The MDR1 Defect in Dogs: ABCB1, P-Glycoprotein and Drug Sensitivity

1. What the Defect Actually Is

1.1 The Gene and the Protein

The gene is ABCB1, formerly called MDR1. It encodes P-glycoprotein, a large transmembrane pump that moves a structurally diverse set of compounds out of cells. At the blood-brain barrier it transports drugs from the brain back into the blood, which is why most mammals tolerate substances that would otherwise be neurotoxic (Mealey, Bentjen, Gay & Cantor, 2001).

1.2 The Mutation

A 4-base-pair deletion produces a frameshift and premature stop codons, terminating P-glycoprotein synthesis early. The result is a severely truncated, nonfunctional protein. The variant is written ABCB1-1Δ in the American literature and nt230(del4) in the European.

Both names describe the same deletion. The inconsistency matters only when reading test reports, where one laboratory may use one convention and another a different one.

1.3 Why It Reaches Beyond the Brain

P-glycoprotein is expressed at several pharmacological barriers, but the evidence is not equally strong for every organ in the dog. Its role at the blood-brain barrier is well established. Canine studies also show a major role in biliary excretion: homozygous deficient dogs can have markedly reduced biliary transport of P-glycoprotein substrates, and heterozygous dogs may show an intermediate effect (Mealey, Owens & Freeman, 2023).

The intestinal story is less tidy. P-glycoprotein is present in the gut, but canine in-vivo studies have not shown a uniform increase in oral bioavailability across substrates; results vary by drug. P-glycoprotein is also expressed in renal tubules, yet its quantitative contribution to renal drug excretion in dogs remains poorly characterized. The safe summary is therefore not "absorption and renal clearance are always increased," but that the mutation can alter drug disposition at several barriers in a substrate-specific way (Mealey et al., 2023).

1.4 Where the Finding Came From

Ivermectin toxicosis in Collies was described in the 1980s, long before anyone could explain it. For years it was a breed-associated clinical phenomenon: some Collies reacted catastrophically to doses that other dogs tolerated, and the mechanism was unknown.

The 2001 work identified the four-base-pair deletion and linked homozygosity for it to the ivermectin-sensitive phenotype (Mealey et al., 2001). It remains one of the clearest examples in veterinary pharmacogenetics of a clinical observation being traced to a defined transport defect.

1.5 Why This Article Exists in a Behavior Library

The genotype can matter when behavioral medication is prescribed, but this requires drug-specific judgment rather than a blanket "behavior drugs are MDR1 drugs" rule. In a canine cell model, fluoxetine and clomipramine inhibited P-glycoprotein in vitro, whereas selegiline did not; that finding demonstrates interaction potential, not an automatic clinical dose adjustment for every dog receiving those medications (Schrickx & Fink-Gremmels, 2014).

The broader behavioral relevance is differential diagnosis. A dog that develops neurological or behavioral change after drug exposure may be experiencing a pharmacological adverse effect rather than a training problem (how behavioral medication is chosen).

1.6 What P-Glycoprotein Normally Does

P-glycoprotein is an ATP-dependent efflux transporter. At the blood-brain barrier it can move susceptible compounds back toward the circulation before they accumulate in nervous tissue. In the liver it contributes to biliary excretion of some substrates. At other sites, including the intestine, the practical effect depends strongly on the specific compound and cannot be assumed from transporter expression alone (Mealey et al., 2023).

This is why "P-glycoprotein substrate" is not synonymous with "forbidden drug." Clinical risk depends on how much that drug relies on P-glycoprotein, its therapeutic margin, the dose and route, concurrent medications and the dog's genotype.

1.7 P-Glycoprotein Also Interacts With the HPA Axis

P-glycoprotein also restricts cortisol transport across the blood-brain barrier. In a 2022 study, urine from 23 homozygous mutant dogs and 16 wild-type dogs showed lower concentrations of several cortisol-related metabolites in the mutant group, supporting altered cortisol metabolism and HPA-axis regulation (Gramer et al., 2022).

A 2026 mini-review summarizes evidence that chronic P-glycoprotein deficiency can suppress HPA-axis activity and discusses possible clinical consequences during severe illness, including susceptibility to relative adrenal insufficiency (Mealey, 2026). This does not establish a temperament phenotype or mean that an otherwise well MDR1-affected dog is chronically ill. It does mean that "healthy until a substrate is given" is now too narrow a description of the biology.

