Portal Hypertension and Cirrhosis Complications: Portal Hypertension: Ascites and Complications

0:00Welcome to Board Pearls. This is episode two of three of the Portal Hypertension and Cirrhosis Complications chapter, in the Liver Disease module. This episode is ascites and the complications that grow out of it. A man with new abdominal swelling shows why ascites forms and why sodium is the target, how to read the paracentesis, how the diuretics are dosed, and which drugs to stop. A woman with tense ascites shows why a large tap needs albumin, and another woman whose ascites stops responding shows who can have a transjugular intrahepatic portosystemic shunt, TIPS, and how to protect her from encephalopathy afterward. A man whose sodium falls on diuretics shows how hyponatremia is managed and how fast it can be corrected. A confused man without fever shows spontaneous bacterial peritonitis, from the count that makes the diagnosis to the albumin that protects the kidneys, and a woman with a very high count shows the peritonitis that needs a surgeon. A woman with a rising creatinine shows how hepatorenal syndrome is diagnosed and treated. And a man who sleeps all day shows hepatic encephalopathy, the search for its precipitant, and why protein restriction is wrong.

1:07Here's the first patient. A sixty-three-year-old man with alcohol-related cirrhosis, abstinent for six months, has new abdominal swelling and ankle edema. His sodium is one hundred thirty-four and his creatinine is normal. He takes ibuprofen for back pain. He's been drinking extra water because he heard it was good for his kidneys, and his doctor told him to cut back on fluids.

1:29New swelling in a man with cirrhosis is ascites until a tap says otherwise, and both pieces of advice he's been given miss how it forms. Portal hypertension drives splanchnic vasodilation through nitric oxide and other vasodilators, which leaves the rest of the circulation underfilled. The body reads that as low volume and activates the renin-angiotensin-aldosterone system and antidiuretic hormone. Sodium and water retention then exceeds what the kidney can excrete, and fluid collects in the peritoneum, where the hydrostatic pressure is lowest. Water follows sodium, so treatment targets sodium, not water.

2:05His ascites is new, so he has a diagnostic paracentesis, as every patient with new or worsening ascites does. A serum-ascites albumin gradient, SAAG, of one point one grams per deciliter or more supports portal hypertension as the cause. A SAAG under one point one points away from it, toward peritoneal carcinomatosis, tuberculous peritonitis, pancreatic ascites, or nephrotic ascites. Ascitic protein refines a high SAAG: under two and a half grams per deciliter supports cirrhosis, while a high protein suggests heart failure or constrictive pericarditis.

2:40Other tests are sent when the SAAG and history point to them. A high amylase supports pancreatic ascites, a high adenosine deaminase supports tuberculous peritonitis, cytology looks for malignancy, and the cell count looks for infection. His SAAG is one point nine, his protein one point two, and his neutrophil count normal, so he has cirrhotic ascites without infection.

3:03His treatment is sodium restriction and diuretics. Sodium is restricted to under two grams a day, and fluid is restricted only when the serum sodium falls under one hundred twenty-five, so his doctor's fluid advice was wrong, and he can simply drink normally. Spironolactone and furosemide are combined in a ratio of one hundred to forty, starting at one hundred milligrams of spironolactone with forty of furosemide, and increased together to a maximum of four hundred and one hundred sixty. Secondary hyperaldosteronism drives the sodium retention, so spironolactone is the primary drug. Furosemide adds loop diuresis and offsets the hyperkalemia and metabolic alkalosis that spironolactone causes alone. Furosemide by itself targets the wrong mechanism and drops the potassium.

3:49His ibuprofen has to stop. Nonsteroidal anti-inflammatory drugs, NSAIDs, block the prostaglandins that keep the kidneys perfused against an activated renin-angiotensin system, and aminoglycosides are directly toxic to the tubules, so both are forbidden in cirrhotic ascites, because both precipitate acute kidney injury and hepatorenal syndrome. Two weeks later he's lost four kilograms, with a normal potassium and creatinine. Had he not lost weight, his adherence to two grams of sodium would be checked and both drugs increased together.

