Published ahead of print on July 24, 2008, doi:10.1164/rccm.200711-1657OC
© 2008 American Thoracic Society doi: 10.1164/rccm.200711-1657OC
Human Metapneumovirus in Lung Transplant Recipients and Comparison to Respiratory Syncytial Virus1 Queensland Heart-Lung Transplant Unit, The Prince Charles Hospital, Brisbane, Australia; 2 Queensland Paediatric Infectious Diseases Laboratory, Royal Children's Hospital, Brisbane, Australia; and 3 Clinical Medical Virology Centre, University of Queensland, Brisbane, Australia Correspondence and requests for reprints should be addressed to Dr. Peter Hopkins, M.D., Department of Respiratory Medicine, The Prince Charles Hospital, Rode Road, Chermside, Brisbane, Australia 4032. E-mail: peterwakatipu{at}hotmail.com
Rationale: Human metapneumovirus is a newly described virus isolated in 2001 from children with acute respiratory viral infection. It has subsequently been reported globally, although there are limited data in lung transplant recipients. Objectives: To prospectively analyze whether human metapneumovirus was circulating in our adult lung transplant community and assess the morbidity of this infection and to compare the clinical presentation and outcome after intravenous ribavirin of human metapneumovirus with that of respiratory syncytial virus (RSV). Methods: Lung transplant patients with clinical features of respiratory viral infection underwent nasopharyngeal aspirates. Patients with a positive specimen for RSV or human metapneumovirus by reverse transcriptase–polymerase chain reaction analysis and graft dysfunction received intravenous ribavirin and pulse steroid therapy. Measurements and Main Results: Eighty-nine patients had 199 visits for aspirate studies. A viral cause was determined for 62 visits in 47 patients (19 human metapneumovirus, 18 RSV, 13 parainfluenza, 9 influenza A, 2 adenovirus, and 1 influenza B). A significant percentage of patients with metapneumovirus (63%) and RSV (72%) developed graft dysfunction, with average declines in FEV1 of 30 ± 12.4% and 25.9 ± 11.2%, respectively. In these patients, bronchiolitis obliterans syndrome onset or progression occurred in no patients with human metapneumovirus compared with 5 of 13 (38%) patients with RSV at 6 months. Conclusions: Human metapneumovirus is a leading cause of acute respiratory tract illness in lung transplant recipients. The incidence and clinical spectrum at presentation are similar to RSV, although the latter seems to be associated with a higher risk of chronic rejection. We recommend testing of nasopharyngeal aspirates for human metapneumovirus with polymerase chain reaction to assess local epidemiologic patterns.
Key Words: respiratory • virus • solid organ transplant
Respiratory viral infections (RVIs) are common in immunocompromised patients and have been associated with significant morbidity and mortality approaching 20% (1, 2). Lung transplant recipients have a unique predisposition and response to infection because of diminished cough reflex, abnormal lymphatic drainage, impaired mucociliary clearance, and preexisting airways damage with obliterative bronchiolitis (3). Clinical manifestations may range from acute self-limiting pharyngitis to the more severe spectrum of bronchiolitis, including viral pneumonitis and acute lung injury with respiratory failure. Secondary bacterial infection is well recognized, along with a predisposition to acute allograft rejection and bronchiolitis obliterans syndrome (BOS) through local immune up-regulation. Early diagnosis is essential to direct therapy, identify epidemic trends in the local transplant community, and prevent nosocomial acquisition of infection. However, in a substantial proportion of lung transplant recipients with acute respiratory tract infections, no virus is isolated. A combined analysis of studies by Hopkins and colleagues and Garbino and colleagues suggests that 42 to 63% of such illnesses in transplant recipients have no cause ascribed (4–6). This evidence suggests the presence of unidentified pathogens and is consistent with the pediatric literature on the epidemiology of RVI. In 2001, researchers in the Netherlands isolated a new virus from children with acute RVI (7). During a 20-year period, this pathogen was isolated from 28 infants and children with respiratory tract disease. Biochemical and genetic evidence suggested that this new agent was the first human pathogen in the Metapneumovirus genus. This new human pathogen was named human metapneumovirus (hMPV) and was classified as belonging to the family of Paramyxoviridae. The hMPV is characterized as a nonsegmented, enveloped, negative stranded RNA virus with a plus sense genome approximately 30 kb in length. The closest genetic relative is avian metapneumovirus, and the closest clinical relative respiratory syncytial virus (RSV). In 2002, hMPV infection was reported in three Australian children, and there was evidence to support worldwide distribution of the virus (9). Although the precise epidemiology remains poorly described, upward of 20% of RVIs which are negative on standard viral detection panels may be attributable to hMPV (8, 10). The aims of this study were as follows: (1) to prospectively analyze whether hMPV was circulating in our adult lung transplant community, (2) to characterize the potential short- and long-term morbidity and mortality of this infection, to (3) assess the impact of RSV treatment protocols in recovery from illness, and (4) to compare the clinical presentation and outcome with RSV.
