Lower Respiratory Viral Illnesses Improved Diagnosis by Molecular Methods and Clinical Impact

Transcription

1 Lower Respiratory Viral Illnesses Improved Diagnosis by Molecular Methods and Clinical Impact Jorge Garbino, Margaret W. Gerbase, Werner Wunderli, Christelle Deffernez, Yves Thomas, Thierry Rochat, Beatrice Ninet, Jacques Schrenzel, Sabine Yerly, Luc Perrin, Paola M. Soccal, Laurent Nicod, and Laurent Kaiser Division of Infectious Diseases; Central Laboratory of Virology, Division of Infectious Diseases; Division of Pulmonary Medicine; Central Laboratory of Bacteriology, Division of Infectious Diseases; and Clinic of Thoracic Surgery, University Hospitals of Geneva, Geneva, Switzerland We assessed the frequency and the potential role of respiratory viruses on disease outcomes in hospitalized patients and lung transplant recipients who underwent a bronchoalveolar lavage (BAL) for an acute respiratory infection. BAL specimens (148) were analyzed by reverse transcription-polymerase chain reaction for the presence of 11 different viruses, as well as Mycoplasma pneumoniae, Chlamydophila pneumoniae, and Legionella pneumophila. Respiratory viruses were identified in 34 of 117 BAL specimens (29%) obtained in patients with a suspected respiratory infection and in only 2 of 31 control subjects (7%) (p 0.01). M. pneumoniae was identified in five additional cases. Only 30% of cases that were virus positive by molecular methods were also positive by cell culture analysis. Rhinovirus was the most frequently identified virus (56% of cases) followed by respiratory syncytial virus (27%). In lung transplant recipients, the rate of viral infections was 55% in cases with respiratory symptoms compared with only 4% in control subjects (p 0.001). In these cases, respiratory viral infections were associated with significant lung function abnormalities. By using reverse transcription-polymerase chain reaction assays, we frequently identified respiratory viruses in BAL specimens of patients hospitalized with lower respiratory tract infections. These viruses were associated with high morbidity, particularly in lung transplant recipients. Keywords: bronchoalveolar lavage; lung transplantation; respiratory syncytial virus; respiratory viruses; rhinovirus Lower respiratory tract infections (LRTIs) are a leading cause of morbidity, hospitalization, and antibiotic use in patients with immunosuppression and/or chronic lung diseases; however, the etiology remains undetermined in 48 to 70% of cases (1 3). This lack of diagnosis exists because appropriate lower respiratory tract samples are often not available and routine diagnostic procedures are limited both in their sensitivity and in the number of agents routinely targeted. Among undiagnosed agents, respiratory viruses are thought to contribute to a substantial number of LRTIs in hospitalized patients, especially when such viruses are circulating in the community. Hence, there is a need to establish the precise impact of respiratory viruses in patients hospitalized with an LRTI. Although nucleic acid detection by reverse transcription-polymerase chain reaction (RT-PCR) has considerably improved our ability to diagnose these infections (4, 5), the use of this tool is limited by the large number of viral types and subtypes and by the difficulty to collect appropriate lower respiratory specimens. Therefore, few studies have at- (Received in original form June 21, 2004; accepted in final form September 4, 2004) Supported by a grant from the University Hospitals of Geneva. Correspondence and requests for reprints should be addressed to Laurent Kaiser, M.D., Department of Medicine, Division of Infectious Diseases, Central Laboratory of Virology, University Hospitals of Geneva, 24, Rue Micheli-du-Crest, 1211 Geneva 14, Switzerland. laurent.kaiser@hcuge.ch Am J Respir Crit Care Med Vol 170. pp , 2004 Originally Published in Press as DOI: /rccm OC on September 10, 2004 Internet address: tempted to evaluate the role of respiratory viruses by systematically using sensitive RT-PCR assays on appropriate samples such as bronchoalveolar lavage (BAL) specimens. In particular, we lack information on the potential role of viruses that are difficult to grow, such as rhinovirus, human coronavirus, and human metapneumovirus. The detection of viral RNA in a respiratory sample suggests a viral contribution to respiratory symptomatology and pathology. This issue is of particular importance in immunocompromised patients who may be infected with several agents and who are prone to bacterial superinfections. In this study, we addressed this question by correlating clinical findings with the presence or absence of a proven viral infection. In addition to respiratory viruses, Mycoplasma pneumoniae, Chlamydophila pneumoniae, and Legionella pneumophila also cause community outbreaks, share similar clinical features, and present diagnostic difficulties. Their impact in immunocompromised hosts likewise needs to be established in large studies. The purpose of this study was to employ molecular methods to detect and identify 11 different respiratory viruses and three atypical agents in patients from whom a BAL and a viral culture were performed after an acute respiratory event. Our goal was to assess the incidence and the potential impact of respiratory viral infections on disease outcome in a patient population that was at high risk of complications. This knowledge should enable a more rational use of new antivirals, antibiotics, and vaccines when available. Portions of the results of this study have been previously reported in the form of an abstract (6). METHODS Patients and BAL Specimens We selected specimens over a 1-year period ( ) to cover all four seasons, as respiratory viruses have a marked seasonal distribution that is specific for each virus type. During the study period, 678 BAL procedures were performed in our institution. In 348 procedures, viral cultures were performed, based on the decision by the physician in charge, and independent of this study. Among these 348 samples, we excluded samples