Bronchoalveolar lavage for the evaluation of interstitial lung disease: is it clinically useful?

TECHNICAL ASPECTS OF PERFORMING BAL

BAL technique is of paramount importance in obtaining a useful specimen [10, 11]. To retrieve a specimen that is likely to reflect the types of inflammatory cells that are present in distal airspaces, a sufficient amount of isotonic saline should be instilled to adequately sample bronchoalveolar secretions from distal airspaces. The bronchoscopist should avoid contamination of the sample with proximal large airway secretions by maintaining the distal end of the bronchoscope in a wedged position in a segmental or subsegmental bronchus throughout the period of time required for the instillation and retrieval of saline aliquots. Additionally, dwell time should be minimised (aliquots should be aspirated immediately once the entire aliquot volume has been instilled). Many different BAL protocols have been published and consist of multiple aliquots that usually range from five or six aliquots of 20 mL each to four aliquots of 60 mL each. Some investigators consider the first aliquot to represent predominantly airway cells and secretions, and have recommended that this aliquot be kept separate and used for microbiological analysis, while subsequent sequential aliquots, usually three to four in number, should be pooled and used for cellular analysis because these are more representative of distal airspaces (alveolar sampling) [12]. Others combine all aliquots and do not differentiate between “bronchial” and “alveolar” sampling. There are no data that support one approach over the other, and many centres pool all retrieved BAL aliquots prior to submitting BAL for laboratory analysis. Additionally, there is no consensus on the optimal magnitude of negative pressure to retrieve BAL fluid once an aliquot has been instilled.

The right middle lobe (RML) and lingula of the left upper lobe are the areas of the lung that are more easily accessible and likely to allow good return of BAL fluid from patients when lavaged while in a supine position [11]. These areas have traditionally been used for lavage based on the perception that ILD is a “diffuse” parenchymal lung disease, as demonstrated on routine chest radiographs. Because patients with ILD are now routinely evaluated with HRCT images of the chest, HRCT images may be used to target areas of the lung that may yield cellular findings more reflective of the disease process, such as areas of ground glass attenuation, more prominent nodular profusion or fine reticulation. Since HRCT is sensitive in detecting patterns of otherwise occult segmental and/or patchy alveolar opacities or ground glass attenuation in patients with ILD, the HRCT images should be used as a guide to determine the site of BAL in order to include sampling from involved segments as indicated by the HRCT scan. However, no studies have been performed that demonstrate that the use of HRCT to target regions that show more prominent parenchymal change in certain geographic lung regions compared with traditionally used sites, such as the RML or lingula, improve the diagnostic yield of BAL. Ideally, the percentage of BAL fluid that is retrieved should be ≥30% for a patient with ILD. However, if ILD is superimposed on emphysema, airways may readily collapse with negative pressure and compromise fluid retrieval. Post-bronchoscopy fever (fever, chills and extreme malaise) may occur within hours of the BAL procedure as a consequence of pro-inflammatory mediator release, and these symptoms are more likely to occur if larger volumes of fluid are used for BAL [13].

PROCESSING OF BAL FLUID FOR CELLULAR ANALYSES

An accurate and statistically meaningful enumeration of the BAL nucleated immune cell profile requires examination of ≥300 of these BAL cells randomly on a single slide, with accuracy further enhanced by analysing 400–500 cells (especially if more than one slide is analysed with the differential counts from each slide averaged). The relative numbers of bronchial and squamous epithelial cells should also be noted. Although many investigators have published BAL cell profiles for cohorts of healthy volunteers, laboratories that perform BAL analysis would ideally study a sizable group of normal volunteers and establish normal values in their own laboratories that correspond to their subject population and analytical methods. BAL slide preparations can be analysed in clinical haematopathology laboratories, but it is of key importance that personnel who analyse the BAL slide preparations and determine differential counts are adequately trained in the proper identification of BAL cells and in recognising various atypia that may be present on such preparations [11]. However, pulmonologists familiar with BAL cell patterns and ILD should interpret the BAL cell patterns and other findings.

