Bad Performance of Lung Cryobiopsy in the Diagnosis of Interstitial Lung Diseases: Don't Throw the Baby Out with the Bathwater.

To the Editor:

Multidisciplinary management is the gold standard for interstitial lung disease (ILD) diagnosis and treatment (1). In the recent American Thoracic Society/European Respiratory Society/Japanese Respiratory Society/Latin American Thoracic Association guidelines, experts did not make any recommendation for or against transbronchial lung cryobiopsy (TBLC), mostly due to a lack of strong data and the absence of standardization for the procedure. To date, most experts agree that TBLC provides a proper diagnosis in 80% of cases (2, 3), and data suggest that cryobiopsy can have a significant impact when performed in the setting of multidisciplinary management of ILD (4). In a recent issue of the Journal, Romagnoli and colleagues reported the first study to directly compare surgical lung biopsy (SLB) with TBLC for the diagnosis of ILD (5). After samples were read and a diagnosis was made by local pathologists, the samples were blinded and read by an external expert pathologist. The results revealed poor concordance between the two techniques as compared with the final diagnosis retained by local teams, which clearly casts a shadow on the spreading use of TBLC.

The authors must be acknowledged for conducting the first study on sequential SLB and TBLC. However, some points should be noted to preclude any hasty conclusions. First, the fact that only 21 patients were included does not allow for a strong statistical analysis. In addition, when we look at each case, some of the apparent discrepancies were expected, as we know that some patients have two different pathology patterns in their lungs (6). On the other hand, some differences between SLB and TBLC in a single patient are quite surprising (e.g., in patient #15 in the study, TBLC showed Langerhans histiocytosis and SLB showed usual interstitial pneumonia).

The authors report that TBLC would have led to a different treatment if SLB had not been performed in 11 of 21 cases (52%). However, when we analyze Table E3 in the online supplement of their study, this seems to be the case for only eight of 21 cases (38%).

Finally, the most striking result of the study is that neither SLB nor TBLC analysis by the blinded pathologist achieved good agreement with the final diagnosis (62% and 48%, respectively, with a wide confidence interval). This point illustrates the fact that we should consider the pathologist as one actor, among others, in the multidisciplinary assessment of ILD (7). In this regard, it would have been interesting to have the agreement between the final diagnosis and TBLC or SLB analyzed by the local pathologist, and to compare the diagnostic performance of the local pathologist (taking part in the multidisciplinary discussion) with that of the blinded expert pathologist.

In conclusion, although the study demonstrates low agreement between blinded analyses of SLB and TBLC and the final diagnosis, the results should not prevent specialists from performing TBLC in the setting of specialized multidisciplinary management of ILD. We also think that this important work by Romagnoli and colleagues paves the way for future trials comparing SLB and TBLC in a multidisciplinary setting.