Patient Population During a 42-month prospective study period from July 2003 until December 2006, lung transplant patients presenting with a clinical syndrome consistent with RVI at the Prince Charles Hospital underwent nasopharyngeal aspirates (NPAs). We targeted patients with symptoms of influenza-like illness, defined as any combination of sore throat, nasal irritation, low-grade fever, myalgia, and arthralgia with or without lower respiratory tract symptoms of cough, dyspnea, or wheeze. All patients post-transplant were encouraged to telephone the transplant physician on call with new symptoms and to present to medical outpatients promptly for medical review. Selected patients with significant lower respiratory tract (LRT) symptoms, radiographic changes (reticulonodular opacities and evidence of hyperinflation), or acute allograft dysfunction on spirometry proceeded to early fiberoptic bronchoscopy for transbronchial lung biopsy and bronchoalveolar lavage (BAL). Respiratory NPA specimens were obtained in the outpatient department and screened by indirect fluorescent antibody test and polymerase chain reaction (PCR) for common respiratory viruses, including RSV, influenza A and B, parainfluenza 1–3, and adenovirus. Samples were routinely screened for hMPV by reverse transcriptase (RT)–PCR analysis. NPAs were performed by physiotherapists specifically trained in this technique to ensure specimen adequacy, and results were available within 24 hours. Baseline FEV1 was defined as the average of two measurements recorded in the 4 months preceding clinical presentation separated by at least 4 weeks. Acute bronchiolitis was defined clinically as an acute respiratory illness characterized by diffuse wheeze or inspiratory squeaks with or without peribronchial thickening and hyperinflation on the plain chest radiograph. BOS was defined according to standard International Society for Heart and Lung Transplantation (ISHLT) criteria (11). Study participants received similar immunosuppressive therapy early post-transplantation consisting of oral cyclosporine, prednisone (maintenance dose, 0.1–0.2 mg/kg/d), and a nucleotide blocking agent of azathioprine (1–2 mg/kg/d) or mycophenolate mofetil (trough level target 2.0–4.5 mg/L). Acute rejection was treated with intravenous methylprednisolone 15 mg/kg/day for 3 days and followed by an oral prednisolone taper starting at 1 mg/kg. Patients were switched from cyclosporine to tacrolimus (trough target level 5–15 µg/L) for recurrent or persistent rejection. Upon confirmed detection of hMPV, patients were hospitalized, and a treatment protocol was initiated if there was a decline in FEV1 of 10% or more from baseline. Hospitalized patients were nursed in single-room isolation under infection control precautions for droplet transmission. The mainstay of treatment consisted of intravenous ribavirin (1-B-D-ribofuranosyl-1,2,4-triazole-3-carboxamide) (ICN Pharmaceuticals, Costa Mesa, CA) at a starting dose of 33 mg/kg/day for the first 24 hours, then 20 mg/kg/day thereafter. Duration of therapy was determined by resolution of clinical symptoms and sustained improvements in respiratory function. Patients with hMPV received 200 mg intravenous methylprednisolone once daily for 3 days; broad-spectrum, β-lactam–based antibiotic therapy, and nebulized bronchodilators. Pulse methylprednisolone was followed by oral prednisolone at a dosage of 1 mg/kg, tapering by 5 mg/day to baseline. This hMPV treatment protocol was identical to our RSV regimens. Patients with self-limiting upper respiratory tract symptoms were not hospitalized or treated but were followed within 48 hours of diagnosis and at Day 7 to ensure no development of graft dysfunction. Infected patients were instructed to keep self-recorded diaries of portable spirometric values and report a decline in lung function greater than 10%.
