Y Chang and JI Kim contributed equally as co-first authors.
Editor: Ja Kyung Kim, Yonsei University College of Medicine, Korea
Liver biopsy (LB) remains the gold standard for the evaluation of liver disease. However, over the past two decades, many noninvasive tests have been developed and utilized in clinical practice as alternatives to LB. The aim of this study was to evaluate the clinical use and safety of LB in the era of noninvasive assessment of liver fibrosis.
This retrospective study included 1,944 consecutive cases of LB performed between 2001 and 2018 in a tertiary hospital. All of the LBs were conducted under ultrasonography guidance with 18-gauge cutting needles.
LBs were performed an average of approximately 108 times per year during the study period. Chronic hepatitis B (25.3%) and suspected malignancy (20.5%) were the two most common indications for LB. The use of LB for nonalcoholic fatty liver disease increased from 8.1% to 17.2% in the past 5 years compared to the last 10 years, while that for viral hepatitis decreased from 40.3% to 18.9%. Discordance rate between the suspected diagnosis and the final diagnosis was 2.6% (51 cases). The overall rate of major adverse events was 0.05% (one case), which involved delayed bleeding at the biopsy site. Liver cirrhosis was observed in 563 cases (28.9%), and the presence of cirrhosis did not affect the frequency of complications (
LB is widely used in clinical practice as an irreplaceable diagnostic tool, even in the era of noninvasiveness. Ultrasonography-guided LB can be performed safely in patients with liver cirrhosis.
Liver biopsy has many clinical usefulness until recently, and when the procedure is performed using ultrasound, complications are low even in patients with cirrhosis.
Percutaneous liver biopsy was first introduced in 1883 [
Recently, many noninvasive methods have been developed to replace liver biopsies [
In this study, we evaluated the changes in liver biopsy indications over the past 18 years and the safety of ultrasonography-guided liver biopsies in the era of noninvasive assessments of liver fibrosis.
Between January 2001 and December 2018, we retrospectively collected data from patients undergoing ultrasonography-guided percutaneous liver biopsies at the Gastroenterology and Hepatology Department of a single tertiary center. A liver biopsy was performed when each investigator deemed it necessary to determine the cause and severity of the patient’s liver disease. During this period, five patients underwent blind biopsies and 97 patients underwent computed tomography (CT)-guided biopsies. The 102 patients who underwent blind or CT-guided liver biopsies were excluded from the study. The final number of eligible patients included in the study was 1,944. The clinical indications for liver biopsy, the suspected diagnosis before the biopsy, and the final diagnosis after the biopsy were collected. The medications taken at the time of biopsy and complications after the biopsy were also reviewed. Liver cirrhosis was assessed by ultrasound, defined as the presence of a nodular liver surface, round edges, and hypoechoic nodules in the liver parenchyma, which represent regenerative nodules [
The study protocol was approved by the Institutional Review Board of SoonChunHyang University Bucheon Hospital (IRB number SCHBC 2019-10-022). The study protocol conformed to the ethical guidelines of the World Medical Association Declaration of Helsinki.
Ultrasonography-guided liver biopsies were performed by four expert hepatologists experienced with more than 1,000 ultrasound procedures and over 200 liver biopsies. The patients were admitted to the hospital the day before the procedure and fasted for at least 12 hours. Prior to the procedure, routine laboratory tests, including coagulation parameters, chest X-rays, and abdominal ultrasounds were performed. During the procedure, a transthoracic approach was routinely used with the patient in the supine position. Local anesthesia using a 2% lidocaine solution was administered, and the biopsy was performed using an 18-gauge Tru-cut needle. After the liver biopsy, the standard compressive dressing was immediately applied to the wound and bed rest was advised for more than 6 hours for hemostasis. The patients remained under observation for one day following the procedure. The biopsy specimens were fixed in formalin and embedded in paraffin. Each biopsy specimen was analyzed by three experienced hospital pathologists. The histologic grading and staging of chronic hepatitis were performed based on standardized guidelines proposed by the Korean Study Group for the Pathology of Digestive Diseases [
Frequencies and percentages were used for the descriptive statistics. Statistical differences between the groups were investigated using the chi-squared test and Student’s
In all 1994 patients, hepatitis B virus (HBV) was the most common indication for a liver biopsy (25.3%), followed by suspected malignancy (20.5%), nonalcoholic fatty liver disease (NAFLD) (13.0%), drug-induced liver disease (DILI) (11.7%), and alcohol-related liver disease (8.2%) during the overall study period (
In addition, there were two noticeable discordant cases. A 58-year-old male was suspected to have liver cirrhosis but it was difficult to identify whether the cause of the cirrhosis was HBV or alcohol by serological and radiological testing. However, after the liver biopsy, the pathologic analysis indicated that the main cause of the chronic liver injury was alcohol rather than HBV. Valuable information was obtained from the liver biopsy, and without the liver biopsy, it would not have been possible to accurately diagnose the patient using only his history or laboratory results. The second noticeable discordant case was a 48-year-old female diagnosed with a benign mass, although the suspected diagnosis was focal fat-sparing before the liver biopsy. She had a hypoechoic lesion in the background of fatty liver by ultrasonography and was suspected to have a fatty liver with a focal fat-sparing zone. Additional CT and magnetic resonance imaging (MRI) tests could not differentiate between the focal fat-sparing zone vs. the benign mass in aspects such as focal nodular hyperplasia. A targeted liver biopsy was performed for the differential diagnosis of the focal lesion and it was finally confirmed as a hemangioma.
Among the 1,944 patients included in the study population, 563 (30.0%) had liver cirrhosis at the time of their liver biopsy. In these patients, a liver biopsy was conducted to determine the etiology of the liver cirrhosis and accurately determine the degree of liver fibrosis. Most of the pre-biopsy diagnoses were consistent with the final diagnoses, including cryptogenic liver cirrhosis. However, in contrast, the histological results of these patients revealed that 16% (90 of 563) had no liver cirrhosis. The histological results of the patients diagnosed with cirrhosis by ultrasound were: F1 (nine patients, 1.60%), F2 (six patients, 1.07%), F3 (75 patients, 13.32%), and F4 (473 patients, 84.01%). As shown in
Painkillers were prescribed during hospitalization for 116 patients (6.0%) because of pain and there was no record of hypotension or pneumothorax. However, this was a retrospective study and thus, these minor complications may have been underestimated.
The major complication rate was 0.05%. A single major complication associated with a liver biopsy occurred during the study period. The patient was a 61-year-old female who had cryptogenic liver cirrhosis. She had thrombocytopenia (71,000/µL) and prolonged prothrombin time (15.5 seconds). Her baseline Child-Pugh score was 6, and her MELD score was 13. The liver biopsy was done without immediate complications and the patient was discharged without procedure-related complications by abdominal ultrasonography. However, four days post-liver biopsy, the patient was re-admitted due to abdominal pain and delayed bleeding from the biopsy site was found. The patient’s vital signs and laboratory findings, including hemoglobin, were stable and no procedure was required for hemostasis. The patient was discharged successfully after supportive care for 5 days.
In this study, we retrospectively analyzed all of the cases of ultrasonography-guided percutaneous liver biopsies for the past 18 years. Currently, percutaneous liver biopsy under ultrasonography guidance has been the most generally used method. Only when there were specific reasons, such as high bleeding tendency or large amounts of perihepatic ascites, were transjugular or CT-guided liver biopsies performed instead of ultrasonography-guided percutaneous liver biopsies. We included only patients who underwent ultrasonography-guided percutaneous liver biopsies to exclude those extreme cases and analyze a homogenous and standardized liver biopsy study population. The number of liver biopsies decreased from 2013 to 2018 and the indications for liver biopsies also changed by the period. The most common indication for liver biopsy during the entire study period was chronic HBV, followed by malignancy, and NAFLD. However, since 2013, the rate of NAFLD has increased rapidly, making it the most common indication. The rate of liver biopsies for chronic HBV and malignancy has declined, whereas the incidence of DILI, alcohol-related liver disease, and autoimmune hepatitis/PBC has increased sharply in the recent 5 years. Overall, a major adverse event of delayed bleeding at the biopsy site was identified, which was resolved by conservative management.
The decrease in liver biopsies for viral hepatitis was mainly attributed to the development of noninvasive methods for the assessment of liver fibrosis; the introduction of potent antiviral agents, such as entecavir, tenofovir, and direct-acting agents for hepatitis C; and the expansion of indications for antiviral treatment. In addition, advances in radiologic imaging techniques have reduced the number of liver biopsies for malignancies. Fatty liver has recently emerged as a major etiological factor underlying liver disease, which has rapidly increased the number of liver biopsies for NAFLD. These dynamic changes in indications for liver biopsy showed epidemiological changes over time in the study population.
Liver biopsy has been considered the most specific test to assess liver disease [
Liver biopsies, however, have their own unique role in the differential diagnosis and management of various liver diseases. A liver biopsy is essential for the diagnosis of autoimmune hepatitis, PBC—especially anti-mitochondrial antibody-negative—and several rare infiltrative diseases, such as amyloidosis and the hepatic involvement of lymphoma [
In terms of safety, our study reported a substantially lower rate of major complications (0.05%) than previous reports. In addition, we found that liver biopsies were safe in patients with cirrhosis and/or ascites, contrary to the concerns of clinicians. Some studies have reported that perihepatic ascites did not significantly affect the major and minor complication rates of image-guided percutaneous liver biopsies [
There were several limitations associated with this study mainly due to its retrospective format. Minor complications, such as pain, could not be fully investigated because of the retrospective design. Therefore, the incidence of complications may be underestimated. Second, since this study was conducted in a single institution, it may not represent all patient types. However, some generalizations are possible because several doctors had performed liver biopsies independently even in a single institution and we examined all patients who underwent liver biopsies in the institution over 18 years, achieving a large study population. To represent the dynamic changes in liver biopsy indications in South Korea, further multicenter studies to validate our results are necessary.
In conclusion, a liver biopsy is an irreplaceable diagnostic method, even in this era of noninvasive techniques. Also, we found that ultrasonography-guided liver biopsy was safe for patients with liver cirrhosis.
Conceptualization: JJ Yoo; methodology: Y Chang and JI Kim; formal analysis: B Lee; investigation: JI Kim; resources: SW Jeong and JJ Jang; data curation: SG Kim; writing – original draft preparation, Y Chang and JJ Yoo; writing – review & editing, JJ Yoo; visualization, B Lee; and supervision, SG Kim and YS Kim.
This work was supported by Soonchunhyang University research fund.
computed tomography
drug-induced liver disease
hepatitis Bvirus
hepatocellular carcinoma
model for end-stage liver disease
magnetic resonance imaging
nonalcoholic fatty liver disease
primary biliary cholangitis
Indications for percutaneous liver biopsies during the overall study period (2001 to 2018). Chronic hepatitis B was the most common indication for liver biopsy, followed by malignancy, non-alcoholic fatty liver disease, drug-induced liver disease, and alcohol-related liver disease (8.2%). LC, liver cirrhosis; HBV, hepatitis B virus; NAFLD, non-alcoholic fatty liver disease; PBC, primary biliary cholangitis; HCV, hepatitis C virus.
Changes in liver biopsy indications over time. Liver biopsies for viral hepatitis and malignancy decreased; whereas liver biopsies for non-alcoholic fatty liver disease, drug-induced liver disease, alcohol, and autoimmune hepatitis increased. LC, liver cirrhosis; PBC, primary biliary cholangitis; NAFLD, non-alcoholic fatty liver disease; HCV, hepatitis C virus; HBV, hepatitis B virus.
Baseline characteristics of the study population
Total (n=1,944) | Before 2008 (n=456) | 2008 to 2012 (n=859) | After 2013 (n=629) | ||
---|---|---|---|---|---|
Age (years) | 48 (37–57) | 47 (35–57) | 46 (37–54) | 51 (41–60) | <0.001 |
Males | 1,135 (58.4) | 336 (73.7) | 481 (56.0) | 318 (50.5) | <0.001 |
Laboratory findings | |||||
Hemoglobin (g/dL) | 13.05±1.97 | 13.26±2.02 | 13.19±1.88 | 12.7±2.02 | <0.001 |
Platelets (103/µL) | 202.5±85.74 | 193.78±85.23 | 201.56±87.13 | 210.1±83.64 | 0.007 |
Total bilirubin (mg/dL) | 0.82 (0.59–1.26) | 0.78 (0.55–1.19) | 0.85 (0.62–1.27) | 0.82 (0.6–1.31) | 0.011 |
Albumin (g/dL) | 3.92±0.56 | 3.93±0.53 | 4.01±0.54 | 3.79±0.58 | <0.001 |
AST (U/L) | 51 (31–94) | 52 (31–98) | 48 (30–91) | 52 (32–95) | 0.195 |
ALT (U/L) | 48 (24–110) | 45 (25–93.5) | 47 (24–108.5) | 50 (24–118.75) | 0.701 |
Creatinine (mg/dL) | 1.02±0.97 | 1.05±0.83 | 1.00±0.89 | 1.02±1.15 | 0.576 |
PT (INR) | 1.08±0.16 | 1.09±0.14 | 1.08±0.19 | 1.09±0.15 | 0.560 |
Liver cirrhosis | 563 (28.96) | 141 (30.92) | 244 (28.41) | 178 (28.30) | 0.567 |
Medication | |||||
Antiplatelet | 56 (2.88) | 9 (1.97) | 32 (3.73) | 15 (2.38) | 0.129 |
Warfarin | 2 (0.10) | 1 (0.22) | 0 (0.00) | 1 (0.16) | 0.432 |
NSAID | 73 (3.76) | 16 (3.51) | 35 (4.07) | 22 (3.50) | 0.803 |
Values are presented as mean±standard deviation, median (range) for continuous variables, or frequency (percentage) for categorical variables. The proportions are presented as percentages for categorical variables.
AST, aspartate aminotransferase; ALT, alanine aminotransferase; PT, prothrombin time; INR, international normalized ratio; NSAID, non-steroidal antiinflammatory drugs.
Trends in the indications for liver biopsy over time
Total (n=1,944) | Before 2008 (n=456) | 2008 to 2012 (n=859) | After 2013 (n=629) | ||
---|---|---|---|---|---|
HBV | 492 (25.3) | 121 (26.5) | 269 (31.3) | 102 (16.2) | <0.001 |
Malignancy | 398 (20.5) | 158 (34.6) | 135 (15.7) | 105 (16.7) | <0.001 |
NAFLD | 253 (13.0) | 37 (8.1) | 108 (12.6) | 108 (17.2) | <0.001 |
Drug-induced liver disease | 228 (11.7) | 25 (5.5) | 109 (12.7) | 94 (14.9) | <0.001 |
Alcohol-related liver disease | 160 (8.2) | 32 (7.0) | 51 (5.9) | 77 (12.2) | <0.001 |
Autoimmune hepatitis or PBC | 158 (8.1) | 25 (5.5) | 46 (5.4) | 87 (13.8) | <0.001 |
HCV | 114 (5.9) | 20 (4.4) | 77 (9.0) | 17 (2.7) | <0.001 |
Others | 107 (5.5) | 27 (5.9) | 54 (6.3) | 26 (4.1) | <0.001 |
Etiology of liver cirrhosis | 34 (1.7) | 11 (2.4) | 10 (1.2) | 13 (2.1) | <0.001 |
Values are presented as number (%).
HBV, hepatitis B virus; NAFLD, nonalcoholic fatty liver disease; PBC, primary biliary cholangitis; HCV, hepatitis C virus.
Diagnostic concordance before and after biopsy
Final diagnosis | Suspected diagnosis before liver biopsy |
Total | ||||||||
---|---|---|---|---|---|---|---|---|---|---|
HBV | HCV | Alcohol | NAFLD | AI/PBC | DILI | Malignancy | LC | Etc. | ||
Normal | 1 | 35 | 36 | |||||||
HBV | 491 | 3 | 494 | |||||||
HCV | 113 | 113 | ||||||||
Alcohol | 1 | 160 | 1 | 1 | 163 | |||||
NAFLD | 245 | 3 | 5 | 1 | 254 | |||||
Autoimmune hepatitis | 1 | 56 | 7 | 2 | 66 | |||||
PBC | 55 | 2 | 1 | 58 | ||||||
Overlap syndrome | 24 | 2 | 26 | |||||||
DILI | 1 | 1 | 5 | 200 | 1 | 2 | 210 | |||
HCC | 278 | 278 | ||||||||
IHCC | 53 | 53 | ||||||||
Other cancer | 40 | 1 | 41 | |||||||
Benign mass | 1 | 10 | 4 | 15 | ||||||
Cryptogenic LC | 1 | 3 | 32 | 2 | 38 | |||||
Rejection | 3 | 3 | ||||||||
Infection | 6 | 6 | ||||||||
Unknown | 5 | 15 | 11 | 8 | 51 | 90 | ||||
Total | 492 | 114 | 160 | 253 | 158 | 228 | 398 | 34 | 107 | 1,944 |
HBV, hepatitis B virus; HCV, hepatitis C virus; NAFLD, non-alcoholic fatty liver disease; AI, autoimmune hepatitis; PBC, primary biliary cholangitis; DILI, drug-induced liver injury; LC, liver cirrhosis; HCC, hepatocellular carcinoma; IHCC, intrahepatic cholangiocarcinoma.
Baseline characteristics of the study population with or without a diagnosis of liver cirrhosis
Total (n=1,944) | Liver cirrhosis (-) (n=1,382) | Liver cirrhosis (+) (n=562) | ||
---|---|---|---|---|
Age (years) | 48 (37–57) | 45 (34–55) | 53 (45–61) | <0.001 |
Males | 1,135 (58.35) | 763 (55.21) | 372 (66.07) | <0.001 |
WBC (103/µL) | 5,700 (4,500–7,040) | 5,995 (4,800–7,300) | 5,070 (3,905–6,500) | <0.001 |
Hemoglobin (g/dL) | 13.05±1.97 | 13.29±1.92 | 12.46±1.99 | <0.001 |
Platelets (103/µL) | 200 (146–251) | 220 (175–263) | 137 (92–194) | <0.001 |
Total bilirubin (mg/dL) | 0.82 (0.59–1.26) | 0.76 (0.56–1.13) | 0.97 (0.72–1.47) | <0.001 |
Albumin (g/dL) | 3.92±0.56 | 4.03±0.51 | 3.64±0.57 | <0.001 |
AST (U/L) | 51 (31–94) | 51 (29–100) | 50 (34–86) | 0.537 |
ALT (U/L) | 48 (24–110) | 58 (26–147) | 34 (21–57) | <0.001 |
Creatinine (mg/dL) | 1.02±0.97 | 1.02±0.97 | 1.03±0.97 | 0.820 |
PT (INR) | 1.08±0.16 | 1.05±0.14 | 1.17±0.19 | <0.001 |
Ascites | <0.001 | |||
None | 1,784 (91.72) | 1,335 (96.6) | 449 (79.75) | |
Mild | 137 (7.04) | 43 (3.11) | 94 (16.7) | |
Moderate to severe | 24 (1.23) | 4 (0.29) | 20 (3.55) | |
Child-Pugh class | ||||
Class A | N/A | N/A | 413 (73.36) | |
Class B or C | N/A | N/A | 150 (26.64) | |
MELD | N/A | N/A | 8.75 (7.6–10.59) | |
Complication | 1 | 0 | 1 | 0.289 |
Values are presented as mean±standard deviation, median (range) for continuous variables, or frequency (percentage) for categorical variables. The proportions are presented as percentages for categorical variables.
WBC, white blood cell; AST, aspartate aminotransferase; ALT, alanine aminotransferase; PT, prothrombin time; INR, international normalized ratio; N/A, not applicable; MELD, model for end-stage liver disease.