2. Genotype and What It Means

2.1 Three Results

A test reports one of three genotypes: homozygous normal (two intact copies), heterozygous (one intact, one deleted), and homozygous mutant (two deleted copies). The variant is inherited autosomally, but clinical drug sensitivity cannot be reduced to a single dominant-versus-recessive rule. The size of the genotype effect varies by substrate: homozygous mutants generally have the largest loss of P-glycoprotein function, while heterozygous dogs can be clinically relevant for some drugs and not for others.

2.2 What the Original Study Found

In the study that identified the mutation, dogs homozygous for the deletion displayed the ivermectin-sensitive phenotype, while homozygous normal and heterozygous dogs did not show increased sensitivity to ivermectin (Mealey et al., 2001).

That is worth stating precisely, because the common summary — "carriers are intermediately sensitive" — is a generalization drawn from later clinical experience with other substrates rather than from this finding about ivermectin.

2.3 Why Carriers Still Matter

For drugs with a narrow margin, or where P-glycoprotein contributes substantially to elimination rather than only to brain exclusion, heterozygous dogs can show clinically relevant effects. Current veterinary guidance therefore treats the carrier state as relevant to dosing decisions for a subset of drugs rather than as harmless across the board (Mealey et al., 2023).

The practical form of that: a carrier is not a normal dog for every substrate, and is not an affected dog either. Which of the two it resembles depends on the drug.

A published case makes the point concretely. In a dog treated with cyclosporine for immune-mediated hemolytic anemia, pharmacodynamic monitoring revealed an unexpectedly high degree of T-cell suppression at a relatively low dose, with blood cyclosporine concentrations inside the expected target range. The dog turned out to be heterozygous for the deletion, and the authors suspected the genotype had contributed to the excessive immunosuppression, which was followed by a suspected secondary infection (Mackin et al., 2020).

Two things follow from that report. Blood concentration within target does not guarantee an expected effect in a carrier, and the problem surfaced only because someone was measuring what the drug did rather than how much of it was present.

2.4 Why the Phenotype Depends on the Drug

P-glycoprotein can influence brain penetration and, for some compounds, hepatobiliary elimination. Intestinal effects are substrate-dependent, and the quantitative importance of renal P-glycoprotein in dogs is still uncertain (Mealey et al., 2023).

That is why one genotype produces different degrees of risk across drugs. A compound with a narrow therapeutic margin and strong dependence on P-glycoprotein at the blood-brain barrier can become dangerous in a homozygous mutant dog, while another substrate may remain usable at an evidence-based dose. The drug-specific evidence matters more than the label "substrate" by itself.

2.5 One Test, Many Decisions

The genotype is a single stable fact that feeds into an open-ended series of future decisions: the next antiparasitic, a sedation for dentistry, an antiemetic, a chemotherapy protocol years later. Testing therefore buys information that keeps paying out rather than answering one question.

That framing matters for owners weighing the cost. The comparison is not the test against one treatment, but the test against every prescribing decision in the dog's remaining life.

2.6 Two Naming Conventions, One Variant

ABCB1-1Δ and nt230(del4) appear in different literatures and describe the same four-base-pair deletion. Laboratory reports may use either, and some use the older gene name MDR1 alongside the current ABCB1.

For an owner comparing a test result with an article, that is a needless source of confusion and worth knowing: four names, one thing. A report mentioning any of them is reporting on this variant.

3. Which Dogs Carry It

3.1 Not Only Collies

The deletion is strongly associated with Collie-lineage and related herding populations, but it is not confined to the Rough Collie. Reports and surveys have identified the variant in Australian Shepherds, Shetland Sheepdogs, Old English Sheepdogs/Bobtails, White Swiss Shepherds, Longhaired Whippets, Silken Windhounds, English Shepherds, McNabs and several other populations, as well as mixed-breed dogs (Mealey & Meurs, 2008; Mealey et al., 2023). European survey data describe the same picture under the nt230(del4) designation for breeds in Germany (Gramer et al., 2011).

That list should not be read as a universal prevalence table. Frequencies vary sharply between breeds, countries and sampled populations, and some published samples were enriched precisely because dogs were submitted for genetic testing.

3.2 Frequencies Are Population-Dependent

The deletion can be common in some Collie populations and much less common in other breeds. Border Collies are a useful caution: published European data have found very low frequencies in some cohorts, so their risk should not be presented as equivalent to Collies or Australian Shepherds.

