EUTECNIQAtlas

Every page of EUTECNIQ on one mapArrow keys move through results · Enter opens · Esc closes

Reference library Primary physiology and guidance Peer-reviewed human data Schematics and calculations

The liver, from segments and sinusoids to drug metabolism and liver injury

How the liver is built, what it makes and clears, which enzymes and transporters handle drugs and plant compounds, how liver tests are read, and how clinicians and regulators judge whether a drug or supplement injured the liver. Each value is tied to the source it came from.

75–80%of the liver's blood volume arrives through the portal vein; the hepatic artery supplies the other 20–25%
≈600 mLof bile a day, about 75% from hepatocytes and 25% from the cells lining the bile ducts
27%of P450-catalyzed drug reactions are attributed to CYP3A4 (4,058 reactions, 860 drugs)
7% → 20%share of U.S. Drug-Induced Liver Injury Network cases attributed to herbal and dietary supplements, 2004–05 to 2010–12

Vernon H, et al. StatPearls, Liver anatomy, updated Nov 26, 2022; Hundt M, et al. StatPearls, Bile secretion, updated Sep 26, 2022; Rendic S, Guengerich FP. Chem Res Toxicol 2015;28:38–42; Navarro VJ, et al. Hepatology 2014;60:1399–1408.

Status. Education, not medical advice. EUTECNIQ is in formation and runs no laboratory, clinical service or studies. Figures marked as calculations are educational calculations from published thresholds, not measurements or predictions for any product or person.

Related: ADME, PK and PBPK (first pass, extraction ratio, PBPK structure, midazolam interaction data) · Risk assessment · Contaminants · Supplement–drug interactions

Three-dimensional rendering of a hexagonal liver lobule: plates of hepatocytes radiating from a central vein, with a portal vein, hepatic artery and bile duct at each corner.
  1. 1 Central vein drains to the hepatic veins
  2. 2 Portal triad: portal vein, hepatic artery, bile duct
  3. 3 Plates of hepatocytes; blood flows toward the center

Rendered by EUTECNIQ; generalized and not to scale.

Anatomy

Eight segments, set by blood supply, divide a liver that looks like four lobes

The liver sits under the diaphragm in the right upper abdomen. On its surface it shows four lobes: right, left, caudate and quadrate. The falciform ligament divides it into anatomic right and left along the front, and its free edge carries the round ligament, the remnant of the fetal umbilical vein.

Surgeons and radiologists use a different map. The Couinaud classification divides the liver into eight functionally independent segments, each a wedge with its apex at the hilum. A single branch of the hepatic artery, portal vein and bile duct enters each segment, and the hepatic veins run between segments to the inferior vena cava.

The hepatic veins and the falciform ligament split the liver vertically; the plane of the portal vein splits it into upper and lower segmentsSchematic
Anterior view of the liver: eight Couinaud segments bounded by the hepatic veins, the falciform ligament and the plane of the portal vein, with the inferior vena cava, gallbladder and the vessels and bile ducts at the hilumSegment VIISegment VISegment VIIISegment VSegment IVSegment IISegment IIIVIIVIVIIIVIVIIIIIIInferior vena cavaRight hepatic veinMiddle hepatic veinLeft hepatic veinFalciform ligamentCaudate lobe (segment I),on the posterior surfaceRound ligament(ligamentum teres)Hepatic artery properPortal veinCommon bile ductGallbladder (fundusbelow the margin)Portal vein (inflow)Hepatic arteryBile ductsHepatic veins, IVC (outflow)Faded: inside the liver. Dashed lines: segment boundaries. Schematic, not to scale.Anterior view of the liver: eight Couinaud segments bounded by the hepatic veins, the falciform ligament and the plane of the portal vein, with the inferior vena cava, gallbladder and the vessels and bile ducts at the hilumSegment VIISegment VISegment VIIISegment VSegment IVSegment IISegment IIIVIIVIVIIIVIVIIIIII12345678910111Inferior vena cava2Right hepatic vein3Middle hepatic vein4Left hepatic vein5Falciform ligament6Caudate lobe (segment I), on the posterior surface7Round ligament (ligamentum teres)8Hepatic artery proper9Portal vein10Common bile duct11Gallbladder (fundus below the margin)Portal vein (inflow)Hepatic arteryBile ductsHepatic veins, IVC (outflow)Faded: inside the liver. Dashed lines: segment boundaries.Schematic, not to scale.
Anterior view, schematic, not to scale; the patient's right is on the viewer's left. Segment boundaries drawn from the Couinaud description in Azizaddini S, et al. Diagnostic Imaging of the Liver, StatPearls, updated Feb 20, 2023, and Prater S, Zayas JO, Percutaneous Radiofrequency Ablation of Liver Tumors, StatPearls, updated May 22, 2023; segment II upper and III lower from Sharma M, et al. World J Gastrointest Endosc 2018;10:326–339; lobes and ligaments from Vernon H, et al., StatPearls, updated Nov 26, 2022.

On the functional map, the middle hepatic vein, running from the gallbladder fossa to the vena cava, separates right from left liver; the falciform ligament separates the medial and lateral parts of the left lobe.

Text version: the eight segments and what bounds them
Couinaud segments, their position and the structures that bound them
SegmentPart of the liverBounded by
ICaudate lobe, on the posterior surfaceCan receive blood from both right and left portal vein; drains directly into the inferior vena cava (imaging chapter; the anatomy chapter says it drains into the middle hepatic vein in most people)
IILeft lobe, lateral, upperFalciform ligament (medially); portal plane (below)
IIILeft lobe, lateral, lowerFalciform ligament (medially); portal plane (above)
IVLeft lobe, medialMiddle hepatic vein (right side); falciform ligament (left side)
VRight lobe, anterior, lowerMiddle and right hepatic veins; portal plane (above)
VIRight lobe, posterior, lowerRight hepatic vein; portal plane (above)
VIIRight lobe, posterior, upperRight hepatic vein; portal plane (below)
VIIIRight lobe, anterior, upperMiddle and right hepatic veins; portal plane (below)

Azizaddini S, Tafti D, Mani N. Diagnostic Imaging of the Liver. StatPearls, updated Feb 20, 2023: "The middle hepatic vein runs from the inferior vena cava to the gallbladder fossa and divides the liver into the right and left lobes. The right hepatic vein divides the right lobe into the anterior and posterior segments, and the falciform ligament divides the left lobe into medial and lateral segments. The portal vein divides segments horizontally into superior and inferior segments." Segments II, III and IV form the left lobe, V to VIII the right; the segments are numbered clockwise (Prater S, Zayas JO, StatPearls, updated May 22, 2023). Within the lateral left lobe, segment II is posterior-superior and segment III anterior-inferior, separated by the plane through the portal vein bifurcation (Sharma M, Somani P, Rameshbabu CS, et al. Stepwise evaluation of liver sectors and liver segments by endoscopic ultrasound. World J Gastrointest Endosc 2018;10:326–339).

Blood supply, venous drainage and bile ducts of the liver
StructureWhat the source statesSource
Dual blood supply75% to 80% of blood volume from the portal vein, 20% to 25% from the hepatic arteryStatPearls, Liver anatomy (2022)
OxygenThe portal vein provides 75% of the liver's blood supply but 50% of its oxygen supplyStatPearls, Diagnostic imaging (2023)
Portal veinFormed by the superior mesenteric and splenic veins behind the neck of the pancreas; runs with the proper hepatic artery and common bile duct in the hepatoduodenal ligamentStatPearls, Liver anatomy
Hepatic veinsRight 39.6%, middle 32.7%, left 20.7% of venous drainage; the caudate lobe drains to the middle hepatic vein in most people, sometimes straight into the vena cavaStatPearls, Liver anatomy
Bile ductsRight and left hepatic ducts join as the common hepatic duct; with the cystic duct from the gallbladder it forms the common bile duct, which enters the duodenum at the ampulla of Vater. The left hepatic duct drains segments II, III and IVStatPearls, Biliary ducts (2023)
Countercurrent flowIn each portal triad bile leaves while portal and arterial blood arrive, a countercurrent that the source links to exchange of electrolytes and bile acidsStatPearls, Liver anatomy

Vernon H, Wehrle CJ, Alia VSK, Kasi A. Anatomy, Abdomen and Pelvis: Liver. StatPearls, updated Nov 26, 2022; Azizaddini S, et al. StatPearls, 2023; Hundt M, Wu CY, Young M. Anatomy, Abdomen and Pelvis: Biliary Ducts. StatPearls, updated Jul 15, 2023. The three hepatic-vein shares add to 93% as published. Liver blood flow as a share of cardiac output in the ICRP reference adult (25.5%) is on ADME, PK and PBPK.

Microanatomy

Blood runs from the portal triads to the central vein, and bile runs the other way

Hepatocytes form plates around a central vein in a roughly hexagonal lobule, with a portal triad at each corner: a portal-vein branch, a hepatic-artery branch and a bile duct. The blood of both vessels mixes in the sinusoids, the capillaries between the plates, and drains to the central vein, a branch of the hepatic vein.

The portal acinus describes the same tissue by blood flow. It is the rhombus between two portal triads and the two central veins they face. Zone 1 lies next to the triads, zone 3 around the central veins and zone 2 between them. Oxygen falls along this path, which is one reason the zones do different work.

Periportal blood carries about twice the oxygen tension of blood leaving past the central veinSchematic and published ranges
Two classic hexagonal lobules with a central vein each and portal triads at the corners; the portal acinus joins two triads and two central veins and is divided into zones 1, 2 and 3; published oxygen tension falls from about 60 to 65 mm Hg in periportal blood to about 30 to 35 mm Hg in perivenous bloodZone 3 (pericentral, perivenous)Central vein (terminal hepatic venule)Central vein (terminal hepatic venule)Portal venule (branch of the portal vein)Hepatic arterioleBile ductulePortal venule (branch of the portal vein)Hepatic arterioleBile ductulePortal venule (branch of the portal vein)Hepatic arterioleBile ductulePortal venule (branch of the portal vein)Hepatic arterioleBile ductulePortal venule (branch of the portal vein)Hepatic arterioleBile ductulePortal venule (branch of the portal vein)Hepatic arterioleBile ductulePortal venule (branch of the portal vein)Hepatic arterioleBile ductulePortal venule (branch of the portal vein)Hepatic arterioleBile ductulePortal venule (branch of the portal vein)Hepatic arterioleBile ductulePortal venule (branch of the portal vein)Hepatic arterioleBile ductulePortal venule (branch of the portal vein)Hepatic arterioleBile ductulePortal venule (branch of the portal vein)Hepatic arterioleBile ductule12323Central veinPortal triadBlood: triad → central veinBile: canaliculi → triad bile ductuleHexagon: classic lobuleRhombus: portal acinus, zones 1–3Oxygen tension, mm Hg010203040506070Blood, periportal: 60–65 mm HgBlood, perivenous: 30–35 mm HgHepatocytes, periportal: 45–50 mm HgHepatocytes, perivenous: 15–20 mm HgZone 1Zone 3periportalperivenousBloodHepatocytesTwo classic hexagonal lobules with a central vein each and portal triads at the corners; the portal acinus joins two triads and two central veins and is divided into zones 1, 2 and 3; published oxygen tension falls from about 60 to 65 mm Hg in periportal blood to about 30 to 35 mm Hg in perivenous bloodZone 3 (pericentral, perivenous)Central vein (terminal hepatic venule)Central vein (terminal hepatic venule)Portal venule (branch of the portal vein)Hepatic arterioleBile ductulePortal venule (branch of the portal vein)Hepatic arterioleBile ductulePortal venule (branch of the portal vein)Hepatic arterioleBile ductulePortal venule (branch of the portal vein)Hepatic arterioleBile ductulePortal venule (branch of the portal vein)Hepatic arterioleBile ductulePortal venule (branch of the portal vein)Hepatic arterioleBile ductulePortal venule (branch of the portal vein)Hepatic arterioleBile ductulePortal venule (branch of the portal vein)Hepatic arterioleBile ductulePortal venule (branch of the portal vein)Hepatic arterioleBile ductulePortal venule (branch of the portal vein)Hepatic arterioleBile ductulePortal venule (branch of the portal vein)Hepatic arterioleBile ductulePortal venule (branch of the portal vein)Hepatic arterioleBile ductule12323Blood: triad → central veinBile: canaliculi → triad bile ductuleHexagon: lobule. Rhombus: acinus, zones 1–3Oxygen tension, mm Hg010203040506070Blood, periportal: 60–65 mm HgBlood, perivenous: 30–35 mm HgHepatocytes, periportal: 45–50 mm HgHepatocytes, perivenous: 15–20 mm HgZone 1Zone 3periportalperivenousBloodHepatocytes
Lobule and acinus: schematic, not to scale, after Kietzmann T. Redox Biol 2017;11:622–630, Fig. 1, and Vernon H, et al., StatPearls 2022. Oxygen: ranges as stated by Kietzmann 2017 (blood 60–65 mm Hg periportal, 30–35 mm Hg perivenous; cells about 15 mm Hg lower). The dashed lines only join the two ranges; the profile between them is not drawn from data.
Show data: oxygen tension along the acinus
Oxygen partial pressure in blood and hepatocytes at the periportal and perivenous ends of the acinus
CompartmentPeriportal (zone 1)Perivenous (zone 3)
Blood60–65 mm Hg (84–91 µmol/L)30–35 mm Hg (42–49 µmol/L)
Hepatocytes (intracellular)45–50 mm Hg15–20 mm Hg

Kietzmann T. Metabolic zonation of the liver: the oxygen gradient revisited. Redox Biol 2017;11:622–630, section 3.

