses. RT-PCR testing was performed without knowledge of serological test results. RNA extracts were prepared from 100 μL of specimen, using the automated NucliSens extraction system (bioMérieux). Primers used for RT-PCR were designed from conserved regions of the hMPV nucleoprotein (MPVN-F1-FAM [+] 5′-CAACAGGAAGCAAAGCAGAAAG-3′ and MPVN-R1 [−] 5′-CAGATTCAGGACCCATTTCTC-3′) and fusion protein (MPVF-F1-FAM [+] 5′-GAGCAAATTGAAAATCCCAGACA-3′ and MPVF-R1 [−] 5′-GAAAACTGCCGCACAACATTTAG-3′) genes. One primer for each set was 5′ end-labeled with fluorescein (6-FAM), to facilitate GeneScan analysis. All specimens were also tested for human β-actin mRNA, to control for RNA extraction and RT-PCR inhibition. One-step amplification reactions were performed using the Access RT-PCR System (Promega). Separate 50-μL reactions were prepared for each primer set by the addition of 5 μL of RNA extract to 45 μL of nuclease-free water that contained 1× reaction buffer, 0.2 mM each dNTP, 1.5 mM MgSO4, 0.1 U/μL avian myeloblastosis virus reverse transcriptase, 0.1 U/μL Tf1 DNA polymerase, and 1.0 μM each primer. Thermocycling was performed on a MicroAmp 9600 (Perkin Elmer) programmed for 48°C for 45 min for RT and 94°C for 2 min for RT denaturation; 40 cycles of 94°C for 1 min, 54°C for 1 min, and 72°C for 1 min for cDNA amplification; and 5 min at 72°C for the final extension. hMPV-positive and -negative controls containing standardized viral RNA extract and nuclease-free water, respectively, were included in each assay. Amplification products were analyzed on an ABI Prism 310 Genetic Analyzer with GeneScan software (version 3.1.2; both from Applied Biosystems). Amplification products identified within 1 nt of the expected size (195 nt for nucleoprotein gene and 347 nt for the fusion protein gene) were considered “preliminary” positives. Specimens were designated “true” positives if the same amplification product was identified on repeat testing

hMPV Isolation

Nasopharyngeal swab specimens were transported to the laboratory on wet ice and held at 4°C for up to 4 h, at which point they were cultured for hRSV and influenza viruses. Residual samples were aliquoted and placed at −20°C and later stored at −70°C. Frozen nasal secretion samples from 16 seropositive subjects who were evaluated within 5 days of symptom onset were selected for viral isolation on LLC-MK2 cells. Nine of 16 were RT-PCR positive. Cultures were observed for 16 days for the development of the CPE

Statistical Methods

Means were compared with Student’s t test, and proportions were compared using the χ2 and Fisher’s exact tests, as appropriate

Results

A total of 938 adults were prospectively enrolled, and 626 subjects were evaluated at admission to the hospital. From the 1564 subjects enrolled, 1495 illnesses were evaluated, of which 984 had paired serum samples available for hMPV testing. In addition, 217 pre- and postseason serum sample pairs from asymptomatic prospective subjects were also analyzed for hMPV infection

Overall, 42 (4.3%) of 984 illnesses were associated with a ⩾4-fold increase in hMPV antibody, and another 10 subjects had >2-fold but <4-fold increases in titers. Of the 52 samples with >2-fold increases in antibody titers, 25 (48%) were positive by RT-PCR (23/42 with ⩾4-fold and 2/10 with >2-fold increases). None of the 54 samples from seronegative patients were RT-PCR positive. Either a ⩾4-fold increase in hMPV IgG or a >2-fold increase in titer with a positive RT-PCR was considered to be definite hMPV infection. Thus, a total of 44 definite hMPVs were identified from 984 illnesses (4.5%) (table 1). In addition, 9 (4.1%) of 217 infections were identified by serological responses in pre- and postseason blood samples from subjects with no reported illnesses. hMPV could not be cultured from any of the 16 frozen nasal specimens from ill subjects identified as being hMPV infected by serologic tests or RT-PCR