A 2022 Belgian clinical-setting study illustrates the sampling problem particularly well. Among 286 dogs sampled from routine veterinary practice, the deletion allele frequency was only 0.2%, whereas it was 21.6% among 599 samples specifically submitted for genetic testing. The variant in the clinical sample was found only in an at-risk breed (Beckers et al., 2022). Referral and testing datasets therefore cannot be treated as general-population prevalence.

3.2a Why Appearance Is a Poor Screen

The old veterinary mnemonic — white feet, don't treat — captured the association with herding breeds in a memorable and easily misapplied form. Coat pattern is not the variant, herding ancestry is not always visible, and a dog with no white on it can be affected (why breed is a weak proxy).

The mnemonic did useful work in an era without a test. With a cheek swab available, an appearance-based rule is a reason to test rather than a substitute for testing.

3.3 Mixed Breeds and Unknown Ancestry

Mixed-breed dogs can carry the deletion, and ancestry is not always obvious from appearance. That supports genotyping when there is known or plausible risk ancestry, or when a planned medication would materially change depending on genotype. It does not mean every dog of unknown background requires an ABCB1 test before every routine treatment.

3.4 Why Appearance Is a Poor Screen

The old veterinary mnemonic — "white feet, don't treat" — was memorable but imprecise. Coat color is not genotype, and visible breed resemblance is an imperfect ancestry test. Modern genotyping is preferable whenever the result will change a prescribing decision.

3.5 Sampling Shapes Prevalence Figures

Published frequencies depend heavily on who was sampled. Breed-club testing, referral laboratories, general-practice populations and country-specific surveys answer different questions. Precise percentages should therefore always be attached to the population in which they were measured rather than generalized to a breed worldwide (Mealey & Meurs, 2008; Beckers et al., 2022).

3.6 Exposure Risk Depends on Environment, Not Breed Stereotypes

A practical high-dose exposure risk exists wherever concentrated horse or livestock antiparasitic products are stored or used. A susceptible dog may ingest paste, drench, spilled product or contaminated material. The important variable is access to the product, not whether the dog happens to work livestock.

For households, farms and equestrian settings, secure storage and rapid identification of the product after accidental ingestion are more useful safeguards than relying on breed folklore.

4. The Drugs

4.1 The Original Case: Ivermectin

Ivermectin sensitivity in Collies was the phenotype that led to discovery of the deletion. The important lesson is not "all macrocyclic lactones are forbidden" but that exposure and dose matter. The dose used in a licensed preventive product can differ by orders of magnitude from doses used for other indications or from accidental ingestion of concentrated large-animal formulations (Mealey et al., 2001; Mealey et al., 2023).

Product-specific safety data therefore matter. Prospective target-animal studies have been performed in homozygous mutant Collies. For example, a 2025 randomized, blinded study administered a moxidectin-containing monthly combination to 32 homozygous MDR1-mutant Collies in four groups — placebo, one, two and five times the maximum recommended label dose — on three occasions 28 days apart, scoring the standard avermectin-sensitivity signs through 72 hours after each treatment (Riggs, Wang & Wiseman, 2025). That finding applies to that tested product and regimen, not automatically to every formulation containing the same drug.

4.2 Loperamide

Loperamide remains one of the clearest owner-facing examples because it is an over-the-counter antidiarrheal that can produce central opioid toxicity when P-glycoprotein protection is deficient. The classic Collie case is well documented, and later work has expanded the evidence beyond a single anecdote (Sartor et al., 2004; Mealey et al., 2023).

It is therefore an important household warning, but not a scientifically established ranking of "the highest everyday risk." Concentrated livestock antiparasitics, prescription substrates and drug-drug interactions can also be clinically important.

4.3 Substrate Does Not Mean Contraindicated

P-glycoprotein transports chemically diverse compounds. Some are unsafe in deficient dogs at ordinary therapeutic doses, some require dose adjustment or monitoring, and some products have direct safety data supporting labeled use. Therapeutic margin, dose, route and the organ protected by P-glycoprotein all matter (Mealey et al., 2023).

This article therefore does not reproduce a static "safe/unsafe" dosing table. The durable rule is to pair the genotype with current, drug-specific veterinary evidence at the time of prescribing.

4.4 Why Lists Go Out of Date

New substrates, interaction data and product-specific safety studies continue to appear. A list also tends to erase the difference between an active ingredient, its dose and the formulation in which it is given.