A sinusoid is lined by perforated endothelium; the space behind it holds stellate cells and hepatocyte microvilliSchematic
Longitudinal section of a hepatic sinusoid: plates of hepatocytes on both sides, the space of Disse with hepatocyte microvilli and a stellate cell, the fenestrated endothelium, a Kupffer cell in the lumen, and bile canaliculi between neighboring hepatocytes; blood flows toward the central vein and bile toward the portal triadHepatocyteNucleusHepatocyteNucleusHepatocyteNucleusHepatocyteNucleusHepatocyteNucleusHepatocyteNucleusHepatocyteNucleusHepatocyteNucleusHepatocyteNucleusHepatocyteNucleusHepatocyteNucleusHepatocyteNucleusBile canaliculusBile canaliculusBile canaliculusBile canaliculusBile canaliculusBile canaliculusBile canaliculusBile canaliculusBile canaliculusBile canaliculusSpace of DisseSpace of DisseSinusoidal endothelial cellSinusoidal endothelial cellSinusoidal endothelial cellSinusoidal endothelial cellSinusoidal endothelial cellSinusoidal endothelial cellSinusoidal endothelial cellSinusoidal endothelial cellSinusoidal endothelial cellSinusoidal endothelial cellSinusoidal endothelial cellSinusoidal endothelial cellSinusoidal endothelial cellSinusoidal endothelial cellSinusoidal endothelial cellSinusoidal endothelial cellSinusoidal endothelial cellSinusoidal endothelial cellSinusoidal endothelial cellSinusoidal endothelial cellSinusoidal endothelial cellSinusoidal endothelial cellSinusoidal endothelial cellSinusoidal endothelial cellSinusoidal endothelial cellSinusoidal endothelial cellSinusoidal endothelial cellSinusoidal endothelial cellSinusoidal endothelial cellSinusoidal endothelial cellSinusoidal endothelial cellSinusoidal endothelial cellSinusoidal endothelial cellSinusoidal endothelial cellSinusoidal endothelial cellSinusoidal endothelial cellSinusoidal endothelial cellSinusoidal endothelial cellSinusoidal endothelial cellSinusoidal endothelial cellSinusoidal endothelial cellSinusoidal endothelial cellSinusoidal endothelial cellSinusoidal endothelial cellSinusoidal endothelial cellSinusoidal endothelial cellSinusoidal endothelial cellSinusoidal endothelial cellSinusoidal endothelial cellSinusoidal endothelial cellSinusoidal endothelial cellSinusoidal endothelial cellSinusoidal endothelial cellSinusoidal endothelial cellSinusoidal endothelial cellSinusoidal endothelial cellSinusoidal endothelial cellSinusoidal endothelial cellSinusoidal endothelial cellSinusoidal endothelial cellHepatic stellate cellLipid droplet (retinoid store)Lipid droplet (retinoid store)Lipid droplet (retinoid store)Lipid droplet (retinoid store)Kupffer cell (resident macrophage)blood → central vein (zone 3)bile → portal triad bile ductuleHepatocyte plateBile canaliculusSpace of DisseStellate cell withlipid dropletsFenestratedendotheliumKupffer cellPortal triad sideCentral vein sideSchematic, not to scale. Fenestra diameter in human liver without pathology: 107 ± 1.5 nm.Longitudinal section of a hepatic sinusoid: plates of hepatocytes on both sides, the space of Disse with hepatocyte microvilli and a stellate cell, the fenestrated endothelium, a Kupffer cell in the lumen, and bile canaliculi between neighboring hepatocytes; blood flows toward the central vein and bile toward the portal triadHepatocyteNucleusHepatocyteNucleusHepatocyteNucleusHepatocyteNucleusHepatocyteNucleusHepatocyteNucleusHepatocyteNucleusHepatocyteNucleusBile canaliculusBile canaliculusBile canaliculusBile canaliculusBile canaliculusBile canaliculusSpace of DisseSpace of DisseSinusoidal endothelial cellSinusoidal endothelial cellSinusoidal endothelial cellSinusoidal endothelial cellSinusoidal endothelial cellSinusoidal endothelial cellSinusoidal endothelial cellSinusoidal endothelial cellSinusoidal endothelial cellSinusoidal endothelial cellSinusoidal endothelial cellSinusoidal endothelial cellSinusoidal endothelial cellSinusoidal endothelial cellSinusoidal endothelial cellSinusoidal endothelial cellSinusoidal endothelial cellSinusoidal endothelial cellSinusoidal endothelial cellSinusoidal endothelial cellSinusoidal endothelial cellSinusoidal endothelial cellSinusoidal endothelial cellSinusoidal endothelial cellSinusoidal endothelial cellSinusoidal endothelial cellSinusoidal endothelial cellSinusoidal endothelial cellSinusoidal endothelial cellSinusoidal endothelial cellSinusoidal endothelial cellSinusoidal endothelial cellHepatic stellate cellLipid droplet (retinoid store)Lipid droplet (retinoid store)Lipid droplet (retinoid store)Lipid droplet (retinoid store)Kupffer cell (resident macrophage)blood → central vein (zone 3)bile → portal triad bile ductule1234561Hepatocyte plate2Bile canaliculus3Space of Disse4Stellate cell with lipid droplets5Fenestrated endothelium6Kupffer cellSchematic, not to scale. Human fenestrae: 107 ± 1.5 nm.
Schematic, not to scale, after Kietzmann 2017, Fig. 1B; Braet F, Wisse E. Comp Hepatol 2002;1:1; Wisse E, et al. Gene Ther 2008;15:1193–1199 (human fenestra diameter); Kalra A, et al. Physiology, Liver, StatPearls 2023.
Cells and structures of the hepatic sinusoid and what the sources state about each
StructureWhat it is and doesNumbers
Sinusoidal endothelial cellsLine the sinusoid; holes (fenestrae) grouped in sieve plates let fluid, solutes and particles smaller than the holes reach the space of DisseHuman liver without pathology: fenestra diameter 107 ± 1.5 nm. Review figure for scanning EM in general: 150–175 nm, 9–13 per µm², 6–8% of the surface, with porosity rising from 6% periportal to 8% centrilobular
Space of DisseGap between endothelium and hepatocytes, filled with hepatocyte microvilli and extracellular matrix; hepatocytes have no true basement membrane—
Stellate (Ito) cellsSit in the space of Disse and store vitamin A in lipid droplets; when activated by injury they lose the droplets, and they secrete TGF-β in cirrhosisMost of the body's retinoid is held in these droplets
Kupffer cellsResident macrophages found in the sinusoid—
Bile canaliculiChannels formed by the apical membranes of neighboring hepatocytes; they make a "chicken-wire" lattice and carry bile toward the portal triadThe canalicular membrane is about 1% of the hepatocyte surface
Chylomicron remnantsMust pass the fenestrae to reach hepatocytesMean diameter 90–250 nm

Braet F, Wisse E. Comp Hepatol 2002;1:1 (sieve plates, scanning EM values, remnants); Wisse E, Jacobs F, Topal B, Frederik P, De Geest B. Gene Ther 2008;15:1193–1199; Blaner WS, et al. Biochim Biophys Acta 2009;1791:467–473; Kietzmann 2017; Kalra A, et al. StatPearls, Physiology, Liver, updated May 1, 2023; Hundt M, et al. StatPearls, Bile secretion, 2022.

In human liver, glutamine synthetase and several drug-metabolizing P450s are pericentral; enzymes the authors link to urea production are periportalPublished human data
Fold enrichment of selected transcripts in zone 1 or zone 3 of normal human liver, measured by laser capture microdissection and RNA sequencing: urea-cycle feeder enzymes periportal, glutamine synthetase and drug-metabolizing enzymes pericentral← higher in zone 1 (periportal)higher in zone 3 (pericentral) →2×2×4×4×8×8×16×16×32×32×64×64×HAL (histidine ammonia-lyase): 29.12-fold higher in zone 129.12×HALhistidine ammonia-lyaseGLS2 (glutaminase 2): 19.87-fold higher in zone 119.87×GLS2glutaminase 2SDS (serine dehydratase): 11.5-fold higher in zone 111.50×SDSserine dehydrataseCYP2A7 (cytochrome P450 2A7): 9.0-fold higher in zone 19.00×CYP2A7cytochrome P450 2A7RHBG (ammonium transporter Rh B): 37.93-fold higher in zone 337.93×RHBGammonium transporter Rh BCYP1A2 (cytochrome P450 1A2): 13.58-fold higher in zone 313.58×CYP1A2cytochrome P450 1A2SLCO1B3 (OATP1B3 uptake transporter): 12.8-fold higher in zone 312.80×SLCO1B3OATP1B3 uptake transporterCYP3A4 (cytochrome P450 3A4): 11.0-fold higher in zone 311.00×CYP3A4cytochrome P450 3A4CYP2C19 (cytochrome P450 2C19): 9.88-fold higher in zone 39.88×CYP2C19cytochrome P450 2C19GLUL (glutamine synthetase): 8.91-fold higher in zone 38.91×GLULglutamine synthetaseUGT1A3 (UDP-glucuronosyltransferase): 4.02-fold higher in zone 34.02×UGT1A3UDP-glucuronosyltransferaseCYP2E1 (cytochrome P450 2E1): 3.67-fold higher in zone 33.67×CYP2E1cytochrome P450 2E11×Fold change, zone-enriched over the other zone (log scale)Fold enrichment of selected transcripts in zone 1 or zone 3 of normal human liver (laser capture microdissection and RNA sequencing)Higher in zone 1 (periportal)HAL, histidine ammonia-lyaseHAL: 29.12-fold29.12×GLS2, glutaminase 2GLS2: 19.87-fold19.87×SDS, serine dehydrataseSDS: 11.5-fold11.50×CYP2A7, cytochrome P450 2A7CYP2A7: 9.0-fold9.00×Higher in zone 3 (pericentral)RHBG, ammonium transporter Rh BRHBG: 37.93-fold37.93×CYP1A2, cytochrome P450 1A2CYP1A2: 13.58-fold13.58×SLCO1B3, OATP1B3 uptake transporterSLCO1B3: 12.8-fold12.80×CYP3A4, cytochrome P450 3A4CYP3A4: 11.0-fold11.00×CYP2C19, cytochrome P450 2C19CYP2C19: 9.88-fold9.88×GLUL, glutamine synthetaseGLUL: 8.91-fold8.91×UGT1A3, UDP-glucuronosyltransferaseUGT1A3: 4.02-fold4.02×CYP2E1, cytochrome P450 2E1CYP2E1: 3.67-fold3.67×Bar length: fold change on a log scale (2× to 64×)
McEnerney L, et al. Exp Mol Med 2017;49:e413, Tables 1 and 2: laser capture microdissection and RNA sequencing of zones 1 and 3 from normal liver removed at hepatectomy for an isolated metastasis; 139 transcripts differed between zones 1 and 3. Values are mRNA ratios, not protein amounts or enzyme activity.

Zone 3 receives the least oxygen and, in this human study, carries more CYP1A2, CYP3A4, CYP2C19 and CYP2E1 mRNA than zone 1.

