- Email: [email protected]
Computed tomography of the liver after biopsy
Clinical Radiology (1987) 38, 97-99
Computed Tomography of the Liver after Biopsy D. E. MALONE, P. A. McCORMICK, D. P. O'DONOGHUE and D. P. MacERLEAN Departments of Radiology and Gastroenterology, St Vincents Hospital, Elm Park, Dublin 4, Ireland Angiographic studies suggest that needle biopsy of the liver is invariably associated with some degree of haemorrhage (Wallace et al., 1972). The appearances of the liver on computed tomography after percutaneous biopsy have not been described. As an aid to the interpretation of scans in patients referred because of clinical complications of biopsy, we performed unenhanced computed tomography of the liver in 30 patients within 72 h of biopsy. None of these patients would ordinarily have been referred for computed tomography. Only one of the scans showed a parenchymal abnormality attributable to biopsy. We conclude that post-biopsy changes on unenhanced computed tomography scans of the liver are uncommon and that the needle tracks and vascular anomalies commented on by angiographers are not usually shown.
Needle biopsy of the liver is invariably associated with some degree of haemorrhage, usually minor and of no clinical significance (Wallace et al., 1972). Angiography and isotope scans have both been used to investigate the appearances of the liver after percutaneous biopsy. Needle tracks, haematomas, arterioportal fistulae and aneurysms have been reported (Preger, 1967; Wallace et al., 1972; Okuda etal., 1978; Forssell et al., 1981). Many of these abnormalities occur in asymptomatic patients and would normally go undetected (Wallace et al., 1972; Okuda et al., 1978; Forssell et al., 1981). The resolution of both of these methods is inferior to that of computed tomography which can show lesions 0.5 cm in diameter (Bernardino and Lewis, 1982) and pleural or peritoneal fluid. We feel that knowledge of the usual appearance of the liver on unenhanced computed tomography will be of value to those interpreting scans in patients referred because of clinically complicated biopsies.
Unenhanced scans were performed on a Somatom DR3 scanner (Siemens) using 8 mm thick contiguous sections through the liver. All scans were reviewed by the same consultant radiologist, who was not supplied with clinical information. RESULTS No serious complications occurred. Two patients developed severe pain and two mild pain within 1 h of the biopsy. One patient developed immediate transient hypotension considered to be vasovagal in origin. An asymptomatic peripherally-located haematoma 0.5 cm in diameter (attenuation 70 Hounsfield units (HU)) was detected in one patient (Fig. 1). Another, who had severe pain, had right pleural fluid subsequently shown to be due to a haemothorax. Both of these patients had normal coagulation before the biopsy. No other scan showed any abnormality attributable to the biopsy. Parenchymal liver abnormalities consistent with the primary disease were shown in four patients. Five patients had ascites and three pleural effusions. These findings were known to be due to their primary disease. A 0.5 cm high density lesion (129 HU) was found in one patient. This is consistent with a calcified granuloma. Three days after the biopsy one patient, whose prebiopsy ultrasound and post-biopsy computed tomography were normal, developed right upper quadrant pain and tenderness. Unfortunately residual barium in the colon from a barium enema prevented computed tomography being done. Ultrasound showed a 2 cm lesion in the right lobe of the liver consistent with a haematoma. Computed tomography (unenhanced) was performed 3 days later by which time her symptoms had resolved. No abnormality was shown. DISCUSSION
PATIENTS AND M E T H O D S
A prospective study was performed. Ten female and 20 male patients had computed tomography within 72 h of diagnostic percutaneous liver biopsy. None of the patients would ordinarily have been referred for postbiopsy computed tomography. The age range of the group was 25 to 78 years, with a mean age of 59. Biopsy was performed at the site of maximum hepatic dullness in the mid-axillary line using a 'Tru-Cut' needle. Coagulation screening tests were performed before biopsy. The mean prothrombin time was 16 s (control 12-15 s), range 11.5-19.5 s. The mean platelet count (normal 150-400x 109/litre) was 283x 109/litre, range 70-571x 109/litre. Two patients received fresh frozen plasma before the biopsy because of disturbed coagulation. Table 1 shows the histological diagnoses.