The safer procedure is stable even when the pharmacology changes: record the genotype, disclose it before prescribing, and check the current evidence for the exact drug and product.

4.5 Dose, Product and Route All Matter

The same active substance can be safe in one labeled regimen and hazardous at a much higher exposure. A dog having tolerated a preventive dose in the past therefore does not establish that a different product, indication or dose is safe.

Conversely, the presence of an MDR1/ABCB1 warning should not be interpreted as a reason to withhold effective parasite prevention when an appropriately studied option is available.

4.6 Accidental Exposure

A recurring emergency involves no veterinary prescription at all: access to horse or livestock wormers. These formulations can contain doses far above those intended for dogs. For a homozygous mutant dog, concentrated macrocyclic-lactone exposure can be particularly dangerous.

The practical prevention is environmental: secure storage, prevent access during dosing, and retain the package or product name so an emergency veterinarian knows what was ingested.

4.7 Psychopharmacology: Interaction Is Not the Same as Contraindication

Behavioral medicine illustrates why mechanistic labels need care. In a canine cell model, fluoxetine inhibited P-glycoprotein by 41% and clomipramine by 59% at the tested concentration, while selegiline did not inhibit it (Schrickx & Fink-Gremmels, 2014). These are in-vitro inhibition data. They do not by themselves prove that either drug is clinically unsafe in an ABCB1 carrier or mutant dog, nor do they establish a specific dose adjustment.

The practical conclusion is narrower: polypharmacy and transporter interactions deserve veterinary review, and "substrate," "inhibitor" and "clinically significant interaction" should not be used as synonyms.

4.8 Drug Interactions Can Reduce P-Glycoprotein Function

Some drugs can inhibit P-glycoprotein and thereby change the disposition of another substrate even in a genetically normal dog. This acquired reduction in transporter function is clinically relevant, but it should not be equated with complete homozygous deficiency (Mealey et al., 2023).

That distinction matters when interpreting an adverse reaction in a genetically normal dog: the genotype does not exclude a transporter-mediated drug-drug interaction.

5. Recognizing a Reaction

5.1 What It Looks Like

Reported signs include lethargy and depression, ataxia, tremors, hypersalivation, dilated pupils, vomiting, disorientation, seizures and coma. Onset follows drug administration rather than occurring spontaneously, which is the feature that makes the history decisive.

5.2 Why the Exposure History Changes the Differential Diagnosis

Nothing about the clinical picture is unique to P-glycoprotein deficiency. Neurological signs have many differential diagnoses. A recent exposure to a known or suspected P-glycoprotein substrate can make drug toxicity much more plausible, but the exposure history is evidence for a working diagnosis rather than the diagnosis itself.

An owner who can identify the product, strength, amount and timing gives the veterinarian information that can change triage and treatment immediately. That is a practical argument for keeping a record of prescription, over-the-counter and non-prescription products.

5.3 What to Do

Treat it as an emergency, contact a veterinarian immediately, and bring the packaging of whatever was given. There is no antidote to a P-glycoprotein deficiency; management is supportive and depends on the drug involved and the severity.

5.4 Why Owners Underreport

Owners may not spontaneously classify over-the-counter products, supplements or something borrowed from another household as "medication." In a suspected adverse reaction, asking about these categories explicitly can therefore recover clinically important exposure information.

Asking specifically — what was given, in what quantity, from which package, and when — retrieves information that an open question does not, which is a familiar problem in any history-taking (why the question determines the answer).

5.5 Severity Is Not Predictable From Genotype Alone

Two affected dogs given the same drug at the same dose do not necessarily react identically. Body condition, concurrent medication, liver and kidney function and the route of administration all contribute, and the genotype sets a susceptibility rather than a fixed outcome.

For clinical purposes that means an affected dog that tolerated something once has not demonstrated safety, and it means a reaction can be more severe than the last one under different circumstances.

5.6 Recovery

Dogs that receive supportive care can recover from substrate toxicity, and the course depends on the drug, the dose and how quickly treatment began. That is the argument for treating any suspected reaction as urgent rather than waiting to see whether it resolves.

There is no treatment that restores the missing transporter during an acute intoxication. Management depends on the compound and exposure and may include decontamination when appropriate, supportive care and treatment of complications under veterinary supervision.

5.7 Documenting What Happened

After a suspected reaction, what is worth recording is mundane: the product name and strength, the amount given, the time, the time signs began, and what the dog did. Photographs of the packaging take seconds and outlast anyone's memory of a stressful evening.

