Pulmonary Gas Exchange Abnormalities in Liver Transplant Candidates

Pulmonary Gas Exchange Abnormalities in Liver Transplant Candidates Rosmawati Mohamed,* Jonathan W. Freeman, Peter J. Guest, Michael K. Davies, and James M. Neuberger* Abnormal diffusing capacity is the commonest pulmonary dysfunction in liver transplant candidates, but severe hypoxemia secondary to hepatopulmonary syndrome and significant pulmonary hypertension are pulmonary vascular manifestations of cirrhosis that may affect the perioperative course. We prospectively assessed the extent of pulmonary dysfunction in patients referred for liver transplantation. A total of 57 consecutive patients with chronic liver disease were evaluated. All patients had a chest radiograph, standing arterial blood gas on room air, pulmonary function testing, and Doppler echocardiogram. Those patients with arterial hypoxaemia (PaO 2 < 10 kpa) also underwent 99m Tc-macroaggregated albumin lung scan, and nine patients had agitated normal saline injection during echocardiography to define further the existence of pulmonary vascular dilatation. Reduced diffusing capacity for carbon monoxide less than 75% of the predicted value was found in 29 of 57 (51%) patients. Although elevated alveolar-arterial oxygen tension difference was detected in 35% (20/57) of the patients, only four (7%) patients had hypoxemia. We were unable to find evidence of intrapulmonary vascular dilatation either on the lung scan or saline-enhanced echocardiography in any of these patients. Reduction in diffusing capacity for carbon monoxide was noted in 75% (18/24) of patients who were transplanted for primary biliary cirrhosis and was accompanied by widened alveolar-arterial oxygen tension in 10 out of 18 (56%) of patients. This study shows that in liver transplant candidates, diffusion impairment and widened alveolar-arterial oxygen tension difference were frequently detected, especially in patients with primary biliary cirrhosis. (Liver Transpl 2002;8:802-808.) A wide spectrum of pulmonary gas exchange abnormalities can be found in patients with advanced liver disease. 1,2 In many patients, the compliance of the lung is decreased by massive hepatomegaly, ascites, basal atelectasis, or pleural effusions. If none of these From the *Liver Unit, Featherstone Department of Anaesthetics and Intensive Care, and the Department of Radiology, Queen Elizabeth Hospital, Birmingham, and the Department of Cardiology, Selly Oak Hospital, Raddlebarn Road, Birmingham, England. Address reprint requests to James Neuberger, DM, Liver Unit, Queen Elizabeth Hospital, Birmingham, B15 2TH, UK. Telephone: (44) 121-627-2414; FAX: (44) 121-627-2449; E-mail: James.Neuberger@ uhb.nhs.uk Copyright 2002 by the American Association for the Study of Liver Diseases 1527-6465/02/0809-0072$35.00/0 doi:10.1053/jlts.2002.33746 factors or a recognizable pulmonary disorder is present, the three main mechanisms that have been shown to contribute to the development of arterial hypoxemia are ventilation-perfusion mismatch, alveolar-capillary limitation of oxygen diffusion, and intrapulmonary shunts. 3 In the preoperative assessment of the liver transplant candidate, the impact of long-standing liver disease on other organ systems must be assessed in the light of the stresses that will be imposed by the liver transplantation. Arterial hypoxemia secondary to hepatopulmonary syndrome and significant pulmonary hypertension is a pulmonary manifestation of cirrhosis that may contribute to the morbidity and mortality in the perioperative period. 4,5 Hepatopulmonary syndrome is defined as hypoxemia and intrapulmonary shunting in patients with liver disease in the absence of intrinsic cardiopulmonary disease. 5,6 The intrapulmonary right-to-left shunt occurs as a consequence of the presence of dilated small precapillary pulmonary blood vessels termed intrapulmonary vascular dilatations (IPVD). Reports from the United States found abnormal arterial oxygenation in up to 69% of liver transplant candidates and IPVD present in 13% to 47%. 1,7-9 The development of pulmonary hypertension in patients with portal hypertension is well described. 10 Pulmonary hypertension is defined as elevation of pulmonary vascular resistance and mean pulmonary artery pressure greater than 30 mm Hg with a normal pulmonary capillary wedge pressure. Pulmonary hypertension is unusual as a complication of liver disease, but may pose a difficult management problem, especially in the consideration of liver transplantation. 11 This study was undertaken to investigate the incidence of pulmonary gas exchange abnormalities in consecutive patients referred to one center for orthotopic liver transplantation (OLT). Patients and Methods Fifty-eight consecutive adult patients with chronic liver disease referred and subsequently listed for OLT were prospectively evaluated. Inclusion criteria were age 16 years and over and the ability to perform a full lung function test satisfactorily. Exclusion criteria were the presence of coronary artery or 802 Liver Transplantation, Vol 8, No 9 (September), 2002: pp 802-808

