Transjugular liver biopsy in the 1990s: a 2-year audit

Authors


Patch Dr Liver Transplantation and Hepatobiliary Medicine, Royal Free Hospital, Pond Street, London NW3 2QG, UK.

Abstract

Background

: In view of the changing nature of transjugular liver biopsy, we performed an audit of the safety, adequacy and clinical impact of such biopsies in our centre over a 2-year period from 1995 to 1997.

Methods

: One hundred and fifty-seven transjugular biopsies were carried out in 145 patients, with prothrombin time >5 s over control and/or platelet count <50 × 109/L and/or gross ascites.

Results

: Major complications were two (1.3%) capsular perforations, which were easily plugged with coils without sequelae. Biopsy sample was adequate for histological diagnosis in 90%, inadequate in 6% and technically unsuccessful in 4% of cases. Mean biopsy size was 14.8 ± 7.7 (1–51) mm. Adequacy did not differ between cases with and without cirrhosis. Transjugular biopsy had a clinical impact (specific diagnosis or influence on patient’s management) in 50% of acute liver disease, 62% of chronic liver disease and 87% of transplant patients (< 0.001). In chronic liver disease, it had a significantly greater clinical impact in cases trying to establish the stage rather than diagnosis (84% vs. 35%, < 0.001).

Conclusions

: Transjugular liver biopsy is a safe procedure for high-risk patients providing an adequate liver sample even in cirrhosis. It has a clinical impact in more than 80% of transplant patients and for staging chronic liver disease, but in only 50% (acute) or 35% (chronic) of liver disease when a diagnosis is sought.

INTRODUCTION

Histology is the diagnostic gold standard for most liver diseases. Routine percutaneous liver biopsy is the procedure of choice in most cases,1 but there is morbidity (2%) and mortality (0.13–0.33%).2 The procedure is contraindicated in patients with increased risk of haemorrhage, such as those with marked coagulopathy, thrombocytopaenia or ascites. The alternative non-surgical techniques are plugged percutaneous liver biopsy3, 4 and transjugular liver biopsy (TJLB)5[6][7]–8 which allow histological examination of liver tissue with relative safety in almost any patient. We have previously shown that plugged percutaneous liver biopsy and TJLB have a similarly high success rate, but the former may be associated with an increased risk of haemorrhage.9 We therefore use plugged percutaneous biopsy only in cases with moderately impaired haemostasis (prothrombin time between 3 and 5 s over control value and/or platelet count between 50 and 80 × 109/L) and TJLB in cases with worse parameters.

The transjugular route of liver biopsy was first described in experimental dog studies by Dotter10 and successfully performed in patients by Weiner & Hanafee.11 Subsequently, several studies have confirmed its safety and efficacy.5[6][7]–8, 12[13][14][15][16]–17 TJLB has the advantage that bleeding from the liver is back into the venous circulation and is therefore without haemodynamic compromise. The procedure is not only used when impaired haemostasis is present but also when there is gross ascites or measurement of hepatic venous pressures is needed.8 In addition, we and others use carbon dioxide for retrograde visualization of the portal vein via the hepatic vein.18, 19 However, the diagnostic adequacy of specimens from TJLB has been a concern, particularly in cirrhosis, because of relatively small liver specimens compared to those obtained by the percutaneous route,5, 14, 20 while the clinical impact of performing the biopsy has been questioned in some cases. Moreover, TJLB has been considered to be more time-consuming and has been associated, albeit rarely, with serious or even fatal complications.5[6][7]–8, 21 On the other hand, technical experience is improving, new needles have been developed which may influence the technical success rate, while new clinical settings needing the transjugular approach have appeared. Liver transplantation has increased the demand for TJLB because histology is frequently required in these patients who initially have coagulation disorders.22, 23

In view of the changing nature of TJLB, we performed an audit of the safety, adequacy and clinical impact of the TJLBs in our centre over a 2-year period.

MATERIALS AND METHODS

Between June 1995 and May 1997, 157 TJLBs were attempted in 145 patients who had one or more of the following criteria: prothrombin time greater than 5 s over control value and/or platelet count less than 50 × 109/L and/or gross ascites (Table 1).

Table 1.  .  Characteristics of patients undergoing transjugular liver biopsy Thumbnail image of

TJLBs were performed in patients with acute liver disease in 22 (15%), chronic liver disease in 66 (41%) and post-transplant disorder in 69 (44%) of the cases. TJLBs were performed to establish a diagnosis in 120 (76.4%) and to stage the liver disease in 37 (23.6%). In particular, diagnostic indication was present in all cases with acute liver disease or post-transplant disorder and in 29 (44%) of the 66 patients with chronic liver disease. Staging of disease (presence of cirrhosis, and/or acute alcoholic hepatitis in patients with alcoholic liver disease prior to the use of steroids) was the indication in the remainder (37/66 or 56%).