4:23The second patient shows what a large tap does to the circulation. A fifty-nine-year-old woman with tense ascites has eight liters removed. Any paracentesis over five liters needs albumin, six to eight grams for each liter removed. Removing the fluid suddenly withdraws the intra-abdominal pressure that was supporting splanchnic venous return and unmasks effective hypovolemia, which is called post-paracentesis circulatory dysfunction, and albumin restores oncotic pressure and circulating volume. She gets fifty-six grams and stays stable. A four-liter tap wouldn't require albumin.

4:55Some ascites stops responding, and the definition decides when to stop escalating. Diuretic-resistant ascites fails to respond to maximum doses with sodium restriction. Diuretic-intractable ascites can't be treated with diuretics because they cause hyponatremia, acute kidney injury, hyperkalemia, or encephalopathy, which can happen well below maximum doses. Either way the ascites is refractory, the diuretics stop climbing, and treatment becomes serial large-volume paracentesis with albumin every two to four weeks, or TIPS in selected patients, plus a referral for transplant. In pregnancy, the diuretic plan changes too, because spironolactone is teratogenic and is replaced with amiloride, and chapter thirty-five covers pregnancy.

5:40The third patient has refractory ascites. A fifty-seven-year-old woman with cirrhosis from metabolic dysfunction-associated steatohepatitis, MASH, needs a six-liter tap every two weeks despite maximum diuretics, which now drop her sodium. Her Model for End-Stage Liver Disease, MELD, score is fourteen, her bilirubin is one point eight, and her international normalized ratio, INR, is one point four, and she's never had overt encephalopathy.

6:12TIPS creates a controlled shunt between a hepatic vein and a branch of the portal vein, and in an individual-patient meta-analysis it reduced the recurrence of ascites and improved transplant-free survival in selected patients. But the same shunt that lowers portal pressure also sends gut blood past the liver's ammonia clearance and raises the preload on the right heart, and that explains every contraindication. TIPS is ruled out by a MELD over about eighteen to twenty-two, depending on the center, a bilirubin over three to five, an INR over two point five, recurrent severe encephalopathy, congestive heart failure, and moderate to severe pulmonary hypertension. An echocardiogram before TIPS is standard, to find right ventricular dysfunction that the extra preload would unmask. She passes every criterion and her echo is normal.

6:58Her TIPS needs protection against its most common complication, encephalopathy. In a randomized trial, rifaximin started two weeks before the procedure and continued for about six months reduced encephalopathy after TIPS, by reducing the ammonia-producing bacteria in the colon. An under-dilated eight-millimeter covered stent lowers portal pressure only partly, which treats the ascites while diverting less blood past the liver. She gets both. Her ascites resolves over two months, her taps stop, and her transplant evaluation continues, because TIPS treats a complication, not the cirrhosis.

7:33She won't get routine long-term albumin. One trial of weekly outpatient albumin in refractory ascites showed a survival benefit at eighteen months, another of routine albumin in hospitalized patients showed none, and the twenty twenty-one American Association for the Study of Liver Diseases, AASLD, ascites guidance endorses routine long-term albumin in neither setting. Albumin keeps its specific indications: spontaneous bacterial peritonitis, hypovolemic hyponatremia, acute kidney injury, where it's the volume expander of choice, and large-volume paracentesis.

8:10The fourth patient shows the problem diuretics bring. A sixty-one-year-old man with cirrhosis and ascites on spironolactone and furosemide has a sodium of one hundred eighteen and mild confusion, which is severe hyponatremia. In cirrhosis, hyponatremia is usually hypervolemic and dilutional, from release of antidiuretic hormone triggered by effective hypovolemia rather than by osmolality. Sodium tracks how disturbed the circulation is and predicts mortality, which is why it was added to the MELD score used for allocation. It also worsens encephalopathy, by adding osmotic stress to astrocytes already strained by ammonia.

8:49His management follows the AASLD twenty twenty-one thresholds. A sodium of one hundred twenty-six to one hundred thirty-five is monitored without changing the diuretics. One hundred twenty to one hundred twenty-five gets fluid restriction to about one to one and a half liters a day and reduction or withdrawal of the diuretics. Under one hundred twenty, concentrated albumin may be used, with hypertonic saline reserved for severe symptoms in the intensive care unit, ICU, or around transplant. Tolvaptan, if it's used at all, is limited to thirty days because of hepatotoxicity.

