Outcomes of liver transplantation for hepatopulmonary syndrome in patients with concomitant respiratory disease.

BACKGROUND & AIMS: Concomitant respiratory disease is a common finding in patients with hepatopulmonary syndrome (HPS). Among patients who underwent liver transplantation (LT) for HPS, we compared characteristics and outcome of patients with versus without concomitant respiratory disease. METHODS: This single center retrospective observational study included patients with HPS who underwent LT between 1999 and 2020. RESULTS: During the study period, 32 patients with HPS received a LT; nine (28%) with concomitant respiratory disease of whom one required a combined lung-liver transplantation. Patients with concomitant respiratory disease had higher PaCO2 (38 vs. 33 mm Hg, p = .031). The 30-day postoperative mortality was comparable, but the estimated cumulative probability of resolution of oxygen therapy after LT in HPS patients with versus those without concomitant respiratory disease was lower: 63% versus 91% at 12 months and 63% versus 100% at 18 months (HR 95% CI .140-.995, p = .040). In addition to the presence of concomitant respiratory disease (p = .040), history of smoking (p = .012), and high baseline 99mTcMAA shunt fraction (≥20%) (p = .050) were significantly associated with persistent need of oxygen therapy. The 5-year estimated cumulative probability of mortality in patients with concomitant respiratory disease was worse: 50% versus 23% (HR 95% CI .416-6.867, p = .463). CONCLUSIONS: The presence of a concomitant respiratory disease did not increase the short-term postoperative mortality after LT in patients with HPS. However, it resulted in a longer need for oxygen therapy.

Endovascular closure of a congenital extrahepatic portosystemic shunt for the treatment of hepatopulmonary syndrome in an infant.

Congenital portosystemic shunts may result in the development of hepatopulmonary syndrome, typically presenting with progressive hypoxemia in later childhood. We describe a case of a 5-month-old male with heterotaxy with polysplenia presenting with new onset hypoxemia. Subsequent evaluation identified an extrahepatic portosystemic shunt arising from the confluence of the main portal and superior mesenteric veins draining into the left renal vein. To treat his hypoxemia and prevent future complications of shunting, the patient underwent a successful single-stage endovascular closure.

Perioperative Extracorporeal Membrane Oxygenation Support for Acute Respiratory Distress Syndrome Aggravated by Hepatopulmonary Syndrome in Deceased Donor Liver Transplantation: A Case Report.

Background: Extracorporeal membrane oxygenation (ECMO) is an accommodation of the cardiopulmonary bypass technique that can support gas exchange and hemodynamic stability. It is used as a salvage maneuver in patients with life-threatening respiratory or cardiac failure that does not respond to conventional treatment. There are few case reports of successful perioperative use of ECMO, especially preoperatively, in liver transplantation (LT). Here, we report an experience of successful anesthetic management in deceased donor liver transplantation (DDLT) by applying perioperative veno-venous (VV) ECMO support in the setting of acute respiratory distress syndrome (ARDS) aggravated by hepatopulmonary syndrome (HPS). Case: A 25-year-old female (156.0 cm, 65.0 kg), without any underlying disease, was referred to our emergency department for decreased mentality. Based on imaging and laboratory tests, she was diagnosed with acute liver failure of unknown cause combined with severe ARDS aggravated by HPS. Since the patient faced life-threatening hypoxemia with a failure of conventional ventilation maneuvers, preoperative VV ECMO was initiated and maintained during the operation. The patient remained hemodynamically stable throughout DDLT, and ARDS showed gradual improvement after the administration of VV ECMO. As ARDS improved, the patient's condition alleviated, and VV ECMO was weaned on postoperative day 6. Conclusions: This case demonstrates that VV ECMO may be a useful therapeutic option not only during the intraoperative and postoperative periods but also in the preoperative period for patients with liver failure combined with reversible respiratory failure.

Model for End-Stage Liver Disease/Pediatric End-Stage Liver Disease exception policy and outcomes in pediatric patients with hepatopulmonary syndrome requiring liver transplantation.