obtained from the same patient that were collected less than 1 week apart, as well as the few samples that were inadequate to conduct all the assays (26 of 348 [8%]). Of the remainder, we selected every other sample, a total of 148 specimens. Although some patients had multiple BAL procedures during the study period, each sample and episode were analyzed independently. Samples were subsequently analyzed for the presence of respiratory viruses and atypical bacteria by technicians blinded to any previous microbiological results. Based on the clinical chart review with microbiological results still blinded, patient BAL specimens were distributed into two groups according to the likelihood of a respiratory infection. Group 1 consisted of BAL samples obtained from patients with any clinical suspicion of respiratory infection. Patients included in this group had at least one of the following acute respiratory symptoms: rhinorrhea, cough, sputum, dyspnea, unexplained lung function decline, or new chest X-ray abnormalities. This group also included immunocompromised patients presenting with persistent fever. Group 2 included BAL samples from

2 1198 AMERICAN JOURNAL OF RESPIRATORY AND CRITICAL CARE MEDICINE VOL patients who underwent fiberoptic bronchoscopies for reasons other than suspicion of a respiratory infection (i.e., routine follow-up of lung transplant recipients, investigation of noninfectious interstitial lung diseases, or pulmonary nodules) and were considered as the control group. Because lung transplant recipients constituted a large and well-defined subgroup of patients enrolled in this study, we performed an additional analysis of this subgroup. According to the previously mentioned criteria, BAL specimens from lung transplant recipients were further distributed into subgroups 1 and 2. Subjects were considered immunocompromised if at least one of the following conditions was present: hematopoietic stem cell transplantation, solid organ transplantation, hematologic malignancies, chemotherapy for cancer, autoimmune or collagen diseases, long-term use of 10 mg or more of prednisone per day, any other immunosuppressive treatment, human immunodeficiency virus infection, liver cirrhosis, or diabetes. The presence of one of the following diseases was considered an additional comorbidity: cardiovascular disease, renal insufficiency, alcoholism, chronic respiratory disease, or solid cancer. Data Collection A case report form recorded patient characteristics, underlying disease, comorbidities, immunosuppressive therapy, survival at 1 month after the infectious episode, and radiologic findings. Report forms for each episode were completed in parallel by investigators blinded to microbiological investigations. In all lung transplant recipients, the results of sequential spirometries and transbronchial biopsies performed as part of the clinical follow-up were also recorded and analyzed in both groups. The ethics committee of the University Hospitals of Geneva approved the study. Microbiological Sampling and Analysis BAL procedures were performed following a standardized protocol. Accordingly, 30 to 50 ml of sterile saline solution were instilled three times into the distal bronchial tree, either at the site of the radiographic abnormality or in the middle lobes. All BAL specimens were separated into aliquots and processed similarly for subsequent analysis. Gram stain, acridine orange, auramine, and Giemsa colorations were performed for direct identification of bacteria, mycobacteria, fungi, and parasites. Cultures for bacterial identification were inoculated under standard aerobic conditions on four different media as well as on specific media for mycobacterium detection. Pulmonary bacterial infection was considered only when quantitative BAL sample results were 10 4 cfu/ ml or more for the pathogen in association with clinical symptoms; otherwise, it was not considered an infection. For virus detection, four different cell culture lines (human embryonic fibroblasts, A549, Medin-Darby canine kidney [MDCK], and LLC- MK2 cells) were inoculated in tubes at two different temperatures (37 C and 33 C). Sample Processing Qualitative RT-PCR assays were performed on all specimens for the following 10 different respiratory viruses: influenza A and B, respiratory syncytial virus (RSV) A and respiratory syncytial virus B, parainfluenza 1 and 3, human rhinovirus, human metapneumovirus, and coronaviruses OC43 and E229. We also completed this series by determining the presence of the recently discovered human coronavirus NL63 in a subgroup of 120 specimens for which we had stored DNA remaining. Adenovirus was detected by cell culture only. These 120 cases were equally distributed across all groups and subgroups, representing 78%, 81%, 87%, and 78% of cases in Group 1, Group 2, Subgroup 1, and Subgroup 2, respectively. We also performed PCR assays for the detection of M. pneumoniae, C. pneumoniae, and L. pneumophila. After collection, approximately 5 ml of BAL specimen were immediately placed in 2 ml of viral transport media. Crude BAL mixed in the viral transport media conserved at 4 C were used to inoculate cells within 4 hours, and 2 ml were aliquoted and frozen at 80 C. The frozen aliquot was untouched and thawed only to conduct the RT-PCR/PCR assays. PCR analysis for the human coronavirus NL63 was performed during a second time period using stored cdna. Viral RNA Extraction RNA was extracted from 200 l of each specimen by using 400 l of lysis buffer (HCV Amplicor Specimen Preparation Kit; Roche Diagnostics Corporation, Indianapolis, IN) followed by incubation at room temperature for 10 minutes. Propanol (650 l) was added, and the tubes were centrifuged at 5,000 g for 15 minutes. The resulting pellet was washed with 1 ml of ethanol (70%) followed by another centrifugation at 5,000 g for 5 minutes. The ethanol was then removed, and the pellet was dried at room temperature for 15 minutes. RNA was dissolved in 30 l of nuclease-free water (Promega Product, Catalys AG, Wallisellen, Switzerland) and immediately processed for reverse