Diagnostic Accuracy of Transbronchial Biopsy in Acute Farmer’s Lung Disease: Discussion

21 Mar

TBB is often recommended to “rule out” an alternative diagnosis in the clinical setting of FL. We challenge this recommendation. First, the results of our analysis showed that as much as 48.6% of the TBB specimens were read as nonspecific. Second, even when an alternate diagnosis (such as sarcoidosis) was made, FL could not always be excluded (Table 3). Although TBB can identify a high proportion of patients with sarcoidosis, its operating characteristics (both sensitivity and specificity) for the diagnosis of sarcoidosis have not been well characterized. For instance, a number of case series reporting on pathologic abnormalities in sarcoidosis have been published. In all these case series, granulomas were described in a high percentage (approximately 90%) of cases of stage III sarcoidosis, which may be difficult to distinguish from hypersensitivity pneumonitis. Although these results suggest that TBB is highly sensitive for the diagnosis of sarcoidosis, its specificity and, consequently, its likelihood ratio remain uncertain because of the absence of a control group.
As a matter of fact, the diagnostic accuracy (sensitivity and specificity) of TBB for most interstitial lung diseases remains undetermined, with few exceptions pertaining to specific clinical circumstances including the investigation of pulmonary infections in patients with AIDS and of acute lung rejection in lung transplant recipients. canadian neightbor pharmacy

Like Wall et al, we were also particularly concerned with the high rate of nonspecific test results that we obtained (as much as 48.6%). To this high rate of nonspecific test results, we must add the proportion of patients on whom TBB was attempted without providing the pathologist with adequate material (<10% in our institution; unpublished data; 1994). This figure is, however, not very different from many previous publications reporting on the diagnostic accuracy of TBB and from others articles in which this issue of management of patients after such a result has been addressed.
The results reported in this study question our practice to perform TBB on most of our patients with suspected FL. The likelihood ratios associated with either the overall assessment of the biopsy specimens or the specific pathologic criteria are rather modest, with the exceptions of diffuse lymphocytic infiltration and the combination of diffuse lymphocytic infiltration and granuloma. The low proportion of cases of FL in which diffuse lymphocytic infiltration is found, however, limits the usefulness of this criterion. Nevertheless, we believe that, under certain circumstances, TBB may still be useful in suggesting or supporting a diagnosis of FL: it should be reserved primarily for clinically equivocal cases (ie, for patients with intermediate pretest probability of FL) in order to significantly increase or decrease the likelihood that FL is present depending on the test result. Immunocytochemical analysis of TBB specimens may be useful to assist differentiating FL from sarcoidosis; the adjunction of new immunologic markers to the usual hematoxylin-eosin staining for the analysis of TBB might further increase diagnostic accuracy of the test without additional discomfort or risk to the patients undergoing the biopsies.