Hepatic diseases are relatively common in cats.

J Vet Intern Med 2006;20:1327 1333 High Complication Rate of an Automatic Tru-Cut Biopsy Gun Device for Liver Biopsy in Cats S.J.M. Proot and J. Rothuizen Background: Liver biopsies taken with an automatic Tru-Cut biopsy gun device caused unexpected fatal shock reactions in cats. The goal of the present study was to determine if this biopsy device caused more frequent fatal complications than did a semiautomatic device. Animals: All cats referred to the Utrecht University, between October 1, 2002, and October 31, 2004, in which ultrasoundguided Tru-Cut liver biopsies were taken. The indications for liver biopsy were increased liver enzyme activity, increased bile acid concentrations, ultrasonographic abnormalities of the liver, ultrasonographic abnormalities of the bile ducts, or some combination of these findings. Coagulation parameters were normal. Methods: From October 1, 2002, until October 31, 2003, 26 cats were biopsied with an automatic biopsy device. Between November 1, 2003, and October 31, 2004, 19 cats underwent liver biopsy with a semiautomatic biopsy device. Results: In the first period, 5 of the 26 cats (19%) developed severe shock within 15 minutes. Resuscitation was not successful. In the second period, none of the 19 cats experienced any major adverse effect. There were no significant differences between the 2 groups with respect to diagnosis, clinical signs, clinicopathologic findings, or the use of anesthetics. Conclusions and Clinical Importance: We conclude that the difference in complication rate is explained by the biopsy technique used. The pressure wave, which occurs when firing the automatic device, may have caused intense vagotonia and shock. Use of this automatic biopsy device should be avoided in cats because of the high risk of fatal complications. Key words: Fatal risk; Cat; Liver disease; Vagal shock. Hepatic diseases are relatively common in cats. Depending on the geographic location, cholangiohepatitis or idiopathic hepatic lipidosis is regarded as the most common hepatic disorder of cats. In the United States, hepatic lipidosis seems to be most common, while in the UK and Europe, neutrophilic or lymphocytic cholangitis is most common. 1 Other hepatic disorders of cats are feline infectious peritonitis (FIP), neoplasia, and amyloidosis. Ultrasonographic findings can be suggestive but are not diagnostic for any of these diseases. Fine needle aspiration biopsies often are helpful in diagnosing lipidosis, steroid-induced hepatopathy, and lymphoma, but histology of liver biopsies is required to diagnose most liver diseases. 2 Biopsies can be performed under ultrasound guidance or during laparoscopy or laparotomy. A surgical approach allows macroscopic evaluation of the liver, yields larger biopsy samples, and enables control of postbiopsy hemorrhage, but these techniques are more invasive. In contrast, ultrasoundguided biopsy is much less invasive and also permits visualization of the liver, bile ducts, biopsy tract, and the occurrence of hemorrhage. At the Companion Animal Clinic of Utrecht University, most liver biopsies in cats are performed under ultrasound guidance unless surgery is warranted for a specific reason. In the past, we used a semiautomatic 16G VET-core biopsy needle a (Fig 1). Because of the high costs, we From the Department of Clinical Sciences of Companion Animals, Faculty of Veterinary Medicine, University Utrecht, Utrecht, The. Reprint requests: Sarah J.M. Proot, DVM, Department of Clinical Sciences of Companion Animals, Faculty of Veterinary Medicine, University Utrecht, P.O. Box 80.154, 3508 TD Utrecht, The ; e-mail: S.J.M.Proot@vet.uu.nl. Submitted March 3, 2003; Revised May 9, 2006; Accepted June 30, 2006. Copyright E 2006 by the American College of Veterinary Internal Medicine 0891-6640/06/2006-0006/$3.00/0 have replaced this method with a less expensive automatic biopsy device, the Pro-Mag Ultra Automatic Biopsy Instrument b (Fig 2). This device has been used for many years to perform biopsies in dogs and has been a safe and suitable instrument in our experience. The better size and quality of the biopsy samples obtained with this device was another reason to consider this method in cats. 3 However, experience indicated that some cats developed fatal hypotensive shock reactions after biopsies were performed with the16g device, in contrast to the semiautomatic device. The aim of this study was to evaluate whether collection of 16G Tru-Cut biopsy samples with an automatic biopsy device causes more complications than collection of biopsy samples with a semiautomatic device of identical diameter. Materials and Methods Cases We reviewed all cases from October 1, 2002, until October 31, 2004, in which ultrasound-guided Tru-Cut liver biopsies were performed in cats referred to the University of Utrecht. Reasons for referral were