9:25He's overloaded, so his diuretics stop, his fluid is restricted, and he gets concentrated albumin. The exception is hypovolemic hyponatremia after overdiuresis, with rapid weight loss, a low blood pressure, a rising creatinine, and little ascites, which gets saline and albumin with the diuretics stopped. Chronic hyponatremia is corrected by four to six milliequivalents per liter in twenty-four hours and never more than eight, to avoid osmotic demyelination. His rise from one hundred eighteen to one hundred twenty-three is within target.

10:01Ascites is also where infection takes hold, and the fifth patient shows how quietly it can arrive. A fifty-two-year-old man with hepatitis C cirrhosis and ascites is admitted with new confusion and mild abdominal discomfort, without fever. His creatinine is one point three, his bilirubin three point six, and he weighs eighty kilograms. The team plans lactulose and wants to skip the tap because he isn't febrile.

10:26New confusion in a man with ascites can be spontaneous bacterial peritonitis, so skipping the tap is wrong, and every patient with cirrhosis and ascites who's admitted is tapped. Spontaneous bacterial peritonitis carries a hospital mortality of twenty to forty percent, and it can present as confusion without fever, because infection precipitates encephalopathy. His ascitic neutrophil count is six hundred eighty. A neutrophil count over two hundred fifty per cubic millimeter in cirrhotic ascites makes the diagnosis whatever the symptoms or the culture, and treatment starts now. Culture-negative neutrocytic ascites is treated the same way, because culture is positive in only fifty to eighty percent of true infections, since the inoculum is small and partly controlled by complement and innate immunity in the fluid.

11:12He needs an antibiotic now. Empiric therapy is intravenous ceftriaxone, two grams daily, or cefotaxime, two grams every eight hours, for five to seven days, which covers E. coli, Klebsiella, and Streptococcus pneumoniae without needless breadth. A carbapenem is used instead for infection acquired in hospital, healthcare exposure, recent antibiotics, or septic physiology. The treatment dose of ceftriaxone is two grams, double the one gram used as prophylaxis in a gastrointestinal bleed. Aminoglycosides are forbidden.

11:46He also needs albumin, which matters as much as the antibiotic. Intravenous albumin, one and a half grams per kilogram on day one and one gram per kilogram on day three, reduces hepatorenal syndrome and short-term mortality. The infection and its inflammatory cascade drop the effective circulating volume and constrict the kidneys, and albumin restores oncotic pressure and binds the cytokines that drive that cascade. He gets one hundred twenty grams on day one and eighty on day three, plus lactulose for his confusion.

12:18His response is checked rather than assumed. A repeat paracentesis at forty-eight hours documents a fall in the neutrophil count of at least twenty-five percent. Failure suggests a resistant organism, inadequate dosing, or secondary peritonitis, which call for broader coverage guided by culture, a dose check, or abdominal imaging. His culture grows E. coli, his count falls by more than half, his creatinine returns to one, and his confusion clears.

12:46He'll need prophylaxis from here on. After an episode, secondary prophylaxis with norfloxacin four hundred milligrams daily, ciprofloxacin five hundred milligrams daily, or trimethoprim-sulfamethoxazole continues indefinitely, until the ascites resolves or he's transplanted, and he starts ciprofloxacin. Primary prophylaxis is given to any patient with cirrhosis and gastrointestinal, GI, bleeding, as seven days of ceftriaxone. It's also given to patients with an ascitic protein under one and a half grams per deciliter plus advanced disease: Child-Pugh B or C, a creatinine over one point two, a blood urea nitrogen, BUN, over twenty-five, a sodium under one hundred thirty, or a bilirubin over three. With his low ascitic protein, his bilirubin and creatinine would have qualified him before this episode. If he's on a beta-blocker, it's held while his systolic pressure is under ninety, his sodium under one hundred thirty, or his kidneys injured, and resumed once they recover.