Gas Exchange in OLT Candidates 803 Table 1. Demographic and Clinical Characteristics (n 57) Median age in years (range) 51(16-69) Female : male ratio 30 F : 27 M Child-Pugh grade Grade C 31 Grade B 21 Grade A 5 Cause of liver disease Primary biliary cirrhosis 24 Sclerosing cholangitis 11 Alcoholic liver disease 9 Autoimmune hepatitis 5 Wilson s disease 2 Cryptogenic cirrhosis 2 Budd Chiari syndrome 1 Secondary biliary cirrhosis 1 Familial hepatic fibrosis 1 Alpha1antitrypsin deficiency 1 NOTE. Values represent number of patients unless otherwise noted. valvular heart disease, symptomatic lung disease caused by conditions other than the liver, such as moderate or severe asthma or chronic obstructive airways disease requiring regular bronchodilator therapy. One patient in grade 2 encephalopathy who underwent transplantation within 24 hours of being on the liver transplant waiting list was unable to perform the lung function test and therefore excluded from the study. During the study period, there was no patient who was excluded or deemed not eligible for liver transplantation on the basis of pulmonary problem. Informed consent was obtained from each patient and the study was approved by the local Research Ethics Committee. Patients A total of 27 men and 30 women, aged 16 to 69 years (median, 51years) were studied. The causes of liver disease are shown in Table 1. The severity of liver disease was Child- Pugh grade C in 31cases, B in 21, and A in 5. Smokers were defined as current if they had smoked tobacco within 1year of transplant assessment. A positive smoking history included patients who were previous smokers but stopped more than a year before transplant assessment. Methods Chest radiograph, standing arterial blood gas on room air, pulmonary function tests, and two-dimensional, Doppler, and color-flow echocardiograms were carried out as part of the pretransplant workup. The chest radiographs were reviewed by one radiologist (PJG) who had no knowledge of the patients clinical status. All patients with a PaO 2 of less than 10 kpa (75 mm Hg) had blood gas analysis during 100% oxygen breathing and a macroaggregated albumin (MAA) lung scan to assess quantitatively the degree of intrapulmonary shunting. Nine patients (including two patients with hypoxemia) randomly were assigned to have agitated normal saline injection during echocardiography to define the existence of pulmonary vascular dilatations. The alveolar-arterial oxygen tension difference, PA-aO 2, was calculated as follows: PAO 2 FiO 2 Pb 47 (PaCO 2 R FiO 2 1 R PaCO 2 R where FiO 2 is the inspired oxygen fraction, Pb is the barometric pressure and R is respiratory exchange ratio, assumed to be 0.8. 12 MAA lung scan. Transmission scan was carried out in the supine position with a Toshiba GCA 902 A/S camera using a LEAP collimator (Toshiba, Tokyo, Japan). Immediately after injection of 100 Mbq of Tc-99 m labeled MAA (CIS Bio international), anterior and posterior whole body scans were performed. In the normal individual, the MAA (95% are between 10 to 100 m with none larger than 150 m) are trapped in the capillary bed (normally, 8 to 15 m in diameter). In the presence of IPVD (where the pulmonary vessels are 15 to 500 m in diameter), the MAAs pass directly through the dilated capillaries into the systemic circulation enabling quantitative assessment of intrapulmonary shunting. Saline microbubble injection echocardiography. For qualitative assessment of right to left shunting, 20 ml of agitated 0.9% saline solution was injected into a peripheral intravenous line. This was preceded by transthoracic echocardiography with an apical four-chamber view and continued for at least one minute after the saline injection. Microbubbles of agitated saline, 60 to 90 m in diameter are identified as echoes in the right atrium and ventricle. Normally, these microbubbles are then cleared by passage through the pulmonary circulation. In the presence of right to left intracardiac shunting, there is immediate opacification of the left atrium. Patients with IPVD show delayed appearance of echoes in the left atrium approximately 3 to 6 beats after their appearance in the right ventricle. 5 Pulmonary function tests. Full pulmonary function tests included dynamic and static lung volume determinations and single-breath carbon monoxide transfer factor according to the British Thoracic Society and the Association of Respiratory Technicians and Physiologists guidelines. 13 Total lung capacity was determined from the helium dilution method. Measurements of diffusing capacity of carbon monoxide (DLco) were adjusted at a standard haemoglobin of 14.6 g/dl. The additional index of transfer coefficient, KCO, was derived from the equation, KCO DLco V A eff, where V A eff is the effective alveolar volume estimated from the dilution of helium during the single breath-hold. Refer-