transcription. For M. pneumoniae, C. pneumoniae, and L. pneumophila identification, 500 l of the sample was used. Extraction was performed using a phenol chloroform isoamyl alcohol procedure followed by ethanol precipitation. Reverse Transcription Reverse transcription was performed using Superscript II RNase H Reverse Transcriptase (Invitrogen, Life Technologies, Basel, Switzerland) in a reaction mixture containing 5 l of extracted RNA, 4 l of 5 First-Strand Buffer (250-mmol Tris-HCl [ph 8.3], 375-mmol KCl, 15-mmol MgCl 2 ), 2 l of dithiothreitol (0.1 M), 100 U Superscript II RNase H Reverse Transcriptase, 2 l of deoxynucleoside triphosphate (1.5 mmol; Amersham Biosciences Europe GmbH, Otelfingen, Switzerland), 20 U of RNase Inhibitor (Roche Diagnostics GmbH, Mannheim, Germany), 1.5 l of Primer Random p(dn) 6 (Roche Diagnostics GmbH, Rotkrenz, Switzerland), and 4 l of nuclease-free water. Reverse transcription was performed for 60 minutes at 42 C followed by incubation at 95 C for 10 minutes. The tubes were placed on ice and were immediately processed for PCR. Real-time PCR PCR (total reaction of 25 l per well) was performed using 5 l cdna, Taqman Universal Mastermix containing Rox passive reference (Applied Biosystems, PE Europe B.V., Rotkrenz, Switzerland) and appropriate concentrations of the primers and probes described in Table 1. Amplification and detection were performed with an ABI Prism 7,900 HT (Applied Biosystems, PE Europe B.V.) under two different sets of conditions according to the virus type: 2 minutes at 50 C, 10 minutes at 95 C, and 55 cycles of 15 seconds at 95 C and 1 minute at 55 C for rhinovirus, parainfluenza 1, and coronaviruses OC43, E229, and NL63. For influenza A and B, RSV A and RSV B, parainfluenza 3, human metapneumovirus, M. pneumoniae, C. pneumoniae, and L. pneumophila, the conditions were 2 minutes at 50 C, 10 minutes at 95 C, and 55 cycles of 15 seconds at 95 C and 1 minute at 60 C. Samples were considered positive if the number of cycles for amplicon production was less than 45. The specificity of all PCR assays has been confirmed by carefully conducted experiments that have tested each PCR assay for all other respiratory viruses and heterologous viruses likely to be isolated in the respiratory tract (herpes simplex viruses, cytomegalovirus, and severe acute respiratory syndrome [SARS] coronavirus). None of these assays tested positive for other viruses. In particular, the rhinoviral PCR did not detect any enteroviruses (7). The analytical RT-PCR sensitivity, including extraction, was evaluated by using cell culture supernatants of known concentration for influenza A and influenza B, RSV A and RSV B, rhinovirus, coronavirus OC43, E229, and enterovirus. Based on these experiments, the sensitivity ranged from % tissue culture infective dose depending on the virus studied. When compared with culture of 357 clinical specimens obtained during the influenza season, the influenza RT-PCR detected 183 of 184 culture positive specimens, and 41 (11%) additional samples were PCR positive but were culture negative (8). Similarly, the rhinoviral RT-PCR performed on 161 clinical specimens showed a higher sensitivity compared with cell culture (14% vs. 4% of positive results, respectively) (7). For difficult-to-grow viruses, including the human metapneumovirus and the new coronavirus NL63, we used plasmids and RNA as PCR controls. The human metapneumovirus control was designed from plasmid DNA provided by R. A. Fouchier and A. Osterhaus (Department of Virology, Erasmus Medical Center, Amsterdam, The Netherlands). The sensitivity of the PCR was two DNA copies per microliter. A stock of 10 4 to 10 5 copies/ l of human coronavirus NL63 RNA was kindly provided by L. van der Hoek (Department of Human Retrovirology, Academic Medical Center, University of Amsterdam, Amsterdam, The Netherlands). The

3 Garbino, Gerbase, Wunderli, et al.: Lower Respiratory Viral Illnesses 1199 TABLE 1. PRIMERS AND PROBES USED IN REAL-TIME POLYMERASE CHAIN REACTION ASSAYS FOR IDENTIFICATION OF RESPIRATORY VIRUSES OR ATYPICAL BACTERIA Virus (Target Gene) Primer Forward Primer Reverse Probes Influenza A (matrix) 5 -GGA CTG CAG CGT AGA CGC TT-3 5 -CAT CCT GTT GTA TAT GAG GCC 5 -CTC AGT TAT TCT GCT GGT CAT-3 GCA CTT GCC A-3 Influenza B (hemagglutinin) 5 -AAA TAC GGT GGA TTA AAT AAA AGC 5 -CCA GCA ATA GCT CCG AAG AAA-3 5 -CAC CCA TAT TGG GCA ATT AA-3 TCC TAT GGC-3 RSV A (N gene) 5 -CTC AAT TTC CTC ACT TCT CCA GTG T-3 5 -CTT GAT TCC TCG GTG TAC CTC 5 -TCC CAT TAT GCC TAG GCC TGT-3 AGC AGC A-3 RSV B (N gene) 5 -TTC CTA ACT TCT CAA GTG TGG TCC 5 -CTG GTT TCT TGG CGT ACC TCT ATA 5 -TCC CAT TAT GCC TAG ACC TA-3 C-3 TGC TGC ATT G-3 Parainfluenza 1 (hemagglutinin-neuraminidase) 5 -CAT TAT CAA TTG GTG ATGC-3 5 -CTT AAA TTC AGA TAT GTA TCC TG-3 5 -CTT AAT CAC TCA AGG ATG TGC AGA TAT A-3 Parainfluenza 3 (hemagglutinin-neuraminidase) 5 -CTC GAG GTT GTC AGG ATA TAG-3 5 -CTT GGG AGT TGA ACA CAG TT-3 5 -AAT AAC TGT AAA CTC AGA CTT GGT ACC TGA CTT-3 Rhinovirus (5 noncoding region) 5 -GCA CTT CTG TTT CCC C GGC AGC CAC GCA GGC T-3 5 -AGC CTC ATC TGC CAG GTC TA-3 5 -AGC CTC ATC TGC CAG GTC TG-3 Human metapneumovirus (polymerase) 5 -TGC TCA TGC CCA CTA TAA AAG GT-3 5 -TCT GTT AAT ATC CCA CAC 5 -CCA TGG AAA TAA TTC TCT CAA TGA C-3 CTC TTG TTC AGG AAC T-3 Coronavirus OC43 (polymerase) 5 -CGC CGC CTT ATT AAA GAT GTT G-3 5 -GGC ATA GCA CGA TCA CAC TTA GG-3 5 -AAT CCT GTA CTT ATG GGT T GGG ATT-3 Coronavirus 229 E (polymerase) 5 -TGG AGC GAG GAT CGT GTT C-3 5 -TAG GCT GTG ACA GCT TTT GCA-3 5 -TGT TCT CAC GCT GCT GTT GAT TCG CT-3 Coronavirus NL63 (replicase) 5 -TGT TGT AGT AGG TGG TTG TGT AAC 5 -AAT TTT TGT GCA CCA GTA 5 -ATG TTT CAC CAA TTG TTA ATCT-3 TCA AGT TT-3 GTG AGA AAA TTT CTG TTA TGG A-3 Mycoplasma pneumoniae (Cytadhesin P1) 5 -AAG TTA AAC CCG CAA ACG CC-3 5 -GGG ACC TTG TTT TTG ACC TCG-3 5 -TCA CCT TTA ACC CCT TTG GCG GGC T-3 Chlamydophila pneumoniae (Pst1) 5 -TGG AGA TAA AAT GGC TGG ACG-3 5 -TAT GGC ATA TCC GCT TCG G-3 5 -CAC GGA AAT AAA GGT GTT GTT TCC AAA ATC G-3 Legionella pneumophila (16S rdna) 5 -CGT AAG GGC CAT GAT GAC TTG-3 