Transjugular intrahepatic portosystemic shunt for refractory ascites in patients with high model for end‐stage liver disease scores

Watch a video presentation of this article

Watch the interview with the author

Abbreviations

HRS

hepatorenal syndrome

LVP

large‐volume paracentesis

MELD

Model for End‐Stage Liver Disease

TIPS

transjugular intrahepatic portosystemic shunt

Ascites is a common complication of cirrhosis, developing in approximately 50% of compensated cirrhotic patients within 10 years of diagnosis.1 Ascites develops as a direct result of sinusoidal portal hypertension and associated splanchnic vasodilation causing an increasing portion of the circulating blood volume to be in the splanchnic bed. This leads to effective hypovolemia and a hyperdynamic circulatory state. Arterial pressure in this state is maintained through the activation of vasoconstrictor systems, including the renin‐angiotensin‐aldosterone system, the sympathetic nervous system, and in later stages, hypersecretion of arginine vasopressin.2 Starling forces in the hepatic sinusoids drive excess sodium and water into hepatic lymph. When the capacity of hepatic lymphatics to return the lymph into circulation is exceeded, the excess sodium and water spills into the peritoneal cavity to form ascites. This detailed knowledge of the genesis of ascites provides rational targets for treatment.

Refractory Ascites

In about 5% to 10% of cirrhotic patients per year and as liver disease progresses, response to sodium restriction and diuretics diminishes, leading to refractory ascites.2 Refractory ascites is defined by the failure to mobilize ascites despite high‐dose diuretics, typically about 400 mg/day spironolactone or equivalent dose of a distal‐acting diuretic and 160 mg furosemide or equivalent dose of loop‐acting diuretic.2 Ascites is also considered to be refractory when effective doses of diuretics cannot be used because of development of diuretic‐associated complications. Refractory ascites is often associated with dilutional hyponatremia, type 2 hepatorenal syndrome (HRS), spontaneous bacterial peritonitis, and muscle wasting.3

Role of Transjugular Intrahepatic Portosystemic Shunt in Refractory Ascites

The management refractory ascites is currently limited to large‐volume paracenteses (LVPs) and transjugular intrahepatic portosystemic shunt (TIPS). TIPS is a procedure whereby a shunt is created between the portal vein and the hepatic vein. The goal of TIPS creation is to reduce the portosystemic pressure gradient, and in the case for refractory ascites, to less than 12 mm Hg.4 Although a gradient less than 8 mm Hg is considered optimal by some, there is a tradeoff in terms of risk for encephalopathy with greater portosystemic shunting.4 TIPS improves ascites via increased natriuresis through reduction in both proximal tubular sodium reabsorption and activity in the renin‐angiotensin‐aldosterone system.5 Notably, even after TIPS creation, patients must continue to restrict their sodium intake, and a majority of patients will require low‐dose diuretics to maintain an ascites‐free state.

Numerous randomized controlled trials have compared LVP and TIPS (Table 1).6 TIPS has been found to improve ascites in about 64% of patients versus 24% in those undergoing LVP, but had increased rates of post‐TIPS hepatic encephalopathy.6 Even though rigorously performed clinical trials have failed to demonstrate a survival advantage for TIPS, one meta‐analysis suggested an overall benefit (49% versus 32.5% 2‐year survival, respectively).7 There is, however, substantial heterogeneity in terms of design, methods, and quality across the trials limiting the interpretation of the meta‐analysis. Notably, other meta‐analyses, such as a well‐done Cochrane review8 of the same data, have failed to clearly demonstrate a survival advantage with TIPS. Some of the key confounding factors across trials were the varying and often small sample sizes, entry criteria, and variable use of albumin administration after paracentesis and methods to use sodium restriction uniformly in LVP versus TIPS arms.

Table 1.

Large‐Volume Paracentesis Versus Transjugular Intrahepatic Portosystemic Shunt for Refractory Ascites

Study	Patients (n)	Child‐Pugh Class C (%)	Average MELD Score	Improvement of Ascites (%)	Post‐TIPS HE (%)
Lebrec et al.
LVP	12	33	Not reported	0	0
TIPS	13	31	Not reported	38	23
Rossle et al.
LVP	31	23	Not reported	43	35
TIPS	29	38	Not reported	84	51
Gines et al.
LVP	35	43	Not reported	17	66
TIPS	35	37	Not reported	51	77
Sanyal et al.
LVP	57	Not reported	Not reported	16	23
TIPS	52	Not reported	Not reported	53	39
Salnero et al.
LVP	33	73	11.1	42	39
TIPS	33	39	11.1	79	61
Narahara et al.
LVP	30	30	10.6	27	17
TIPS	30	37	9.6	80	66

Adapted with permission from Clinical Liver Disease.6 Copyright 2014, The American Association for the Study of Liver Diseases.