liver-related clinical signs (eg, general malaise, lethargy, anorexia, vomiting, weight loss, jaundice), increased liver enzyme activities, increased bile acid concentrations, ultrasonographic abnormalities, or some combination of these findings. Between October 1, 2002, and October 31, 2003, biopsies were performed with an automatic biopsy device b on 26 occasions. One cat underwent biopsy twice with an interval of 1 month. In another cat, a hepatic mass was biopsied, and 3 weeks later other locations within the diffusely abnormal liver also were biopsied. We considered each biopsy as a separate case for statistical analysis. Between November 1, 2003, and October 31, 2004, biopsies were performed with a semiautomatic VET-core biopsy needle a on 19 occasions. During this period, 1 cat underwent biopsy 3 times with intervals of 1 and 2 months, respectively. Clinical Database All cats were examined physically. In most cats, CBC and biochemical analysis (especially liver enzyme activities and bile acid

1328 Proot and Rothuizen Fig 2. gun). Pro-Mag Ultra Automatic Biopsy Instrument (biopsy Fig 1. Needle used for hepatic biopsies. Semiautomatic 16-gauge VET-core biopsy needle. concentrations) were performed at the referring practice or at the University. Determination of coagulation parameters (ie, prothrombin time [PT], activated partial thromboplastin time [APTT], and fibrinogen concentration) was always performed before performing Tru-Cut biopsies. In each cat, ultrasonographic examination of the abdomen, with special attention to hepatobiliary structures, was performed. If the cat was not cooperative enough to safely perform the biopsy procedure under local anesthesia, it was anesthetised, usually with propofol c (4 mg/kg IV as the initial dose, with additional dosing depending on the effect). Biopsy Procedures and Devices The biopsy site was surgically prepared. Lidocaine, 2 to 3 ml, was injected SC and into the abdominal wall before a small skin incision was made with a number 11 surgical blade. The biopsy instrument then was introduced through the abdominal wall and advanced into the liver under ultrasound guidance to the location of interest. Before firing the device, the distance between the tip of the needle and important structures, such as blood vessels or bile ducts, was measured. A distance of at least 2 cm always was allowed because inner needle advancement was 2 cm when the device was fired. In general, 2 biopsy samples were taken from each cat. After each attempt, Doppler ultrasonography was used to detect bleeding. Skin incisions did not require suturing. 4 The procedure was performed by the hepatologist (JR) or the resident rotating in the hepatology department at that time. This procedure was followed consistently during the entire study period. If the resident performed the procedure, it was done so after intensive training in the method and always under the direct supervision of the hepatologist. Resuscitation In the first period (when the automatic biopsy device was used), a high incidence of hypotensive or bradycardic shock, occurring within 10 to 15 minutes of taking the biopsy, was observed. Resuscitation was initiated immediately in these cats. Oxygen was supplied, a peripheral venous catheter was placed for fluid therapy, an electrocardiogram was monitored, heart massage was applied, and resuscitative drugs (eg, atropine, d epinephrine, e dobutamine f ) were administered IV as necessary. Statistical Analysis For statistical analysis, SPSS 13.0 software g was used. First, the normality of the different variables was evaluated by the Kolmogorov-Smirnov test. Normally distributed variables were then evaluated with Levene s test for equality of variances and t- test for equality of means (2-tailed). Non normally distributed parameters were evaluated by a 2-tailed Mann-Whitney test. Significant differences between the 2 groups for the different clinical signs, sex, histology, and anesthesia were evaluated by the 2-sided Fisher s exact test. A P value of,.05 was considered significant. Results Forty-five ultrasound-guided liver biopsies were performed. In the first period, biopsy specimens were collected at 26 occasions with the automatic device. Of these cats, 18 were neutered males, 7 were neutered females, and 1 was an intact female. The median age was 10 years (95% confidence interval [CI], 1 13; in 2 instances the age was not available), and median weight was 4.1 kg (95% CI, 2.4 6.4; in 3 instances weight was not recorded). This group consisted of 14 mixed breed cats, 8 European shorthairs, 2 British shorthairs, 1 Siamese, and 1 Tonkinese cat. Of the 19 biopsy samples collected in the second period, 15 were taken from neutered males, 1 from an intact male, and 3 from neutered females. During this period, the median age of the cats was 11 years (95% CI, 5 14 years), and the median weight was 3.8 kg (95% CI, 2.9 6.2; in 