Practice questions

Pharmacokinetics describes what the body does to a drug. Four processes determine what happens to an oral dose: absorption, distribution, metabolism and excretion. Together, these four processes are called ADME. ADME determines how much of the drug reaches its site of action, and for how long. Watch at 0:00

The tablet swells and breaks into smaller particles; this step is called disintegration. The drug then dissolves: molecules leave the surface of each crystal and enter the fluid. Only dissolved drug can cross the gut wall. Watch at 1:08

The stomach empties its contents into the small intestine, where most absorption takes place. The small intestine is about two thirds of the five-meter length of the gut, and its contents take about three hours to pass through it. Together, the villi and microvilli increase the surface area 60 to 120 times. Watch at 1:43

Most drugs cross by passive diffusion: they dissolve in the cell membrane and move down their concentration gradient, from the gut lumen into the cell and on into the blood. Small water-soluble molecules can also pass between the cells. Tight junctions seal most of the space between the cells, so this paracellular route carries only a small fraction of most doses. Watch at 2:33

Uptake transporters carry some drugs into the cell. PEPT1 normally absorbs small peptides from food, and it also carries the antiviral drug valacyclovir. Efflux transporters, such as P-glycoprotein, pump drug from the cell back into the gut lumen, using energy from ATP. Watch at 3:08

Drug that leaves the enterocytes enters the capillaries of the villi, and the portal vein carries it to the liver. Drug that is not absorbed moves on to the colon and leaves in the feces. Watch at 3:35

The fraction of the dose that crosses from the gut lumen into the gut wall is called the fraction absorbed, Fa. Watch at 3:53

Distribution is the reversible movement of a drug from the blood into the tissues and back. It decides how much of the drug reaches the tissue where it acts. Watch at 0:00

Organs receive very different shares of the blood that the heart pumps each minute, the cardiac output. The kidneys, the liver and the brain receive a large blood flow for their weight, so drug reaches them quickly. Fat is a large tissue with a small blood flow for its weight, so drug reaches it slowly. Watch at 0:32

In muscle, skin and fat, the endothelial cells form a continuous wall, and small water-soluble molecules pass only through narrow gaps between the cells. In the gut and the kidney, the endothelial cells have pores called fenestrae. In the liver, the capillaries, called sinusoids, have large openings, so even large molecules reach the liver cells. Watch at 1:11

In plasma, many drugs bind to proteins, mainly albumin and alpha-1-acid glycoprotein. The binding is reversible, and only the unbound drug can cross cell membranes and act on the target. Watch at 1:58

Drugs also bind inside tissues. Chloroquine binds so extensively in the tissues that almost all of the chloroquine in the body is outside the plasma. Watch at 2:32

In the brain, tight junctions close the gaps between the endothelial cells almost completely. These cells also carry P-glycoprotein, the same efflux transporter as in the gut wall. Together, they form the blood-brain barrier. Watch at 2:47

The placenta is a second barrier. Nearly all drugs given in pregnancy cross the placenta and reach the fetus to some degree. Efflux transporters in the placenta, such as P-glycoprotein and BCRP, limit the transfer of some drugs. Watch at 3:16

Metabolism is the chemical change of a drug by enzymes, mainly in the liver. Metabolism usually ends the action of the drug. It also usually makes the drug more water-soluble, so the kidneys can excrete the products in the urine, or the liver can excrete them in the bile. Watch at 0:00

Phase 1 reactions add or expose a functional group, most often by oxidation. Phase 2 reactions join the drug, or its phase 1 metabolite, to an endogenous molecule, such as glucuronic acid, sulfate or glutathione. Watch at 0:41

Cytochrome P450 enzymes, or CYPs, carry out most of these reactions. About a dozen CYPs metabolize 70 to 80 percent of the drugs in clinical use, and CYP3A4 alone carries out about a quarter of the CYP reactions with drugs. Watch at 0:52

For an oral dose, metabolism starts before the drug reaches the systemic circulation. Enzymes in the enterocytes, mainly CYP3A4, can metabolize part of the absorbed drug. On this first passage through the liver, liver enzymes metabolize another part. Watch at 1:42

The oral bioavailability, F, is the fraction of the dose that reaches the systemic circulation unchanged. Watch at 2:27

Almost all of an oral dose of midazolam is absorbed. The gut wall then removes a little over 40 percent of the absorbed drug, and the liver removes a little over 40 percent of the rest, so F is only about 30 percent. Watch at 2:38

Grapefruit juice decreases the amount of CYP3A4 in the gut wall, and it increases the plasma concentrations of felodipine, an antihypertensive drug. Watch at 2:56

Ketoconazole and itraconazole inhibit CYP3A4: in one study, they increased the exposure to oral midazolam 10 to 15 times. Rifampicin induces CYP3A4: it increases the amount of the enzyme in the gut wall and the liver. In another study, rifampicin reduced the exposure to oral midazolam by 96 percent. Watch at 3:19

Metabolism can also activate a drug. Valacyclovir is a prodrug: it is hydrolyzed to the antiviral drug acyclovir, and 54 percent of an oral dose reaches the systemic circulation as acyclovir. Watch at 3:45

CYP2D6 converts a small part of a dose of codeine into morphine, which is more active than codeine. Watch at 4:02

Excretion removes the drug and its metabolites from the body. The first main route is the kidneys, which excrete drug into the urine. The second is the liver, which excretes drug into the bile and then the feces. Watch at 0:00

Other routes carry a small fraction of most drugs. Inhaled anesthetics such as desflurane are excreted almost entirely in exhaled air, and many drugs pass into breast milk. Watch at 0:22

Each kidney contains many small units called nephrons. Each nephron starts with a filter, the glomerulus, and continues as a tubule, where the filtered fluid becomes urine. Watch at 0:40

Three processes in the nephron determine renal excretion. In the glomerulus, unbound drug is filtered from the blood into the tubule. In the tubule, transporters secrete some drugs from the blood into the urine. Lipophilic drug in the tubule can diffuse back into the blood, which is tubular reabsorption. Watch at 0:56

Lipophilic drug in the tubule can diffuse back into the blood, which is tubular reabsorption. Polar metabolites are reabsorbed less, so they stay in the urine. This is why metabolism increases renal excretion. Watch at 1:19

The second main route is the bile. In the liver, transporters in the membrane of the liver cells secrete some drugs and metabolites into the bile. The bile drains into the small intestine. Watch at 1:29

Part of the drug can be reabsorbed from the gut and return to the liver; this is enterohepatic circulation. The rest leaves the body in the feces. Watch at 1:42

Paracetamol is also called acetaminophen. It is absorbed rapidly, and 70 to 90 percent of an oral dose reaches the systemic circulation, so its F is 0.7 to 0.9. The liver metabolizes most of the dose by phase 2 reactions. Watch at 0:00

Paracetamol is also called acetaminophen. It is absorbed rapidly, and 70 to 90 percent of an oral dose reaches the systemic circulation, so its F is 0.7 to 0.9. Watch at 0:21

The rate of absorption depends mainly on the rate of gastric emptying. Food slows gastric emptying, so food delays the absorption of paracetamol. Watch at 0:41

Paracetamol binds little to plasma proteins, and it distributes evenly through most tissues. It also crosses the blood-brain barrier, and the central nervous system is thought to be its main site of action. Watch at 0:49

The liver metabolizes most of the dose by phase 2 reactions. UGTs form the glucuronide, and sulfotransferases form the sulfate. CYP2E1 and other CYPs convert a small fraction to a reactive metabolite, NAPQI, which glutathione normally inactivates. Watch at 1:04

After an overdose, sulfation and then glucuronidation reach their maximum rate, so a larger fraction of the dose forms NAPQI. NAPQI uses up the glutathione and then damages the liver cells. The antidote, acetylcysteine, restores glutathione. Watch at 1:29

The kidneys then excrete the metabolites. Within 24 hours, 85 to 95 percent of a therapeutic dose appears in the urine, mainly as the glucuronide and the sulfate. Only about 4 percent is excreted unchanged. Watch at 1:51

After an oral dose, the concentration rises while absorption is faster than elimination. It reaches a peak, Cmax, at the time tmax, and then falls. The area under the curve, AUC, measures the exposure: the plasma concentration added up over time. Watch at 2:10

The ratio of the AUC after an oral dose to the AUC after an equal intravenous dose is the bioavailability, F. Watch at 2:42

The fall of the curve shows elimination. For paracetamol, the plasma concentration falls by half about every two hours: its half-life is about two hours. Watch at 2:53

For paracetamol, the concentration falls by half every two hours, which is why a dose is repeated every 4 to 6 hours. Watch at 3:31

A drug in a tablet or capsule must dissolve in the gut fluid before it can pass through the gut wall. Dissolution is the process in which molecules leave the solid drug and enter the fluid. Watch at 0:06

The Biopharmaceutics Classification System, or BCS, sorts drugs into four classes. It uses two properties that limit oral absorption: solubility, and permeability through the gut wall. Watch at 0:37

By this solubility criterion, about 40 percent of marketed drugs are poorly soluble. Among drugs in development, the share is about 90 percent. Watch at 1:09

The Noyes-Whitney equation describes the dissolution rate. The rate rises with three terms: the diffusion coefficient, the surface area of the solid, and the solubility minus the drug concentration in the fluid. A larger surface area or a higher solubility therefore gives a faster rate. Watch at 1:54