Hepatopulmonary syndrome (HPS) is associated with increased waitlist mortality in liver transplantation (LT) candidates. Children with HPS are granted Model for End-Stage Liver Disease (MELD)/Pediatric End-Stage Liver Disease (PELD) exception points for waitlist prioritization in the United States based on criterion developed for adults. In this study, the impact of this MELD/PELD exception policy on post-LT survival in children was examined. A retrospective cohort of patients aged younger than 18 years with a MELD/PELD exception request who underwent LT between 2007 and 2018 were identified in the Scientific Registry of Transplant Recipients. Patients were stratified by waitlist partial pressure of arterial oxygen (PaO 2 ) to assess risk factors for waitlist mortality and post-LT survival. Among 3082 pediatric LT recipients included in the study, 124 patients (4%) received MELD/PELD exception points for HPS. Patients with HPS were a median age of 9 years (interquartile range: 6, 12 years), 54.8% were girls, and 54% were White. Most patients (87.9%) were listed with laboratory MELD/PELD scores <15. Waitlist mortality for patients with HPS exception points was rare and not different from patients without HPS. When stratified by pre-LT PaO 2 , hypoxemia severity was not associated with differences in 1-, 3-, or 5-year survival rates after LT ( p = 0.13). However, patients with HPS showed a slightly lower survival rate at 5 years compared with patients without HPS (88.7% vs. 93.4%; p = 0.04). MELD/PELD exceptions for children with HPS mitigated waitlist mortality, and recipients with HPS experienced excellent 5-year survival after LT, although slightly lower than in patients without HPS. Unlike adults with HPS, the severity of pre-LT hypoxemia in children does not impact post-LT survival. These data suggest that adult criteria for granting MELD/PELD exception points may not appropriately capture HPS severity in pediatric patients. Further prospective multicenter studies to examine the risk factors predicting negative survival outcomes in children with HPS are warranted.

Extracorporeal membrane oxygenation in patients with hepatopulmonary syndrome undergoing liver transplantation: A systematic review of the literature.

BACKGROUND: Isolated cases of extracorporeal membrane oxygenation (ECMO) in patients with hepatopulmonary syndrome (HPS) undergoing liver transplantation (LT) have been reported with increasing frequency. We aimed to systemically review and synthesize the available literature on ECMO use in this population. METHODS: A systematic literature review of the PubMed, Web of Science, Cochrane Library, and Embase databases (end-of-search date: November 14, 2021) was conducted in accordance with the PRISMA statement. Eligible studies presented clinical parameters and outcomes of adult or pediatric patients with HPS receiving ECMO support at the time of, or following, LT. RESULTS: Sixteen studies from 4 continents reporting on 17 patients who were initiated on ECMO prior to (n = 2), during (n = 1) or after LT (n = 14) were included. Nine of the 16 studies were published between 2019 and 2021. The median pre-LT PaO2 was 38.0 mmHg (IQR 35.0-52.0). The median time from LT to ECMO initiation was 7 days (IQR, 3-12). Six patients (50%, n = 6 of 12) were extubated post-LT, before deterioration, development of refractory hypoxemia, and initiation of ECMO. Most patients were cannulated with a venovenous configuration (75%, n = 12 of 16). Most patients cannulated with a venoarterial or veno-arterial-venous strategy (75%, n = 3 of 4) had concurrent hemodynamic instability. The median total time on ECMO was 13 days (IQR 10-29). Using linear regression, for patients cannulated postoperatively, each day between LT and ECMO initiation was associated with a 3.5-day increase in total ECMO duration (95%CI: 2.23-4.73, p < 0.001, R2 = 73.7%). The median postoperative intensive care unit length of stay was 40 days (IQR, 37-61) and hospital length of stay was 59.5 days (IQR 42-77). 82.4% of patients (14 of 17) survived to discharge. CONCLUSIONS: ECMO is feasible in patients with HPS undergoing LT and appears to be associated with better outcomes compared to other causes of cardiopulmonary failure in LT patients. As the volume of experience grows, ECMO may become a central part of perioperative support in LT patients with severe HPS.

ICU and Hospital Outcomes in Patients with Hepatopulmonary Syndrome Undergoing Liver Transplantation.