804 Mohamed et al Table 2. Details on the 4 Patients With Hypoxemia Patient PaO 2 on Air (kpa/mm Hg) PaO 2 on 100% O 2 (kpa/mm Hg) FEV 1 / FVC (%) RV/TLC (%) DLco (% predicted) MPAP (mm Hg) PVR (dynes s/cm 5 ) PCWP (mm Hg) 18.3 / 62 53.1/ 398 59% 35% 44 32 54 14 2 9.1/ 68 73.9 / 554 58% 53% 44 28 59 23 3 9.2 / 69 41.3 / 310 79% 37% 69 18 48 17 4 9.9 / 74 75.7 / 568 49% 43% 36 32 224 17 Abbreviations: FEV 1 /FVC, ratio of forced expiratory volume in 1second to forced vital capacity; RV/TLC, ratio of residual lung volume to the total lung capacity; DL CO, transfer factor for carbon monoxide corrected for haemoglobin (% predicted value); MPAP, mean pulmonary artery pressure (mm Hg); PVR, pulmonary vascular resistance (dynes s/cm 5 ); PCWP, pulmonary capillary wedge pressure (mm Hg); CI, cardiac index (l/min/m 2 ). ence values were obtained from the European Community for Steel and Coal summary equations. 14 Results were expressed as a percentage of the reference value for each of the lung function indices adjusted for age, height and weight. Measurement of pulmonary hemodynamic profiles. Baseline pulmonary hemodynamic profiles were obtained in these patients using radial arterial line and flow-directed pulmonary artery catheter (Swan-Ganz, Baxter Edwards Laboratories, California) after induction of anaesthesia for OLT. The pulmonary artery catheter is routinely inserted in all patients undergoing liver transplantation. Statistics Correlations were examined with Pearson correlation coefficient for normally distributed measurements and Spearman method for nonnormal distributions. The Mann-Whitney U test and Wilcoxon rank sum tests were used for direct comparison between 2 groups with continuous variables. -square test was used for comparisons between 2 groups with categorical or noncontinuous variables. P values of less than.05 was considered significant. Results There were 10 current smokers and 7 previous smokers. Dyspnoea, either at rest or on exertion, was reported in 21patients. A total of 20 patients had cutaneous spider naevi. Ascites was present in 39 patients. Arterial Blood Gas Values The median arterial PCO 2 was 4.2 kpa (range 2.7 to 5.1 kpa) for the study group. There was a weak but statistically significant inverse correlation between PaCO 2 and Child-Pugh score (r 0.4; P.008). The PaO 2 ranged from 8.3 kpa to 15.2 kpa, with a median of 12.6 kpa. Lower PaO 2 levels tended to occur in patients with a positive smoking history, dyspnoea, or spider naevi, but these associations did not reach statistical significance (P.051, 0.08 and 0.06, respectively). There was no significant correlation between PaO 2 level and Child-Pugh score (r 0.1, P.9). Arterial hypoxemia (PaO 2 10 kpa) was found in only 4 patients (7%). Three of these four patients were smokers who had an obstructive pattern of lung function test and 2 had associated hyperinflation (Table 2). One hundred percent O 2 breathing resulted in normal response (PaO 2 500 mm Hg or 67 kpa) in all but two patients with moderate hypoxemia. Alveolar-arterial oxygen tension difference was elevated ( 20 mm Hg or 2.67 kpa) in 20/57 (35%) patients. There was no significant correlation between PaO 2 gradient and the Child-Pugh score (r 0.5, P.74). The association between alveolar-arterial oxygen tension difference and presence of dyspnoea was statistically significant (P.027). MAA lung perfusion scan. In the 4 patients who were hypoxemic and had the lung perfusion scan, MAA particles were not detected in the brain or the kidney, suggesting no evidence of intrapulmonary shunting through IPVD. Chest radiograph. Interstitial shadowing was the isolated abnormality on chest radiographs in 4 patients. Unilateral basal consolidation was noted in 5 patients. Two patients had small pleural effusions. Increased distal pulmonary vascularity was present unilaterally in 2 patients. Echocardiography. Echocardiography showed no evidence of pulmonary hypertension in any of the patients as assessed by tricuspid regurgitation jet or right ventricle size. The nine patients (including the two hypoxemic patients with negative MAA lung perfusion scan) who had saline microbubble injection which was clearly shown in the right heart did not have detectable intrapulmonary shunting.