When a manufacturer mills a drug to smaller particles, the same mass has a larger total surface area. Very small particles also have a thinner diffusion layer. Because the equation divides by h, a thinner layer gives a faster rate. Watch at 2:53

When the pH is below the pKa, most molecules of a weak base take up a proton and become cations, positively charged ions. The charged form is more soluble in water than the neutral form. When the pH is above the pKa, most molecules of a weak acid give up a proton and become anions, negatively charged ions. A weak acid is therefore more soluble at higher pH, and a weak base at lower pH. Watch at 3:18

Compared with the neutral drug, a salt usually has a higher solubility and a faster dissolution rate. Watch at 4:53

In the control treatment, the mean ketoconazole AUC was 17.9 milligram hours per liter. After omeprazole with water, the mean AUC was 3.5, about one fifth of the control value. When they took ketoconazole with an acidic cola drink at pH 2.5 after omeprazole, absorption rose to about 65 percent of the control mean. Watch at 5:18

Permeability is the rate at which a drug crosses a barrier for each unit of area and each unit of concentration difference. Its unit is centimeters per second. Watch at 0:52

A drug can cross the gut wall by three main routes. The first is passive diffusion through the cells, the transcellular route. The drug dissolves in the lipid membrane of the enterocyte and moves from high to low concentration. It crosses the membrane that faces the lumen, moves through the cell, and then crosses the membrane that faces the blood. Watch at 1:23

The second route is between the cells, the paracellular route. The tight junctions let small hydrophilic drugs through and restrict larger hydrophilic molecules. Watch at 2:37

Uptake transporters carry drugs into the cell, and efflux transporters pump drugs out of the cell. The efflux transporters P-glycoprotein, or P-gp, and BCRP, the breast cancer resistance protein, pump drugs back into the lumen. For drugs that P-gp or BCRP pump, efflux can lower the fraction absorbed and the oral bioavailability. Watch at 3:09

One study compared the permeability of 20 drugs and peptides in Caco-2 cells with their absorption in humans. Drugs with complete absorption had a permeability above 1 times 10 to the minus 6 centimeters per second. Drugs and peptides with less than 1 percent absorption had 1 times 10 to the minus 7 centimeters per second or less. Watch at 4:27

Rifampin raised the amount of P-gp in the duodenum, the first part of the small intestine, 4.2-fold. During rifampin, the AUC of oral talinolol was 35 percent lower than before, and the AUC of intravenous talinolol was 21 percent lower. Watch at 5:12

After an oral dose, most of the absorbed drug passes through two organs before it reaches the systemic circulation. Enzymes in both organs can metabolize the drug, that is, change it chemically into other compounds. This loss of drug in the gut wall and the liver, on this first pass through the two organs, is first-pass metabolism. Watch at 0:06

Fa, the fraction absorbed, is the fraction of the dose that enters the gut wall. Fg is the fraction of this drug that escapes metabolism in the gut wall. Fh is the fraction of the drug that enters the liver and escapes removal there, mainly by metabolism. Fg and Fh each equal 1 minus the extraction ratio, the fraction of the drug that enters an organ and that the organ removes. Watch at 1:15

In the human small intestine, CYP3A makes up on average about 80 percent of the measured cytochrome P450 protein. CYP3A is a small group of enzymes, and in the gut wall it is mainly CYP3A4. The enterocytes also attach sulfate or glucuronic acid to some drugs. Watch at 1:41

The second organ is the liver. There, CYP3A makes up about 40 percent of the measured P450 protein. Across the two organs, the enzymes most often involved in first-pass metabolism are CYP3A4, CYP2D6, CYP1A2, CYP2C9 and CYP2C19. CYP1A2 is present in the liver, but the gut wall contains little or none. Watch at 2:21

When an organ removes most of a drug, a small change in its extraction ratio gives a large change in the fraction that escapes. For example, a fall in the extraction ratio from 0.9 to 0.8 doubles the fraction that escapes, from 0.1 to 0.2. A drug with a large first-pass metabolism therefore often has a variable oral bioavailability. Watch at 3:37

Some routes avoid the first pass. The lining of the mouth has a dense blood supply. Drug that crosses this lining therefore enters the systemic circulation without a passage through the gut wall or the liver. A buccal dose rests between the cheek and the gum, and a sublingual dose rests under the tongue. Watch at 4:08

The intravenous dose gave the extraction ratio of the liver, 0.44, so Fh was 0.56. From this difference, the study estimated that the gut wall removed 43 percent of the midazolam that entered it, so Fg was 0.57. The two organs had similar extraction ratios, and the gut-wall value varied more between people. Watch at 4:35

In a second study, eight healthy volunteers received 5 milligrams of midazolam by the buccal route and, on another occasion, intravenously. The buccal AUC was 74.5 percent of the intravenous AUC, so the buccal bioavailability was 74.5 percent. Watch at 6:04

Bioavailability, F, is the fraction of a dose that reaches the systemic circulation unchanged. An intravenous dose enters the systemic circulation directly, so its F is 1 by definition. A dose by any other route must first be absorbed, and part of the dose can be lost on the way. Its F can then be less than 1. Watch at 0:06

Fa, the fraction absorbed, is the fraction of the dose that crosses into the gut wall. Fg is the fraction of drug that enters the gut wall and escapes metabolism there. Fh is the fraction of drug that enters the liver and escapes metabolism there. F equals Fa times Fg times Fh. Watch at 0:47

The first is absolute bioavailability, F measured against an intravenous dose. In a crossover study, the same people receive both doses on separate occasions. F is therefore the oral AUC divided by the intravenous AUC, after each AUC is divided by its dose. Watch at 1:49

The second type is relative bioavailability. It compares two products of the same drug, or two ways of giving it, without an intravenous dose. The calculation uses the same ratio of dose-normalized AUC values, with one product as the reference. A relative bioavailability of 100 percent means that the two products deliver the same fraction of the dose to the systemic circulation. Watch at 3:05

Two products are bioequivalent when their rate and extent of absorption agree within set limits. Regulators use the AUC as the measure of the extent, and the peak plasma concentration, Cmax, as the measure of the rate. For most products, the 90 percent confidence interval of this ratio must lie between 80 and 125 percent. Watch at 3:41

In one study, 20 healthy adults received 5 milligrams intravenously and 10 milligrams orally. The absolute oral bioavailability was 49 percent. In one study, the relative bioavailability of an oral solution against the tablet was 105 percent. A crushed tablet in water met the bioequivalence limits. Watch at 4:53

A food effect is a change in exposure when a person takes a dose with a meal instead of in the fasted state. A positive food effect increases the exposure, and a negative food effect decreases it. Watch at 0:05

The first is gastric emptying, the passage of stomach contents into the small intestine. A meal slows it. Absorption from a retained tablet therefore starts later. Watch at 0:49

Bile salts are detergent molecules in bile, and the gallbladder releases more bile into the duodenum after a meal. Bile salts and lipids form micelles, small clusters with a lipid core. Lipophilic drugs, which dissolve better in fat than in water, dissolve in this core. Micelles therefore raise the solubility of poorly soluble, lipophilic drugs. Watch at 1:25

After a solid and liquid meal, the median pH in the stomach rose from 1.7 to a peak of 6.7. It returned to the fasted value within 2 hours. This short rise lowers the solubility of some weak bases, drugs that dissolve better in acid than at a neutral pH. In the duodenum, the median pH fell from 6.1 to 5.4, and it stayed lower for the whole 4-hour observation period. Watch at 2:08

In six healthy subjects, a high-protein meal raised the estimated liver blood flow by 34 percent. The oral bioavailability of propranolol rose from 27.2 to 45.5 percent. With a faster blood flow, the liver removes a smaller fraction of such a drug in one passage. Watch at 2:47

A fatty meal also increases the transport of some drugs through the intestinal lymph. After a fatty meal, the enterocytes, the cells that line the gut wall, release lipoproteins, particles of fat with a protein coat, into the lymph. Some highly lipophilic drugs, typically with a log P above 5, bind to these lipoproteins and enter the lymph with them. The fraction of the dose that enters the lymph avoids metabolism in the liver on the first pass. Watch at 3:46

Positive food effects came mainly from solubilization: food raised the solubility of the drug in the gut. Negative food effects occurred mainly with highly hydrophilic drugs. These drugs dissolve well in water, and they are often absorbed only in a short part of the intestine. Watch at 4:37

The low-fat breakfast increased the lapatinib AUC 2.67-fold and the Cmax 2.42-fold. The high-fat breakfast increased the AUC 4.25-fold and the Cmax 3.03-fold. The high-fat breakfast gave the larger increase in exposure. Watch at 5:24

The route of administration is the place where a dose enters the body. The formulation is the product that holds the drug, for example a solution, a patch or an extended-release tablet. The route and the formulation determine which barriers the drug crosses and how fast it enters the blood. Watch at 0:06

A subcutaneous injection places the drug in the tissue under the skin, and an intramuscular injection places it in a muscle. In rats, salicylate, a small drug molecule, entered the blood through the vessels under the injection site. It entered the blood about four times faster after an intramuscular injection than after a subcutaneous one. In sheep, molecules larger than 16,000 daltons, the unit of molecular weight, were absorbed mainly into the lymphatic vessels. Watch at 1:33

Only part of an inhaled dose reaches the lung. The lung has a large surface area for absorption and a low enzyme activity, so little drug is metabolized there. Drug absorbed from the lung reaches the systemic circulation without a first pass through the liver. The swallowed part passes through the gut wall and the liver first. Watch at 2:28