Table 1

No. of subjects with human metapneumovirus virus (hMPV) infection

Table 1

No. of subjects with human metapneumovirus virus (hMPV) infection

There was a significant difference between years 1 and 2 in the rate of symptomatic hMPV infection (7/452 [1.5%] vs. 37/532 [7.0%]; P<.0001) and asymptomatic infections (1/104 [1%] vs. 8/113 [7.1%]; P=.01). In both years, the peak activity was during February (figure 1). Year 1 was dominated by influenza A and hRSV, with hMPV accounting for a low percentage of overall infections, whereas, in year 2, hMPV was more common than either influenza A or B and nearly equaled the number of hRSV infections identified (data not shown). Of the 984 illnesses evaluated, 162 had other viral infections documented by culture, RT-PCR, or serological testing. Six (3.7%) of 162 showed evidence of hMPV infection (hRSV-3, influenza A-1, or influenza B-2). Thirty-eight of (4.6%) 822 patients with no other pathogen identified tested positive for hMPV infection. hMPV illnesses identified by serological testing only were more likely to have another pathogen identified than RT-PCR–positive illnesses (5/19 vs. 1/25; P=.07)

Figure 1

Histogram of Metapneumovirus infections during the 2 winter seasons. The no. of cases identified in 2-week periods is indicated

Figure 1

Histogram of Metapneumovirus infections during the 2 winter seasons. The no. of cases identified in 2-week periods is indicated

Among the prospectively enrolled subjects, hMPV illness rates were highest among the young adults (2.9% and 9.1%) (table 1). This group also accounted for 7 of 9 asymptomatic infections. During the second year, 6 (15%) of 41 young adults with no reported illness had evidence of hMPV infection. All had regular contact with young children: most lived with children. Overall, infection rates were similar in the healthy elderly and high-risk adults, at 1.7% and 2.9%, respectively. Two (1.4%) of 143 asymptomatic elderly subjects had seroresponses, both in the second year. Seventy-two percent of the older persons had some contact with children, but few lived with children. A single hMPV infection was identified in the LTCF each winter. Among the hospitalized subjects, there was a striking difference in hMPV infection rates between years 1 and 2. During the first winter season, we identified only 2 (1.4%) infections of 142 subjects evaluated, whereas, during the second season, hMPV accounted for 18 (10.8%) of 167 illnesses in hospitalized patients (P<.0008). Because the hospitalized subjects were all high risk or elderly, the asymptomatic infection rate of 1.4% (2/143) for elderly and high-risk prospective subjects over 2 seasons was compared with the symptomatic infection rate of 20 (6.5%) of 309 in the hospitalized group and was found to be significantly lower (P=.003)

The rate of RT-PCR–positive illness varied with the group studied. Of those with ⩾4-fold antibody responses, 100% of the healthy elderly and 73% of the young subjects were RT-PCR positive, compared with 33% of the high-risk and 47% of the hospitalized patients. Overall, the RT-PCR positive rate was lower in hospitalized patients than in outpatients (47% vs. 70%; P=.01) and was associated with a significantly longer duration of symptoms before evaluation, at 6.1±5.6 for inpatients, versus 2.8±1.8 days for outpatients (P=.01)

The clinical illnesses associated with hMPV infection in adults were not distinctive (table 2). Infection was associated with high rates of cough (100%), nasal congestion (85%), dyspnea (69%), and wheezing (62%). Fever was uncommon, with only 1 subject having a temperature ⩾38° C. When the elderly groups (healthy, high-risk, and nursing home residents) were combined, the older adults experienced significantly more dyspnea and wheezing, compared with the young people, whereas the young experienced hoarseness more frequently. The impact of hMPV infection was greatest in the high-risk subjects, who were ill for nearly twice as long as the young adults (17.4±9.4 vs. 8.5±3.4 days; P=.01). In addition, the high-risk elderly subjects sought medical attention more frequently than did the young or healthy elderly subjects; 5 of 7 visited their doctors, 2 were hospitalized, and 1 subject with dual hRSV-hMPV died. Only 1 healthy older person and none of the young people visited their physicians. Because hRSV and hMPV infections appear to be indistinguishable in young children, the symptoms of the young adults were compared with those of a similar group with documented hRSV infection, to identify unique features of hMPV infection in adults (table 3). Although there did not appear to be a distinct syndrome, hoarseness was significantly more common with hMPV, compared with hRSV (91% vs. 42%; P=.01), with a trend toward a more productive cough

Table 2

Clinical illness associated with human metapneumovirus virus

Table 2

Clinical illness associated with human metapneumovirus virus

Table 3

Symptoms of young subjects with human metapneumovirus virus (hMPV) vs. human respiratory syncytial virus (hRSV) infection

Table 3

Symptoms of young subjects with human metapneumovirus virus (hMPV) vs. human respiratory syncytial virus (hRSV) infection