Procedure for NPA Specimens—Molecular Analysis
Statistics
The referral population of transplant subjects was community based with no cases of nosocomial acquisition during the study period. Eighty-nine patients had 199 visits for NPA studies, and 20 (22.5%) underwent further bronchoscopic evaluation. A viral cause was determined for 62 visits in 47 patients (19 hMPV, 18 RSV, 13 parainfluenza, 9 influenza A, 2 adenovirus, and 1 influenza B). There were no episodes in which more than one viral pathogen was simultaneously identified. There were 137 samples (68%) with no viral cause identified on extended testing. Therefore, 19 of 156 samples with no virus detected on standard viral panel of RSV, adenovirus, parainfluenza, and influenza tested positive for hMPV (12.2%). The most frequently detected virus was hMPV, with an occurrence rate of 30.6% among those with a confirmed viral pathogen. The characteristics of lung transplant recipients with hMPV and RSV infection are shown in Table 1. There were no differences in age, sex, type of transplant, immunosuppression, or prior diagnosis of BOS grade between the hMPV and the RSV groups.
Infection with hMPV peaked in August to October, with 12 of 19 cases occurring during this period. In contrast, RSV infection occurred earlier in the reporting season, with 11 of 18 cases occurring from June to August. There was no significant variation in occurrence of viral infection each year during the evaluation period. Patients with RVI ranged in age from 29 to 65 years and were diagnosed 57 to 4,482 days post-transplant. No patient received augmented immunosuppression or antirejection treatment in the month preceding the diagnosis. Seven (37%) hMPV-positive subjects suffered only mild self-limiting upper respiratory tract symptoms and were not hospitalized. The remaining 12 hMPV-positive patients experienced significant morbidity from the infection with LRT symptoms and acute decline in allograft function requiring hospital admission. Physical manifestations of hMPV infection in these subjects resembled RSV, although intense bronchospasm typified by high-pitched inspiratory squeaks or polyphonic wheeze was less common (P = 0.03). Fever and abnormal chest radiographs (focal infiltrates, peribronchial cuffing) indicative of bronchiolitis were occasionally seen with hMPV and RSV. The mean decrease in absolute volume and percentage of baseline FEV1 was similar for hMPV and RSV. Transbronchial lung biopsies in eight patients with LRT involvement (RSV, n = 3; hMPV, n = 5) revealed intense airway-centered inflammation with a mixed infiltrate of neutrophils, lymphocytes, and plasma cells. Other histologic features included aggregates of foamy macrophages and plugs of organizing pneumonia which is indicative of more proximal bronchiolar obstruction (Figure 1). One procedure confirmed moderate acute allograft rejection with associated lymphocytic bronchiolitis grade B2 (15).
