Outpatient Liver Biopsy: How Safe Is It?

Percutaneous liver biopsy is being done more frequently as a result of major changes in the therapy of liver diseases.The ease and safety of outpatient, blind liver biopsy has been well established. Complication rates have ranged from 0.9% to 3.7%; most complications have become manifest within the first 3 hours after biopsy, and none has resulted in death. Nevertheless, differing guidelines may affect relative degrees of safety in different Centers. A review of reported series suggests that the Menghini technique may be the preferred first choice, with the Tru-Cut technique reserved for patients in whom an adequate sample cannot be obtained. The requirement of a normal bleeding time appears to decrease the incidence of hemorrhagic complications. The decision to use outpatient biopsy must be based on a careful assessment of the risks and benefits for each individual patient. Percutaneous needle biopsy of the liver has been performed for more than 100 years. This procedure was first done by Paul Ehrlich in Germany in 1883 [1] for the determination of glycogen in the livers of diabetic patients, and later it was used in tropical medicine to aid in the diagnosis of hepatic abscesses; however, needle biopsy was not widely used because of the high complication rate. Even after the first large series were reported in the 1930s by Huard and colleagues [2] in France and Baron [3] in the United States, the procedure was still regarded with apprehension. Percutaneous liver biopsy was not widely adopted until 1957 when Menghini reported his 1-second aspiration technique [4], which introduced a needle that shortened the duration of the intrahepatic phase of the biopsy to a fraction of a second. Since then, percutaneous trans-thoracic needle biopsy has been the standard method for obtaining liver tissue. The technique is done blindly, usually using either a standard, thin-bore (1.4 to 1.8 mm in diameter) aspiration Menghini-type needle or a Tru-Cut needle (2.05 mm in diameter) (Travenol Laboratories Inc., Deerfield, Illinois) where tissue is obtained by cutting. Histologic examination of the liver remains essential for the diagnosis of most liver diseases. Liver test abnormalities of prolonged duration or unexplained hepatomegaly are still the main diagnostic indications for liver biopsy. However, a biopsy may not be necessary in an obese patient with aminotransferase elevations within twice the normal limit, since fatty infiltration of the liver, which is almost always the cause, can usually be confirmed by ultrasound. Liver biopsies are now being used frequently to assess rejection in liver transplant patients and to distinguish these abnormalities from other causes of liver dysfunction. Liver biopsy measurements of iron or copper are the most definitive way to diagnose hemochromatosis or Wilson disease. Biopsies may help define the diagnosis in patients with fever of unknown origin or in immunocompromised patients with hepatomegaly or elevated liver enzyme levels. With advances in the knowledge of the natural history of many diseases and the development of new therapeutic modalities, the indications for liver biopsy have expanded to include prognostic and therapeutic evaluations. Liver histology has been established as an important factor in primary biliary cirrhosis and primary sclerosing cholangitis, as well as in alcoholic liver disease. Therapeutically, advances in antiviral therapy for chronic hepatitis B and C have led to an increased use of liver biopsy both to establish disease severity and to judge the potential effectiveness of interferon. Also, trials with drugs such as ursodeoxycholic acid and methotrexate have stimulated a need for the histologic staging of patients with primary biliary cirrhosis and sclerosing cholangitis, for both selection and follow-up. Even though indications have increased, liver biopsy should be done only after a complete clinical evaluation, including appropriate noninvasive studies, which suggests that the patient may benefit if liver histology is available. For many years, liver biopsy was done as an in-patient procedure because of fear of complications. Reported complication and mortality rates for inpatient procedures [5-10] ranged from 0.1% to 0.6% and from 0% to 0.12%, respectively, with the latter averaging 0.03% (Table 1). Frank and Leodolter, who reported the first series of 204 outpatient liver biopsies in 1966 [11], observed only one complication. Since then, studies have confirmed that outpatient liver biopsy is associated with a low incidence of complications Table 2, most of which occur within 3 hours of the procedure [8, 9, 12-17]. Differences in complication rates between in-patient and outpatient series are probably due to differences in the severity of the underlying illness (more bleeding diatheses, tumors, and so forth) and to differences in the severity of complications reported: Whereas 32 of 369 complications (8.7%) reported in inpatient