1 instance weight was not recorded). The second group comprised 9 mixed breed cats, 4 European shorthairs, 3 Persians, 1 Burmese, 1 East European Shorthair, and 1 Main Coon. There were no differences between groups in regard to sex, age, breed, or weight. Clinical Signs Clinical information about 2 cats in the first group was limited, likely because the cats were referred only for biopsy and not for full diagnostic evaluation. Similar clinical signs were present in both groups of cats. Jaundice and vomiting were more frequent in cats seen in the first period (Table 1). However, none of the differences was significant. Clinicopathologic Findings In the first period, 6 cats (23%) had anemia, 8 cats (31%) had leukocytosis, and 1 had leukopenia. Bile acid concentrations were increased in 11 cats (42%); alanine aminotransferase (ALT) activity, in 16 cats (62%); total protein concentration, in 16 cats (62%); serum albumin concentration, in 4 cats (15%); and in the 18 cats (of 26) in which serum protein electrophoresis was performed, gamma globulin concentrations were increased in 16 cats. In the second period, 5 cats (26%) had anemia, 4 (21%) had leukocytosis, but no cat had leukopenia. Bile

Risks of Biopsy Guns in Cats 1329 Table 1. Clinical signs in cats of both groups. No. with Automatic Biopsy Device No. with Semiautomatic Biopsy Device Clinical Sign n 5 24 n 5 19 P Value Jaundice 12 5.133 Anorexia/decreased appetite 20 14.477 Weight loss 17 15.728 Lethargy 21 15.680 Ascites 4 2 1.00 Vomiting 14 7.223 PU/PD 3 3 1.00 Fever 3 4.680 PU, polyuria; PD, polydipsia. acid concentrations were increased in 8 cats (42%); ALT activity, in 10 cats (53%); total protein concentration, in 6 cats (32%); serum albumin concentration, in 3 cats (16%); and in 8 of the 11 cats in which serum protein electrophoresis was performed, gamma globulin concentrations were increased (Table 2). In all cats of both periods, blood coagulation parameters (ie, PT, APTT, and fibrinogen concentration) were within the reference range. Statistically, there were no differences in PCV, leukocyte counts, bile acid concentrations, ALT activity, albumin, or gamma globulin concentrations between the 2 groups. Only the difference in total protein concentration was statistically significant, with mean total protein concentration being higher in cats of the first period (7.5 g/dl) compared with those of the second period (6.7 g/dl; P,.05). Clinicopathologic findings in surviving and nonsurviving cats in which the automatic biopsy device was used were within comparable ranges (Table 3). It was not possible to statistically evaluate these results because of insufficient numbers. Anesthesia In 25 of the 45 cats, anesthesia was required to safely perform the biopsy procedure. In the period from October 1, 2002, until October 31, 2003, 12 of the 26 (46%) were anesthetized. In 6 of these, propofol was used. Medetomidine h was used in 3 cats. Midazolam i combined with buprenorphine j or ketamine k were used once each. The anesthetics used could not be identified in 1 case. Between November 1, 2003, and October 31, 2004, 13 of the 19 cats (68%) required anesthesia. In 9 of these cats, propofol was used. Medetomidine was used twice, and the combination of acepromazine l with methadone m was used in 1 cat. In 1 cat, the anesthetic used could not be identified. There was no significant difference in the number of cats that needed anesthesia in the first period compared with the second period. In addition, the anesthetic protocol used was not changed between study periods, and the anesthetics used were selected based on the needs of each individual cat. Biopsy Procedure and Outcome Ultrasound-guided biopsies were taken in all 45 cats. Care was taken to avoid bile ducts and large blood vessels in all cats. After completion of the biopsy procedure, ultrasonography was performed and no bleeding could be detected in any of the cats. During the period in which the automatic biopsy device was used, 5 cats developed severe shock within 15 minutes after the biopsy. All of these cats experienced severe bradycardia accompanied by very weak pulse, respiratory impairment, and loss of consciousness (ie, hypotensive shock). None survived despite prompt resuscitation as described previously. Table 2. Clinicopathologic findings in cats of both groups. Automatic Biopsy Device a Semiautomatic Biopsy Device a Value Units Reference Range Median 95% CI Median 95% CI PCV % 27 50 35 18 43 29 16 39 Leucocytes 10 3 /ml 4.3 16.2 14.8 7.0 31.5 10.8 5.7 28.4 Bile acids mm 1 13 14 1 540 10 2 68 ALT U/L 30 73 302 25 1114 97 23 1111 Total protein* g/dl 5.4 7.0 7.5 5.6 9.6 6.9 5.7 7.7 Albumin g/dl 2.5 3.4 3.0 2.2 4.2 2.9 2.0 3.7 Gamma globulins g/dl 0.4 0.8 1.7 0.6 5.7 1.1 0.6 2.8 CI, confidence interval. a Median values and 95% CI of the 2.5th and 97.5th percentiles. * indicates significant difference between the 2 groups.