Explanation, details and sources

Zonation separates opposing pathways and is not fixed: its pattern changes with nutrition, drugs and hormones, and even glutamine synthetase, long treated as static, can extend under thyroid hormone or R-spondin signals (Kietzmann 2017). In the human study, zone 1 enrichment of serine dehydratase (SDS) and glutaminase 2 (GLS2) and zone 3 enrichment of glutamine synthetase (GLUL) and the ammonium transporter RHBG led the authors to conclude that zone 1 specializes in urea production while zone 3 uses ammonia to make glutamine, as in mice. StatPearls places the biotransformation of drugs mainly in zone 3. Germander injury, for example, shows prominent centrilobular necrosis on biopsy (LiverTox).

Fold enrichment of selected transcripts in zone 1 or zone 3 of human liver
GeneEncodesHigher inFold change
HALHistidine ammonia-lyaseZone 129.12
GLS2Glutaminase 2Zone 119.87
SDSSerine dehydrataseZone 111.50
CYP2A7Cytochrome P450 2A7Zone 19.00
RHBGRh family B glycoproteinZone 337.93
CYP1A2Cytochrome P450 1A2Zone 313.58
SLCO1B3OATP1B3 uptake transporterZone 312.80
CYP3A4Cytochrome P450 3A4Zone 311.00
CYP2C19Cytochrome P450 2C19Zone 39.88
GLULGlutamine synthetase (glutamate-ammonia ligase)Zone 38.91
UGT1A3UDP-glucuronosyltransferase 1A3Zone 34.02
CYP2E1Cytochrome P450 2E1Zone 33.67

McEnerney L, Duncan K, Bang BR, et al. Dual modulation of human hepatic zonation via canonical and non-canonical Wnt pathways. Exp Mol Med 2017;49:e413 (RNA-seq data GSE83990); Kietzmann 2017; Vernon H, et al. StatPearls 2022; LiverTox, Germander, updated Mar 12, 2018. UniProt lists OATP1B1 and OATP1B3 at the basolateral membrane of hepatocytes near the central vein.

Functions

One organ makes plasma proteins and bile, stores vitamins and clears heme

The liver makes up around 2% of adult body weight. It is both exocrine, secreting bile, and endocrine, taking part in glucose control with insulin and glucagon.

What the liver does, with the published numbers
FunctionWhat happensPublished numbersSource
Protein synthesisMakes nearly every plasma protein, including albumin, binding globulins, protein C, protein S and every clotting factor except factor VIII. Factors II, VII, IX and X and proteins C and S need vitamin K for γ-carboxylationAlbumin: 10–15 g a dayStatPearls, Physiology, Liver; Physiology, Albumin (updated Jan 31, 2026)
BileHepatocytes secrete bile acids, cholesterol, phospholipid and bilirubin into the canaliculi; bile is stored and concentrated in the gallbladder; most bile acids return from the ileum to the liver≈600 mL a day (75% hepatocytes, 25% cholangiocytes); about 225 mL a day of the hepatocyte part is bile-salt-dependent; about 5% of bile acids are excretedStatPearls, Bile secretion (2022)
Bile-acid synthesisP450 oxidation of cholesterol; cholesterol 7α-hydroxylase (CYP7A1) is the rate-limiting stepThis classic pathway is 95% of bile-acid synthesisStatPearls, Physiology, Bile Acids (2023)
Heme and bilirubinTakes up albumin-bound unconjugated bilirubin, conjugates it with glucuronic acid (UGT1A1) and secretes it into bileAbout 4 mg/kg of bilirubin is produced daily, roughly 80% from hemoglobin of old red cellsStatPearls, Physiology, Bilirubin (2022); Physiology, Liver
CarbohydrateStores glycogen, at the highest concentration per gram of any tissue; glucose-6-phosphatase lets it release free glucose into the blood; makes glucose by gluconeogenesis—StatPearls, Biochemistry, Glycogen (2026); Liver anatomy
NitrogenConverts ammonia to urea; zone 3 hepatocytes also fix ammonia as glutamine—StatPearls, Liver anatomy; McEnerney 2017
Storage and activationStores vitamin A (in stellate cells), iron and copper; 25-hydroxylates vitamin D by P450; deiodinates T4 to T3—StatPearls, Physiology, Liver
Clearance of foreign compoundsTurns lipophilic compounds into water-soluble ones by phase I and phase II reactions, mostly in the smooth endoplasmic reticulum, then moves them into bile or bloodSee the next three sectionsStatPearls, Physiology, Liver

Kalra A, Yetiskul E, Wehrle CJ, Tuma F. Physiology, Liver. StatPearls, updated May 1, 2023; Moman RN, et al. Physiology, Albumin, updated Jan 31, 2026; Hundt M, Basit H, John S. Physiology, Bile Secretion, updated Sep 26, 2022; Chen I, Cassaro S. Physiology, Bile Acids, updated May 1, 2023; Kalakonda A, et al. Physiology, Bilirubin, updated Sep 12, 2022; Daghlas SA, Rahimi N. Biochemistry, Glycogen, updated Mar 21, 2026 (all NCBI Bookshelf).

Phase I

Five P450s carry most drug oxidation, and CYP3A4 the largest share

Cytochrome P450 enzymes (CYPs; EC 1.14.14.1) insert one oxygen atom from O2 into a substrate and reduce the other to water, using electrons from NADPH. Of 57 putatively functional human CYPs, about a dozen in families 1, 2 and 3 handle most foreign compounds, including 70–80% of drugs in clinical use. The basics of first-pass metabolism and the extraction ratio are on ADME, PK and PBPK.

CYP3A4 leads for drugs and for natural compounds, but natural compounds spread across many more P450sPublished literature survey
Percentage of P450-catalyzed reactions attributed to each human P450, for 860 drugs and for 331 natural and physiological compounds; CYP3A4 leads both0%10%20%30%Drugs (4,058 reactions)Natural and physiological compounds (1,601)CYP3A4CYP3A4, drugs: 27%27%CYP3A4, natural and physiological compounds: 15%15%CYP2D6CYP2D6, drugs: 13%13%CYP2D6, natural and physiological compounds: 6%6%CYP2C9CYP2C9, drugs: 10%10%CYP2C9, natural and physiological compounds: 8%8%CYP2C19CYP2C19, drugs: 9%9%CYP2C19, natural and physiological compounds: 6%6%CYP1A2CYP1A2, drugs: 9%9%CYP1A2, natural and physiological compounds: 8%8%CYP3A5CYP3A5, drugs: 6%6%CYP3A5, natural and physiological compounds: 3%3%CYP1A1CYP1A1, drugs: 5%5%CYP1A1, natural and physiological compounds: 7%7%CYP2C8CYP2C8, drugs: 5%5%CYP2C8, natural and physiological compounds: 3%3%CYP2B6CYP2B6, drugs: 4%4%CYP2B6, natural and physiological compounds: 3%3%CYP2E1CYP2E1, drugs: 3%3%CYP2E1, natural and physiological compounds: 4%4%CYP2A6CYP2A6, drugs: 2%2%CYP2A6, natural and physiological compounds: 4%4%CYP1B1CYP1B1, drugs: 1%1%CYP1B1, natural and physiological compounds: 4%4%Other P450sOther P450s, drugs: 6%6%Other P450s, natural and physiological compounds: 29%29%Share of reactions, not of compounds; one compound can count under several P450sPercentage of P450-catalyzed reactions attributed to each human P450, for 860 drugs and for 331 natural and physiological compounds; CYP3A4 leads both0%10%20%30%DrugsNatural and physiological compoundsCYP3A4CYP3A4, drugs: 27%27%CYP3A4, natural and physiological compounds: 15%15%CYP2D6CYP2D6, drugs: 13%13%CYP2D6, natural and physiological compounds: 6%6%CYP2C9CYP2C9, drugs: 10%10%CYP2C9, natural and physiological compounds: 8%8%CYP2C19CYP2C19, drugs: 9%9%CYP2C19, natural and physiological compounds: 6%6%CYP1A2CYP1A2, drugs: 9%9%CYP1A2, natural and physiological compounds: 8%8%CYP3A5CYP3A5, drugs: 6%6%CYP3A5, natural and physiological compounds: 3%3%CYP1A1CYP1A1, drugs: 5%5%CYP1A1, natural and physiological compounds: 7%7%CYP2C8CYP2C8, drugs: 5%5%CYP2C8, natural and physiological compounds: 3%3%CYP2B6CYP2B6, drugs: 4%4%CYP2B6, natural and physiological compounds: 3%3%CYP2E1CYP2E1, drugs: 3%3%CYP2E1, natural and physiological compounds: 4%4%CYP2A6CYP2A6, drugs: 2%2%CYP2A6, natural and physiological compounds: 4%4%CYP1B1CYP1B1, drugs: 1%1%CYP1B1, natural and physiological compounds: 4%4%Other P450sOther P450s, drugs: 6%6%Other P450s, natural and physiological compounds: 29%29%Share of reactions, not of compounds; onecompound can count under several P450s
Rendic S, Guengerich FP. Chem Res Toxicol 2015;28:38–42, Table 1 (drugs: 4,058 reactions, 860 drugs; natural and physiological compounds: 1,601 reactions, 331 compounds). Shares of reported reactions; an enzyme's share does not equal its contribution to clearance.

For natural and physiological compounds, 29% of reported P450 reactions fall outside the twelve P450s listed, against 6% for drugs.

Explanation, details and sources

The highest-expressed liver P450s are CYP3A4, 2C9, 2C8, 2E1 and 1A2; 2A6, 2D6, 2B6, 2C19 and 3A5 are less abundant, and 2J2, 1A1 and 1B1 are mainly extrahepatic (Zanger and Schwab 2013). Earlier analyses of the 200 most prescribed drugs found that about 75% are cleared mainly by P450s, about 90% of P450 reactions belong to CYP1A2, 2C9, 2C19, 2D6 and 3A4, and about 46% to CYP3A enzymes (as summarized by Rendic and Guengerich). StatPearls gives CYP2D6 and CYP3A4/5 roles in about 20% and 30% of clinically used drugs.

The authors' caveats: the analysis covers only oxidation and reduction; counts reflect the compounds studied most; listing an enzyme does not give its share of clearance; CYP3A4 and 3A5 overlap; and the fractions may change as orphan P450s are characterized.

Percentage of P450-catalyzed reactions by enzyme for drugs and for natural and physiological compounds
P450Drugs, %Natural and physiological compounds, %
CYP3A42715
CYP2D6136
CYP2C9108
CYP2C1996
CYP1A298
CYP3A563
CYP1A157
CYP2C853
CYP2B643
CYP2E134
CYP2A624
CYP1B114
Other P450s629

Rendic S, Guengerich FP. Survey of human oxidoreductases and cytochrome P450 enzymes involved in the metabolism of xenobiotic and natural chemicals. Chem Res Toxicol 2015;28:38–42; Zanger UM, Schwab M. Pharmacol Ther 2013;138:103–141 (abstract); Patel T, Rahimi N, Cassagnol M. Biochemistry, Cytochrome P450. StatPearls, updated Jan 10, 2026; UniProt P11712, P05177 (EC numbers).

Phase I, beyond P450

Flavin monooxygenases, amine oxidases, dehydrogenases and esterases handle much of the rest

In the same survey, P450s catalyzed about 96% of drug oxidation and reduction reactions; FMO, monoamine oxidase and aldo-keto reductases shared the remainder. Outside oxidation and reduction, UDP-glucuronosyltransferases and esterases were the main non-P450 enzymes in an earlier analysis of drugs.