Gas Exchange in OLT Candidates 805 Lung function test. Reduced DLco of less than 75% of the predicted value was found in 29/57 (51%) patients. In 3 patients, reduction in DLco was not accompanied with a similar reduction in KCO. Air flow obstruction (defined as the ratio of forced expiratory volume in 1second to the forced vital capacity less than 70% of predicted value associated with an elevated ratio of residual volume to the total lung capacity) was noted in 7/57 (12%) patients. Ventilatory restriction (defined as reduction in total lung capacity less than 80% of predicted value and an increase of forced expiratory volume in 1second/forced vital capacity) was noted in only 5/57 (9%) patients. DLco correlated poorly with severity of liver damage assessed by the Child-Pugh score and PaO 2 level at rest (r 0.07 and 0.15, respectively). However, DLco was significantly reduced in patients with ascites (P.041). Dlco tended to be lower in those who were dyspnoeic but this did not reach statistical significance (P.058). Six current and two previous smokers were amongst those with reduced transfer factor. However, there was no significant difference in DLco between those with and without a postive smoking history (P.42). Of the 29 patients, 14 (48%) with diffusion impairment had widened PaO 2 gradient. However, alveolar-arterial oxygen tension difference correlated poorly with the transfer factor (r 0.27, P.072). Obstructive lung function test (reduced forced expiratory volume in 1second /forced vital capacity ratio) was significantly associated with a positive smoking history (P.02). A reduced DLco was observed in all patients with a restrictive defect. The presence of ascites was not associated with ventilatory restriction (P.51). Primary Biliary Cirrhosis In this study, 24 patients were transplanted for primary biliary cirrhosis. Reduced DL co of less than 75% of the predicted value was noted in 18 of 24 (75%) patients with primary biliary cirrhosis: DL co is significantly lower in patients with primary biliary cirrhosis (P.022). In 2 patients, reduction in transfer factor was associated with restrictive ventilatory defect. Diffusion impairment was accompanied by elevated PaO 2 gradient in 10 out of 18 (56%) patients. Liver Transplantation All the patients in this study underwent transplantation while they were on the elective waiting list. Six patients were on the liver transplantation waiting list for longer than 6 months: the Child-Pugh score at the time of assessment and liver transplantation was the same in these six patients. Table 3. The Median Values and Range of Pulmonary Hemodynamics and Measurements of Lung Function Median Minimum Maximum PO 2 (mm Hg) 12.6 8.3 15.2 PCO 2 (mm Hg) 4.2 2.7 5.1 RV/TLC 34 11 53 FEV 1 /FVC 77 49 89 DL CO 69 36 135 MPAP 18.5 10 32 PCWP 13.5 1 23 PVR 59 8 224 Abbreviations: RV/TLC, ratio of residual lung volume to the total lung capacity; FEV 1 /FVC, ratio of forced expiratory volume in 1second to forced vital capacity; DL CO, transfer factor for carbon monoxide corrected for hemoglobin (% predicted value); MPAP, mean pulmonary artery pressure (mm Hg); PVR, pulmonary vascular resistance (dynes s/cm 5 ); PCWP, pulmonary capillary wedge pressure (mm Hg). Pulmonary hemodynamics. The median values and range of pulmonary hemodynamics and measurements of lung function are shown in Table 3. Mean pulmonary artery pressure of less than 30 mm Hg was detected in 2 patients who both had hypoxemia. However, only one patient had true pulmonary hypertension with an associated elevated pulmonary vascular resistance of 224 dynes/s/cm 5 (Table 2). The patient with pulmonary hypertension did not have respiratory symptoms, electrocardiographic evidence of right ventricular hypertrophy, or prominent pulmonary arteries on chest radiograph before the transplant. Intraoperatively, the patient did not have hemodynamic instability. Postoperative course. None of the patients with hypoxemia had any ventilatory problem intraoperatively, but in the immediate postoperative period, one patient needed prolonged mechanical ventilation and had difficulty in maintaining the arterial PaO 2. Despite good liver function, he continued to have low arterial PaO 2 and developed seizures 15 days after liver transplant. CT scan revealed left midbrain hemorrhage with surrounding edema and left parietal infarct, and he died on day 24 after transplant. One patient who developed unexplained acute lung injury/adult respiratory distress syndrome after liver transplantation succumbed to it 31 days later. The patient with the greatest degree of hypoxemia before OLT had normal PaO 2 (13 kpa or 97.5 mm Hg) 6 months after transplant. The other patient with preoperative hypoxemia who was alive at 6 months after

