Liver Test Interpretation and Workup: Liver Test Interpretation: Workup and Staging

0:00Welcome to Board Pearls. This is episode two of two of the Liver Test Interpretation chapter, in the Liver Disease module. This episode covers the single abnormal value and then staging the liver. A young man with yellow eyes after fasting shows why an isolated bilirubin is split into direct and indirect first, how hemolysis is excluded, and how the inherited syndromes are told apart, and a woman on the pill shows the two benign causes of an isolated direct bilirubin. A woman with a high alkaline phosphatase and an aching thigh shows which test proves whether the liver is the source, and a man with weight loss shows what a hepatic source can hide. A man with fatty liver shows how fibrosis is staged without a biopsy, with the Fibrosis-4 index, FIB-4, and elastography, and why one stiffness number doesn't fit every disease. And a woman with ascites and an unclear diagnosis shows when biopsy still earns its place, percutaneous versus transjugular, and what the hepatic venous pressure gradient measures that no image can.

1:05Here's the first patient. A twenty-two-year-old man noticed yellow eyes on the third day of a viral gastroenteritis, when he'd barely eaten. His total bilirubin is three point four, almost all of it indirect. His alanine aminotransferase, ALT, aspartate aminotransferase, AST, alkaline phosphatase, albumin, international normalized ratio, INR, and blood count are normal. He takes no medications. He remembers his eyes looking yellow once before, after a long fast before an exam.

1:41Jaundice after fasting and during an illness, with every other liver test normal, suggests Gilbert syndrome, but the diagnosis is made only after the alternatives are excluded. A raised bilirubin with a normal aminotransferase, alkaline phosphatase, albumin, and INR is isolated hyperbilirubinemia, and the first split is whether it's mostly indirect, meaning unconjugated, or direct, meaning conjugated. Each side has a short list driven by mechanism, and its commonest causes are benign, so fractionating the bilirubin spares a broad liver workup. His is indirect.

2:17Indirect bilirubin rises when heme breakdown outpaces conjugation, in three ways. The first is increased production, from hemolysis, ineffective erythropoiesis, resorption of a hematoma, or transfusion. The second is decreased uptake by the liver, from heart failure, a portosystemic shunt, or drugs that compete for uptake, such as rifampin and probenecid. The third is decreased conjugation, from Gilbert syndrome, Crigler-Najjar syndrome types one and two, or drug inhibition.

2:48The level helps narrow it. Indirect hyperbilirubinemia rarely exceeds twelve milligrams per deciliter without severe hemolysis or Crigler-Najjar syndrome, and Gilbert syndrome rarely exceeds five, so an indirect bilirubin of nine isn't called Gilbert syndrome without looking harder. He takes no competing drugs and has no heart failure or shunt, which leaves hemolysis or a conjugation defect.

3:13Hemolysis is excluded with a short panel. Haptoglobin is low in intravascular hemolysis, because it's consumed binding free hemoglobin. Reticulocytes are high, as the marrow replaces lost cells. The smear looks for schistocytes, spherocytes, or target cells. Lactate dehydrogenase, LDH, is raised, and a direct Coombs test looks for autoimmune hemolysis. His are all normal.

3:37That leaves a conjugation defect, and in him it's Gilbert syndrome, the most common cause of indirect hyperbilirubinemia. An extra TA repeat in the TATA box promoter of UDP-glucuronosyltransferase one A one, called UGT1A1 star twenty-eight, reduces the enzyme to about thirty percent of normal. That's enough on an ordinary day, but fasting, illness, dehydration, surgery, or other stress raises the bilirubin load beyond it, and the level falls back within days.

4:10Gilbert syndrome affects five to ten percent of the population, needs no treatment, and has no clinical consequence apart from a mild association with irinotecan toxicity, since that drug is cleared by the same enzyme. It's a clinical diagnosis, made without genetic testing. He's reassured and told to mention it if he ever needs irinotecan, and two weeks later, eating normally, his bilirubin is one point four. The label would be doubted if the indirect bilirubin were well above five, the hemolysis screen were positive, any other liver test were abnormal, or a competing drug were present.