Non-P450 phase I enzymes: reaction, location and published share
Enzyme (gene, EC)ReactionWhere in the cellNotes from the sources
Flavin-containing monooxygenase 3 (FMO3; EC 1.14.13.8)Oxygenation of nitrogen- and sulfur-containing drugs and dietary compoundsEndoplasmic reticulum (microsomes); liverFMO: about 2% of drug oxidation–reduction reactions
Monoamine oxidase A and B (MAOA, MAOB; EC 1.4.3.4)Oxidative deamination of amines, including neurotransmitters and exogenous aminesOuter mitochondrial membraneMAO: about 1% of drug oxidation–reduction reactions
Aldo-keto reductases (AKR)NADPH-dependent reduction of carbonyl compounds to alcohols (AKR1B1 as example)—About 1% for drugs, about 4% for natural and physiological compounds
Aldehyde oxidase (AOX1; EC 1.2.3.1)Oxidation of a broad range of drug and xenobiotic substratesCytoplasm; abundant in liver—
Alcohol and aldehyde dehydrogenases (ADH, ALDH2)ADH and ALDH are the primary enzymes of alcohol metabolism; ALDH2 oxidizes acetaldehyde derived from ethanolADH cytoplasm; ALDH2 mitochondriaAn inactive ALDH2 form, common in Asian populations, impairs acetaldehyde metabolism and causes flushing, nausea and vomiting
Carboxylesterases (CES1, CES2)Hydrolysis of esters; activation of ester and amide prodrugsEndoplasmic reticulum lumenCES1 mainly in liver; CES2 mainly in the small intestine, with moderate liver expression
Microsomal epoxide hydrolase (EPHX1; EC 3.3.2.9)Adds water to reactive epoxides, giving less reactive dihydrodiolsEndoplasmic reticulum—

Rendic and Guengerich 2015 (shares, Table 1, and the earlier analysis by Williams et al. that they cite); UniProt entries P31513 (FMO3), P27338 (MAOB), Q06278 (AOX1), P00325 (ADH1B), P05091 (ALDH2), P23141 (CES1), O00748 (CES2), P07099 (EPHX1), P15121 (AKR1B1), accessed Oct 29, 2026; Wen AG, et al. Pharmacogenomics Overview, StatPearls (ADH and ALDH).

Phase II

Conjugation makes a molecule more water-soluble for secretion into bile or blood

Phase II enzymes join a drug or its phase I metabolite to glucuronic acid, sulfate, glutathione, an acetyl or a methyl group. Glucuronidation happens in the endoplasmic reticulum; sulfation and most glutathione conjugation happen in the cytosol. When reduced glutathione is used up, toxic metabolites can build up, as in acetaminophen overdose.

Phase II enzyme families, donors, locations and clinically documented variants
Family (example, EC)Group transferred (donor)WhereTypical substrates and variants
UDP-glucuronosyltransferases (UGT1A1; EC 2.4.1.17)Glucuronic acid (UDP-glucuronic acid)Endoplasmic reticulum; liver, colon, small intestineBilirubin, steroids, thyroid hormones, morphine, acetaminophen. Reduced UGT1A1 activity causes the mild hyperbilirubinemia of Gilbert syndrome; UGT1A1*28/*28 impairs conjugation of SN-38, which links UGT1A1 to irinotecan toxicity
Sulfotransferases (SULT1A1; EC 2.8.2.1)Sulfonate (PAPS)Cytosol; liver, lung, adrenal, brain, platelets, skinMolecules with a hydroxyl or amine group
Glutathione S-transferases (GSTA1, GSTM1; EC 2.5.1.18)GlutathioneCytosol (a minority of conjugation in mitochondria)Electrophilic compounds
N-acetyltransferases (NAT1, NAT2; EC 2.3.1.5)Acetyl (acetyl-CoA)CytosolHydrazine and arylamine drugs such as isoniazid; NAT2 genotype separates rapid, intermediate and poor (slow) acetylators
Catechol O-methyltransferase (COMT; EC 2.1.1.6)Methyl (S-adenosylmethionine)Cytosol and membrane; brain, liver, placenta, blood cellsCatecholamines, catechol hormones, L-DOPA, α-methyldopa
Thiopurine S-methyltransferase (TPMT; EC 2.1.1.67)Methyl (S-adenosylmethionine)CytosolAzathioprine, mercaptopurine, thioguanine. Low-activity variants raise the risk of myelosuppression; CPIC estimates 0.2% of Europeans are poor metabolizers and 8.7% intermediate

Kalra A, et al. StatPearls, Physiology, Liver (2023); Wen AG, Wen S, Patel P. Pharmacogenomics Overview. StatPearls (NBK617055); UniProt P22309, P50225, P08263, P09488, P18440, P11245, P21964, P51580, accessed Oct 29, 2026; CPIC TPMT frequency table (file of Oct 2026) and Relling MV, et al. Clin Pharmacol Ther 2019;105:1095–1105.

Phase III

Transporters decide what enters the hepatocyte and which way metabolites leave

The hepatocyte is polarized. Its basolateral (sinusoidal) membrane faces the blood and carries uptake carriers of the solute carrier (SLC) family; its canalicular membrane faces bile and carries ATP-binding cassette (ABC) pumps that secrete against a concentration gradient. Some authors call this transport step phase III.

Uptake on the blood side, metabolism inside, pumping into bile on the other sideSchematic
A hepatocyte between the sinusoidal blood and the bile canaliculus: uptake transporters on the basolateral membrane, phase I and phase II enzymes inside, efflux transporters on the canalicular membrane and back to bloodSinusoidal blood and space of DisseBlood (sinusoid)HepatocyteHepatocyteBile canaliculusBile (canaliculus)basolateral (sinusoidal) membranecanalicular (apical) membraneNTCPNTCPconjugated bile acids,with Na⁺OATP1B1OATP1B1organic anions, e.g.statinsOATP1B3OATP1B3organic anionsOCT1OCT1organic cationsMRP3MRP3conjugates back tobloodBSEPBSEPbile saltsMRP2MRP2organic anionconjugatesP-gp (MDR1)P-gp (MDR1)drugsBCRPBCRPdrugs, conjugatesMDR3MDR3phospholipidPhase I: oxidation, reduction, hydrolysisCYPs, FMO3, epoxide hydrolase (endoplasmicreticulum); MAO (mitochondria); AOX1, ADH(cytosol); CES1 esterasePhase II: conjugationUGTs (endoplasmic reticulum); SULTs, GSTs,NAT1/2, COMT, TPMT (cytosol)Parent compound or metabolite leaves by thecanalicular pumps (bile) or MRP3 (blood)A hepatocyte between the sinusoidal blood (top) and the bile canaliculus (bottom): uptake and efflux transporters on its two membranes and the phase I and II enzymes insideBlood (sinusoid)NTCPNTCP↓ conjugated bile acids, with Na⁺OATP1B1OATP1B1↓ organic anions, e.g. statinsOATP1B3OATP1B3↓ organic anionsOCT1OCT1↓ organic cationsMRP3MRP3↑ conjugates back to bloodHepatocytePhase I: oxidation, reduction, hydrolysisCYPs, FMO3, epoxide hydrolase (endoplasmicreticulum); MAO (mitochondria); AOX1, ADH(cytosol); CES1 esterasePhase II: conjugationUGTs (endoplasmic reticulum); SULTs, GSTs,NAT1/2, COMT, TPMT (cytosol)Basolateral membrane faces the blood; canalicularmembrane faces the bileBile (canaliculus)BSEPBSEP↓ bile saltsMRP2MRP2↓ organic anion conjugatesP-gp (MDR1)P-gp (MDR1)↓ drugsBCRPBCRP↓ drugs, conjugatesMDR3MDR3↓ phospholipid
Schematic. Membrane locations from UniProt (SLCO1B1 Q9Y6L6, SLCO1B3 Q9NPD5, SLC22A1 O15245, ABCC3 O15438, ABCB11 O95342, ABCC2 Q92887, ABCB1 P08183, ABCG2 Q9UNQ0) and Hundt M, et al., StatPearls, Bile secretion; enzyme locations from UniProt; roles from Galetin A, et al. Nat Rev Drug Discov 2024;23:255–280.
Hepatic drug and bile-acid transporters, their membrane and documented clinical relevance
Transporter (gene)MembraneMovesDocumented relevance
NTCP (SLC10A1)BasolateralConjugated bile acids into the cell, driven by the sodium gradient—
OATP1B1 (SLCO1B1)Basolateral, centrilobular hepatocytesOrganic anions into the cellDecreased-function variants raise exposure to most statins and can lead to myopathy; the review notes that the flavonoid quercetin inhibits OATP1B1-mediated statin transport, while increases in pravastatin exposure with quercetin in healthy participants were modest
OATP1B3 (SLCO1B3)Basolateral, near the central veinOrganic anions into the cellListed with OATP1B1 among transporters to evaluate for drug interactions (ICH M12)
OCT1 (SLC22A1)BasolateralOrganic cations into the cellEMA guidance asks sponsors to consider OCT1 inhibition
MRP3 (ABCC3)BasolateralConjugates back to the bloodHigher plasma glucuronides in NASH are consistent with less MRP2 and more MRP3 efflux
BSEP (ABCB11)CanalicularBile salts into bileInherited defects cause PFIC type 2; inhibition by drugs is associated with cholestasis and drug-induced liver injury
MDR3 (ABCB4)CanalicularPhospholipid into bileInherited defects cause PFIC type 3
MRP2 (ABCC2)Canalicular (apical)Multispecific organic anions, including conjugates—
P-glycoprotein, MDR1 (ABCB1)Canalicular (apical)Drugs out of the cellInducible through PXR, the receptor that also regulates CYP3A4; St John's wort raised duodenal P-gp 1.4-fold in men
BCRP (ABCG2)ApicalDrugs and conjugatesListed among transporters to evaluate for drug interactions (ICH M12)

Galetin A, Brouwer KLR, Tweedie D, et al. Membrane transporters in drug development and as determinants of precision medicine. Nat Rev Drug Discov 2024;23:255–280 (PMC11464068); ICH M12 Drug Interaction Studies, Step 4, May 21, 2024; UniProt subcellular-location entries (accessed Oct 29, 2026); Hundt M, et al. StatPearls, Bile secretion; Kalra A, et al. StatPearls, Physiology, Liver; Dürr D, et al. Clin Pharmacol Ther 2000;68:598–604. The transporter roster for intestine and blood–brain barrier is on ADME, PK and PBPK.

Genetics

Inherited variants make one person a poor metabolizer and another ultrarapid

CPIC, the Clinical Pharmacogenetics Implementation Consortium, translates genotypes into metabolizer phenotypes. For CYP2D6 each allele gets an activity value; the sum is the activity score: 0 is a poor metabolizer, above 0 and below 1.25 intermediate, 1.25 to 2.25 normal and above 2.25 ultrarapid; an allele carrying several copies of a functional gene has its value multiplied by the copy number.