Treatment for subjects with LRT involvement secondary to RSV and hMPV consisted of intravenous ribavirin generally for at least 10 days (range, 5–28 days). After intravenous methylprednisolone, patients received antiinflammatory doses of oral prednisone commencing at 1 mg/kg and tapering by 5 mg/day to their usual maintenance dose. Routine bacterial cultures of sputum or BAL were positive in 91.5% of patients with hMPV and in 84.5% of patients with RSV with graft dysfunction. The most common pathogen was Pseudomonas aeruginosa at 77% of LRT samples, followed by other gram-negative pathogens, including Stenotrophomonas, Achromobacter, and Haemophilus. Given the difficulties in distinguishing colonization from true infection, all hospitalized patients received concurrent broad-spectrum antibiotics. Coexisting isolation with fungal pathogens was ascertained in four patients with hMPV and three patients with RSV on culture. Approximately 10% of patients (n = 2 in both viral groups) developed probable aspergillosis in the first month after viral infection as per the Mycosis Study Group definition of invasive fungal infection (16). This demanded the institution of aggressive combination antifungal therapy. Standard infection control practices were implemented for all hospitalized patients, with RVI involving isolation with droplet precautions.
Table 2 presents the evaluation of outcomes for patients with hMPV and RSV with graft dysfunction in the 6 months after diagnosis. Outcome was favorable for all patients with hMPV, with no cases of progressive respiratory failure requiring intubation or assisted ventilation. One patient did not return to within 10% of baseline but did not evolve to diagnostic criteria (
We report the largest series of hMPV infection in a lung transplant population presenting with acute respiratory tract infection that was negative on standard viral panel testing. A prospective evaluation was performed over two reporting seasons and confirmed an isolation rate of approximately 10% among all respiratory NPA samples, making hMPV infection as common as RSV. This is comparable to epidemiologic studies from the Netherlands, where 7 of 68 samples (10.3%) from a pediatric cohort over a single season were positive for hMPV (7). A series from the United States in children younger than 5 years has determined a detection rate of 6.4% over a 4-month evaluation period (10). In addition, a collaborative study with four provincial public health laboratories in Canada in all age groups has reported an isolation rate of 14.8% during the 2001–2002 reporting season (17). The age range of patients with hMPV infection was 2 months to 93 years (median age, 25 years). hMPV is a ubiquitous pathogen of global distribution (18–20), with an isolation rate in our lung transplant population suggesting parallels to that of the general community. In our series, infection occurred throughout the entire observation period (although peaking in winter and spring), a distribution consistently described in other studies (7, 10, 17). Over 60% of suspected RVI remain undiagnosed in the pediatric and lung transplant community. A study by the University of Colorado Health Sciences Center evaluated 116 nasal washes and BAL samples in 72 lung transplant recipients with upper or lower respiratory tract infections. Only 31 specimens (27%) yielded a recognizable respiratory virus (21). A single-season prospective study of RVIs by Milstone and colleagues reported that only 34% of 49 symptomatic lung translant recipients had detectable virus on PCR testing for RSV; flu A and B; PIV 1, 2, and 3; and adenovirus (22). This evidence suggests that unidentified pathogens account for the majority of clinical presentations with suspected RVI. Researchers have previously postulated that patients with negative diagnostic tests had common respiratory viruses, such as rhinovirus and coronavirus (23). In addition, atypical pathogens, such as Chlamydia pneumonia and Mycoplasma pneumonia, may mimic RVI. Our large series confirms that hMPV was responsible for 19 cases of 156 NPA samples negative for the standard viral panel as detailed above. Several additional factors may explain our relatively high prevalence of infection, such as seasonal and geographic variation along with heightened efforts to diagnose infection. Our patients are encouraged through ongoing education to present early after the onset of influenza-like symptoms for medical assessment. The possibility of viral isolation is enhanced when specimens are collected within 7 days of clinical onset of symptoms (4). hMPV is an emerging respiratory pathogen and should be included in the standard viral panel of nasopharyngeal specimens for lung transplant recipients. hMPV and RSV produced