Interstitial lung diseases comprise a heterogeneous group of conditions with variable natural history and treatment response. Through the integration of comprehensive clinical, serological, and radiological data within a multidisciplinary discussion (MDD), an underlying interstitial lung disease subtype can usually be identified.1, 2, 3 In particular, the high-resolution CT scan can provide detailed information on the probable disease pattern, with specific clinical context allowing for refinement of potential differential diagnoses. In up to 30% of cases, however, the high-resolution CT and clinical findings are not sufficient to allow for confident clinical diagnosis, requiring a surgical lung biopsy (SLB) for histopathological evaluation.4 The current accepted standard practice for obtaining SLB is through video-assisted thoracoscopic surgery (VATS), done by a thoracic surgeon in the operating room under general anaesthesia. This procedure poses substantial risks, including a reported 1·7% mortality in patients subjected to elective SLB.5 Although this mortality risk might be lower at centres with staff experienced in SLB, the decision to proceed with invasive biopsy should not be undertaken lightly. Poor cardiopulmonary reserve, advancing age, and comorbid disease often render patients with interstitial lung disease unsuitable candidates for SLB. Thus, diagnostic uncertainty will remain in a substantial proportion of patients with interstitial lung disease. For these patients requiring histopathological assessment, the benefit of obtaining the lung biopsy with a less invasive procedure than SLB is evident.
Research in context
Evidence before this study
Transbronchial lung cryobiopsy (TBLC) is an emerging technique to obtain lung tissue for diagnosis of interstitial lung disease. As a minimally invasive technique, TBLC has been adopted into clinical practice in many centres; however, it requires validation as an accurate diagnostic test. Due to an absence of high-quality evidence, diagnostic guidelines for idiopathic pulmonary fibrosis do not recommend for or against TBLC in its diagnostic algorithms. The safety of TBLC has been raised as a potential concern, with varying amounts of reported cases of pneumothorax and airway bleeding. Furthermore, the diagnostic accuracy of TBLC has not been verified in adequately powered prospective studies comparing TBLC against the accepted histopathological standard of surgical lung biopsy (SLB). We searched PubMed using the terms “cryobiopsy” or “cryoprobe” and “interstitial lung disease” or “diffuse parenchymal lung disease” or “pulmonary fibrosis”, for all clinical trials published from database inception up until July 8, 2019, with no language restrictions. Most of the publications were retrospective single-centre case series. The four systematic reviews of TBLC reported data for diagnostic yield and safety, but none assessed diagnostic accuracy. We identified only one small study that directly compared TBLC and SLB sampled sequentially from the same patients. The study retrospectively used a single pathologist to analyse the histopathological patterns after initial unmasked assessment and multidisciplinary discussion for clinical diagnosis, showing low concordance between the findings of the two forms of biopsy. Because of the very small sample size and other methodological limitations of this study, the issue of diagnostic accuracy of TBLC in interstitial lung disease diagnosis remains unresolved.
Added value of this study
The COLDICE study was a prospective, multicentre, investigator-initiated study designed to evaluate diagnostic accuracy of TBLC in interstitial lung disease diagnosis. The study was adequately powered to compare diagnostic agreement between TBLC and SLB obtained from the same patients at the same time from the same lobes, for both masked histopathological analysis, and for clinical diagnosis at multidisciplinary discussion. The study showed high concordance between the paired biopsy specimens for both histopathological pattern and multidisciplinary discussion diagnosis. The data from TBLC specimens were informative and reliable, particularly when high-confidence patterns were reported by the pathologist. Although our study was not designed to address the true safety aspects of TBLC independent of SLB, we did not find any new safety signals.
Implications of all the available evidence
To our knowledge, the COLDICE study is the first comparative study showing a high agreement between TBLC and SLB for interstitial lung disease diagnosis. Together with the data from case series, the evidence suggests that TBLC is a valid first-line diagnostic tool for patients with interstitial lung disease deemed to require histopathological diagnosis. Although further studies of a similar design would enrich the existing data, we appreciate that larger studies will be difficult to do in the clinical setting. Studies focusing on safety and standardisation of the TBLC procedure will be important adjuncts to clinical practice.
Transbronchial lung cryobiopsy (TBLC) has emerged over the past decade as an alternative diagnostic technique, with increasing use across many centres. The diagnostic yield, accuracy, potential safety, and health resource use advantages of TBLC over SLB are important considerations that need to be addressed in well designed studies. Before widespread implementation of TBLC can take place, it is necessary to directly compare histopathological interpretation of this small tissue sampling with larger SLB specimens obtained from the same patients. Although it is apparent that the diagnostic yield of TBLC is lower than that of SLB, TBLC might be established as a potentially safer, but reliable substitute if accuracy can be shown. To date, the literature reports a diagnostic yield for an identifiable histopathological pattern in 73–81% of TBLC specimens, compared with around 95% for SLB specimens.6, 7, 8 Although the TBLC yield is less than that of SLB, there is evidence to suggest a similar clinical utility, at a lower risk to the patient. For example, TBLC has been shown to affect diagnostic confidence to a similar degree as SLB, within the context of MDD.9 The diagnostic accuracy of TBLC, however, has not been addressed in a robust manner. This important issue can only be assessed through the direct verification of TBLC findings against SLB specimens, obtained from the same anatomical sites, from the same patients. Romagnoli and colleagues10 attempted to address this issue in a small interstitial lung disease cohort, observing poor agreement for TBLC and SLB histopathological analyses with a κ-concordance coefficient of only 0·22. Because of methodological limitations, including underpowering and the use of a single pathologist review after MDD diagnosis, these findings should be interpreted with caution. In the absence of rigorous direct comparison with the SLB, the role of TBLC in interstitial lung disease diagnostic algorithms remains unclear. The increasing dichotomy between the European and North American perceived role of TBLC (ie, generally, the technique has been used more in Europe than in North America), highlights the urgent need to settle the issue of the clinical utility of TBLC for interstitial lung disease diagnosis.11, 12 We therefore did the cryobiopsy versus open lung biopsy in the diagnosis of interstitial lung disease alliance (COLDICE) study, designed to evaluate the agreement between TBLC and SLB as a means of assessing diagnostic accuracy, at both histopathological assessment and at MDD.