BAL fluid obtained from healthy, never-smoking individuals should contain, on average, a majority of alveolar macrophages (AMs) (80–90%), some lymphocytes (5–15%), and very few neutrophils (≤3%) or eosinophils (<1%). While it is preferential to have laboratories establish normal values (including T-lymphocyte subsets), normal values reported by others may be used, especially if the BAL procedure that is used conforms to published techniques and values obtained for normal subjects. The presence of squamous epithelial cells suggests that the BAL fluid has been contaminated by oropharyngeal secretions, which may be a consequence of poor technique in performing the BAL or because the subject aspirated upper airway secretions. BAL cell differentials and total cell counts for ex-smokers should be similar to that of never-smokers. Smokers, however, generally have a significantly increased total BAL cell count, as well as total macrophages and neutrophils per microlitre of BAL fluid, but the BAL differential cell count for smokers appears to vary little from that for never-smokers or ex-smokers, except for a somewhat lower percentage of lymphocytes [5, 14]. Elderly subjects have been reported to have increases in lymphocytes and neutrophils in their differential cell counts, suggesting that advanced age may affect BAL cell differentials [15, 16]. Additionally, the total volume of retrieved fluid declines with advanced age due to the loss of elastic recoil and increased compliance of the aged lung. Some investigators have shown that total volumes of instilled saline that range from 100 to 250 mL appear to give similar cell differentials in individual patients with ILD [17, 18].

Infection can cause the subacute onset of diffuse lung infiltrates or coexist with noninfectious ILD. Therefore, BAL should be examined and screened for mycobacterial or fungal infection if clinical suspicion of an infectious aetiology warrants such testing. If bacterial infection is a possibility, an uncentrifuged BAL fluid specimen can be sent for quantitative bacterial culture. Uncentrifuged BAL fluid can also be sent for viral culture, although a centrifuged specimen may enhance virus detection via stains or viral nucleic acid probes. Centrifuged BAL specimens can also be used to culture Legionella, Mycoplasma and Chlamydia spp., mycobacteria and fungi, and cytocentrifuged specimens can be stained to detect the presence of intracellular bacteria, Pneumocystis jiroveci, mycobacteria or fungi [7].

BAL CHARACTERISTICS AND THE DIAGNOSIS OF SPECIFIC TYPES OF ILD

When performed with a standardised technique, expertly examined, and combined with clinical and imaging data, BAL differential cell counts and other characteristics can provide important information that contributes significantly to the diagnosis of specific ILDs (table 1) [1–3, 10, 19–22]. Additionally, cytopathological examination may be useful to detect malignancies (e.g. lymphangitic carcinoma and lymphomas) that can present as diffuse lung disease with a similar appearance to nonmalignant ILD on thoracic imaging [23]. Other observations, including the gross appearance of BAL fluid and identification of unusual cells (e.g. plasma cells, mast cells and foamy AMs), or foreign material can yield important information that may lead to a specific diagnosis. In the appropriate clinical setting, certain gross and cellular findings in BAL are highly suggestive or even virtually diagnostic of specific ILD entities (table 2), but these observations must be interpreted in the context of the patient’s clinical presentation and radiological findings. Additionally, clinicians should recognise that radiographic imaging, including HRCT, may not appear particularly abnormal when certain forms of ILD (e.g. non-IPF interstitial pneumonias or HP) are present. If BAL is performed on a symptomatic patient who has radiographic imaging that is not particularly suspicious for the presence of ILD, an abnormal cell profile consistent with the presence of an “alveolitis” can indicate the need for additional investigation including lung tissue biopsy.

When BAL fluid is obtained and analysed from patients with ILD, differential cell counts obtained from slides prepared via cytocentrifugation usually show variations in the BAL cell differential that differ from patterns found in normal subjects. These patterns tend to reflect inflammatory cell profiles in affected lung tissues [1], but cell patterns may be obtained that are not typical for a specific ILD diagnosis, show little or no change from normal profiles or show mixed patterns with significant changes in the relative percentage of more than one type of nucleated immune cell [2]. Additionally, altered BAL cell profiles can be observed in various airway disorders (e.g. bronchiolitis, bronchitis, asthma, eosinophilic bronchitis, bronchiectasis and allergic bronchopulmonary aspergillosis). Nonetheless, certain cell patterns with significant changes in one or more BAL cell type can be highly consistent with or strongly suggest the presence of specific types of ILD. A recently published study of the predictive value of BAL for ILD diagnosis for a large cohort of patients (n=3,118) suggested that BAL cell counts are most useful for relatively common ILD diagnoses, such as sarcoidosis, in contrast to relatively rare forms of ILD where BAL differential cell counts tend to have less diagnostic utility [24].