significant clinical infection in lung transplant patients, necessitating hospitalization, single bed isolation, and supportive care in the majority of affected individuals. Radiographic anomalies were not predictive of more severe graft dysfunction or of the development of respiratory failure. Only one patient with RSV required ventilatory support and died as a result of acute respiratory failure. A significant minority of patients with RSV and hMPV experienced self-limiting upper respiratory tract symptoms only, resembling the common cold. Histologic assessment of transbronchial lung tissue in those requiring biopsy generally disclosed acute airway inflammation with a mixed cellular infiltrate and organizing pneumonia. Ribavirin was considered the most critical component of the treatment strategy for RSV and hMPV. However, we lack an understanding of the natural history of these viruses—especially hMPV—in the absence of treatment. Ribavirin is a broad-spectrum nucleoside analog with reported antiviral activity against RSV and hMPV when evaluated in vitro with LLC-MK2 cells (24). Hamelin and colleagues (25) have reported that early ribavirin treatment for 4 days starting 12 hours postinfection drastically reduced hMPV replication in lungs of mice with a parallel decrease in pulmonary inflammation. Glanville and colleagues have reported on 18 lung transplant patients with RSV and acute bronchiolitis who received intravenous ribavirin (26). All but one patient recovered lung function and returned to baseline BOS grade when followed for 3 months. Mean duration of ribavirin therapy was 8 days, whereas our study averaged 10 days for hMPV and 11.6 days for RSV.
The rationale for pulse corticosteroid therapy was suppressing the innate immune response to viral infection and the consequent lung inflammatory reaction. Acute RSV bronchiolitis is characterized by infection of respiratory epithelial cells with the release of neutrophil and eosinophil chemoattractants and chemokines, including macrophage inflammatory protein-1 Our RSV treatment protocol specifies the intravenous route of administration for ribavirin. However, in Australia, ribavirin is only approved for clinical use via the inhaled route for severe lower respiratory tract RSV infection in hospitalized children less than 2 years of age. Aerosolized ribavirin has been recommended in allogeneic stem cell and lung transplant recipients with paramyxovirus infections. Nonetheless, a number of limitations have been associated with aerosol delivery, and we deemed this not suitable for our transplant patient group. First, aerosolized ribavirin is resource intensive, requiring a small particle aerosol generator, an isolation tent or a respiratory circuit, isolation rooms with negative pressure flow, and continuous drug administration for 16 of 24 hours (28). Worsening of respiratory status may occur in patients requiring assisted ventilation, precluding the successful use of aerosolized ribavirin in patients with acute respiratory failure. Finally, McCurdy and colleagues (29) have reported that one-third of lung transplant recipients with lower respiratory paramyxoviral infection treated with aerosol ribavirin died or failed to return to baseline FEV1. It has been proposed that ribavirin therapy should continue until nasopharyngeal swabs are negative on IFA or PCR technique. PCR assays are increasingly used in the diagnosis of community-acquired RVIs. The clinical relevance of the detected nucleic acid sequence is not always clear because it may not indicate active replicating virus (30). For example, respiratory viruses may persist in the nasopharyngeal tract after symptom resolution. Noninfective viral nucleic acids may be detected for months after an RVI, as has been demonstrated with rhinoviruses (31). Mice experimentally infected with RSV were able to clear the virus from BAL within 2 weeks, but viral RNA could be detected in lung homogenates for 100 days or more (32). With the potential for the persistence of nucleic acids from nonviable or nonreplicating virus causing a false-positive result on repeat NPA, we determined the duration of therapy according to clinical response. Regardless, patients with RVIs secondary to RSV or hMPV generally require 3 to 4 weeks to return to baseline lung function, and this is achieved invariably after discharge from hospital. An alternative diagnostic test is traditional cell culture, although such techniques are time consuming and resource intensive. After the first 12 months of our study, routine cell culture for hMPV was ceased due to the low positive rate compared with PCR technique. Only two of eight patients (25%) with hMPV on PCR had a corresponding positive cell culture with no cases detected on culture and not PCR. Our