1330 Proot and Rothuizen Table 3. Clinicopathologic findings in the surviving and deceased cats of the period in which the automatic biopsy gun was used. Survivors a Deceased Cats a Value Units Reference Range Median Range Median Range PCV % 27 50 36 19 45 23 13 39 Leucocytes 10 3 /ml 4.3 16.2 16 7.0 47.0 9.4 2.5 10.9 Bile acids mm 1 13 9 1 540 47 5 578 ALT U/L 30 73 311 15 1513 292 28 743 Total protein g/dl 5.4 7.0 7.5 4.7 9.6 8.0 5.6 10.2 Albumin g/dl 2.5 3.4 3.0 2.2 4.2 2.9 2.0 3.7 Gamma globulins g/dl 0.4 0.8 1.6 0.5 7.2 1.7 1.5 5.7 ALT, alanine aminotransferase. a Median values with lowest and highest values. Because of this high complication rate, as of November 2003 we stopped using the automatic biopsy device and began using the semiautomatic VET-core biopsy needle. During the next 12 months, 19 liver biopsies were performed in cats. None of them experienced any major adverse effects. The complication rate in the first period in which the automatic biopsy device was used was 19.2%, and the complication rate was 0.0% in the second period during which the semiautomatic biopsy device was used. The difference in complication rates had a P value of.06. Histopathology Lymphocytic and neutrophilic cholangitis was equally distributed between the 2 groups (Table 4). In the group in which the automatic biopsy device was used, there were 2 cats with neoplasia (1 adenocarcinoma and 1 hepatocellular adenoma). In the other group, there were 5 tumors: 1 adenocarcinoma, 1 (likely metastatic) carcinoma, 1 malignant lymphoma, 1 histiocytosis, 1 erythroid myeloma. In 2 of the 3 cats with normal liver histopathology, bile duct dilatation was seen with ultrasonography. However, culture of bile samples from these cats was negative and no bacteria or leukocytes were seen on cytologic examination. Fatal Cases Of the 5 cats that died, 2 were European Shorthairs, 2 were mixed breed, and 1 was a Tonkinese cat. The median age was 11 years (range, 5 months 13 years). Reasons for referral were weight loss (in all cats), jaundice (in 4 cats), decreased appetite or anorexia (in 4 cats), fever and ascites (in 1 cat), and vomiting and abdominal pain (in 1 cat). In 4 cats, the PCV was low with a median of 23% (range, 13 39%). In all but 1 cat, bile acid concentrations were increased (median, 47 mm; range, 5 578 mm). In 3 cats ALT activity was increased (median, 292 U/L; range, 28 743 U/L). In 3 cats total protein concentration was increased (median, 8 g/dl; range, 5.6 10.2 g/dl). One cat had high serum albumin concentration (without high PCV), and in all 5 cats gamma globulin concentrations were increased (median, 1.7 g/dl; range, 1.5 5.7 g/dl). Coagulation parameters were normal in all cats. To perform ultrasonography and biopsy, 2 of the 5 nonsurviving cats had required anesthesia. In both instances propofol was used. Two cats were diagnosed with FIP, 1 with lymphocytic cholangitis, 1 with acute local necrotic hepatitis, and 1 with macro-vesicular and micro-vesicular steatosis suggestive of hepatic lipidosis. Of the 2 cats that required anesthesia, 1 had FIP and the other had lymphocytic cholangitis. In 1 cat (with FIP) postmortem examination was performed, confirming the diagnosis and absence of hemorrhage. Discussion Fatal shock reactions were only observed after the use of the automatic biopsy device b and did not occur when using the semiautomatic biopsy device. a The difference in complication rate between the 2 groups had a P value of.06. It is important to evaluate whether factors other than the biopsy device used could explain this difference. Massive bleeding (leading to hypotensive shock) or injury of larger bile ducts could have been complicating factors. However, ultrasound-guidance was used as the standard procedure in all cats in this study. The distance between the tip of the Tru-Cut needle and vital structures (eg, blood vessels, bile ducts, gall bladder) always was measured before taking the biopsy sample, and enough space (minimum of 2 cm) was allowed to avoid such structures. After the biopsy was performed, ultrasonography was used to check for the presence of hemorrhage. Clinically relevant bleeding was not observed in any of the cats in this study. Postmortem examination was performed in 1 of the 5 nonsurviving cats, and no trauma to vital structures or evidence of bile peritonitis was found in this cat. Anesthesia also may have been a factor in the outcome. The number of cats that needed anesthesia for the procedure during the first period was not substantially different from the number requiring anesthesia during the second period. Of the 5 nonsurviving cats, 2 were anesthetized with propofol. The course of anesthesia was smooth and uneventful until the time of biopsy, and fatal shock occurred immediately after taking the biopsy sample. Therefore, there is no indication that anesthesia could explain the difference between the 2 groups.

