Portopulmonary hypertension in cirrhotic patients: Prevalence, clinical features and risk factors

The incidence and clinical features of portopulmonary hypertension (POPH) have not been adequately described and it is currently unknown whether an association exists between the severity of POPH and liver function. Additionally, POPH risk factors are yet to be identified. The aim of this study was to determine the prevalence, describe the clinical features and investigate the potential risk factors of POPH. We conducted a study of 100 cirrhotic patients hospitalized between March 2011 and May 2012 at Tongji Hospital in Shanghai. The clinical characteristics of patients with and without POPH were analyzed. Clinical variables with a possible association with POPH were measured and pulmonary artery systolic pressure (PASP) was determined by cardiac Doppler echocardiography. Of the 100 patients enrolled in this study, 10 were diagnosed with POPH. Seven of the cases were mild, two were moderate and only one was severe; eight were attributed to viral infections. POPH was not detected in patients with schistosomal or alcoholic cirrhosis. Hemoglobin (Hb) levels were lower in patients with POPH compared to those without POPH (P<0.01) and the severity of POPH was not significantly correlated with Child-Pugh grade (R=−0.06, P=0.09). Hb levels, incidence of hepatitis C virus (HCV) infection and portal vein thrombosis differed between the two groups (P<0.05). Hb levels were identified as an independent risk factor associated with POPH and portal vein thrombosis may play an important role during the development of POPH. However, the severity of POPH was not associated with liver function.


Introduction
Portopulmonary hypertension (POPH) refers to the condition of simultaneous pulmonary arterial and portal hypertension. Hemodynamically, it is defined as a mean pulmonary artery pressure (MPAP) >25 mmHg and normal volume status or pulmonary capillary wedge pressure (PCWP) <15 mmHg in patients with chronic liver disease and portal hypertension (1)(2)(3). POPH has been reported in ~6-9% of patients with advanced liver disease, that are referred for liver transplants (4,5). This variation in prevalence results from the different groups of cirrhotic patients studied and different diagnostic procedures used.
Clinical manifestations of portal hypertension typically precede those of pulmonary artery hypertension by 2-15 years. The most common symptoms of POPH include dyspnea upon exertion, syncope, chest pain, fatigue, hemoptysis and orthopnea. The presence of molecular markers is also likely, as one study supported the hypothesis that pulmonary vasculature may be exposed to either cytokines or excess circulating vasoconstrictors, including interleukin-6 and endotoxin produced by the diseased liver (6). However, to date, the predictors of and biological mechanism responsible for the development of this complication remain unknown.
There are no known clinical factors that determine the risk of POPH in patients with advanced liver disease. Similarly, the mechanism for pulmonary vascular obliteration in patients with portal hypertension, characterized by systemic vasodilation remains unclear. It follows that the identification of patient characteristics associated with an increase or decrease in the probability of developing POPH may not only be clinically useful, but may also shed light on the etiology of this relatively common comorbidity of portal hypertension. Therefore, the aims of this study were to explore the clinical features of POPH and to identify risk factors associated with POPH in cirrhotic patients.
with intrinsic liver cirrhosis. Patients with significant obstructive lung disease, restrictive ventilatory defects, human immunodeficiency virus, severe aortic or mitral stenosis, regurgitation or significant left ventricular dysfunction were excluded. Also excluded were individuals with hepatocellular carcinoma or any other malignancy, known hemostatic disorders other than liver disease, bacterial infection, a clinical history of peripheral venous thrombosis or Budd-Chiari syndrome, spleen resection, lung and liver transplantation and those receiving endoscopic treatment or anticoagulation therapy. A total of 100 patients were considered acceptable and were enrolled in our study. Informed consent was obtained from all patients and the study was carried out according to the principles of the declaration of Helsinki and the guidelines of the institutional ethics committee.
Liver cirrhosis was diagnosed by clinical findings or morphology and liver function of the POPH patients was scored using the Child-Pugh classification (7). Past medical history and social history were recorded. Non-invasive screening for POPH was performed and evaluated by Doppler echocardiog-  raphy. In order to identify risk factors associated with POPH, the patients were classified into two groups: a POPH group [pulmonary artery systolic pressure (PASP) ≥40 mmHg] and a non-POPH group (PASP <40 mmHg) (8,9).

Diagnosis of POPH. Pulmonary vascular resistance (PVR)
is considered an essential feature of POPH. Diagnostic measurements of PVR are traditionally obtained by right heart catheterization (RHC) (10,11); however, since bleeding complications are a concern among cirrhotic patients, other forms of diagnosis are preferred. Results of Doppler echocardiography are comparable to RHC in non-cirrhotic patients with pulmonary hypertension (12,13). For cirrhotic patients undergoing liver transplantation (14,15), it decreases the requirement for repeated invasive measurements. In our study, PASP (measured by Doppler echocardiography) was used to diagnose POPH. Noninvasive color Doppler echocardiographies were performed on all patients. The MPAP was calculated from PASP (MPAP = 0.61 PASP + 2 mmHg). PASP values of 38-54 mmHg were considered to represent mild POPH, 55-69 mmHg were moderate and values ≥70 mmHg were severe (16).
Statistical analysis. The SPSS software package for Windows (SPSS version 11.0, SPSS Inc., Chicago, IL, USA) was used for statistical analysis. Continuous data were summarized using mean ± standard deviation (SD) or median (interquartile  In this multivariate analysis, we used a number of variables, including those that were not significant univariate predictors, since they may contribute to a multiple regression model in unforeseen ways due to complex intercorrelations among them. The coefficients obtained from the logistic regression analyses were also expressed in terms of odds ratios (ORs) with 95% confidence intervals. Considering the number of cases, we determined that the final multivariate model should include ≤4 predictors to prevent overfitting. A two-tailed P-value ≤0.05 was considered to indicate a statistically significant difference.

