Portal Hypertension and Cirrhosis Complications: Cirrhosis Complications: Lungs and Vessels

0:00Welcome to Board Pearls. This is episode three of three of the Portal Hypertension and Cirrhosis Complications chapter, in the Liver Disease module. This episode covers the lungs, clotting, and the blood vessels of the liver. A woman whose breathlessness eases when she lies down shows hepatopulmonary syndrome, and a man with a high pulmonary pressure shows portopulmonary hypertension, two opposite vascular syndromes with opposite effects on transplant. A woman with a right-sided effusion shows hepatic hydrothorax and the one procedure never to do. A man about to get plasma before a paracentesis shows why the international normalized ratio, INR, misleads in cirrhosis, which procedures need no correction, and why his heparin should continue, and two more patients show when to anticoagulate and why the heart is checked before a shunt. Then a young woman with sudden ascites shows how vascular liver disease is managed by where the obstruction sits, from the hepatic veins to the sinusoids to the portal vein, ending with portal hypertension in a nearly normal liver.

1:03Here's the first patient. A forty-eight-year-old woman with cirrhosis from autoimmune hepatitis is breathless, worse sitting or standing and better lying down. Her room-air saturation is ninety-four percent lying and eighty-eight percent sitting. Her arterial oxygen pressure, PaO2, is fifty-six with a high alveolar-arterial gradient, and her chest film is clear. Her team has her on home oxygen and thinks that's the treatment.

1:32Hypoxemia that worsens upright, with a clear chest film, in a patient with cirrhosis suggests hepatopulmonary syndrome. Cirrhosis has two vascular syndromes in the lungs that are physiologic opposites, plus a mechanical complication, hepatic hydrothorax, and they're managed very differently. Hepatopulmonary syndrome is intrapulmonary vascular dilation causing hypoxemia. The cirrhotic liver fails to clear nitric oxide and other vasoactive mediators, which reach the lungs and dilate the pulmonary capillaries. Deoxygenated blood then crosses the dilated bed too fast to equilibrate with the alveolus, which behaves like a right-to-left shunt. Its signature is platypnea-orthodeoxia, breathlessness and hypoxemia that worsen upright and improve lying down, exactly as hers do.

2:22Her suspicion is confirmed with two tests. Contrast-enhanced echocardiography, a bubble study, shows a delayed right-to-left shunt, with bubbles appearing in the left heart three to six cardiac cycles after the right, because they have to cross the dilated lung capillaries. Bubbles within one to three cycles mean a shunt inside the heart, which isn't hepatopulmonary syndrome. Confirmation then comes from a macroaggregated albumin scan or from oxygenation criteria, a room-air PaO2 under eighty millimeters of mercury with a raised gradient. Her bubbles appear at the fifth cycle and her PaO2 is fifty-six, so she has hepatopulmonary syndrome.

3:02Her team's plan is incomplete. Supplemental oxygen is supportive but doesn't touch the shunt, and the only cure is liver transplant, because only a new liver stops the mediators. A PaO2 under sixty earns Model for End-Stage Liver Disease, MELD, exception points, so she's listed with exception points, and her oxygen continues while she waits. Chapter twenty-four covers exceptions. The same physiology is why terlipressin, used for hepatorenal syndrome, can cause respiratory harm: the pulmonary vasoconstriction it causes worsens the ventilation-perfusion mismatch in a patient with intrapulmonary shunting.

3:37The second patient has the opposite syndrome. A fifty-three-year-old man with cirrhosis has breathlessness on exertion and a high estimated pulmonary pressure on echocardiography, which suggests portopulmonary hypertension. Portopulmonary hypertension is pulmonary arterial hypertension complicating portal hypertension. The pulmonary arterioles remodel and constrict and raise resistance, the reverse of the capillary dilation in hepatopulmonary syndrome. The echo can only raise the question. Diagnosis requires right heart catheterization: a mean pulmonary artery pressure of at least twenty-five millimeters of mercury, a pulmonary vascular resistance over two hundred forty dyne, or three Wood units, and a wedge pressure of fifteen or less, which excludes left heart disease. His mean pressure is forty-two, his resistance five Wood units, and his wedge ten.