PURPOSE: There are limited data regarding hospital and intensive care unit (ICU) outcomes in patients with hepatopulmonary syndrome (HPS) following liver transplantation (LT). METHODS: Data were retrospectively collected from consecutive HPS adult patients who underwent LT and were immediately admitted to the ICU at three transplant centers with shared management protocols, from 2002 to 2018. Demographic, clinical, surgical, laboratory, and outcome data were extracted. RESULTS: We identified 137 patients (74 male, 54%), with a median age at LT of 58 years (IQR: 52-63). One hundred and 31 (95.6%) patients were admitted to the ICU on invasive mechanical ventilation (MV). The median time on invasive MV in the ICU was 12 hours (IQR: 5-28) and 97 patients (74%) were extubated within 24 hours of ICU admission. The median highest positive end expiratory pressure and fraction of inspired oxygen (FiO2) were 7 (IQR: 5-8) and 0.6 (IQR: 0.5-0.7), respectively. 7 patients (5%) developed severe post-transplant hypoxemia. Of all patients, 42 (30.4%) required vasopressors and the median ICU and hospital length of stay (LOS) were 3 (IQR: 1-5) and 10 (IQR: 7-20) days, respectively. The in-hospital mortality rate was 3.6% (5/137). HPS severity was not associated with hospital mortality. CONCLUSION: Most HPS patients have short durations of MV, ICU, and hospital LOS post-LT. HPS severity does not impact hospital mortality.

Extracorporeal membrane oxygenation as rescue therapy in a pediatric liver transplant recipient with very severe hepatopulmonary syndrome.

BACKGROUND: In children with cirrhosis, the prevalence of HPS ranges from 3% to 20%, resulting in impaired gas exchange due to alterations in pulmonary microvasculature. LT is the gold-standard cure for cirrhosis complicated by HPS and should ideally be performed prior to the development of severe HPS due to increased risk for post-transplant hypoxia, right heart failure, and outflow obstruction. METHODS: We present a case of a 13-year-old man, who underwent pediatric LT for severe HPS complicated by postoperative respiratory collapse, requiring a 92-day course of veno-venous ECMO. RESULTS: Post-transplant, despite BiPAP, inhaled nitric oxide and isoproterenol infusion, he remained hypoxic postoperatively and acutely decompensated on postoperative day 25, requiring veno-venous ECMO. After 84 days on ECMO, a persistent large splenorenal shunt was identified that was embolized by interventional radiology, and 8 days after shunt embolization and ASD closure, he was successfully weaned off ECMO. CONCLUSIONS: This case describes the longest known duration of ECMO in a pediatric LT recipient and a unique improvement in hypoxemia following a portosystemic shunt closure. ECMO presents a heroic rescue measure for pediatric LT recipients with HPS that develops acute respiratory failure postoperatively refractory to alternative measures.

Predictors of outcomes following liver transplant in hepatopulmonary syndrome: An OPTN database analysis.

Hepatopulmonary syndrome (HPS) is a type of pulmonary vascular disease occurring exclusively in those with underlying liver disease, associated with significant mortality in patients awaiting liver transplantation (LT). LT is curative in HPS, and these patients are granted Model for End Stage Liver Disease (MELD) exception points to expedite LT. The purpose of this study is to use multivariable competing risk Accelerated Failure Time models and propensity matching to examine the relationship between pre-LT hypoxemia and post-LT outcomes in HPS. We performed a retrospective cohort study of UNOS/OPTN database of all adult patients undergoing LT between January 1, 2006 and January 12, 2020. Pre-LT PaO2 was significantly associated with post-LT mortality in HPS, with each 1 mmHg increase in PaO2 significantly decreasing the risk of post-LT mortality (coefficient 0.039, HR = 0.95, p = 0.001). HPS patients with a pre-LT PaO2 < 54 mmHg demonstrated increased mortality following LT as compared to matched non-HPS cirrhotic patients. We conclude that HPS patients with a PaO2, 54 mmHg are at increased risk of post-LT mortality and may identify high-risk patients who would benefit from additional resources during LT, and that the effects of HPS MELD exception points to optimize post-LT outcomes should be continuously re-evaluated.

[Hepatopulmonary syndrome: a rare manifestation of cirrhosis in patient with diencephalic obesity and nonalcoholic fatty liver disease after surgery for craniopharyngioma].

We describe a 15-year girl, who developed panhypopituitarism and diencephalic obesity after surgical excision of craniopharyngioma, followed by nonalcoholic fatty liver disease and cirrhosis 5 years after surgery. Cirrhosis in this case manifested by hypoxia due to hepatopulmonary syndrome, and despite cure of craniopharyngioma by surgery and radiosurgery treatment and adequate hormonal substitution therapy patient died 9 years after surgery. Growth hormone substitutional therapy in patients with hypopituitarism, and steatohepatitis may decrease liver triglyceride accumulation and prevent end-stage liver disease.

