Portopulmonary hypertension (POPH) is defined as pulmonary arterial hypertension in the setting of portal hypertension, regardless of the presence of underlying liver cirrhosis. POPH greatly affects morbidity and mortality in liver disease, and it may constitute a contraindication for liver transplantation (LT) in its moderate and severe forms.[1-4] The diagnosis of POPH is based exclusively on right heart catheterization (RHC) parameters according to well-defined criteria from expert consensus.[3-5]
Despite clear diagnostic criteria for POPH, there is still considerable variation among tertiary referral centers concerning which patients should proceed to RHC to arrive at a definitive POPH diagnosis. The right ventricular systolic pressure (RVSP), as determined by transthoracic echocardiography (TTE), has been used as the screening tool defining the need to proceed with RHC. A few basic concepts need to be remembered whenever we are considering adapting the use of screening tools in medical practice: (1) sensitivity and specificity are useful in defining the operational characteristics of a screening test, but the positive predictive value (PPV) and the negative predictive value (NPV) are the preferred parameters needed to guide a clinical decision; (2) because of the low prevalence of the disease, the NPV will tend to be high, regardless of the sensitivity; and (3) when we define the level of error allowed in the screening strategy, other health-related issues (ie, the financial consequences of a false negative on screening and the invasiveness and side effects of the diagnostic tool) should be taken into account in order to have a cost-effective cutoff point.
The debate about the RVSP cutoff point triggering the need for RHC in patients undergoing an LT evaluation is without a doubt a true work in progress. Using standard TTE, our colleagues at the Mayo Clinic in Rochester determined that an RVSP cutoff of 50 mm Hg could reliably identify patients with moderate to severe POPH. Their PPV and NPV were 74% and 97%, respectively. Using a different approach, investigators at Hospital Beaujon, Clichy, France aimed to determine a cutoff value that could identify all forms of POPH (mild to severe). They found that an RVSP of 30 mm Hg in TTE had a PPV of 59% and an NPV of 100%. On the basis of these results, using an RVSP of 50 mm Hg to rule out the need for RHC is reasonable, but there is still a very low risk of LT cancellation at the time of surgery. However, this risk is completely abolished with a cutoff point of 30 mm Hg, but at the cost of a fair number of false-positive results leading to unneeded RHC. In a time of monetary crisis and stronger regulations in health care, avoiding unnecessary tests is a matter of utmost importance.
There is certainly no ideal screening test, and many transplant centers use an RVSP between 30 and 50 mm Hg (eg, the University of California San Francisco uses a cutoff value of 40 mm Hg), either in isolation or along with other parameters, before they consider the need for RHC. In this issue of Liver Transplantation, Raevens et al. evaluate different RVSP cutoff values (30-50 mm Hg) according to TTE and aim to determine the most accurate cutoff value for defining the need for RHC to detect all forms of POPH. This study represents a commendable effort from the investigators to update and improve their previous results. With a cutoff value of 30 mm Hg, 74 of 152 patients met the criteria for probable POPH by TTE, although the diagnosis was confirmed in only 7 patients during RHC; this resulted in a sensitivity of 100%, a specificity of 54%, a PPV of 10%, and an NPV of 100%. As expected, increasing the cutoff value from 30 to 50 mm Hg caused a marked improvement in specificity with acceptable losses in sensitivity (the NPV remained at 99%). Notably, using a new cutoff value of 38 mm Hg (determined via a receiver operating characteristic curve analysis) resulted in an improved specificity of 82% with no changes in sensitivity, although the PPV was still very low at 22%. The authors were able to further improve the accuracy of this cutoff value by adding the presence of right ventricular (RV) dilatation [defined as a right ventricular end-diastolic diameter (RVEDD) > 3.3 cm]. Thus, through the consideration of a composite endpoint of an RVSP > 38 mm Hg and an RVEDD > 3.3 cm, the number of false positives could be dramatically reduced, and a sensitivity of 100%, a specificity of 93%, a PPV of 41%, and an NPV of 100% would be yielded. This approach would cause a significant reduction in the number of RHC procedures and limit unnecessary exposure to an invasive and expensive procedure (proportionate decrement = 57%).
This study by Raevens et al. is thus particularly relevant because it provides the reader with an in-depth analysis of the different clinical scenarios that can occur with the different cutoff values that can be selected for RVSP when one is screening for POPH. It also underscores RVEDD as another objective criterion that, when combined with RVSP, can improve the accuracy of TTE. Most transplant centers would likely feel uncomfortable with the need to send all their patients with an isolated RVSP > 38 mm Hg to RHC because of the high number of false positives, but the use of both RVSP and RVEDD brings some objectivity when one is deciding on the need for RHC in patients with an RVSP between 30 and 50 mm Hg. We believe that there is a need to reanalyze and refine the use of TTE as an effective screening method. Raevens et al. correctly allude to the importance of identifying RV dysfunction, which occurs as a direct result of the increased pulmonary vascular resistance affecting the RV afterload. They briefly refer to tricuspid annular plane systolic excursion measurements (a surrogate for the RV ejection fraction). There are precedents in the pulmonary arterial hypertension literature (not related to portal hypertension) on the usefulness of tricuspid annular plane systolic excursion and other RV parameters (eg, the Tei index or myocardial performance index) that could very well complement the determination of RVSP in the screening for POPH. A more sophisticated RV assessment such as speckle-based strain imaging might be promising for further improving the accuracy of TTE in the evaluation of RV function in patients for LT.
Collaboration with our cardiology colleagues in the use of evolving TTE methods to assess RV function in POPH, in addition to RVSP, seems prudent. Screening aside, we believe that further noninvasive studies of POPH should include RV functional surrogates/parameters in addition to RVSP measurements. Finally, because of the low prevalence of POPH, multicenter studies are to be encouraged in order to arrive at more robust conclusions.
Michael J. Krowka3
1Division of Hospital Medicine, Department of Internal Medicine, Paul L. Foster School of Medicine, Texas Tech University Health Sciences Center, El Paso, TX
2Division of Gastroenterology and Hepatology, University of Arkansas for Medical Sciences, Little Rock, AR
3Division of Pulmonary and Critical Care Medicine, von Liebig Transplant Center, Mayo Clinic, Rochester, MN