5 -TTG GGT TAA GTC CCG TAA CGA-3 5 -ACC ATC ACA TGC TGG CAA CTA AGG AT-3 Definition of abbreviation: RSV respiratory syncytial virus. limit of detection of our RT-PCR assay was a dilution of 10 4 of this stock. The limit of detection of the M. pneumoniae, C. pneumoniae, and L. pneumophila PCR was 0.01 pg of purified DNA spiked into the PCR reaction. For each specimen, an inhibition control was conducted during all procedures, including the extraction. Positive and negative control subjects were conducted for each PCR assay. End Points Based on previous studies, the expected number of samples positive by cell culture for at least one respiratory virus is less than 5% (9, 10), and the expected number of positive cases by molecular amplification is estimated as between 15 and 35%. For the primary analysis, we identified the number of respiratory viral infections and/or atypical bacteria observed by molecular tests in the two different groups, as well as the 1-month survival according to the presence of respiratory viruses and/or atypical bacteria. The number of RT-PCR positive versus culture-positive cases was also compared. In the subset of lung transplant recipients, the FEV 1 was measured at the time of the BAL procedure and again 3 months later. Values were expressed as the percentage change according to the baseline values defined as the previous best FEV 1 value obtained within 3 months before the BAL episode. The presence or absence of acute rejection, as defined by the International Society of Heart and Lung Transplantation, was documented in all but two patients (one in each group) with an available lung biopsy. Only acute rejections A 2 or more were considered for comparison between groups. The outcome of lung recipients was assessed by recording the occurrence of bronchiolitis obliterans or bronchiolitis obliterans syndrome after the infectious episode, according to International Society of Heart and Lung Transplantation criteria and mortality after infection. Statistical Analysis Data are presented as absolute values, percentage values, percentagepredicted values (FEV 1 ), mean SD, and median (range), as appropriate. Unpaired t tests were used to compare demographic data between groups. Chi-square tests were used to compare rates of infection, acute rejection, and bronchiolitis obliterans and/or bronchiolitis obliterans syndrome between groups. Mortality rates between groups were compared using the Kaplan-Meier log rank method. These analyses were performed using the PRISM version 3.0 for Windows (GraphPad Software, San Diego, CA). A p value of 0.05 or less was considered statistically significant. Role of the Funding Source The funding source had no role in the design, data collection, analysis, or interpretation of the study or in the decision to submit the article for publication. RESULTS During a 1-year study period, 148 BAL specimens from 111 patients were analyzed. Multiple BAL procedures (mean, 2.8; range, 2 6) were performed on 21 patients. The mean age of patients was 54 years (range, 1 83), and 83 (56%) were male. Primary clinical conditions present at the time of the BAL proce-

4 1200 AMERICAN JOURNAL OF RESPIRATORY AND CRITICAL CARE MEDICINE VOL TABLE 2. MAIN CHARACTERISTICS AND CLINICAL CONDITIONS PRESENT AT THE TIME OF THE BRONCHOALVEOLAR LAVAGE PROCEDURE TABLE 3. RESPIRATORY VIRUSES AND ATYPICAL BACTERIA IDENTIFIED FROM BRONCHOALVEOLAR LAVAGE EPISODES PERFORMED IN GROUP 1 n(% ) Immunosuppressive therapy* 86 (58) Transplant recipients 75 (51) HIV positive (median CD4 78/mm 3, range ) 15 (13) Solid cancer 15 (13) Autoimmune and collagen diseases 16 (11) Cirrhosis 4 (3) Chemotherapy during hospitalization 13 (11) At least one immunosuppressive condition* 113 (76) At least one additional comorbidity* 147 (99) Bacterial and/or fungal infection in BAL specimens 29 (20) New chest X-ray infiltrate 79 (68) Antibiotic treatment 113 (76) Definition of abbreviations: BAL bronchoalveolar lavage; HIV human immunodeficiency virus. n 148. * See METHODS for definitions. dure are shown in Table 2. A comorbidity was present in all but one subject, and an underlying immunosuppressive condition was present during 76% of the episodes. Graft recipients (lung, 57; kidney, 8; heart, 4; liver, 4; and hematopoietic stem cell transplantation, 2) received 75 (51%) BAL procedures. Other immunosuppressive conditions included human immunodeficiency virus infection in 15 patients (10%), solid cancer in 15 (10%), autoimmune or collagen disease in 16 (11%), diabetes in 19 (13%), and liver cirrhosis in 4 (3%). Of the 148 BAL specimens analyzed, 117 were from Group 1 and 31 from Group 2. Respiratory viruses and/or atypical bacteria were identified in 39 of 117 BAL specimens (33%) in Group 1 compared with 2 of 31 BAL specimens (7%) in Group 2 (p 0.01). Group 1 positive BAL specimens were distributed as follows: 34 were positive for a respiratory virus by RT-PCR and/ or cell culture (by cell culture only: two adenovirus, one RSV A), and M. pneumoniae infection was identified in five additional cases. Overall, a double infection was present in six cases (two viruses in five cases and one virus and M. pneumoniae in one case). The two positive cases in Group 2 included one case of rhinovirus in a lung transplant recipient and one case of influenza in a young patient with cerebral death after a cranial trauma during the influenza season. Given the circumstances for the latter patient, we did not have a complete medical history, and the BAL was performed for screening purposes before lung donation for transplantation. Among the BAL samples positive by RT-PCR for a respiratory virus, only 30% were culture positive (p 0.001). The distribution of the different virus types identified is detailed in Table 3. Of the respiratory viral episodes, 65% occurred during the autumn and winter seasons. In group 1, the mortality rate at 30 days was 13%. Mortality was similar between those positive for a respiratory virus and/or an atypical agent compared with those with negative results, 4 of 39 (10%) and 11 of 78 (14%), respectively (p 