13:45Not every infected ascites is spontaneous, and the sixth patient shows the difference. A sixty-four-year-old woman with cirrhosis and ascites has abdominal pain, an ascitic neutrophil count of fourteen thousand, a polymicrobial culture, a glucose under fifty, a protein over one, and a lactate dehydrogenase, LDH, above the serum upper limit. These point to secondary bacterial peritonitis from a perforated or inflamed bowel, so she needs imaging and surgery, not just antibiotics and albumin, and her computed tomography, CT, shows free air from a perforated diverticulum. The pleural version of spontaneous peritonitis follows the same rules. A fifty-eight-year-old man with a right-sided hepatic hydrothorax, fever, and a pleural neutrophil count of four hundred has spontaneous bacterial empyema, treated with the same antibiotics and followed by secondary prophylaxis, and episode three covers hepatic hydrothorax.

14:42The kidneys are the next organ the circulation fails, and the seventh patient shows how carefully that has to be proven. A sixty-year-old woman with alcohol-related cirrhosis and ascites on spironolactone and furosemide has a creatinine rising from zero point nine to one point nine over five days. Her urine sodium is under ten, her urinalysis has no protein or blood, and her kidney ultrasound is normal. She isn't in shock and has had no contrast, NSAIDs, or aminoglycosides. Her saturation is ninety-six percent and she weighs seventy kilograms. The team wants to start terlipressin today.

15:18Her rising creatinine with a low urine sodium suggests hepatorenal syndrome, but she can't have that label, or terlipressin, yet. The terms have changed. What was type one hepatorenal syndrome is now hepatorenal syndrome with acute kidney injury, HRS-AKI, and type two is now hepatorenal syndrome with chronic kidney disease, HRS-CKD, or with non-acute kidney injury, HRS-NAKI. HRS-AKI is functional kidney injury in structurally normal kidneys, caused by splanchnic vasodilation and renal vasoconstriction, and without transplant its mortality is fifty to eighty percent.

15:55The revised International Ascites Club criteria for HRS-AKI require cirrhosis with ascites, and a rise in creatinine of more than zero point three milligrams per deciliter within forty-eight hours or more than fifty percent within seven days. They require no structural kidney disease, meaning no proteinuria over five hundred milligrams a day, no microscopic hematuria, and a normal ultrasound, and no shock and no nephrotoxins. And they require no improvement after two days of diuretic withdrawal plus albumin, one gram per kilogram a day. She hasn't had that albumin challenge, so terlipressin waits. Her diuretics could be causing a prerenal injury, and a vasoconstrictor is the wrong treatment for a kidney that needed volume.

16:39The albumin challenge separates her from the other causes. Prerenal azotemia also has a low urine sodium, and only its improvement with volume tells it apart. Acute tubular necrosis has a high urine sodium and granular casts, and it doesn't improve. Her diuretics stop, and she gets seventy grams of albumin daily for two days, as the volume expander of choice rather than saline. On day three her creatinine is two point one, so she has HRS-AKI. A fall to one point two would have meant a prerenal injury.

17:10A different woman shows acute tubular necrosis. A sixty-two-year-old woman has a rising creatinine after two days of hypotension from a GI bleed, with muddy brown granular casts and a urine sodium of sixty. She has acute tubular necrosis, and her shock alone excludes hepatorenal syndrome.

17:30Now the first woman can be treated. First-line therapy in the United States is terlipressin, approved by the Food and Drug Administration in twenty twenty-two, plus albumin. Terlipressin is a V1 receptor agonist that constricts the splanchnic bed, reducing portal flow and shifting blood back to the systemic and renal circulation, while albumin fills the underfilled circulation. It starts at zero point eight five milligrams intravenously every six hours. On day four, a creatinine fall of less than thirty percent means increasing to one point seven milligrams every six hours, and a creatinine at or above baseline means stopping. Albumin is given at twenty to forty grams a day, and treatment continues until twenty-four hours after two creatinine values of one point five or less, for at most fourteen days.

18:18Terlipressin has firm limits. It isn't started with a saturation under ninety percent, or with ongoing coronary, peripheral, or mesenteric ischemia. In its registration trial it caused an excess of respiratory events, because pulmonary vasoconstriction worsens the ventilation-perfusion mismatch in patients with intrapulmonary shunting. It's avoided in grade three acute-on-chronic liver failure, ACLF, and used cautiously with a creatinine over five or a MELD of thirty-five or higher, because there the survival benefit narrows and the respiratory risk widens. She qualifies. Her creatinine falls thirty-five percent by day four, so her dose stays the same, and after values of one point three and one point two, terlipressin stops twenty-four hours later.