Risks of Biopsy Guns in Cats 1331 Table 4. Histologic diagnosis in cats of both groups. Histologic Diagnosis No. with Automatic Biopsy Device a No. with Semiautomatic Biopsy Device Lymphocytic cholangitis 5 (1) 4 Neutrophilic cholangitis 3 2 Hepatic lipidosis 1 (1) 3 Neoplasia 2 5 Feline infectious peritonitis 2 (2) 0 Steroid-induced Hepatopathy 2 0 Polycystic kidney disease 3 1 Arteriovenous fistula 1 0 Secondary reactive hepatitis 3 1 Local necrotic hepatitis/necropurulent hepatitis 1 (1) 1 Ischemic hepatitis 0 1 Normal 3 1 Total 26 19 a Number in parentheses indicates the number of cats that died with this diagnosis. Clinical differences existed in the cats before biopsy. Whenever possible potential risks were avoided. Anesthesia was also avoided, unless the biopsy procedure could not be safely performed without it. The number of cats was too small to statistically compare the clinicopathologic features of the survivors to those of the nonsurviving cats during the first period. Table 3 compares the clinicopathologic features of the groups. The PCV of the nonsurviving cats was somewhat lower than that of the survivors. Although the ranges are quite comparable, the median bile acid concentrations differ between the groups. Cats that were more anemic and had more severe hepatic dysfunction may have been more susceptible to complications, which may have contributed to their deaths. Coagulation was evaluated before biopsy and was normal in all 45 cats. Finally, specific subgroups of liver diseases may be associated with specific risk factors during the biopsy procedure. The diagnoses in the 2 groups were comparable, but not identical. Two of the 5 nonsurvivors had FIP, but the clinical relevance of this observation is unknown. In the second group of cats in which the semiautomatic VET-core biopsy needle was used, none had FIP. Fatal complications were not limited to cats with FIP, and other nonsurviving cats in the first group had hepatobiliary diseases that also occurred in the second group in which no complications occurred. We therefore do not believe that the presence of FIP in the first group explains the difference in complication rates. Moreover, no clinically relevant variations were identified with respect to clinical signs or clinicopathologic findings between the groups. Only total protein concentration was significantly different between the 2 groups, which may be explained by the presence of FIP in only 1 group. We conclude that the difference in biopsy procedures between the groups is the only logical explanation for the different occurrence of fatal postbiopsy shock reactions. Thus, the most probable explanation for the occurrence of shock after the automatic biopsy procedure is the occurrence of intense vagotonia leading to shock. All 5 cats that died had developed severe bradycardia with weak pulse, respiratory impairment, and loss of consciousness. All of these cats died because of cardiac arrest despite immediate resuscitation. Such clinical signs may be seen when a strong increase in vagal tone occurs (eg, during surgical procedures in which vagal reflexes occur secondary to visceral traction). 5 Such a reaction may be explained when the 2 biopsy devices are compared. The automatic biopsy needle is put in place under ultrasound guidance. When the device is activated, an inner needle with stylet and specimen notch is advanced 2 cm into the liver and then immediately covered by an outer cannula. This splitsecond event transfers a 14 pound-force (lbf), which equals 62.28 N, to the tissue which is specifically produced by the stylet (written communication from Inter-V, tjolsen@interv.net, Jan 23, 2006). In the case of the semiautomatic biopsy needle, the stylet (with specimen notch) is manually advanced into the liver under ultrasound guidance. When the plunger is fully depressed, the outer cutting cannula is quickly advanced to cover the stylet. Here the manual advancement of the stylet transfers much less force, and the thin cutting sheath, which is fired automatically, also represents less force than the much quicker movement of the stylet of the automatically driven biopsy needle. It is likely that the strong pressure wave from the automatic biopsy device is not tolerated by cats. Vagotonia after liver biopsy is a reported complication in human patients. In humans, complications of percutaneous liver biopsy are pain; vagal reaction with hypotension; trauma to other abdominal organs; perforation of the diaphragm, gallbladder, or blood vessels; fracture or dislocation of the needle or parts of it; and seeding of microorganisms or tumor cells. 