0.49). In addition to respiratory viruses, cytomegalovirus was isolated in 15 of 117 samples (13%) in Group 1 and in 5 of 31 (16%) in Group 2. Herpes simplex virus was isolated in 9 of 148 samples (6%) in Group 1 and in none of Group 2. Concomitant bacterial and/or fungal respiratory infections (Haemophilus influenza, Moraxella catarrhalis, Staphylococcus aureus, Streptococcus pneumoniae, Pseudomonas aeruginosa, Nocardia asteroides, and/or Aspergillus fumigatus) were documented at the time of BAL in 8 of 39 virus-positive cases (21%) and in 12 of 78 virus-negative cases (15%) from Group 1. Similarly in Group 2, concomitant infec- n(% ) Positive for at least one respiratory virus and/or atypical bacteria* 39 (33) Positive for at least one respiratory virus* 34 (29) Human rhinovirus 19 (56) RSV A 5 (15) RSV B 4 (12) Adenovirus 2 (6) Influenza B 2 (6) Coronavirus 229 E 2 (6) Parainfluenza 1 2 (6) Influenza A 1 (3) Parainfluenza 3 1 (3) Coronavirus OC43 1 (3) Human metapneumovirus 0 (0) Coronavirus NL63 0 (0) Positive for at least one atypical bacteria* 6 (5) Mycoplasma pneumoniae 6 (100) Legionella pneumophila 0 (0) Chlamydophila pneumoniae 0 (0) For definition of abbreviations see Table 1. n 117. * Dual infections were observed in six episodes: two respiratory viruses in five cases (rhinovirus with coronavirus 229 E in two, and rhinovirus with RSV A, RSV B, or parainfluenza 1, one in each case) and a dual viral/atypical bacteria infection in one case (influenza B with M. pneumoniae). Tested on 120 specimens for which we had enough sample remaining (78% of cases of group 1, 81% of group 2, 87% of subgroup 1, and 77% of subgroup 2). tions were detected in 1 of 2 (50%) virus positive cases and in 8 of 29 (28%) virus negative cases. We next focused our analysis on the 57 BAL procedures performed in lung transplant recipients: 31 cases in Subgroup 1 and 26 control subjects in Subgroup 2. A new chest X-ray infiltrate was present at the time of the BAL in 13 (42%) episodes in Subgroup 1 and none in Subgroup 2. The number of episodes positive for a respiratory virus was 17 (55%) and 1 (4%) in Subgroups 1 and 2, respectively, (p 0.001). The rate of bacterial and/or fungal infections diagnosed in BAL specimens was 16% in Subgroup 1 and 23% in Subgroup 2, a difference that was not statistically significant. The rate of cytomegalovirus infection was similar in both subgroups (Table 4). We also analyzed lung function changes over time by calculating the percentage change in FEV 1 between the best absolute values obtained 3 months before and the values obtained during the suspected infection (Figure 1). Compared with Subgroup 2, a significant decrease of FEV 1 was found in Subgroup 1 after the viral infection, and this functional loss persisted for more than 3 months in several patients. The median change (range) in the FEV 1 was 145 ml ( 1,580 to 390 ml) in Subgroup 1 and 65 ml ( 770 to 690 ml) in Subgroup 2. The proportions of cases with an FEV 1 lower than 70% of the predicted value at the time of the BAL procedure were 53% and 33% in Subgroups 1 and 2, respectively. The rate of acute rejections and bronchiolitis obliterans/bronchiolitis obliterans syndrome were similar in both subgroups as shown in Table 4. The 1-month survival rate was 89% (n 9) in lung transplant recipients who experienced at least one respiratory viral infection and 100% (n 9) in those who were negative in all BAL procedures. Finally, to assess whether respiratory virus could be shed chronically and asymptomatically in lung transplant recipients, we analyzed available BAL samples before and after a positive viral episode. We identified eight positive cases that fulfilled these

5 Garbino, Gerbase, Wunderli, et al.: Lower Respiratory Viral Illnesses 1201 TABLE 4. CLINICAL CHARACTERISTICS AND OUTCOME OF LUNG TRANSPLANT RECIPIENTS WITH (SUBGROUP 1) AND WITHOUT (SUBGROUP 2) CLINICAL SUSPICION OF RESPIRATORY INFECTIONS Subgroup 1n(% ) Subgroup 2n(% ) (n 31) (n 26) p Respiratory virus 17 (55) 1 (4) 0.01 Rhinovirus 9 1 na RSVAorB 4 0 na Influenza A or B 2 0 na Parainfluenza na Adenovirus 1 0 na Respiratory bacterial and/or fungal infection* 5 (16) 6 (23) 0.4 Respiratory CMV infection* 4 (13) 4 (15) 0.54 Respiratory HSV infection* 0 (0%) 0 (0%) Blood lymphocytes, g/l FEV 1 (% predicted), mean SD Acute rejection ( A 2 ), n (%) 2/20 (10) 5/21 (24) 0.32 BO/BOS, n (%) 2/20 (10) 1/21 (4) 0.55 Definition of abbreviations: BO bronchiolitis obliterans; BOS bronchiolitis obliterans syndrome; CMV cytomegalovirus; HSV herpes simplex virus; na not applicable; RSV respiratory syncytial virus. * Positive culture in the bronchoalveolar lavage specimen. Because some subjects had more than one bronchoalveolar lavage procedure, we analyzed patients individually and not each bronchoalveolar lavage procedure alone. Number of acute rejection episodes ( A 2 ) diagnosed during hospitalization using the International Society of Heart and Lung Transplantation criteria. BO and/or BOS diagnosed during hospitalization using the International Society of Heart and Lung Transplantation criteria. criteria and in which BAL, performed as a routine procedure after transplantation, was available (a median of 12 weeks before and after the initial positive BAL). All of these samples tested negative for the virus identified at the time of the positive episode. DISCUSSION This study analyzed the presence of respiratory viruses in BAL specimens to determine their frequency and impact on LRTIs. Data showed that in hospitalized subjects (mainly immunocompromised patients) who needed a BAL procedure for an LRTI, the incidence of respiratory viral infections was very high. Indeed, RT-PCR assays targeting 11 different respiratory viruses Figure 1. Change of FEV 1 in three groups of lung recipients. Percentage change of FEV 1 according to baseline in three groups of lung recipients. Samples were obtained during a confirmed respiratory viral infection (Subgroup 1 virus positive; circles), during a suspected but not confirmed respiratory viral infection (Subgroup 1 virus negative; triangles), and during a