Results
Clinical features.  Tables II and III). The levels of BPC, ALB, TBIL, Hs-CRP, D-dimer, thrombin time (TT), Fib, ICAM-1, IFN-α and TFN-α did not differ significantly between POPH and non-POPH patients. There were no significant differences in the levels of APTT and PT between the two groups. However, the levels of Hb were lower in the POPH group than those in the non-POPH groups (P=0.001). Hb was determined to be a risk factor of POPH by univariate analysis (Table IV).  Correlation between Child-Pugh class and severity of POPH. In our study, a correlation between Child-Pugh class and severity of POPH (reflected by the value of PASP) was investigated. The results revealed that the PASP values were not significantly correlated with Child-Pugh class (R=-0.06, P=0.09; Fig. 1).
Multivariate analysis. The Hb, HCV, D-dimer and portal vein thrombosis were included in multivariate logistic regression analysis. Only Hb was identified as an independent factor associated with an increased risk of POPH (OR=0.952; Table V).

Discussion
In the present study, POPH was diagnosed in 10% of patients with cirrhosis. This is within the range of prevalence reported in the literature (20). In previous studies, POPH has been estimated to occur with a prevalence of 16.1% in patients with cirrhosis and refractory ascites and 0.25-4% in patients with cirrhosis without refractory ascites (18,19). Although our results demonstrated that there are higher percentages of females with cirrhosis that develop POPH, no significant difference was observed in the incidence of POPH across the gender (results not shown). This indicates that males and females with cirrhosis are at risk of developing POPH. These results contradict the findings of a previous study that identified that females are at a higher risk of POPH (11).
In our study, the majority of the POPH patients had viral cirrhosis (HBV and HCV). Viral cirrhosis was the most common type of cirrhosis among our Chinese population; however, our results demonstrated that a higher incidence of POPH exists among patients with viral cirrhosis. While previous research identified that HCV infection was negatively correlated with POPH (11), this was not supported by our data. In contrast, the high incidence of POPH among patients with HCV infection indicates that patients with HCV cirrhosis may be at higher risk of developing POPH.
With regard to the severity of POPH, our results demonstrated that the severity of POPH is unrelated to liver function, as we were unable to identify an association between POPH severity and Child-Pugh class. However, this conclusion is based on a small number of cases and larger studies are required to verify this correlation.
Previously hypothesized risk factors of POPH were not associated with POPH in our study. Age, gender and Child-Pugh class did not differ between the POPH and non-POPH groups. Also, hepatic complications, including GI hemorrhage, hepatorenal syndrome ascites and encephalopathy did not differ between the groups. Similarly, HBP, DM, coronary artery disease, drug use and blood transfusion did not appear to affect the risk of developing POPH and smoking and alcohol use were common in the two groups.
With respect to patient medical histories, only portal vein thrombosis differed between the two groups, indicating that portal vein thrombosis is associated with POPH. These results are supported by a previous study that described microembolism of the lungs in hepatic fibrosis due to recurrent cholangitis. The passage of small emboli to the lungs was attributed to the presence of small hepatic arteriovenous fistulae (21). The correlation between POPH and portal vein thrombosis may be explained by a similar anatomical mechanism. Portal venous obstruction secondary to portal vein thrombosis would likely enlarge such communications, resulting in repeated microembolism of the lungs. This may lead to pulmonary hypertension and, ultimately, congestive failure due to chronic cor pulmonale (21).
Among the laboratory parameters examined in our study, levels of Hb were lower in the POPH group than in the non-POPH group. To our knowledge, a correlation between POPH and Hb has not been described previously. However, it is known that a decrease in Hb leads to a significant increase in cardiac output and exacerbated hyperdynamic splanchnic circulation (22). The hyperdynamic splanchnic circulation is a major contributor to portal hypertension. Therefore, it is not surprising that Hb levels were significantly lower in the POPH group and that Hb level was the only variable independently associated with POPH. These findings indicate that Hb level is an independent risk factor and plays a key role in the development of POPH. While HCV and PVT were also associated with a higher incidence of POPH, these factors were not independent predictors of POPH. Perhaps they contributed to the formation of POPH through interactions with other unknown factors. However, significant differences between the two groups were not observed for any of the other markers that we investigated.
In conclusion, within our study population POPH was present in 10% of cirrhotic patients. It was most common among patients with viral cirrhosis and absent among schistosomal and alcoholic cirrhosis patients. Notably, the severity of POPH was unrelated to liver function (Child-Pugh classification). HCV infection and portal vein thrombosis may play important roles during the development of POPH; however, Hb level is the only significant, independent predictor of POPH. Future studies should examine the mechanistic role of these factors in the development of POPH.