low rate of positive viral culture is consistent with the literature on RVIs, with reported sensitivities in detection ranging from 22 to 45% (3, 4). Poor viral culture results may reflect in vitro viral instability or suboptimal specimen handling. Although all samples were sent to an outside reference laboratory, inoculation into cell culture and PCR assay were achieved within a few hours. No sample was frozen in transport, which may otherwise promote cellular degradation and reduce viral stability (4). Retrospective clinical reviews support an association between RVIs and the development of BOS (33). Seasonal clustering of the onset of BOS suggests that infectious triggers such as respiratory viruses may be involved in the natural history (34). Obliterative bronchiolitis has been described shortly after influenza pneumonia in pediatric and adult lung transplant recipients (35, 36). Experimentally, the intratracheal instillation of parainfluenza 1 in rat lung allografts induces the typical lesion of obliterative bronchiolitis within 56 days on histologic section (37). Palmer described four of eight patients developing BOS immediately after RVI or within several months of infection (33). Kumar and colleagues (38) reported 6 of 50 patients developing BOS over a 12-month period after the diagnosis of community-acquired RVI. In a small series of nine patients with hMPV, none of whom received ribavirin, Larcher and colleagues (30) described two patients developing rapid-onset obliterative bronchiolitis who died several months later. In our study, patients with hMPV experienced a reversible though significant decline in lung function with no progression to BOS or worsening of BOS grade on follow-up. In contrast, 5 of 13 patients with RSV with graft dysfunction developed progressive BOS, with two deaths occurring at 6-months follow-up. Both patients with RSV who died were BOS grade 3 at diagnosis, which suggests that patients with severe preexisting bronchiolar epithelial damage have a limited repertoire of defense mechanisms to this virulent virus. The remaining three patients with RSV had previously returned to baseline lung function but developed BOS within 6 months, posing the question of virus induced up-regulation of alloimmune cell reactivation. Overall, patients with RSV were more likely than patients with hMPV to develop BOS despite similarities in clinical presentation and treatment regimens. Inspiratory squeaks indicative of small airways dysfunction were more common with RSV at presentation, suggesting a more intense bronchiolitis. Of the five patients with RSV who developed progressive or new-onset BOS, only one patient had delayed clinical presentation before diagnosis. Nonetheless, we believe early diagnosis and management are essential to prevent severe airway epithelial injury and subsequent BOS. In conclusion, hMPV is a leading cause of acute respiratory tract illness in lung transplant recipients, accounting for approximately one in three viral infections that can be identified. The incidence and clinical spectrum at presentation are similar to RSV, ranging from mild, self-limiting upper respiratory tract symptoms to severe bronchiolitis. Therapy with intravenous ribavirin and high-dose corticosteroid therapy may reduce the subsequent risk of obliterative bronchiolitis with hMPV and, to a lesser extent, RSV infection. However, one must caution against making strong treatment recommendations given the limitations from this study of lack of a comparative control group and small numbers of patients. Nonetheless, we believe there is a strong argument for formally evaluating the role of ribavirin in the context of a randomized, controlled trial. Overall, RSV seems to be associated with a higher risk of BOS despite early clinical presentation and aggressive treatment protocols. We recommend testing of NPA samples for hMPV PCR to assess local epidemiologic patterns and facilitate early identification.
Originally Published in Press as DOI: 10.1164/rccm.200711-1657OC on July 24, 2008 Conflict of Interest Statement: None of the authors has a financial relationship with a commercial entity that has an interest in the subject of this manuscript. Received in original form November 8, 2007; accepted in final form July 22, 2008
This article has been cited by other articles:
|
|||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
TOP
ABSTRACT
AT A GLANCE COMMENTARY
20% decline) for the development of BOS. A return to baseline pulmonary function occurred at a mean of 29.5 days (range, 7–44 d) for hMPV and 22.2 days (
and IL-8. Corticosteroid therapy in the form of hydrocortisone (200–1,000 ng/ml) attenuates production of these factors in PIV-infected cells but has minimal effect on RSV-infected cells (