4:43The same enzyme has a severe end. An infant with severe jaundice in the first days of life, at risk of kernicterus, has Crigler-Najjar type one, with no UGT1A1 activity at all. Type two has minimal residual activity and responds to phenobarbital, which induces it, while type one doesn't. So Gilbert syndrome, Crigler-Najjar type two, and type one run from reduced, to minimal, to absent enzyme.

5:09The direct side has its own short list, and the second patient shows it. Isolated direct hyperbilirubinemia is unusual. Early or partial obstruction rarely raises the bilirubin without changing the alkaline phosphatase or aminotransferases, and pure obstruction tends to plateau under fifteen milligrams per deciliter, so a level above fifteen suggests parenchymal disease. With a truly isolated direct elevation and a normal biliary tree, the list comes down to two inherited defects of canalicular transport. A thirty-year-old woman has a direct bilirubin of three point eight found a few months after she started an oral contraceptive, with a normal aminotransferase, alkaline phosphatase, and biliary tree.

5:50The first is Dubin-Johnson syndrome, an autosomal recessive defect in multidrug resistance-associated protein two, MRP2, also called ABCC2, the canalicular transporter that exports conjugated bilirubin into bile. Conjugated bilirubin refluxes into the plasma instead, giving a chronic, mild elevation of two to five milligrams per deciliter with a benign course. Its pathognomonic finding is a black liver, from melanin-like pigment in the lysosomes. The urinary total coproporphyrin is normal, but the ratio is inverted, with more coproporphyrin one, and hepatobiliary scintigraphy shows delayed clearance. Pregnancy and oral contraceptives can unmask it.

6:33The second is Rotor syndrome, from defects in OATP1B1 and OATP1B3, which take conjugated bilirubin back into the hepatocyte from the sinusoidal blood. The liver is a normal color, the urinary total coproporphyrin is markedly raised, and the scintigraphy pattern is different. It's rarer and similarly benign. In both syndromes conjugation is normal, and the defect is in moving conjugated bilirubin, the opposite of Gilbert syndrome. Her urine shows a normal total coproporphyrin with an inverted ratio, so she has Dubin-Johnson syndrome. Neither syndrome needs treatment, and she can expect her bilirubin to rise in pregnancy.

7:12A single abnormal value can also come from outside the liver, and the third patient shows it. A seventy-year-old woman has an alkaline phosphatase of three hundred ten on routine labs, with a normal ALT, AST, bilirubin, and albumin. She feels well apart from an aching right thigh and a sense that her hat fits more tightly. She takes no regular medications and drinks rarely. She's been referred to liver clinic.

7:40Her thigh pain and her tightening hat suggest that her alkaline phosphatase may not be coming from the liver at all. Alkaline phosphatase comes from the liver, bone, intestine, placenta, and kidney, with liver and bone accounting for almost all of it in adults. Pregnancy raises it through the placental form, growing children and adolescents through bone growth, and high bone turnover through the bone form, as in Paget disease, healing fractures, bone metastases, hyperthyroidism, and hyperparathyroidism. Macro-alkaline phosphatase, the enzyme bound to an immunoglobulin, causes a persistent elevation of no significance. So a high value in a pregnant woman or a growing teenager isn't liver disease.

8:19One test settles the source: gamma-glutamyl transpeptidase, GGT, which is induced in cholestatic liver disease but isn't made by bone. A high alkaline phosphatase with a high GGT confirms a hepatic source and starts the cholestatic workup from episode one, while a normal GGT in an adult points to bone. Alcohol and the enzyme-inducing drugs phenytoin, phenobarbital, and carbamazepine raise GGT on their own, so in those patients a raised GGT doesn't prove a hepatic source, and a bone fraction asks the question directly. Her GGT is normal.