Patches deliver drug through the skin at a steady rate over many hours. The stratum corneum, the thin outermost layer of dead cells in the skin, is the main barrier. The 500 dalton rule proposes that a compound must have a molecular weight below 500 daltons to cross the skin. Watch at 3:05

Drug that crosses the lining of the mouth enters the systemic circulation directly. It avoids the gut wall and the liver on the first pass. The patient can swallow part of the dose with saliva. Watch at 3:38

The fifth is the extended-release tablet, which releases the drug over many hours. After it leaves the stomach, a tablet passes through the small intestine in about 3 hours. Then it enters the colon, and drug that the tablet releases later must be absorbed there. A long release time can therefore lower the bioavailability. Watch at 4:05

For a few highly lipophilic drugs, which dissolve better in fat than in water, part of the dose enters the intestinal lymph. In the enterocytes, the cells that line the gut wall, these drugs associate with lipoproteins called chylomicrons. The chylomicrons carry the drug into the lymph, and the lymph returns it to the blood without a first pass through the liver. Watch at 4:44

In eight surgical patients, 92 percent of the fentanyl that left the patch reached the systemic circulation. A buccal tablet gave a bioavailability of 65 percent in 26 healthy subjects. When the subjects swallowed the tablet, the bioavailability was 31 percent. Watch at 5:24

The volume of distribution, Vd, is the amount of drug in the body divided by the plasma concentration, after distribution is complete. Watch at 0:18

Vd is an apparent volume, and it can be larger than the whole body. Chloroquine, a drug for malaria, had a Vd of 116 to 285 liters per kilogram. In a 70-kilogram adult, this is more than 8,000 liters, almost 200 times the body water. Watch at 0:58

Only the unbound drug crosses cell membranes and enters the cells of the tissue. Watch at 1:40

Many drugs bind reversibly to plasma proteins, mainly albumin and alpha-1-acid glycoprotein. Acids are drugs that can give up a proton, a hydrogen ion, and become negatively charged. They bind strongly to albumin and weakly in tissue, so their Vd is usually 0.1 to 0.25 liters per kilogram. Watch at 2:24

Bases are drugs that can take up a proton and become positively charged. They bind strongly to the negatively charged phospholipids, so they tend to have a larger Vd than acids and neutral drugs. When a drug binds more in tissue than in plasma, its Vd is larger than the body water, 0.6 liters per kilogram. Watch at 3:02

A weak base exists in a neutral and a charged form. The neutral form crosses the phospholipid membrane of the lysosome. The charged form crosses the membrane only slowly, so the drug collects inside, a process called ion trapping. Like binding in tissue, ion trapping raises Vd. Watch at 3:44

Plasma proteins bind only about 60 percent of chloroquine, so its large Vd comes from tissue. Chloroquine is a weak base, so lysosomes can trap it. Watch at 4:23

The terminal half-life is the time for the plasma concentration to fall by half after distribution is complete. A larger Vd gives a longer half-life at the same clearance, and a higher clearance gives a shorter one. Watch at 5:23

A drug can bind to plasma proteins and separate from them again. The fraction unbound, fu, is the unbound concentration divided by the total concentration, bound plus unbound. Watch at 0:23

Two plasma proteins bind many drugs. Acidic drugs, which can give up a proton, a hydrogen ion, tend to bind to albumin. The second, alpha-1-acid glycoprotein, is present at 0.05 to 0.1 grams per deciliter. Basic drugs, which can take up a proton, tend to bind to it. Watch at 1:01

Disease changes the concentrations of both: albumin decreases during inflammation. Alpha-1-acid glycoprotein can rise to more than 4 times its normal concentration in infection, cancer and myocardial infarction, a heart attack. Watch at 1:36

The first is the action at the target, the protein on which the drug acts, such as a receptor. Only unbound drug binds to the target. Watch at 2:07

Plasma binding and tissue binding set the volume of distribution, the amount of drug in the body divided by the total plasma concentration. High binding in plasma gives a small volume. High binding in tissue gives a large volume. Watch at 2:56

A low-clearance drug has a liver clearance much smaller than the liver blood flow. For such a drug, the liver clearance is close to fu times CLint. Suppose fu doubles for such a drug, and the dose rate, the dose given per unit time, and CLint stay the same. The clearance doubles, and the total concentration at steady state falls to half. Watch at 3:48

Warfarin shows how fu changes the clearance in patients. In the 31 patients, warfarin clearance ranged from 1.16 to 4.35 milliliters per hour per kilogram and increased with fu. Watch at 4:29

Suppose fu doubles for such a drug, and the dose rate, the dose given per unit time, and CLint stay the same. The clearance doubles, and the total concentration at steady state falls to half. The unbound concentration stays the same, because, for such a drug, it depends only on the dose rate and CLint. For warfarin at steady state, at the same dose rate and CLint, a higher fu gives a lower total concentration but the same unbound concentration. Watch at 5:44

The blood-brain barrier is the wall of the brain capillaries, the smallest blood vessels. Endothelial cells form this wall. Watch at 0:13

Kp is the total brain concentration, bound plus unbound, divided by the total concentration in plasma, the liquid part of the blood. Kp,uu is the same ratio for unbound drug. Watch at 0:32

Tight junctions close the gaps between the endothelial cells, so a drug must cross the cells. It can cross by passive diffusion, movement from a higher to a lower concentration without energy use. Lipophilicity, the tendency to dissolve in lipids, mainly sets the rate, because cell membranes are made of lipids. For this, a drug needs a molecular weight below 400 daltons and fewer than 8 hydrogen bonds. Watch at 1:16

Efflux transporters are proteins that carry their substrates, the molecules they transport, out of a cell. P-glycoprotein, or P-gp, and breast cancer resistance protein, or BCRP, lie in the membrane that faces the blood. They use energy from ATP, adenosine triphosphate, the molecule that stores energy in cells. Watch at 1:56

At steady state, when concentrations stay constant, Kp,uu equals the clearance into the brain divided by the clearance out of the brain. If a drug crosses by passive diffusion alone and has little metabolism in the brain, the clearances are similar, and Kp,uu is close to 1. Kp,uu is below 1 when active efflux dominates and above 1 when active entry dominates. Watch at 2:53

Binding to brain tissue raises the total brain concentration and Kp. At steady state, it leaves the unbound brain concentration and Kp,uu unchanged. Watch at 3:50

A study in 37 healthy volunteers measured how much P-gp limits the brain uptake of carbon-11 N-desmethyl-loperamide, a radioactive P-gp substrate. In 12 volunteers, the substrate was injected during an infusion of tariquidar, a drug that blocks P-gp, into a vein. Its brain uptake was then about 5 times the uptake without tariquidar. Watch at 4:06

Loperamide is an opioid, a drug that acts on the same receptors as morphine. It treats diarrhea and acts outside the brain, and it is a P-gp substrate. In mice without the mdr1a gene for P-gp, the brain uptake of loperamide and its metabolites rose 7 times. In these mice, oral loperamide had opioid effects in the brain and spinal cord. Watch at 4:55

A continuous layer of fused cells, the syncytiotrophoblast, is the main barrier between the maternal blood and the fetus, the unborn child. Passive diffusion is movement from a higher to a lower concentration, without energy use. Nearly all drugs reach the fetal blood to some extent by passive diffusion. Watch at 0:25

Drug bound to plasma proteins cannot cross the placenta, so binding slows the transfer. The efflux transporters P-glycoprotein and breast cancer resistance protein carry drugs back into the maternal blood. Fetal Kp,uu is the unbound drug concentration in fetal plasma divided by that in maternal plasma. At the end of pregnancy, it was below 0.25 for darunavir and lopinavir, two drugs that P-glycoprotein carries. Watch at 0:59

Fetal plasma is slightly more acidic than maternal plasma. At a lower pH, a larger share of a weak base takes up a proton, a hydrogen ion, and becomes charged. The charged form crosses membranes slowly, so more of it stays in the fetal blood. This ion trapping can raise the fetal concentration above the maternal concentration. Watch at 1:38

The milk-to-plasma ratio is the drug concentration in milk divided by that in maternal plasma. The infant dose is the milk concentration times the daily milk intake per kilogram of infant weight. The relative infant dose is this dose as a percentage of the daily maternal dose per kilogram of maternal weight. Watch at 2:12

A relative infant dose below 5 to 10 percent is usually taken as a safety threshold. Below 10 percent, infant plasma concentrations stay below those that treat adults, unless infants remove the drug much more slowly per kilogram than adults. Watch at 2:40

The relative infant dose depends on the milk-to-plasma ratio and the maternal clearance. At a given dose, a low maternal clearance gives high plasma and milk concentrations. Phenobarbital has a low clearance. Its milk-to-plasma ratio is below 1, but its relative infant dose is 50 to 70 percent. Watch at 3:44

In 9 women, the lamotrigine concentration in umbilical cord plasma, which comes from the fetal blood, was similar to the maternal plasma concentration at delivery. At 72 hours after birth, the median infant plasma concentration was still 75 percent of the cord concentration. In the same study, the infant dose was 0.51 milligrams per kilogram per day, and the relative infant dose was 9.2 percent. The infant plasma concentration, 18.3 percent of the maternal concentration, was about twice the relative infant dose. Watch at 4:22

Metabolism is the chemical change of a drug by enzymes, proteins that make chemical reactions faster. The product is a metabolite. For many drugs, metabolism is the main route of elimination. Metabolism usually ends the action of the drug. Watch at 0:08

