Model for end-stage liver disease (MELD) exception for hepatopulmonary syndrome


Hepatopulmonary syndrome (HPS) occurs when intrapulmonary vascular dilatation leads to abnormal arterial gas exchange in the setting of liver disease and/or portal hypertension.1 Over the past 15 years, HPS has emerged as a relatively common complication of cirrhosis, occurring in 15 to 20% of cases, and is an indication for liver transplantation (LT) in patients with significant hypoxemia.1 However, many clinical features of HPS that may influence exception to the Model for End-Stage Liver Disease (MELD) scoring system, including standardized diagnostic criteria, pre- and post-LT mortality rates, and rate of progression of hypoxemia, are not fully characterized.

The diagnosis of HPS is based on the presence of hypoxemia on room air due to intrapulmonary vascular dilatation in the setting of liver disease and/or portal hypertension.1 The prevalence of HPS (range, 12 to 32%) varies depending on whether abnormalities in arterial gas exchange are defined by an abnormal alveolar-arterial oxygen gradient or arterial hypoxemia (PaO2).2–7 For the purpose of considering HPS for MELD exception, arterial hypoxemia (PaO2 < 60 mm Hg) detected in the sitting position (to avoid effects of positional changes on PaO2) is a reasonable standard.

In a large prospective study of 200 LT candidates, pulse oximetry was found to be a useful screening technique for hypoxemia in patients with cirrhosis.3 The use of a screening oximetry threshold of ≤96% as a trigger for obtaining an arterial blood gas would have detected all patients with a PaO2 < 60 mm Hg and resulted in screening of 14% of the cohort. A proposed screening algorithm is provided in Figure 1. The most sensitive and appropriate screening test to detect intrapulmonary vascular dilatation is microbubble transthoracic contrast echocardiography.2 Although this is a qualitative test, it can also be used to screen for pulmonary hypertension and frequently distinguishes between intracardiac and intrapulmonary shunting. Radionuclide scanning with Tc99m-macroaggregated albumin (MAA), although quantitative, should not be used as a screening test because the MAA scanning technique is not standardized across centers, it is less sensitive than transthoracic contrast echocardiography, and it cannot distinguish between intracardiac and intrapulmonary shunting.8, 9 In addition, up to 20 to 30% of patients with HPS have comorbidities that may contribute to hypoxemia (i.e., ascites, hydrothorax, pneumonia, chronic obstructive pulmonary disease).10 Therefore, it is appropriate to review cardiopulmonary data (i.e., chest x-ray, pulmonary function tests, computed tomographic scan of the chest) in all patients being considered for MELD exception. The major diagnostic use of the MAA scan, if performed and interpreted by means of standardized methodology, is in defining the contribution of HPS to hypoxemia in patients with comorbidities and PaO2 < 60 mm Hg.8 If the MAA scan is positive for increased shunting, HPS is likely an important component of hypoxemia.

Figure 1.

Proposed screening algorithm for hepatopulmonary syndrome.

The natural history and outcomes before and after LT in patients with HPS are incompletely characterized. Two recent single-center studies suggest that mortality in well-characterized patients with cirrhosis not undergoing LT is significantly increased in those with HPS compared with those without HPS, and that the degree of hypoxemia and severity of liver disease adversely influence outcome.7, 11 However, the numbers of patients are small, and in one study, a subset of patients with HPS did not undergo LT because of comorbidities that may have influenced survival. Preliminary United Network for Organ Sharing data on MELD exception for HPS, which do not include data on pulmonary parameters, progression, or cause of death, have not shown reduced pre-LT survival in listed patients compared with those who were not granted MELD exception.12 This apparent disparity underscores the need to standardize, collect, and interpret MELD exception criteria for HPS and to follow and collect data on listed patients with HPS. The outcome of LT in patients with HPS is generally favorable, and survival is better in patients who undergo LT than in those not transplanted.1, 7, 11 However, after LT, mortality appears to be increased relative to patients without HPS, particularly in those with severe hypoxemia or advanced liver disease.7, 9, 11, 13 A single prospective study found that significant hypoxemia (PaO2 < 50 mm Hg) and increased intrapulmonary shunting documented by MAA scan (>20%) are associated with a marked increase in post-LT mortality.9 A recent retrospective, single-center report supports the concept that HPS outcomes are worse in patients with progressive hypoxemia.11 However, at present, there are insufficient data to definitively establish the level of severity of HPS at which LT should not be performed.

The progression of hypoxemia in patients with HPS and the factors that influence mortality are also not well characterized. No studies have addressed whether certain medications, specific complications of cirrhosis (i.e., bleeding, subacute bacterial peritonitis), or rapid changes in hepatic synthetic function may influence oxygenation and outcomes in patients with cirrhosis with HPS. In a small cohort, PaO2declined in 12 (85%) of 14 patients over time with an average decline of 5 mm Hg per year.11 To properly allocate MELD exception points for patients with HPS and to assess factors that influence pre- and post-LT outcomes, additional data from patients with HPS, followed after listing for LT, are required.


There are sufficient data to justify additional priority for patients with HPS. However, standardizing diagnostic data collection and review for MELD exception, including more detailed data collection on pre- and post-LT outcomes and incorporating assessment of progression, to modify MELD exception points will facilitate optimal LT outcomes in patients with HPS and avoid biasing allocation to or away from these patients.


HPS, hepatopulmonary syndrome; MELD, Model for End-Stage Liver Disease; LT, liver transplantation; MAA, Tc99m-macroaggregated albumin.


On the basis of the information presented above and the lack of detailed data (pre-LT severity and post-LT outcome) on patients with HPS currently collected by the United Network for Organ Sharing, revision of the MELD exception protocol is recommended with a focus on capturing and using data to optimize timing of LT in patients with HPS within the context of optimizing outcomes for all patients awaiting LT.

One approach would be to link eligibility for MELD exception to specific diagnostic data and serial collection of clinical data within the framework of ongoing and future National Institutes of Health–sponsored clinical studies of pulmonary vascular complications of cirrhosis. Specific components of such a strategy would include an infrastructure to manage data collection, and would provide an interval data review process with evidence-based initiation and modification of MELD exception criteria based on data analysis. In addition, inclusion of approved protocols for recipient blood and explant tissue samples for pathophysiologic studies and potential medical therapies could be included for a subset of participating centers.


Proposed diagnostic criteria for MELD exception in patients with HPS are listed in Table 1.

Table 1. Proposed Criteria for MELD Exception for Liver Transplant Candidates With Hepatopulmonary Syndrome
  1. Abbreviations: MELD, Model for End-Stage Liver Disease; MAA, Tc99m-macroaggregated albumin.

• Evidence of liver disease and/or portal hypertension.
• PaO2 < 60 mm Hg on room air at rest in the sitting position, arterial blood gas result provided.
• Pulmonary vascular dilatation documented by a positive transthoracic contrast echocardiography (full report included); MAA scan alone not acceptable.
• Absence of significant alternative pulmonary disease to explain hypoxemia (chest x-ray, pulmonary function tests, and chest computed tomography reports included) or if moderate or severe pulmonary function tests changes or significant chest x-ray abnormalities, an MAA scan positive for intrapulmonary shunting (using published protocol, result included).


Once patients are identified as having HPS with a PaO2 < 60 mm Hg and meet the proposed MELD exception criteria, the allocation of exception points is stratified by severity of hypoxemia and evidence of progression over time, as shown in Figure 2.

Figure 2.

Allocation of exception points for hepatopulmonary syndrome.