Liver biopsy is currently considered the gold standard for assessing hepatic fibrosis or cirrhosis. However, it is an invasive procedure, with rare, but potentially life-threatening, complications.1 In addition, the accuracy of liver biopsy in assessing fibrosis has limitations because of well-known sampling errors and interobserver variability.2-6 Transient elastography (FibroScan [FS]) is a rapid, noninvasive, and reproducible method for measuring liver stiffness that is increasingly explored to assess liver fibrosis.7 FS measurements can be carried out in about 95% of patients but are limited in those with severe obesity and ascites.8 Liver stiffness as measured by FS is associated strongly with the degree of liver fibrosis in patients with chronic hepatitis mainly due to hepatitis C virus and hepatitis B virus.9-12 On the basis of these studies, a cutoff value of about 12.5 kPa is considered to be optimal for discrimination of liver fibrosis from liver cirrhosis in patients with chronic hepatitis C virus infection.10 Liver stiffness values ranging from 2.4 to 5.5 are considered as normal although no large F0 validated control groups have been studied so far.10 Based on a recent meta-analysis, it has been even suggested that patients could be spared from histologic assessment if FS measurements exceed cut-off values.12 However, liver stiffness also increases during alanine aminotransferase flares in patients with chronic viral hepatitis.13, 14 Moreover, it has been established recently that FS frequently yields pathologically high values in patients with acute liver damage rendering FS unsuitable for detecting cirrhosis in these patients.15 The aim of the present study was to determine if the measurement of liver stiffness by FS is altered in patients with extrahepatic cholestasis.
Transient elastography (FibroScan [FS]) is a novel non-invasive tool to assess liver fibrosis/cirrhosis. However, it remains to be determined if other liver diseases such as extrahepatic cholestasis interfere with fibrosis assessment because liver stiffness is indirectly measured by the propagation velocity of an ultrasound wave within the liver. In this study, we measured liver stiffness immediately before endoscopic retrograde cholangiopancreatography and 3 to 12 days after successful biliary drainage in patients with extrahepatic cholestasis mostly due to neoplastic invasion of the biliary tree. Initially elevated liver stiffness decreased in 13 of 15 patients after intervention, in 10 of them markedly. In three patients, liver stiffness was elevated to a degree that suggested advanced liver cirrhosis (mean, 15.2 kPa). Successful drainage led to a drop of bilirubin by 2.8 to 9.8 mg/dL whereas liver stiffness almost normalized (mean, 7.1 kPa). In all patients with successful biliary drainage, the decrease of liver stiffness highly correlated with decreasing bilirubin (Spearman's ρ = 0.67, P < 0.05) with a mean decrease of liver stiffness of 1.2 ± 0.56 kPa per 1 g/dL bilirubin. Two patients, in whom liver stiffness did not decrease despite successful biliary drainage, had advanced liver cirrhosis and multiple liver metastases, respectively. The relationship between extrahepatic cholestasis and liver stiffness was reproduced in an animal model of bile duct ligation in landrace pigs where liver stiffness increased from 4.6 kPa to 8.8 kPa during 120 minutes of bile duct ligation and decreased to 6.1 kPa within 30 minutes after decompression. Conclusion: Extrahepatic cholestasis increases liver stiffness irrespective of fibrosis. Once extrahepatic cholestasis is excluded (e.g., by liver imaging and laboratory parameters) transient elastography is a valuable tool to assess liver fibrosis in chronic liver diseases. (HEPATOLOGY 2008.)
