Liver Biopsy Sampling

天狼星

Semin Liver Dis 2004;  89-97
DOI: 10.1055/s-2004-823103
  


Liver Biopsy Sampling in Chronic Viral Hepatitis
  
Maria Guido1, 2 M.D. , Massimo Rugge2 M.D.  
1 Pathologist, Università Degli Studi di Padova, Azienda Ospedale, Padova, Italy
2 Dipartimento di Scienze Chirurgiche ed Oncologiche-Anatomia Patologica 2, Università Degli Studi di Padova, Azienda Ospedale, Padova, Italy


  
MORPHOLOGICAL APPROACH TO CHRONIC VIRAL HEPATITIS
TYPES OF BIOPSY SAMPLE
COULD A RANDOM SAMPLE REFLECT DAMAGE TO THE ORGAN AS A WHOLE?
  The Historical Perspective
  The Present Perspective
WOULD THE SAMPLE SIZE AFFECT THE HISTOLOGICAL ASSESSMENT OF CHRONIC HEPATITIS IN TERMS OF GRADE AND STAGE?
  The Historical Perspective
  The Present Perspective
CONCLUSIONS
REFERENCES
  

  

In chronic viral hepatitis, liver biopsy is performed to grade and stage liver damage. Liver biopsy samples are usually taken with a needle using a percutaneous procedure. This method produces a core of tissue representing approximately 1/50,000th of the liver mass, which raises concern about how representative a needle biopsy can be. A critical review of the literature reveals methodological limits unacceptable in this era of evidence-based medicine. Integrating earlier experience with more recently produced data indicates that a biopsy sample 2 cm or more in length containing at least 11 complete portal tracts should be reliable for grading and staging chronic viral hepatitis.

Top of Page

KEYWORDS
Liver biopsy - chronic hepatitis - viral hepatitis - histopathology - sampling error


  

In the past 20 years, pathologists have radically changed the way they generate their reports. In the fields of neoplastic and inflammatory diseases, increasingly sophisticated information, such as the results of molecular studies, is needed for diagnosis and treatment. The introduction of new methods for analyzing tissues has prompted an ever-closer interaction between clinician and pathologist to ensure the most appropriate sampling in a given clinical context. This is also true of liver diseases, particularly in chronic viral hepatitis, which means taking a new look at the old liver biopsy procedure.

Liver biopsy was introduced by Paul Ehrlich[1] more than 100 years ago and has since been widely used, enabling substantial progress in our understanding of liver histology, pathophysiology, and pathology. The diffusion of the use of liver biopsy was mainly because of the Menghini “one-second liver biopsy” technique,[2] which provided samples suitable for various morphological studies, including the histochemical, immunohistochemical, ultrastructural and, more recently, molecular biology studies. Low mortality and relatively low morbidity rates have facilitated the widespread use of liver biopsy procedures.[3]

Because of advances in imaging techniques and the development of reliable serological and virological tests, the indications for liver biopsy have changed dramatically, and the technical biopsy procedures have also been refined.[4] Different methods for obtaining liver biopsy samples are available,[5] and the choice of technique is dictated by the clinical setting (e.g., associated coagulation disorders, presence or absence of ascites, results of prior ultrasound examination), the operator's preference and experience, and the nature of the disease prompting the histological investigation.

The purpose of this article is not to review the topic of “liver biopsy,” which includes many relevant (and still debated) issues, such as clinical indications and contraindications, complications, costs, and benefits. Our aim is to focus on the problem of “sampling errors” and when a sample can be considered to be “adequate” if a liver biopsy is needed to evaluate the severity of chronic hepatitis, especially in cases due to the most frequent hepatotropic viruses, in other words, hepatitis B virus (HBV) and hepatitis C virus (HCV).

As we have learned from epidemiological studies, sampling errors occur when the samples considered fail to represent the target population (e.g., tissue in our setting) from which it is derived. Considering that a biopsy sample taken from an adult corresponds to a fraction of just 1/50,000th to 1/100,000th of the whole liver, a biopsy specimen would seem to be inadequate in the case of nonfocal diseases, such as a chronic viral hepatitis, in which the liver changes may be unevenly distributed.

There are several facets to the problem, but, from a practical point of view, it can be brought down to two partially overlapping questions: (1) could a random sample reflect damage to the organ as a whole and (2) would the sample size affect the histological assessment of chronic hepatitis in terms of grade and stage?

Top of Page

MORPHOLOGICAL APPROACH TO CHRONIC VIRAL HEPATITIS
Because of the identification of HCV and the effectiveness of antiviral treatments, the diagnostic approach to chronic hepatitis has undergone a “Copernican revolution” in the last two decades.[6][7][8][9][10] The unfavorable prognostic value of the “time-honored concept of chronic active hepatitis” (CAH)[10] based on the presence of interface hepatitis (formerly piecemeal necrosis), as opposed to the favorable diagnosis of “chronic persistent hepatitis” (without interface hepatitis by definition), is no longer accepted. The new reporting system integrates etiology with morphological findings to define prognosis and treatment indications.[11] The grading and staging system, borrowed from oncological practice,[11][12] takes into account (along with interface hepatitis) the whole spectrum of morphological lesions affecting progression to cirrhosis. Grading reflects necroinflammatory severity of the different elementary histological lesions (e.g., portal inflammation, interface hepatitis, focal and confluent lobular necrosis). Staging quantifies the extent of fibrosis and defines disease progression toward its potential cirrhotic end point.