A lipophilic drug dissolves well in fat. The kidney filters such a drug, but much of it crosses back into the blood. Without metabolism, the drug therefore stays in the body longer and can cause toxicity. Metabolism usually forms products that are more water-soluble than the drug. Watch at 0:38

A phase 1 reaction adds or exposes a functional group, a small group of atoms to which phase 2 can attach a molecule. Examples are the hydroxyl, carboxyl, amine and thiol groups. Phase 1 has three main types of reaction: oxidation, reduction and hydrolysis. Watch at 1:30

Oxidation removes electrons from a molecule and often adds an oxygen atom. A typical CYP oxidation uses one oxygen molecule and one NADPH molecule, which supplies two electrons. The enzyme adds one oxygen atom to the drug, and the other oxygen atom, with the two electrons and two hydrogen ions, forms water. Watch at 2:05

Cytochrome P450 enzymes, or CYP enzymes, carry out most phase 1 oxidations. In the 200 most prescribed US drugs of 2014, CYP enzymes were 48 percent of the enzymes on major metabolic routes. Watch at 2:36

Reduction, the second type, adds electrons to a molecule. Hydrolysis, the third type, uses water to split a bond. An ester bond joins an acid to an alcohol. Esterases, such as the carboxylesterases, split this bond and leave a carboxyl group and a hydroxyl group. Watch at 3:12

A phase 2 reaction attaches a small molecule or group of atoms to a functional group. The product is a conjugate. The main ones attached are glucuronic acid, a sulfate group, glutathione, an acetyl group and a methyl group. Conjugates with glucuronic acid, sulfate or glutathione are more water-soluble than the drug. Watch at 3:52

Phase 2 enzymes are mainly transferases, which move a group from a donor molecule to the drug. Five classes carry out most phase 2 reactions. They are the UDP-glucuronosyltransferases, or UGTs, and the sulfotransferases, or SULTs. The other three are the N-acetyltransferases, the glutathione S-transferases and the methyltransferases. Watch at 4:39

Paracetamol already has a hydroxyl group, so phase 2 enzymes act on most of the dose with no phase 1 step. Four UGTs, UGT1A1, UGT1A6, UGT1A9 and UGT2B15, add glucuronic acid. In adults, mainly SULT1A1 and SULT2A1 add a sulfate group. After a recommended dose, the glucuronic acid conjugate is 52 to 57 percent of urinary metabolites, and the sulfate conjugate is 30 to 44 percent. Watch at 5:35

In a phase 1 reaction, CYP enzymes, mainly CYP2E1, oxidize a small part of the dose to N-acetyl-p-benzoquinone imine, or NAPQI. After a highly toxic dose, the glucuronic acid route also reaches its maximum rate, and more than 15 percent of the dose forms NAPQI. NAPQI then depletes the glutathione in the liver, binds to proteins in the liver cells and kills these cells. Watch at 6:59

Cytochrome P450 enzymes, or CYP enzymes, form a superfamily, a large group of proteins from genes with a common origin. Each CYP enzyme has a heme group, a ring structure with an iron atom at its center. The iron atom binds oxygen. In most reactions, the enzyme oxidizes the substrate: it adds one oxygen atom to the molecule. Watch at 0:38

In each name, the number after CYP gives the family, the letter gives the subfamily, and the last number gives the individual gene. CYP3A4 is the enzyme from gene 4 of subfamily A in family 3, and CYP3A is the whole subfamily. Watch at 1:14

The liver contains the highest amounts of CYP3A4, CYP2C9, CYP2C8, CYP2E1 and CYP1A2. CYP2A6, CYP2D6, CYP2B6, CYP2C19 and CYP3A5 occur in lower amounts. Watch at 1:58

The human genome contains 57 functional CYP genes. Of these 57, about a dozen enzymes from families 1, 2 and 3 metabolize 70 to 80 percent of all drugs in clinical use. A review listed the major metabolic enzymes of the 200 most prescribed drugs in the United States in 2014. Among the major metabolic routes of these drugs that run through CYP enzymes, CYP3A4 accounts for 40 percent. Watch at 2:30

CYP2D6 prefers basic molecules, which take up a proton. CYP2C9 prefers slightly acidic molecules, which give up a proton. CYP3A4 accepts a wide range of lipophilic molecules, which dissolve well in fat. Watch at 3:33

An inhibitor is a substance that lowers the activity of an enzyme. The antifungal drug ketoconazole binds to CYP3A4 and lowers the metabolism of other CYP3A4 substrates. For a drug that CYP3A4 mainly metabolizes, the concentration in plasma, the liquid part of the blood, then rises. Watch at 4:01

An inducer is a substance that raises the amount of an enzyme. Most inducers activate a nuclear receptor, a protein that binds to the DNA of the enzyme gene. The cell then makes more of the enzyme. Rifampicin induces CYP3A4, and the plasma concentration of CYP3A4 substrates can fall. Watch at 4:39

Nine volunteers took part in three phases, four weeks apart. They took 400 milligrams of ketoconazole, 200 milligrams of itraconazole or placebo, a dose with no active drug, once a day for four days. Both antifungal drugs raised the midazolam AUC to 10 to 15 times the placebo value. They raised the peak concentration, the highest plasma concentration, to 3 to 4 times the placebo value. Watch at 5:46

In a second study, ten healthy volunteers took 600 milligrams of rifampicin or placebo once a day for five days. Compared with placebo, rifampicin lowered the midazolam AUC by 96 percent and the peak concentration by 94 percent. The half-life, the time in which the plasma concentration falls by half, fell from 3.1 to 1.3 hours. Watch at 6:39

A prodrug is a compound with little or no activity of its own. In the body, an enzyme or a chemical reaction converts it to the active compound, the active metabolite. Chemists design many prodrugs to correct a poor property of the active compound, such as poor absorption. Watch at 0:37

A prodrug is a compound with little or no activity of its own. In the body, an enzyme or a chemical reaction converts it to the active compound, the active metabolite. A drug with activity of its own can also form an active metabolite. Watch at 0:59

The first is valacyclovir, an antiviral prodrug. Valacyclovir is acyclovir joined to the amino acid valine by an ester bond, a bond that water can split. Enzymes such as valacyclovir hydrolase split this bond and release acyclovir. Watch at 1:29

Oral bioavailability is the fraction of the dose that reaches the systemic circulation unchanged, through the gut wall and the liver. Twelve healthy volunteers received valacyclovir by mouth and, as the 100 percent reference, acyclovir into a vein. The oral bioavailability of acyclovir from valacyclovir was 54.2 percent, three to five times the value for oral acyclovir. Watch at 2:36

Carboxylesterase-1, a hydrolase, converts 85 percent of each dose to inactive acids. The other 15 percent is available for activation by two oxidation steps. Both steps need cytochrome P450 enzymes, or CYP enzymes, which oxidize many drugs in the liver. The first step forms 2-oxo-clopidogrel, and the second forms the active metabolite. Watch at 3:31

A reduced-function allele of CYP2C19 lowers the enzyme activity. The carriers had a 32.4 percent lower exposure to the active metabolite than the other volunteers. The drop in platelet aggregation with clopidogrel was 9 percentage points smaller in the carriers. Carriers thus had less active metabolite and a smaller effect on their platelets. Watch at 5:15

The third is codeine, a drug with weak activity of its own. CYP2D6 converts a small part of each dose to morphine, and morphine produces most of the pain relief. Watch at 6:49

The fourth is paracetamol, a drug for pain and fever. It forms a toxic metabolite. At a therapeutic dose, CYP enzymes oxidize 5 to 10 percent of the dose to this metabolite, NAPQI. After a highly toxic dose, more than 15 percent forms NAPQI, and the glutathione in the liver falls. Watch at 7:04

A gene, a section of DNA, holds the code for a protein, such as an enzyme. Pharmacogenetics studies how inherited DNA differences change drug response. Each person inherits one version of most genes from each parent, and each version is an allele. The pair of alleles is the genotype. Watch at 0:36

The phenotype is the enzyme activity of a person, and the genotype predicts it. Of the four main groups, a poor metabolizer has little or no activity, and an intermediate metabolizer has decreased activity. A normal metabolizer has normal activity, and an ultrarapid metabolizer has increased activity from extra gene copies or increased-function alleles. Watch at 1:10

Each CYP2D6 allele receives an activity value: 0 for no function, 0.25 or 0.5 for decreased function, and 1 for normal function. For an allele with extra gene copies, the value is multiplied by the number of copies. The activity score is the sum of the values of the two alleles. A poor metabolizer has a score of 0, and an intermediate metabolizer has a score of 0.25 to 1. Watch at 2:10

In one compilation, the no-function allele CYP2D6 star 4 makes up 18 percent of alleles in Europeans and 0.6 percent in East Asians. From the allele frequencies, it predicts 0.4 to 5.4 percent poor metabolizers and 1 to 21 percent ultrarapid metabolizers across the major population groups. Europeans have the highest frequency of poor metabolizers, and East Asians and Oceanians have the lowest. Oceanians have the highest frequency of ultrarapid metabolizers, at 21.2 percent, and East Asians have the lowest, at 1.4 percent. Watch at 3:09

An inhibitor is a drug that lowers the activity of an enzyme. For a patient who takes a strong CYP2D6 inhibitor, the CPIC guideline sets the activity score to 0. For a moderate inhibitor, it multiplies the score by 0.5. The genotype stays the same, and the phenotype changes. Watch at 4:16