4:29His numbers decide his transplant candidacy. A mean pressure of fifty or more is a contraindication to transplant, because the right ventricle fails during the operation. A mean pressure of thirty-five to forty-five that responds to vasodilators is a relative contraindication that becomes acceptable once treatment lowers it. Mild to moderate disease responds to the pulmonary hypertension drugs: phosphodiesterase-five, PDE-five, inhibitors, endothelin receptor antagonists, and prostacyclins. He starts a PDE-five inhibitor and an endothelin receptor antagonist, his mean pressure falls to thirty-two, and he becomes eligible for evaluation. A pressure of fifty-four that didn't respond to treatment would rule transplant out.

5:12The two syndromes pull in opposite directions. Hepatopulmonary syndrome moves a patient up the list, because transplant cures it, while portopulmonary hypertension can keep a patient off it. Moderate to severe pulmonary hypertension also contraindicates a transjugular intrahepatic portosystemic shunt, TIPS, because the shunt delivers more blood to the right heart, which is why every TIPS is preceded by an echocardiogram.

5:37The third lung problem is mechanical. A sixty-one-year-old woman with cirrhosis has a large right pleural effusion and only a little ascites. That's hepatic hydrothorax until the tap says otherwise. Hepatic hydrothorax is a transudative effusion that forms through microscopic defects in the diaphragm, through which ascitic fluid is drawn into the pleural space on inspiration. More than eighty percent are right-sided, because the right hemidiaphragm has more fenestrations and less muscular reinforcement. The ascites doesn't need to be detectable, because fluid can cross directly without collecting in the abdomen first, so her small amount of ascites doesn't argue against it.

6:15A diagnostic thoracentesis comes first. The fluid mirrors ascites, with low protein and a high serum-ascites albumin gradient, SAAG. A pleural neutrophil count over two hundred fifty defines spontaneous bacterial empyema, which is treated with the same antibiotics as spontaneous bacterial peritonitis, followed by secondary prophylaxis. Her fluid has a neutrophil count of ninety, without infection.

6:40Her treatment starts the way ascites treatment does, with sodium restriction and diuretics, plus therapeutic thoracentesis for symptoms. If the effusion becomes refractory, TIPS controls it in about seventy-five percent and is preferred over chronic indwelling drainage, and a hydrothorax meeting specific criteria can earn a MELD exception. A chest tube and pleurodesis are contraindicated. The tube tract becomes infected, the pleural space is contaminated, and the resulting empyema can rule out transplant. Small-bore pigtail catheters are reserved for palliation.

7:13The next patient shows that the same liver that fails to clear mediators also changes how blood clots. A fifty-nine-year-old man with decompensated cirrhosis is admitted with tense ascites. His INR is two point two and his platelets are fifty-two thousand. The team is about to order four units of fresh frozen plasma and a platelet transfusion before his paracentesis, and plans to hold his heparin prophylaxis because of his INR.

7:38His INR doesn't mean what the team thinks. Cirrhosis produces rebalanced hemostasis. The liver makes fewer procoagulant factors, two, five, seven, nine, ten, and eleven, but also fewer anticoagulants, protein C, protein S, and antithrombin, and the fibrinolytic proteins change in parallel. The INR and prothrombin time, PT, measure only the procoagulant side, so, as the twenty twenty-one American Association for the Study of Liver Diseases, AASLD, guidance states, they don't predict bleeding risk in cirrhosis. Many patients are actually prothrombotic, which explains their raised rates of venous thromboembolism and of portal vein thrombosis. In cirrhosis the INR measures synthetic function, which is why it sits inside MELD, not bleeding risk.

8:30So his paracentesis doesn't need correction. Paracentesis, thoracentesis, and endoscopy without intervention need no prophylactic plasma or platelets, whatever the INR or platelet count. Their bleeding risk is low to begin with, and transfusion doesn't change bleeding outcomes, while it adds volume, which raises portal pressure, and adds the risk of transfusion-related lung injury. Routine plasma for an INR over one and a half isn't evidence-based. The order is stopped, and his paracentesis goes ahead without bleeding.