19:03Some patients can't have terlipressin. A fifty-four-year-old man in the ICU with HRS-AKI and a saturation of eighty-eight percent can't start it, so he gets norepinephrine with albumin, which works through the same vasoconstriction and is effective when an arterial line and intensive monitoring are available. Where neither drug is available, the historical ward fallback is midodrine, an oral alpha agonist, plus octreotide, which reduces splanchnic vasodilation, plus albumin. TIPS isn't routine for HRS-AKI, because the acute illness doesn't allow time for the kidneys to recover through lower portal pressure, but it may be considered in HRS-CKD with refractory ascites that meets TIPS criteria. Transplant is the only cure, and renal replacement therapy bridges refractory cases to it.

19:53The last patient shows the brain's version of the same failure. A sixty-six-year-old man with MASH cirrhosis is brought in by his wife because he sleeps all day and is awake at night, and today he couldn't tell her the date. He has asterixis and three days of constipation, and he's been taking more oxycodone for back pain. His potassium is two point nine and his BUN is up from baseline. His wife has heard he should eat less protein.

20:17A reversed sleep cycle, disorientation, and asterixis in a man with cirrhosis is overt hepatic encephalopathy. Ammonia from the gut bypasses the liver through portosystemic shunts and crosses the blood-brain barrier. Astrocytes convert it to glutamine, which is osmotically active, so they swell, and neurotransmission is disrupted. Cerebral inflammation amplifies the injury in advanced disease, hyponatremia adds osmotic stress, and manganese deposition contributes to chronic features. Unlike in acute liver failure, the brain in cirrhosis has adapted to chronic ammonia and doesn't develop significant cerebral edema.

20:57His grade comes from the West Haven classification, which runs from minimal, or covert, encephalopathy, found only on psychometric testing, to grade four. Grade one is mild confusion, a reversed sleep-wake cycle, and inattention. Grade two is lethargy, asterixis, and disorientation to time. Grade three is somnolence, marked asterixis, and disorientation to place. Grade four is coma. He knows where he is but not the date, so he's grade two.

21:28The first job is finding his precipitant, because lactulose alone won't hold if the cause continues. The precipitants are infection, especially spontaneous bacterial peritonitis; GI bleeding, a protein load that gut bacteria turn into ammonia; hypokalemia, which raises the kidney's ammonia production; metabolic alkalosis, which shifts ammonium to ammonia that diffuses into the brain; dehydration; constipation; sedatives such as benzodiazepines and opioids; and renal failure. His tap shows no infection and his stool no blood, so his rising BUN fits dehydration, and his precipitants are constipation, opioids, low potassium, and dehydration, all of which need reversing. A positive stool test would have made the encephalopathy the first sign of a bleed, calling for ceftriaxone, octreotide, and endoscopy within twelve hours.

22:25His treatment reverses each precipitant and then clears ammonia. His potassium is replaced, his oxycodone stopped, and his dehydration treated. Lactulose, thirty to sixty milliliters by mouth titrated to two to three soft stools a day, acidifies the colon, trapping ammonia as charged ammonium that can't diffuse back, and shortens transit. Too little fails to clear the ammonia, and too much causes diarrhea and dehydration, which is itself a precipitant. He's oriented in two days. After a first overt episode, rifaximin five hundred fifty milligrams twice daily is added to prevent recurrence. It reduces ammonia-producing bacteria in the colon, and since less than zero point four percent is absorbed, it's well tolerated long term. He goes home on both, with his sodium checked.

23:13His wife has heard the wrong advice about protein. Protein isn't restricted, and the target is one point two to one point five grams per kilogram a day, with a snack at night. Muscle is the largest site of ammonia detoxification outside the liver, through glutamine synthetase. Restricting protein worsened sarcopenia, and lost muscle means more ammonia, not less. The bedtime snack shortens the overnight fast so muscle isn't broken down for fuel. Branched-chain amino acids are sometimes used, but the evidence is mixed. His wife is told he should eat plenty of protein and something before bed.