6 Vasovagal episodes and pain are the most common reasons for hospitalization after outpatient liver biopsy. 7 Janes and Lindor 8 found that 3.2% of all patients undergoing outpatient liver biopsy required hospitalization afterwards. In 92% of patients, hospitalization was for pain or hypotension. Piccinino et al 9 analyzed the complications of percutaneous liver biopsy in humans in relation to the type of needle used and the underlying disease in a retrospective study of 68,276 liver biopsies from 1973 to 1983. In this study, liver biopsies were performed with

1332 Proot and Rothuizen a Menghini needle (60,611 biopsies), a Tru-Cut needle (7,372 biopsies), or a Vim-Silverman needle (293 biopsies). Death, serious hemorrhagic complications, pneumothorax, and biliary peritonitis occurred most frequently when the Tru-Cut needle was used. The use of a semiautomatic device, however, was not evaluated. The underlying disease in human patients who did not survive the procedure was always cirrhosis or hepatic neoplasia. Dupas et al 10 investigated the complications of percutaneous liver biopsy guided by fluoroscopy, ultrasonography, and computed tomography in human patients. Liver biopsy was associated with complications in,2% of patients and consisted of transient hepatic pain and vagal reactions that were not fatal but required treatment with atropine. Unfortunately, none of these studies provided an explanation for the vasovagal episodes or hypotension. In the veterinary literature, similar immediate adverse effects have not been reported. 11,12 Léveillé et al 13 reviewed 233 ultrasound-guided biopsies of abdominal structures in anesthetized cats and dogs in which an automated Biopty biopsy device n with an 18-gauge needle was used. In 177 cases (cats and dogs), the liver was biopsied, and in 2 cats major complications occurred. These cats developed bile peritonitis 24 hours after the procedure, although the gall bladder was not punctured. Both cats were diagnosed with hepatic lipidosis, and the friability of the lipidotic liver was thought possibly to have led to bile duct rupture. Such an event could only explain 1 of the 5 deaths in our study. Furthermore, in the study of Léveillé et al, the cats did not develop complications immediately as was the case in our study. The absence of vagotonic shock in the previous study may have been due to less pronounced forces developed by the Biopty device n (9.0 6 0.9 lbf or 40.0 6 4.0 N) (written communication from Frank Borremans of CR Bard Inc, Frank.Borremans@ crbard.com, April 19, 2006). An alternative explanation may be the smaller gauge needle used (ie, 18 gauge compared with 16 gauge in this study). However, the bevelled edge of the stylet of the 16-gauge needle does not exceed that of an 18-gauge needle by more than 15%. Consequently, we believe the force exerted by the automatic device is the major factor contributing to the fatal complications in this cat series. Apparently, cats are much more susceptible to developing vagotonia than are dogs, because we have used the same automatic biopsy device for many years in dogs without the complications reported here in cats. As in humans, we have observed discomfort and vagotonic reactions that had to be treated with atropine and IV fluid administration in a few dogs with hepatic carcinomas. These vagotonic reactions were treated successfully and were never fatal. The use of anticholinergic premedication could