A BAL lymphocyte differential count >25% is quite likely to be caused by ILD associated with granuloma formation (e.g. sarcoidosis and HP) or drug toxicity, if other possibilities, such as mycobacterial or fungal infection, are excluded. However, prominent BAL lymphocytosis can also be observed with other entities, such as cryptogenic organising pneumonia (COP), cellular nonspecific interstitial pneumonia (NSIP), lymphoid interstitial pneumonia (LIP) or non-LIP lymphoproliferative disorders. Extreme increases in BAL neutrophils are probably caused by infection, or relatively acute and diffuse lung injury. Eosinophil differential cell counts ≥25% are highly likely to be caused by eosinophilic lung disease, especially eosinophilic pneumonia (EP) if the presentation is acute [25, 26]. Increased numbers of mast cells have been associated with HP, drug reactions, sarcoidosis, ILD associated with collagen vascular disease, IPF, COP, EP and malignancy. Plasma cells have been observed in BAL in HP, drug reactions, EP, malignancy and infection [27]. AMs may also display certain morphological changes, such as: a foamy appearance in HP; markedly vacuolated cytoplasm with positive staining of vacuoles for fat in chronic aspiration pneumonitis; cytoplasmic inclusions associated with viral infection (e.g. cytomegalovirus pneumonia); ingested red blood cells (RBCs), and RBC fragments and haemosiderin with diffuse alveolar haemorrhage (DAH); ingested asbestos bodies; or other dust particles.

Grossly bloody lavage fluid, which may range from pink to red with acute haemorrhage, or to orange-brown if subacute, is characteristic of DAH, especially if sequentially retrieved BAL fluid aliquots do not show any decrease in the amount of bloody discoloration of the lavage fluid or (especially) if the discoloration increases [28]. If DAH is present, RBCs should be identifiable on the cytospin and AMs will stain positively for haemosiderin when stained with an iron stain if the onset of haemorrhage has preceded the time of BAL by 24–48 h. Freshly retrieved BAL fluid that has a milky or light brown to whitish, cloudy appearance with flocculent debris that settles out of the fluid to the bottom of its container prior to centrifugation suggests pulmonary alveolar proteinosis (PAP) as the likely diagnosis [29]. The diagnosis can be confirmed if the sediment, which is surfactant-derived lipoproteinaceous material, stains positively with periodic acid–Schiff staining.

Flow cytometric analyses of BAL cells have been reported by numerous investigators [2, 24, 27] and alterations in BAL lymphocyte subsets have been examined extensively, especially for sarcoidosis [1, 2, 30–32]. The presence of a high CD4+/CD8+ T-lymphocyte ratio increases the likelihood of diagnosis of sarcoidosis when combined with BAL lymphocytosis, although the clinician must consider age as a factor in the elevated CD4+/CD8+ ratio if the patient is elderly [16]. Nonetheless, it is debatable whether BAL lymphocyte subset data provides a clinical tool that contributes significantly to ILD diagnosis. The BAL CD4+ T-lymphocyte subset and CD4+/CD8+ T-lymphocyte ratio tend to be elevated in clinically active pulmonary sarcoidosis [31]. CD4+/CD8+ lymphocyte ratios in BAL from clinically healthy, younger adults usually range from 1.0 to 3.5, with average values between 1.5 and 2.0. An elevated CD4+/CD8+ ratio of ≥3.5 is relatively specific for sarcoidosis [32, 33]. However, the sensitivity of an increased CD4+/CD8+ ratio is relatively low for sarcoidosis [32, 33] and many patients do not have an elevated ratio or may even have a low ratio [34]. A depressed CD4+/CD8+ ratio has been observed in HP, drug-induced lung disease, COP, EP and IPF [27], but this ratio has not been shown to contribute meaningfully to the diagnosis of non-sarcoid ILD. Use of the CD4+/CD8+ ratio as a routine component of BAL analysis is questionable and may increase the cost of the procedure considerably.

Flow cytometry may be useful to support a diagnosis of pulmonary Langerhans cell histiocytosis (PLCH) if clinical data and imaging suggest this entity. PLCH is associated with increased (>4–5% of total cells) CD1a+ cells in BAL, although this may no longer be the case in later stages of disease [35, 36]. However, macrophage autofluorescence may present insurmountable difficulties in obtaining reliable data and immunohistochemistry on prepared slides is more likely to provide reliable enumeration of the percentage of positively stained Langerhans cells [36]. Flow cytometry and/or immunocytochemistry can be useful in identifying monoclonal lymphocyte populations in order to support the diagnosis of a lymphoid malignancy, and should be obtained if clinical data suggest the possibility of pulmonary lymphoma and the clinical laboratory can perform appropriate analyses. Although flow cytometric analysis of BAL cells may provide useful information in certain situations, the routine use of flow cytometry to analyse BAL cell subpopulations to assist in the diagnosis of specific forms of ILD is of unclear benefit. However, flow cytometry of BAL cells can be obtained if it is likely to provide important diagnostic information.