While an etiologic diagnosis of chronic hepatitis is easily obtained in most cases using serological and virological tests, biopsy is currently indicated for the grading and staging of liver damage.[13][14][15][16][17] In practice, pathologists establish the grade and stage of disease by scoring each elementary lesion according to semiquantitative numerical systems.[6][11][18][19][20][21] Several studies have evaluated intraobserver and interobserver consistency using such scoring systems[22][23][24] and have shown an acceptable consistency for necroinflammatory lesions and an almost perfect agreement for fibrosis, whatever the scoring system used. In clinical practice, numerical scores may be translated into descriptive severity categories (mild, moderate, and severe) and have been shown to have prognostic and therapeutic implications.[25][26] Scoring has been routinely applied in clinical trials on the treatment of chronic hepatitis B and C[27][28][29][30] to evaluate treatment eligibility and effects. Although much effort has been made (and must still be made) to identify alternative, noninvasive methods,[31][32][33][34] liver biopsy remains the gold standard for grading and staging chronic viral hepatitis.[35][36][37] Unfortunately, none of the different scoring systems went so far as to evaluate the minimum sample size needed for the scores to be reliable.

Top of Page

TYPES OF BIOPSY SAMPLE
Two different types of liver tissue sample can be obtained: wedge and needle biopsy specimens.[38] A wedge biopsy can be taken during laparoscopic or open surgery. All pathologists would agree, however, that a wedge biopsy (which is a subcapsular sample) should not be taken to evaluate diffuse disease and chronic hepatitis in particular. In fact, fibrous septa spreading from Glisson's capsule to the adjacent parenchyma may mimic cirrhosis, which means that the stage of disease is overestimated. In addition, because of the surgical procedure, the subcapsular region is often affected by nonspecific necroinflammatory lesions, which can result in an overestimation of the grade of the disease's inflammatory component. So, even when laparoscopy or laparotomy is performed for clinical reasons, unless a focal lesion is to be tested, a needle biopsy (sampling the parenchyma in depth) should be preferred. Wedge biopsies are more suitable for biochemical studies, given the larger amount of tissue obtained. The risks and complications of wedge biopsies relate to the laparoscopy or laparotomy procedure.

Needle biopsy samples can be taken (1) during laparoscopic or open surgery; (2) percutaneously, either “blindly” or under radiological or ultrasound (US) guidance; and (3) by transjugular catheter.[3] This last procedure is reserved for subjects with severe coagulation disorders and is performed under cardiac monitoring, due to the risk of arrhythmia. Even in the most expert hands, it often produces only fragmented specimens.

Laparoscopic (or laparotomic) needle biopsy has declined considerably and is now used mainly when focal lesions are accidentally found during routine surgery. It can also be used in patients with focal lesions and coagulopathy instead of the transvenous procedure.

In current practice, percutaneous needle biopsy is the quickest, safest, and most-used technique for assessing diffuse diseases and for grading and staging chronic viral hepatitis.[13] Many centers use US or computed tomography (CT) guidance because of the lower rate of complications reported in some studies.[39][40] Whether, when, and how radiological or US guidance is indicated in liver biopsy is a matter that is still under debate.[39][40][41][42][43]

Two main types of needle are currently used: the Tru-Cut, which is a cutting needle, and the Menghini needle, which uses a suction technique.[40] The cutting needle usually produces a larger sample but has been associated with a higher risk of complications,[44] probably because the needle remains in the liver longer. A suction needle tends to produce more fragmented samples. As emphasized in the original article,[2] the advantage of the Menghini technique is that the needle remains in the parenchyma just “one second,” which minimizes the risk of rupturing the capsule. The size of the specimen varies, depending on the type of needle: a biopsy obtained with 14-gauge (G) to 21G needle is usually defined as “large,” whereas needles less than 21 G result in “thin” or “fine” biopsy samples with a core less than 1 mm in diameter. The thin biopsy is universally recommended for the diagnosis of focal lesions; its use for grading and staging is under debate.[45]

Complications after percutaneous liver sampling are related to the patient's clinical conditions, the operator's expertise, the type of needle, and the number of passes.[46] Elderly patients with malignant hepatic tumors or severe liver failure, or both, are at the greatest risk for developing complications.[4][47][48][49] A detailed discussion of liver biopsy contraindications and complications is beyond the purpose of this article, and readers are referred to specific references.[4][46][47][48][49][50] Guidelines have also been published on the practice of liver biopsy.[40][51][52][53]

Top of Page

COULD A RANDOM SAMPLE REFLECT DAMAGE TO THE ORGAN AS A WHOLE?
Answering this question is complicated by difficulties in designing satisfactory study models. Establishing a gold standard would involve testing the entire liver to compare different types of needle biopsy. Various approaches have been used to overcome this drawback in clinical studies.

Top of Page

The Historical Perspective
Earlier studies focused mainly on diagnosing cirrhosis, which is still a relevant clinical issue. The diagnosis of cirrhosis has important prognostic implications in terms of quality of life or life expectancy, or both. Cirrhosis has a significantly lower response to interferon treatment. Moreover, a diagnosis of cirrhosis prompts a follow-up program with a view to the early diagnosis of hepatocellular carcinoma.[54]