806 Mohamed et al Table 4. Clinical Characteristics of Patients Who Died After Liver Transplant (n 6) Patient Age (yr) Gender PaO 2 Pre-OLT (kpa/mm Hg) Cause of Death Time of Death (After OLT) 1 63 F 14.5 / 109 Multiple organ failure Day 27 2 49 F 11.5 / 86 Acute lung injury/adult respiratory Day 31 distress syndrome 3 37 M 15.1 / 113 Intraperitoneal bleed Day 12 4 69 M 9.2 / 69 Intracerebral bleed Day 24 5 52 F 13.1 / 98 Massive haemorrhaghic necrosis of Day 10 the liver 6 50 F 9.1/ 68 Multiple organ failure Day 147 liver transplant had an improvement in PaO 2 (from 9.9kPa/74 mm Hg to 11.1kPa/83 mm Hg). Mortality The interval between hospitalization for assessment and liver transplantation ranged from less than 1week to 47 weeks (median, 7). None of the patients died awaiting liver transplant. Six patients died after liver transplant. Five deaths occurred during the transplant admission, whereas one patient died 4 months after transplant. The causes of death are listed in Table 4. Although 2 patients with moderate hypoxemia were among those who died, their deaths were not directly attributable to abnormal arterial oxygenation. Furthermore, there was no significance difference in the pretransplant PaO 2 (P.35) between survivors and nonsurvivors. Discussion This study shows that diffusion limitation is a frequent finding in liver disease, which is in agreement with data available in the literature. 1,2 However, although arterial hypoxemia was not common in our group of patients with chronic liver disease referred for liver transplantation, a widened PaO 2 gradient was present in about one third of our patients, which is consistent with the reported incidence of abnormal arterial oxygenation in liver disease. 6 Alveolar-arterial oxygen tension difference is a better indicator of arterial oxygenation abnormality, because characteristic hyperventilation, which was found in a large number of patients, may alter the PaO 2. However, from the practical standpoint of assessing candidacy for liver transplant, minor abnormalities in arterial oxygenation do not usually affect the perioperative course. Arterial blood sampling was taken with the patient standing because this position is most likely to detect hypoxemia as hepatopulmonary syndrome, because PaO 2 usually worsens on standing (orthodeoxia) in those with IPVD. We did not formally test for orthodeoxia (the arterial blood samples taken with the patient supine and standing) because this would involve two arterial punctures or insertion of an indwelling arterial catheter and would provide very little additional useful information. Furthermore, even in the presence of orthodeoxia either MAA lung scan or contrast-enhanced echocardiogram would be needed to make a diagnosis of hepatopulmonary syndrome. In patients with arterial hypoxemia, we were unable to confirm the presence of IPVD using either MAA lung scan or contrast-enhanced transthoracic echocardiography, although evidence of IPVD was found on both methods in some patients with liver disease not involved in this study. IPVD has been shown to occur despite negative lung scan. 15,16 It is possible that IPVD does exist as MAA lung scan is specific in subjects with a PaO 2 of less than 60 mm Hg 17 : the pretransplant PaO 2 in patients in this study cohort is above 60 mm Hg. In patients with cirrhosis, transesophageal echocardiography appears to be more sensitive than transthoracic echocardiography in the detection of IPVD. 18 The cause of hypoxemia in the four patients is unknown: it is difficult to differentiate intrinsic lung disease from pulmonary complications of cirrhosis. Smoking is the most likely contributing factor in the 3 patients with obstructive ventilatory pattern on the pulmonary function test. In addition, reduced transfer factor for carbon monoxide is noted in these 3 patients. The patient who was a nonsmoker underwent both MAA lung perfusion scan and saline-enhanced echocardiography, which elicited negative results for IPVD. Nevertheless, hepatopulmonary syndrome cannot be ruled out in these four patients. The patients in our study were referred for liver transplantation and, therefore, were selected but are not different from other transplant centers in the United

