Liver Test Interpretation and Workup: Isolated Abnormalities, Imaging, and Biopsy

0:00Welcome to Board Pearls. This is episode two of two of the Liver Test Interpretation chapter, in the Liver Disease module. This episode is what to do when the pattern is an isolated abnormality or the labs themselves aren't enough: isolated unconjugated and conjugated hyperbilirubinemia, isolated alkaline phosphatase elevation, the imaging and noninvasive fibrosis tools that come after the labs, and the biopsy indications where tissue still changes management.

0:28The prior episode leaned on a pattern across multiple liver tests. These cases are different: a single value is up with everything else normal, or the labs aren't the bottleneck and imaging and fibrosis assessment are. Each forces the same question in a different form: what's the next test, and does it commit me to an invasive workup or excuse me from it?

0:49Start with isolated hyperbilirubinemia, where the bilirubin is up and the transaminases, alkaline phosphatase, albumin, and INR are normal. The workup splits at the very first line on the panel: is the elevation predominantly indirect, the unconjugated fraction, or direct, the conjugated fraction? The two branches share almost nothing, and once you know which side you're on, the differential is short and mostly benign. Indirect hyperbilirubinemia means heme breakdown is outpacing conjugation, in three categories: increased production from hemolysis, ineffective erythropoiesis, or hematoma resorption; decreased uptake from heart failure, shunting, or drugs that compete for the uptake transporter; and decreased conjugation, the Gilbert and Crigler-Najjar end. The total in indirect hyperbilirubinemia rarely climbs high, and Gilbert itself rarely passes five.

1:42The clinical task on the indirect side is to separate hemolysis from Gilbert, since both produce unconjugated bilirubin and both can show scleral icterus during stress. The screen runs five values: a low haptoglobin when intravascular hemolysis consumes it, an elevated reticulocyte count as the marrow replaces destroyed cells, the smear morphology, schistocytes in microangiopathy, spherocytes in autoimmune or hereditary spherocytosis, bite cells and Heinz bodies in G6PD, an elevated LDH, and a positive direct Coombs in autoimmune hemolysis. A clean hemolysis screen in the right vignette is Gilbert. The G6PD vignette returns most often: a young person of Mediterranean or African descent started on trimethoprim-sulfamethoxazole a few days ago who develops fatigue, dark urine, and icterus, with unconjugated hyperbilirubinemia, a dropping hemoglobin, reticulocytosis, low haptoglobin, high LDH, and bite cells and Heinz bodies on the smear, where the mechanism is acute oxidative hemolysis, the trigger is the sulfa drug, and the answer is to stop the drug and support, not chase a liver etiology. Gilbert is the diagnosis when the hemolysis screen is clean: reduced activity of the conjugating enzyme from a promoter polymorphism leaves patients running on about a third of normal conjugation capacity, which keeps up as long as substrate flux is normal but fails when the patient fasts, gets a viral illness, dehydrates, or undergoes any physiologic stress, so the classic stem is a young patient who noticed icterus during a viral illness or a long fast, with bilirubin returning to baseline within days. It affects a substantial minority of the population, needs no treatment, and the only clinical consequence worth knowing is increased irinotecan toxicity because the same enzyme conjugates its active metabolite, with the diagnosis clinical, not genetic. Crigler-Najjar sits at the severe end of the same enzyme spectrum, with the complete-absence type presenting in the neonate with severe unconjugated hyperbilirubinemia and kernicterus risk and the partial type responding to phenobarbital.

3:47Direct hyperbilirubinemia in true isolation, with normal alkaline phosphatase and transaminases, is unusual, because pure obstruction rarely produces an isolated conjugated bilirubin without dragging up the other tests. When it's genuinely isolated with a normal biliary tree on imaging, the differential narrows to two inherited canalicular transport defects. Dubin-Johnson is a defect in the transporter that exports conjugated bilirubin into bile, so it accumulates and refluxes into plasma, giving chronic mild conjugated hyperbilirubinemia with normal alkaline phosphatase and transaminases and a benign course, with the pathognomonic black-pigmented liver at biopsy, an inverted urinary coproporphyrin isomer ratio despite a normal total, and unmasking by oral contraceptives or pregnancy. Rotor is similar in phenotype but from a defect in the sinusoidal uptake transporters, with a normal-colored liver, no black pigment, and a markedly elevated urinary total coproporphyrin rather than the inverted ratio, equally benign. So the young woman with mild conjugated hyperbilirubinemia after starting oral contraceptives, with normal alkaline phosphatase, transaminases, and ultrasound, is the favored Dubin-Johnson vignette, and the answer is reassurance, not biopsy and not a viral panel.