A first group of studies compared the morphological findings of multiple needle biopsy samples obtained from different portions of the liver.[55][56][57][58] On the whole, these studies varied considerably in terms of the number of patients involved, the study design (including biopsy and autopsy series), the type of needle (the diameter of which is often not mentioned), and the morphological criteria adopted. Diseases of different etiologies were included in the groups studied, which made it difficult to transfer the conclusions to the context of chronic viral hepatitis. In addition, these studies did not evaluate the extent of noncirrhotic fibrosis, so no information can be gleaned on sampling errors in precirrhotic stages, which is currently important in the management of chronic viral hepatitis. Finally, most of these studies did not include any evaluation of intraobserver and interobserver agreement, which is crucial in dealing with morphology implicating subjective interpretations. Overall, these studies reported a sampling error ranging from 10 to 70%.[55][56][57][58][59]

Among the “historical” studies, the one by Soloway et al[56] is one of the most quoted to support the inadequacy of percutaneous biopsy. In this study, based on the assumption that cirrhosis is irreversible, three to five needle biopsies were taken in 19 patients at 6- or 12-month intervals after the detection of cirrhosis. Multiple simultaneous biopsies from adjacent areas were also taken in 13 subjects. The authors concluded that “…evaluation of multiple simultaneous biopsies showed that the presence and grade of hepatitis in a single biopsy accurately reflected hepatitis in adjacent areas,” yet “there were many discrepancies in the presence and grade of cirrhosis.”[56] For several reasons, the results of this study can no longer be considered reliable: (1) the number of cases was limited; (2) the series included patients with not otherwise specified “chronic active liver diseases”; (3) there is no indication of the gold standard used to measure how representative the liver samples were; (4) there is no mention of the sample size, and particularly whether paired, multiple, or consecutive biopsies were of equivalent length and width (which is also a critical point in this type of study); (5) cirrhosis was “graded” on three levels, grade 1 indicating early cirrhosis or doubtful cases and grades 2 and 3 indicating definite cirrhosis (unfortunately the differences between grades 2 and 3 were not explained).

Three biopsy samples were obtained prior to autopsy using a modified Menghini method in the study by Abdi et al.[57] An 80% accuracy for the diagnosis of cirrhosis was demonstrated on the strength of one sample, and this rose to 100% when all three samples were evaluated jointly. Specimens were considered adequate if they were no less than 25 mm long, but no information was given on the diameter of the biopsy cores.

天狼星

In the study by Baunsgaard et al,[58] two specimens were simultaneously taken from the right lobe, 4 to 6 cm apart, in 70 patients using a Menghini needle. The indications for liver biopsy were not reported. To be considered in the study, biopsy samples had to be no shorter than “1 cm for specimens composed of 1 to 3 fragments and 1.5 cm for specimens comprising 4 or more fragments.” Diagnostic consistency was evaluated for cirrhosis and different types and grades of lesions, including steatosis, Mallory bodies, acidophilic bodies, portal and septal fibrosis, portal inflammation, focal intralobular necrosis, bile duct proliferation, alcoholic hepatitis, and cholestasis. The highest diagnostic consistency between paired samples was demonstrated for steatosis (τ = 0.82), cholestasis (τ = 0.91), and cirrhosis (τ = 0.83). Diagnostic inconsistency, however, is not necessarily due to sampling error: intraobserver variation was not evaluated, and it is not clear whether the minimum size referred to the sum of the various fragments or to the length of the longest one. The study did not mention the diameter of the sample and did not explain whether paired samples from the same subjects were of equivalent size.

The superiority of multiple samples over a single biopsy was also reported by Maharaj et al,[59] but the study did not give any information on the sample size or on the etiology of the cirrhosis. Although multiple samples were obtained from the same patient, no complications were observed in any of the 75 subjects enrolled.

The conclusion that can be drawn from the previous studies is that the more tissue there is available, the greater is the accuracy of the diagnosis.

A different approach to the issue of sampling errors has been to compare the diagnostic accuracy of laparoscopy (with or without associated laparoscopic biopsy) versus blind needle biopsy. Generally speaking, most such studies suggest that laparoscopy (alone or associated with biopsy) is more accurate than blind percutaneous biopsy is, which produces false-negative diagnoses of cirrhosis in 9 to 32% of cases.

Pagliaro et al[60] designed a prospective and randomized study to compare the diagnostic accuracy of blind percutaneous biopsy versus laparoscopic biopsy, which showed that liver cirrhosis was missed by percutaneous biopsy in 20% of cases. For various methodological reasons, this result is no longer convincing: (1) the gold standard for testing the strength of the two methods in diagnosing cirrhosis is not clear because the study enrolled patients with “clinical evidence of chronic diffuse well-compensated liver diseases,” which did not necessarily mean cirrhosis; (2) blind percutaneous biopsies and laparoscopic biopsies were performed in different subjects; (3) specimens were considered adequate if they were at least 1 cm long, but the mean length ± standard deviation of the specimens from the whole group was not reported; and (4) both the laparoscopy and the histological evaluation were performed by two different operators, but interobserver reproducibility was only reported for the laparoscopic diagnosis. A lesson applicable to clinical practice can still be found in the Pagliaro et al study: a final diagnosis can be reached by accurately integrating all the clinical and morphological findings. In the proper clinical context, even “minor” histological lesions can acquire “major” diagnostic significance.

In a more recent study by Poniachik et al,[61] a 32% histological sampling error was reported in patients with a laparoscopic diagnosis of cirrhosis. The study included primary biliary cirrhosis and primary sclerosing cholangitis, diseases known to feature different stages at the same time.[62]

More encouraging results were reported in a well-designed study comparing a blind needle biopsy with two additional biopsy samples obtained at laparoscopy from the right and left lobe of the same patient.[63] This study involved 60 patients, and various findings (other than cirrhosis) were assessed, including normal liver, minimal changes, chronic persistent hepatitis, and CAH. The diagnosis was based on a single percutaneous biopsy and coincided in all cases with the final diagnosis emerging from a joint evaluation of the laparoscopic biopsy findings. It is worth noting that all the biopsy specimens considered in this study were ≥2 cm long.