Patients and Methods
Fifteen patients presenting with extrahepatic cholestasis at the Salem Medical Center/University of Heidelberg from June 2007 to June 2008 were analyzed. The underlying disease was established according to standard criteria using laboratory tests, ultrasound, endoscopic ultrasound, computed tomography (CT) imaging, and endoscopic retrograde cholangiopancreatography (ERCP). Extrahepatic cholestasis was due mostly to neoplastic invasion of the biliary tree and included pancreas carcinoma, gastrointestinal stroma tumor (GIST), metastatic liver disease, and Klatskin tumor. One patient had gallstones of the common bile duct (Table 1). With the exception of Patient 7, all pancreas carcinomas were restricted to the distal common bile duct without any apparent hepatic invasion. All patients underwent therapeutic ERCP to re-establish biliary drainage by stone extraction or stent implantation. Liver stiffness was measured by FS before ERCP and 3 to 12 days after ERCP depending on the decreases of cholestasis parameters. Because extrahepatic cholestasis is a contraindication for liver biopsy and liver stiffness was clearly dependent only on cholestasis parameters, no histologic assessment of liver fibrosis was carried out in the majority of patients. Three patients had to be excluded from the study because no valid liver stiffness measurements could be obtained due to extreme obesity, gallbladder hydrops, and progressed chronic obstructive pulmonary disease with pronounced pulmonary emphysema. The study was approved by the Ethical Committee of the University of Heidelberg.
|Patient||Sex||Age (Years)||Etiology of Obsctructive Jaundice||Signs of Hepatic Tumor Invasion||Signs of Liver Cirrhosis||Intervention||Bilirubin before Intervention (mg/dL)||Bilirubin after Intervention (mg/dL)||Sufficient Drainage||Liver Stiffness before Intervention (kPa)||Liver Stiffness after Intervention (kPa)||Observation Interval (days)|
|2||M||62||GIST Tumor||No||No||Stent Implantation||8,2||2||Yes||10||5,4||3|
|3||M||77||Pancreas Carcinoma||No||No||Stent Implantation||8,9||5,8||Yes||14,3||10,4||12|
|4||F||88||Pancreas Carcinoma||No||No||Stent Implantation||16,5||6,6||Yes||12,1||7,6||3|
|5||M||41||Gastric Cancer with Peritoneal Carcinomatosis||No||No||Stent Implantation||4||1,5||Yes||10,1||7,9||7|
|6||F||78||Pancreas Carcinoma||No||No||Stent Implantation||6,5||1,6||Yes||10,8||3,8||6|
|7||M||83||Pancreas Carcinoma||Yes||No||Stent Implantation||19,9||12,1||Yes||36,3||17,3||3|
|8||W||87||Gallbladder Carcinoma||No||No||Stent Implantation||13,3||9,1||Yes||15,9||7,1||5|
|9||M||71||Gallbladder Carcinoma||No||No||Stent Implantation||3,8||1,4||Yes||7,1||4,3||3|
|10||M||37||Chronic Pancreatitis||No||No||Stent Implantation||6,5||3,6||Yes||10,2||6,1||5|
|11||F||90||Klatskin Tumor Type II (Bismuth)||Yes||No||Stent Implantation||7,5||7,2||Yes||21,6||19,1||4|
|12||M||72||Pancreas Carcinoma||No||No||Stent Implantation||10,2||2,7||Yes||6,9||6,4||8|
|13||M||62||Pancreas Carcinoma||No||No||Stent Implantation||3,5||2||Yes||5,7||5,3||7|
|14||M||69||Colon Cancer with Liver Metastasis||Yes||No||Stent Implantation||14,8||6,5||Yes||7,8||8,3||3|
|15||M||69||Pancreas Carcinoma||No||Alcoholic Liver Cirrhosis||Stent Implantation||8,3||7,2||Yes||75||75||7|
Liver Stiffness Measured by FS.
Liver stiffness was measured by FS as described recently in detail by Sandrin et al.7 The tip of the probe transducer was placed on the skin between the rib bones and the level of the right lobe of the liver. The measurement depth was between 25 and 65 mm below the skin surface. Ten measurements were carried out with success rates of at least 60%. In 12 patients (74%), the interquartile range (IQR) was in all measurements below 30%. The IQR did not exceed 40% in any of the measurements. The results were expressed in kilopascals (kPa). The median value was taken as representative.
Ultrasound examination was carried out routinely to confirm extrahepatic cholestasis after patients had fasted for at least 6 hours. Liver cirrhosis was suspected when two of the following criteria were present: (1) nodular aspect of the liver surface; (2) portal vein diameter >12 mm; (3) collateral circulation; and (4) hypertrophy of segment 4 (quadrate lobe).16
Experimental Animals and Treatment.
