The Case

A 42-year-old woman with a history of multiple malignancies, including osteosarcoma and recurrent breast cancer, presented to an emergency department with jaundice and epigastric pain. An abdominal ultrasound revealed several liver masses and subsequent imaging was consistent with metastatic malignancy of unknown primary source. The patient was admitted directly for an ultrasound-guided liver biopsy. Prior to the procedure, the patient required platelet transfusions, which increased her admitting platelet count of 5000/μL to 71,000/μL (reference range: 150,000–400,000/μL) prior to biopsy. Immediately after the procedure, she was transferred to the floor and began complaining about new abdominal pain. Her blood pressure was noted to be lower than baseline at 88/55 mm Hg, so a call was placed to the covering in-house physician. The physician believed that pain was common after such biopsies and ordered a dose of analgesics, which improved the patient's symptoms.

Over the next 2 hours, the patient's pain worsened and she became increasingly somnolent. When the bedside nurse returned to assess her pain, she found the patient unresponsive and called a code blue. The patient had pulseless electrical activity and the initial assessment also revealed a hematocrit of 14%, a decrease from 28% before the procedure. A massive transfusion protocol was initiated. The patient had a prolonged resuscitation and was transferred to the intensive care unit where she later died of multi-organ failure. The delay in recognizing the post-procedure intraperitoneal hemorrhage led to a detailed review by the hospital's quality committee. The protocols for managing patients following a liver biopsy were noted to be clearer in the outpatient setting where most procedures take place. No such protocols were in place for the less common inpatient liver biopsy at this facility.

The Commentary

Ensuring Safety With Liver Biopsy

The use of liver biopsy to obtain tumor or liver histology remains an important diagnostic tool in the management of patients with liver disease or suspected malignancy. The indications for liver biopsy are manifold and can be grouped into the following categories: (i) diagnosis, (ii) prognosis or staging, and (iii) management. As in this particular patient, liver biopsy is frequently used to clarify the diagnosis of intrahepatic mass lesions. While this indication was reasonable, the severe bleeding and rare outcome of death encountered in this patient emphasize several critical points, including a need for better understanding of available liver biopsy techniques, common complications, the limitations of biopsy, management of coagulopathies, and the recommended monitoring post-procedure to mitigate such outcomes.

While liver biopsy can provide critical diagnostic and prognostic information, and thus is often helpful to make management decisions, its risks and benefits must be carefully weighed each time it is considered. For example, biopsy for evaluation of focal liver disease detected by imaging is often difficult because biopsy of normal surrounding tissue rather than the area of interest is possible. Evaluation of focal liver disease is further complicated because liver lesions have many different radiographic appearances. Thus, the use of liver biopsy must always be individualized. For example, in this particular case, details are missing about the nature of the patient's jaundice and coagulopathy. Perhaps they weren't reflective of liver disease alone. Jaundice is rarely caused by tumor infiltration of the liver, except after advanced and extensive infiltration of the tumor. A platelet count of 5000/μL is also rarely a result of sequestration of platelets caused by splenomegaly due to cirrhosis and portal hypertension, but rather suggests a consumptive process such as disseminated intravascular coagulation. Putting aside the clinical elements that would warrant further investigation, the teaching point is that it's always critical to consider the yield and value of pursuing liver biopsy. In this case, additional studies may have suggested systemic disease beyond the liver that might have made the biopsy unnecessary in the first place.

There are certain situations in which liver biopsy is contraindicated, largely because in these instances, it is simply unsafe. Contraindications to liver biopsy are listed in Table 1. Perhaps the most important consideration has to do with the ability of the patient to cooperate. A patient who is not able to follow commands poses a risk to both the patient and the proceduralist.