9:00Correction makes sense only when it's matched to the procedure's risk, not to a lab value. Viscoelastic testing, thromboelastography or rotational thromboelastometry, ROTEM, measures clot formation, propagation, and lysis together on whole blood. Before high-risk procedures, it reduces transfusion by identifying the dominant defect, whether low fibrinogen, low platelets, or hyperfibrinolysis. Fibrinogen may be replaced to over one hundred twenty milligrams per deciliter and platelets transfused to over fifty thousand, but only with documented bleeding or for the highest-risk procedures, such as neuraxial anesthesia. Thresholds for prophylactic platelets in severe thrombocytopenia run from under twenty thousand for very low-risk procedures to under fifty thousand for high-risk ones. In active variceal bleeding, the targets are a hematocrit over twenty-five percent, platelets over fifty thousand, and fibrinogen over one hundred twenty, without routine plasma.

10:00His heparin shouldn't be held either. Hospitalized patients with cirrhosis have an elevated risk of venous thromboembolism and should receive standard prophylaxis, so his heparin continues.

10:12If he needs an endoscopy, his liver changes the sedation too. Cirrhosis prolongs the effects of midazolam and fentanyl, and sedation can unmask encephalopathy, so midazolam doses are halved or benzodiazepines are avoided altogether. Propofol is generally preferred, but it's titrated carefully, because it lowers vascular resistance in a patient who's already vasodilated. Elective banding in a patient who isn't bleeding doesn't need antibiotic prophylaxis. And in a patient with cirrhosis on warfarin, the baseline INR already reflects synthetic dysfunction, so it can't be read as a measure of anticoagulation alone.

10:50Full anticoagulation is a separate decision, and the next patient shows how it's made. A sixty-three-year-old woman with Child-Pugh A cirrhosis develops atrial fibrillation that needs stroke prevention. Anticoagulating a patient with cirrhosis, whether for atrial fibrillation, venous thromboembolism, or portal vein thrombosis, weighs the prothrombotic tendency against the bleeding risk. The direct oral anticoagulants rivaroxaban and apixaban are reasonable in Child-Pugh A and some Child-Pugh B patients, but they're avoided in Child-Pugh C, in severe coagulopathy, and after a recent variceal hemorrhage. She starts apixaban, with variceal screening by the usual rules.

11:34The heart matters before any shunt. A fifty-five-year-old man being evaluated for TIPS has a prolonged QT interval and a blunted contractile response on stress testing. He has cirrhotic cardiomyopathy, defined by a prolonged corrected QT, QTc, a blunted contractile response to stress, and diastolic dysfunction. It raises his risk around surgery and after TIPS, because a heart that can't raise its output under stress may not tolerate the extra preload a shunt delivers. It's screened with an electrocardiogram, ECG, and an echocardiogram together, the ECG for the QTc and the echo for diastolic dysfunction.

12:12The last group of patients has obstruction in the liver's blood vessels themselves. A thirty-four-year-old woman has had abdominal pain, an enlarged liver, and ascites for three weeks. Her alanine aminotransferase, ALT, is two hundred twenty. She takes an oral contraceptive. Doppler shows no flow in her hepatic veins, and computed tomography, CT, shows an enlarged caudate lobe in a patchy, congested liver.

12:38She has Budd-Chiari syndrome, obstruction of hepatic venous outflow. Blood enters her liver but can't leave, which produces the pain, the enlarged liver, the ascites, and the raised aminotransferases. The caudate lobe enlarges in about seventy-five percent, because it drains directly into the inferior vena cava and is spared while the rest of the liver congests. A young patient with sudden ascites, a painful enlarged liver, and a big caudate has Budd-Chiari syndrome, not cirrhosis.

13:09Her clot needs a cause, so a thrombophilia workup is mandatory. The most common identifiable cause is a myeloproliferative neoplasm with the JAK2 V617F mutation. Others are antiphospholipid syndrome, paroxysmal nocturnal hemoglobinuria, factor V Leiden, and oral contraceptives. Her contraceptive doesn't end the search, and her JAK2 V617F is positive. She stops the contraceptive and is referred to hematology.