Percutaneous liver biopsy is considered a relatively safe diagnostic technique. Mortality rates of less than one in 5000 have been reported in large series (Conn, 1974). Major complications can be expected in 0.32% and include haemorrhage, bile leak, pneumothorax, haemothorax and inadvertant penetration of abdominal viscera (Hegarty and Williams, 1984). Minor complications, usually transient pain or hypotension soon after biopsy, occur in about 6% (Perrault et al., 1978). Despite the low clinical incidence of complications, radiological abnormalities are frequently found in the post-biopsy liver. They often occur in asymptomatic patients (Worobetz et al., 1983; Raines et al., 1974). Angiography and isotope scans done after biopsy have shown haematomas, needle tracks, arterioportal fistulae and aneurysms (Preger, 1967; Wallace et al., 1972; Raines et al., 1974; Hellekant, 1976; Okuda et al., 1978; Forssell et al., 1981). The frequency with which
98
CLINICAL RADIOLOGY
Table 1 - Histological diagnoses
Diagnosis'
Number
Normal Cirrhosis Granulomatous hepatitis Hepatoma Fatty liver Miscellaneous (alcoholic hepatitis, leukaemia, minor drug-induced change etc.)
12 4 2 2 2 8
Fig. 1 - The CT Scan (W280 C50) shows a small area of increased attenuation in the right lobe of the liver (arrow). The mean attenuation value was 70 HU, consistent with a haematoma.
angiography has shown abnormalities has led to the statement that needle biopsy of the liver is invariably accompanied by some degree of h a e m o r r h a g e , usually minor and of no clinical significance (Wallace et al., 1972). The number of vascular complications found after biopsy reflects the diligence of the search (Worobetz et al., 1983). Because of the uniquely high resolution of computed t o m o g r a p h y and its increasing usage we performed this study to investigate the appearance of unenhanced post~biopsy computed t o m o g r a p h y of the liver. Ninety-seven per cent of our scans showed no parenchymal abnormality attributable to biopsy. Four (12%) of our patients had had post-biopsy pain and one (3%) transient hypotension. The only parenchymal abnormality shown was a small peripheral h a e m a t o m a in an asymptomatic patient. One symptomatic patient had a collection of pleural fluid (later found to be a haemothorax) without associated parenchymal abnormality shown on computed tomography. Previous studies using isotope scans have shown that intrahepatic h a e m a t o m a s occur in 2 - 7 % of patients within 72 h of biopsy (Raines et al., 1974; Forssell et al., 1981). Isotope scans will show lesions 2 cm in diameter whereas computed t o m o g r a p h y can show lesions 0.5 cm in diameter (Bernardino and Lewis, 1982; Kandel and Marcon, 1984). C o m p u t e d t o m o g r a p h y can also differentiate between h a e m a t o m a s and other lesions on
the basis of their attenuation values. The single h a e m a f o m a we found measured 0.5 cm in diameter. Lesions of this size are unlikely to be visible on an isotope scan (Bernardino and Lewis, 1982; Kandel and Marcon, 1984). The 3% rate of h a e m a t o m a formation shown in our series suggests that, despite their limitations with regard to lesion size and type, isotope studies did not significantly underestimate the incidence of post-biopsy h a e m a t o m a s . The finding of this 0.5 cm h a e m a t o m a on the day of biopsy suggests that unenhanced computed tomography can show fresh bleeding into the liver. Considering the patient with late onset symptoms, we feel that there are two possible explanations of the normal computed tomography appearance despite a firm clinical and sonographic diagnosis of intrahepatic haemorrhage 3 days earlier. These are either that the 2 cm h a e m a t o m a had been resorbed in the 3 days or that a h a e m a t o m a may at some stage be isodense with liver. In view of the latter possibility enhancement by intravenous contrast medium may be advisable in the evaluation of a symptomatic post-biopsy patient if the unenhanced scan appears normal, especially if scanning takes place after the day on which symptoms arise. We conclude that the majority of unenhanced postliver biopsy computed tomography scans show no biopsy-related abnormality. The needle tracks and vascular anomalies c o m m e n t e d on by angiographers are not usually shown. The liver may be normal in appearance despite the presence of post-biopsy pain. Unenhanced computed tomography can show very small h a e m a t o m a s on the day of biopsy and the identification of extrahepatic fluid collections may assist in the evaluation of the complicated biopsy. Late intrahepatic h a e m o r r h a g e may rarely occur in patients whose post-biopsy unenhanced computed tomography appearances were normal. Acknowledgements. Our thanks to the Radiographers in the Department of CT scanning for their assistance, to Ismay FitzGerald for her help in preparing the manuscript and to the Department of Medical Photography, St Vincent's Hospital.