Anesthetic Management Using Low Fraction of Inspiratory Oxygen for Living Donor Liver Transplantation in a Patient With Hepatopulmonary Syndrome Complicated by Interstitial Pneumonia: A Case Report.

BACKGROUND: Hepatopulmonary syndrome frequently complicates end-stage liver disease. It causes hypoxemia and requires oxygen administration. Additionally, interstitial pneumonia causes hypoxemia; however, it is known to be aggravated by high-concentration oxygen administration. CASE PRESENTATION: A 71-year-old woman with hepatopulmonary syndrome and interstitial pneumonia underwent living donor liver transplantation, requiring conflicting management in terms of the inspiratory oxygen concentration. We achieved a low intraoperative fraction of inspiratory oxygen by increasing the cardiac output with intravenous catecholamines. As a result, the transplanted liver functioned well postoperatively, and the patient was discharged without exacerbation of the interstitial pneumonia. CONCLUSION: We suggest that patients with hepatopulmonary syndrome complicated with interstitial pneumonia can undergo successful living donor liver transplantation without the use of high inspiratory oxygen concentration by using catecholamines to maintain a high mixed venous oxygen saturation.

[Liver transplantation for hepatopulmonary syndrome complicating cirrhosis]. / Transplantation hépatique pour syndrome hépatopulmonaire compliquant une cirrhose.

Hepato-pulmonary syndrome (HPS) is a pulmonary vascular complication of cirrhosis quite frequent but often under-diagnosed, and characterized by intra-pulmonary capillary and pre-capillary vascular dilatations that may lead to severe hypoxemia. HPS is often asymptomatic but may induce a progressive dyspnea. HPS diagnosis is based on arterial gasometry that proves the hypoxemia and contrast-enhanced echo-cardiography revealing the vascular dilatations. Screening of HPS is recommended in every cirrhotic patient complaining of dyspnea or in every liver transplantation candidate. Indeed, the only effective treatment of HPS is liver transplantation; HPS patients receive exception-points in the MELD (Model for End-Stage Liver Disease) liver allocation score. The authors report herein the case of a 39-year-old male patient with a cirrhosis of unknown origin complicated by HPS which appeared as a disabling dyspnea. This patient underwent liver transplantation a year after HPS diagnosis and recovered completely.

Le syndrome hépatopulmonaire est une complication vasculaire pulmonaire de la cirrhose relativement fréquente et sous-diagnostiquée, caractérisée par des vasodilatations capillaires et pré-capillaires intrapulmonaires pouvant entraîner une hypoxémie sévère. Souvent asymptomatique, ce syndrome se révèle le plus souvent par une dyspnée d'apparition progressive. Le diagnostic est réalisé par une gazométrie artérielle prouvant l'hypoxémie et une échographie cardiaque de contraste démontrant l'existence de vasodilatations intrapulmonaires. Le dépistage du syndrome hépatopulmonaire est préconisé chez tout patient atteint de cirrhose présentant de la dyspnée et chez tout patient candidat à une greffe hépatique. En effet, le seul traitement efficace est la transplantation hépatique, et ces patients bénéficient d'ailleurs de points d'exception dans le calcul du score de MELD («Model for End-Stage Liver Disease¼). Nous rapportons ici le cas d'un patient de 39 ans atteint d'une cirrhose d'origine indéterminée compliquée d'un syndrome hépatopulmonaire qui s'est révélé par une dyspnée devenue rapidement invalidante. Ce patient a pu bénéficier d'une transplantation hépatique un an après le diagnostic de syndrome hépatopulmonaire, permettant ainsi une guérison complète tant sur plan hépatique que pulmonaire.

Improved survival in children with HPS: Experience from two high volume liver transplant centers across continents.