13:39Her treatment escalates in a fixed order, so that each step takes on more risk only when the last one fails. Anticoagulation comes first, with heparin and then a direct oral anticoagulant or warfarin to an INR of two to three, to stop the clot extending. Angioplasty with a stent treats a focal web or a short-segment stenosis directly. TIPS bypasses the blocked hepatic veins when disease is refractory. And transplant replaces the liver in fulminant or end-stage disease. She starts heparin, then a direct oral anticoagulant. Venography shows a short-segment stenosis of a hepatic vein, which is dilated and stented, her ascites resolves, and she stays on long-term anticoagulation. Diffuse obstruction with nothing to dilate, and recurrent ascites, would have meant TIPS.

14:27The obstruction can also sit inside the sinusoids. A forty-two-year-old man, two weeks after a myeloablative stem cell transplant with busulfan and cyclophosphamide, has painful hepatomegaly, a seven percent weight gain, ascites, and a bilirubin of four. He has sinusoidal obstruction syndrome, formerly called veno-occlusive disease. Injury to the sinusoidal endothelium leads to occlusion of the central venules downstream. It follows myeloablative regimens with busulfan, cyclophosphamide, or melphalan, or exposure to the pyrrolizidine alkaloids in comfrey and other herbal teas.

15:03The classical criteria are a bilirubin over two plus two of painful hepatomegaly, weight gain over five percent, or ascites, within twenty-one days of transplant, and he has all three. Defibrotide, which is antithrombotic, anti-inflammatory, and fibrinolytic and helps repair the endothelium, is approved by the Food and Drug Administration for severe disease, and he starts it. Ursodiol prophylaxis, twelve milligrams per kilogram in divided doses, reduces the incidence of the syndrome in transplant recipients. His obstruction lies beyond the reach of a balloon or stent, so his treatment targets the endothelium instead.

15:41The obstruction can sit before the liver, in the portal vein. A sixty-four-year-old woman with cirrhosis has a new portal vein thrombosis on ultrasound. Portal vein thrombosis develops in up to a quarter of patients with cirrhosis over time, as portal flow stagnates and the clotting balance tips, and it also occurs without cirrhosis, from thrombophilia or local inflammation such as pancreatitis or intra-abdominal infection. Before she's anticoagulated, hepatocellular carcinoma with tumor thrombus has to be excluded, because tumor thrombus is metastatic disease and isn't anticoagulated. A thrombus that enhances in the arterial phase is tumor, and a bland clot doesn't enhance.

16:21Her thrombus doesn't enhance and fills more than half of the main portal vein. Anticoagulation is decided by extent, meaning more than half of the main portal vein, extension into the mesenteric veins, or transplant candidacy, weighed against bleeding risk, so she starts it. Cavernous transformation is different. Collaterals that form through and around a chronically occluded vein usually don't warrant anticoagulation, because the vein can't be recanalized, and the portal hypertension is managed instead.

16:51The last patient shows portal hypertension with a nearly normal liver. Idiopathic non-cirrhotic portal hypertension, now called porto-sinusoidal vascular disease, is a group of obliterative portal venopathies that cause presinusoidal portal hypertension without cirrhosis. The clue is a mismatch. Synthetic function is preserved and liver stiffness is only mildly raised, despite clinically significant portal hypertension, because the obstruction sits upstream of the scarred tissue that elastography detects. So varices and splenomegaly are believed over a low stiffness. Biopsy with a reticulin stain shows nodular regenerative hyperplasia or obliterative portal venopathy, without bridging fibrosis.

17:36Its associations include human immunodeficiency virus, HIV, infection with past didanosine, oxaliplatin, azathioprine, hereditary hemorrhagic telangiectasia, and several genetic syndromes. Management targets the complications, with variceal screening, a non-selective beta-blocker, and banding, rather than any treatment of the liver tissue itself.

17:59One more vascular injury affects the whole liver at once. Shock liver, or ischemic hepatitis, follows hypotension or hypoxemia, with aminotransferases often over five thousand and a disproportionately high lactate dehydrogenase, LDH, that fall quickly once perfusion returns. Chapter seventeen covers how to recognize it.

