A definitive diagnosis of liver disease often requires a representative specimen of the affected tissue. Methods used to obtain liver specimens range from those that are minimally invasive, such as fine-needle aspira- tion (FNA), to more invasive methods such as surgical biopsy. 1-5 Fine-needle aspiration and cytologic examina- tion can be a useful diagnostic tool and is commonly used for liver specimens. 3,6,7 It can be performed expedi- ently, with low risk to the patient, and usually without sedation or general anesthesia. However, FNA and cyto- logic examination of the liver does have limitations, including failure to identify the primary disease and inability to provide information on parenchymal archi- tecture. 8-10 The accuracy of cytologic examination as a diagnostic tool has been evaluated in the veterinary liter- ature; however, limited data are available regarding the liver. Roth 8 reported either partial or complete agreement between cytologic and histologic diagnosis for the liver in 45 of 56 (80%) cases in dogs and cats, although the method of obtaining the cytology specimens in that study was not mentioned. In another study 2 that evaluat- ed the accuracy of cytologic examination via FNA or impression smears, compared with histopathologic find- ings, there was agreement between cytologic and histopathologic findings in 29 of 44 (66%) cases. However, of the cytologic specimens collected by FNA rather than as impression smears, there was agreement with histopathologic findings in only 2 of 11 cases. A recent report by Weiss et al 10 indicated good correlation between cytologic findings with FNA and histopatholog- ic findings when active inflammatory liver disease was present in dogs. Chronic active and suppurative inflam- mation were readily identified in dogs, whereas lympho- cytic hepatitis was commonly misdiagnosed. The accura- cy of cytologic examination may therefore vary depend- ing on the underlying type of hepatic disease. Noninflammatory hepatic diseases in dogs were misdiag- nosed via FNA cytologic examination more often than they were correctly diagnosed. 10 In addition, the presence of hepatic vacuolar change in cats with concurrent infil- trative liver disease may lead to a diagnosis of hepatic lipidosis, although the primary disease is overlooked. 9 Most liver FNAs are performed with ultrasound guid- ance, yet there are few reports comparing cytologic find- ings of this technique with histopathologic findings. The purpose of the study reported here was to evaluate the accuracy of ultrasound-guided FNA and cytologic find- ings of the liver of dogs and cats, compared with histopathologic findings, in the clinical setting. Criteria for Selection of Cases The medical records of cats and dogs evaluated from 1990 to 2000 at the Virginia-Maryland Regional College of Veterinary Medicine were reviewed for cases in which hepatic tissue was evaluated by use of ultra- sound-guided FNA and cytologic examination, as well as histopathologic examination. Procedures All cytologic and histopathologic specimens had been examined by a pathologist, faculty clinical pathol- JAVMA, Vol 224, No. 1, January 1, 2004 Scientific Reports: Retrospective Study 75 SMALL ANIMALS Accuracy of ultrasound-guided fine-needle aspiration of the liver and cytologic findings in dogs and cats: 97 cases (1990–2000) Kelly Y. Wang, DVM; David L. Panciera, DVM, MS, DACVIM; Raida K. Al-Rukibat, DVM, MS, DACVP; Zaher A. Radi, DVM, PhD Objective—To evaluate the accuracy of ultrasound- guided fine-needle aspiration of the liver and cytolog- ic findings in dogs and cats. Design—Retrospective study. Animals—56 dogs and 41 cats. Procedure—Medical records of dogs and cats evaluated from 1990 to 2000 by use of cytologic and histopatho- logic examination of the liver were reviewed. Histologic and cytologic diagnoses were categorized as vacuolar hepatopathy, inflammation, neoplasia, cirrhosis, primary cholestasis, shunt, normal, and other. Results—Overall agreement between the histopatho- logic diagnosis and cytologic diagnosis was found in 17 of the 56 (30.3%) canine cases and 21 of the 41 (51.2%) feline cases. Vacuolar hepatopathy was the category with the highest percentage of agreement. Vacuolar hepatopathy was identified via cytologic examination in 7 of 11 and 15 of 18 dogs and cats, respectively, in which histopathologic examination revealed that it was the predominant disease process. However, it was also the category that was most commonly misdiagnosed via cytologic examination. Inflammatory disease was accurately identified cytologically in 5 of 20 and 3 of 11 dogs and cats, respectively. Conclusions and Clinical Relevance—Acknowledging the limitations of cytology and the extent of discrepan- cies between cytologic and histopathologic findings in dogs and cats will help clinicians make better decisions in diagnosing liver disease. (J Am Vet Med Assoc 2004; 224:75–78) From the Departments of Small Animal Clinical Sciences (Wang, Panciera) and Biomedical Sciences and Pathobiology (Al-Rukibat, Radi), Virginia-Maryland Regional College of Veterinary Medicine, Virginia Tech, Blacksburg, VA 24061. Dr. Wang’s present address is the Department of Veterinary Clinical Sciences, College of Veterinary Medicine, The Ohio State University, Columbus, Ohio 43210. Dr. Al- Rukibat’s present address is the Department of Veterinary Pathology, Jordan University of Science and Technology, Irbid 22110, Jordan. Dr. Radi’s present address is Veterinary Diagnostic and Investigational Laboratory, University of Georgia, Tifton, GA 31793-3000. Address correspondence to Dr. Wang.