The length of tissue specimen was taken as its size for this audit. If there was more than one tissue sample the cumulative length was measured. TJLBs were defined as technically unsuccessful if liver tissue was not obtained for any reason. Biopsies were defined as inadequate if no portal tracts were identified or the size itself made it impossible to make a diagnosis. Histology reports were classified into three main categories: (i) specific diagnosis (ii) non-specific diagnosis (despite an adequate liver specimen), and (iii) inadequate liver specimen. The clinical notes of all cases with a non-specific diagnosis were reviewed and then classified as (a) influencing the patient’s management (mainly because of disease exclusion) and (b) not influencing the patient’s management. Biopsies were considered as having a clinical impact when there was a specific diagnosis or when a non-specific diagnosis was associated with an influence on a patient’s management.

Patients without liver transplants were also classified into those with or without cirrhosis according to histological findings. If there was an inadequate specimen, the classification was made according to clinical signs, and laboratory and radiological findings.

TJLB technique

A 7F or 10F sheath introducer (Arrow International, Inc.) was introduced into the right internal jugular vein by a Seldinger technique under local anaesthesia and often intravenous diazepam sedation. A 7F Cobra catheter (Cordis Europa N.V.) was introduced into the hepatic vein and the sheath advanced over it. The catheter was removed and the biopsy needle was inserted via the sheath into the hepatic vein. A standard Cooks type transjugular biopsy needle or a Quick-core Tru-cut type biopsy needle was always used (Figure 1). The standard Cooks needle was used during the first year (in 81 cases) and the Quick-core needle during the second year of the audit (in 76 cases) following a change in sterilization policy in our hospital. The Quick-Core needle was also tried in five cases in which the Cooks needle failed to obtain an adequate specimen.

Figure 1.

.  Transjugular liver biopsy needles. A standard Menghini type Cooks needle on the right and a Quick-core Tru-cut type needle (William Cook Europe A/S) on the left.

The two needles differ in technique. The standard Cooks transjugular needle is a hollow tube with a sharp cutting edge. Specimens are obtained via the Menghini (aspiration) technique. It has the disadvantage that cirrhotic livers may ‘bounce’ away from the needle, and a core may not be obtained. In addition, the depth of puncture is difficult to control, increasing the risk of inadvertent capsular puncture. Nonetheless, good length specimens may be obtained and the needle is re-usable. The Quick-Core needle is a thin spring-loaded Tru-cut type device. Because of its sharpness, specimens can always be obtained from cirrhotic livers, and in addition it does not require a hard push in order to penetrate the liver parenchyma. Multiple specimens can be obtained (e.g. for tissue culture, base metal analysis). Disadvantages are that it may be difficult to enter hepatic veins that come off the inferior vena cava at right angles, is not re-usable and is more expensive.

Following the biopsy, size was immediately assessed and another needle pass attempted when the first specimen was considered inadequate. A maximum of four passes was allowed. After each pass, contrast medium was injected in order to detect possible capsular perforation and in such cases embolization coils (William Cook Europe A/S) were injected via the transjugular catheter until the leak was sealed.

Statistical analysis

Statistical analysis was carried out by corrected chi-squared test, two-tailed Fisher’s exact test and t-test, when appropriate.

RESULTS

Safety

No death occurred attributable to TJLB. Major complications were capsular perforations in two (1.3%) cases which were diagnosed during the procedure and managed as described above. The standard Cooks needle was used in one and the Quick-Core needle in the other case with capsular perforation. No clinical sequelae were noted and no blood transfusion for haemorrhage required.

Mean procedure time was 40.6 ± 18.4 (7–120) min and mean screening time 14.8 ± 7.7 (1–37) min.

Adequacy

The biopsy specimen was adequate for histological diagnosis in 142 (90.4%) and inadequate in 9 (5.7%) of the 157 cases; no portal tract was identified in four and portal tracts were less than 5 in five of these nine cases. TJLB was technically unsuccessful in the remaining 6 (3.8%) cases, due to failure to puncture jugular vein (n = 4; despite ultrasound guidance), blocked superior vena cava (n = 1) and no liver tissue obtained (n = 1).

Mean biopsy size in technically successful biopsies was 14.8 ± 7.7 (1–51) mm with a mean number of passes of 1.8 ± 1.0 (1–5). Liver specimen size was significantly smaller in cirrhotics (12.8 ± 5.9 mm) compared to non-cirrhotics or transplant patients (15.9 ± 8.3 mm) (= 0.02). However, diagnostic adequacy (specimen adequate for histological diagnosis) did not significantly differ between cases with and without cirrhosis (47/54 or 87% vs. 95/103 or 92%, = 0.44).

There was no significant difference in the technical success or specimen adequacy rate between the standard Cooks and the Quick-core transjugular biopsy needle (data not shown). However, we obtained an adequate liver specimen using the Quick-core needle in five cases in which the standard Cooks needle had previously failed.

Clinical impact

The histological diagnoses for each type of liver disease are presented in Table 2.

Table 2.  .  Histological diagnoses of 157 transjugular liver biopsies in relation to the type of liver disease Thumbnail image of

TJLB had a clinical impact in 112 (71.3%) of the 157 cases (specific diagnosis: 47.8% and non-specific diagnosis but with an influence on patient’s management: 23.5%). A clinical impact was seen in 87% of transplant patients compared to 62% of those with chronic liver disease and only 50% of those with acute liver disease (= 0.001 for both comparisons). Specific diagnosis was obtained more frequently in patients with chronic liver disease (51.5%) or post-transplant disorders (52.2%) compared to those with acute liver disease (22.7%) (= 0.025 for both comparisons) (Table 3).

Table 3.  .  Clinical impact of transjugular liver biopsy (TJLB) in relation to type of liver disease Thumbnail image of

In patients with chronic liver disease, TJLB had a significantly greater clinical impact in cases trying to establish the stage (83.8%) rather than the diagnosis of the disease (34.5%) (< 0.001) (Table 4).

Table 4.  .  Clinical impact of transjugular liver biopsy (TJLB) in patients with chronic liver disease in relation to whether diagnosis or stage of disease is sought Thumbnail image of

DISCUSSION

No deaths occurred in our study population which included only patients with severely impaired haemostasis and/or gross ascites. We have had no death related to the procedure amongst over 400 TJLBs performed in our centre within the last 6 years. The safety of transjugular liver biopsy has been well established.5[6][7][8]–9 Mortality ranges from 0% to 0.5% which is notable, considering that the procedure is performed mostly in high-risk patients.5[6][7]–8 The most common major complication is capsular perforation with bleeding5[6][7]–8 which can be detected and treated during the procedure by injecting gelatin material or coils at the site of the leak.7, 9 We had two (1.3%) capsular perforations which were easily managed with coils without any sequelae.

Although a transjugular approach is always performed by trained personnel, its success rate in the literature ranges from 64% to 97%.5[6][7]–8, 13[14]–15 In our study, the technical success rate was more than 95% and the specimen obtained was adequate for histological diagnosis in 90%. Technical improvements of the biopsy needles may be responsible for the very high success rates in this series. Our study was not designed to compare the two different types of needles used during the study period. The overall success rate did not differ between the two needles. However, we managed to obtain adequate biopsy specimen using the Quick-Core needle in five cases when the standard Cooks biopsy needle had previously failed. Although the standard Cooks needle was used in both cases with complications, we generally found both needles to be quite safe. Subjectively, the Quick-core needle was technically easier to use.

Adequacy of the specimens has been a concern for TJLBs particularly in cirrhotic patients. The mean liver specimen length has been reported less than 1 cm in most series including a previous one from our centre.8, 9, 16, 17 The best results have been reported by Bull et al. with use of the Tru-Cut needle with a mean specimen length of 1.8 cm.6 In our study, the mean specimen length was about 1.5 cm and a specimen adequate for histological diagnosis was obtained in 90% of all cases and in 94% of the technically successful attempts. The specimen size was smaller in cirrhotics but adequate for diagnosis in 87% of them, and the mean biopsy length of 1.3 cm was longer than that reported by others irrespective of presence of cirrhosis.8, 9, 16, 17 The biopsy needles recently used in our centre and using up to four needle passes may explain the high rate of adequate specimens even in cirrhotics.

In the non-transplant setting, TJLB had a clinical impact in 41% of cases when the diagnosis of the (acute or chronic) liver disease was sought and in 84% of cases trying to establish the stage of chronic liver disease. In particular in the latter cases, absence of clinical impact was due to inadequate liver specimen in 11% and due to non-specific histological findings in 5% (two cases with known Budd–Chiari syndrome in which the histological examination of an adequate specimen could not confirm or exclude cirrhosis).

Liver transplantation has increased demand for TJLB, because histological examination of the liver is very often essential for differential diagnosis of post-transplant disorders in patients with severely impaired haemostasis.23 In our centre, which is a supra-regionally designated liver transplant centre, graft dysfunction was the indication for biopsy in 44% of cases. TJLB was extremely useful in the post-transplant period, because it had a clinical impact in 87% of cases, significantly more frequently than in any other setting. It gave a specific diagnosis in half of the post-transplant cases and it influenced the patient’s management in another third of cases mainly by excluding cellular rejection and subsequently avoiding the need for increasing immunosuppression.

In conclusion, transjugular liver biopsy in the 1990s is a safe procedure even in patients with very disturbed clotting, which gives adequate samples of liver tissue for histological diagnosis. Use of new improved biopsy needles and the feasibility of multiple passes has increased the specimen length; adequate tissue samples can be provided in about 90% of cases even in cirrhosis. It has a clinical impact in more than 80% of transplant patients and for staging liver disease, but in only 35–50% of liver disease when performed for a diagnosis. Transjugular liver biopsy is more time-consuming and expensive than the conventional or the plugged percutaneous liver biopsy [cost of consumable material: for transjugular biopsy using the standard Cooks needle (not resterilizing) £198 and the Quick-core needle £163, for plugged biopsy £60 and for conventional percutaneous liver biopsy £6–8]. However, such comparisons do not seem appropriate as transjugular and percutaneous approaches are used in different patient settings. The transjugular biopsy is part of the ‘one stop liver shop’ during which we can also perform hepatic venography, CO2 portography, wedged hepatic venous pressure and caval pressure measurements using the same minimal access.

Ancillary