18:21Here are the traps the boards set in these patients. Hepatopulmonary syndrome and portopulmonary hypertension are opposites and aren't managed the same way. Bubbles that appear early mean a shunt inside the heart, not hepatopulmonary syndrome. Oxygen doesn't cure hepatopulmonary syndrome, and a PaO2 under sixty earns exception points. Portopulmonary hypertension isn't diagnosed by echo alone, and a wedge over fifteen points to the left heart instead. A mean pressure of fifty or more rules out transplant, and significant pulmonary hypertension rules out TIPS.

18:55The hydrothorax and clotting traps come next. A small amount of ascites doesn't exclude hepatic hydrothorax, every effusion is tapped to exclude empyema, and a chest tube is never placed. Plasma before a paracentesis because of the INR, or platelets for a low-risk procedure, adds risk without benefit. Viscoelastic testing, not the INR, guides correction before high-risk procedures, and plasma isn't given routinely in variceal bleeding. A high INR isn't a reason to withhold clot prophylaxis. Direct oral anticoagulants are avoided in Child-Pugh C and after a recent variceal bleed. And cirrhotic cardiomyopathy should be looked for before TIPS or surgery.

19:39The vascular traps close the list. A young patient with sudden ascites and an enlarged caudate has Budd-Chiari syndrome until proven otherwise. A contraceptive doesn't end the thrombophilia workup, and JAK2 is always checked. Budd-Chiari syndrome is treated in order, so TIPS doesn't come before anticoagulation and angioplasty. Sinusoidal obstruction syndrome follows myeloablation or herbal teas, and ursodiol prevents it. Tumor thrombus and cavernous transformation aren't anticoagulated. And porto-sinusoidal vascular disease isn't cirrhosis, even though it causes portal hypertension, and a low stiffness doesn't exclude it.

20:21So, start to finish. Hepatopulmonary syndrome is intrapulmonary vasodilation: hypoxemia worse upright, bubbles delayed to three to six cycles, and a PaO2 under eighty, with exception points under sixty and cure only by transplant. Portopulmonary hypertension is arteriolar constriction confirmed by right heart catheterization, with a mean pressure of twenty-five or more, resistance over three Wood units, and a wedge of fifteen or less. Treat it with pulmonary hypertension drugs, and a mean pressure of fifty or more, or one that doesn't respond, rules out transplant.

20:58Hepatic hydrothorax is ascites crossing the diaphragm, usually on the right. Tap it, treat it like ascites, use TIPS if it's refractory, and never place a chest tube.

21:08Don't treat the INR in cirrhosis, because hemostasis is rebalanced and patients clot as well as bleed. Perform paracentesis, thoracentesis, and routine endoscopy without plasma or platelets, use viscoelastic testing for high-risk procedures, and give hospitalized patients clot prophylaxis. Reserve direct oral anticoagulants for Child-Pugh A and some Child-Pugh B patients, and screen for cirrhotic cardiomyopathy before TIPS and surgery.

21:35Let the level of obstruction guide vascular disease. Budd-Chiari syndrome needs a thrombophilia workup with JAK2, then anticoagulation, angioplasty, TIPS, and transplant in that order. Sinusoidal obstruction syndrome is a bilirubin over two with two of painful hepatomegaly, weight gain, or ascites within twenty-one days, treated with defibrotide and prevented with ursodiol. Exclude enhancing tumor thrombus before anticoagulating a portal vein thrombosis, and don't anticoagulate cavernous transformation. Porto-sinusoidal vascular disease causes portal hypertension with a near-normal stiffness and is managed through its complications.

22:17That completes the Portal Hypertension and Cirrhosis Complications chapter. The next chapter is liver tumors and transplant.

22:24For the full chapter, the practice vignettes, and the topic-tagged question bank, head to board pearls dot com. You'll find the rest of the series on Apple Podcasts, Spotify, or wherever you listen to podcasts. That brings us to the end of episode three of three of chapter twenty-three, and I'll see you in the next one.