TWO of the more severe and challenging manifestations of end-stage liver disease include hepatopulmonary syndrome (HPS) and portopulmonary hypertension (PPHTN). Defined as the presence of intrapulmonary vasodilatation in conjunction with hypoxemia (oxygen tension in arterial blood < 70 mmHg with an inspiratory fraction of oxygen = 0.21), HPS may be seen in 6 to 23% of patients undergoing orthotopic liver transplantation, whereas PPHTN, defined as a mean pulmonary artery pressure greater than 25 mmHg with a normal pulmonary capillary wedge pressure in a patient with portal hypertension, may be seen in as many as 4% of such patients. [2,3]With the exception of an isolated case in the Japanese literature of intrapulmonary vasodilatation in a patient with PPHTN, there is no reported case of a patient with HPS and PPHTN simultaneously, thereby leading to the idea that these disorders may be mutually exclusive. However, as shown in the case we report, these syndromes may coexist.
A 46-yr-old woman with end-stage liver disease secondary to hepatitis C and alcoholism was evaluated for liver transplant 15 months before the operation. Her symptoms of liver failure included refractory ascites and encephalopathy. Evaluation revealed a 15 pack-yr history of cigarette smoking, decreased breath sounds at the lung bases bilaterally, a room air oxygen saturation of 95%, and digital clubbing. The history and physical examination indicated no teleangectasia, orthodeoxia, or platypnea. Screening echocardiography revealed mild mitral regurgitation, mild tricuspid regurgitation, an ejection fraction more than 70%, and estimated pulmonary artery systolic pressure greater than 45 mmHg. Mild pulmonary hypertension was confirmed with right heart catheterization, which revealed a pulmonary artery pressure of 48/25 mmHg (mean = 33 mmHg, transpulmonary gradient = 15 mmHg, pulmonary vascular resistance = 109 dyne [middle dot] s [middle dot]-cm-5), pulmonary capillary wedge pressure of 18 mmHg, and cardiac output of 11 l/min. Chest radiography revealed evidence of previous granulomatous disease. The findings of electrocardiography were normal, with the exception of sinus bradycardia. Table 1shows liver function tests and arterial blood gas values. At the time of transplantation, the patient's pulmonary artery pressure was 58/34 mmHg, with a pulmonary capillary wedge pressure of 15 mmHg immediately after induction of anesthesia and initiation of controlled ventilation. The patient's oxygen tension in arterial blood remained in the range of 70 to 80 mmHg despite inspired oxygen concentrations of 60 to 100% until the abdomen was opened, at which time the oxygen tension in arterial blood increased to 117 mmHg with 100% oxygen. Throughout the initial transplant procedure and immediately after operation, the patient remained relatively hypoxemic, and pulmonary artery pressures remained elevated; however, reperfusion occurred without hemodynamic consequences. Unfortunately, primary nonfunction developed on two separate occasions, necessitating three liver transplants in 4 days. Throughout the process, her pulmonary artery pressures remained elevated, with systolic pressures frequently more than 55 mmHg, and her oxygenation level was difficult to manage, frequently requiring periods of an inspiratory fraction of oxygen of 1.0. The day after the third liver transplant, transthoracic echocardiography was performed to try to determine the cause of hypoxemia. The study revealed a right-to-left shunt that was thought to be caused by a patent foramen ovale. However, intraoperative transesophageal echocardiography performed during an exploratory laparotomy the next day revealed a significant intrapulmonary shunt with microbubbles entering the left atrium via the pulmonary veins, confirming the presence of HPS (Figure 1). During her treatment in the intensive care unit, her pulmonary artery pressures remained in the range consistent with moderate to severe pulmonary hypertension, although oxygenation gradually improved. However, the third transplant graft eventually failed and the patient died.
Controversy exists regarding the diagnosis of PPHTN. The stated definition for primary pulmonary hypertension, a mean pulmonary artery pressure more than 25 mmHg with a pulmonary capillary wedge pressure less than 15 mmHg, does not account for the extremely elevated cardiac output manifest by patients with liver failure. Some have suggested using transpulmonary gradient (mean pulmonary artery pressure - pulmonary capillary wedge pressure) greater than 12 mmHg as the definition, but this still ignores the effects of such enormous blood flow. Pulmonary vascular resistance does take cardiac output into account, but to what degree increased flows effect the pulmonary circulation is unknown. Furthermore, as widespread vasodilatation occurs in patients with liver failure leading to decreased systemic vascular resistance and relatively low systemic blood pressure despite elevated cardiac output, could “normal” pulmonary vascular resistance and “normal” pulmonary artery pressures be, in fact, abnormally high in these patients? In addition, Kuo et al. have shown that patients with cirrhosis and “normal” pulmonary artery pressures respond to a fluid challenge with abnormally high pulmonary artery pressures compared with controls, indicating either an abnormal pulmonary circulatory response, left ventricular dysfunction, or both. Clearly more research must be done to understand the pulmonary circulation and its implications in patients with liver disease. We are comfortable with a diagnosis of mild PPHTN in this patient based on a mean pulmonary artery pressure of 33 mmHg, a transpulmonary gradient of 15 mmHg, and a significantly elevated pulmonary vascular resistance during her hospital course, despite an initially normal pulmonary vascular resistance.