For some heart and brain blood vessel conditions, the CYP2C19 genotype guides the choice between clopidogrel and another antiplatelet drug. The CYP2C9 gene, for an enzyme that metabolizes warfarin, and the VKORC1 gene, for the warfarin target, guide the warfarin dose. The TPMT and NUDT15 genes code for two enzymes that inactivate thiopurine drugs, a group of immunosuppressant and anticancer drugs. These genes guide the initial dose of two thiopurine drugs, azathioprine and mercaptopurine. Watch at 4:45

Codeine relieves pain mainly through morphine, which CYP2D6 forms. In studies of healthy volunteers, poor metabolizers had a 96 percent lower mean morphine exposure in serum than normal and intermediate metabolizers. In this test, codeine gave poor metabolizers no more pain relief than placebo, a dose with no active drug. The median morphine exposure was 45 percent higher in ultrarapid metabolizers than in normal metabolizers. Watch at 5:35

For normal metabolizers, the CPIC guideline recommends the dose on the product label. For intermediate metabolizers, it recommends the same dose and close checks of pain relief. For poor and ultrarapid metabolizers, it recommends another pain drug. Watch at 7:28

The extraction ratio is the fraction of the drug that enters the liver and that the liver removes in one pass of the blood. A ratio of 1 means complete removal in one pass, and a ratio of 0 means no removal. Watch at 0:40

In the usual model, three quantities set the extraction ratio. The first is hepatic blood flow, the volume of blood that passes through the liver per unit of time. The second is the unbound fraction in blood, the concentration of drug that is not bound to proteins, divided by the total concentration in blood. The third is intrinsic clearance, the rate of metabolism in the liver divided by the unbound drug concentration at the enzymes. Watch at 0:59

The well-stirred model combines the three quantities in one equation. The model gives the extraction ratio as a fraction. The numerator is the product of the unbound fraction and intrinsic clearance. The denominator is the sum of that product and hepatic blood flow. Watch at 1:56

The first is a high-extraction drug, for which the product of the unbound fraction and intrinsic clearance is much larger than hepatic blood flow. Hepatic clearance then approaches hepatic blood flow, and the extraction ratio approaches 1. The clearance therefore depends mainly on hepatic blood flow, and a small change in intrinsic clearance or the unbound fraction has little effect on it. Watch at 2:46

The second is a low-extraction drug, for which that product is much smaller than hepatic blood flow. Hepatic clearance then approaches the product itself, and the extraction ratio is small. The clearance therefore changes in proportion to the unbound fraction and to intrinsic clearance, and hepatic blood flow has little effect on it. Watch at 3:18

Three days of propranolol lowered hepatic blood flow from 1,275 to 957 milliliters per minute, by 25 percent. With less hepatic blood flow and no significant change in the product, the extraction ratio rose from 0.86 to 0.90, as the equation predicts. The lidocaine clearance from plasma fell less, from 1,030 to 866 milliliters per minute, or by 16 percent. Propranolol therefore lowered the clearance of lidocaine mainly through its effect on hepatic blood flow, as the model predicts for a high-extraction drug. Watch at 3:50

CYP2C9, a cytochrome P450 enzyme, metabolizes most of S-warfarin, the more potent of its two mirror-image forms. A change in CYP2C9 activity changes the clearance of warfarin and its unbound concentration. The total clearance is the clearance calculated from the total plasma concentration, bound plus unbound. A change in the unbound fraction changes the total clearance, and the faster or slower elimination offsets most of the change in the unbound concentration. Watch at 5:39

Plasma is the liquid part of the blood. Clearance is the volume of plasma from which the body removes all of the drug per unit of time, for example in liters per hour. The body removes only the drug, and the plasma stays in the body. Watch at 0:40

The rate of elimination, the amount of drug that the body removes per unit of time, equals the clearance times the plasma concentration. With a clearance of 52 liters per hour, the body removes 52 milligrams per hour at a concentration of 1 milligram per liter. At half that concentration, it removes 26 milligrams per hour. Watch at 0:56

The area under the whole concentration-time curve, the AUC, measures the total exposure to the drug. The body eliminates the whole dose, and the amount eliminated equals the clearance times the AUC. Clearance therefore equals the intravenous dose divided by the AUC. Watch at 1:58

The oral bioavailability, F, is the fraction of the dose that reaches the systemic circulation unchanged. Only F times the dose enters the systemic circulation. The oral dose divided by the AUC therefore gives the clearance divided by F, the apparent oral clearance. Watch at 2:38

The kidneys excrete unchanged drug into urine, and their clearance is the renal clearance. The liver removes drug by metabolism and by excretion into bile, and its clearance is the hepatic clearance. Watch at 3:13

An organ removes drug only from the blood that flows through it. Its blood flow therefore sets the upper limit of its blood clearance, the clearance based on the drug concentration in blood instead of plasma. Watch at 3:35

The organs remove drug at the same time and independently, so their clearances add. The total clearance of the body equals the renal clearance plus the hepatic clearance plus the clearance by any other route. Watch at 3:52

In healthy subjects given intravenous ciprofloxacin, the total clearance was 52.2 liters per hour. The renal clearance was 26.5 liters per hour, and the nonrenal clearance was 25.8, which add to about the total. About half of the dose appeared unchanged in urine. Watch at 4:32

Reduced kidney function lowers mainly the renal clearance. Its effect on the total clearance therefore depends on the size of the renal part. In patients with reduced kidney function, the renal clearance of ciprofloxacin was lower when the creatinine clearance was lower. The nonrenal clearance showed no correlation with the creatinine clearance, and the total clearance showed a weaker correlation than the renal clearance. Watch at 5:44

Renal clearance is the volume of plasma from which the kidneys remove all of the drug per unit of time. It equals the amount of unchanged drug that enters the urine over a time interval, divided by the AUC for the same interval. The AUC is the area under the plasma concentration-time curve. Watch at 0:20

Each kidney forms urine in many small units, the nephrons. Each nephron starts at a glomerulus, a network of small blood vessels inside a capsule. The glomerulus filters plasma into Bowman's space, the space inside the capsule. The filtered fluid then flows along a tubule and becomes urine. Watch at 0:45

Three processes set the renal clearance: filtration, secretion and reabsorption. Each process has a clearance: its rate of drug transfer divided by the plasma concentration. Renal clearance equals the filtration clearance plus the secretion clearance minus the reabsorption clearance. Watch at 1:18

The glomerular filtration rate, GFR, is the volume of plasma that passes from the glomeruli into Bowman's space per unit of time. Only drug that is not bound to plasma proteins passes into Bowman's space. The filtration clearance therefore equals the fraction unbound in plasma times the GFR. Watch at 1:42

Organic anions, with a negative charge, enter the tubule cells from the blood through OAT1 and OAT3. Efflux pumps such as MRP2 then move them into the tubule fluid. Organic cations, with a positive charge, enter from the blood through OCT2, and MATE1 and MATE2-K move them into the tubule fluid. Secretion can make the renal clearance larger than the filtration clearance. Watch at 2:17

Drug that is lipophilic, or fat-soluble, and un-ionized, without a charge, diffuses back across the lipid membranes of the tubule cells. The ionized fraction of weak acids and weak bases depends on pH, a measure of acidity. Urine pH therefore changes their reabsorption and their renal clearance. At a higher urine pH, a weak acid is more ionized, so less of it diffuses back, and its renal clearance increases. Watch at 3:12

Clinicians estimate kidney function from creatinine, a waste product mainly from muscle. The glomerulus filters creatinine freely, and the proximal tubule also secretes it. Creatinine clearance is therefore about 10 to 20 percent higher than the GFR, especially when kidney function is low. Watch at 3:59

At body pH, more than 99.9 percent of metformin is a cation, so the tubule reabsorbs little of it. Its renal clearance is about 510 milliliters per minute, about 4 times the creatinine clearance. Filtration accounts for about 120 milliliters per minute, and secretion by OCT2, MATE1 and MATE2-K for the other 390. Watch at 4:42

Tucatinib, a breast cancer drug, inhibits OCT2, MATE1 and MATE2-K in cells, and in the kidney mainly MATE1 and MATE2-K. In 17 healthy adults, tucatinib lowered the renal clearance of metformin from about 500 to 294 milliliters per minute. The GFR, measured with the filtration marker iohexol, did not change, so the fall came from less secretion. Creatinine clearance fell by 23 percent, because tucatinib also inhibits the secretion of creatinine. Watch at 5:26

In chronic kidney disease, the renal clearance of metformin falls approximately in proportion to creatinine clearance, so the same dose gives higher concentrations. The dose should therefore fall in proportion to creatinine clearance. One study gave daily doses of 1500 and 1000 milligrams in stages 3A and 3B. These doses kept blood metformin below 5 milligrams per liter, the generally accepted safe upper limit. Watch at 6:09

Bile is a fluid that liver cells make, and biliary excretion is the transport of a drug or its metabolites into bile. Bile flows from small ducts in the liver through the common bile duct into the duodenum, the first part of the small intestine. Watch at 0:21

Transporters, proteins that carry drugs across cell membranes, affect how much of a drug enters bile. Polarity, the uneven spread of electric charge in a molecule, and molecular weight, the mass of the molecule, also affect this amount. In humans, anions are more likely to enter bile when their molecular weight is above about 475 daltons, the unit of molecular mass. Watch at 0:50

