Respiratory Viruses in Mechanically Ventilated Patients

A Pilot Study

Raquel Nazareth; Maria-Jesus Chasqueira; Maria-Lúcia Rodrigues; Carolina Paulino; Catarina Conceição; Lia Lêdo; Úrsula Segura; Madalena Santos; António Messias; Pedro Póvoa; Paulo Paixão


BMC Pulm Med. 2020;20(39) 

In This Article


In our study it was possible to identify common respiratory viruses in samples of the lower respiratory tract of adults under invasive mechanical ventilation, regardless of the presence or absence of acute lower respiratory infection, through real-time PCR techniques. In the field of pulmonary virology investigation, several factors make this study innovative: the age and type of patients, the inclusion of two groups of patients, with (WRI group) and without acute lower respiratory infection (WORI group), and also the type of samples (protected samples from the distal lower respiratory tract). According to Luyt et al, BAL is the respiratory sample of choice to evaluate lung infection.[17]

Nucleic acid extraction was performed with the same commercial kit used in the study of Wang et al,[13] who analized the respiratory virome of healthy and severe acute respiratory infection children through metagenomic analysis. We used "in-house" real-time PCR and RT-PCR techniques that were tested and validated long before.[16]

Thirty percent of the patients belonging to WORI group had common respiratory viruses in the distal lower respiratory tract. According to data from the pediatric population,[13] lung virome of asymptomatic individuals is less diverse than that of infected individuals, and consists mostly, of members of the Anelloviridae family, with a lower percentage of common epidemic respiratory viruses. In the study of Wang et al,[13] samples from infected patients had six to seven-fold more viral pathogens than the WRI group. These authors suggested that the viral infection may be asymptomatic and, occasionally prolonged, making controversial the interpretation of a positive PCR for some viruses.

. In a study conducted by Choi et al,[18] the main respiratory viruses associated with severe pneumonia in adults admitted in ICU were HRV (23.6%), HPIV (20.8%), HMPV (18.1%), influenza (16.7%) and RSV (13.9%), according to data obtained from real-time PCR analysis in bronchoalveolar lavage samples. In the study by Xu et al,[14] 368 samples collected with nasopharyngeal swab from infected children were analyzed with real-time PCR, using a panel of 18 respiratory viruses. The percentage of positive samples was 58.97%.

In our study, half of the WRI group patients had positive samples. Comparing the two groups, influenza AH3 was the most prevalent virus, coinciding with the peak of influenza in Portugal. RSV, HMPV and HRV were common to both groups. Differences were found with HPIV 1/3, which was more prevalent in uninfected patients, and HEV and HBoV that were, only found in WRI group. The analysis of the Ct values showed similar viral loads in both groups, although with a slightly tendency for lower Ct values in WORI group. The small number of samples did not allow us to draw conclusions about possible differences in viral loads between infected and uninfected individuals.

It may be hypothesized that some respiratory viruses, such as influenza, RSV, HMPV and HRV, may transiently colonize the mucosa of the tracheobronchial tract at times of increased viral activity, whereas others, such as HPIV 1/3, might be prolonged colonizers. However, only a longitudinal study could determine the extent of the colonization period and thus solve this important issue. In addition, we need to know the meaning of its presence, if they are only bystanders, even during an acute respiratory infection, or if they are responsible for symptomatic infections of the lower respiratory tract.

Our study has some limitations, including a small sample size and a limited number of centers involved, which does not allow us to make inferences about mortality between the groups and subgroups. In addition, all the patients in both groups were severely ill. Therefore, the observations made in this study may not be generalizable to other groups, in particular to healthy populations.

Although signs and symptoms of respiratory infection were excluded at admission in the WORI group, viral respiratory infections in the last weeks preceding hospitalization cannot be excluded, and therefore viral detection in some of the cases could be the result of a recent infection and not an extended stay in the lower respiratory tract.

In this study a protected mini-BAL (with a double catheter, Combicath® kit) was used in order to reduce the upper level contamination. However, it is possible that contamination from upper respiratory secretions, mainly due to the ventilation process, may have had an important contribution to the detection rate observed in this study.

Another limitation relates with the methodology used: although real-time PCR is currently the gold-standard for the diagnosis of viral infections,[19] this methodology is limited to specific target sequences. In fact, only the most common respiratory viruses were investigated. According to Willner et al,[9] PCR-based studies confer an incomplete airway virome picture and little opportunity for the discovery of new agents, as compared to metagenomics, a technique independent of genomic sequences. In addition, viruses like herpesvirus simplex, cytomegalovirus or torque teno virus can be present in the distal airway mucosa,[13,20–22] and therefore may have been missed with our PCR strategy. However, despite this limitation, the main respiratory viruses were properly searched, and this is the group of major concern, when dealing with respiratory infections. Another important advantage of this study was the focus on a population undergoing invasive mechanical ventilation, allowing the collection of samples from the distal airways, although blindly, which is usually a limitation in this type of studies.