More than half of Oceanian people are predicted CYP2C19 poor metabolizers, and NAT2 slow acetylators are common in most groupsConsortium frequency tables
Estimated frequencies of poor, intermediate, normal, rapid and ultrarapid metabolizer phenotypes for CYP2D6, CYP2C19 and NAT2 by biogeographic group, from the CPIC frequency tablesCYP2D6poorAfrican American/Afro-CaribbeanCYP2D6, African American/Afro-Caribbean: Poor 2.3%CYP2D6, African American/Afro-Caribbean: Intermediate 35.9%CYP2D6, African American/Afro-Caribbean: Normal 53.8%CYP2D6, African American/Afro-Caribbean: Ultrarapid 4.1%CYP2D6, African American/Afro-Caribbean: Indeterminate 3.9%2.3%AmericanCYP2D6, American: Poor 2.0%CYP2D6, American: Intermediate 23.4%CYP2D6, American: Normal 66.9%CYP2D6, American: Ultrarapid 5.2%CYP2D6, American: Indeterminate 2.5%2.0%Central/South AsianCYP2D6, Central/South Asian: Poor 2.4%CYP2D6, Central/South Asian: Intermediate 28.1%CYP2D6, Central/South Asian: Normal 58.1%CYP2D6, Central/South Asian: Ultrarapid 1.5%CYP2D6, Central/South Asian: Indeterminate 10.0%2.4%East AsianCYP2D6, East Asian: Poor 0.8%CYP2D6, East Asian: Intermediate 38.5%CYP2D6, East Asian: Normal 53.2%CYP2D6, East Asian: Ultrarapid 0.9%CYP2D6, East Asian: Indeterminate 6.7%0.8%EuropeanCYP2D6, European: Poor 6.5%CYP2D6, European: Intermediate 38.3%CYP2D6, European: Normal 49.2%CYP2D6, European: Ultrarapid 2.3%CYP2D6, European: Indeterminate 3.7%6.5%LatinoCYP2D6, Latino: Poor 3.1%CYP2D6, Latino: Intermediate 29.1%CYP2D6, Latino: Normal 59.6%CYP2D6, Latino: Ultrarapid 4.1%CYP2D6, Latino: Indeterminate 4.2%3.1%Near EasternCYP2D6, Near Eastern: Poor 2.2%CYP2D6, Near Eastern: Intermediate 30.1%CYP2D6, Near Eastern: Normal 56.5%CYP2D6, Near Eastern: Ultrarapid 7.4%CYP2D6, Near Eastern: Indeterminate 3.7%2.2%OceanianCYP2D6, Oceanian: Poor 0.3%CYP2D6, Oceanian: Intermediate 9.5%CYP2D6, Oceanian: Normal 63.6%CYP2D6, Oceanian: Ultrarapid 17.8%CYP2D6, Oceanian: Indeterminate 8.7%0.3%Sub-Saharan AfricanCYP2D6, Sub-Saharan African: Poor 2.0%CYP2D6, Sub-Saharan African: Intermediate 38.1%CYP2D6, Sub-Saharan African: Normal 46.4%CYP2D6, Sub-Saharan African: Ultrarapid 4.5%CYP2D6, Sub-Saharan African: Indeterminate 9.0%2.0%CYP2C19poorAfrican American/Afro-CaribbeanCYP2C19, African American/Afro-Caribbean: Poor 4.8%CYP2C19, African American/Afro-Caribbean: Intermediate 34.2%CYP2C19, African American/Afro-Caribbean: Normal 32.8%CYP2C19, African American/Afro-Caribbean: Rapid 23.7%CYP2C19, African American/Afro-Caribbean: Ultrarapid 4.3%CYP2C19, African American/Afro-Caribbean: Indeterminate 0.2%4.8%AmericanCYP2C19, American: Poor 1.5%CYP2C19, American: Intermediate 21.4%CYP2C19, American: Normal 62.8%CYP2C19, American: Rapid 13.6%CYP2C19, American: Ultrarapid 0.7%1.5%Central/South AsianCYP2C19, Central/South Asian: Poor 8.2%CYP2C19, Central/South Asian: Intermediate 40.8%CYP2C19, Central/South Asian: Normal 29.6%CYP2C19, Central/South Asian: Rapid 18.6%CYP2C19, Central/South Asian: Ultrarapid 2.9%8.2%East AsianCYP2C19, East Asian: Poor 13.0%CYP2C19, East Asian: Intermediate 46.0%CYP2C19, East Asian: Normal 38.1%CYP2C19, East Asian: Rapid 2.5%CYP2C19, East Asian: Indeterminate 0.3%13.0%EuropeanCYP2C19, European: Poor 2.4%CYP2C19, European: Intermediate 26.2%CYP2C19, European: Normal 39.6%CYP2C19, European: Rapid 27.1%CYP2C19, European: Ultrarapid 4.6%2.4%LatinoCYP2C19, Latino: Poor 1.2%CYP2C19, Latino: Intermediate 19.4%CYP2C19, Latino: Normal 52.5%CYP2C19, Latino: Rapid 24.1%CYP2C19, Latino: Ultrarapid 2.8%1.2%Near EasternCYP2C19, Near Eastern: Poor 1.9%CYP2C19, Near Eastern: Intermediate 23.5%CYP2C19, Near Eastern: Normal 45.2%CYP2C19, Near Eastern: Rapid 25.7%CYP2C19, Near Eastern: Ultrarapid 3.7%1.9%OceanianCYP2C19, Oceanian: Poor 57.1%CYP2C19, Oceanian: Intermediate 36.9%CYP2C19, Oceanian: Normal 3.5%CYP2C19, Oceanian: Rapid 2.1%CYP2C19, Oceanian: Ultrarapid 0.3%57.1%Sub-Saharan AfricanCYP2C19, Sub-Saharan African: Poor 4.7%CYP2C19, Sub-Saharan African: Intermediate 34.2%CYP2C19, Sub-Saharan African: Normal 37.0%CYP2C19, Sub-Saharan African: Rapid 21.1%CYP2C19, Sub-Saharan African: Ultrarapid 3.0%4.7%NAT2poorCentral/South AsianNAT2, Central/South Asian: Poor 59.8%NAT2, Central/South Asian: Intermediate 34.4%NAT2, Central/South Asian: Rapid 5.4%NAT2, Central/South Asian: Indeterminate 0.1%59.8%East AsianNAT2, East Asian: Poor 22.4%NAT2, East Asian: Intermediate 50.5%NAT2, East Asian: Rapid 26.7%22.4%EuropeanNAT2, European: Poor 59.5%NAT2, European: Intermediate 35.1%NAT2, European: Rapid 5.4%59.5%LatinoNAT2, Latino: Poor 46.5%NAT2, Latino: Intermediate 42.8%NAT2, Latino: Rapid 10.0%NAT2, Latino: Indeterminate 0.7%46.5%Sub-Saharan AfricanNAT2, Sub-Saharan African: Poor 44.5%NAT2, Sub-Saharan African: Intermediate 41.2%NAT2, Sub-Saharan African: Rapid 9.0%NAT2, Sub-Saharan African: Indeterminate 5.2%44.5%PoorIntermediateNormalRapidUltrarapidIndeterminateEstimated frequencies of poor, intermediate, normal, rapid and ultrarapid metabolizer phenotypes for CYP2D6, CYP2C19 and NAT2 by biogeographic group, from the CPIC frequency tablesCYP2D6poorAfrican American/Afro-CaribbeanCYP2D6, African American/Afro-Caribbean: Poor 2.3%CYP2D6, African American/Afro-Caribbean: Intermediate 35.9%CYP2D6, African American/Afro-Caribbean: Normal 53.8%CYP2D6, African American/Afro-Caribbean: Ultrarapid 4.1%CYP2D6, African American/Afro-Caribbean: Indeterminate 3.9%2.3%AmericanCYP2D6, American: Poor 2.0%CYP2D6, American: Intermediate 23.4%CYP2D6, American: Normal 66.9%CYP2D6, American: Ultrarapid 5.2%CYP2D6, American: Indeterminate 2.5%2.0%Central/South AsianCYP2D6, Central/South Asian: Poor 2.4%CYP2D6, Central/South Asian: Intermediate 28.1%CYP2D6, Central/South Asian: Normal 58.1%CYP2D6, Central/South Asian: Ultrarapid 1.5%CYP2D6, Central/South Asian: Indeterminate 10.0%2.4%East AsianCYP2D6, East Asian: Poor 0.8%CYP2D6, East Asian: Intermediate 38.5%CYP2D6, East Asian: Normal 53.2%CYP2D6, East Asian: Ultrarapid 0.9%CYP2D6, East Asian: Indeterminate 6.7%0.8%EuropeanCYP2D6, European: Poor 6.5%CYP2D6, European: Intermediate 38.3%CYP2D6, European: Normal 49.2%CYP2D6, European: Ultrarapid 2.3%CYP2D6, European: Indeterminate 3.7%6.5%LatinoCYP2D6, Latino: Poor 3.1%CYP2D6, Latino: Intermediate 29.1%CYP2D6, Latino: Normal 59.6%CYP2D6, Latino: Ultrarapid 4.1%CYP2D6, Latino: Indeterminate 4.2%3.1%Near EasternCYP2D6, Near Eastern: Poor 2.2%CYP2D6, Near Eastern: Intermediate 30.1%CYP2D6, Near Eastern: Normal 56.5%CYP2D6, Near Eastern: Ultrarapid 7.4%CYP2D6, Near Eastern: Indeterminate 3.7%2.2%OceanianCYP2D6, Oceanian: Poor 0.3%CYP2D6, Oceanian: Intermediate 9.5%CYP2D6, Oceanian: Normal 63.6%CYP2D6, Oceanian: Ultrarapid 17.8%CYP2D6, Oceanian: Indeterminate 8.7%0.3%Sub-Saharan AfricanCYP2D6, Sub-Saharan African: Poor 2.0%CYP2D6, Sub-Saharan African: Intermediate 38.1%CYP2D6, Sub-Saharan African: Normal 46.4%CYP2D6, Sub-Saharan African: Ultrarapid 4.5%CYP2D6, Sub-Saharan African: Indeterminate 9.0%2.0%CYP2C19poorAfrican American/Afro-CaribbeanCYP2C19, African American/Afro-Caribbean: Poor 4.8%CYP2C19, African American/Afro-Caribbean: Intermediate 34.2%CYP2C19, African American/Afro-Caribbean: Normal 32.8%CYP2C19, African American/Afro-Caribbean: Rapid 23.7%CYP2C19, African American/Afro-Caribbean: Ultrarapid 4.3%CYP2C19, African American/Afro-Caribbean: Indeterminate 0.2%4.8%AmericanCYP2C19, American: Poor 1.5%CYP2C19, American: Intermediate 21.4%CYP2C19, American: Normal 62.8%CYP2C19, American: Rapid 13.6%CYP2C19, American: Ultrarapid 0.7%1.5%Central/South AsianCYP2C19, Central/South Asian: Poor 8.2%CYP2C19, Central/South Asian: Intermediate 40.8%CYP2C19, Central/South Asian: Normal 29.6%CYP2C19, Central/South Asian: Rapid 18.6%CYP2C19, Central/South Asian: Ultrarapid 2.9%8.2%East AsianCYP2C19, East Asian: Poor 13.0%CYP2C19, East Asian: Intermediate 46.0%CYP2C19, East Asian: Normal 38.1%CYP2C19, East Asian: Rapid 2.5%CYP2C19, East Asian: Indeterminate 0.3%13.0%EuropeanCYP2C19, European: Poor 2.4%CYP2C19, European: Intermediate 26.2%CYP2C19, European: Normal 39.6%CYP2C19, European: Rapid 27.1%CYP2C19, European: Ultrarapid 4.6%2.4%LatinoCYP2C19, Latino: Poor 1.2%CYP2C19, Latino: Intermediate 19.4%CYP2C19, Latino: Normal 52.5%CYP2C19, Latino: Rapid 24.1%CYP2C19, Latino: Ultrarapid 2.8%1.2%Near EasternCYP2C19, Near Eastern: Poor 1.9%CYP2C19, Near Eastern: Intermediate 23.5%CYP2C19, Near Eastern: Normal 45.2%CYP2C19, Near Eastern: Rapid 25.7%CYP2C19, Near Eastern: Ultrarapid 3.7%1.9%OceanianCYP2C19, Oceanian: Poor 57.1%CYP2C19, Oceanian: Intermediate 36.9%CYP2C19, Oceanian: Normal 3.5%CYP2C19, Oceanian: Rapid 2.1%CYP2C19, Oceanian: Ultrarapid 0.3%57.1%Sub-Saharan AfricanCYP2C19, Sub-Saharan African: Poor 4.7%CYP2C19, Sub-Saharan African: Intermediate 34.2%CYP2C19, Sub-Saharan African: Normal 37.0%CYP2C19, Sub-Saharan African: Rapid 21.1%CYP2C19, Sub-Saharan African: Ultrarapid 3.0%4.7%NAT2poorCentral/South AsianNAT2, Central/South Asian: Poor 59.8%NAT2, Central/South Asian: Intermediate 34.4%NAT2, Central/South Asian: Rapid 5.4%NAT2, Central/South Asian: Indeterminate 0.1%59.8%East AsianNAT2, East Asian: Poor 22.4%NAT2, East Asian: Intermediate 50.5%NAT2, East Asian: Rapid 26.7%22.4%EuropeanNAT2, European: Poor 59.5%NAT2, European: Intermediate 35.1%NAT2, European: Rapid 5.4%59.5%LatinoNAT2, Latino: Poor 46.5%NAT2, Latino: Intermediate 42.8%NAT2, Latino: Rapid 10.0%NAT2, Latino: Indeterminate 0.7%46.5%Sub-Saharan AfricanNAT2, Sub-Saharan African: Poor 44.5%NAT2, Sub-Saharan African: Intermediate 41.2%NAT2, Sub-Saharan African: Rapid 9.0%NAT2, Sub-Saharan African: Indeterminate 5.2%44.5%PoorIntermediateNormalRapidUltrarapidIndeterminate
CPIC frequency tables for CYP2D6 (file updated Oct 17, 2024), CYP2C19 (Mar 11, 2022) and NAT2 (created Jul 19, 2026), downloaded Oct 29, 2026. CYP2D6 and CYP2C19 phenotypes are estimated from published allele frequencies assuming Hardy–Weinberg equilibrium; NAT2 from the UK Biobank 200k phased data. Likely poor and likely intermediate CYP2C19 calls are counted with poor and intermediate. Right column: poor-metabolizer share.
Show data: phenotype frequencies, %
CPIC estimated metabolizer phenotype frequencies by biogeographic group
Gene and groupPoorIntermediateNormalRapidUltrarapidIndeterminate
CYP2D6, African American/Afro-Caribbean2.335.953.8—4.13.9
CYP2D6, American2.023.466.9—5.22.5
CYP2D6, Central/South Asian2.428.158.1—1.510.0
CYP2D6, East Asian0.838.553.2—0.96.7
CYP2D6, European6.538.349.2—2.33.7
CYP2D6, Latino3.129.159.6—4.14.2
CYP2D6, Near Eastern2.230.156.5—7.43.7
CYP2D6, Oceanian0.39.563.6—17.88.7
CYP2D6, Sub-Saharan African2.038.146.4—4.59.0
CYP2C19, African American/Afro-Caribbean4.834.232.823.74.30.2
CYP2C19, American1.521.462.813.60.70.0
CYP2C19, Central/South Asian8.240.829.618.62.90.0
CYP2C19, East Asian13.046.038.12.50.00.3
CYP2C19, European2.426.239.627.14.60.0
CYP2C19, Latino1.219.452.524.12.80.0
CYP2C19, Near Eastern1.923.545.225.73.70.0
CYP2C19, Oceanian57.136.93.52.10.30.0
CYP2C19, Sub-Saharan African4.734.237.021.13.00.0
NAT2, Central/South Asian59.834.4—5.4—0.1
NAT2, East Asian22.450.5—26.7—0.0
NAT2, European59.535.1—5.4—0.0
NAT2, Latino46.542.8—10.0—0.7
NAT2, Sub-Saharan African44.541.2—9.0—5.2

CPIC frequency tables (files.cpicpgx.org), sheet "Phenotype frequency". CYP2D6 values are sums of activity-score rows grouped by the CPIC ranges; rows may not add to exactly 100% because of rounding and, for NAT2, diplotypes without an assigned function. StatPearls (Pharmacogenomics Overview) gives slow-acetylator prevalence of 10–20% in Asian populations and more than 50% in Caucasian and African populations.

