Lung biopsies for interstitial lung disease: the limits of the traditional methods of microbiological identification

Interstitial lung disease (ILD) is a broad category of lung pathologies that includes more than 100 disorders. The progression of ILD varies from disease to disease and from person to person. It is important to determine the specific form of ILD in patients because prognosis over time and treatment strategies may differ, depending on the cause. In this setting, surgical lung biopsy (LB) is needed to get enough lung tissue to make a specific diagnosis. If LB is strongly recommended by recent guidelines to obtain a confident pathological diagnosis [1,2], what constitute the best methods of microbiological identification remain to date undefined.

In an article published in this issue of the journal, Fibla et al. [3] report on 296 consecutive patients, who underwent LB for ILD during a 7-year period. The first aim of the study was to assess the rationale of microbiological analyses of such biopsies, including bacterial, mycobacterial, viral, and fungal cultures. A total of 592 specimens were assessed (range 1–4 per patient) from open LB in 8% and from thoracoscopic LB in 92%. The most common pathologic diagnoses were idiopathic pulmonary fibrosis (41%), cryptogenic organizing pneumonia (10%) and respiratory bronchiolitis ILD (5%). Microbiology testing was negative in 174 patients (59%) and positive in the remaining 122 patients. Among the positive specimens, 97% were clinically considered to be contaminants and resulted in no change in clinical management. The LB resulted in change in clinical management in only four patients (1.4%). The cost of microbiology studies per specimen was €709, with a total cost for the study cohort of €420 000. Because of high level of inconclusive microbiological results and because of the weak impact on clinical management, the authors have concluded that the routine use of such analyses remains questionable and should be avoided. Their routine use should be limited to those cases of ILD with a high suspicion of infection.

The authors are to be congratulated in tackling this important issue neglected in the existent literature. However, the article deserves several points of discussion.

First, in the broad spectrum of the different etiologies, ILD can be idiopathic but can also be wrongly considered as idiopathic because of the non-convincing results of microbiological analysis. Infectious diseases with different multiple microorganisms have been well documented with ILD [4]. However, the clear relation between infection and ILD is complex. On the one hand, pulmonary infections can cause extensive pulmonary damage with patterns of an ILD. On the other hand, microorganisms can trigger the immune system and provoke an abnormal response – not directed against the causative pathogen – that may result in ILD and fibrosis with inconclusive microbiological results [4]. Moreover, patients with ILD often are susceptible to infection, and infections can importantly influence the course of ILD. Thus, microorganisms can initiate and/or influence the course of ILD. As a result, microbiological assessment of LB seems to be an unavoidable part of the diagnosis procedure. A negative result in this context represents major information, as important as a positive result. At last, the resulting cost of such analysis does not represent a strong argument when taking into account the global cost of such diseases over a long period.

Second, the article from Fibla emphasizes mostly the current limitation of the methods of microbiological identification. Knowledge of microorganisms in lung-tissue biopsy and more generally in the environment has depended in the past mainly on studies of pure cultures in the laboratory. However, there are several limitations with traditional culture methods used in this study. First, depending on the type of sample, isolation and/or detection of potential microbial pathogens could be difficult in complex microbial samples contaminated by other bacteria or fungi. Second, traditional culture techniques are usually limited to a few selective or non-selective media on which many uncultured pathogens and anaerobes could be missed and thus could not be detected. Recent data have demonstrated that 99% of organisms seen microscopically are not cultivated by routine techniques and required modern and innovating techniques to be identified [5]. It seems critical in this context to develop new media that could be more permissive for the isolation of new bacteria. Thus, development of new molecular techniques, such as 16S ribosomal rRNA (16S rRNA) gene amplification and sequence directly from samples may be useful and will provide a new point of view of the problem [6]. This has been successfully used for pleural infection [7] and in pneumonia [8]. Moreover, evaluation of the potential microbial diversity in these samples may be also assessed using molecular techniques including 16S rRNA gene clonal library sequencing, 16S rRNA gene pyrosequencing, or metagenomic of the human microbiome, as recently exemplified to characterize respiratory-tract DNA-viral communities in cystic fibrosis and non-cystic fibrosis patients [6,9]. To date, there are no specific data on genetic and molecular approach in patients presenting ILD. This pleads to systematically conserve a piece of lung specimen in RNA later and to store it at −90 °C for subsequent analysis in selected patients.

The last point of discussion should be the lack of systematic viral analysis. Viruses, such as herpes viruses, have long been implicated as either causative agents or propagators of inflammation in ILDs. To date, current cultures of lung-tissue specimens have failed to provide evidence of viable viral organisms. Indirect evidence comes from studies reporting serologic evidence of exposure to numerous viruses in individuals with a variety of ILDs. A link between viral infection and ILDs has been suggested by some studies that amplified viral DNA from lung-biopsy specimens [10]. It is possible that some viruses may precipitate hypersensitivity pneumonitis in predisposed individuals and/or occasional cases of ILD to occur during the course of well-documented viral infections, indicating that, at least in some situations, viruses may be the primary cause of ILD. The lack of definitive evidence implicating viral infection as a cause of ILDs does not rule out a potential role of viruses as a triggering event that may be important for the development of ILD in predisposed individuals.

Beyond the economic point of view, the article by Fibla et al., clearly underlines the limits of the traditional phenotypic method of culture applied in ILD, prioritizing bacteria rather than virus and leading to potential diagnosis errors. Considering the severity of ILD and its prognosis, microbiological analyses should be dramatically improved by the way of modern and innovating techniques of microbiological identification.