5:08Now isolated alkaline phosphatase elevation. Alkaline phosphatase is a multi-tissue enzyme, produced by liver, bone, intestine, placenta, and kidney, so an isolated elevation doesn't necessarily mean liver disease: pregnancy raises it through the placental form in the third trimester, growing children have elevated levels from bone growth, and high bone turnover from Paget disease, healing fractures, bone metastases, or hyperparathyroidism raises the bone fraction, with a rare immunoglobulin-bound complex a benign cause of persistent unexplained elevation. The discriminator the boards expect is gamma-glutamyl transpeptidase, which is induced in cholestatic disease and by alcohol and enzyme-inducing drugs but is not produced by bone, so an elevated alkaline phosphatase with an elevated GGT confirms a hepatic source, while an elevated alkaline phosphatase with a normal GGT in an adult points to bone, and the next step is calcium, phosphorus, vitamin D, and a bone fraction or bone scan. The split is binary and decisive. The Paget vignette runs cleanly through it: an older patient with progressive deep bone pain and a creeping hat size, with an isolated alkaline phosphatase several times normal, a normal GGT, and normal calcium and phosphate, where the hepatic workup stops, the bone fraction confirms the source, and a bisphosphonate is the treatment. The pregnancy vignette is even cleaner: a third-trimester woman with alkaline phosphatase a couple of times normal who feels well with no pruritus or jaundice and normal bile acids gets reassurance, with the trap being intrahepatic cholestasis of pregnancy, which requires pruritus plus total bile acids over ten, severity above forty triggering ursodiol and consideration of early delivery. And the isolated GGT vignette deserves its own beat: a patient on phenytoin with an isolated GGT elevation on an insurance screen and everything else normal is showing enzyme induction, not liver disease, so the answer is to identify the inducer and observe.

7:03Once alkaline phosphatase is confirmed hepatic by an elevated GGT, the workup follows the cholestatic algorithm: ultrasound first for dilation, masses, and infiltration, then the mitochondrial antibody and IgM for primary biliary cholangitis, the canonical cause of asymptomatic elevation in a middle-aged woman, then MRCP for sclerosing cholangitis especially with inflammatory bowel disease, IgG4 if the cholangiogram fits, and cross-sectional imaging for infiltrative disease. The trap on sclerosing cholangitis is that its alkaline phosphatase is often normal at any single check, so a normal lab doesn't exclude it when the picture fits.

7:43That brings us to imaging and noninvasive fibrosis, where the post-lab workup lives. The split that matters is between modalities that image structure, ultrasound, CT, MRI, and MRCP, and modalities that quantify fibrosis, the elastography techniques and the serum panels, and a complete workup uses both. Ultrasound is the first-line liver imaging because it's fast, safe, cheap, and answers the early questions, biliary dilation, a bright steatotic liver, masses, splenomegaly, ascites, and portal vein patency on Doppler, limited by operator dependence and poor fibrosis quantification. CT is next for an extraductal mass or staging, MRI gives better soft-tissue characterization and is the modality of choice for hepatocellular carcinoma and benign lesion characterization, and MRCP is the noninvasive standard for the biliary tree. For a focal lesion in a cirrhotic liver, the answer is multiphase contrast MRI or CT read through the standardized imaging criteria: a new lesion with arterial enhancement, washout, a capsule, and threshold growth or qualifying size is hepatocellular carcinoma by imaging alone without biopsy, and biopsy of a definite cancer in a cirrhotic risks seeding the needle tract and isn't needed, the one place in oncology where imaging substitutes for tissue.