We used 10 German landrace pigs (22.8 ± 2.9 kg) and divided them into control (n = 5) and bile duct ligation (n = 5) groups. Preoperative preparation included fasting for 12 hours, allowing free access to water, a standardized narcotic protocol (premedication: azaperone 8 mg/kg intramuscularly, midazolam 0.5-0.7 mg/kg intramuscularly, ketamine 5 mg/kg intravenously, atropin 1 mg intravenously), which was followed by endotracheal intubation. Pressure controlled ventilation was done in a half-closed system (ventilation parameters: frequency, 12/minute; volume, 300 mL; room air, 1.5-2 L/minute; O2, 0.5-1.0 L/minute, N2O, 1.5-2.0 L/minute; Isoflurane, 0.75-1.5%). Arterial blood gases were controlled within a strict limit (pO2, 100-150 mm Hg, pCO2, 35-42 mm Hg). Cardiocirculatory parameters such as central venous and mean arterial pressure were measured by indwelling catheters into the internal jugular vein and common carotid artery, respectively. The control group was operated on accordingly without doing bile duct ligation. Heart rate and rhythm were monitored continuously by a surface electrocardiogram. The body temperature was measured continuously using a rectal temperature probe. During the experimental procedure body temperature could be kept above 37°C by using heated blankets (Warm-Touch, Mallinckrodt Medical) as well as infrared heating lamps on the operation table. At the end of the experiment, while animals were in deep anesthesia, they were sacrificed by a central vein injection of 20 mmol potassium chloride. Approval for the experimental procedure was obtained from the German Committee on Animal Care, Regierungspräsidium Karlsruhe and the Medical Faculty Ethics Committee, University of Heidelberg. During the experiments, all animals received human care in compliance with the European Regulations for Animal Experiments and the United States National Research Council's criteria for human care, as outlined in “Guide for the Care and Use of Laboratory Animals” prepared by the National Institution of Health.
Correlations between laboratory findings and liver stiffness in patients were calculated as bivariate regression analysis for nonparametric variables according to Spearman (regression coefficient r, r2, P). Differences were considered significant at P < 0.05. All statistical analyses were carried out with SPSS, version 12.0.1 (SPSS, Munich, Germany). For bile duct ligation experiments in animals, average and standard deviations were used and calculated by Excel (Microsoft, Redmond, WA).
Longitudinal Correlation of Liver Stiffness with Bilirubin and Reversibility in a Single Patient with Obstructive Jaundice.
At first presentation, a patient with obstructive jaundice due to GIST occluding the terminal common bile duct had increased bilirubin of 3.5 mg/dL and an increased liver stiffness with 5.7 kPa (Fig. 1). After successful stent implantation, bilirubin decreased to 2.0 mg/dL and liver stiffness to 5.0 kPa (see Fig. 1, arrow). Two weeks later, the patient presented again with abdominal discomfort. Bilirubin was still at 2 mg/dL and the liver stiffness at 5.4 kPa. While hospitalized, however, the patient again developed obstructive jaundice due to stent occlusion. Bilirubin rose to 8.0 mg/dL that was accompanied by a concomitant increase of liver stiffness to 10 kPa. Thus, within several days, the patient had developed liver stiffness values that would otherwise suggest liver fibrosis, almost reaching the cut-off value for F4 fibrosis (cirrhosis) without presenting any signs and laboratory findings for liver cirrhosis.
Liver Stiffness Increases with Extrahepatic Cholestasis.
The observation above prompted us to study the correlation between liver stiffness and cholestasis parameters in patients with extrahepatic cholestasis. A total of 15 patients (male/female, 9/6) with extrahepatic cholestasis and successful biliary drainage were seen during the study period. The mean age was 71 ± 15 years. Patients' characteristics and etiology of cholestasis are shown in Table 1. Patients with no successful biliary drainage were excluded. Extrahepatic cholestasis before ERCP intervention was confirmed by laboratory testing and ultrasound imaging. Initial serum bilirubin levels ranged from 3.5 to 19.9 mg/dL (mean, 9.5 ± 4.7 mg/dL). All patients underwent abdominal sonography and showed dilated intra- and extrahepatic bile ducts. None of the patients had criteria for liver cirrhosis in the ultrasound examination except Patient 15, who showed nodular aspects of the liver surface, splenomegaly in ultrasound and CT scan, and typical laboratory findings (thrombocytopenia). Patients 7, 12, and 14 showed overt liver metastasis. Liver stiffness was measured before and after ERCP. In all patients with extrahepatic cholestasis, initial liver stiffness was increased ranging from 5.7 to 75 kPa (Table 1). The mean success rate of measurements was 82% (range, 60%-100%) and the mean observation interval was 3-12 days (mean, 5.3 ± 2.5).
