Material: FL vs Control Group
Pathologic specimens obtained between 1980 and 1993 from consecutive patients who had TBBs when initially investigated during an acute phase of FL were selected. The FL specimens were matched to TBB specimens from patients with diffuse parenchymal diseases other than hypersensitivity pneumonitis. These control cases were randomly selected by a research assistant not otherwise involved in the study using the 1980 to 1993 records of the endoscopy unit. Our assumption was that a random sample of control subjects among all the patients who had TBBs would lead to comparison of FL with an appropriate spectrum of patients to whom the diagnostic test is usually applied in clinical practice. The diagnosis and the number of cases (FL and control subjects) were known only by the investigator who gathered the material (Y.L.). Only biopsy specimens with adequate material were considered. Adequate was defined as at least three small biopsy fragments with identifiable lung parenchyma, one medium size and one small fragment, or one large fragment. sildenafil citrate pink
Routine protocol for biopsy and processing of the material was similar for all patients, whatever the suspected diagnosis. Bronchoscopies were done under topical anesthesia with 4% lidocaine and oxygen supplement. Under fluoroscopic guidance, four to eight TBB specimens were taken, generally in the right lower lobe, with a 2.6-mm forceps (Olympus FB-19; Lake Success, NY). Each biopsy fragment was fixed in Bouin’s solution, cut at 4 μm, and stained with hematoxylin-eosin. When indicated, special stains were utilized to clarify the nature of the pathologic abnormality. However, the pathologists were provided only with material stained with hematoxylin-eosin.
Interobserver Agreement: Two pathologists (R.S.F. and M.F.) independently assessed the biopsy specimens. Both pathologists were blinded to the patients’ final diagnoses.
Pathologic Criteria: The selection of criteria was first based on the description of FL histopathology by Reyes et al. We elected to analyze only lymphocytic interstitial infiltrates and granulomas, the two most common findings in the series of Reyes et al, as pathologic criteria (Fig 1). Lymphocytic infiltrates were further subdivided into “focal” or “diffuse.” Other pathologic findings reported by Reyes et al were rejected as diagnostic criteria for different reasons related to the nature of TBB: pleural fibrosis cannot be observed on TBB specimens; edema is patchy in distribution; and bronchiolitis obliterans is a rare finding on TBB specimens.
Figure 1. Diffuse lymphocytic infiltration and clusters of epithelioid cells (arrows) in farmer’s lung (hematoxylin-eosin staining; original magnification X300).