9:15Transient elastography measures liver stiffness by sending a shear wave through the liver, and stiffness correlates with fibrosis, but the cutoffs are etiology-dependent because inflammation, congestion, cholestasis, and steatosis all confound it. In fatty liver disease, a value under eight effectively rules out advanced fibrosis while values in the low teens suggest cirrhosis, and the cutoffs differ in hepatitis C and again in hepatitis B, so the threshold isn't universal and must be anchored to the etiology. Its failure rate is meaningful and worst in obese patients, with confounders including obesity, severe steatosis, active inflammation, hepatic congestion, and biliary obstruction, each of which can raise the apparent stiffness, and it also measures a steatosis parameter. It crosses into portal hypertension too: a stiffness of twenty-five or more rules in clinically significant portal hypertension, twenty to under twenty-five with a platelet count under a hundred fifty thousand indicates probable significant portal hypertension, and a stiffness of fifteen or less with platelets of a hundred fifty thousand or more rules it out, which triggers variceal screening. MR elastography is the most accurate noninvasive fibrosis test, with a very high accuracy and a very low failure rate, using an acoustic driver during MRI to generate shear waves converted to a stiffness map, with the advantages of accuracy, a low failure rate, and regional visualization, and the disadvantages of cost and availability, so it's the choice when transient elastography is unreliable in severe obesity, ascites, or narrow intercostal spaces, with iron overload the main confounder. Shear-wave elastography on conventional ultrasound is a reasonable bedside alternative.

10:51The serum biomarkers are the screening layer above elastography. FIB-4 is the main one because it's calculated from age, AST, ALT, and platelets, all on routine labs, and the boards test three cutoffs: under one point three rules out advanced fibrosis with high negative predictive value, over two point six seven supports advanced fibrosis, and the indeterminate zone between triggers a second-tier test, with the age threshold mattering because in patients over sixty-five the lower cutoff shifts up to two point zero, since age is a heavy weight in the formula. The AST-to-platelet ratio is simpler but slightly less accurate, the fatty-liver fibrosis score adds body mass index, diabetes, and albumin and runs slightly better, and the enhanced liver fibrosis panel indicates advanced fibrosis above one threshold and a high risk of liver-related events above a higher one. So the fatty-liver sequence the boards build their vignettes from is FIB-4 first as the screen, with a low result reassured and retested in a few years, an indeterminate or high result triggering transient or MR elastography, concordant low results ruling out advanced fibrosis, concordant high results making biopsy unnecessary, and discordant results triggering MR elastography or biopsy, so the candidate who reaches for FIB-4 before elastography and elastography before biopsy is following the modern algorithm. And a patient with established cirrhosis moves into cancer surveillance with ultrasound and alpha-fetoprotein every six months plus variceal screening, so the fibrosis test doesn't just answer a fibrosis question, it opens a surveillance program.

12:26That brings us to liver biopsy, which has narrowed in indication as noninvasive tools improved but remains the answer in four situations: uncertain staging after serology and noninvasive tests; overlapping diagnoses that serology can't separate, like autoimmune overlap, antibody-negative primary biliary cholangitis, small-duct sclerosing cholangitis, or unexplained granulomatous hepatitis; an etiology still unclear after a complete workup; and when transjugular access provides the pressure-gradient information imaging can't. The default is percutaneous biopsy under ultrasound guidance, advancing the needle through the right intercostal space at end-expiration under direct visualization to avoid the pleura above and the gallbladder, kidney, and colon below, taking a core of a couple of centimeters that should contain at least eleven portal tracts for adequate staging, since an inadequate sample can miss focal fibrosis. Complications are uncommon, with bleeding the dominant concern, and the serious ones being significant hemorrhage, pneumothorax, hemothorax, gallbladder puncture, and rare hemobilia, with a low mortality. The post-biopsy vignette to recognize is right shoulder pain, increasing right-upper-quadrant discomfort, tachycardia, and a falling hemoglobin within hours, where the shoulder pain reflects diaphragmatic irritation from blood, and most serious events present within hours, which is why observation periods are standard.