23:48Two more patients show the edges of the disease. A fifty-nine-year-old woman with three overt episodes this year, despite lactulose and rifaximin, and refractory ascites has been offered TIPS. Prior overt encephalopathy is a relative contraindication, because the shunt sends more gut ammonia to the brain, so she stays on serial paracentesis with albumin. Encephalopathy that's refractory and disabling is itself a reason for transplant evaluation. A sixty-two-year-old accountant with cirrhosis who still drives has been making unusual errors at work, with a normal exam and no asterixis. He may have covert encephalopathy, which matters for complex work and driving and is found only with psychometric testing.

24:29Here are the traps the boards set in these patients. Every new or worsening ascites is tapped. A SAAG under one point one isn't cirrhotic ascites, and a high SAAG with high protein points to heart failure. Fluid isn't restricted when the sodium is normal. Furosemide isn't used alone, and the ratio of one hundred to forty is kept. NSAIDs and aminoglycosides are forbidden. A tap over five liters needs albumin. Diuretics aren't pushed past their maximum.

24:57The TIPS and sodium traps follow. TIPS isn't placed with a high MELD or bilirubin, an INR over two point five, recurrent encephalopathy, heart failure, or pulmonary hypertension, or without an echocardiogram first and rifaximin for protection. Weekly albumin isn't given routinely. Dilutional hyponatremia isn't treated with saline, tolvaptan isn't used beyond thirty days, and sodium isn't corrected by more than eight in a day.

25:23The infection traps come next. The absence of fever isn't a reason to skip the tap, and treatment doesn't wait for a positive culture. Infection acquired in hospital needs broader coverage. Albumin on days one and three is part of treatment. A very high count, a polymicrobial culture, or a low glucose means secondary peritonitis. And prophylaxis doesn't stop after the first episode.

25:48The kidney traps are about proof. Hepatorenal syndrome isn't diagnosed without the two-day albumin challenge, or in the presence of proteinuria, hematuria, an abnormal ultrasound, or shock. Casts and a high urine sodium mean acute tubular necrosis. Terlipressin isn't started with a saturation under ninety or in grade three ACLF, and its dose is reviewed on day four. And TIPS isn't a treatment for HRS-AKI.

26:14The encephalopathy traps close the list. Lactulose isn't given without looking for the precipitant. Protein isn't restricted. Lactulose is titrated, not overshot into dehydration. Rifaximin is added after the first episode. Benzodiazepines aren't given for agitation. And a normal exam doesn't exclude covert encephalopathy.

26:34So, start to finish. Ascites comes from splanchnic vasodilation and hormonal sodium retention. Tap every new or worsening ascites: a SAAG of one point one or more means portal hypertension, and low protein points to cirrhosis. Restrict sodium to two grams and fluid only below one hundred twenty-five, give spironolactone and furosemide at one hundred to forty up to four hundred to one hundred sixty, avoid NSAIDs and aminoglycosides, and give six to eight grams of albumin per liter after taps over five liters.

27:07Refractory ascites means serial taps or TIPS, after an echocardiogram and with rifaximin and a small stent, plus a transplant referral. Manage hyponatremia by severity, and correct it by no more than eight a day.

27:21A neutrophil count over two hundred fifty is spontaneous bacterial peritonitis. Give ceftriaxone two grams daily for five to seven days, or a carbapenem for infection acquired in hospital, with albumin one and a half grams per kilogram on day one and one on day three, and long-term prophylaxis afterward. A count over ten thousand, a low glucose, a high LDH, or polymicrobial growth means secondary peritonitis and surgery.

27:47Diagnose HRS-AKI only after two days of diuretic withdrawal and albumin without improvement, and treat it with terlipressin plus albumin, increasing the dose on day four if needed, or with norepinephrine when the saturation is under ninety. For encephalopathy, find and reverse the precipitant, give lactulose to two to three soft stools, add rifaximin after the first episode, give one point two to one point five grams of protein per kilogram with a bedtime snack, avoid TIPS after recurrent episodes, and test for covert disease when function slips.

28:19Episode three is the lungs, coagulation and procedures, and the vascular liver diseases.

28:25For 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 two of three of chapter twenty-three, and I'll see you in the next one.