REFERENCES
Bernardino, ME & Lewis, E (1982). Imaging hepatic neoplasms. Cancer, 50, 2666-2671. Conn, HO (1974). Intrahepatic haematoma after liver biopsy. Gastroenterology, 67, 375-381. Forssell, PL, Bonkowsky, HL, Anderson, PB & Howell, DA (1981). Intrahepatic haematoma after aspiration liver biopsy. A prospective randomized trial using two different needles. Digestive Diseases and Sciences, 26, 631 635. Hegarty, JE & Williams, R (1984). Liver biopsy: techniques, clinical applications and complications. British Medical Journal, 288, 1254-1256. Hellekant, C (1976). Vascular complications following needle puncture of the liver - clinical angiography. Acta Radiologica: Diagnosis', 17, 209-222. Kandel, G & Marcon, NE (1984) Pyogenic liver abscess: new concepts of an old disease. American Journal of Gastroenterology, 79, 65-71. Okuda, K, Musha, H, Nakajima, Y, Takayasu, K, Suzuki, Y, Morita, M, etal. (1978). Frequency of intrahepatic arteriovenous fistulaeas a sequela to percutaneous needle puncture of the liver. Gastroenterology, 74, 1204-1207. Pcrrault, J, McGill, DB, Ott, BJ & Taylor, WF (1978). Liver Biopsy: Complications in 1000 inpatients and outpatients. Gastroenterology, 74, 103-106.
COMPUTED TOMOGRAPHY OF THE LIVER AFTER BIOPSY Preger, L (1967): Hepatic arteriovenous fistula after percutaneous liver biopsy. American Journal of Roentgenology, Radium Therapy and Nuclear Medicine, 101,619-620. Raines, DR, Van Heertum, RL & Johnson, LF (1974). Intrahcpatic haematoma: a complication of percutaneous liver biopsy. Gastroenterology, 67,284-289,
99
Wallace, S, Medellin, H & Nelson, RS (1972). Angiographic changes due to needle biopsy of the liver. Radiology, 11)5, 13-1~. Worobetz, LJ, Passi, RB & Sullivan SN (1983). Haemobilia after percutaneous liver biopsy: role of endoscopic retrograde cholangiopancreatography and sphincterotomy. American Journal of Gastroenterology, 78, 182-184.
Computed Tomography of the Liver after Biopsy D. E. MALONE, P. A. McCORMICK, D. P. O'DONOGHUE and D. P. MacERLEAN Departments of Radiology and Gastroenterology, St Vincents Hospital, Elm Park, Dublin 4, Ireland Angiographic studies suggest that needle biopsy of the liver is invariably associated with some degree of haemorrhage (Wallace et al., 1972). The appearances of the liver on computed tomography after percutaneous biopsy have not been described. As an aid to the interpretation of scans in patients referred because of clinical complications of biopsy, we performed unenhanced computed tomography of the liver in 30 patients within 72 h of biopsy. None of these patients would ordinarily have been referred for computed tomography. Only one of the scans showed a parenchymal abnormality attributable to biopsy. We conclude that post-biopsy changes on unenhanced computed tomography scans of the liver are uncommon and that the needle tracks and vascular anomalies commented on by angiographers are not usually shown.
Needle biopsy of the liver is invariably associated with some degree of haemorrhage, usually minor and of no clinical significance (Wallace et al., 1972). Angiography and isotope scans have both been used to investigate the appearances of the liver after percutaneous biopsy. Needle tracks, haematomas, arterioportal fistulae and aneurysms have been reported (Preger, 1967; Wallace et al., 1972; Okuda etal., 1978; Forssell et al., 1981). Many of these abnormalities occur in asymptomatic patients and would normally go undetected (Wallace et al., 1972; Okuda et al., 1978; Forssell et al., 1981). The resolution of both of these methods is inferior to that of computed tomography which can show lesions 0.5 cm in diameter (Bernardino and Lewis, 1982) and pleural or peritoneal fluid. We feel that knowledge of the usual appearance of the liver on unenhanced computed tomography will be of value to those interpreting scans in patients referred because of clinically complicated biopsies.