8:55The bone workup is calcium, phosphorus, and vitamin D, then a bone alkaline phosphatase fraction or a bone scan, with a scan preferred when there's focal pain. Her calcium, phosphorus, and vitamin D are normal, and her bone scan shows intense uptake in the right femur and skull, which is Paget disease of bone. She's referred to endocrinology. One normal GGT spared her an ultrasound, an anti-mitochondrial antibody, AMA, and a magnetic resonance cholangiopancreatography, MRCP.

9:29When the GGT is high, the source is the liver, and the fourth patient shows what a hepatic source can hide. A confirmed hepatic elevation follows the cholestatic sequence: ultrasound, then AMA and immunoglobulin M, IgM, then MRCP and immunoglobulin G4, IgG4, as the findings suggest. An isolated hepatic alkaline phosphatase adds computed tomography, CT, or magnetic resonance imaging, MRI, to look for infiltrative disease, which can raise the alkaline phosphatase markedly while sparing the aminotransferases and bilirubin until late. That means sarcoidosis, with granulomatous hepatitis and often mediastinal or hilar nodes; amyloidosis, with an enlarged liver, a markedly high alkaline phosphatase, near-normal aminotransferases, and proteinuria; and lymphoma or metastases.

10:26A fifty-nine-year-old man has weight loss, an enlarged liver, an alkaline phosphatase of six hundred ninety-nine, near-normal aminotransferases and bilirubin, protein in his urine, a high GGT, a non-dilated biliary tree, and a negative AMA. He goes to cross-sectional imaging, and his picture points to amyloidosis, with the proteinuria showing that the process isn't confined to the liver. Imaging rarely proves infiltrative disease, so if it doesn't settle the question, a biopsy does.

10:58The other hepatic causes have their own clues. Primary biliary cholangitis and primary sclerosing cholangitis are the classic causes. Drug-induced cholestasis, from amoxicillin-clavulanate, anabolic steroids, oral contraceptives, parenteral nutrition, and other antibiotics, is identified by its timing. And sepsis-associated cholestasis appears in critically ill patients, driven by cytokines that suppress the canalicular transporters. A persistent unexplained hepatic elevation warrants a biopsy, which finds primary biliary cholangitis, small-duct primary sclerosing cholangitis with its normal cholangiogram, granulomatous hepatitis, and infiltrative disease.

11:44Once a liver disease is named, the next question is how much scar it has caused, and the fifth patient shows how that's answered. A fifty-eight-year-old man has metabolic dysfunction-associated steatotic liver disease, MASLD, with obesity, type two diabetes, and a bright liver on ultrasound. His AST is forty-four, his ALT fifty-two, and his platelets one hundred eighty thousand. He drinks rarely and asks how much damage there is.

12:09His MASLD is diagnosed, and what he needs now is a fibrosis stage, which takes the right kind of test. Imaging shows structure, while elastography and blood scores quantify scar. Ultrasound is the first-line liver imaging, fast, safe, and cheap, for biliary dilation, echotexture, with steatosis appearing bright, masses, splenomegaly, ascites, and portal vein patency and flow on Doppler, but it's operator dependent and insensitive for small masses and for fibrosis. CT follows for a mass outside the ducts, for staging, or for a technically limited ultrasound. MRI gives better soft tissue characterization and is the modality of choice for diagnosing hepatocellular carcinoma and characterizing benign lesions, and MRCP is the noninvasive gold standard for primary sclerosing cholangitis and for duct stones when ultrasound is non-diagnostic. His ultrasound shows steatosis but can't measure his scar.

13:11Staging starts with FIB-4, calculated from age, AST, ALT, and platelets. A value under one point three rules out advanced fibrosis with a high negative predictive value, a value over two point six seven means high risk, and the range between is indeterminate and needs a second test. Age weighs heavily in the formula, so over sixty-five the lower cutoff shifts to two point zero. His FIB-4 is one point eight, which is indeterminate at fifty-eight.