Liver Stiffness Decreases after Successful Biliary Drainage.
Successful biliary drainage was established by stent implantation in 14 cases and stone extraction in one case. In 13 of 15 patients, successful biliary drainage led to a significant decrease of liver stiffness. In the majority of patients (n = 10, Patients 1-10; see Table 1), liver stiffness markedly decreased after intervention by 2.2 to 9.1 kPa (mean, 6.6 ± 4.9). In these patients, liver stiffness declined by 40% to 64% within the mean observation time of 5.5 days. Three patients completely normalized and had values between 3.5 and 5.4 kPa (Patients 2, 6, and 9). In four patients, liver stiffness was elevated above the cut-off values of 12.5 kPa for liver cirrhosis. In three patients, liver stiffness fell below this cut-off value within the observation period. Patient 7 had multiple liver metastasis due to pancreas carcinoma. Nevertheless, liver stiffness fell significantly after biliary drainage from 36.3 to 17.3 kPa but remained above the cut-off value for liver cirrhosis.
Liver Stiffness Significantly Correlates with Bilirubin Levels in Patients with Extrahepatic Cholestasis.
In the patient group with a significant decrease in bilirubin (Patients 1-10), a mean decrease of liver stiffness by 1.2 ± 0.56 kPa was measured per 1 g/dL decrease in bilirubin. In these patients, Spearman correlation of liver function tests and cholestasis parameters with liver stiffness showed a significant correlation for bilirubin and liver stiffness of r = 0.67 with a significance level of P < 0.05 (Table 2). However, liver stiffness did neither correlate significantly with serum ALT activities nor with levels of alanine aminotransferase (AP) or gamma glutamyltransferase (GGT). Patients 11-13 showed mildly decreased liver stiffness of 1.1 kPa after successful biliary drainage despite a sufficient decrease of bilirubin of 3.2 mg/dL and an observation interval of 6.5 days. However, in two cases, initial liver stiffness was increased only mildly (∼6 kPa; Patients 12 and 13) and in one case intrahepatic metastasis could explain the sustained increase of stiffness (Patient 11).
Persistently Elevated Liver Stiffness Values Despite Successful Biliary Drainage.
Liver stiffness did not decrease in two patients although biliary drainage was established successfully and confirmed by ultrasound and decrease in bilirubin (Patients 14 and 15). However, the sustained and increased liver stiffness in these two patients was explained by metastasizing liver disease and liver cirrhosis, respectively. Patient 14 had multiple liver metastases due to colon carcinoma as evidenced by liver imaging and confirmed by ultrasound guided biopsy. Patient 15 had advanced alcoholic liver cirrhosis and showed liver stiffness values of 75 kPa despite ERCP. Despite decline of histologic analysis, cirrhosis was unambiguously confirmed by two independent imaging studies (CT and ultrasound) showing definite signs of liver cirrhosis such as nodular aspect of liver surface and splenomegaly. In addition, the patient's medical history showed continuous alcohol abuse in the last 2 decades whereas blood tests showed thrombocytopenia and mildly increased international normalized ratio despite vitamin K substitution.
Experimental Bile Duct Ligation in a Pig Model Increases in Liver Stiffness.
To prove the correlation between extrahepatic cholestasis and liver stiffness, we carried out bile duct ligation for 120 minutes in five German landrace pigs. This period is sufficient to induce a firm gall bladder and a palpable and visible swelling of the liver. Controls were sham operated without bile duct ligation (n = 5). Of note, liver stiffness values were comparable to those of humans before bile duct ligation (mean, 4.6 kPa). The results are shown in Figure 2. Liver stiffness increased significantly during the 2 hours of bile duct ligation from 4.6 to 8.8 kPa, reaching cut-off values that are considered to reflect F3 fibrosis. Control animals did not show any change in liver stiffness. Reversibility of liver stiffness was tested by restoration of bile flow. Liver stiffness rapidly decreased within 30 minutes after removal of bile duct ligation and declined to 6.1 kPa.