More recently, abdominal US has been deemed more accurate than percutaneous biopsy is in the diagnosis of cirrhosis in a study of 212 subjects.[64] Histology, however, revealed cirrhosis in 10 cases found negative after US, whereas US suggested cirrhosis in 32 histologically negative cases. The two procedures were not tested against any external gold standard, so it is not clear whether histology underestimated or US overestimated the presence of cirrhosis.[64]

Top of Page

The Present Perspective
Few published studies have addressed sampling error in the context of a modern morphological approach, which involves evaluating the grade and stage of hepatitis (Table [1]).


Table 1 Sampling Error in Grading and Staging of Chronic Hepatitis C

--------------------------------------------------------------------------------

Authors No. of Cases Needle Criteria for Adequacy Length of Samples Diameter of Samples No. of Portal Tracts Observer Agreement

--------------------------------------------------------------------------------

Fanning et al[65] 16 - > 5 PT - - - -
Regev et al[66] 124 16G Tru-Cut 1.5 cm > 5 PT - - - >90%
Persico et al[67] 52 18 G Menghini > 1.5 cm 2.8 + 1.1 cm

2.5 + 0.9 cm** - - 90%
Siddique et al[68] 29 1.7 Bard > 1.5 cm/> 4 PT 16.4 + 0.7 mm

16.6 + 0.6 mm[†] - Range 4-5 -

--------------------------------------------------------------------------------

PT, portal tracts
*
right lobe

†
left lobe

Fanning et al[65] performed a study based on double biopsies taken from subjects with chronic hepatitis C, their main goal being to evaluate viral load variability. Although viral load proved consistent between samples, 33% (4 of 12) and 25% (3 of 12) of the paired biopsies showed a discrepancy of ≥2 grades of inflammation and fibrosis, as evaluated using Ishak's score. The study included only 16 cases; however, it did not mention the size of the samples or the mean number of portal tracts observed in the series as a whole (a minimum of five portal tracts was among the inclusion criteria), and it provided no information on observer variations.

In the study by Regev et al,[66] 124 patients with chronic HCV hepatitis underwent liver needle biopsy from the right and left lobes during laparoscopy. A sample was considered adequate if it was “at least 1.5 cm long and with 5 or more portal zones.” Grading and staging were done according to a modified Scheuer score system. The comparison between right and left lobes showed a 2-point difference in grade in 1 case (1.6%), and 30 patients (24.2%) had a difference of 1 point. As for the stage, differences in fibrosis score were observed in 33% of cases (with a 1-point difference in 30.6%). The authors emphasized that the difference in stage amounted to 9.7% when stages 3 and 4 were considered as one stage (because they have similar clinical implications). Interobserver agreement proved to be high. It would have been interesting, however, to know the specimens' length and also the number of portal tracts in the pairs of biopsies yielding different scores.

Different conclusions were reached by Persico et al,[67] who evaluated 52 consecutive chronic hepatitis C patients by means of a double, right, and left lobe percutaneous liver biopsy obtained with a modified Menghini 18-G needle. Two independent pathologists whose general interobserver agreement was 90% performed grading and staging based on Ishak's scoring system. No difference was observed between the two lobes in either grade or stage. The mean length of specimens was 2.8 + 1.1 cm in the right lobe and 2.5 + 0.9 cm in the left.

In 2003, Siddique et al[68] reported a surprisingly high variability in the samples: 69 and 62% for activity and fibrosis, respectively. This study analyzed 29 paired biopsies using the Knodell histological activity index: 20 of 29 paired biopsies showed a difference in necroinflammatory lesions ≥2, and 10 revealed a difference ≥4; the difference in fibrosis score was ≥1 in 11 cases and ≥2 in 6. All samples contained four to five portal tracts. The differences in the histological scores may stem from the discontinuous numerical scale of the Knodell scoring system, which tends to highlight any scoring inconsistencies.

Top of Page

WOULD THE SAMPLE SIZE AFFECT THE HISTOLOGICAL ASSESSMENT OF CHRONIC HEPATITIS IN TERMS OF GRADE AND STAGE?
Before evidence-based medicine and before grading and staging, Sherlock postulated that “a satisfactory biopsy is 1-4 cm long and weighs 10-50 mg.”[3] This was a general principle applicable to all kinds of liver disease, but Sherlock noted that “in macronodular cirrhosis it is possible to aspirate a large nodule and find normal architecture” and that “there is sampling variability in the diagnosis of cirrhosis and chronic active hepatitis.”[3] Authoritative literature[13] suggests that a biopsy 1.5 cm long suffices for the histological assessment of chronic hepatitis and that pathologists are “satisfied” with samples containing four to six portal tracts (we might add that it is a matter of satisfying not the pathologists' needs but the patients' needs), and these standards have been extensively adopted in clinical studies.

Top of Page

The Historical Perspective
Two studies performed by the same group investigated the relationship between sample variability and sample size.

The first evaluated 100 liver needle biopsies from patients with a clinically established diagnosis of acute or chronic hepatitis and cirrhosis of various etiology.[69] Biopsies longer than 2.5 cm were selected, but no attention was paid to their diameter. All samples were reviewed blindly at five different sessions, changing the length of the sample with the aid of opaque paper tape. The study demonstrated that, although acute viral hepatitis was reliably diagnosed, even in samples 5 mm long, only 30 and 40% of cases of CAH and cirrhosis, respectively, were diagnosed in 5-mm samples. Predictive positive and negative values increased significantly with increasing sample lengths. The study concluded that “a considerable number of biopsies less than 25 mm long are not diagnostic” for cirrhosis.