A conjugate forms when an enzyme joins a drug or metabolite to a polar group. For example, glucuronic acid, a sugar acid, gives a glucuronide, an anion about 176 daltons heavier than the drug. Conjugates cross cell membranes poorly, so transporters carry them. Watch at 1:21

The first step is uptake into the hepatocyte, the liver cell that makes bile. A drug must cross two membranes of the hepatocyte, one on the blood side and one on the bile side. On the blood side, uptake transporters, for example the organic anion transporting polypeptides, OATPs, carry the drug into the cell. Watch at 2:02

The membrane on the bile side faces the bile canaliculus, a small channel between adjacent hepatocytes that drains into the bile ducts. Efflux transporters in this canalicular membrane carry the drug or its conjugate out of the cell into bile. Three efflux transporters carry most drugs and conjugates into bile: P-glycoprotein, multidrug resistance-associated protein 2, MRP2, and breast cancer resistance protein, BCRP. MRP2 carries anions such as glucuronides. Watch at 2:28

Bile carries the drug or its conjugate into the intestine. There, gut bacteria can hydrolyze a conjugate back into the parent drug, the drug before conjugation. The intestine absorbs the parent drug again, and the portal vein carries it from the intestine to the liver. This cycle of biliary excretion and intestinal reabsorption is enterohepatic recycling. Watch at 3:54

Recycling returns the drug to the plasma, the liquid part of the blood. Recycling can produce more than one peak and make the apparent half-life longer. Watch at 4:23

MPAG is the only form of the drug found in bile, so the cycle depends on bacteria in the colon, the large intestine. These bacteria convert MPAG back to MPA, and the colon absorbs MPA again. Plasma MPA peaks within 1 to 2 hours after a dose, and the reabsorbed MPA often gives a second peak at 6 to 12 hours. Watch at 4:45

Ciclosporin, another immunosuppressant used with mycophenolate mofetil, inhibits the transport of MPAG into bile by MRP2, so less MPA recycles. With ciclosporin, the MPA AUC is about 30 to 40 percent lower than with mycophenolate mofetil alone or with tacrolimus, an immunosuppressant without this effect. In kidney transplant patients who stopped ciclosporin, the MPA AUC was 20 to 30 percent higher than in patients who continued low-dose ciclosporin. Watch at 7:07

Half-life is the time for the plasma concentration to fall by half. After one half-life, the plasma concentration is 50 percent of its first value. After two half-lives, it is 25 percent, and after five half-lives, it is about 3 percent. Watch at 0:28

The rate of elimination is the amount of drug that the body removes per unit of time. For many drugs, the rate is proportional to the amount of drug in the body. The rate divided by the amount is the elimination rate constant, k. The body thus removes the same fraction of the drug in each hour. Watch at 0:41

Half-life equals the natural logarithm of 2, which is 0.693, divided by k. It therefore equals 0.693 times the volume of distribution divided by the clearance. Half-life thus depends on two quantities: it falls with the clearance and rises with the volume of distribution. Watch at 1:14

Clearance depends on the blood flow to the liver and the kidneys and on their enzymes. At the same volume, a higher clearance removes a larger fraction of the drug per hour, and the half-life is shorter. Watch at 2:45

The liver and the kidneys remove drug only from the blood that flows through them. With a large volume, most of the drug is in the tissues, so the same clearance removes a small fraction of the drug per hour. The half-life is then long. The volume depends on binding to plasma proteins and to tissues. Watch at 3:25

In young men, midazolam had a volume of distribution of 1.3 liters per kilogram and a half-life of 2.1 hours. Its clearance was 7.8 milliliters per minute per kilogram. Midazolam clearance was nearly 20 times the diazepam clearance, and the diazepam half-life was 32 hours. Watch at 4:04

Diazepam is lipophilic, or fat-soluble, and lipophilic drugs distribute well into adipose tissue, or body fat. At normal weight, its volume of distribution was 70 liters and its half-life was 32 hours. In obesity, the volume was 228 liters and the half-life was 82 hours. Clearance was similar: 1.56 liters per hour at normal weight and 1.92 in obesity. Watch at 4:44

The longer half-life delays steady state, the condition in which the dose rate equals the rate of elimination. For a given dose rate, clearance sets the average concentration at steady state, and half-life sets the time to reach it. Watch at 5:38

A compartment is a space in which the drug concentration is the same everywhere at a given time. Watch at 0:26

The volume of this compartment is the volume of distribution, Vd: the amount of drug in the body divided by the plasma concentration. The one-compartment model treats the whole body as one compartment, so it assumes that distribution is instant. Watch at 0:44

In first-order elimination, the rate of elimination, the amount removed per unit of time, is proportional to the amount of drug in the body. Watch at 1:14

The elimination rate constant, k, is the rate of elimination divided by the amount. Its unit is 1 per unit of time, such as 1 per hour. Watch at 1:26

Under these assumptions, the plasma concentration falls along an exponential curve. The equation is C equals C zero times e to the power of minus k times t. C is the plasma concentration at time t after the bolus, and C zero is its value at time zero. In each hour, the body thus removes the same fraction of the drug left in the body. Watch at 1:38

Half-life, the time for the plasma concentration to fall by half, equals 0.693, the natural logarithm of 2, divided by k. Watch at 2:16

In this model, k equals CL divided by Vd. Their ratio, 20.3 liters per hour divided by 58.3 liters, is also about 0.35 per hour. Watch at 2:37

The concentration falls by the same fraction in each equal interval of time. On a plot with a logarithmic concentration axis, equal distances show equal ratios of concentration. On such a plot, the curve therefore becomes a straight line. Watch at 3:34

On a logarithmic scale, the four medians lie close to one straight line, which agrees with the one-compartment model. Watch at 5:08

A two-compartment model describes the body as two spaces, and drug moves between them in both directions. Watch at 0:47

The central compartment is the plasma and the organs that the drug reaches fast, such as the liver. The dose enters it, blood samples come from it, and drug leaves the body from it. The peripheral compartment is the tissues that the drug reaches more slowly. Watch at 0:57

The distribution phase is the early, fast fall of the plasma concentration, while drug moves into the peripheral compartment. The elimination phase is the later, slow fall, in which the plasma and tissue concentrations fall in parallel. Watch at 1:25

The plasma concentration, C, at time t after the dose is the sum of two exponential terms. C equals A times e to the minus alpha t, plus B times e to the minus beta t. Alpha and beta are rate constants: the larger the constant, the faster its term falls. A and B are the values of the two terms at time zero. Watch at 1:46

On a logarithmic scale, each exponential term is a straight line, steep for alpha and shallow for beta. The alpha term soon becomes small, and the curve then lies on the beta line, whose slope gives minus beta. When alpha is much larger than beta, the curve shows two nearly straight phases. Watch at 2:36

In the distribution phase, the plasma concentration falls for two reasons: drug moves into the tissues, and the body also eliminates drug. At the same time, the concentration in the tissues rises. Watch at 3:09

The third is the terminal half-life. It is the half-life of the elimination phase, and it equals 0.693, the natural logarithm of 2, divided by beta. The terminal half-life is longer when a larger share of the drug is in the tissues, and shorter when clearance is higher. Watch at 3:47

The mean distribution half-life was 0.35 hours, about 21 minutes. The mean terminal half-life was 27.9 hours, about 80 times longer. Watch at 4:52

Distribution of digoxin into the tissues takes 6 to 8 hours. A sample before then gives an unreliable measure of the effect and of toxicity. Blood samples for digoxin are therefore taken at least 6 hours after a dose, or just before the next dose. Watch at 5:36

If a new dose comes before the body eliminates all of the previous dose, the drug that is left and the new dose add together. Accumulation is the amount of drug left in the body from earlier doses. Watch at 0:35

The peaks and troughs approach a level at which each dose replaces the drug eliminated in one interval. This condition is called steady state. Watch at 1:22

The accumulation ratio, R, is most often the AUC over one interval at steady state, divided by the AUC over the first interval. The area under the plasma concentration curve, AUC, measures the exposure to the drug over a period of time. Watch at 1:36

In the simplest model, the body is one well-mixed space that eliminates the same fraction of the drug in each hour. Each dose enters the blood at once, as with an injection into a vein. If f is the fraction of a dose left at the end of one interval, R equals 1 divided by 1 minus f. Watch at 2:04

When the interval equals the half-life, half of each dose is left when the next dose comes, so R is 2. With an interval of half the half-life, about 71 percent is left, and R is about 3.4. With an interval of two half-lives, 25 percent is left, and R is about 1.3. In this model, R is the same for every dose size. Watch at 2:26

The second is fluctuation, the change of the concentration within one interval. The peak-to-trough ratio, the peak divided by the trough, measures it. In the simplest model, at an interval equal to the half-life, the trough is half the peak. Watch at 3:05

In this model, the same dose per day gives the same average concentration at steady state for any interval. Smaller doses given more often therefore keep the same average with a smaller fluctuation. Watch at 3:34

In the simplest model, the half-life predicts that a little less than half of each dose is left at the next dose. The equation then gives an R of about 1.9. The measured R was 1.7, and the authors found it consistent with the half-life and the twice-daily doses. The concentration fell faster in the first hours after each peak, and such a fall can make the predicted R too high. Watch at 4:00

With 10 milligrams once daily, the average peak-to-trough ratio was about 8. With 5 milligrams twice daily, the same dose per day, it was about 3. Watch at 5:29