The consequences depend on the drug. In healthy volunteers given codeine, CYP2D6 poor metabolizers had a 96% lower mean morphine AUC than normal and intermediate metabolizers, and ultrarapid metabolizers a 45% higher median morphine AUC than normal metabolizers. For clopidogrel, whose active metabolite is formed less in CYP2C19 intermediate and poor metabolizers, CPIC advises avoiding the standard 75 mg dose in intermediate metabolizers and avoiding clopidogrel in poor metabolizers if possible, using prasugrel or ticagrelor when there is no contraindication (acute coronary syndrome or PCI).

Crews KR, et al. Clin Pharmacol Ther 2021;110:888–896 (CYP2D6 activity score, Table 1; codeine data); Lee CR, et al. Clin Pharmacol Ther 2022;112:959–967 (CYP2C19 and clopidogrel, Table 2); Patel T, et al. StatPearls, Cytochrome P450 (2026).

Induction and inhibition

Some botanicals change how fast the liver clears drugs, and human studies show which ones

Induction raises the expression of an enzyme or transporter; St John's wort does this in people, and in primary human hepatocytes its constituent hyperforin acts through the pregnane X receptor. Inhibition lowers the activity of enzyme already present, as goldenseal does for CYP2D6 and CYP3A4/5. Both are measured in people with probe drugs. The midazolam data for St John's wort and goldenseal, plotted against FDA potency categories, are on ADME, PK and PBPK; this section adds the mechanism and other probes.

Of six botanical extracts taken for 14 days, only goldenseal lowered CYP2D6 activity, to about halfHuman study
Geometric mean ratio of the debrisoquine urinary recovery ratio after versus before 14 days of each botanical extract, with 90% confidence intervals: goldenseal 0.53, the other five between 0.97 and 1.030.40.50.60.811.2no changeGoldensealGoldenseal: 0.53 (90% CI 0.44–0.64)0.53 (0.44–0.64)Milk thistleMilk thistle: 0.97 (90% CI 0.93–1.1)0.97 (0.93–1.10)Black cohoshBlack cohosh: 0.99 (90% CI 0.94–1.04)0.99 (0.94–1.04)KavaKava: 1.01 (90% CI 0.96–1.06)1.01 (0.96–1.06)St. John's wortSt. John's wort: 1.02 (90% CI 0.97–1.07)1.02 (0.97–1.07)EchinaceaEchinacea: 1.03 (90% CI 0.98–1.07)1.03 (0.98–1.07)Post/pre ratio of CYP2D6 phenotypic ratio (log scale; below 1 = less CYP2D6 activity)Geometric mean ratio of the debrisoquine urinary recovery ratio after versus before 14 days of each botanical extract, with 90% confidence intervals: goldenseal 0.53, the other five between 0.97 and 1.030.40.50.60.811.2no changeGoldensealGoldenseal: 0.53 (90% CI 0.44–0.64)0.53 (0.44–0.64)Milk thistleMilk thistle: 0.97 (90% CI 0.93–1.1)0.97 (0.93–1.10)Black cohoshBlack cohosh: 0.99 (90% CI 0.94–1.04)0.99 (0.94–1.04)KavaKava: 1.01 (90% CI 0.96–1.06)1.01 (0.96–1.06)St. John's wortSt. John's wort: 1.02 (90% CI 0.97–1.07)1.02 (0.97–1.07)EchinaceaEchinacea: 1.03 (90% CI 0.98–1.07)1.03 (0.98–1.07)Post/pre ratio of CYP2D6 phenotypic ratio (log scale;below 1 = less CYP2D6 activity)
Gurley BJ, et al. Mol Nutr Food Res 2008;52:755–763, Table 1: three open-label randomized studies in healthy volunteers (18 per study, 9 women, in the abstract; 16 per study, 8 women, in the methods), 14 days per extract with 30-day washouts; debrisoquine 5 mg; 8-hour urinary recovery ratio; geometric mean post/pre ratios with 90% CI.
Show data: CYP2D6 phenotypic ratios before and after each extract
Debrisoquine urinary recovery ratios before and after 14 days of six botanical extracts
ExtractBefore, mean (95% CI)After, mean (95% CI)Post/pre, geometric mean (90% CI)
Goldenseal0.708 (0.632–0.785)0.374 (0.285–0.463)0.53 (0.44–0.64), significant
Milk thistle0.694 (0.621–0.767)0.672 (0.602–0.743)0.97 (0.93–1.10)
Black cohosh0.684 (0.610–0.759)0.679 (0.593–0.765)0.99 (0.94–1.04)
Kava0.654 (0.568–0.739)0.669 (0.573–0.765)1.01 (0.96–1.06)
St John's wort0.606 (0.522–0.691)0.620 (0.537–0.703)1.02 (0.97–1.07)
Echinacea0.583 (0.514–0.653)0.601 (0.530–0.673)1.03 (0.98–1.07)

Gurley BJ, Swain A, Hubbard MA, et al. Clinical assessment of CYP2D6-mediated herb–drug interactions in humans. Mol Nutr Food Res 2008;52:755–763 (PMC2562884), Table 1.

Studies of St John's wort and goldenseal on drug-metabolizing enzymes, by design
StudyDesignFindingEvidence type
Moore et al. 2000Receptor binding; primary human hepatocytesHyperforin from St John's wort binds the pregnane X receptor (PXR) with Ki = 27 nM; extracts or hyperforin markedly induced CYP3A4 expression in hepatocytesIn vitro
Dürr et al. 20008 healthy men, St John's wort extract for 14 daysDuodenal P-glycoprotein 1.4-fold and CYP3A4 1.5-fold; hepatic CYP3A4 activity (14C-erythromycin breath test) 1.4-fold; digoxin exposure after a single 0.5 mg dose 18% lowerHuman
Markowitz et al. 200312 healthy volunteers (6 women), 300 mg three times daily for 14 days; alprazolam and dextromethorphan probesAlprazolam AUC fell 2-fold and clearance rose 2-fold; half-life 12.4 to 6.0 h. Dextromethorphan urinary ratio (CYP2D6) did not change significantlyHuman
Gurley et al. 200512 healthy volunteers (6 women), each extract for 28 days; probe cocktailGoldenseal inhibited CYP2D6 and CYP3A4/5 by about 40%; kava lowered CYP2E1 by about 40%; black cohosh lowered CYP2D6 by about 7%, judged not clinically relevant; valerian had no effectHuman
Gurley et al. 20083 studies of healthy volunteers, 2 extracts each, 14 days per extractGoldenseal halved the debrisoquine recovery ratio; the five other extracts had no significant effect (figure above)Human

Moore LB, et al. Proc Natl Acad Sci USA 2000;97:7500–7502; Dürr D, et al. Clin Pharmacol Ther 2000;68:598–604; Markowitz JS, et al. JAMA 2003;290:1500–1504; Gurley BJ, et al. Clin Pharmacol Ther 2005;77:415–426 (PMC1894911); Gurley 2008. Supplement pages: St John's wort, goldenseal, kava. Interaction overview: Supplement–drug interactions.

Liver tests

Liver tests measure injury, bile flow and synthetic reserve, and each has sources outside the liver

The American College of Gastroenterology prefers the term liver chemistries, because most of the panel reflects injury or cholestasis rather than function. ALT and AST are released when hepatocytes are injured; ALP and GGT point to the bile ducts; bilirubin, albumin and the prothrombin time with its INR say more about how much working liver is left.

ALT, albumin and the clotting test point mainly to the liver; AST, ALP and GGT also come from other tissuesSourced classification
Matrix of seven liver tests against the tissues in which their analyte is made or foundLiverBoneMuscle,heartKidneyGutPlacentaBlood cellsOtherALTALT · Liver: highALT · Muscle, heart: lowALT · Kidney: lowASTAST · Liver: lowAST · Muscle, heart: lowAST · Kidney: lowAST · Blood cells: lowAST · Other: lowALPALP · Liver: highALP · Bone: highALP · Kidney: lowALP · Gut: lowALP · Placenta: lowGGTGGT · Liver: lowGGT · Kidney: lowGGT · Gut: lowGGT · Placenta: lowGGT · Other: lowBilirubinBilirubin · Liver: highBilirubin · Blood cells: highAlbuminAlbumin · Liver: highPT / INRPT / INR · Liver: highMain origin, or the organ that makes or clears itAlso foundMatrix of seven liver tests against the tissues in which their analyte is made or foundALTALT · : highALT · : lowALT · : lowASTAST · : lowAST · : lowAST · : lowAST · : lowAST · : lowALPALP · : highALP · : highALP · : lowALP · : lowALP · : lowGGTGGT · : lowGGT · : lowGGT · : lowGGT · : lowGGT · : lowBilirubinBilirubin · : highBilirubin · : highAlbuminAlbumin · : highPT / INRPT / INR · : highLiverBoneMuscleKidneyGutPlacentaBlood cellsOtherMain origin, or the organ that makes or clears itAlso found
From the tissue statements in StatPearls (Liver Function Tests, updated Jun 8, 2026; ALT test; Serum Alkaline Phosphatase; Physiology, Bilirubin; Physiology, Liver) and UniProt P19440 (GGT1 tissue expression). For bilirubin, red cells are the main origin of the heme and the liver conjugates and excretes it.
Liver tests: what each measures, other sources and the reference interval stated by StatPearls
TestWhat it reflectsSources other than the liverReference interval (StatPearls)
ALT, alanine aminotransferaseHepatocyte injury; mostly cytosolic; transfers an amino group from L-alanine to α-ketoglutarate with pyridoxal 5′-phosphate; half-life about 47 hKidney, heart, muscle4–36 IU/L
AST, aspartate aminotransferaseHepatocyte injury; cytosolic and mitochondrial formsSkeletal muscle, myocardium, kidney, brain, leukocytes, erythrocytes5–30 IU/L
ALP, alkaline phosphataseConcentrated in the canalicular membrane and bile-duct epithelium; rises with impaired bile flowBone (over 80% of serum ALP comes from liver and bone), placenta, intestine, kidney30–120 IU/L
GGT, γ-glutamyltransferaseConfirms a hepatobiliary source for a raised ALP; induced by alcohol and some medicinesKidney, pancreas, stomach, intestine, placenta, lung (UniProt)6–50 IU/L
Bilirubin, total and directProduction, uptake, conjugation and excretion together; the direct fraction approximates conjugated bilirubin and is partly method-definedHeme from red cells; raised unconjugated bilirubin also in hemolysis and Gilbert syndromeTotal 2–17 µmol/L; direct 0–6 µmol/L
AlbuminSynthetic capacity; changes slowly because of its long half-lifeLowered also by inflammation, poor nutrition, renal or gut protein loss35–50 g/L
Prothrombin time and INRLiver-made clotting factors; responds faster than albumin because the factors have shorter half-livesProlonged also by vitamin K deficiency, anticoagulants, malabsorptionPT 10.9–12.5 s; INR 1.1 or below

Jogu P, Zubair M, Minter DA. Liver Function Tests. StatPearls, updated Jun 8, 2026 (intervals, tissue sources, interpretation); Moriles KE, et al. Alanine Aminotransferase (ALT) Test. StatPearls; Lowe D, et al. Serum Alkaline Phosphatase. StatPearls, updated Mar 23, 2026; Zubair M, et al. Prothrombin Time. StatPearls, updated Sep 8, 2026 (INR); UniProt P19440; Kwo PY, et al. Am J Gastroenterol 2017;112:18–35.

