To the Editor.—Although many pulmonologists are adopting transbronchial cryobiopsy, there is a remarkable dearth of information on the specific pathologic features used to diagnose various forms of interstitial lung disease in such specimens. Cooper et al1 have recently published pathologic criteria for usual interstitial pneumonia (UIP); they suggest that the combination of fibroblast foci and patchy fibrosis, along with the absence of features suggesting an alternative diagnosis, strongly support a diagnosis of UIP in a cryobiopsy.
We2 reported in Archives of Pathology & Laboratory Medicine an in-silico exercise in which we created “cryobiopsies” by outlining multiple circles on the pathology slides of 15 video-assisted thoracoscopic surgery (VATS) biopsies from cases that had been given a 60% or greater probability of fibrotic hypersensitivity pneumonitis (fibrotic HP) during a specialty devised multidisciplinary discussion exercise.3 We found a low probability of detecting granulomas or giant cells, features acknowledged to be markers in this context of fibrotic HP and not UIP. However, the probability of finding peribronchiolar metaplasia affecting 50% or greater of bronchioles, a feature that also supports a diagnosis of fibrotic HP,3 was considerably higher, and the frequency of these features within an individual case generally increased with increasing numbers of “cryobiopsies.”
Here, we went back to our previously used VATS biopsy slides and asked whether the features described by Cooper et al1 as favoring a diagnosis of UIP could be found in our fibrotic HP-derived “cryobiopsies.” We used the same “cryobiopsies” previously selected for analysis of granulomas/giant cells and peribronchiolar metaplasia, and counted the number of samples showing fibroblast foci, and patchy fibrosis in 1 to 8 “cryobiopsies” for each case; these features had not been evaluated in our previous study.2 For comparison, the total biopsy area of 86.4 mm2 for 4 of our “cryobiopsies” is identical to the mean total area of 87 mm2 reported in the COLDICE study.4
Results are shown in the Table. Fibroblast foci were frequently found, from 8 of 15 (53%) cases if a single “cryobiopsy” was evaluated for each case, to 12 of 15 (80%) if 4 biopsies were evaluated, and 12 of 13 (92%) with 8 biopsies (2 cases did not have enough VATS biopsy area to create 8 “cryobiopsies”). Patchy fibrosis ranged from 4 of 15 (27%) single biopsies, to 8 of 15 (53%) with 4 biopsies, and 12 of 13 (92%) with 8 biopsies. The combination of fibroblast foci and patchy fibrosis was found in 1 of 15 (6.7%) cases with a single biopsy and increased to 7 of 15 (47%) cases when 4 biopsies were reviewed and 10 of 13 (77%) with 8 biopsies, suggesting the potential to misclassify almost half of fibrotic HP biopsies as UIP with 4 biopsies and the majority of patients with 8 biopsies using these criteria alone. This situation is improved by also looking for giant cells/granulomas or peribronchiolar metaplasia affecting more than 50% of bronchioles, but still leaves a significant number of fibrotic HP cases misclassified as UIP (Table).
This is obviously a small series that uses a somewhat artificial system. Nonetheless, these data emphasize the idea that a combination of fibroblast foci and patchy fibrosis is a feature not only of UIP but also of fibrotic HP, and used by themselves, these features will lead to misclassification of a proportion of fibrotic HP cases as UIP.
Although it may be surprising to nonspecialist pathologists (and to many clinicians), the exact pathologic features that define UIP, even in VATS biopsies, are not universally agreed upon, as discussed in detail in 2 recent position papers,5,6 and this is a problem that needs to be addressed. That being so, it is even harder to determine the features that should be used to separate conditions, such as fibrotic HP, which often have considerable morphologic overlap with UIP.
Granulomas/giant cells are generally accepted as features against a diagnosis of UIP; extensive peribronchiolar metaplasia, in our view, favors fibrotic HP,3 and fibrosis that is predominantly peribronchiolar rather than subpleural also is in favor of fibrotic HP. But how to quantify and apply these criteria to actual cryobiopsies, and where one draws the line between UIP and fibrotic HP in such biopsies, is an important unanswered question, a problem also acknowledged by Cooper et al.1
Churg received honoraria from Boeringher Ingelheim Canada and Hoffmann LaRoche Canada for speaking at meetings on the diagnosis of interstitial lung disease. The other authors have no relevant financial interest in the products or companies described in this article.