Unenhanced scans were performed on a Somatom DR3 scanner (Siemens) using 8 mm thick contiguous sections through the liver. All scans were reviewed by the same consultant radiologist, who was not supplied with clinical information. RESULTS No serious complications occurred. Two patients developed severe pain and two mild pain within 1 h of the biopsy. One patient developed immediate transient hypotension considered to be vasovagal in origin. An asymptomatic peripherally-located haematoma 0.5 cm in diameter (attenuation 70 Hounsfield units (HU)) was detected in one patient (Fig. 1). Another, who had severe pain, had right pleural fluid subsequently shown to be due to a haemothorax. Both of these patients had normal coagulation before the biopsy. No other scan showed any abnormality attributable to the biopsy. Parenchymal liver abnormalities consistent with the primary disease were shown in four patients. Five patients had ascites and three pleural effusions. These findings were known to be due to their primary disease. A 0.5 cm high density lesion (129 HU) was found in one patient. This is consistent with a calcified granuloma. Three days after the biopsy one patient, whose prebiopsy ultrasound and post-biopsy computed tomography were normal, developed right upper quadrant pain and tenderness. Unfortunately residual barium in the colon from a barium enema prevented computed tomography being done. Ultrasound showed a 2 cm lesion in the right lobe of the liver consistent with a haematoma. Computed tomography (unenhanced) was performed 3 days later by which time her symptoms had resolved. No abnormality was shown. DISCUSSION
PATIENTS AND M E T H O D S
A prospective study was performed. Ten female and 20 male patients had computed tomography within 72 h of diagnostic percutaneous liver biopsy. None of the patients would ordinarily have been referred for postbiopsy computed tomography. The age range of the group was 25 to 78 years, with a mean age of 59. Biopsy was performed at the site of maximum hepatic dullness in the mid-axillary line using a 'Tru-Cut' needle. Coagulation screening tests were performed before biopsy. The mean prothrombin time was 16 s (control 12-15 s), range 11.5-19.5 s. The mean platelet count (normal 150-400x 109/litre) was 283x 109/litre, range 70-571x 109/litre. Two patients received fresh frozen plasma before the biopsy because of disturbed coagulation. Table 1 shows the histological diagnoses.
Percutaneous liver biopsy is considered a relatively safe diagnostic technique. Mortality rates of less than one in 5000 have been reported in large series (Conn, 1974). Major complications can be expected in 0.32% and include haemorrhage, bile leak, pneumothorax, haemothorax and inadvertant penetration of abdominal viscera (Hegarty and Williams, 1984). Minor complications, usually transient pain or hypotension soon after biopsy, occur in about 6% (Perrault et al., 1978). Despite the low clinical incidence of complications, radiological abnormalities are frequently found in the post-biopsy liver. They often occur in asymptomatic patients (Worobetz et al., 1983; Raines et al., 1974). Angiography and isotope scans done after biopsy have shown haematomas, needle tracks, arterioportal fistulae and aneurysms (Preger, 1967; Wallace et al., 1972; Raines et al., 1974; Hellekant, 1976; Okuda et al., 1978; Forssell et al., 1981). The frequency with which
98
CLINICAL RADIOLOGY
Table 1 - Histological diagnoses
Diagnosis'
Number
Normal Cirrhosis Granulomatous hepatitis Hepatoma Fatty liver Miscellaneous (alcoholic hepatitis, leukaemia, minor drug-induced change etc.)
12 4 2 2 2 8
Fig. 1 - The CT Scan (W280 C50) shows a small area of increased attenuation in the right lobe of the liver (arrow). The mean attenuation value was 70 HU, consistent with a haematoma.