BACKGROUND: Severe HPS increases morbidity and mortality after LT in children. We reviewed the combined experience of LT for HPS in children from two LT centers in Europe and Asia. METHODS: All children with "proven" HPS as per ERS Task Force criteria (detailed in manuscript) who underwent LT were categorized into M (PaO2 ≥80 mmHg), Mo (PaO2  = 60-79 mmHg), S (50-59 mmHg), and VS (PaO2 <50 mmHg) HPS, based on room air PaO2 . RESULTS: Twenty-four children with HPS underwent 25 LT (one re-transplantation) at a median age of 8 years (IQR, 5-12), after a median duration of 8 (4-12) months following HPS diagnosis. Mechanical ventilation was required for a median of 3 (1.5-27) days after LT. Ten children had "S" post-operative hypoxemia, requiring iNO for a median of 5 (6-27) days. "VS" category patients had significantly prolonged invasive ventilation (median 35 vs. 3 and 1.5 days; p = .008), ICU stay (median 39 vs. 8 and 8 days; p = .007), and hospital stay (64 vs. 26.5 and 23 days; p < .001) when compared to "S" and "M/Mo" groups, respectively. The need for pre-transplant home oxygen therapy was the only factor predicting need for re-intubation. Patient and graft survival at 32 (17-98) months were 100% and 95.8%. All children ultimately had complete resolution of HPS. CONCLUSIONS: VS HPS is associated with longer duration of mechanical ventilation and hospital stay, which emphasizes the need for early LT in these children.

Outcomes of liver transplantation in patients with hepatopulmonary syndrome in the pre and post-MELD eras: A systematic review.

BACKGROUND: The lack of large hepatopulmonary syndrome cohorts undergoing liver transplantation (LT) has resulted in limited information about post-LT outcomes and expectations. METHODS: The long and short-term outcomes of LT in patients with hepatopulmonary syndrome (HPS) were evaluated before and after the implementation of Model for Endstage Liver Disease (MELD) score in 2002, granting exception points for patients with HPS. PubMed/Medline, Embase, Web of Science and Scopus databases were searched for published and unpublished studies from 01/1990 to 04/2019. Studies that included HPS patients who underwent LT and reported post-LT outcomes and HPS severity were reviewed. After reviewing the full text of 1421 articles, 30 were included in the pre-MELD era (before 2002) and 60 in the post-MELD era. RESULTS: A total of 598 patients (210 children and 388 adults) with HPS who underwent LT were included in this systematic review. In children, 5-year survival probability was similar in the pre and post-MELD groups (85.7% vs. 97.4; p = 0.09). Median post-transplant PaO2 in room air was higher in the post-MELD group (71 [53-87] vs. 97 [80-108] mmHg: p = 0.008). In adults, 5-year survival probability was higher in the post-MELD era (73 vs. 87.3%; p = 0.008). Median post-transplant PaO2 in room air was higher in post-MELD group (75 [63-85] vs. 87 [75-95] mmHg; p = 0.001).. CONCLUSIONS: After MELD exception implementation, survival rates and post-transplant oxygenation improved in adult patients with HPS who underwent liver transplantation, whereas only post-transplant oxygenation improved in children.

Non-invasive assessment of intrapulmonary shunt and ventilation to perfusion ratio in children with hepatopulmonary syndrome before and after liver transplantation.

OBJECTIVES: To use the oxyhaemoglobin dissociation curve (ODC) to non-invasively measure the ventilation perfusion ratio (VA/Q) and right-to-left intrapulmonary vascular shunt before and after liver transplantation (LT) in children with hepatopulmonary syndrome (HPS). To investigate whether the right-to-left shunt derived by ODC correlated with the shunt derived by technetium-99 labelled macroaggregated albumin lung perfusion scan (MAA). METHODS: A retrospective cohort study at King's College Hospital NHS Foundation Trust, London, UK was performed between 1998 and 2016. The VA/Q and right-to-left shunt were non-invasively measured pre and post LT. The pre-LT right-to-left intrapulmonary shunt was also measured by MAA. The non-invasively derived pre-LT shunt was correlated with the shunt derived by MAA. RESULTS: Fifteen children with HPS were studied with a median (IQR) age at LT of 8.8 (6.6-12.9) years. The median (IQR) pre-LT VA/Q [0.49 (0.42-0.65)] was lower compared to the post-LT VA/Q [0.61 (IQR 0.54-0.72), p = 0.012]. The median (IQR) pre-LT shunt was 19 (3-24) % which decreased to zero in all but one children post-LT, (p = 0.001). The MAA-derived shunt was significantly positively correlated with the ODC-derived shunt (r = 0.783, p = 0.001). The mean (SD) difference between shunt derived by ODC and shunt derived by MAA was 0.5 (7.2) %. CONCLUSIONS: Ventilation/perfusion impairment reverses but not completely resolves after liver transplantation in children with hepatopulmonary syndrome. The non-invasive method for estimating intrapulmonary shunting could be used as an alternative to the macroaggregated albumin scan in this population.