13:41Other blood scores exist and do less. The AST-to-platelet ratio index, APRI, uses only the AST and platelets and is less accurate. The nonalcoholic fatty liver disease, NAFLD, fibrosis score adds body mass index, BMI, diabetes, and albumin. FibroTest is a proprietary panel that includes alpha-two macroglobulin and apolipoprotein A1. And the enhanced liver fibrosis, ELF, test, made of hyaluronic acid, procollagen three N-terminal peptide, PIIINP, and tissue inhibitor of metalloproteinase one, TIMP-1, indicates advanced fibrosis above nine point eight and a high risk of liver-related events above eleven point three.

14:33An indeterminate FIB-4 needs a second test, usually vibration-controlled transient elastography, FibroScan. It sends a low-frequency shear wave through the liver and reports stiffness in kilopascals, with a controlled attenuation parameter for fat. The cutoffs depend on the disease, because inflammation, congestion, cholestasis, and steatosis all confound stiffness. In MASLD, a stiffness under eight kilopascals effectively rules out advanced fibrosis, stage three to four, and over twelve to fourteen suggests cirrhosis. In hepatitis C, advanced fibrosis is roughly nine and a half and cirrhosis around twelve and a half, and hepatitis B has its own cutoffs.

15:14FibroScan has limits. It fails in three to twenty-seven percent of patients, worst in obesity, and obesity, severe steatosis, inflammation, congestion, and biliary obstruction each raise the stiffness regardless of fibrosis. His stiffness is ten point five, in the middle.

15:32The same stiffness speaks to portal hypertension. Under the Baveno seven criteria, twenty-five kilopascals or more rules in clinically significant portal hypertension, fifteen or less with platelets of one hundred fifty thousand or more rules it out, and twenty to under twenty-five with platelets under one hundred fifty thousand makes it probable. Ruling it in triggers variceal screening, covered in chapter twenty-three. His values rule it out.

15:58His results are still unclear, and that's when a more accurate test comes in. Magnetic resonance, MR, elastography is the most accurate noninvasive fibrosis test. An acoustic driver on the right abdomen sends shear waves that MRI converts into a stiffness image, with higher accuracy than FibroScan, a failure rate under two percent, and regional detail, at the cost of expense, availability, and breath-holds, and iron overload disrupts its signal. It's the test of choice when FibroScan is unreliable, as with severe obesity, ascites, or narrow intercostal spaces. Point or two-dimensional shear wave elastography, done during conventional ultrasound, has an accuracy similar to MR elastography, with a failure rate of two to thirteen percent and less standardized quality criteria.

16:48His FIB-4 and his FibroScan both sit in their indeterminate ranges, and his obesity and steatosis can raise his stiffness. His MR elastography, which is less affected by obesity, shows significant but not advanced fibrosis. His treatment is covered in chapter twenty-one, and his noninvasive tests will be repeated over time.

17:09The sequence in MASLD is FIB-4 first. A value under one point three, or under two point zero over sixty-five, is reassuring, with retesting in one to three years, and higher values go to FibroScan, MR elastography, or ELF. Concordant low results rule out advanced fibrosis. Concordant high results, a FIB-4 over two point six seven with a stiffness over twelve, make a biopsy unnecessary in many cases. And discordant results trigger MR elastography or a biopsy. A sixty-nine-year-old woman with MASLD and a FIB-4 of one point seven is low risk, because her age-adjusted cutoff is two point zero.

17:50Sometimes the noninvasive tests can't answer the question, and the last patient shows when biopsy comes in. A sixty-one-year-old woman has ascites, an enlarged liver, an alkaline phosphatase of four hundred, and an ALT of ninety. Her ultrasound shows no ductal dilation, her AMA is negative, and her MRCP is normal. Her antinuclear antibody, ANA, is weakly positive with a high immunoglobulin G, IgG. Her INR is one point seven, and her platelets are seventy-two thousand. Her team can't tell whether she has autoimmune hepatitis overlapping a cholestatic disease, an infiltrative process, or something else.