The second study[70] mainly addressed the accuracy of a diagnosis of CAH using the same method as before, except for the needle, which was a 16-G Menghini needle in all cases. The gold standard was, here again, the original liver sample, which had to be no less than 25 mm long. The study showed that the specificity of the diagnosis of CAH with or without cirrhosis rose significantly when the length of the biopsied tissue changed from 5 to 15 mm. Moreover, 65% of all diagnostic errors were made on biopsies up to 10 mm long, and the error was always on the side of less severe conditions. The study concluded that a specimen at least 15 mm long is needed for an acceptable accuracy in the diagnosis of CAH, but larger needle biopsy samples are warranted when cirrhosis is suspected.

Expanding on earlier data,[71] a study by Colombo et al[72] demonstrated the superiority of the Tru-Cut (89.5% accuracy) over the Menghini needle (65.5% accuracy) in diagnosing cirrhosis. Differences were attributed to the greater fragmentation of liver samples obtained with the Menghini needle. Although this was not specifically highlighted in the study, it emerged that diagnostic accuracy depended on the number of complete portal tracts within the biopsy samples and that 34.8% of the portal tracts in the Menghini biopsy specimens were incomplete, compared with 19.6% in the Tru-Cut biopsy specimens.

A weakness of the previous studies lay in the lack of intraobserver and interobserver agreement tests, which was partially overcome by the experience of the pathologists who performed the histological examination. In fact, diagnostic consistency improves when the assessment is performed by experienced pathologists specializing in liver diseases.

天狼星

The Present Perspective
A more recent approach to the problem of specimen size has been to compare “large” versus “thin” liver biopsies in the diagnosis of diffuse diseases. The rationale behind these studies is that taking biopsies with larger needles can cause more complications. Fine-needle biopsies are easier to perform in outpatients under local anesthesia, and patients are more compliant. Using a finer needle is also an advantage in clinical conditions in which large-needle biopsy is contraindicated.

Rocken et al[73] compared thin and large biopsy samples and concluded that the amount of tissue available correlated strongly with the type of needle, although thin biopsies (20G) yield sufficient material to make a diagnosis of diffuse liver disease of various etiologies. The study, however, did not address the grading and staging of chronic hepatitis.

This was the aim of a more recently published study by the same group,[74] in which large (17G Menghini needle) and thin (20G modified cutting Menghini needle) biopsies from 88 patients with chronic type B or C hepatitis were retrospectively re-evaluated for grading and staging purposes. The study concluded that a thin biopsy is as good as a large one for the grading and staging of chronic viral hepatitis. It is important to note, however, that (1) the thin and large samples came from different patients; (2) there was no clearly defined gold standard for assessing diagnostic accuracy; and (3) because the study was retrospective, patients were not randomized to undergo large or thin biopsy, and patients with thin biopsies were more often found to have low platelet counts and high alanine aminotransferase levels, which introduce a selection bias. It is also worth noting that, in the small subgroup of patients whose clinical diagnosis of cirrhosis was taken as the gold standard, cirrhosis was histologically diagnosed by thin samples in 44% of cases and by large biopsies in 60%, which means that biopsy size becomes relevant when a gold standard is established.

More recently, a study was designed to evaluate the effect of core length and diameter on the grading and staging of chronic types B and C hepatitis.[75] The method was similar to the one used in previous studies[69][70] and consisted of progressively reducing the length and width of the original samples, which were all at least 30 mm long and 1.4 mm wide. The same pathologist reviewed all the slides throughout the various sessions and intraobserver agreement proved to be high. This study provided evidence that both the length and the diameter of the biopsy core affect grading and staging and (as in the previous studies[69][70]) that examining shorter and thinner samples leads to an underestimation of the severity of disease. Disease activity and fibrosis were underestimated in thin biopsies (i.e., 1 mm wide) regardless of the length of the biopsy, suggesting that the main problem lies in the lower number of complete portal tracts in the smaller samples. Portal areas are the elective sites of damage in chronic hepatitis, and they must be intact for disease grading and staging. The study demonstrated that 11 to 15 complete portal tracts was the critical number below which disease grade and stage were significantly underestimated and that a liver biopsy 2 cm long and 1.4 mm wide guaranteed this number of portal tracts in 94% of cases. The study also suggested that the number of complete portal tracts should be mentioned in the histological report to focus the clinicians' attention on the adequacy of their samples and on the limits of histological interpretation.

Top of Page

CONCLUSIONS
Clinicians having to decide whether to perform a liver biopsy will gain little from reading this article, which is concerned exclusively with what happens after such a decision has been made.

For this article, we tried to answer two questions: (1) Can a random sample reflect damage to the organ as a whole? Sherlock had the answer[3]: “It is surprising that such a small biopsy should so often be representative of changes in the whole liver.” Available data confirm that, in diffuse processes such as acute and chronic hepatitis, a needle biopsy is representative of the liver disease, ensuring an accurate diagnostic classification.[56][69][70] In practice, needle liver biopsy is still indicated to diagnose chronic viral hepatitis when clinical or serological data, or both, produce inconsistent results or when confirmation is required for clinical purposes (e.g., in the liver transplantation setting) or research.

(2) Can the sample size affect the histological assessment of chronic hepatitis in terms of grade and stage? Our earlier analyses showed that most of the available data predated the recent grading and staging system of chronic viral hepatitis. Almost all studies agreed that the more tissue there is, the greater is the diagnostic accuracy-a fact that has to be balanced against the patient's safety (primum non nocere). The traditional assumption that a sample 1.5 mm long or containing four to six portal tracts, or both, is adequate is no longer true for the grading and staging of chronic hepatitis. In fact, none of the studies using this standard length produced consistent results[57][58][64][66][67][68][69][75] and when different-sized samples were compared, the smaller specimens resulted in a diagnosis of less severe disease.[69][70][75] Samples at least 2 cm long can ensure greater diagnostic accuracy.[63][67][69][70] Below this length, grade and stage are significantly underestimated.[75] As for the portal tracts, their number clearly correlates with sample size,[73][74][75] and there is evidence that four to five portal tracts are not enough, not for grading and staging at least; with fewer than 11 to 15 portal tracts, grade and stage are significantly underestimated.[75] The study by Colloredo et al[75] introduced the concept of a “minimum number of complete portal tracts.” The lower number of complete portal tracts may explain the lower diagnostic accuracy obtained with smaller samples.[68][71][73]

Although technological advances are making medical practice progressively less bloody and painful (e.g., with the wider diffusion of mini-invasive surgery and refinement of endoscopic techniques), many medical procedures in daily practice nevertheless remain invasive. Invasiveness is a feature of both therapeutic and diagnostic procedures, including morphological analyses, and liver biopsies are no exception. Their potential harmfulness is often stressed in the current lively debate on whether it is appropriate to take liver biopsies in chronic viral hepatitis. “First, do no harm.”[76] No one would disagree with this ethic, but this is not necessarily the same as saying “Do no biopsy.” Harm can also be done by failing to perform a biopsy[77] or by basing action on the strength of an inadequate sample.

In the current scenario of chronic viral hepatitis, the main goal of liver biopsy is to grade and stage the liver damage for prognosis and treatment. An adequate (although probably still imperfect) sample, pending the availability of reliable noninvasive methods, needs to be at least 2 cm long and to contain no fewer than 11 complete portal tracts.

Procedures for obtaining such an “adequate” biopsy sample are currently available and in routine use at most liver units.

While this article was in press, two additional articles have been published dealing with the same topic. In the first one, Bedossa et al, (Hepatology 2003; 38:1446-57) addressed the problem of sampling variability by using a virtual, computer assisted, analysis. In the second, Brunetti et al (J Hepatol 2004; 40:501-6) evaluated the diagnostic yield of a series of paired, fine and coarse liver blopsies obtained in 149 consecutive patients with chronic hepatitis C. Both these studies further support our conclusions on the characteristics of liver sample to be considered adequate for assessing grading and staging in chronic hepatitis.

Top of Page

ABBREVIATIONS
CAH chronic active hepatitis

CT computed tomography

HBV hepatitis B virus

HCV hepatitis C virus

US ultrasound

Top of Page

REFERENCES
1 von Frerichs FT. Uber den Diabetes. Hirschwald Berlin: 1884

2 Menghini G. One-second needle biopsy of the liver. Gastroenterology 1958; 35: 190-195

3 Sherlock S, Dooley J. Diseases of the Liver and Biliary System, 4th ed. Blackwell Scientific London: 1997

4 Van Leeuwen DJ, Wilson L, Crowe DR. Liver biopsy in the mid-1990s: questions and answers. Semin Liver Dis 1995; 15: 340-358

5 Lebrec D. Various approaches to obtaining liver tissue: choosing the biopsy technique. J Hepatol 1996; 25 (Suppl 1): 20-24

6 Scheuer PJ. Classification of chronic viral hepatitis: a need for a reassessment. J Hepatol 1991; 13: 372-374

7 Schmid M, Flury R, Buhler H. Chronic viral hepatitis B and C: an argument against the conventional classification of chronic hepatitis. Virchows Arch 1994; 425: 221-226

8 Ludwig J. The nomenclature of chronic active hepatitis: an obituary. Gastroenterology 1993; 105: 274-278

9 Batts KB, Ludwig J. Chronic hepatitis. An update on terminology and reporting. Am J Surg Pathol 1995; 19: 1409-1417

10 Gerber MA. Chronic hepatitis C: the beginning of the end of a time honored nomenclature?. Hepatology 1992; 15: 733-734

11 Desmet VJ, Gerber M, Hoofnagle JH. Classification of chronic hepatitis: diagnosis, grading and staging. Hepatology 1994; 19: 1513-1520

12 Brunt EM. Grading and staging the histopathological lesions of chronic hepatitis: the Knodell histology activity index and beyond. Hepatology 2000; 31: 241-246

13 Bravo AA, Sheth SG, Chopra S. Liver biopsy. N Engl J Med 2001; 344: 495-500

14 Saadeh S, Cammell G, Carey WD. The role of liver biopsy in chronic hepatitis C. Hepatology 2001; 33: 196-200

15 Lok AS. Hepatitis B infection: Pathogenesis and management. J Hepatol 2000; 32 (Suppl 1): 89-97

16 Booth JC, O'Grady J, Neuberger J. Clinical guidelines on the management of hepatitis C. Gut 2001; 49 (Suppl 1): 1-21

17 Dienstang JL. The role of liver biopsy in chronic hepatitis C. Hepatology 2002; 36: S152-S160

18 Knodell RG, Ishak KG, Black WC. Formulation and application of a numerical scoring system for assessing histological activity in asymptomatic chronic active hepatitis. Hepatology 1981; 1: 431-435

19 The French METAVIR Cooperative Study Group. Intraobserver and interobserver variations in liver biopsy interpretation in patients with chronic hepatitis C. Hepatology 1994; 20: 15-20

20 Ishak K, Baptista A, Bianchi L. Histological grading and staging of chronic hepatitis. J Hepatol 1995; 22: 696-699

21 Guido M, Faa G. Morphological diagnosis of chronic hepatitis. Minimal requirements for histopathological referral [in Italian]. Pathologica 1998; 90: 176-182

22 Goldin RD, Goldin GJ, Burt AD. Intra-observer and inter-observer variation in the histopathological assessment of chronic viral hepatitis. J Hepatol 1996; 25: 649-654

23 Westin J, Lagging LM, Weistal R. Interobserver study of liver histopathology using the Ishak score in patients with chronic hepatitis C virus infection. Liver 1999; 19: 183-187

24 Gronbaek K, Christensen PB, Hamilton-Dutoit S. Interobserver variation in interpretation of serial liver biopsies from patients with chronic hepatits C. J Viral Hepat 2002; 9: 443-449

25 Yano M, Kumada H, Kage M. The long-term pathological evolution of chronic hepatitis C. Hepatology 1996; 23: 1334-1340

26 Gebo KA, Herlong HF, Torbenson MS. Role of liver biopsy in management of chronic hepatitis C: a systematic review. Hepatology 2002; 36: S161-S172

27 Zeuzem S, Feinman SV, Rasenack J. Peg-interferon alpha-2a in patients with chronic hepatits C. N Engl J Med 2000; 343: 1666-1672

28 Manns MP, Mc Hutchinson JG, Gordon SC. Peg-interferon alpha 2b plus ribavirin compared with interferon alpha 2b plus ribavirin for initial treatment of chronic hepatitis C: a randomized trial. Lancet 2001; 358: 958-965

29 Saracco G, Ciancio A, Olivero A. A randomized 4-arm multicenter study of interferon alpha -2b plus ribavirin in the treatment of patients with chronic hepatits C not responding to interferon alone. Hepatology 2001; 34: 133-138

30 Shiratori I, Imazeki F, Moriyama M. Histologic improvement of fibrosis in patients with hepatits C who have sustained response to interferon therapy. Ann Intern Med 2000; 132: 517-524

31 Oberti F, Valsesia E, Pilette C. Non-invasive diagnosis of hepatic fibrosis and cirrhosis. Gastroenterology 1997; 113: 1609-1616

32 Imbert-Bismut F, Ratziu V, Pieroni L. Biochemical markers of liver fibrosis in patients with hepatits C virus infection: a prospective study. Lancet 2001; 357: 1069-1075

33 Fontana RJ, Lok ASF. Noninvasive monitoring of patients with chronic hepatitis C. Hepatology 2002; 36: S57-S64

34 Forns X, Ampurdanes S, Llovet JM. Identification of chronic hepatitis C patients without hepatic fibrosis by a simple predictive model. Hepatology 2002; 36: 986-992

35 Oh S, Afdhal NH. Hepatic fibrosis: are any of the serum markers really useful?. Curr Gastroenterol Rep 2001; 3: 12-18

36 Friedman SL. Liver fibrosis-from bench to bedside. J Hepatol 2003; 38: S38-S53

37 Afdhal NH. Diagnosing fibrosis in hepatitis C: is the pendulum swinging from biopsy to blood tests?. Hepatology 2003; 37: 972-974

38 Geller SA, Pitman MB. In: McSween RNM, Burt AD, Portmann BC, et al., Eds.; Morphological diagnostic procedures (liver biopsy)Pathology of the Liver. Churchill Livingston London: 2002, p. 943-960

39 Farrel RJ, Smiddy PF, Pilkington RM. Guided versus blind liver biopsy for chronic hepatis C: clinical benefits and costs. J Hepatol 1999; 30: 580-587

40 Grant A, Neuberger J. Guidelines of the use of liver biopsy in clinical practice. Gut 1999; 45 (Suppl 1): 1-11

41 Ahmad M, Riley TR. Can one predict when ultrasound will be useful with percutaneous liver biopsy?. Am J Gastroenterol 2001; 96: 547-549

42 Lindor KD, Bru C, Jorgensen RA. The role of ultrasonography and automatic-needle biopsy in outpatients percutaneous liver biopsy. Hepatology 1996; 23: 1079-1083

43 Farrel RJ, Smiddy PF, Pilkington RM. Guided versus blind liver biopsy for chronic hepatis C: clinical benefits and costs. J Hepatol 1999; 30: 580-587

44 Piccinino F, Sagnelli E, Pasquale G, Giusti G. Complications following percutaneous liver biopsy. A multicentre retrospective study on 68,276 biopsies. J Hepatol 1986; 2: 165-173

45 Demetris AJ, Ruppert K. Pathologist's perspective on liver needle biopsy size?. J Hepatol 2003; 39: 275-277

46 McGill DB, Rakela J, Zinsmeister AR, Ott BJ. A 21-year experience with major hemorrhage after percutaneous liver biopsy. Gastroenterology 1990; 99: 1396-1400

47 Tobkes AI, Nord HJ. Liver biopsy: review of methodology and complications. Dig Dis 1995; 13: 267-274

48 Terjung B, Lemnitzer I, Ludwig-Dumoulin F. Bleeding complications after percutaneous liver biopsy. An analysis of risk factors. Digestion 2003; 67: 138-145