be considered when using the automatic biopsy device in cats. However, we question whether it would be ethical to expose any other cats, even with anticholinergic premedication, to the high risk of fatal complications. Consequently, we have not pursued this approach. Use of anticholinergics, however, may be a consideration when using other automated devices, because cats appear to be more susceptible to vagotonia than are dogs. We conclude that the use of the automatic biopsy device evaluated in this study (which develops a force of 14 lbf) should be avoided for liver biopsy in cats because of the unacceptably high frequency of severe complications. Instead, semiautomatic or automatic biopsy devices that do not develop such high forces may be employed, taking standard safety precautions for liver biopsy. Footnotes a Vet-Core biopsy needle (16 gauge; length, 9 mm; throw length, 20 mm), Global Veterinary Products Inc, New Buffalo, MI b Pro-Mag Ultra Automatic Biopsy Instrument, Manan Medical Products, PBN Medicals, Stenløse, Denmark. (22-mm stroke, serial no. 6527) c PropoVet, Abbott Laboratories, Queensborough, Kent, UK d Atropinesulfaat, Pharmachemie BV, Haarlem, Noord-Holland, The e Adrenaline, Pharmachemie BV, Haarlem, Noord-Holland, The f Dobutamine, Centrafarm, BV, Etten-Leur, Noord-Brabant, The g SPSS 13.0; SPSS Inc, Chicago, IL h Domitor, Pfizer Animal Health BV, Cappelle a/d IJssel, Zuid- Holland, The i Dormicum, Roche Nederland BV, Woerden, Utrecht, The j Temgesic, Schering-Plough, Utrecht, The k Ketamine HCl, AST farma BV, Oudewater, Utrecht, The l Acepromazine, Apotheek faculteit diergeneeskunde, Utrecht, The m Methadon HCl, Eurovet Animal Health BV, Bladel, Noord- Brabant, The n Bard Urological Division, CR Bard Inc, Covington, GA Acknowledgments We hereby want to thank Dr E. Teske (DVM, PhD, Dipl ECVIM-CA) for his support in the statistical analysis of the data and J. Fama for providing digital images. References 1. Caney SMA, Gruffed-Jones TJ. Feline inflammatory liver disease. In: Ettinger SJ, ed. Textbook of Veterinary Internal Medicine, 6th ed. Philadelphia, PA: WB Saunders; 2005: 1448 1453. 2. Webster CRL. History, clinical signs, and physical findings in hepatobiliary disease. In: Ettinger SJ, ed. Textbook of Veterinary Internal Medicine, 6th ed. Philadelphia, PA: WB Saunders; 2005:1422 1434. 3. Hoppe FE, Hager DA, Poulos PW, et al. A comparison of manual and automatic ultrasound-guided biopsy techniques. Vet Radiology 1986;27:99 101. 4. Rothuizen J, Desmet VJ, van den Ingh TSGAM, et al. In: WSAVA Standards for Histological and Clinical Diagnosis of

Risks of Biopsy Guns in Cats 1333 Canine and Feline Liver Disease. Edinburgh, UK: Churchill Livingstone; 2006:5 14. 5. Briggs BA, Cullen DJ. In: Complications in Anaesthesiology. Philadelphia, PA: Lippincott; 1983:239 250. 6. Desmet V, Fevery J. Liver biopsy. Bailliere Clin Gastr 1995;9:811 828. 7. Flora KD, Ingram K. Diagnostic liver biopsy. Available at: http://www.emedicine.com. Accessed March 9, 2004. 8. Janes CH, Lindor KD. Outcome of patients hospitalized for complications after outpatient liver biopsy. Ann Intern Med 1993;118:96 98. 9. Piccinino F, Sagnelli E, Pasquale G. Complications following percutaneous liver biopsy: A multi-centre study. J Hepatol 1986;2:165 173. 10. Dupas B, Frampas E, Leaute F, et al. Complications de gestes interventionnels percutanés sous contrôle radioscopique, échographique ou scanographique. J Radiol 2005;86:586 598. 11. Bigge LA, Brown DJ, Penninck DG. Correlation between coagulation profile findings and bleeding complications after ultrasound-guided biopsies: 434 cases (1993 1996). J Am Anim Hosp Assoc 2001;37:228 233. 12. Cole TL, Center SQ, Shannon NF, et al. Diagnostic comparison of needle and wedge biopsy specimens of the liver in dogs and cats. J Am Vet Med Assoc 2002;220:1483 1490. 13. Léveillé R, Partington BP, Biller DS et al. Complications after ultrasound-guided biopsy of abdominal structures in dogs and cats: 246 cases (1984 1991). J Am Vet Med Assoc 1993; 203:413 415.