One application of BAL that evolved over the past decade was the suggested incorporation of the BAL cellular findings into the major and minor criteria for the clinical diagnosis of IPF when surgical lung biopsy is not performed [37]. However, since the positive predictive value of the HRCT findings of the usual interstitial pneumonia (UIP) pattern for a diagnosis of IPF/UIP is ≥85% and a confident diagnosis of UIP by experienced radiologists can be made in 95% of cases with the characteristic pattern of UIP on HRCT images [38–41], invasive procedures including bronchoscopy, BAL and transbronchial lung biopsy may not be required to make a confident diagnosis of IPF in the appropriate clinical setting [9]. In fact, the features of the criteria for the UIP pattern has been precisely described in the new joint American Thoracic Society (ATS)/European Respiratory Society (ERS)/Japanese Respiratory Society (JRS)/Asociación Latinoamericana del Tórax (ALAT) statement; the new guideline states that the presence of the described UIP pattern on HRCT imaging is sufficient to make a diagnosis of IPF in the appropriate clinical setting and eliminates the need for BAL cellular analyses [42, 43]. However, an increased differential lymphocyte cell count in BAL from patients whose HRCT demonstrates a UIP pattern may raise the possibility of chronic HP or other diagnoses associated with BAL lymphocytosis. In a recent study, 8% of patients with a seemingly confident clinical and HRCT diagnosis of IPF had >30% lymphocytes in BAL, an atypical finding for the diagnosis of UIP/IPF that suggested HP or another disorder as the actual diagnosis [44]. The new evidence-based guideline for diagnosis and management of IPF remarks on this point and defers the decision for the need of BAL cellular analyses to individual experts with established BAL laboratories (that provide reliable cell differential counts) who may opt to use BAL cellular analyses in patients undergoing evaluation for the diagnosis of IPF at their centres to ensure the lack of findings that suggest a non-IPF form of ILD as the diagnosis [42].With the increasing use of HRCT to evaluate patients during the early stages of ILD, it is likely that patients with ILD will not manifest the precise pattern of UIP on HRCT. In this setting, many clinicians may routinely perform BAL for cellular analyses as a part of the initial evaluation of patients presenting with early- and new-onset ILD. Thus, the clinical utility of BAL cellular analyses as an useful intervention in narrowing the differential diagnoses of ILD in this setting is evident. It must, however, be recognised that for an accurate diagnosis, biopsy of the tissue with adequate sampling may still be required. The performance of transbronchial biopsy in the peripheral lesions of UIP, NSIP and desquamative interstitial pneumonia is not advised.

BAL can also assist in the diagnosis of acute-onset ILD (defined as a combination of illness ≤4 weeks in duration, shortness of breath, hypoxaemia and diffuse radiographic infiltrates in a patient with no history of prior lung disease and no obvious risk factors for acute respiratory distress syndrome, such as sepsis or trauma). Diagnostic considerations in acute-onset ILD include infection and noninfectious ILD (acute interstitial pneumonia, acute EP, DAH, acute HP, acute COP, drug toxicity or acute exacerbation of previously undiagnosed IPF). Examination of BAL fluid can detect infection or haemorrhage, and the BAL cell profile may show large numbers of eosinophils, which are strongly supportive of a diagnosis of acute EP. Additionally, large numbers of lymphocytes would suggest entities such as acute HP or drug toxicity, especially if accompanied by plasma cells and an appropriate exposure history. Bronchoscopy with BAL at the time of acute presentation may facilitate diagnosis and minimise procedural risk if a diagnosis can be made, and obviate the need to progress to a surgical lung biopsy.

BAL AND THE MANAGEMENT OF ILD

A role for BAL cellular analyses in the treatment of ILD (e.g. to initiate, direct and/or monitor pharmacological therapies) has not been established. One application of BAL to disease management, however, is its use to evaluate acute changes in symptoms and lung function. Several adverse events may occur in the course of interstitial diseases, including respiratory infection, drug reactions, haemorrhage or an acute exacerbation of the disease process, and BAL may help to identify the cause of clinical deterioration and facilitate timely intervention.

