There are wide variations in the preparation of histologic sections from endoscopic esophageal biopsy specimens. We evaluated serial step sections from 261 esophageal or gastroesophageal junction biopsies at 4 levels to determine the first level at which goblet cell metaplasia (GCM) was detected. Deeper step sections of 152 paraffin blocks also were obtained to determine whether additional sections are useful in detecting GCM not seen in initial levels. GCM was identified in 95.3% of blocks in 3 levels. GCM was seen at level 4 in 12 blocks (4.7%). In the blocks that did not reveal intestinal metaplasia in the initial 4 levels, deeper sections disclosed GCM in only 1 (0.8%) of 120 blocks. However, deeper sections revealed initially undetected GCM in 4 of 32 blocks from patients with a history of documented Barrett esophagus. We conclude that 4 levels of step sections are adequate in routine processing of esophageal biopsy specimens for demonstration of GCM. Deeper sections may be obtained for patients with known Barrett esophagus to better evaluate for dysplasia or find additional foci of GCM.

The value of cultures of tissue obtained by image-directed core needle biopsy of lung nodules has not been determined. Of the 250 biopsies performed during a 5-year period in an area endemic for coccidioidomycosis, 225 (90%) were diagnostic. Granulomas were identified in 75 specimens, whereas 3 specimens revealed abscess. Ziehl-Neelsen stain was positive for acid-fast bacilli (AFB) in 2 cases. Spherules of Coccidioides immitis were seen in 54 of the biopsy specimens with granulomas. Microbiologic cultures were positive for C immitis in 5 (9.6%) of 52 biopsy specimens.Image-guided percutaneous core needle biopsy (CNB) is being increasingly performed in the diagnosis and treatment of solitary and multiple pulmonary nodules. Fine-needle aspiration (FNA) biopsy, which yields material for cytologic evaluation, is an established method with high sensitivity and specificity in the diagnosis of malignancy 1 ; cytologic examination alone, however, may not be definitive in categorizing a benign disease process.2 Since the initial report on the efficacy of percutaneous CNB in the United States by Parker et al 3 in 1989, many radiologists have advocated the use of the automated biopsy gun to obtain tissue samples for definitive histologic diagnosis. [4][5][6] Because the main differential diagnosis of a localized pulmonary lesion is an infectious process or a neoplasm, a portion of the CNB tissue is usually submitted for microbiologic cultures in an attempt to isolate and identify the causative infectious agent. tion, 10 have been found valuable, the usefulness of culturing a CNB specimen has not been evaluated.This study was conducted to determine the relative sensitivities of cultures and special stains for microorganisms in histologic sections of CNB specimens of pulmonary nodules. MATERIALS AND METHODSBetween July 1990 and June 1995, 250 imagedirected percutaneous CNBs were performed in the radiology department of the Carondelet St Mary's Hospital in Tucson, Ariz, an area endemic for coccidioidomycosis. A 20-gauge spring-loaded automated biopsy gun (Monopty; Bard, Covington, Ga) was used to obtain specimens for histologic evaluation. All biopsy procedures were performed by the coaxial technique, with initial localization accomplished by an 18-gauge thin-walled guiding needle. Fluoroscopic guidance was used in 237 procedures, and 13 were directed by computed tomography.The biopsy samples were collected and delivered in saline solution to the surgical pathology laboratory. A portion of the tissue was submitted to the microbiology laboratory for culturing; the remaining specimen was fixed in formalin and routinely processed. During sectioning, two slides containing contiguous ribbon of tissue were prepared at each level. One of the slides, representing one level of sectioning, was stained with hematoxylin-eosin,