49 Froehlich F, Lamy O, Fried M, Gonvers JJ. Practice and complication of liver biopsy: result of a national survey in Switzerland. Dig Dis Sci 1993; 38: 1480-1484

50 McGill DB, Rakela J, Zinsmeister AR, Ott BJ. A 21-year experience with major haemorrhage after percutaneous liver biopsy. Gastroenterology 1990; 99: 1396-1400

51 Jacobs WH, Goldberg SB. Statement on outpatient percutaneous liver biopsy. Dig Dis Sci 1989; 34: 322-323

52 Nousbaum JB, Cadranel JF, Bonnemaison G. Clinical practice guidelines on the use of liver biopsy. Gastroenterol Clin Biol 2002; 26: 848-878

53 Buckley A, Petrunia D. Practice guidelines for liver biopsy. Canadian Association of Gastroenterology. Can J Gastroenterol 2000; 14: 481-482

54 Shalm SW. The diagnosis of cirrhosis: clinical relevance and methodology. J Hepatol 1997; 27: 1118-1119

55 Waldstein S, Szanto PB. Accuracy of sampling by needle biopsy in diffuse liver disease. Arch Pathol 1950; 50: 326-329

天狼星

56 Soloway RD, Baggenstoss AH, Schoenfield LJ, Summerskill WHJ. Observer error and sampling variability tested in evaluation of hepatitis and cirrhosis by liver biopsy. Dig Dis Sci 1971; 16: 1082-1086

57 Abdi W, Milln JC, Mezey E. Sampling variability on percutaneous liver biopsy. Arch Intern Med 1979; 139: 667-669

58 Baunsgaard P, Sanchez GC, Lundborg CJ. The variation of pathological changes in the liver evaluated by double biopsies. Acta Pathol Microbiol Scand [A] 1979; 87: 51-57

59 Maharaj B, Maharaj RJ, Leary WP. Sampling variability and its influence on the diagnostic yield of percutaneous needle biopsy of the liver. Lancet 1986; 1: 523-525

60 Pagliaro L, Rinaldi F, Craxi A. Percutaneous blind biopsy versus laparoscopy with guided biopsy in diagnosis of cirrhosis. A prospective, randomized trial. Dig Dis Sci 1983; 28: 39-43

61 Poniachik J, Bernstein DE, Reddy KR. The role of laparoscopy in the diagnosis of cirrhosis. Gastrointest Endosc 1996; 43: 568-571

62 Scheuer PJ, Lefkowitch JH. Liver Biopsy Interpretation 6th ed. WB Saunders Philadelphia: 2000

63 Picciotto A, Ciravegna G, Lapertosa G, Celle G. Percutaneous or laparoscopic needle biopsy in the evaluation of chronic liver disease?. Am J Gastroenterol 1984; 79: 567-568

64 Gaiani S, Gramantieri L, Venturoli N. What is the criterion for differentiating chronic hepatitis from compensated cirrhosis? A prospective study comparing ultrasonography and percutaneous liver biopsy. J Hepatol 1997; 27: 979-985

65 Fanning L, Loane J, Kenny-Walsh E. Tissue viral load variability in chronic hepatitis C. Am J Gastroenterol 2001; 96: 3384-3389

66 Regev A, Berho M, Jeffers LJ. Sampling error and intraobserver variation in liver biopsy in patients with chronic HCV infection. Am J Gastroenterol 2002; 97: 2614-2618

67 Persico M, Palmentieri B, Vecchione R. Diagnosis of chronic liver disease: reproducibility and validation of liver biopsy. Am J Gastroenterol 2002; 97: 491-492

68 Siddique I, El-Naga HA, Madda JP, Memon A, Hasan F. Sampling variability on percutaneous liver biopsy in patients with chronic hepatitis C virus infection. Scand J Gastroenterol 2003; 38: 427-432

69 Holund B, Poulsen H, Schlichting P. Reproducibility of liver biopsy diagnosis in relation to the size of the specimen. Scand J Gastroenterol 1980; 15: 329-335

70 Schlichting P, Holund B, Poulsen H. Liver biopsy in chronic aggressive hepatitis. Diagnostic reproducibility in relation to size of specimen. Scand J Gastroenterol 1983; 18: 27-32

71 Vargas-Tank L, Martinez V, Jiron MI. Tru-Cut and Menghini needles: different yield in histological diagnosis of liver diseases. Liver 1985; 5: 178-181

72 Colombo M, Del Ninno E, De Franchis R. Ultrasound-assisted percutaneous liver biopsy: superiority of the Tru-cut over the Menghini needle for diagnosis of cirrhosis. Gastroenterology 1988; 95: 487-489

73 Rocken C, Meier H, Klauck S. Large-needle biopsy versus thin-needle biopsy in diagnostic pathology of liver diseases. Liver 2001; 21: 391-397

74 Petz D, Klauck S, Rohl FW. Feasibility of histological grading and staging of chronic viral hepatitis using specimens obtained by thin-needle biopsy. Virchows Arch 2003; 442: 238-244

75 Colloredo G, Guido M, Sonzogni A, Leandro G. Impact of liver biopsy size on histological evaluation of chronic viral hepatitis: the smaller the sample, the milder the disease. J Hepatol 2003; 39: 239-244

76 Andriulli A, Annese V, Facciorusso D, Giacobbe A. First do no harm: power, oppression, and violence of liver biopsy. Gastroenterology 2003; 125: 272-273

77 Green RM, Flamm SL. First do no harm: power, oppression, and violence of liver biopsy. Gastroenterology 2003; 125: 273-274