Gas Exchange in OLT Candidates 807 Kingdom in the selection of patients. 19 A large percentage of patients with primary biliary cirrhosis were evaluated at the time of this study which may not be reflective of the cases assessed at most transplant centers currently. Most patients with primary biliary cirrhosis in the study cohort have abnormal carbon monoxide transfer and were accompanied by widened alveolararterial oxygen tension difference in the majority of cases. Whether these abnormalities are related to hepatopulmonary syndrome needs further evaluation. To date, published reports on the effect of liver transplantation on hepatopulmonary syndrome suggest a trend towards improvement of arterial oxygenation after hepatic replacement. 20-22 An improvement in PaO 2 was noted in the two patients with pretransplant hypoxemia who survived longer than 6 months. These patients may have had hepatopulmonary syndrome that resolved after liver transplant. Measurement of pulmonary hemodynamic parameters was made at the time of liver transplantation when the pulmonary artery catheter is routinely inserted, because we believed that the patients in the study did not have findings that warranted invasive evaluation at the time of liver transplant assessment. Although the pulmonary hemodynamic parameters was measured at a temporally distant time from the baseline evaluation, none of the patients in this study was either evaluated emergently or received a transplant imminently. Measurement of pulmonary hemodynamics after induction of general anaesthesia may be problematic as many factors may affect the hemodynamic parameters. No adverse hemodynamic effect was seen in the patient with pulmonary hypertension during liver replacement. Liver transplantation alone has been performed successfully in patients with moderate pulmonary hypertension, obviating the need for combined heart-lung and liver transplantation, and reversibility has been reported after OLT. 23-25 In the pretransplant assessment of the liver transplant candidate, the focus lies on the detection of significant pulmonary dysfunction, which may contribute to the morbidity and mortality in the perioperative period. Routine preoperative evaluation should include chest radiography, standing arterial blood gas and Doppler echocardiography. Abnormal arterial oxygenation warrants further study by assessing response to breathing 100% oxygen on standing. Severely hypoxemic patients with at least a moderate response to breathing 100% oxygen on standing (PaO 2 20 kpa or 150 mm Hg) are thought to have an adequate pulmonary reserve and may be safely oxygenated intraoperatively. 26 Patients suspected of having pulmonary hypertension may need cardiac catheterization. Mean pulmonary artery pressure greater than 40 mm Hg and pulmonary vascular resistance greater than 250 dynes/ sec/cm 5 or poor myocardial function is associated with a high risk for liver transplantation. Acknowledgments The authors thank their clinical colleagues, particularly Prof. P. McMaster, A. Mayer, J. Buckels, Prof. E. Elias, Prof. D. Adams, Dr. D. Mutimer, and the medical and nursing staff of the Liver Unit for their help and Dr. Susan L Hill for reviewing the lung function tests. The authors thank Dr. M.J. Krowka for his expert advice. References 1. Hourani JM, Bellamy PE, Tashkin DP, Batra P, Simmons MS. Pulmonary dysfunction in advanced liver disease: Frequent occurrence of an abnormal diffusing capacity. Am J Med 1991; 90:693-700. 2. Martinez GP, Barbera JA, Visa J, Rimola A, Pare JC, Roca J, et al. Hepatopulmonary syndrome in candidates for liver transplantation. J Hepatol 2001;34:651-657. 3. Rodriguez-Roisin R, Agusti AGN, Roca J. The hepatopulmonary syndrome: New name, old complexities. Thorax 1992;47: 897-902. 4. Hobeika J, Houssin D, Bernard O, Devictor D, Grimon G, Chapuis Y. Orthoptic liver transplantation in children with chronic liver disease and severe hypoxemia. Transplantation 1994;57:224-228. 