An extremely controversial issue surrounds the use of liver biopsy in patients with coagulopathy and/or thrombocytopenia. Unfortunately, the evidence base is limited, and there are little good data to guide clinical management.(1,2) This is particularly true in patients with underlying liver disease, where the international normalized ratio (INR) does not correlate well with bleeding tendency.(3) In one study of patients with cirrhosis, a platelet count of less than 60,000/μL was associated with an increased risk of bleeding.(4) However, there are no currently established cutoffs for INR or platelet count, or its impact on the performance of liver biopsy. While some investigators have advocated for the use of viscoelastic studies to predict the risk of bleeding because they provide additional and often useful clinical information, further investigation is required before they can be widely recommended.(5)

For patients on pharmacologic anticoagulation, current expert opinion suggests that antiplatelet (e.g., aspirin, adenosine diphosphate receptor antagonists, IIb/IIIa receptor antagonists, nonsteroidal anti-inflammatory drugs) and/or anticoagulant drugs (e.g., warfarin) should be discontinued from several to 10 days prior to the procedure.(6) The new oral anticoagulants, including dabigatran, rivaroxaban, and apixaban, have a much shorter half-life than warfarin and can probably be continued until closer to the biopsy; however, clinical guidelines are lacking. Best practice now suggests that management of specific agents should be individualized. The risk of discontinuing antiplatelet or anticoagulant medications must be weighed against the potential risk of bleeding during or after liver biopsy; in each case, the pros and cons of medication discontinuation versus the need for assessment of liver histology should be considered carefully.

Once a decision to biopsy is made, several different techniques may be used to obtain liver tissue.(6) These include percutaneous transthoracic palpation/percussion-guided, transthoracic image-assisted, transthoracic real-time image-guided, subcostal real-time image-guided, and transvenous (typically transjugular) biopsy. Historically, liver biopsy was most frequently performed using the transthoracic palpation/percussion approach, often referred to as the "blind" biopsy because it's performed with the guidance of physical examination only. The blind technique is only used to obtain random liver samples and thus provides only parenchymal histology. While simple to perform, because of the emergence of widespread and easy-to-use imaging equipment, it has largely been replaced by one of the image-guided procedures, most often transthoracic image-assisted biopsy. With this technique, the physician evaluates the location of the liver in relation to typical landmarks (in the transthoracic approach, the ribs), marks the optimal site, and then proceeds with the biopsy. Variations of this technique include transthoracic real-time image-guided and subcostal real-time image-guided techniques. Additionally, the use of computed tomography as an image-guidance tool has gained popularity.

In practice today, biopsy of mass lesions is always image guided and is typically performed with real-time imaging, as in the patient in this case. The transjugular approach, which refers to inserting a catheter into a major vein in the neck to access the liver for tissue, is typically reserved for patients in whom there are concerns about bleeding or in whom the liver is not approachable with a percutaneous approach. The transjugular approach also provides the added advantage of allowing portal pressure measurements by assessing the hepatic venous pressure gradient. Preferences for the various techniques vary and are greatly influenced by local experience and expertise. Although evidence is limited, the safety of the different techniques appears to be similar.

There are many potential complications of liver biopsy (Table 2). The most common is pain. Pain may occur locally at the skin incision site, but the most common and concerning type of pain is that caused by puncture of the liver with subsequent bleeding and stretching of its capsule. It should be recognized that all patients have bleeding from the liver after biopsy. This is because the liver is highly vascular, and by definition, the liver must be punctured in the course of the biopsy. Bleeding usually stops within several minutes (7); however, a feared complication of liver biopsy is ongoing bleeding. Severe bleeding is uncommon, occurring in 1 to 2500 to 10,000 patients (6) undergoing biopsy to obtain random liver histology. Further, the most feared complication, death after liver biopsy, is usually related to aggressive hemorrhage. Death occurs in approximately 1 in 10,000 patients.(6) The most important clinical clue to aggressive bleeding is severe or ongoing pain. Thus, this patient's severe abdominal pain after the biopsy was a clue that she had significant bleeding.

In terms of the risk of complications, multiple factors play a role (Table 3). The risk of bleeding and death are likely higher in certain groups of patients, though available data is extremely heterogeneous and inexact. In this author's experience, complications are greater in hospitalized patients, many of whom have substantial underlying co-morbidities. It should also be pointed out that liver biopsy must only be performed in patients who are capable of being cooperative with the procedure. The risk of bleeding also varies with non–patient-related factors such as operator experience, the number of passes used to perform the biopsy, and a variety of technical issues. The latter include the type of needle used, the caliber of the needle, and whether image-assisted guidance was performed.

Finally, while liver biopsy can be performed in essentially any location, ranging from an outpatient surgery center to radiology or gastrointestinal procedural suite or even an inpatient bed, a core principle is that the patient must be carefully and rigorously monitored after the procedure. The risk of bleeding is greatest immediately after biopsy, which makes careful post-procedure observation essential. In the hour following the procedure, this patient was noted to have low blood pressure, which almost certainly was another sign of bleeding. The optimal observation time is currently controversial, but ranges between 1 to 3 hours. This stems in part from a large study of post-biopsy monitoring where the complication rates did not vary after recovery time was shortened from 6 hours to 1 hour.(8) The vast majority of complications were also detected within 1 hour. In keeping with this evidence, the American Association for the Study of Liver Disease (AASLD) guidelines state the following: "Vital signs must be frequently monitored (at least every 15 minutes for the first hour) after liver biopsy."(6) In this author's experience, implementation of standard written guidelines for monitoring after liver biopsy helps ensure that patients and their vital signs are closely monitored. This is true no matter where or how the biopsy is performed and helps to ensure capture of vital signs early enough to detect potential adverse patient outcomes and to mitigate the risk of complications.

Take-Home Points

• Indications for liver biopsy consist of (i) diagnosis, (ii) prognosis or staging, and (iii) management. It's always critical to consider the yield and value of pursuing a liver biopsy, which includes weighing the risks and benefits for each patient.
• A controversial issue in managing patients undergoing liver biopsy is how to address an underlying coagulopathy and/or thrombocytopenia, including for those on pharmacologic anticoagulation. The absence of best practices forces providers to individualize treatment plans.
• The most common complication after liver biopsy is pain. Severe abdominal pain should increase the concern for a significant bleeding complication. Careful monitoring is always required following liver biopsy, particularly in the first hour.

Don C. Rockey, MD Professor and Chairman
Department of Medicine
Medical University of South Carolina Charleston, SC

References

1. Caldwell SH, Hoffman M, Lisman T, et al; Coagulation in Liver Disease Group. Coagulation disorders and hemostasis in liver disease: pathophysiology and critical assessment of current management. Hepatology. 2006;44:1039-1046. [go to PubMed]

2. Tripodi A, Mannucci PM. Abnormalities of hemostasis in chronic liver disease: reappraisal of their clinical significance and need for clinical and laboratory research. J Hepatol. 2007;46:727-733. [go to PubMed]

3. Tripodi A, Caldwell SH, Hoffman M, Trotter JF, Sanyal AJ. Review article: the prothrombin time test as a measure of bleeding risk and prognosis in liver disease. Aliment Pharmacol Ther. 2007;26:141-148. [go to PubMed]

4. Seeff LB, Everson GT, Morgan TR, et al; HALT-C Trial Group. Complication rate of percutaneous liver biopsies among persons with advanced chronic liver disease in the HALT-C trial. Clin Gastroenterol Hepatol. 2010;8:877-883. [go to PubMed]

5. Mallett SV, Chowdary P, Burroughs AK. Clinical utility of viscoelastic tests of coagulation in patients with liver disease. Liver Int. 2013;33:961-974. [go to PubMed]

6. Rockey DC, Caldwell SH, Goodman ZD, Nelson RC, Smith AD; American Association for the Study of Liver Diseases. Liver biopsy. Hepatology. 2009;49:1017-1044. [go to PubMed]

7. Caldwell S, Northup PG. Bleeding complication with liver biopsy: is it predictable? Clin Gastroenterol Hepatol. 2010;8:826-829. [go to PubMed]

8. Firpi RJ, Soldevila-Pico C, Abdelmalek MF, Morelli G, Judah J, Nelson DR. Short recovery time after percutaneous liver biopsy: should we change our current practices? Clin Gastroenterol Hepatol. 2005;3:926-929. [go to PubMed]

Tables

Table 1. Contraindications to liver biopsy

Table 2. Complications of liver biopsy

Table 3. Factors important in determining the risk of complications of liver biopsy