Impact of hypoxemia on pediatric liver transplantation for hepatopulmonary syndrome.

BACKGROUND: The treatment of choice for patients with cirrhosis and HPS is LT. The clinical manifestations associated with hypoxemia result in limitations and a poor health-related quality of life of affected patients. The present report aims to study the differences in outcomes between patients with PaO2  < 50 mm Hg and those with PaO2  ≥ 50 mm Hg. METHODS: This was a retrospective study of 21 patients under 18 years of age conducted from 2001 to 2018; the patients were divided into 2 groups: G1-PaO2  ≥ 50 mm Hg, 11 patients, and G2-PaO2  < 50 mm Hg, 10 patients. Demographic, clinical, laboratory, and perioperative data; outcome variables; and post-transplant survival were compared between the groups. RESULTS: In total, 2/11 (18.2%) patients in G1 and 8/10 (80%) patients in G2 required supplemental oxygen therapy at home (P = .005). Patients in G2 required prolonged MV (median 8.5 days in G2 vs 1 day in G1, P = .015) and prolonged ICU and hospital stays (P = .002 and P = .001, respectively). Oxygen weaning time was longer in G2 (median 127.5 days) than in G1 (median 3 days; P = .004). One (9.1%) patient in G1 and three (30%) patients in G2 died (P = .22). The survival at 90 months was 90.9% in G1 and 70% in G2 (P = .22). CONCLUSION: The survival between groups was similar. Patients with very severe HPS required a longer MV time, longer ICU and hospital stays, and a longer O2 weaning time than those with mild, moderate, or severe HPS.

Combined lung and liver transplantation for noncirrhotic portal hypertension with severe hepatopulmonary syndrome in a patient with dyskeratosis congenita.

DC is caused by defects at the level of telomere maintenance, and cells from patients with this disease have abnormally short telomeres and show premature senescence. One consequence of DC is bone marrow failure. Thus, patients with DC often require HSCT. However, HSCT does not ameliorate other DC-related manifestations. In fact, HSCT can accelerate organ dysfunction due to treatment-related complications, and solid organ transplantation is required in some patients with DC. In this report, we describe the clinical course of a 5-year-old boy who was transferred to our hospital because of progressive dyspnea, 2 years after HSCT. At admission, he had tachypnea and hypoxemia. A liver biopsy was performed for suspected HPS caused by PH, and LT was considered. Eventually, his hypoxemia worsened, and he was transferred to a PICU and started on VA ECMO. He subsequently underwent a CLLT. ECMO was stopped on post-operative day 12, extubation was achieved on post-operative day 29, and the patient recovered well from the surgery. Our results show that CLLT could be a life-saving treatment option for DC patients with very severe HPS in whom a poor outcome is expected after LT.

High-flow nasal cannula oxygen therapy: Alternative respiratory therapy for severe post-transplant hypoxemia in children with hepatopulmonary syndrome.

Severe post-transplant hypoxemia, which is defined as <50 mm Hg of the partial pressure of oxygen in arterial blood/fraction of inspired oxygen (P/F) ratio, is a major post-operative complication with high mortality rates in patients with hepatopulmonary syndrome (HPS). Non-invasive positive pressure ventilation therapy and mechanical ventilation are options for respiratory support of patients with severe post-transplant hypoxemia. However, these therapies are associated with several problems, such as compliance, ventilator-associated pneumonia, and lung injury. We here firstly described two children with HPS who developed severe post-transplant hypoxemia (lowest post-operative P/F ratio, 49.7 and 34.0 mm Hg, respectively) that was successfully managed with high-flow nasal cannula (HFNC) oxygen therapy and vasodilation drugs without adverse complications or necessity of reintubation. We consider that HFNC oxygen therapy could become a safe alternative respiratory therapy or be added to the other such as inhaled nitric oxide (iNO), methylene blue (MB), inhaled epoprostenol, embolization of abnormal pulmonary vessels, and combination of iNO and MB for severe post-transplant hypoxemia in children with HPS.