routine clinical control (Subgroup 2; inverted triangles). The follow-up of the FEV 1 is presented in the same Subgroups 3 months later (open symbols). identified a viral infection in 30% of episodes compared with only 10% by cell culture methods. This sensitive tool better estimates the frequency of respiratory viral infections in hospitalized patients and permits the identification of the cause of a respiratory event in patients who do not respond to conventional empirical antibiotic treatment. Although mild or subclinical illness may follow a respiratory viral infection, it is generally agreed that the presence of viral RNA in respiratory secretions denotes a recent infection because these viruses are not known to cause latent infections. In healthy subjects, the infection is generally limited to the upper respiratory tract with viral clearance occurring within a few days (11, 12). Conversely, patients with comorbidities and/or impaired immune function are at a higher risk of severe LRTI complications (9 10, 13 16). To understand better whether these respiratory viruses contributed to the symptoms that prompted the BAL procedure, we specifically analyzed lung transplant recipients (Subgroups 1 and 2) because this largest subgroup is the most prone to complications. We found a high rate of viral infections (55%) in transplant recipients who received the BAL procedure. This differs significantly with the 4% of viral infections found in the control group. After the acute episode of infection, we observed a worsening of pulmonary function that persisted for more than 3 months in a substantial number of cases. This is consistent with previous reports showing a deterioration of lung function in lung transplant recipients with respiratory viral infections (10, 17 20). To assess further the significance of this finding, we identified all positive episodes in which BAL samples were available in the months before and after the viral episode. All pre- and post-bal specimens tested negative for the virus identified during the period of respiratory symptoms, strongly suggesting that it is unusual to detect respiratory viruses without the presence of respiratory symptoms. Other events that could produce symptoms similar to viral respiratory infections such as bacterial or cytomegalovirus infection, or acute rejection in lung transplant recipients, were only identified in a very small number of cases. Taken together, the majority of our findings strongly suggest that respiratory viral infections are frequent in lung

6 1202 AMERICAN JOURNAL OF RESPIRATORY AND CRITICAL CARE MEDICINE VOL transplant recipients, particularly during the autumn winter season when community outbreaks occur. Viral infections played an important role in the onset of respiratory symptoms, and asymptomatic shedding was uncommon. The most frequent virus recovered was rhinovirus. This is in agreement with the knowledge that rhinovirus is the primary cause of acute viral respiratory illnesses (21). Such a high rate of rhinoviral infection in a population of hospitalized subjects has not been reported previously, although rhinovirus is a major cause of asthma exacerbation (22, 23) and has been linked to hospital outbreaks (24). This could be explained by the fact that few studies have addressed this issue by systematically applying a sensitive RT-PCR method. Rhinovirus replication is mostly limited to the upper respiratory tract, but experimental data have shown that it also replicates in the lower respiratory tract (25), and severe rhinoviral pneumonia has been described in immunocompromised hosts (9, 26, 27). Our findings corroborate these previous reports and strongly suggest that rhinovirus is a leading cause of respiratory disease in hospitalized immunocompromised hosts (28, 29). Similar to rhinovirus, human coronavirus is generally not identified by cell culture and is not targeted by commonly used RT-PCR panels. In this study, human coronaviruses OC43 and 229E were identified in three cases. Coronavirus has been associated with nosocomial respiratory outbreaks (30) and with acute cardiopulmonary illnesses in older patients (31). The identification of coronavirus in lower respiratory tract specimens in adults has not been reported in earlier case studies. In a recent report of hematopoietic stem cell transplantation recipients, RT-PCR assays performed in 46 BAL specimens of patients with new chest X-ray abnormalities were all negative for coronavirus (27). These differences might be explained by patient selection, assay sensitivity, and seasonal pattern (32). Whether human coronavirus can easily replicate in the lower respiratory tract of immunocompromised adults needs to be confirmed in further studies. However, the SARS outbreak together with the recent discovery of a new human coronavirus that causes bronchiolitis (33) and the evidence that some animal species suffer from pneumonia (34) illustrates the potential of these viruses. We also tested 80% of specimens for the presence of the recently discovered human coronavirus NL63 (33), but all were negative. The first reports have shown that this virus can be recovered at a relatively low frequency in upper or lower respiratory tract specimens from children and adults with respiratory symptoms (33, 35). Our observation suggests that during the study period this coronavirus was an infrequent cause of respiratory diseases in our group of patients; however, our knowledge of the epidemiology and impact of this virus is still in an early stage and deserves further investigation. In recent years, human metapneumovirus has been shown to cause respiratory diseases similar to RSV in children (36 38). However, the morbidity of human metapneumovirus in adults and in immunocompromised patients (39), as well as its seasonal distribution (40), has not been fully evaluated. Our study addressed a population of hospitalized patients during one winter season. In these patients, human metapneumovirus was not isolated. This specific seasonal characteristic is also illustrated by a relatively low frequency of influenza