5. Krowka MJ, Cortese DA. Hepatopulmonary syndrome: An evolving perpective in the era of liver transplantation. Hepatology 1990;11:138-142. 6. Sherlock S. Disorders of the Liver and Biliary System (ed 8). Oxford: Blackwell; 1989:82-85. 7. Fahy JV, Kerr KM, Lake JR, Gold WM. Pulmonary function before and after liver transplantation. Am Rev Respir Dis 1992; 143:A303. 8. Krowka MJ, Dickson ER, Wiesner RH, Krom RAF, Atkinson B, Cortese DA. A prospective study of pulmonary function and gas exchange following liver transplantation. Chest 1992;102:1161-1166. 9. Hopkins WE, Waggoner AD, Barzilai B. Frequency and significance of intrapulmonary right-to-left shunting in end-stage hepatic disease. Am J Cardiol 1992;70:516-519. 10. McDonnell PJ, Toye PA, Hutchins GM. Primary pulmonary hypertension and cirrhosis: are they related? Am Rev Respir Dis 1983;127:437-441. 11. Cheng EY, Woehlek HJ. Pulmonary artery hypertension complicating anesthesia for liver transplantation. Anesthesiology 1992;77:389-392. 12. Murray JF. Gas exchange and oxygen transport. In: The Normal Lung: The Basis for Diagnosis and Treatment of Pulmonary Disease (ed 2). Philadelphia: Saunders; 1986:183-210. 13. Guidelines for the measurement of respiratory function. Recommendations of the British Thoracic Society and the Association of Respiratory Technicians and Physiologists. Respiratory Medicine 1994;88:165-194. 14. Quanjer PhH (Ed). Standardization of Lung Function Tests:

808 Mohamed et al 1993 Update. Report Working Party for the European Community for Steel and Coal. Eur Respir J 1993;6:Suppl.16. 15. Mimidis KP, Vassilakos PI, Mastorakou AN, Spiropoulos KV, Lambropoulou-Karatza CA, Thomopoulos KC, et al. Evaluation of contrast echocardiography and lung perfusion scan in detecting intrapulmonary vascular dilatation in normoxemic patients with early liver cirrhosis. Hepatogastroenterology 1998;45: 2303-2307. 16. Abrams GA, Jaffe CC, Hoffer PB, Binder HJ, Fallon MB. Diagnostic utility of contrast echocardiography and lung perfusion scan in patients with hepatopulmonary syndrome. Gastroenterology 1995;109:1283-1288. 17. Abrams GA, Nanda NC, Dubovsky EV, Krowka MJ, Fallon MB. Use of macroaggregated albumin lung perfusion scan to diagnose hepatopulmonary syndrome: a new approach. Gastroenterology 1998;114:305-310. 18. Aller R, Moya JL, Moreira V, Boixeda D, Cano A, Picher J, et al. Diagnosis of hepatopulmonary syndrome with contrast transesophageal echocardiography: Advantages over contrast transthoracic echocardiography. Dig Dis Sci 1999;44:1243-1248. 19. Hartley P, Petruckevitch A, Reeves B, Rolles K. The National Liver Transplantation audit: An overview of patients presenting for liver transplantation from 1994 to 1998. On behalf of the Steering Group of the UK Liver Transplantation Audit. Br J Surg 2001;88:52-58. 20. Stoller JK, Moodie D, Schiavone WA, et al. Reduction of intrapulmonary shunt and resolution of digital clubbing associated with primary biliary cirrhosis after liver transplantation. Hepatology 1990;11:54-58. 21. Eriksson LS, Soderman C, Ericzon BG, Eleborg L, Wahren J, Hedenstierna G. Normalization of ventilation/perfusion relationships after liver transplantation in patients with decompensated cirrhosis: Evidence for a hepatopulmonary syndrome Hepatology 1990;12:1350-1357. 22. Laberge JM, Brandt ML, LeBecque P, Moulin D, Veykemans F, Paradis K, et al. Reversal of cirrhosis-related pulmonary shunting in two children by orthotopic liver transplantation. Transplantation 1992;53:1135-1138. 23. Plevak D, Krowka M, Rettke S, Dunn W, Southorn P. Successful liver transplantation in patients with mild to moderate pulmonary hypertension. Transplant Proc 1993;25:1840. 24. Scott V, De Wolf A, Kang Y, Martin M, Selby R, Fung J, et al. Reversibility of pulmonary hypertension after liver transplantation: A case report. Transplant Proc 1993:25:1789-1790. 25. Koneru B, Ahmed S, Weisse AB, Grant GP, McKim KA. Resolution of pulmonary hypertension of cirrhosis after liver transplantation. Transplantation 1994;58:1133-1135. 26. Poterucha JJ, Krowka MJ, Dickson ER, Cortese DA, Stanson AW, Krom RAF. Failure of hepatopulmonary syndrome to resolve after liver transplantation and successful treatment with embolotherapy. Hepatology 1995;21:96-100.