Distribution Phase Drug Interactions

Drug Interactions

Drug interactions occur when one drug’s pharmacokinetic, pharmacodynamic or toxicological characteristics are influenced by a second substance. Usually, the second substance is another drug—termed a “drug-drug interaction”—but it can also be a component of the diet (such as proteins, fats, minerals, vitamins). In describing these interactions we use the terms “victim” and “perpetrator”. The victim is the drug whose PK, PD or tox characteristics are changed whereas the perpetrator is the drug (or sometimes a nutrient) who causes the change. In many instances, both drugs are affected and the victim/perpetrator model breaks down.

Distribution Phase Drug Interactions Overview

Distribution phase drug interactions occur when one drug alters the distribution characteristics of another drug. These interactions can involve displacement from plasma protein binding sites, competition for tissue binding sites, or interactions with the membrane transports at the interface of blood and the tissue. Protein binding displacement interactions were historically thought of as highly clinically significant but modern experience suggests they are mostly transient interactions and not clinically significant for drugs with low extraction ratios. However, interactions involving high extraction drugs or tissue distribution can have important clinical consequences.

Protein Binding Displacement Interactions

Protein binding displacement occurs when two drugs compete for the same binding sites on plasma proteins, typically albumin or alpha-1-acid glycoprotein. When displaced from plasma protein, the immediate consequence is an increase in free (unbound) drug concentration. This might seem alarming—after all, only free drug is pharmacologically active. However, the clinical significance depends on several factors.

The Classic Displacement Interaction

Consider a classic scenario: warfarin (victim) displacement by phenylbutazone (perpetrator), both highly bound to Sudlow site I of albumin. Warfarin is normally ~99% bound, meaning f_u is 0.01. When phenylbutazone is added and displaces some warfarin from albumin, binding might decrease to 97% causing f_u to triple to 0.03!

This tripling of free warfarin concentration could theoretically triple anticoagulant effect and cause the patient to bleed uncontrollably. A compelling story that worried a lot of pharmacologists in the past. However, the full picture is more complex.

When warfarin unbinds from albumin, the amount that is available for metabolism and excretion increases. There may be an immediate slight disturbance in plasma levels but clearance increases and a new steady-state is established. Total warfarin concentration decreases (because less is sequestered on the albumin molecule) but the free concentration returns close to the original level. The net result of these displacement interactions is often clinically insignificant.

Why Most Displacement Interactions Are Not Clinically Significant

For drugs with low hepatic extraction ratios (E < 0.3), where f_u × CL_int << Q_H, clearance depends on the unbound fraction (f_u) and intrinsic clearance (CL_int):

$$ CL \approx f_u \times CL_{int} $$

The reader is referred to Metabolism for further information on this relationship.

This relationship tells us that when displacement causes an increase in f_u, CL increases proportionally.

This is why displacement interactions are generally not clinically significant at steady-state for low extraction drugs.

When Displacement Interactions Matter

Despite the compensatory mechanisms described here, certain displacement interactions can be clinically important. The following red-flags may help to identify high-risk situations:

• Narrow therapeutic window
• High protein binding
• Low extraction ratio
• Small volume of distribution
• Rapid onset of action or toxicity

Drugs where small increases in free concentration cause serious toxicity (ie narrow therapeutic window) are more susceptible to clinically significant displacement because there will always be an immediate slight pertubation in plasma concentration before the new steady-state is established. Warfarin, phenytoin, and methotrexate fall into this category. Even transient increases in free concentration can cause bleeding (warfarin), seizures (phenytoin), or bone marrow suppression (methotrexate).

Drugs that are >90% bound are more vulnerable because displacement causes larger proportional changes in the unbound fraction. If a drug is 50% bound and displacement reduces binding to 40%, free fraction increases from 50% to 60% (20% increase). But if a drug is 95% bound and displacement reduces binding to 90%, free fraction increases from 5% to 10%—a 100% increase.

Trivially, drugs confined mainly to plasma (small V_d) are more affected by protein binding changes because a larger proportion of the total drug in the body is in plasma bound to proteins.

If the displaced drug acts quickly or toxicity develops rapidly, the transient increase in free concentration becomes more clinically relevant. For drugs with slow onset, compensatory mechanisms engage and may restore plasma concentration to normal before toxicity develops.

Clinically Important Displacement Examples

Warfarin is involved in numerous documented displacement interactions. Phenylbutazone, aspirin (at high doses), and sulfonamides can all displace warfarin from albumin, transiently increasing INR and bleeding risk.

Phenytoin, an anticonvulsant, is ~90% protein bound and has a narrow therapeutic window. Valproic acid can displace phenytoin from albumin, increasing free phenytoin concentrations. Since phenytoin exhibits dose-dependent (zero-order) kinetics, small increases in concentration can cause large increases in half-life, potentially leading to phenytoin toxicity (ataxia, nystagmus, confusion). Clinicians monitoring total phenytoin concentrations (the usual test) might miss this interaction; free phenytoin measurement should be considered in such cases.