That record serves the treating veterinarian first and the dog's future prescribers second, because a documented reaction is a durable piece of clinical information in a way that "he had a bad reaction to something once" is not.

6. Testing

6.1 What the Test Is

Genotyping for ABCB1-1Δ/nt230(del4) is a straightforward laboratory test from a cheek swab or blood sample. The inherited genotype is stable for life, so a valid result does not need to be repeated.

6.2 When Testing Is Most Useful

Testing is most informative in known at-risk breeds, dogs with plausible at-risk ancestry, and before a medication decision for which genotype would change drug selection, dose or monitoring. General-clinic data suggest that routine indiscriminate testing of every dog is not necessary: in one 2022 clinical population the deletion was rare and was found only in an at-risk breed (Beckers et al., 2022).

A dog that has tolerated one P-glycoprotein substrate has not been cleared for every other substrate, because risk is drug- and dose-specific.

6.3 When the Status Is Unknown

If an untested dog has clear risk ancestry and a drug with known ABCB1 relevance is being considered, a conservative prescribing approach is reasonable until the status is known. That is not the same as delaying all routine care. Where a licensed preventive has direct safety data in homozygous mutant dogs, parasite protection need not wait for genetic testing (Riggs et al., 2025).

6.4 Where the Result Should Go

A genotype that nobody knows about is of little practical value. It belongs in the veterinary record and should be available to emergency clinics, boarding facilities and household members who might administer medication.

6.5 What a Result Does Not Do

A genotype does not choose a drug. It changes the information available to the prescriber. The indication, current evidence for the exact medication, dose, route, co-medications and the individual dog's health still determine the plan.

6.6 Testing After a Suspected Reaction

Genotyping after a suspected reaction can identify whether the dog has a known inherited susceptibility. A positive result supports that vulnerability but does not prove that the specific clinical episode was caused by P-glycoprotein deficiency; a negative result does not exclude a drug reaction or an acquired transporter interaction.

6.7 One Common Deletion Is Not Every ABCB1 Question

The routine MDR1 test targets the well-established four-base-pair deletion. Other ABCB1 variants have been described, but their clinical meaning is not automatically established. In 2026, for example, the promoter-near c.-6-180T>G polymorphism was found widely across 17 breeds, yet its functional and clinical significance remained unclear (Palócz, Wágner & Csikó, 2026).

A negative ABCB1-1Δ result therefore answers an important, specific question. It should not be reinterpreted as proof that every possible P-glycoprotein-related interaction is excluded.

7. Breeding

7.1 What the Genetics Imply

Two carriers can produce affected offspring. Because frequencies are high in some breeds, eliminating the variant entirely would mean excluding a large share of the breeding population, which has its own costs for genetic diversity.

7.1a Where Testing Sits Among Other Screening

Genotyping for this variant is part of a broader shift toward testing individual dogs rather than reasoning from breed, and the same caution applies to all of it: a panel result is information for a clinician, not a verdict about an animal (how another canine variant is handled).

What distinguishes this one is how directly it converts into a decision. Many genetic findings are merely interesting; this one changes what is safe to prescribe next week.

7.2 The Pragmatic Approach

The common recommendation is to test breeding animals and avoid pairings that can produce homozygous affected puppies, rather than to remove all carriers from breeding. That is a population-management judgment rather than a research finding, and it depends on how common the variant is in the specific breed.

7.3 What Buyers Can Ask

In an at-risk breed, the genotypes of both parents are a reasonable question, and a breeder who cannot answer it has not done a test that costs very little. The answer also tells a buyer what to expect for their own puppy before any test is run.

7.4 Why Eradication Is Not the Goal

Where a variant reaches high frequency in a breed, removing every carrier from breeding narrows the gene pool substantially and trades one problem for others. That is a general principle in canine population genetics rather than something specific to this variant.

The alternative — test, then plan pairings so that no litter is expected to be homozygous mutant — can reduce high-risk genotypes while retaining genetic diversity. The appropriate population strategy depends on allele frequency, pedigree structure and other health priorities within the breed.

7.5 What Puppy Buyers Should Do Anyway

Whether an individual puppy needs its own test depends on what can be inferred securely from verified parental genotypes and parentage. If the puppy's genotype is not genetically determined by the mating or documentation is incomplete, individual testing provides a definitive result.

7.6 What a Breeding Decision Cannot Fix

Even a breed in which no affected puppies were produced for a generation would still contain carriers, and carriers still matter for a subset of substrates. Breeding decisions reduce the number of high-risk individuals; they do not remove the prescribing question from the breed.