Many investigators have examined the role of BAL cellular profiles for monitoring disease activity and response to therapies in specific ILDs, and BAL cellular findings at the time of diagnosis have been reported by some investigators to reflect disease severity and predict the likelihood of disease progression. The degree of increase in BAL neutrophils has been correlated with disease severity and prognosis for both HP [45, 46] and IPF [37, 47–49], and BAL eosinophilia has been linked to more severe disease and worse prognosis in IPF [50, 51]. Similarly, the presence of increased neutrophils in BAL from patients with sarcoidosis has been associated with more progressive disease that is less likely to respond to immunosuppressive therapy [52]. BAL lymphocytosis has been linked to prognosis and response to therapy for some forms of idiopathic interstitial pneumonia (IIP). BAL findings from patients diagnosed with IPF when this diagnostic term did not necessarily exclude other IIPs, such as NSIP, suggested that higher BAL lymphocyte differential cell counts correlated with better prognosis and/or response to therapy [37, 48–50, 53]. A more recent study, although retrospective, demonstrated that BAL lymphocytosis in patients who appear to have fibrotic IIP suggests that the diagnosis is not IPF but rather NSIP, and BAL lymphocytosis was found to correlate with a better prognosis [54].

Lung lavage was initially used as a therapy for septic lung disease and PAP [55, 56]. Although its therapeutic application is quite limited at this time, whole-lung lavage (WLL) is considered standard of care for PAP [57]. However, standardisation of technique and controlled clinical trials to determine the efficacy and durability of WLL as a therapy for PAP are lacking [57].

FUTURE DIRECTIONS AND RESEARCH QUESTIONS

BAL holds the potential to play an increasingly useful role in both the diagnosis and monitoring of patients with ILD, but many important issues have not been adequately resolved. These unresolved questions include the following. 1) What is the best technique for performing BAL (aliquot volume and number of aliquots; type and magnitude of negative pressure to retrieve instilled fluid; and pooling of aliquots for subsequent analysis)? 2) When and how should HRCT be used as a guide to determine the site of BAL in patients demonstrating diffusely distributed lesions in routine chest radiographs? 3) Should microbiological testing be performed whenever BAL is performed for cytological analysis to determine BAL cell total and differential cell count? 4) Should cytological analysis to detect malignant cells be performed whenever BAL is performed to evaluate diffuse lung disease? 5) Who should interpret BAL cellular preparations? 6) Can BAL combined with transbronchial lung biopsy synergistically increase diagnostic power and diminish the need for surgical lung biopsy? BAL may prove particularly helpful as treatments become available that may alter the profiles of BAL cells and/or noncellular components of lower respiratory tract bronchoalveolar surface liquid; if patients have favourable responses to such therapies, detection of these changes may support continued, ongoing treatment with specific agents. Additionally, multicentre studies that prospectively compare BAL with HRCT and histological patterns in large groups of patients will be facilitated by improvements in BAL standardisation [11]. Such studies have the potential to better define the underlying inflammatory and histological events associated with different HRCT appearances, which may reduce the need for biopsies and provide information that can guide effective therapies.

New tools that use genomic and proteomic characterisation of BAL cells and soluble components may eventually prove extremely useful in making an accurate ILD diagnosis, choosing and implementing potentially effective therapies, monitoring disease activity, and assessing the effect of therapeutic pharmacologic interventions. The recent development of DNA microarrays and other technologies that identify and monitor expression patterns of vast numbers of genes promises to provide important information about the pathobiology of various types of ILD [58, 59]. Similarly, recent advances in proteomics that allow mapping and identification of multiple protein expression patterns by two-dimensional (2D) gel electrophoresis, image analysis, protein spot transfer and mass spectrometry now raise the real hope that the products of genes linked to disease pathogenesis can be identified and quantitated [60].