Hospitalized subjects with hMPV infection were mainly elderly (mean age, 75 years), and 85% had chronic heart or lung disease. The group appeared to be very similar to patients hospitalized during the same time period with acute cardiopulmonary disorders without a specific diagnosis (table 4). Most were previous or active smokers (75%), and one-fifth used long-term oral steroids and home oxygen. Although most subjects were admitted from the community, they were frail, with mild functional impairment

Table 4

Demographic characteristics of hospitalized group

Table 4

Demographic characteristics of hospitalized group

The most common admitting diagnoses for hMPV infection were chronic obstructive pulmonary disease and bronchitis (n=9) and pneumonia (n=6). Similar to outpatients, the clinical features of hMPV in hospitalized persons were not distinctive (table 5). Cough was universal, and wheezing, dyspnea, and sputum production were common, but none distinguished hMPV-infected subjects from those without this infection. Twenty-five percent had infiltrates on chest radiographs. Unilateral patchy densities in the lower lobes were noted in 3 of 5 patients, and 2 others had bilateral lower lobe infiltrates. The mean length of stay for patients with hMPV was 8.1±4.9 days. All patients received antibiotics, 85% were given steroids, and 65% were treated with bronchodilators. One subject required mechanical ventilation and survived, and 1 patient died. No patient had a bacterial pathogen isolated from blood or sputum

Table 5

Clinical features of hospitalized subjects with and without human metapneumovirus virus (hMPV) infection

Table 5

Clinical features of hospitalized subjects with and without human metapneumovirus virus (hMPV) infection

Discussion

Acute respiratory-tract infections are a major cause of disability in all age groups. In most studies, the cause of illness remains unidentified in half, even after extensive laboratory investigations [5]. The discovery of hMPV may represent a missing piece of the puzzle of respiratory infections. Our data suggest that, in some years, hMPV infection is common, although the exact proportion of illnesses due to hMPV is not yet clear

Because it has only recently been discovered, the epidemiology and clinical impact of hMPV have yet to be fully defined. hMPV infections have been documented in the Netherlands, Australia, Canada, and the United Kingdom, which suggests worldwide distribution [1–4 , 6, 7]. The present article is the first to document its occurrence in the United States. Van den Hoogen et al. [1] found that ∼10% of illnesses in children testing negative for other viruses were RT-PCR–positive for hMPV. Those authors suggested that hMPV may contribute significantly to the clinical burden of respiratory disease in children. In addition, they found 100% seroprevalence, with stable neutralizing titers, in young adults and older persons, which indicates that reinfection throughout life may be common. Our study confirms that hMPV infection occurs in young adults as well as elderly persons. Although the present study spanned only 2 winters in Rochester, New York, it appears that the virus circulates primarily during the late winter and early spring. Because year-round surveillance was not done, summer infections cannot be excluded. The variability in intensity between the 2 years differs from the predictable occurrence of hRSV infections and is more reminiscent of the periodicity of parainfluenza virus or the sporadic occurrence of influenza epidemics. The true temporal pattern of hMPV infection will need to be determined in studies performed over longer periods of time. Not unexpectedly, the highest infection rate was in young adults, which is likely explained in part by their extensive contact with children

The methods of diagnosis used in the study were serologic testing and RT-PCR. Because the seroresponse rate was similar in asymptomatic and ill subjects and those with other viral infections compared with those without a specific diagnosis, concern about the specificity of the EIA assay could be raised. However, we believe that the assay is accurate for several reasons. The distinct clustering of illnesses during certain months and the marked discrepancy between years 1 and 2 are consistent with the biology of an epidemic respiratory virus. In addition, there was good correlation between RT-PCR and serologic testing, with 50% of seroresponders having positive RT-PCR samples, and no positive RT-PCR samples were found in seronegative illnesses. Perhaps equally as important as the question of specificity of the tests used is whether hMPV infection is causally related to respiratory illness in adults. The similar rates of symptomatic and asymptomatic infection in our study raise the possibility that hMPV infection, although real, is an epiphenomenon and not the true cause of illness. Previous studies have focused only on symptomatic individuals without evidence of other viruses and may not have given an accurate picture of the spectrum of hMPV infection. However, it should be recognized that, in our study, the asymptomatic rate was based on seroconversion over a 5–6-month period, whereas the illnesses were identified with sera spanning only 4–6 weeks. Thus, it is likely that the asymptomatic hMPV infection rate for comparison to the symptomatic infection rate during the same time period was considerably lower and that most of the illnesses were causally related to hMPV infection. In addition, the finding of a significantly lower asymptomatic infection rate in the elderly subjects followed prospectively compared with the infection rate in the group hospitalized with respiratory symptoms is evidence that hMPV infection was causally related to these illnesses. However, studies using RT-PCR and serologic testing in comparably collected specimens from both persons with respiratory illnesses and nonill control groups during years when hMPV is prevalent will need to be done to fully answer this question