Methotrexate, used in cancer chemotherapy and autoimmune diseases, is ~50% protein bound. NSAIDs, particularly salicylates, can displace methotrexate from albumin. The combination can cause severe methotrexate toxicity including bone marrow suppression, mucositis, and renal failure. Importantly, NSAIDs also reduce renal clearance of methotrexate (discussed in Excretion Phase Drug Interactions) making it unclear whether interactions are due to protein displacement, renal excretion inhibiton or both. High-dose methotrexate therapy requires careful monitoring and NSAID avoidance.

Displacement of bilirubin, a breakdown product of haem metabolism, in neonates deserves special mention. Neonates (premature neonates in particular) have immature hepatic conjugation systems and can develop hyperbilirubinemia. Bilirubin is highly albumin-bound (~99%). Drugs like sulfonamides and ceftriaxone can displace bilirubin from albumin, increasing free bilirubin concentrations. Unbound bilirubin is then free to cross the blood-brain barrier (which is also less tight in premature neonates) and deposit in the basal ganglia, causing kernicterus—a devastating neurological condition. For this reason, sulfonamides and ceftriaxone are generally avoided in neonates with jaundice.

Tissue Distribution Interactions

Beyond plasma protein binding, drugs can interact at the level of tissue distribution through competition for binding to tissue or alterations in membrane transport.

P-glycoprotein Interactions

P-glycoprotein (P-gp), also known as multidrug resistance protein 1 (MDR1), is an ATP-dependent efflux transporter expressed in many tissues including the intestinal epithelium, blood-brain barrier, placenta, liver canaliculi, and renal tubular cells. At the blood-brain barrier, P-gp actively pumps drugs from brain back into blood, limiting CNS exposure.

Many commonly used drugs are P-gp substrates, inhibitors, or inducers. When a P-gp inhibitor is co-administered with a P-gp substrate, efflux from the brain decreases, increasing brain drug concentrations. This can have significant clinical consequences.

Digoxin is a classic P-gp substrate. Co-administration of P-gp inhibitors like verapamil, quinidine, or clarithromycin increases digoxin tissue distribution, particularly to cardiac muscle (the site of action). This increases both therapeutic effects and toxicity risk. Digoxin doses should be reduced by 30-50% when these inhibitors are added. Monitoring digoxin plasma concentrations is essential during combination therapy.

Loperamide, an opioid used for diarrhoea, is normally excluded from the brain by P-gp at the blood-brain barrier. This is why loperamide doesn't cause CNS opioid effects at therapeutic doses. However, when high doses of loperamide are combined with P-gp inhibitors (a combination abused by some individuals seeking opioid effects), CNS penetration increases dramatically, causing respiratory depression and potentially fatal cardiotoxicity. This has become a growing public health concern.

Conversely, P-gp inducers like rifampicin and St John's wort increase P-gp expression and activity, reducing tissue distribution of P-gp substrates. This can lead to therapeutic failure. For example, rifampicin reduces digoxin concentrations by increasing P-gp-mediated intestinal and renal elimination, potentially necessitating digoxin dose increases.

Other Transporter Interactions

While P-glycoprotein is the most extensively studied, other transporters also mediate clinically relevant distribution interactions. Organic anion transporting polypeptides (OATPs) facilitate hepatic uptake of many drugs including statins. OATP inhibitors like ciclosporin reduce hepatic uptake of simvastatin and atorvastatin, paradoxically increasing systemic exposure (because less drug enters the liver where it would be metabolised). This increases statin toxicity risk, particularly myopathy.

Organic cation transporters (OCTs) and organic anion transporters (OATs) in the kidneys mediate tubular secretion and will be discussed further in Excretion Phase Drug Interactions, but they can also affect tissue distribution.

Altered Tissue Binding

Some drugs bind extensively to specific tissues, creating large volumes of distribution and drug reservoirs. Competition for tissue binding sites can alter distribution patterns, though these interactions are less common than protein binding displacement.

Quinacrine, an antimalarial (now rarely used), displaces chloroquine from tissue binding sites. Since chloroquine has an enormous volume of distribution (~15,000 L) due to extensive lysosomal accumulation, displacement significantly increases plasma chloroquine concentrations, potentially causing toxicity.

Haemodynamic Interactions

Changes in blood flow can affect distribution, particularly for highly perfused tissues. Drugs that alter cardiac output or regional blood flow can influence the distribution of other drugs, though these interactions are generally less predictable and less clinically significant than protein binding or transporter interactions.

Beta-blockers reduce cardiac output, potentially decreasing drug delivery to tissues. Vasodilators increase regional blood flow. In practice, these effects are usually modest and rarely require dose adjustments, except in extreme haemodynamic situations like cardiogenic shock.

Quiz