Liver stiffness correlates well with the stage of liver fibrosis by FS as shown for patients with chronic hepatitis of different etiologies (hepatitis C virus, hepatitis B virus, alcohol, and human immunodeficiency virus/hepatitis C virus coinfection). A cut-off value of ∼12.5 kPa has been described to discriminate between liver cirrhosis and liver fibrosis. In our own study group comprising 60 patients with alcoholic liver disease, a similar cutoff value of 13.7 kPa was established.17
In this study, we analyzed 15 patients with extrahepatic cholestasis due mostly to neoplastic invasion of the biliary tree (pancreas carcinoma, Klatskin tumor, liver metastases, and GIST). All patients underwent liver stiffness measurements by FS before and after successful biliary drainage by ERCP. Initial liver stiffness values above 12.5 kPa, which are otherwise suggestive of liver cirrhosis, were measured in four patients (26.6%) although none of these patients showed any signs of liver cirrhosis (clinical setting, liver imaging). With two exceptions, liver stiffness decreased in all patients after successful biliary drainage or stone extraction. Liver stiffness correlated significantly with a decrease in bilirubin but not with GGT, AP, glutamic-oxaloacetic transaminase, or glutamic-pyruvic transaminase. The two patients in whom liver stiffness did not fall despite successful biliary drainage had both other causes that explained the persistently increased liver stiffness. One patient had progressive alcoholic liver cirrhosis and the other had multiple liver metastasis due to colon carcinoma.
To further confirm the impact of extrahepatic cholestasis on liver stiffness measurements by FS, we carried out bile duct ligation for 120 minutes on pigs and compared the change in liver stiffness with sham operated animals. The ligation period of 120 minutes led to a significant swelling of the liver and a tightly palpable gall bladder as compared to humans. Liver stiffness values doubled during bile duct ligation and reached values suggesting F3 fibrosis. After removal of the bile duct ligation and a recovery period of 30 minutes, liver stiffness values returned to almost normal values around 6.1 kPa.
The present study shows that bile duct obstruction interferes with the determination of liver fibrosis by FS measurements in patients with extrahepatic cholestasis and may erroneously suggest the presence of liver cirrhosis. The reasons underlying the high stiffness in cholestasis are unknown but could be related to tissue swelling, inflammation, edema, and increased intracellular pressure due to impaired bile flow. As shown in the bile duct ligation experiments and as seen from the initial rapid decrease of liver stiffness after biliary drainage, the increased hydrostatic pressure alone seems to contribute to increased liver stiffness during extrahepatic cholestasis. Moreover, cholestasis could also contribute to increased liver stiffness in various other forms of chronic liver disease. Thus, intrahepatic cholestasis has been shown to correlate strongly with liver stiffness in patients with acute hepatitis.15 Intrahepatic cholestasis could be an important superimposed factor, next to fibrosis and inflammation, contributing to the especially high cut-off values of 17.3 kPa for F4 cirrhosis observed in patients with primary biliary cirrhosis and primary sclerosing cholangitis.18
FS examination has been shown repeatedly to be a rapid, noninvasive, and reliable method for assessment of fibrosis/cirrhosis in patients with chronic hepatitis. In patients with extrahepatic cholestasis, however, liver stiffness assessment is unreliable for the diagnosis of coexisting fibrosis or cirrhosis. Four of the patients with cholestasis had liver stiffness values that predicted liver cirrhosis and three of them fell below values suggestive for F3 fibrosis after successful biliary drainage. Thus, FS results need to be interpreted with caution in patients with extrahepatic cholestasis, and high values of liver stiffness do not predict the simultaneous presence of cirrhosis in these patients. Our study emphasizes that extrahepatic cholestasis should be excluded by liver imaging and laboratory testing before interpretation of liver stiffness measurements.