A maintenance dose is a dose given at regular intervals, for example once a day, to keep the concentration near the target. The dose rate is the dose divided by the interval between doses. Watch at 0:25

With repeated doses, each dose adds to the drug that remains from earlier doses, so the concentration rises. As the concentration rises, the rate of elimination rises with it, until it equals the dose rate. This condition is steady state. The concentration still rises and falls between doses, but its average stays the same. Watch at 1:12

A loading dose is a first dose, larger than the maintenance dose, that brings the plasma concentration to the target sooner. Watch at 1:31

At a constant dose rate, in each half-life the concentration rises by half of the difference that remains to its steady-state value. The concentration therefore reaches steady state after about five half-lives. When the rate of elimination is proportional to the concentration, this time is the same at every dose rate. Watch at 2:23

At steady state, the dose rate equals the rate of elimination, which is CL times the plasma concentration. For doses into a vein, the average concentration therefore equals the dose rate divided by CL. The dose rate and CL set this concentration, and the half-life sets the time to reach it. Watch at 2:44

At the target concentration, the amount of drug in the body is the target concentration times Vd. For a dose into a vein, the loading dose therefore equals the target concentration times Vd. Watch at 3:31

The loading dose changes only the early concentrations: the half-life and the average steady-state concentration stay the same. Watch at 3:49

For phenobarbital in newborns, a half-life of about one week would mean about five weeks to steady state without a loading dose. Watch at 4:24

In the trial for the approval, newborns born at full term received one loading dose. In term newborns, one loading dose of 20 milligrams per kilogram gave a mean of 23 milligrams per liter at 1 hour. Watch at 5:44

In linear pharmacokinetics, the parameters stay the same at every dose. Exposure then rises in proportion to the dose. This property is dose proportionality. Watch at 0:38

In nonlinear pharmacokinetics, at least one parameter changes with the dose. Exposure can then rise more, or less, than in proportion to the dose. Watch at 0:53

Enzymes, the proteins that make chemical reactions faster, and transport proteins, which carry drugs across cell membranes, have a limited number of binding sites. Saturation is the condition in which most of these sites are occupied, so the rate rises little when the concentration rises. It is a common cause of nonlinear pharmacokinetics, although few drugs show nonlinear behavior with clinical effects. Watch at 1:20

The Michaelis-Menten equation gives the rate as Vmax times C, divided by the sum of Km and C. C is the drug concentration, Vmax is the maximum rate, and Km is the concentration that gives half of Vmax. When C is much lower than Km, the rate is nearly proportional to C. When C is much higher, the rate stays close to Vmax. Watch at 1:41

For saturable elimination, clearance is therefore Vmax divided by the sum of Km and C. It is nearly constant at low concentrations and falls as the concentration approaches and exceeds Km. Watch at 2:13

When the concentration is near or above Km, clearance falls as the dose rate rises. The steady-state concentration then rises more than in proportion to the dose rate. When saturable metabolism is the only route of elimination, the concentration rises without a steady state at any dose rate above Vmax. Watch at 3:13

Gabapentin, a drug for seizures and nerve pain, crosses the gut wall by a saturable transport process. Its bioavailability fell from 60 to 33 percent as the daily dose rose from 900 to 3,600 milligrams. Its exposure therefore rises less than in proportion to the dose. Watch at 3:44

When most binding sites on the plasma proteins are occupied, the unbound fraction rises with the concentration. The total concentration then rises less than in proportion to the unbound concentration. In six healthy subjects, the unbound fraction of valproic acid, a drug for epilepsy, rose from 7 to 10 percent when the dose doubled. Watch at 4:22

With these values, the Michaelis-Menten equation gives a steady-state concentration of 10 milligrams per liter at 264 milligrams per day, and 20 at 323. A dose increase of about 22 percent thus doubles the concentration. Watch at 5:52

Noncompartmental analysis, or NCA, calculates each person's parameters, such as the AUC and the clearance, from that person's points. It uses no compartments, the spaces of uniform drug concentration in a model, and it makes few assumptions. Watch at 1:00

Population pharmacokinetics, or population PK, fits one model to the data of all patients and explains part of the differences between them. It fits a nonlinear mixed-effects model. Watch at 1:11

The first method is noncompartmental analysis. It needs rich sampling, many samples per person: for most drugs in bioavailability studies, the FDA recommends 12 to 18 for each dose. With few samples per person, the trapezoidal rule can give an AUC that is too low or too high. Rich sampling may be impractical in some patients, for example in children. Watch at 1:38

NCA calculates the AUC by the trapezoidal rule. A straight line joins each pair of consecutive points, and the shape under it is a trapezoid. The area of each trapezoid is the mean of its two concentrations times the time between them. The sum of these areas is the AUC to the last sample. Watch at 1:57

The terminal half-life is the time for the concentration to fall by half in the last phase. On a plot of the natural logarithm of the concentration against time, the points of this phase fall on a straight line. A linear regression line through at least three of these points has a slope of minus lambda z, the terminal rate constant. The terminal half-life is 0.693, the natural logarithm of 2, divided by lambda z. Watch at 2:51

Fixed effects have one value for all patients, such as the typical clearance. Random effects describe how far each patient differs from the typical value. Watch at 4:02

Covariates are measured patient factors, such as body weight or kidney function, that explain part of the between-subject variability. Watch at 4:32

Population PK works with sparse sampling, a few samples from each patient at different times, because it estimates the curve from all patients together. It can therefore use large patient trials and test many covariates at once. Watch at 4:43

The authors concluded that, of the covariates tested, only kidney function warrants a change of dose. Age, sex, heart failure, South Asian origin and creatinine clearance, a measure of kidney function, changed the apparent clearance. Watch at 7:01

Noncompartmental analysis, or NCA, calculates the parameters of a drug from the plasma concentrations of one person at set times. The area under the concentration-time curve, the AUC, measures the total exposure to the drug. Watch at 0:08

The trapezoidal rule adds the areas under straight lines between consecutive points. Watch at 1:19

After absorption ends, the concentration falls by about the same fraction in each hour, and the log-down line follows an exponential decay exactly. A straight line lies above the decay, so the linear rule overestimates the area, more so for a longer time between samples. Watch at 1:50

NCA software such as NonCompart fits a line to the last three points. The software then adds one earlier point at a time, back to the first point after tmax. It keeps the fit with the highest adjusted R squared, the R squared corrected for the number of points. The FDA recommends three or more points. Watch at 2:19

The terminal half-life, the time for the concentration to fall by half in this phase, is 0.693 divided by lambda z. Watch at 2:45

For these studies, the International Council for Harmonisation recommends an AUC to the last sample of at least 80 percent of the AUC to infinity. The extrapolated percentage is then at most 20 percent. Watch at 3:30

The dose divided by the AUC to infinity is the clearance after an injection into a vein, and CL over F after an oral dose. Watch at 3:53

Vss, the volume of distribution at steady state, is CL multiplied by MRT when the clearance is the same at every dose. After an oral dose, the MRT includes the mean absorption time, so this product differs from Vss. Vss is CL multiplied by MRT only after a single rapid injection into a vein, with the same clearance at every dose. Watch at 4:49

CL over F was 3.13 liters per hour, and Vz over F was 38.5 liters. Watch at 6:38

Liver microsomes are fragments of the endoplasmic reticulum, a membrane network inside liver cells. A laboratory separates them from broken-up liver by centrifugation, a fast spin that sorts particles by size and density. They contain the cytochrome P450 enzymes, or CYPs, which metabolize many drugs, and the assay adds NADPH, a cofactor that the CYPs need. Watch at 0:31

Caco-2 cells come from a human colon tumor. On a porous filter, they grow in about 21 days into a single layer like the lining of the small intestine. The filter separates a gut-side chamber from a blood-side chamber. Watch at 1:14

The assay mixes the drug with microsomes and NADPH, or with hepatocytes. It measures the drug left in samples over 60 to 90 minutes. The natural logarithm of the amount left usually lies on a straight line against time. For microsomes, CLint equals the steepness of this line, k, times the mixture volume per milligram of microsomal protein. Watch at 2:37

An efflux ratio of 2 or more suggests efflux by a transporter. If a known inhibitor of the transporter lowers the ratio by more than half, the transporter probably carries the drug. Watch at 3:50

In equilibrium dialysis, a membrane separates plasma with drug from a buffer, a salt solution without protein. Unbound drug crosses the membrane, and protein-bound drug stays in the plasma. After several hours, the unbound concentrations on the two sides are equal. fu is then the buffer concentration divided by the plasma concentration. Watch at 4:08

The IC50 is the concentration of the new drug that halves the probe drug metabolism in microsomes. Watch at 4:55

In human liver microsomes, its metabolism was slow, mainly by CYP3A4 and CYP3A5, two CYP enzymes. In human plasma in vitro, proteins bound about 87 percent, so fu was about 0.13. The efflux ratio was more than 10, in Caco-2 cells and in cells with added P-gp or BCRP, a second efflux transporter. Apixaban did not inhibit seven CYPs, and the IC50 values were above 20 micromolar, a unit of concentration. Watch at 5:23

In 18 healthy subjects, ketoconazole, a strong inhibitor of CYP3A4 and P-gp, doubled the apixaban AUC, the area under the plasma concentration-time curve. One later analysis concluded that CYP3A4 inhibition in the gut wall can explain this increase. In a second clinical study, apixaban did not change digoxin concentrations. Watch at 6:33