18:33Her noninvasive workup is complete and her diagnosis is still unclear, which is the main reason for a liver biopsy. Biopsy is indicated for diagnostic uncertainty after a complete noninvasive workup, such as overlapping autoimmune and metabolic features, possible infiltrative disease that imaging hasn't shown, or suspected drug injury in a complex patient on many drugs. It's also indicated for AMA-negative primary biliary cholangitis, a suspected autoimmune overlap with primary biliary cholangitis or primary sclerosing cholangitis, small-duct primary sclerosing cholangitis, suspected granulomatous hepatitis, staging that will change management, and acute liver failure of uncertain cause, where a transjugular biopsy can identify the cause within hours. She has an unresolved overlap and a possible infiltrative process.

19:25Her ascites and blood counts decide the approach. Percutaneous biopsy is the default. Transjugular biopsy is chosen for tense ascites, which impedes access and risks bleeding into the fluid; for uncorrectable coagulopathy or thrombocytopenia; for morbid obesity; for peliosis hepatis; for suspected vascular liver disease; and whenever the portal pressure is wanted. A catheter passes from the right internal jugular vein through the right atrium and inferior vena cava into a hepatic vein, and the needle is deployed through it. The bleeding risk is lower, because any bleeding returns to the circulation rather than the peritoneum, and the trade-offs are arrhythmia, transient puncture of the capsule, and the risks of central access. She goes transjugular.

20:11Her INR and platelets don't need correcting first. The historical rule required platelets over fifty thousand and an INR under one point five. But in advanced liver disease, the procoagulant factors and the natural anticoagulants fall together, a rebalanced hemostasis, so the INR overstates the bleeding risk. The twenty twenty-one American Gastroenterological Association, AGA, guidance on coagulation in cirrhosis and the twenty twenty-one American Association for the Study of Liver Diseases, AASLD, guidance on procedural bleeding agree that prophylactic plasma, platelets, or cryoprecipitate for an isolated INR elevation or moderate thrombocytopenia isn't supported, and may worsen outcomes, because the volume raises portal pressure. Viscoelastic testing, thromboelastography, TEG, or rotational thromboelastometry, ROTEM, identifies the true coagulopathy that benefits from targeted correction. Hers shows adequate clot formation, so she isn't transfused.

21:10Because she's going transjugular, her portal pressure can be measured in the same procedure, as the hepatic venous pressure gradient, HVPG. A balloon-tipped catheter wedged in a hepatic venule gives the wedged hepatic venous pressure, which reflects sinusoidal pressure and, indirectly, portal pressure. With the balloon deflated, the free hepatic venous pressure is measured, and the HVPG is the wedged minus the free pressure. Normal is one to five millimeters of mercury. Six to nine is portal hypertension that's rarely clinically significant. Ten or more is clinically significant portal hypertension, when varices develop and the risk of decompensation begins. Twelve or more correlates with the risk of variceal hemorrhage.

21:53Her HVPG is fourteen, and her biopsy shows interface hepatitis with plasma cells and early cirrhosis, consistent with autoimmune hepatitis, without infiltrative disease. Chapter twenty covers her treatment, and chapter twenty-three her portal hypertension.

22:09The two pressures also localize the disease. Pre-sinusoidal portal hypertension, from idiopathic non-cirrhotic portal hypertension, schistosomiasis, or portal vein thrombosis, has a low or normal wedged and free pressure and a normal HVPG, because the hepatic vein catheter never sees the pressure upstream. Sinusoidal portal hypertension, dominated by cirrhosis, has a high wedged pressure, a normal free pressure, and a high HVPG, which is her pattern. Post-sinusoidal portal hypertension, from right heart failure, constrictive pericarditis, or Budd-Chiari syndrome, has both pressures high, so the HVPG can be normal. A normal HVPG with obvious portal hypertension means the cause lies before or after the sinusoids.

22:58The default approach fits most patients. A forty-four-year-old man with a cholestatic pattern, a normal MRCP, and suspected small-duct primary sclerosing cholangitis has no ascites, normal clotting, and a normal body habitus. So he has a percutaneous biopsy under ultrasound guidance, through the right intercostal space at end-expiration, avoiding the pleural reflection above and the gallbladder, kidney, and colon below. An adequate core is one and a half to two and a half centimeters long with at least eleven portal tracts, because thinner samples miss patchy fibrosis and diagnoses like small-duct disease.