Determining gene and protein expression patterns over time, as reflected in BAL, can potentially identify the key molecules involved in the initiation and progression of the different ILDs, provide an accurate clinical diagnosis of specific ILDs without resorting to lung biopsy, and indicate targets for new and effective therapies. A global analysis approach has identified many genes that display increased expression in IPF [58]. Additionally, certain gene products that may play an important role in IPF pathogenesis, including pigment epithelium-derived factor [61], matrilysin [62] and osteopontin [63], have been identified via these methods and may serve as disease biomarkers. Thonhofer et al. [64] demonstrated that stimulated BAL cells from patients with sarcoidosis displayed up- or downregulation of >1,000 genes, including selective upregulation of B-MYB, a potent growth factor for lymphocytes and regulator of apoptosis, and FABP4, a regulator of lipid metabolism and arachidonic acid uptake by macrophages. Not only can microarray genetic analysis provide characteristic gene expression patterns for specific ILD, but distinct expression patterns may differentiate one ILD from another. Selman et al. [65] recently validated the concept that gene expression patterns can be characteristic of specific ILDs by demonstrating that distinct gene expression patterns in lung tissue differed for patients with IPF versus those with HP.

A proteomics approach may also revolutionise the utility of BAL analysis for ILD diagnosis and management [60]. Characterising and quantitating proteins in biological specimens allow one to circumvent modifications of RNA that affect protein production and expression, and proteomic analytic techniques can be applied to acellular compartments, such as epithelial surface liquid retrieved via BAL or peripheral blood plasma. Early investigations that used electrophoretic techniques to examine protein profiles in BAL were able to show some differences between IPF and sarcoidosis [66, 67], and later studies using 2D electrophoretic techniques allowed enhanced fingerprinting of digested proteins from BAL supernatant fluids and demonstrated different profiles for IPF, sarcoidosis and HP [68, 69]. More recently, Rottoli et al. [70] used 2D electrophoresis to construct protein maps from BAL fluid that were able to demonstrate differences between IPF, systemic sclerosis and sarcoidosis. The development of techniques that employ mass spectrometry combined with ionisation of peptides via matrix-assisted laser desorption–ionisation or electrospray ionisation, or mass spectrometry combined with the use of surface chromatography to capture proteins on a chip surface (surface-enhanced laser desorption–ionisation) coupled with greatly enhanced computational abilities and proteomic databases hold considerable promise for the study of ILD via this emerging technology, and these techniques have recently been applied to the study of markers in BAL fluid in patients with lung inflammation [71].

SUMMARY AND CONCLUSIONS

Although investigators have recognised the importance of the BAL procedure in clinical pulmonary medicine, unresolved questions concerning the BAL procedure and the clinical utility of BAL cellular analyses for the evaluation and management of ILD need to be addressed, and consensus reached among international experts. The technical aspects of performing BAL and analysing cellular features of BAL fluid need standardisation, and the status of the clinical utilisation of BAL findings in day-to-day clinical practice needs to be clarified so that principles of performing, analysing and interpreting BAL can be applied globally, avoiding wide variations in BAL technique and specimen analysis. BAL cellular patterns and other BAL findings can play an important role in the diagnosis of many forms of ILD, especially when these findings are obtained via a properly performed BAL procedure and analysis, and then combined with clinical data and HRCT imaging. In many instances, the combination of clinical information (history, examination and pulmonary function testing), HRCT imaging and BAL analysis can provide enough information to allow the clinician to be assured of a confident diagnosis and avoid more invasive testing to obtain lung tissue for histopathological diagnostic examination. BAL may be particularly beneficial for the fragile patient or patients at high risk for significant complications if subjected to surgical lung biopsy.

Universal adoption of a standardised BAL procedure will undoubtedly improve the chance of identifying clinically relevant differences between specific disorders and provide results that can be compared from centre to centre. A standardised procedure will enable centres to pool information to achieve the large patient groups necessary to more adequately investigate the value of BAL markers for the identification of specific disease entities, determining clinical disease activity, detecting disease progression, providing a prognosis and guiding treatment strategies. Identifying specific and sensitive markers that can address these problems remains an elusive but vitally important goal. Future application of BAL to the evaluation of ILD in concert with an evolving, improved understanding of the pathogenesis of specific ILDs, and the development of novel testing that uses genomic and proteomic characterisation of BAL fluids holds the potential to provide accurate diagnostic and prognostic information for patients with ILD and potentially play a role in choosing novel therapies for specific forms of ILD. Despite this significant diagnostic potential, very few patients have been investigated so far using gene and protein mapping approaches, and applications of the powerful techniques of gene and protein mapping to the study of ILD are still at a very early stage but have the potential to evolve rapidly in the coming years. Access to large collections of well-characterised clinical samples (BAL, lung biopsies and peripheral blood) to which these techniques can be applied, combined with HRCT or newer forms of thoracic imaging, will greatly facilitate our understanding of basic and clinical aspects of ILD pathobiology, and facilitate the identification of clinically useful biomarkers.