angiography has shown abnormalities has led to the statement that needle biopsy of the liver is invariably accompanied by some degree of h a e m o r r h a g e , usually minor and of no clinical significance (Wallace et al., 1972). The number of vascular complications found after biopsy reflects the diligence of the search (Worobetz et al., 1983). Because of the uniquely high resolution of computed t o m o g r a p h y and its increasing usage we performed this study to investigate the appearance of unenhanced post~biopsy computed t o m o g r a p h y of the liver. Ninety-seven per cent of our scans showed no parenchymal abnormality attributable to biopsy. Four (12%) of our patients had had post-biopsy pain and one (3%) transient hypotension. The only parenchymal abnormality shown was a small peripheral h a e m a t o m a in an asymptomatic patient. One symptomatic patient had a collection of pleural fluid (later found to be a haemothorax) without associated parenchymal abnormality shown on computed tomography. Previous studies using isotope scans have shown that intrahepatic h a e m a t o m a s occur in 2 - 7 % of patients within 72 h of biopsy (Raines et al., 1974; Forssell et al., 1981). Isotope scans will show lesions 2 cm in diameter whereas computed t o m o g r a p h y can show lesions 0.5 cm in diameter (Bernardino and Lewis, 1982; Kandel and Marcon, 1984). C o m p u t e d t o m o g r a p h y can also differentiate between h a e m a t o m a s and other lesions on
the basis of their attenuation values. The single h a e m a f o m a we found measured 0.5 cm in diameter. Lesions of this size are unlikely to be visible on an isotope scan (Bernardino and Lewis, 1982; Kandel and Marcon, 1984). The 3% rate of h a e m a t o m a formation shown in our series suggests that, despite their limitations with regard to lesion size and type, isotope studies did not significantly underestimate the incidence of post-biopsy h a e m a t o m a s . The finding of this 0.5 cm h a e m a t o m a on the day of biopsy suggests that unenhanced computed tomography can show fresh bleeding into the liver. Considering the patient with late onset symptoms, we feel that there are two possible explanations of the normal computed tomography appearance despite a firm clinical and sonographic diagnosis of intrahepatic haemorrhage 3 days earlier. These are either that the 2 cm h a e m a t o m a had been resorbed in the 3 days or that a h a e m a t o m a may at some stage be isodense with liver. In view of the latter possibility enhancement by intravenous contrast medium may be advisable in the evaluation of a symptomatic post-biopsy patient if the unenhanced scan appears normal, especially if scanning takes place after the day on which symptoms arise. We conclude that the majority of unenhanced postliver biopsy computed tomography scans show no biopsy-related abnormality. The needle tracks and vascular anomalies c o m m e n t e d on by angiographers are not usually shown. The liver may be normal in appearance despite the presence of post-biopsy pain. Unenhanced computed tomography can show very small h a e m a t o m a s on the day of biopsy and the identification of extrahepatic fluid collections may assist in the evaluation of the complicated biopsy. Late intrahepatic h a e m o r r h a g e may rarely occur in patients whose post-biopsy unenhanced computed tomography appearances were normal. Acknowledgements. Our thanks to the Radiographers in the Department of CT scanning for their assistance, to Ismay FitzGerald for her help in preparing the manuscript and to the Department of Medical Photography, St Vincent's Hospital.
REFERENCES
Bernardino, ME & Lewis, E (1982). Imaging hepatic neoplasms. Cancer, 50, 2666-2671. Conn, HO (1974). Intrahepatic haematoma after liver biopsy. Gastroenterology, 67, 375-381. Forssell, PL, Bonkowsky, HL, Anderson, PB & Howell, DA (1981). Intrahepatic haematoma after aspiration liver biopsy. A prospective randomized trial using two different needles. Digestive Diseases and Sciences, 26, 631 635. Hegarty, JE & Williams, R (1984). Liver biopsy: techniques, clinical applications and complications. British Medical Journal, 288, 1254-1256. Hellekant, C (1976). Vascular complications following needle puncture of the liver - clinical angiography. Acta Radiologica: Diagnosis', 17, 209-222. Kandel, G & Marcon, NE (1984) Pyogenic liver abscess: new concepts of an old disease. American Journal of Gastroenterology, 79, 65-71. Okuda, K, Musha, H, Nakajima, Y, Takayasu, K, Suzuki, Y, Morita, M, etal. (1978). Frequency of intrahepatic arteriovenous fistulaeas a sequela to percutaneous needle puncture of the liver. Gastroenterology, 74, 1204-1207. Pcrrault, J, McGill, DB, Ott, BJ & Taylor, WF (1978). Liver Biopsy: Complications in 1000 inpatients and outpatients. Gastroenterology, 74, 103-106.
COMPUTED TOMOGRAPHY OF THE LIVER AFTER BIOPSY Preger, L (1967): Hepatic arteriovenous fistula after percutaneous liver biopsy. American Journal of Roentgenology, Radium Therapy and Nuclear Medicine, 101,619-620. Raines, DR, Van Heertum, RL & Johnson, LF (1974). Intrahcpatic haematoma: a complication of percutaneous liver biopsy. Gastroenterology, 67,284-289,
99
Wallace, S, Medellin, H & Nelson, RS (1972). Angiographic changes due to needle biopsy of the liver. Radiology, 11)5, 13-1~. Worobetz, LJ, Passi, RB & Sullivan SN (1983). Haemobilia after percutaneous liver biopsy: role of endoscopic retrograde cholangiopancreatography and sphincterotomy. American Journal of Gastroenterology, 78, 182-184.