13:49Pre-biopsy hemostasis used to be driven by the INR and platelet count, with traditional cutoffs of platelets over fifty thousand and INR under one and a half, but the modern era has moved toward viscoelastic testing and individualized assessment, because conventional coagulation tests don't capture the rebalanced hemostasis of cirrhosis, where procoagulants and anticoagulants are both reduced and offset each other, so the patient isn't as bleeding-prone as the INR alone suggests. The current guidance is direct: prophylactic plasma, platelet, or cryoprecipitate transfusion before biopsy in cirrhotics with isolated INR elevation or moderate thrombocytopenia isn't supported and may worsen outcomes through volume expansion and increased portal pressure, with viscoelastic testing identifying the subset who truly need correction. Transjugular biopsy is the alternative when percutaneous is contraindicated or when the pressure gradient is needed, for tense ascites, uncorrectable coagulopathy, morbid obesity, suspected vascular liver disease, and the requirement for pressure measurement, performed through the right internal jugular into a hepatic vein, with lower bleeding risk because any bleeding is into the vein and reabsorbed, and its unique value is the hepatic venous pressure gradient: a balloon wedged in a venule gives the wedged pressure reflecting sinusoidal and thus portal pressure, and the gradient is wedged minus free, with normal one to five, subclinical portal hypertension six to nine, clinically significant portal hypertension at ten or more where varices develop, and twelve or more correlating with variceal hemorrhage risk. The pattern of wedged and free pressures also localizes the disease and is tested directly: in pre-sinusoidal portal hypertension both pressures are low and the gradient is normal because the obstruction is upstream of the sinusoid, in sinusoidal disease, the cirrhotic pattern, wedged is high and free is normal so the gradient is high, and in post-sinusoidal disease from right heart failure or Budd-Chiari both are high together and the gradient can be normal because the sinusoidal gradient is preserved. So the cirrhotic patient with new ascites and worsening fatigue who needs both tissue and hemodynamics is the textbook transjugular vignette, and the answer is transjugular biopsy with pressure measurement, not percutaneous biopsy with peri-procedural plasma. Surgical wedge biopsy happens during another operation and can overestimate fibrosis because subcapsular fibrosis is heavier, and endoscopic-ultrasound-guided biopsy has emerged for altered anatomy or patients undergoing that procedure anyway.

16:16So the biopsy decision tree is: percutaneous by default for diffuse disease without ascites, without uncorrectable coagulopathy, and with a reachable window; transjugular for tense ascites, severe uncorrectable coagulopathy, suspected vascular disease, morbid obesity, and when the pressure gradient adds value; surgical wedge during another operation; and endoscopic-ultrasound-guided as the advanced-endoscopy alternative. And the short list of indications still driving biopsy is diagnostic uncertainty after a complete noninvasive workup, antibody-negative primary biliary cholangitis, suspected autoimmune overlap, small-duct sclerosing cholangitis with a normal cholangiogram, suspected granulomatous or infiltrative disease, cases where staging will change management, and acute liver failure of uncertain etiology where transjugular biopsy identifies the cause within hours. So the patient with well over a year of cholestatic chemistries, a normal MRCP, and a fully negative serology panel is the textbook tissue-needed vignette, because biopsy is the only test left that separates small-duct sclerosing cholangitis from antibody-negative primary biliary cholangitis from granulomatous hepatitis.

17:28So the pattern across the episode is the same as the first: pick the right reflex test for the pattern in front of you. Isolated indirect bilirubin runs through a hemolysis screen to Gilbert, isolated direct bilirubin runs through normal imaging to Dubin-Johnson or Rotor, and isolated alkaline phosphatase runs through GGT to either the cholestatic or the bone workup. The imaging modality is chosen for the question it answers, structure or fibrosis, the fibrosis tool is chosen for the etiology and body habitus, and biopsy is the answer when serology and noninvasive tests leave a real question, with the percutaneous-versus-transjugular choice turning on ascites, coagulation, and whether you need the pressure gradient.

18:08Next is viral hepatitis, opening on the acute presentation and separating hepatitis A and E as the fecal-oral self-limited viruses, then hepatitis B interpreted on its serologic panel with treatment guided by disease phase and the special cases of pregnancy and reactivation, and hepatitis C diagnosed by reflex antibody and RNA testing and treated with pangenotypic direct-acting antivirals, with the opening question being what each virus does to the liver and how that biology drives diagnosis and therapy.

18:38For 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.