In vitro results show possible mechanisms, and clinical studies test which of them matter. Watch at 7:44

IVIVE, in vitro to in vivo extrapolation, predicts a value in the body from a measurement in vitro. Watch at 0:17

Intrinsic clearance is the rate of metabolism divided by the unbound drug concentration at the enzymes. It measures the removal by the enzymes alone, apart from any limit by blood flow or protein binding. Watch at 0:56

In the assay, the drug is incubated with microsomes or hepatocytes at a low concentration. The concentration then falls exponentially, and the rate constant of this fall is the rate of loss divided by the amount of drug that remains. The rate constant, times the incubation volume, divided by the amount of microsomal protein, is the in vitro intrinsic clearance of total drug. For hepatocytes, the number of cells replaces the amount of protein. Watch at 1:30

The in vitro value is multiplied by the protein or cells per gram of liver and by the liver weight per kilogram of body weight. The result is the whole-liver intrinsic clearance. Watch at 2:07

The assay value uses the total drug concentration, but the microsomes or hepatocytes bind part of the drug. Division by the unbound fraction in the incubation gives the intrinsic clearance of unbound drug. The unbound fraction in blood is the unbound fraction in plasma divided by the blood-to-plasma concentration ratio. Watch at 3:11

The well-stirred model then combines blood flow, binding and intrinsic clearance into hepatic clearance. Watch at 3:57

The fold error is the predicted clearance divided by the observed clearance. A prediction is within twofold when the fold error lies between 0.5 and 2, and the fraction within twofold measures accuracy. Watch at 4:21

With hepatocytes, 31 percent of the predictions were within twofold, 62 percent were below this range and 7 percent above it. With microsomes, the values were 42, 48 and 10 percent. On average, the predictions were 2.7 times too low with hepatocytes and 2.0 times too low with microsomes. Watch at 5:16

Promethazine shows that a prediction within twofold can still put a drug in the wrong clearance range. This prediction was within twofold and 1.7 times too low. The observed value is above 70 percent of hepatic blood flow, but the predicted value falls in the intermediate range, 30 to 70 percent. Watch at 6:30

Allometry is the study of how a property of an animal changes with its body size. Allometric scaling uses this relation to predict human values from animal values. Watch at 0:34

The relation is a power equation: Y equals a times W to the power b. Y is the property, for example clearance, and W is the body weight. The coefficient, a, is the value of Y at 1 kilogram. The exponent, b, sets how fast Y changes with body weight. Watch at 0:49

For clearance, the exponent is often close to 0.75. In published data for 115 compounds, clearance correlated with body weight for 91 of them. For these 91, the mean exponent was 0.74, with a range of 0.29 to 1.2. An exponent below 1 means that a larger species has a lower clearance per kilogram. Watch at 1:19

Clearance is measured in each species after an intravenous dose, an injection into a vein. Clearance is plotted against body weight on logarithmic scales, and the fitted line is extended to a human body weight, such as 70 kilograms. Watch at 2:18

The rule of exponents chooses the correction from the exponent of the simple fit: it keeps simple allometry for 0.55 to 0.70. From 0.71 to 1.0, it multiplies each clearance by the maximum life span of the species, and above 1.0 by the brain weight. The human value of the fit is then divided by the human maximum life span or brain weight. Watch at 2:51

Allometry assumes that body size alone sets the differences between species. Species also differ in the enzymes that metabolize drugs, and in transporters, membrane proteins that carry drugs into or out of cells. Binding to plasma proteins also differs, for example to albumin and alpha-1-acid glycoprotein. Watch at 3:50

An analysis of 81 small-molecule drugs tested simple allometry from three or more species. For drugs that the kidneys excrete unchanged, the prediction was on average 1.8 times too high or too low. For drugs that the liver removes only to a small extent, it was on average 4.5 times too high or too low. Watch at 4:19

In mice, rats and dogs, its clearance exponent was 0.66, in the range where the rule of exponents keeps simple allometry. The authors made the prediction after human values were published. It was 4.68 liters per hour, inside the reported range of 4 to 10.5 liters per hour, near its lower end. Watch at 5:13

Simple allometry predicted a human clearance about 5,000 times too high, and the rule of exponents about 1,750 times too high. The authors call this an extreme example of overprediction. Watch at 6:14

A physiologically based pharmacokinetic model, or PBPK model, makes each compartment a real organ or tissue, with the volume and blood flow of that organ. Watch at 0:34

System parameters describe the body: organ volumes, blood flows, tissue composition and enzyme amounts. They change with the species and the population, and they do not depend on the drug that the model describes. Watch at 0:53

The tissue-to-plasma partition coefficient, Kp, is the total drug concentration in a tissue divided by the total concentration in plasma, when both are constant. Binding to tissue proteins, dissolution in tissue lipids and trapping in lysosomes, small acidic compartments in cells, raise Kp. Kp depends on the drug and on the tissue, so it has one value for each tissue. Watch at 1:22

For each organ, one equation sets the rate of change of the drug amount equal to the rate in minus the rate out. In the usual model, each tissue comes to equilibrium with the venous blood that leaves it, at the ratio that Kp sets. The liver and kidney equations usually also subtract the rate of metabolism or excretion, and the gut wall equation can do the same. Watch at 2:15

Kp values come from equations that combine the drug properties with the lipid and water content of each tissue. In two studies of 67 drugs, these equations predicted 84 to 89 percent of measured values in most rat tissues within a factor of 3. Watch at 3:11

Before the first dose in humans, the model predicts the plasma concentration-time curve from laboratory data, after a check in animals. Exposure is often measured as the AUC, the area under this curve. Watch at 3:35

From 2020 to 2024, 65 of the 245 new drugs that the FDA approved, or 26.5 percent, included a PBPK model in the submission. Watch at 4:29

The model then predicted that rifampicin, a strong CYP3A inducer, would lower the AUC 10-fold. In the study of 17 healthy adults, rifampicin lowered the AUC 10-fold. Watch at 5:36

From these predictions, the authors advised 140 milligrams with a moderate inhibitor, one quarter of 560 milligrams, the highest dose at that time. The drug label at that time adopted this advice. Watch at 6:00

Machine learning is a computer method that learns from examples the relation between a chemical structure and a measured ADME value. A laboratory measures one assay, for example solubility or microsomal stability, for thousands of compounds. The model is fitted so that its predicted values match these measured values. Watch at 0:18

Molecular descriptors are numbers calculated from the structure, such as the molecular weight and the polar surface area. A fingerprint is a row of bits, and the bits record which small substructures the molecule contains. Watch at 0:33

The applicability domain is the part of chemical space where the model predicts reliably. The error rises as the similarity of a compound to its nearest training compounds falls. Watch at 1:21

A random split takes the test compounds at random. A time split trains on the compounds measured before a date and tests on the later ones. The core of a molecule is its ring systems and the chains that link them. A scaffold split holds back whole cores. Watch at 2:08

A random split gives an optimistic accuracy, and a time split gives one closer to the accuracy on compounds not yet measured. Watch at 2:30

Models fail in three situations. The first is new chemical space, where the test compounds are not similar to any training compound. The second is an assay with few measured compounds or a large measurement error. The third is an endpoint that depends on complex biology. Watch at 3:34

In repeated random splits, one public model predicted 52.5 percent of 650 low clearance compounds within twofold, and 10.4 percent of 144 high clearance compounds. High clearance compounds were also rare in the training data. Watch at 4:23

In a time split, a model trained on the 6,448 compounds measured up to 2014 predicted the 2,576 compounds measured in 2015. For nine of the ten largest drug discovery projects, the balanced accuracy rose above 0.75 in at least one of three years. Watch at 6:02

For one of these nine projects the accuracy fell in the third year, and the authors assume that new cores caused the fall. Watch at 6:30

A projected human dose is the dose that a compound is predicted to need in patients, before anyone takes it. Watch at 0:06

The AUC of one dose equals F times the dose, divided by CL. Solved for the dose, this gives the dose equation: the dose equals CL times the target AUC, divided by F. Watch at 0:55

A machine learning model learns the relation between structure and a property from the measured values of earlier compounds. Such models predict, for example, solubility, binding to plasma proteins and the rate of metabolism in liver microsomes, small fragments of liver cells. The only input is the structure, so chemists can use these predictions before they make a compound. Watch at 1:06

The second step is to fill a full ADME profile for each compound. The profile is the set of ADME values that the projection needs. Each value is a measurement where one exists and a prediction elsewhere. Watch at 1:49

In the dose equation, with F and the target AUC fixed, the dose changes in proportion to CL. Therefore, a twofold error in CL gives a twofold error in the dose. Watch at 2:44

Chemists prefer compounds with a lower projected dose. In one study, oral drugs with higher daily doses had more reports of liver failure and death from drug-induced liver injury. The ranked shortlist also records risks that the dose does not show. Watch at 3:09

The first inhibitors in the project caused severe toxicity in rats and dogs, and they also inhibited two related enzymes, DPP-8 and DPP-9. A selective inhibitor of DPP-8 and DPP-9 caused the same toxicity, but a selective DPP-4 inhibitor did not. Watch at 4:17

In healthy volunteers, sitagliptin had a clearance of about 25 liters per hour, about 70 percent by the kidneys, and an F of 87 percent. Watch at 5:01

For this dose, 0.87 times 100 milligrams, divided by 25 liters per hour, gives a daily AUC of about 3.5 milligram hours per liter. Watch at 5:50