The incidence of pulmonary hypertension is reported to be 0.13% in the general population, with a dramatic increase to 0.73% in those with hepatic failure and as much as 4% in those being evaluated for orthotopic liver transplantation. 
Although the cause of PPHTN is not fully understood, its pathophysiologic abnormalities result in hypertrophy and vasoconstriction of the pulmonary arterioles and an increase in pulmonary vascular resistance with an eventual increase in pulmonary artery pressure and right ventricular dysfunction.
This patient, as in most patients with PPHTN, had few if any symptoms of pulmonary hypertension, such as dyspnea on exertion, right heart failure, or chest pain. 
Patients thought to have PPHTN and those thought to have primary pulmonary hypertension are evaluated in a similar manner with chest radiography, electrocardiography, and echocardiography. Ultimately, however, right heart catheterization is required to confirm the diagnosis.
The presence of PPHTN is considered to be a relative contraindication to orthotopic liver transplantation, because the general consensus among transplant centers is that PPHTN is associated with significantly increased intraoperative and postoperative complications and death. . Recently, however, long-term intravenous epoprostenol has been used to reverse pulmonary hypertension and permit safe liver transplantation. [9,10]However, because our patient had no clinically evident right heart dysfunction, and only mild PPHTN before operation, no treatment was attempted. Furthermore, three case reports have documented reversal of PPHTN after orthotopic liver transplantation, [11-13]whereas only one showed a lack of reversibility. 
Hepatopulmonary syndrome can be defined as a clinical triad of liver disease, increased alveolar to arterial oxygen gradient while breathing room air, and evidence of intrapulmonary vascular dilatations. 
As with PPHTN, the pulmonary vascular abnormalities that result in HPS are believed to result from the presence of vasoactive substances abnormally cleared by the failing liver, whereas the signs and symptoms of liver disease comprise the most common presentation in patients subsequently diagnosed with HPS. Although the most common symptom is dyspnea, which may worsen when the patient assumes the upright position (platypnea), one of the most striking physical findings is clubbing of the digits, as was present in this patient, and cyanosis with substantial hypoxemia on room air that worsens when the patient is in the upright position (orthodeoxia). 
Although respirometry and chest radiography are seldom revealing in the diagnosis of HPS, nuclear scanning with technetium-99m-labeled macroaggregated albumin is often used to detect the presence of HPS and to quantify the degree of intrapulmonary shunting. An additional means of diagnostic confirmation is the method we used, which demonstrates, via echocardiography, the appearance of microbubbles (50-100 [micro sign]m) of agitated air in the pulmonary veins approximately five heart beats after injection through a central venous catheter (Figure 1). Pulmonary angiography is also used for imaging and localizing the pulmonary vascular abnormalities in HPS. The discrepancy in this patient between the reports from the transthoracic echocardiography and the transesophageal echocardiography was the results of an error in interpretation. The transthoracic echocardiographer failed to note the delay of four to five heart beats between the injection of microbubbles in the right atrium and their appearance in the left atrium and failed to note the route by which the microbubbles entered the left atrium (pulmonary veins vs. patent foramen ovale). Although methylene blue, a nitric oxide antagonist, has been shown to improve oxygenation in some patients with HPS, only liver transplantation has reversed HPS with moderate to severe hypoxemia. [17,18]Generally, the clinician must simply ensure the ability to oxygenate the patient's blood at the time of diagnosis and transplant. In most cases, increased oxygen concentration is sufficient secondary to increased diffusion pressure across the alveolar-capillary membrane. It is unclear why our patient was so resistant to an increased oxygen concentration, but because we could oxygenate her blood and she had other more critical issues, we did not look further. In extreme cases, embolization of the grossly dilated vessels has been successful in treating hypoxemia; however, we did not attempt that in our patient because of her poor medical condition.
Although the PPHTN was mild in our patient, our purpose in this article was to document, for the first time, the coexistence of HPS and PPHTN. These are two pathologically dissimilar diseases with a common cause that were thought, until now, to be mutually exclusive. We do not know how commonly this occurs; however, the recently initiated national HPS-PPHTN database may address this question. In addition, studies designed to identify the molecular basis of both disease processes and the effect on complications and death during orthotopic liver transplantation should be done. Interestingly, both disease processes previously were thought to be relative contraindications to orthotopic liver transplantation, the only possible long-term treatment for either of them.