infections in this study, whereas in a previous study influenza was the most frequent virus recovered from BAL specimens by cell culture (10). These variations are related to the nature of the influenza virus itself and its ability to cause annual outbreaks of different intensity. The influenza epidemic during the study period was moderate compared with previous years (data not shown). RSV, on the other hand, was the second most frequent virus and is a wellknown cause of severe disease in immunocompromised hosts (12, 41, 42). Despite the use of PCR assays for identification of atypical bacteria, only a small number of M. pneumoniae cases were identified, and no C. pneumoniae or L. pneumophila infections were identified. Thus, these bacteria were an infrequent cause of illness in our study population during the time period studied. These results could be related to epidemiologic patterns but also to the use of wide-spectrum antibiotics that target these microorganisms as empirical therapy. In conclusion, we have shown that by using appropriate and sensitive assays, respiratory viruses are frequently recovered in BAL specimens of patients hospitalized with an LRTI. When present, these viruses are associated with high morbidity, particularly in lung transplant recipients. Rapid identification of respiratory viruses in this latter population could avoid costly procedures and unnecessary treatment. The fact that therapeutic intervention with neuraminidase inhibitors, ribavirin, or immunoglobulins might be effective in cases of influenza, RSV, or other respiratory viral infections should also be considered. New drugs such as pleconaril or protease inhibitors, active against a wide range of rhinoviral strains (43), are also in development. However, very few prospective and controlled studies have been performed in hospitalized patients, and thus, the role of antiviral therapy in this population deserves further investigation (44). Nevertheless, our observation that respiratory viruses are associated with significant clinical impact strongly suggests that the timely use of PCR-based assays, followed by appropriate antiviral treatment, may have a determinant impact on clinical care and outcome, particularly in patients at higher risk for complications. Conflict of Interest Statement : J.G. does not have a financial relationship with a commercial entity that has an interest in the subject of this manuscript; M.W.G. in the subject of this manuscript; W.W. does not have a financial relationship with a commercial entity that has an interest in the subject of this manuscript; C.D. in the subject of this manuscript; Y.T. does not have a financial relationship with a commercial entity that has an interest in the subject of this manuscript; T.R. in the subject of this manuscript; B.N. does not have a financial relationship with a commercial entity that has an interest in the subject of this manuscript; J.S. in the subject of this manuscript; S.Y. does not have a financial relationship with a commercial entity that has an interest in the subject of this manuscript; L.P. in the subject of this manuscript P.M.S. does not have a financial relationship with a commercial entity that has an interest in the subject of this manuscript; L.N. does not have a financial relationship with a commercial entity that has an interest in the subject of this manuscript; L.K. does not have a financial relationship with a commercial entity that has an interest in the subject of this manuscript. References 1. File TM. Community-acquired pneumonia. Lancet 2003;362: Fine MJ, Stone RA, Singer DE, Coley CM, Marrie TJ, Lave JR, Hough LJ, Obrosky DS, Schulz R, Ricci EM, et al. Processes and outcomes of care for patients with community-acquired pneumonia: results from the Pneumonia Patient Outcomes Research Team (PORT) cohort study. Arch Intern Med 1999;159: Garbino J, Sommer R, Gerber A, Regamey C, Vernazza P, Genne D, Dur P, Rothen M, Unger JP, Lew D. Prospective epidemiologic survey of patients with community-acquired pneumonia requiring hospitalization in Switzerland. Int J Infect Dis 2002;6: Freymuth F, Vabret A, Galateau-Salle F, Ferey J, Eugene G, Petitjean J, Gennetay E, Brouard J, Jokik M, Duhamel JF, et al. Detection of respiratory syncytial virus, parainfluenzavirus 3, adenovirus and rhinovirus sequences in respiratory tract of infants by polymerase chain reaction and hybridization. Clin Diagn Virol 1997;8: Fan J, Henrickson KJ, Savatski LL. Rapid simultaneous diagnosis of infections with respiratory syncytial viruses A and B, influenza viruses A and B, and human parainfluenza virus types 1, 2, and 3 by multiplex quantitative reverse transcription-polymerase chain reaction-enzyme hybridization assay (Hexaplex). Clin Infect Dis 1998;s26: Kaiser L, Garbino J, Deffernez C, Thomas Y, Gerbase M, Perrin L, Gasche-Soccal P, Nicod L, Rochat T, Wunderli W. Detection of respiratory virus and atypical bacteria from BAL in hospitalized patients

7 Garbino, Gerbase, Wunderli, et al.: Lower Respiratory Viral Illnesses 1203 with severe respiratory infection and impact on diseases severity [abstract]. American Society of Microbiology, 43th Interscience Conference Antimicrobial Agents and Chemotherapy. Chicago, September 14 17, Deffernez C, Wunderli W, Thomas Y, Yerly S, Perrin L, Kaiser L. Amplicon sequencing and improved detection of human rhinovirus in respiratory samples. J Clin Microbiol 2004;42: Thomas Y, Kaiser L, Hagmann R, Burckhardt-Batista I, Wunderli W. Influenza monitoring by different methods in a surveillance network [abstract]. Influenza vaccine for the world. Lisbon, Portugal, May 24 26, Malcolm E, Arruda E, Hayden FG, Kaiser L. Clinical features of patients with acute respiratory illness and rhinovirus in their bronchoalveolar lavages. J Clin Virol 2001;21: Garbino J, Gerbase MW, Wunderli W, Kolarova L, Nicod L, Rochat T, Kaiser L. Respiratory viruses and severe lower respiratory tract complications in hospitalized patients. Chest 2004;125: Kaiser L, Briones MS, Hayden FG. Performance of virus isolation and Directigen flu A to detect