The clinical syndrome associated with hMPV infection in our study was not sufficiently distinctive to clearly differentiate it from other respiratory viral infections in adults [8, 9]. Similar to the first description of hMPV in children, the symptoms in adults appear similar to hRSV infection [1]. Fever was not common, which may help in distinguishing hMPV from influenza infection. Of note, asymptomatic hMPV infection among young adults was relatively common, in contrast to hRSV [10]. Similar to other respiratory viruses, the clinical impact of hMPV infection was greatest in frail elderly persons, although the infection rate was lowest. Most healthy elderly persons tolerated infection without serious sequelae, whereas those with cardiopulmonary conditions frequently required medical attention

In summary, our data confirm that hMPV infections occur in adults of all ages, both in the community and in nursing homes. The present study suggests that hMPV infection can be severe in frail elderly persons and that, during some years, hMPV may account for a significant portion of the older persons hospitalized with respiratory-tract infections. However, because asymptomatic and dual infection with other respiratory viruses is not uncommon, additional studies using sensitive and specific diagnostic tests and appropriate control groups will be needed to fully determine the clinical and economic impact of this newly described pathogen in adults

Acknowledgments

We thank Patricia Hennessey, Mary Criddle, and Anita Gellert, for patient evaluations; and Teresa Peret, Maria Formica, and Barbara Sikora, for technical assistance

References

1
van den Hoogen
BG
DeJong
JC
Groen
J
, et al.  . 
A newly discovered human pneumovirus isolated from young children with respiratory tract disease
Nat Med
 , 
2001
, vol. 
7
 (pg. 
719
-
24
)
2
van den Hoogen
BG
Besterbroer
TM
Osterhaus
AD
, et al.  . 
Analysis of the genomic sequence of a human metapneumovirus
Virology
 , 
2002
, vol. 
295
 (pg. 
119
-
32
)
3
Peret
TC
Boivin
G
Li
Y
, et al.  . 
Characterization of human metapneumoviruses isolated from patients in North America
J Infect Dis
 , 
2002
, vol. 
185
 (pg. 
1660
-
3
)
4
Stockton
J
Stephenson
I
Fleming
D
, et al.  . 
Human Metapneumovirus as a cause of community-acquired respiratory illness
Emerg Infect Dis
 , 
2002
, vol. 
8
 (pg. 
879
-
901
)
5
Garibaldi
RA
Epidemiology of community acquired respiratory tract infections in adults: incidence, etiology and impact
Am J Med
 , 
1985
, vol. 
78
 (pg. 
32
-
7
)
6
Nissen
MD
Mackay
IM
Withers
SJ
, et al.  . 
Evidence of human metapneumovirus in Australian children
Med J Aust
 , 
2002
, vol. 
176
 pg. 
188
 
7
Pelletier
G
Dery
P
Abed
Y
, et al.  . 
Respiratory tract reinfections by the new human Metapneumovirus in an immunocompromised child
Emerg Infect Dis
 , 
2002
, vol. 
8
 (pg. 
976
-
8
)
8
Falsey
AR
Cunningham
CK
Barker
WH
, et al.  . 
Respiratory syncytial virus and influenza A infections in the hospitalized elderly
J Infect Dis
 , 
1995
, vol. 
172
 (pg. 
389
-
94
)
9
Wald
TG
Miller
BA
Shult
P
, et al.  . 
Can respiratory syncytial virus and influenza A be distinguished clinically in institutionalized older persons
J Am Geriatr Soc
 , 
1995
, vol. 
43
 (pg. 
170
-
4
)
10
Hall
CB
Long
CE
Schnabel
KC
Respiratory syncytial virus infections in previously healthy working adults
Clin Infect Dis
 , 
2001
, vol. 
33
 (pg. 
792
-
6
)
Presented in part: annual meeting of the Infectious Disease Society of America, Chicago, 24–27 October 2002 (abstract 00150)
The study was approved by the Rochester General Hospital Institutional Review Board and the Centers for Disease Control and Prevention, and informed consent was obtained from all subjects or their legal guardians
Financial support: National Institute of Allergy and Infectious Disease (grant RO-1 45969)