23:36The risks are low but real. Symptomatic complications occur in roughly one to three percent, mostly bleeding. Major complications include hemorrhage requiring transfusion or intervention, pneumothorax, puncture of the gallbladder, pneumoperitoneum, and rarely hemobilia, with a mortality of about one in ten thousand to one in a thousand. His biopsy shows concentric periductal fibrosis around small ducts, confirming small-duct primary sclerosing cholangitis.

24:06Two other approaches have narrower roles. A surgical wedge biopsy taken during surgery for another reason is large but superficial, and it can overestimate fibrosis because fibrosis is greater under the capsule, so it should be at least one centimeter deep. A biopsy guided by endoscopic ultrasound, EUS, suits patients with surgically altered anatomy, contraindications to the other approaches, or a planned EUS, with comparable samples in expert hands.

24:35Here are the traps the boards set in these patients. An isolated bilirubin is split into direct and indirect before anything else. Gilbert syndrome isn't diagnosed without a hemolysis screen, doesn't need genetic testing, and doesn't explain an indirect bilirubin of nine. Rifampin and heart failure are causes to remember. Crigler-Najjar type one doesn't respond to phenobarbital. The coproporphyrin patterns of Dubin-Johnson and Rotor syndromes are opposite, and neither syndrome needs treatment.

25:04The alkaline phosphatase traps follow. A liver workup for an isolated alkaline phosphatase doesn't start before the GGT. A high alkaline phosphatase in pregnancy or adolescence isn't liver disease. Alcohol and enzyme inducers raise the GGT by themselves. And a very high alkaline phosphatase with proteinuria can be amyloidosis.

25:26The staging traps come next. Imaging doesn't estimate fibrosis. The FIB-4 cutoff of one point three doesn't apply over sixty-five. One stiffness cutoff doesn't fit every disease, and stiffness isn't trusted during hepatitis, congestion, or obstruction. Iron disrupts MR elastography. And a stiffness of twenty-five kilopascals rules in clinically significant portal hypertension.

25:50The biopsy traps close the list. A biopsy isn't done when noninvasive tests answer the question. Tense ascites calls for the transjugular approach, not the percutaneous one. An INR or platelet count alone isn't a reason to transfuse. Too few portal tracts, or a shallow wedge, can mislead. And a normal HVPG doesn't mean there's no portal hypertension.

26:13So, start to finish. Split an isolated bilirubin into indirect and direct. An indirect elevation with a negative hemolysis screen is Gilbert syndrome, from reduced UGT1A1, rising with fasting or illness and needing nothing. An isolated direct elevation with a normal biliary tree is Dubin-Johnson syndrome, with a black liver and a normal total but inverted urine coproporphyrin, or Rotor syndrome, with a high total, and both are benign.

26:44Confirm the source of an isolated alkaline phosphatase with the GGT. A normal GGT in an adult means bone, and a high one means liver, which calls for the cholestatic workup plus imaging for infiltrative disease, with a biopsy for a persistent unexplained elevation.

27:00Match the test to the question: imaging shows structure, and elastography and blood scores estimate scar. Start MASLD with FIB-4, where under one point three, or under two point zero over sixty-five, is low risk, then FibroScan, where under eight rules out advanced fibrosis and the cutoffs differ by disease, and twenty-five or more rules in clinically significant portal hypertension. Use MR elastography when FibroScan fails or the results disagree.

27:31Biopsy only when it changes management. Go percutaneous by default, with at least eleven portal tracts, and transjugular for ascites, coagulopathy, obesity, vascular disease, or a wanted pressure gradient, without transfusing for numbers alone. An HVPG of ten or more is clinically significant portal hypertension, twelve or more carries a bleeding risk, and a normal HVPG with portal hypertension points before or after the sinusoids.

27:59That completes the Liver Test Interpretation chapter. The next chapter is viral hepatitis.

28:06For 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 two of chapter seventeen, and I'll see you in the next one.