Reference intervals differ between laboratories, methods and populations; a result is read against the interval printed by the laboratory that measured it. StatPearls notes that most intervals cover the central 95% of a reference population, so some healthy people fall outside them. The ACG guideline puts a "true healthy normal" ALT at 29–33 IU/L for men and 19–25 IU/L for women. LiverTox notes typical laboratory ALT upper limits of 40 or 42 U/L, and 55 or 66 U/L in some laboratories.

StatPearls grades aminotransferase rises against the upper limit of normal (ULN): borderline below 2 × ULN, mild 2 to 5 ×, moderate 5 to 15 ×, severe above 15 ×, and massive above 10,000 IU/L. Mildly raised ALT and AST are found in about 10% to 20% of the general population, and up to 30% of asymptomatic elevations resolve within 3 weeks.

Jogu P, et al. StatPearls 2026; Roussel Uclaf Causality Assessment Method (RUCAM) in Drug Induced Liver Injury. LiverTox, updated May 4, 2019.

Liver injury: pattern

The R ratio sorts drug- and supplement-induced injury into hepatocellular, cholestatic and mixed

Drug-induced liver injury (DILI) can mimic almost any liver disease, and no single test proves that a drug, supplement or herb caused it. The first step is the pattern of enzyme rise, captured by the R ratio: ALT divided by ALP, each first expressed as a multiple of its upper limit of normal.

The same ALT can mean hepatocellular or mixed injury depending on how high ALP isCalculated from published thresholds
Pattern of liver injury computed from ALT and ALP as multiples of the upper limit of normal: R = (ALT/ULN) / (ALP/ULN); R of 5 or more hepatocellular, 2 or less cholestatic, between 2 and 5 mixed; values below ALT 5 x ULN and ALP 2 x ULN do not meet the updated RUCAM injury criteriaHepatocellular: R ≥ 5Mixed: 2 < R < 5Cholestatic: R ≤ 2Below ALT 5 × ULN and ALP 2 × ULN0.5125102050ALP ÷ ULN0.5125102050100200ALT ÷ ULNHepatocellularR ≥ 5Mixed2 < R < 5CholestaticR ≤ 2Below injury criteriaR = (ALT ÷ ULN) ÷ (ALP ÷ ULN)Both values from the same day,at first presentation, withthe local laboratory's ULN.Injury criteria (updated RUCAM):ALT ≥ 5 × ULN and/orALP ≥ 2 × ULN.Red line: R = 5Blue line: R = 2Pattern of liver injury computed from ALT and ALP as multiples of the upper limit of normal: R = (ALT/ULN) / (ALP/ULN); R of 5 or more hepatocellular, 2 or less cholestatic, between 2 and 5 mixed; values below ALT 5 x ULN and ALP 2 x ULN do not meet the updated RUCAM injury criteriaHepatocellular: R ≥ 5Mixed: 2 < R < 5Cholestatic: R ≤ 2Below ALT 5 × ULN and ALP 2 × ULN0.5125102050ALP ÷ ULN0.5125102050100200ALT ÷ ULNHepatocellularR ≥ 5CholestaticR ≤ 2Below injury criteriaHepatocellular, R ≥ 5Mixed, 2 < R < 5Cholestatic, R ≤ 2
Calculated from R = (ALT/ULN) ÷ (ALP/ULN) with the thresholds and injury criteria of the updated RUCAM (Danan G, Teschke R. Int J Mol Sci 2016;17:14): hepatocellular R ≥ 5, cholestatic R ≤ 2, mixed 2 < R < 5; liver injury defined as ALT ≥ 5 × ULN and/or ALP ≥ 2 × ULN. An educational calculation, not a measurement or a diagnosis for any person or product.

Doubling ALP at a fixed ALT halves R, so the same aminotransferase rise can move from the hepatocellular to the mixed category.

Explanation, details and sources

The updated RUCAM also classifies a case as hepatocellular when ALT > 5 × ULN with ALP ≤ ULN, as cholestatic when ALP > 2 × ULN with ALT ≤ ULN, and as mixed only when ALT > 5 × ULN, ALP > ULN and 2 < R < 5. When ALT is normal, a raised ALP should be matched by a raised GGT to rule out bone or another non-hepatobiliary origin. Sources word the boundaries slightly differently: the LiverTox clinical-course chapter uses R of 5 or more and 2 or less, its RUCAM chapter and the DILIN analysis by Navarro use R > 5 and R < 2.

LiverTox lists five practical problems: timing (early samples look hepatocellular, late ones cholestatic), ALP from bone, cancer or pregnancy, AST used when ALT is missing, which ULN to use, and liver disease that predates the drug. Its recommendation is to use the first qualifying values and the local laboratory's ULN; if none is given, 40 U/L for ALT and 115 U/L for ALP. In the RUCAM, mixed cases are scored as cholestatic.

Danan G, Teschke R. RUCAM in drug and herb induced liver injury: the update. Int J Mol Sci 2016;17:14, section 5.2; LiverTox: Clinical Course and Diagnosis of Drug Induced Liver Disease; Roussel Uclaf Causality Assessment Method (RUCAM) in Drug Induced Liver Injury (both updated May 4, 2019); Navarro 2014, Table 2 footnote.

Liver injury: severity

Hy's law: hepatocellular injury that also raises bilirubin signals a drug able to cause severe injury

The FDA's 2009 guidance on premarketing evaluation of DILI builds on Zimmerman's observation that drug-induced hepatocellular injury with jaundice carried a 10% to 50% mortality from acute liver failure before transplantation was available. Because the liver has a large reserve for excreting bilirubin, injury that raises bilirubin above 2 × ULN is extensive.

A Hy's law case sits above both thresholds and has no other explanationFDA guidance thresholds
Hy's law thresholds from the 2009 FDA guidance: aminotransferase above 3 times the upper limit of normal together with total bilirubin above 2 times the upper limit of normal, without initial cholestasis and with no other explanationAT > 3 × ULN and TBL > 2 × ULNAT > 3 × ULN, TBL ≤ 2 × ULNTBL > 2 × ULN, AT ≤ 3 × ULN0.3131030100Peak ALT or AST ÷ ULN0.20.51251020Peak total bilirubin ÷ ULNPotential Hy's law caseif ALP is not substantially raised and no other cause is foundAT > 3 × ULN onlyTBL > 2 × ULN onlyHy's law thresholds from the 2009 FDA guidance: aminotransferase above 3 times the upper limit of normal together with total bilirubin above 2 times the upper limit of normal, without initial cholestasis and with no other explanationAT > 3 × ULN and TBL > 2 × ULNAT > 3 × ULN, TBL ≤ 2 × ULNTBL > 2 × ULN, AT ≤ 3 × ULN0.3131030100Peak ALT or AST ÷ ULN0.20.51251020Peak total bilirubin ÷ ULNPotential Hy's law caseif ALP is not substantiallyraised and no other cause isfoundAT > 3 × ULN onlyTBL > 2 × ULNonly
U.S. FDA. Guidance for Industry. Drug-Induced Liver Injury: Premarketing Clinical Evaluation. July 2009 (final guidance; contains nonbinding recommendations). Threshold diagram only; no patient data are plotted.
  1. Hepatocellular injuryThe drug causes a higher incidence of 3-fold or greater elevations above the ULN of ALT or AST than the control drug or placebo.
  2. Bilirubin riseAmong those subjects, one or more also show serum total bilirubin above 2 × ULN, without initial findings of cholestasis (raised serum ALP).
  3. No other reasonNo other cause explains the combination, such as viral hepatitis A, B or C, preexisting or acute liver disease, or another drug capable of causing the injury.

The guidance states that one Hy's law case in a clinical trial database is worrisome and two are considered highly predictive of a drug's potential to cause severe DILI in a larger population. Severe DILI, fatal or requiring transplant, can be estimated to occur at a rate of at least one-tenth the rate of Hy's law cases. Severe injury is usually rare, about 1 in 10,000 or less, so an approval database often contains none.

FDA 2009 guidance, section III (pages 4–5). The guidance addresses drugs in clinical development; LiverTox notes that U.S. supplement manufacturers need not have FDA approval before marketing (Herbal and Dietary Supplements chapter, updated Jul 1, 2025).

Liver injury: causality

RUCAM scores how likely it is that one agent caused the injury

The Roussel Uclaf Causality Assessment Method, introduced in 1993 and updated in 2016, assigns points in seven categories and sums them. Each suspected agent gets its own score, and the course and risk-factor items are scored differently for hepatocellular and for cholestatic or mixed injury.

RUCAM categories and possible points
CategoryPointsWhat earns the most points
1. Time to onset+1 or +2Onset 5 to 90 days after a first exposure while still taking the agent
2. Course after stopping−2 to +3 (hepatocellular); 0 to +2 (cholestatic or mixed)ALT falls by 50% within 8 days; for cholestatic injury, ALP or bilirubin falls by more than 50% within 180 days
3. Risk factors0 or +1 each, two itemsAlcohol use (or pregnancy for cholestatic injury); age 55 or older
4. Concomitant drugs0 to −3No other drug with a compatible time to onset
5. Non-drug causes−3 to +2Hepatitis A, B and C, biliary obstruction, alcoholic liver disease, ischemic hepatitis and complications of underlying disease reasonably ruled out
6. Previous information0, +1 or +2Liver injury stated in the product label
7. Rechallenge−2 to +3ALT doubles (or ALP or bilirubin for cholestatic injury) with the drug alone
RUCAM total score and causality grade
Total score≤ 01–23–56–8≥ 9
CausalityExcludedUnlikelyPossibleProbableHighly probable

The full range is −9 to +14, but LiverTox notes that without rechallenge and without the special risk factors the range for cholestatic or mixed injury is only −5 to +8, so many cases cannot score above "probable". The U.S. Drug-Induced Liver Injury Network instead uses expert opinion: definite (> 95% likelihood), highly likely (75–95%), probable (50–74%), possible (25–49%) or unlikely (< 25%).

Danan G, Teschke R. Int J Mol Sci 2016;17:14 (score bands, section 5.3.8); LiverTox, RUCAM chapter, updated May 4, 2019 (categories, points, ranges); Chalasani N, et al. Gastroenterology 2015;148:1340–1352 (DILIN causality scale).

Liver injury: evidence base

LiverTox grades how firmly the literature links each agent to liver injury

LiverTox, the National Institute of Diabetes and Digestive and Kidney Diseases (NIDDK) reference on drug-induced liver injury, gives each drug, herb or supplement a likelihood score developed by the DILIN. The score is based largely on the published literature: how many cases have been described and whether the injury has a characteristic signature.

