Two recent publications (Willard et al, JAVMA 220:1177-1182, 2002 and Cole et al, JAVMA 220:1483-1490, 2002) addressed the interobserver variation among histopathologic diagnoses of intestinal biopsies from dogs and cats, and the diagnostic comparison of needle and wedge liver biopsies from dogs and cats, respectively.

The rationale for these studies was twofold: to assess the degree of interobserver variation among histopathologic descriptions of biopsy specimens; and the consistency between evaluations of tissue specimens taken by percutaneous needle biopsy at endoscopy, and surgical wedge biopsy at laparotomy. Similar questions have arisen from these procedures in humans.

Intestinal Biopsies. Histologic slides from endoscopic or surgical biopsy specimens of 13 animals (10 dogs, 3 cats) taken from the duodenum, ileum, or colon were evaluated independently by 5 pathologists at 4 institutions. These colleagues, who did not know the tissue of origin, were asked to indicate whether slides were adequate for evaluation, and whether tissues were normal or abnormal. For abnormal specimens, they were to identify the main infiltrating cell type and whether infiltrates were mild, moderate, severe, or neoplastic.

Liver Biopsies. Specimens were evaluated from the livers of 124 animals (98 dogs, 26 cats) by needle biopsy and wedge biopsy from the same liver lobe during laparotomy or postmortem examination. Two needle biopsies were obtained from each animal. Three individuals independently scored the histologic features, and the definitive diagnosis was determined from the wedge biopsy results if at least 2 of the 3 examiners concurred.

Intestinal Biopsies. All 14 slides were judged to be adequate or superior in quality by at least 4 of the 5 pathologists. However, the principle infiltrating cell type was agreed upon on only 6 of the 14 slides, and the severity of the infiltrate was judged concordantly for only 6 of the slides. No consistent pattern of ranking the severity of the lesions among pathologists was found. There was generally poor agreement between the pathologists that was statistically greater than expected by chance alone. Only 2 of the 14 slides were judged uniformly with regard to type and severity of infiltrating cell type, whereas with 3 slides there was nonuniformity of these criteria. For 5 slides, there was discordance between the clinical findings and the pathologists’ description. In 22 of 25 evaluations, tissues were described as abnormal when there was no evidence of intestinal disease.

Liver Biopsies. Tissue sections were excluded from 26 cases because of inadequate or unacceptable specimens; the remaining 124 cases were included in the study. Of these, the wedge biopsies revealed that 10 animals (6 dogs, 4 cats) had hepatic necrosis, 13 (4 dogs, 9 cats) had cholangitis-cholangiohepatitis, 12 (11 dogs, 1 cat) had chronic hepatitis-cirrhosis, 11 (all dogs) had vacuolar hepatopathy, 17 (16 dogs, 1 cat) had portosystemic vascular anomaly and microvascular dysplasia, 10 (8 dogs, 2 cats) had neoplasia, 18 (14 dogs, 4 cats) had miscellaneous hepatic disorders, and 33 (28 dogs, 5 cats) had no evident hepatic disease.

Morphologic diagnoses for needle and wedge biopsies were discordant for 66 of the 133 (50%) needle biopsy specimens, regardless of their length (≤ 5 to ≥ 10 mm long). Median histologic scores were significantly different in paired biopsy specimens, and were significantly lower (less severe changes seen) for needle biopsy than wedge biopsy specimens, except in the case of inflammation. Needle biopsy scores were significantly higher when inflammation was seen than wedge biopsy scores of the same liver lobe.

There was no consistent association between high and low histologic scores and individual evaluators. All 3 individuals agreed on the morphologic diagnosis of needle biopsies 44% of the time and for wedge biopsies 65% of the time. Overall, discordant diagnosis on the paired needle and wedge biopsies was found in 53% of dogs and 50% of cats.