Hepatopulmonary syndrome and liver transplantation: Who, when, and where?


  • Rajasekhar Tanikella M.D,

    1. Internal Medicine Residency Program, Department of Internal Medicine, The University of Texas Medical School at Houston, Houston, TX
    Search for more papers by this author
  • Michael B. Fallon M.D

    Corresponding author
    1. Division of Gastroenterology, Hepatology, and Nutrition, Department of Internal Medicine, The University of Texas Medical School at Houston, Houston, TX
    • Division of Gastroenterology, Hepatology, and Nutrition, Department of Internal Medicine, The University of Texas Medical School at Houston, 6431 Fannin Street, MSB 4.234, Houston, TX 77030
    Search for more papers by this author
    • fax: 713-704-6616

  • See Article on Page 2427

  • Potential conflict of interest: Nothing to report.

Hepatopulmonary syndrome (HPS) is a relatively common pulmonary vascular complication of cirrhosis and/or portal hypertension (PH) found in 5%-32% of patients evaluated for liver transplantation (LT). It is characterized by pulmonary microvascular dilatation and remodeling, which result in impaired oxygenation in the absence of marked intrinsic cardiopulmonary disease.1–3 The range in observed frequency of HPS is influenced by what level of gas exchange abnormality is used to define the presence of the syndrome.2 The European Respiratory Society (ERS) consensus definition establishes an alveolar-arterial oxygen gradient (AaPO2) ≥15 mmHg (age >64 years including ≥20 mmHg) as abnormal and further delineates severity into mild (partial pressure of oxygen [PaO2] ≥80 mmHg), moderate (PaO2 ≥60-<80 mmHg), severe (PaO2 ≥50-<60 mmHg), or very severe (PaO2 <50 mmHg).4 There has been a focus on patients with severe HPS with regards to the Model for End-Stage Liver Disease (MELD) exception for LT based on data that both pre- and postoperative mortality are increased in this group, relative to non-HPS patients without cirrhosis.3, 5, 6 However, the only multicenter prospective study found that the increased mortality in HPS, relative to non-HPS, occurred across the spectrum of oxygenation abnormalities, suggesting that factors other than severity of hypoxemia influence outcome.7

Medical therapy for HPS has generally been ineffective, although a number of promising targets have been identified, including the endothelin system, nitric oxide synthase, tumor necrosis factor alpha, and angiogenesis.8–14 Currently, LT is the only successful treatment for HPS and typically results in complete resolution of gas exchange impairment.1, 15 However, the effect of institution of the MELD exception policy for severe HPS on wait-list and post-transplant mortality remains incompletely defined. An analysis of the United Network for Organ Sharing (UNOS) database in 2008 suggested that wait-list survival was better and post-transplant survival similar between HPS exception patients and non-HPS patients.16 Unfortunately, the UNOS database does not include data on pulmonary parameters and many patients did not undergo HPS screening, making it difficult to be sure which patients had HPS and to define the severity of the process.

Against this backdrop, Iyer et al report on a follow-up retrospective analysis of outcomes in the largest single-center cohort of HPS patients.17 One hundred and six patients with HPS evaluated at the Mayo Clinic from 1986 through 2010 were included and outcomes were compared before and after the institution of MELD exception for HPS (January 1, 2002). There are three major findings of the study: (1) Long-term outcomes in patients transplanted for HPS are excellent; (2) there has been a trend toward improved post-LT outcomes since the inception of the MELD exception for HPS; and (3) transplant outcomes are similar between HPS and non-HPS patients. These findings strongly support employing LT as a therapy for HPS and suggest that the current MELD exception policy may have contributed to the goal of improving LT outcomes in HPS.

A closer look at the results of the study provides some additional insights. First, based on the survival curves, the trend in improved post-LT survival among HPS patients in the MELD exception era appears most prominent in the early postoperative period. This finding corroborates observations from other groups suggesting that post-LT survival in HPS is improving.16–18 Whether these outcomes reflect the application of the MELD exception for HPS and prevention of deterioration of hypoxemia, advancement in surgical techniques, or improved perioperative management of HPS are unclear. The observation in the current study that many patients had very severe hypoxemia and that the degree of hypoxemia and intrapulmonary shunting did not influence post-LT outcomes, in contrast to the earlier analysis from the same group3 and studies from the early MELD exception era,6, 15 supports that standardization and optimization of perioperative care and surgical advances may be important factors.

There are several limitations in the current study that impede generalization of the findings. This is a single-center retrospective analysis of data over a long period of time. The number of patients who declined research authorization is not included, and confounding related to changes in management, surgical technique, and/or comorbidities cannot be assessed. Furthermore, the small numbers of patients in the pre- and post-MELD groups limit the depth of the analysis, as reflected in the inability to use multivariate models. Finally, the diagnostic criteria for HPS in this study differ from the ERS task force. Specifically, all patients with mild to moderate HPS and a PaO2 >70mmHg (approximately 50% of all HPS patients in other studies7) were excluded from the HPS group and are included in the control group. If patients with mild to moderate HPS indeed have increased mortality relative to non-HPS patients, as previously observed,7 then the current approach would have the potential to introduce a marked bias in the results.

What can we conclude currently regarding who, when, and where in LT therapy for HPS?


Current and earlier evidence supports that LT outcomes are excellent in severe and very severe HPS. However, patient selection is critical to obtaining such outcomes and no specific criteria exist to guide the selection of HPS patients for LT. Our approach has been to consider significant cardiopulmonary disease and end-stage renal disease as relative contraindications to LT in very severe HPS because these factors affect early post-LT mortality. More data are needed and might be available from further analysis of the Mayo experience.


Application of the current HPS MELD exception criteria (listing at a MELD score of 22 in LT candidates with evidence of PH and intrapulmonary shunting causing a PaO2 <60 mmHg on room air with a 10% mortality equivalent increase in points every 3 months occurring if PaO2 remains <60 mmHg) achieves a key goal of the MELD exception policy: obtaining similar LT outcomes in HPS and non-HPS patients. Therefore, we appear to be on the right track. On the other hand, the lack of screening for mild to moderate HPS in the UNOS and Mayo data results in misclassification of these patients to the non-HPS group, thereby diminishing recognition of the effect of HPS on outcomes. Defining predictors of increased mortality in mild to moderate HPS is an important future focus for research.


The consistently good outcomes of LT for HPS in the current study suggest that experience and standardized post-LT care may enhance outcomes. Considering LT at experienced centers for the most advanced HPS cases and using established post-LT protocols, such as that outlined in the current study, at all LT centers may further improve outcomes in this disorder.


alveolar-arterial oxygen gradient


European Respiratory Society


hepatopulmonary syndrome


liver transplantation


Model for End-stage Liver Disease


partial pressure of oxygen


portal hypertension


United Network for Organ Sharing