Seven single botanicals reach category A in their LiverTox chapters; examples show categories B to ENIH reference database
LiverTox likelihood categories A to E with the definition of each, every single-plant chapter scored A and examples of botanicals placed in categories B to E by their LiverTox chaptersDefinitionBotanicals (A: every single-plant chapter; B to E: examples)Category AAWell known, well described; more than 50casesBlack cohosh · Germander · Green tea extract · Kava · Polygonum multiflorum ·Tinospora cordifolia · TurmericCategory BBKnown or highly likely; 12 to 50 casesAloe vera (oral) · Ashwagandha · Chaparral · Garcinia cambogia · Pennyroyal oil ·Skullcap (Scutellaria lateriflora)Category CCProbable; fewer than 12 casesButterbur · Centella asiatica · Comfrey · Horsetail · Noni · ValerianCategory DDPossible; fewer than 3 case reportsEchinacea · Cinnamon · Saw palmetto · Senna · Moringa · Yerba mateCategory EEUnlikely despite extensive useMilk thistle · St. John's wort · Ginkgo · Ginger · Berberine · ChamomileAlso defined: E* (unproven but suspected) and X (unknown, too little information).LiverTox likelihood categories A to E with examples of botanicals in eachAWell known, well described; more than 50casesBlack cohosh · Germander · Green tea extract ·Kava · Polygonum multiflorum · Tinosporacordifolia · TurmericBKnown or highly likely; 12 to 50 casesAloe vera (oral) · Ashwagandha · Chaparral ·Garcinia cambogia · Pennyroyal oil · Skullcap(Scutellaria lateriflora)CProbable; fewer than 12 casesButterbur · Centella asiatica · Comfrey · Horsetail ·Noni · ValerianDPossible; fewer than 3 case reportsEchinacea · Cinnamon · Saw palmetto · Senna ·Moringa · Yerba mateEUnlikely despite extensive useMilk thistle · St. John's wort · Ginkgo · Ginger ·Berberine · ChamomileB to E: examples. Also: E* (suspected), X (unknown).
Definitions: LiverTox, Categorization of the Likelihood of Drug Induced Liver Injury, updated May 4, 2019. Placements: the likelihood score printed in each botanical's LiverTox chapter (NLM full-text release of Oct 28, 2026). Category A lists every single-plant chapter scored A; categories B to E show examples (category B also includes the chapters for Ba Jiao Lian, khat and tara flour). Multi-ingredient branded products, single compounds and non-plant supplements are not shown.
LiverTox likelihood categories as defined
CategoryLiverTox definition (abridged)
AWell known, well described and well reported to cause direct or idiosyncratic liver injury, with a characteristic signature; more than 50 cases including case series
BReported and known or highly likely to cause idiosyncratic liver injury, with a characteristic signature; between 12 and 50 cases including small case series
CProbably linked but reported uncommonly, no characteristic signature; fewer than 12 cases without significant case series
DSingle case reports, fewer than 3 cases, no characteristic signature; a possible hepatotoxin and only a rare cause
ENo evidence of liver injury despite extensive use; single reports largely unconvincing
E*Suspected but no convincing cases in the medical literature
XRecently introduced or rarely used; too little information to categorize ("unknown")
Show data: the category A botanicals and the category B examples, as their LiverTox chapters describe them
Botanicals with LiverTox likelihood score A or B and the features their chapters describe
Botanical (chapter, last update)ScoreWhat the chapter describes
Black cohosh (Apr 15, 2025) · supplement pageA"Products sold as black cohosh" are established causes; more than fifty instances; latency usually 2 to 12 weeks; typically hepatocellular. The specific responsible component is unclear; Chinese Actaea species have been found in implicated products labeled black cohosh
Germander (Mar 12, 2018)AFirst reported in France in 1992 after a germander weight-loss product; onset 2 to 18 weeks; hepatocellular; centrilobular necrosis on biopsy; rapid recurrence on reexposure
Green tea (Nov 20, 2020) · supplement pageACases involve mainly green tea extract and, more rarely, large amounts of green tea; drinking green tea has not been associated with injury or enzyme rises. More than 100 cases; onset 1 to 6 months; markedly hepatocellular. In a large prospective trial, ALT rose in 6.7% of women on extract versus 0.7% of controls
Kava (Apr 10, 2018) · supplement pageABetween 50 and 100 cases published or discussed; onset 2 to 24 weeks; hepatocellular; estimated frequency below 1 in 1,000,000 daily doses, with the caveat that spontaneous reporting captures under 1% of severe events
Polygonum multiflorum (Aug 18, 2020) · he shou wu pageAReported in China as the most common cause of herbal liver injury; hepatocellular or mixed; up to 10% of clinically apparent cases fatal or transplanted; HLA-B*35:01 in 70% to 88% of cases in recent series
Tinospora (Jun 27, 2025) · guduchi pageATinospora cordifolia: more than 50 cases reported since 2017, many during the COVID-19 pandemic, mostly from India and Thailand
Turmeric (Jun 16, 2025) · supplement pageAAppears to have become the most common cause of clinically apparent herbal liver injury in the United States; most cases with high-bioavailability curcumin; latency typically 1 to 4 months; HLA-B*35:01 in over 70% of cases versus 10% to 15% of controls; incidence not known, probably 1 in 10,000 to 1 in 100,000 exposed
Aloe vera (May 24, 2022) · supplement pageBOral preparations; at least a dozen cases since 2005; onset 3 to 24 weeks; hepatocellular; fatal cases not reported
Ashwagandha (Dec 3, 2024) · supplement pageBNo enzyme rises in clinical trials, but cases with labeled products; onset 2 to 12 weeks; usually cholestatic or mixed with prolonged jaundice; rare fatal cases or emergency transplantation, particularly with preexisting liver disease or cirrhosis
Chaparral (May 25, 2022)BMore than two dozen cases since 1990; usually 3 to 12 weeks; hepatocellular; some needed emergency transplantation
Garcinia cambogia (Feb 13, 2019)BAt least a dozen reports, many with multi-ingredient products; hepatocellular; frequency likely below 1 in 10,000
Pennyroyal oil (Mar 28, 2020)BAcute, clinically apparent liver injury
Skullcap (Mar 28, 2020)BAmerican skullcap (Scutellaria lateriflora); often with other herbs, and germander adulteration found in some products. The chapter also describes case reports with Chinese skullcap (S. baicalensis; see the Baikal skullcap page), a different species

LiverTox: Clinical and Research Information on Drug-Induced Liver Injury. NIDDK, 2012–; individual chapters as dated; NLM full-text release of Oct 28, 2026. The LiverTox disclaimer in that release reads: “As of September 2026, initial drafts of new and updated LiverTox chapters will be generated using Artificial Intelligence. Each draft will then undergo careful review, verification, and editing by the LiverTox editorial team.” Examples in categories C–E come from the chapters for butterbur, Centella asiatica, comfrey, horsetail, noni and valerian (C); echinacea, cinnamon, saw palmetto, senna, moringa and yerba mate (D); and milk thistle, St John's wort, ginkgo, ginger, berberine and chamomile (E).

Liver injury: who is affected

Herbal and dietary supplements make up a rising share of U.S. drug-induced liver injury cases

The DILIN enrolls consecutive patients with suspected DILI at U.S. referral centers and adjudicates causality by expert panel. Among 839 cases enrolled from 2004 to 2013, 130 (15.5%) were attributed to herbal and dietary supplements (HDS): 45 to bodybuilding products and 85 to other supplements.

The supplement share nearly tripled in a decade, and non-bodybuilding supplements caused hepatocellular injury in 71% of casesPublished registry data
Share of DILIN cases attributed to herbal and dietary supplements rose from 7% in 2004 to 2005 to 20% in 2010 to 2012; pattern of injury at onset by group: non-bodybuilding supplements 71% hepatocellularShare of DILIN cases, %05101520252004–05: bodybuilding HDS 2%2004–05: other HDS 5%7%52004–052010–12: bodybuilding HDS 8%2010–12: other HDS 12%20%8122010–12Other HDSBodybuilding HDSPattern of injury at onset, % of casesBodybuilding HDS (n = 43)Bodybuilding HDS (n = 43): hepatocellular 42%42%Bodybuilding HDS (n = 43): mixed 30%30%Bodybuilding HDS (n = 43): cholestatic 28%28%Other HDS (n = 79)Other HDS (n = 79): hepatocellular 71%71%Other HDS (n = 79): mixed 17%17%Other HDS (n = 79): cholestatic 13%13%Medications (n = 665)Medications (n = 665): hepatocellular 53%53%Medications (n = 665): mixed 23%23%Medications (n = 665): cholestatic 25%25%Hepatocellular (R > 5)MixedCholestatic (R < 2)Share of DILIN cases attributed to herbal and dietary supplements rose from 7% in 2004 to 2005 to 20% in 2010 to 2012; pattern of injury at onset by group: non-bodybuilding supplements 71% hepatocellularShare of DILIN cases, %05101520252004–05: bodybuilding HDS 2%2004–05: other HDS 5%7%52004–052010–12: bodybuilding HDS 8%2010–12: other HDS 12%20%8122010–12Other HDSBodybuilding HDSPattern of injury at onset, % of casesBodybuilding HDS (n = 43)Bodybuilding HDS (n = 43): hepatocellular 42%42%Bodybuilding HDS (n = 43): mixed 30%30%Bodybuilding HDS (n = 43): cholestatic 28%28%Other HDS (n = 79)Other HDS (n = 79): hepatocellular 71%71%Other HDS (n = 79): mixed 17%17%Other HDS (n = 79): cholestatic 13%13%Medications (n = 665)Medications (n = 665): hepatocellular 53%53%Medications (n = 665): mixed 23%23%Medications (n = 665): cholestatic 25%25%Hepatocellular (R > 5)MixedCholestatic (R < 2)
Navarro VJ, et al. Hepatology 2014;60:1399–1408: temporal trend (Results, Figure 2; 2012 enrollees grouped with 2010–11) and Table 2 (pattern at onset; patients with an R value; percentages as published, which add to 100–101).
Show data: DILIN cases by period and by pattern
Share of DILIN cases attributed to supplements by period, and pattern of injury at onset by group
MeasureBodybuilding HDSOther HDSMedicationsAll HDS
Share of cases, 2004–052%5%—7%
Share of cases, 2010–128%12%—20%
Hepatocellular at onset18 (42%)56 (71%)351 (53%)—
Mixed at onset13 (30%)13 (17%)150 (23%)—
Cholestatic at onset12 (28%)10 (13%)164 (25%)—
Liver transplantation0 (0%)11 (13%)24 (3%)—
Death0 (0%)3 (4%)50 (7%)—

Navarro VJ, Barnhart H, Bonkovsky HL, et al. Hepatology 2014;60:1399–1408 (PMC4293199), abstract, Results and Table 2.

Bodybuilding products caused prolonged jaundice (median 91 days) in young men without deaths or transplants; the other supplements caused mostly hepatocellular injury, predominantly in middle-aged women, and required liver transplantation more often than injury from medications (13% versus 3%; deaths 4% versus 7%, Table 2). A later DILIN analysis of 899 cases attributed 145 to supplements. In the national transplant waiting list, supplements rose from 2.9% to 24.1% of drug-induced acute liver failure not caused by acetaminophen, between 1995–2003 and 2013–2020.

The authors caution that the DILIN is not population-based, so a rising share does not prove that the injury itself is becoming more common in the United States. LiverTox adds that supplement products can be contaminated, adulterated or mislabeled, citing germander found in products labeled skullcap and Asian Actaea species in products labeled black cohosh, so chemical analysis of the implicated product can matter for causality.

Navarro 2014; Chalasani N, et al. Gastroenterology 2015;148:1340–1352; Ghabril M, et al. Liver Transpl 2022;28:169–179 (UNOS 1995–2020); LiverTox, Herbal and Dietary Supplements, updated Jul 1, 2025. Botanical identity and contaminant testing: Contaminants.

EUTECNIQ Proposed

Where EUTECNIQ would work

EUTECNIQ would confirm botanical identity and screen for adulterants before any liver-safety question is asked, test extracts against P450 and transporter probes in vitro, and report exposure as parent compounds and conjugates. Liver tests in any human study would be read with the R ratio and Hy's law thresholds above. None of this exists yet.

Related pages: ADME, PK and PBPK · Risk assessment · Contaminants · Supplement–drug interactions · Supplements A–Z

Health information notice

These statements have not been evaluated by the Food and Drug Administration. Nothing on this site is intended to diagnose, treat, cure or prevent any disease. The information is for general education only. It is not medical advice and does not replace advice, diagnosis or treatment from a qualified health professional, or any medication a clinician has prescribed.

Do not start, stop or change any medicine, supplement or herbal product because of something you read here. Talk to your doctor or pharmacist first, especially if you are pregnant or breastfeeding, have a medical condition, take any medicines, are preparing for surgery, or are caring for a child. Research findings, traditional uses and safety data are reported as their sources state them; they are not recommendations or claims about any product. If you think you may have a medical emergency, call your doctor or 911 at once.

EUTECNIQ Sciences sells no supplements or health products today and receives nothing from any product or ingredient mentioned on this site; any commercial interest it later has in an ingredient will be stated on that ingredient’s page.

Ready to check what you take?

Free tools built on published reference values, each with its sources and its limits. Your lists stay in your browser.