Abbreviation: HE, hepatic encephalopathy.

TIPS has also been evaluated in those with HRS, albeit in noncontrolled studies.9 Most subjects in these studies had type 2 HRS. Although the clinical outcomes appeared promising in those who responded, it is cautioned that these were small, uncontrolled studied and carried a risk of selection bias, as well as reporting bias. In a small trial of subjects with type 1 HRS, subjects were initially treated with octreotide and midodrine and responders underwent a TIPS.10 This approach was associated with reasonable short‐term survival, whereas nonresponders to octreotide and midodrine had a very high mortality rate. The lack of control subjects, however, does limit the ability to evaluate this approach because other responders to octreotide and midodrine without a TIPS were not followed. It is therefore difficult to determine how much benefit accrued from the drugs versus TIPS.

Patient Selection

The relative and absolute contraindications for TIPS are listed in Table 2.6 Absolute contraindications include congestive heart failure, severe pulmonary hypertension (mean pulmonary wedge pressure > 45 mm Hg), and severe tricuspid regurgitation. Relative contraindications include severe coagulopathy, obstruction of hepatic veins (such as portal vein thrombosis and hepatic masses), and hepatic encephalopathy. Notably, some relative contraindications, such as anatomic obstruction of hepatic veins, are center specific, because some centers may have the capability of advanced TIPS procedures such as a left hepatic vein to left portal vein approach or an inferior vena cava to right portal vein approach through the caudate lobe with or without the aid of transabdominal or intravascular ultrasound.6

Table 2.

Contraindications for Transjugular Intrahepatic Portosystemic Shunt

Relative

Portal vein thrombosisa

Hepatocellular carcinoma

Moderate pulmonary hypertension

Thrombocytopenia (<20000 cells/m3)a

Coagulopathy (INR <1.5)a

Hepatic encephalopathy

Obstruction of all hepatic veinsa

Absolute

Congestive heart failure

Severe tricuspid regurgitation

Severe pulmonary hypertension

Multiple hepatic cysts

Bacteremia

Infection/Sepsis

Unrelieved biliary obstruction

Adapted with permission from Clinical Liver Disease.6 Copyright 2014, The American Association for the Study of Liver Diseases.

Abbreviation: INR, international normalized ratio.

^aCenter specific.

Models Used for Prediction of Post–Transjugular Intrahepatic Portosystemic Shunt Survival

Numerous models have been used to predict post‐TIPS survival (for any indication), of which the Child‐Pugh‐Turcotte score and Model for End‐Stage Liver Disease (MELD) score have been the most used.11 Of the two, the MELD score has shown to be superior to Child‐Turcotte‐Pugh score as a predictor for short‐term outcomes.12 In a meta‐analysis by Salerno et al.,7 a MELD score between 11 and 19 was found to have a significant survival benefit in patients who underwent TIPS versus patients who underwent LVP.7 In addition, a survival benefit for TIPS was seen across all MELD categories (Table 3). Furthermore, in a recent single‐center study of 100 patients who underwent TIPS placement, with an expanding covered stent, the 1‐year survival rate for a MELD score less than 15, 15 to 18, and greater than 18 was 84%, 67%, and 54%, respectively.13

Table 3.

Estimated 1‐Year Mortality Rate for Various Model for End‐Stage Liver Disease Scores of Transjugular Intrahepatic Portosystemic Shunt Versus Large‐Volume Paracentesis

MELD	TIPS (%)	LVP (%)
<10	28	42
11‐19	44	68
>19	62	84

Adapted with permission from Gut.5 Copyright 2010, British Society of Gastroenterology.

Currently, no randomized controlled trials have evaluating TIPS for refractory ascites in patients with a MELD score greater than 20. As such, in our opinion, each patient with refractory ascites should be evaluated carefully by looking at other clinical factors that can adversely affect outcomes. These factors include older age, high bilirubin, history of hepatic encephalopathy, and low sodium concentration.7 In those without severe hyperbilirubinemia or encephalopathy, TIPS may be considered in highly selected individuals on a case‐by‐case basis and by individual assessment of risks and benefits. It may be beneficial in those in whom the MELD score is being driven by progression to type 2 HRS. In those patients who receive a TIPS for ascites, it is important to continue surveillance for hepatocellular cancer even after ascites improves.

In summary, when compared with LVP, TIPS may be the treatment of choice for patients with refractory ascites. TIPS provides a high rate of response even in those with a high MELD score. Future studies are warranted to further discern the benefits of TIPS for those with a MELD score greater than 20.