series [5-9] required surgical intervention and 50 of 516 complications (9.7%) resulted in death, only 1 of 60 complications (1.6%) reported in outpatient series required surgical intervention and none resulted in death. Thus, in 1989, the Patient Care Committee of the American Gastroenterological Association recommended that performance of liver biopsies utilizing an outpatient facility in which it is possible to follow the patient closely for at least 6 hr is acceptable for many patients [18]. Table 1. Inpatient Liver Biopsy* Table 2 In this issue of Annals, Janes and Lindor [17] report 13 complications in 405 patients undergoing outpatient percutaneous liver biopsy. Their complication rate of 3.2% is similar to the complication rate of 3.7% observed in a previous Mayo Clinic series [14] and to the rates of 3% reported by Ali and colleagues [13] in Australia and by Westaby and colleagues [15] in England. Other series [8, 9, 12, 16, 17] have shown lower complication rates (0.9% to 1.8%); these series include our experience at Yale, where liver biopsies have been done as an outpatient procedure since 1987. In the 30-month period from January 1990 to June 1992, we have done 346 outpatient, blind percutaneous liver biopsies in patients with diffuse liver disease, 5 of whom (1.4%) required admission because of complications. Differences in complication rates could be due to the type of technique used in the different centers. A comparison of needle types and techniques used in out-patient liver biopsy suggests that the number of complications is higher when Tru-Cut needles are used (see Table 2). A multicenter study had previously shown that complications were more frequent after inpatient biopsy with the Tru-Cut needle than after biopsy with the Menghini needle [10]. This is probably due to a longer dwell time in the liver (1 to 3 seconds with the Tru-Cut technique compared with less than 0.5 seconds with the Menghini technique) and the cutting quality of the Tru-Cut needle. Our most serious complication, a gallbladder perforation, occurred with the Tru-Cut needle after two attempts with the Klatskin needle (a Menghini-type needle) had failed. Liver biopsies using the Menghini technique are easier, quicker, and cheaper (because the needles are reusable). However, fragmented specimens are obtained more frequently with the Menghini techniques than with the Tru-Cut technique, particularly in cirrhotic livers [19]. On balance, we suggest that the Tru-Cut technique be used only when the Menghini technique is unsuccessful. Since the procedure is blind, liver biopsies should be done only by or with the assistance of trained physicians (gastroenterologists, hepatologists, transplant surgeons) who do liver biopsies regularly. Patients with more advanced liver disease appear to be at a higher risk for complications from liver biopsies [10]. More than half of the patients described by Janes and Lindor [17] were biopsied for staging of primary biliary cirrhosis and sclerosing cholangitis, but it is not clear whether complications were higher in those with advanced disease. Interestingly, three of our five complications occurred in patients with fibrosis or cirrhosis. Janes and Lindor's finding that hemorrhage is the most frequent complication confirms the experience of others (see Table 2). In contrast, we had three patients with bile peritonitis and two with gallbladder perforations. Differences in hemorrhagic complications may be related to differences in clotting parameter requirements. As shown in Table 2, all centers request a prothrombin time and most request platelet counts; we also require a bleeding time. In our institution, liver biopsy is not done if prothrombin time is prolonged more than 3 seconds, platelet count is less than 60 109/L, and the bleeding time is more than 10 minutes. Where tissue is essential, a transjugular biopsy is done. The combined experience would indicate that it is safe to discharge patients after 4 to 6 hours of observation. Although outpatient biopsies are less expensive than inpatient biopsies [8, 14], the costs can be considerable. For example, in our institution the charges total $1500: $530 for space and equipment rental, $450 for physician fees, and $500 for pathology fees. Retrospective studies with historical controls [20, 21] claim that biopsies done with ultrasound guidance were associated with fewer complications than were blind liver biopsies. However, no differences in complications were observed in a prospective study comparing blind and laparoscopy-guided biopsy [22]. Our two instances of gallbladder perforation would probably have been prevented if ultrasound guidance had been available. However, we recently noticed the occurrence of one hemorrhagic complication in a patient with diffuse liver disease in whom the liver biopsy was done under ultrasound guidance. In patients with a small liver or in whom hepatic dullness is lacking, biopsies should be done under ultrasound guidance. Both u