influenza A virus in experimental human infection. J Clin Virol 1999;14: Hall CB. Respiratory syncytial virus and parainfluenza virus. N Engl J Med 2001;344: Wright PF, Webster RG. Othomyxoviruses. In: Knipe DM, Howley PM, editors. Fields virology, 4th ed. Philadelphia: Lippincott Williams and Wilkins; p Yousuf HM, Englund J, Couch R, Rolston K, Luna M, Goodrich J, Lewis V, Mirza NQ, Andreeff M, Koller C, et al. Influenza among hospitalized adults with leukemia. Clin Infect Dis 1997;24: Randolph AG, Wang EE. Ribavirin for respiratory syncytial virus lower respiratory tract infection: a systematic overview. Arch Pediatr Adolesc Med 1996;150: Whimbey E, Champlin RE, Couch RB, Englund JA, Goodrich JM, Raad I, Przepiorka D, Lewis VA, Mirza N, Yousuf H, et al. Community respiratory virus infections among hospitalized adult bone marrow transplant recipients. Clin Infect Dis 1996;22: Garantziotis S, Howell DN, McAdams HP, Davis RD, Henshaw NG, Palmer SM. Influenza pneumonia in lung transplant recipients: clinical features and association with bronchiolitis obliterans syndrome. Chest 2001;119: Vilchez R, McCurry K, Dauber J, Iacono A, Keenan R, Griffith B, Kusne S. Influenza and parainfluenza respiratory viral infection requiring admission in adult lung transplant recipients. Transplantation 2002; 73: Wendt CH, Fox JM, Hertz MI. Paramyxovirus infection in lung transplant recipients. J Heart Lung Transplant 1995;14: Khalifah AP, Hachem RR, Chakinala MM, Schechtman B, Patterson A, Schuster DP, Mohanakumar T, Trulock EP, Walter MJ. Respiratory viral infections are a distinct risk for bronchiolitis obliterans syndrome and death. Am J Respir Crit Care Med 2004;170: Nicholson KG, Kent J, Hammersley V, Cancio E. Risk factors for lower respiratory complications of rhinovirus infections in elderly people living in the community: prospective cohort study. BMJ 1996;313: Johnston SL, Pattemore PK, Sanderson G, Smith S, Lampe F, Josephs L, Symington P, O Toole S, Myint SH, Tyrrell DA, et al. Community study of role of viral infections in exacerbations of asthma in 9 11 year old children. BMJ 1995;310: Johnston SL, Pattemore PK, Sanderson G, Smith S, Campbell MJ, Josephs LK, Cunningham A, Robinson BS, Myint SH, Ward ME, et al. The relationship between upper respiratory infections and hospital admissions for asthma: a time-trend analysis. Am J Respir Crit Care Med 1996;154(3 Pt. 1): Wald TG, Shult P, Krause P, Miller BA, Drinka P, Gravenstein S. A rhinovirus outbreak among residents of a long-term care facility. Ann Intern Med 1995;123: Papadopoulos NG, Bates PJ, Bardin PG, Papi A, Leir SH, Fraenkel DJ, Meyer J, Lackie PM, Sanderson G, Holgate ST, et al. Rhinoviruses infect the lower airways. J Infect Dis 2000;181: Ghosh S, Champlin R, Couch R, Englund J, Raad I, Malik S. Luna M, Whimbey E. Rhinovirus infections in myelosuppressed adult blood and marrow transplant recipients. Clin Infect Dis 1999;29: Ison MG, Hayden FG, Kaiser L, Corey L, Boeckh M. Rhinovirus infections in hematopoietic stem cell transplant recipients with pneumonia. Clin Infect Dis 2003;36: Hayden FG. Rhinovirus and the lower respiratory tract. Rev Med Virol 2004;14: Papadopoulos NG. Do rhinoviruses cause pneumonia in children? Paediatr Respir Rev 2004;5:S191 S Gagneur A, Sizun J, Vallet S, Legr MC, Picard B, Talbot PJ. Coronavirusrelated nosocomial viral respiratory infections in a neonatal and paediatric intensive care unit: a prospective study. J Hosp Infect 2002;51: Falsey AR, Walsh EE, Hayden FG. Rhinovirus and coronavirus infectionassociated hospitalizations among older adults. J Infect Dis 2002;185: Vabret A, Mourez T, Gouarin S, Petitjean J, Freymuth F. An outbreak of coronavirus OC43 respiratory infection in Normandy, France. Clin Infect Dis 2003;36: Van Der Hoek L, Pyrc K, Jebbink MF, Vermeulen-Oost W, Berkhout RJ, Wolthers KC, Wertheim-van Dillen PM, Kaandorp J, Spaargaren J, Berkhout B. Identification of a new human coronavirus. Nat Med 2004;10: Sirinarumitr T, Paul PS, Halbur PG, Kluge JP. An overview of immunological and genetic methods for detecting swine coronaviruses, transmissible gastroenteritis virus, and porcine respiratory coronavirus in tissues. Adv Exp Med Biol 1997;412: Fouchier RA, Hartwig NG, Bestebroer TM, Niemeyer B, de Jong JC, Simon H, Osterhaus AD. A previously undescribed coronavirus associated with respiratory disease in humans. Proc Natl Acad Sci USA 2004;101: Williams JV, Harris PA, Tollefson SJ, Halburnt-Rush LL, Pingsterhaus JM, Edwards KM, Wright PF, Crowe JE Jr. Human metapneumovirus and lower respiratory tract disease in otherwise healthy infants and children. NEnglJMed2004;350: van den Hoogen BG, van Doornum GJ, Fockens JC, Cornelissen JJ, Beyer WE, de Groot R, Osterhaus AD, Fouchier RA. Prevalence and clinical symptoms of human metapneumovirus infection in hospitalized patients. J Infect Dis 2003;188: Boivin G, De Serres G, Cote S, Gilca R, Abed Y, Rochette L, Bergeron MG, Dery P. Human metapneumovirus infections in hospitalized children. Emerg Infect Dis 2003;9: Cane PA, van den Hoogen BG, Chakrabarti S, Fegan CD, Osterhaus AD. Human metapneumovirus in a haematopoietic stem cell transplant recipient with fatal lower respiratory tract disease. Bone Marrow Transplant 2003;31: Falsey AR, Erdman D, Anderson LJ, Walsh EE. Human metapneumovirus infections in young and elderly adults. J Infect Dis 2003;187: Dowell SF, Anderson LJ, Gary HE Jr, Erdman DD, Plouffe JF, File TM Jr, Marston BJ, Breiman RF. Respiratory syncytial virus is an important cause of community-acquired lower respiratory infection among hospitalized adults. J Infect Dis 1996;174: Simoes EA. Respiratory syncytial virus infection. Lancet 1999;354: Kaiser L, Crump CE, Hayden FG. In vitro activity of pleconaril and AG7088 against selected serotypes and clinical isolates of human rhinoviruses. Antivir Res 2000;47: Ison MG, Gnann JW Jr, Nagy-Agren S, Treannor J, Paya C, Steigbigel R, Elliott M, Weiss HL, Hayden FG, NIAID Collaborative Antiviral Study Group. Safety and efficacy of nebulized zanamivir in hospitalized patients with serious influenza. Antivir Ther 2003;8: