Abstract

Background and Aim: Currently, an increasing number of liver biopsies are performed by radiologists under real-time ultrasound control. A routine ultrasound assessment of a puncture site before performing percutaneous biopsy is reported to increase diagnostic yield and decrease complication rates. It is not clear if real-time ultrasound is superior to marking the puncture site before biopsy as regards reducing biopsy size and avoiding fragmentation and complications. The aim of this study was to compare ultrasound assessment of the puncture site before performing percutaneous liver biopsy with real-time ultrasound liver biopsy for suspected diffuse liver disease.
Methods: Consecutive percutaneous liver biopsies (n = 631) for diffuse liver disease were evaluated. Group A consisted of patients who had real-time guided-ultrasound biopsy performed by radiologists (241 patients; M/F, 35/106; median age 48 year [range, 17-76]; needle 18 G). Group B patients were assessed by radiologists using ultrasound of the puncture site on the same day that biopsies were performed by experienced gastroenterologists/hepatologists on the ward using the marked site (390 patients; M/F, 276/114; median age 43 year [range, 15-75]; needle 16 G).
Results: There were no differences in severity of liver disease, establishing a diagnosis (OR, 1.92 [95% CI, 0.84-4.34]; P = 0.12), length of liver biopsy specimens, number of fragments or complications. Two independent variables were significantly associated with a histological diagnosis: longer biopsy length (P < 0001) and fragment number of two or less (P < 0.001).
Conclusion: Real-time ultrasound did not improve diagnostic yield or result in fewer complications. Marking the puncture site seems adequate and has the practical advantage that it takes up less of the radiologists' time.

Introduction

Traditionally, liver biopsy has been performed by hepatologists or gastroenterologists. Anatomical variation in livers may result in failure to obtain hepatic tissue at liver biopsy without imaging guidance using the so-called blind biopsy.^[1] In everyday clinical practice, an increasing number of liver biopsies are performed by radiologists under ultrasound control.^[2,3] Ultrasonography, as well as being a good screening test for liver disease,^[2] allows selection of the optimal puncture site before performing biopsy. The use of ultrasonography by marking the site for percutaneous biopsy has been reported to increase diagnostic yield and decrease complication rates.^[4]

It is not clear as to whether performing real-time ultrasound (i.e. at the time of liver biopsy) is any more effective than marking the puncture site shortly before biopsy , in terms of biopsy size, fragmentation and complications. Our aim was to assess these factors in a retrospective review of consecutive patients undergoing liver biopsies for suspected liver parenchymal disease

Methods

Patient Selection and Biopsies

Over a period of 10 years (January 1995 to September 2005), we performed percutaneous liver biopsy in 631 consecutive patients at two liver centers in Athens, Greece: Evangelismos General Hospital and Polyclinic General Hospital. All patients gave written informed consent before the procedure. In all patients the indication for biopsy was to establish a diagnosis and to assess severity of a suspected parenchymal chronic liver disease.

Our cohort was divided in two groups: (i) in group A, real-time ultrasound-guided liver biopsy was performed by radiologists (241 patients; M/F, 135/106; median age 48 year (range 17-76). Bard Biopty instrument with a needle of 18 G (Bard Biopsy-Cut biopsy needle, Bard, Murray Hill, NJ, USA) was used (ultrasound devices: GE Logic 9, GE Logic 400 [General Electric, USA], ACUSON 128XP/10 [Acuson, Siemens, Malvern, PA, USA]). All procedures were performed at Evangelismos General Hospital; (ii) in group B an ultrasound identification of a safe biopsy site was performed in the radiological department by a radiologist on the same day as the liver biopsy. The time needed for the patient to be transferred from the radiology department to the ward where the biopsy was performed was approximately between 10 and 35 min. There were 390 patients (M/F, 276/114; median age 43 year [range 15-75]), biopsies were performed using a 16-G Trucut needle (CardinalHeath, McGaw Park, IL, USA). After a plane for optimum direction of the puncture had been selected and a mark placed, the procedure was then undertaken by an experienced gastroenterologist (SM) on the ward using the marked site without the ultrasound facility. All the biopsies in this group of patients were performed at Polyclinic General Hospital.

In both groups a second pass was attempted if the specimen was considered inadequate, by introducing the biopsy needle at the same entry site and redirecting it. Neither group had a third pass attempted. The biopsy procedures and histology were evaluated retrospectively from prospectively collected data at the time of biopsy. The requisite laboratory criteria considered safe for a percutaneous liver biopsy, the liver biopsy procedure itself, as well as the follow-up protocol that was used, have been described elsewhere.^[5] All patients were followed in hospital for 24 h. Further imaging after biopsy was obtained in the case of persistent pain, orthostatic hypotension or decreasing hemoglobin levels.

End Points in the Study

The safety of both techniques of liver biopsy was assessed according to major complications: (i) transfusion of blood; (ii) surgical intervention; and (iii) hospital stay more than 48 h after the procedure. Minor complications were not evaluated as there was inconsistent recording in the notes.

Definitive histological diagnosis was used in order to compare the adequacy of samples between the two groups.

Definitions

Length of biopsy was the aggregate length of all the fragments. The diagnosis was considered definitive if this was stated in the histopathology report or the report contributed to a definitive diagnosis by exclusion of other potential causes of liver disease. Cases in which either the available liver tissue was inadequate or there was no liver tissue at all were considered as having no diagnosis. We did not make any assessment of the histological quality (e.g. number portal tracts).

Statistical Analysis

Analyses were performed in STATA 8.0 (Stata Corporation, College Station, TX, USA). The clinical, biochemical and virologic data are presented as median with range. Continuous variables were compared using the Wilcoxon rank sum test and Mann-Whitney U-test. Categorical variables were compared using X² or Fisher's exact tests as appropriate. Univariate and multivariate logistic regressions were used to address possible relationships between baseline characteristics and the primary outcome. P-values < 0.05 were considered statistically significant. All P-values are two-tailed.

Results

The diagnoses made histologically in the two groups are shown in Table 1 . The patients in both groups were well matched at time of biopsy with respect to proportion of cirrhotics (group A, 62 (25.7%); group B, 84 (21.5%); P = 0.19) and severity of liver disease ( Table 2 ). Patients in group B were younger (group A, 48 (17-76); group B, 43 (15-75); P = 0023) and there was a greater proportion of men (M/F group A, 135/106; group B, 276/114; P < 0.0001).

There was no statistically significant difference in establishing a diagnosis between the two methods (OR, 1.92 [95% CI, 0.84-4.34]; P = 0.12). A definitive pathological diagnosis could not be achieved in eight patients (3.35%) in group A and 24 (6.15%) in group B. No histological diagnosis was possible in eight patients (four in each group) because of excessive fragmentation of the sample, and in 24 (four in group A and 20 in group B) because no liver tissue was obtained. In 15 patients (two in group A) with no liver tissue in the needle a second pass was attempted unsuccessfully, while six refused a second pass. In two patients a second pass was not attempted because of pain post-biopsy. In group A, a second pass was performed in 32 patients (13.3%) and in group B in 48 patients (12.3%; P = 0.42). Twenty-one of 32 (67%) patients with no histological diagnosis were obese (16 were males) and seven were overweight.

There was no difference regarding the length of liver biopsy specimens, whether assessed as means (group A, 1.43 cm ± 0.59 vs group B, 1.43 cm ± 0.61; P = 0.63) or as medians (group A, 1 cm [IQ range 1-2] vs group B, 1 cm [IQ range 1-2]; P = 0.63). There was no difference in the mean number of fragments (group A, 1.41 ± 0.75 vs group B, 1.38 ± 0.87; P = 0.15) or median number (group A, 1 [IQ range 1-2] vs group B, 1 [IQ range 1-1]; P = 0.15).

Using step-wise logistic regression, two independent variables were statistically significantly associated with a confirmed diagnosis: longer biopsy length (P < 0001) and fragment number of two or less (P < 0.001). The interaction of length and fragments was not statistically significant (P = 0.9).

For a biopsy length of more than 1 cm, the likelihood of confirming the diagnosis was 14.72 (95% CI, 1.94-111.72) times higher than if <1 cm (P = 0.009). If the number of fragments at biopsy was three or more, the biopsy was 9.09 times (95% CI, 3.03-24.39; P < 0.001) less likely to be diagnostic.

There was only one major complication consisting of a post-biopsy bleed (group B) in a 25 year old male with chronic hepatitis C requiring 2 units of blood. The histological diagnosis was mild hepatitis without evidence of cirrhosis. There were no deaths.

Discussion

The results of this study confirm that ultrasound-guided liver biopsy is a safe and efficient method in the evaluation of patients with diffuse liver disease. We did not observe any significant difference between the real-time ultrasound biopsy and 'X marks the spot' technique in terms of post-biopsy complications or ability to have a definitive pathological diagnosis. There was, however, a trend toward fewer cases without liver tissue in the biopsy needle in the group of real-time ultrasound, particularly in obese patients.

The wide availability of ultrasonography has changed the practice of performing liver biopsy so that in many centers most biopsies are now carried out under direct ultrasound visualization. Although the benefit of an ultrasound-guided liver biopsy for diffuse disease is not clear, it has been considered safer,^[6-8] more comfortable^[7] and only marginally more expensive^[7] but cost-effective^[9] compared with blind liver biopsies. This change in methodology has had major implications on the training of gastroenterologists in many countries.

The 'X marks the spot' method is a hybrid between 'blind' biopsy and real-time ultrasound-guided biopsy. This has evolved as current clinical practice in centers where gastroenterologists are proficient in ultrasound, or, as in our case, when radiologists do not have sufficient time to perform percutaneous liver biopsies other than for guided biopsies for intrahepatic lesions. Pasha et al. ^[10] are of the opinion that ultrasound performed before biopsy is not as safe as a real-time ultrasound-guided biopsy as the patient may move prior to the biopsy and intrahepatic vessels cannot be avoided, but did not present data to support this. In contrast to this opinion, our data have shown that major complications following biopsy and the efficacy in establishing pathological diagnosis do not differ between the two techniques.

It may be argued that real-time ultrasound might be superior in terms of obtaining an adequate biopsy sample; no liver tissue was obtained in 20 patients in the 'X marks the spot' technique compared with four in the real-time ultrasound group. We re-examined the records of the above patients and we found that the vast majority were obese or overweight at the time the biopsy was undertaken. Real-time ultrasound-guided biopsy may be the preferred method in obese patients, but transjugular biopsy may be even better.^[11] It may be that caution exercised in judging the depth of insertion of the needle during 'blind biopsy' results in inadequate specimens, compared with the real-time technique. Another advantage of real-time ultrasound is that the biopsy obtained is less likely to be subcapsular in location and thus avoids some histological difficulty in assessing fibrosis; however, we did not have sufficient information to compare the two biopsy techniques for this parameter.

Recently it has been reported^[12] that the size of liver biopsy strongly influences the grading and staging of chronic viral hepatitis. It has been proposed that a liver biopsy should be at least 2.5 cm in length to reduce the sampling error.^[13] In our patients, the size of biopsy specimen was similar to other series^[6,14,15] and is thus consistent with clinical practice in several different countries. However, only seven (2.9%) biopsy samples in group A and nine (2.3%) in group B were ≥2.5 cm, and only 25 (10.3%) in group A and 39 (10%) in group B were ≥2.0 cm. The average length was below this partly due to the fact that 25.4% had cirrhosis. In non-cirrhotic patients in our series who had chronic hepatitis, the median biopsy specimen length was 1.4 cm (0.4-3.1); in group A it was 1.3 cm (0.5-2.5) and in group B it was 1.5 cm (0.4-3.1; P = 0.09). The proposed optimal criteria for biopsy length was obtained in only 14% of those with chronic hepatitis.^[16] In a recent systematic review^[11] of length and number of portal tracts obtained with percutaneous liver biopsy, which evaluated 162 studies, the mean length was 13.6 ± 3.2 mm using a Trucut needle with ultrasound guidance.

A potential bias in interpreting the results in our comparative study is that in group A the needle was 18 G and in group B, 16 G. It could be argued that the ability to confirm diagnosis would favor group B. Rocken et al.^[17] evaluated whether thin-needle biopsy gave enough material to study parenchymal diseases. The histopathological diagnoses showed no major differences between 17-G and 20-G needles across all etiologies of liver disease. Recent data show that mean length does not vary significantly with respect to needle diameter,^[11] therefore the different in needle size is unlikely to have biased our interpretation of the results with respect to ultrasound technique. The number of portal tracts is likely to be affected by the gauge of the needle, but in our study, diagnosis rates were not lower in the group with the 18-G needle versus the 16-G needle. Radiologists most often use an 18-G needle as this is satisfactory for biopsy specimens of non-hepatic tissue.

It is unlikely that the difference in size of needle affected complications which are rare.^[18] In our group B only one patient had hemoperitoneum similar to other series.^[4,19] It is well documented that 60% of complications occur during the first 2 h post-biopsy,^[19] so that rapid transfer is needed, if real-time ultrasound technique is used for the patient to have appropriate monitoring on the ward.^[20]

Our data show that both methods of liver biopsy using ultrasound are equally effective and safe; however, there are some disadvantages of real-time ultrasound biopsy performed by radiologists in the radiology suite, which is increasingly more commonplace.^[2,3] First, the time needed to transfer the patient from the radiology department to the ward or monitoring area, and secondly, increasing the workload of radiologists. Gastroenterologists or hepatologists either using 'X marks the spot' technique if they have insufficient ultrasound training, or they themselves using real-time ultrasound on the ward would relieve radiologists' time, and provide adequate specimens for histological diagnosis. Length of specimens obtained by either technique is the major limiting factor in achieving an adequate diagnosis, as has been documented in the literature.^[11] In obese patients, real-time ultrasound may be preferable as in our study, fewer inadequate specimens were obtained.

http://www.medscape.com/viewarticle/561875_4

 

A rigorous method for liver biopsy

 

Liver biopsy is still considered the gold standard for grading, staging and "stad-ging" the chronic liver disease. In addition, it remains a primary source for acquiring new knowledge on the liver pathology. Demand for precise evaluations of the fibrosis and inflammatory tissue detectable in liver biopsy samples has been fuelled by the need to understand the closest-to-real effects of new antiviral molecules on the lesions characterising the histological patterns of chronic viral, toxic, metabolic and autoimmune diseases. The current scoring systems do not quantify these lesions, but only describe subjective classes of severity labelled with ordinal numbers, and the available automated methods based on observer-computer interactions do not abolish observer subjectivity or use an inadequate measurement unit, and also take too long to analyse entire histological sections.

A research article to be published on December 28, 2008 in the World Journal of Gastroenterology addresses this question. The research team led by Nicola Dioguardi from Italy described a quantitative analysis method of liver biopsy sections with a machine called "Metriser" which, at a speed of 0.1 mm2/s, automatically measures the residual hepatocyte mass (including hepatocytes vacuolisation), inflammation, fibrosis and the loss of liver tissue tectonics.

In the absence of any other means of obtaining correct reproducible information concerning the status of liver tissue, the authors explored the possibility of constructing the first totally computer-aided and strictly objective method of rigorously, rapidly and easily obtaining metrical measurements of liver lesions directly from bioptic specimens. The method provides: (1) the metrical extension in two-dimensions of the residual hepatocellular set including the area of vacuoles pertinent to abnormal lipid accumulation; (2) the geometric measure of the inflammation basin, distinguishing intra-basin space and extra-basin dispersed parenchymal leukocytes; (3) he magnitude of collagen islets, which were considered truncated fractals and classified into three classes of magnitude; and (4) the Tectonic Index that quantifies alterations (disorders) in the organization of liver tissue.

This study not only introduced a new kind of liver biopsy measurement but also described a histological picture in verbal and repeatable terms.

 

 

Reference: Dioguardi N, Grizzi F, Fiamengo B, Russo C. Metrically measuring liver biopsy: A chronic hepatitis B and C computeraided morphologic description. World J Gastroenterol 2008; 14(48): 7335-7344 http://www.wjgnet.com/1007-9327/14/7335.asp

Correspondence to: Nicola Dioguardi, MD, Scientific Superintendence, Istituto Clinico Humanitas IRCCS, Via Manzoni 56, Rozzano 20089, Milan, Italy. nicola.dioguardi@humanitas.it

About World Journal of Gastroenterology

World Journal of Gastroenterology (WJG), a leading international journal in gastroenterology and hepatology, has established a reputation for publishing first class research on esophageal cancer, gastric cancer, liver cancer, viral hepatitis, colorectal cancer, and H pylori infection and provides a forum for both clinicians and scientists. WJG has been indexed and abstracted in Current Contents/Clinical Medicine, Science Citation Index Expanded (also known as SciSearch) and Journal Citation Reports/Science Edition, Index Medicus, MEDLINE and PubMed, Chemical Abstracts, EMBASE/Excerpta Medica, Abstracts Journals, Nature Clinical Practice Gastroenterology and Hepatology, CAB Abstracts and Global Health. ISI JCR 2003-2000 IF: 3.318, 2.532, 1.445 and 0.993. WJG is a weekly journal published by WJG Press. The publication dates are the 7th, 14th, 21st, and 28th day of every month. WJG is supported by The National Natural Science Foundation of China, No. 30224801 and No. 30424812, and was founded with the name of China National Journal of New Gastroenterology on October 1, 1995, and renamed WJG on January 25, 1998.

About The WJG Press

The WJG Press mainly publishes World Journal of Gastroenterology

http://www.eurekalert.org/pub_releases/2008-12/wjog-arm123008.php