A tested mating can make some puppy genotypes impossible and others still possible. The useful question is therefore not whether the breeder "tests for MDR1" in general, but what the parental results imply for this particular litter.

8. What This Is Not

8.1 Not Usually a Day-to-Day Clinical Disease

Most homozygous mutant dogs are not neurologically symptomatic in ordinary life simply because of their genotype. The classic clinical risk is still drug sensitivity. However, the variant is not biologically silent: altered cortisol metabolism and HPA-axis regulation have been demonstrated, and the possible relevance during severe illness is now an active clinical question (Gramer et al., 2022; Mealey, 2026).

There is not good evidence here for a shortened lifespan or for a predictable syndrome in otherwise healthy dogs, so those claims should not be invented in either direction.

8.2 Not an Established Behavioral Trait

No established evidence links ABCB1-1Δ to temperament, trainability or a characteristic behavior-problem profile. The HPA-axis finding is a physiological result; it should not be converted into a claim that MDR1 dogs are more anxious, fearful or stress-sensitive without direct behavioral evidence.

The established behavioral-medicine relevance remains pharmacological: medication choice and drug-drug interactions may require additional review (how behavioral medication is selected).

8.3 Not a Reason to Avoid Parasite Prevention

MDR1-affected dogs still need parasite prevention. The correct response is to use current product-specific evidence, not to avoid an entire drug class. Prospective safety studies exist for some licensed products in homozygous mutant Collies, including modern macrocyclic-lactone combinations (Riggs et al., 2025).

8.4 Not a Reason to Refuse Anesthesia or Surgery

Anesthesia and surgery are not automatically contraindicated. The genotype is information for the anesthetic plan. The treating veterinarian can select drugs and doses with the transport defect, concurrent medication and the procedure in mind.

8.5 "Natural" Does Not Bypass Pharmacology

Plant-derived and over-the-counter products can interact with transport proteins in experimental systems, but clinical relevance in dogs has not been established for every supplement marketed as "natural." The defensible household rule is simply to disclose everything the dog receives rather than assuming a product is irrelevant because it is non-prescription.

8.6 Three Everyday Situations

The dog with unknown ancestry. Unknown ancestry alone does not mandate a genetic test before every routine antiparasitic. If risk ancestry is plausible or a planned drug decision depends on ABCB1 status, testing is useful; meanwhile, product-specific safety data can guide necessary routine prevention.

The dog with diarrhea on a Sunday. Loperamide is a well-documented ABCB1 risk and should not be improvised from the household medicine cabinet. The correct response is veterinary advice, not dose guessing (Sartor et al., 2004).

The dog around horse or livestock wormers. Concentrated large-animal antiparasitics are another documented exposure route. Secure storage and keeping the product label available are practical safeguards.

8.7 What to Tell a Household

Four points cover most situations. The inherited genotype is tested once and remains valid for life. Homozygous mutant dogs have an important drug-transport defect and also measurable physiological differences in HPA-axis regulation. A P-glycoprotein substrate is not automatically forbidden; the exact drug, product and dose need current veterinary review. And any suspected neurological reaction after medication or accidental exposure is an urgent veterinary problem.

8.8 The One Sentence That Still Does Most of the Work

"My dog is ABCB1/MDR1 affected" — said before medication is prescribed or administered — remains the most useful practical sentence. The rest of the article explains why the veterinarian then needs to ask a more specific question: what is known about this drug, at this dose, in this genotype?

8.9 Three Everyday Situations

The shelter dog with unknown ancestry. Appearance suggests nothing, the history is empty, and the first routine antiparasitic is due next week. This is the clearest case for testing before treatment rather than after.

The dog with diarrhea on a Sunday. The household has loperamide in the cupboard and the practice is closed. This is the documented household scenario in the literature (Sartor et al., 2004), and the answer is to call the emergency service rather than to reach for the packet.

The tested dog at a new clinic. A genotype recorded at one practice is invisible at another. Carrying the result — on paper, in an app, on the dog's file — is what turns the test into protection.

8.10 Where the Behavioral Overlap Really Sits

Beyond the prescribing question, the practical overlap with behavior work is differential diagnosis. A dog that becomes ataxic, disoriented or unusually withdrawn after a treatment has a pharmacological problem, and no amount of training addresses it (where such medication is used).

Trainers are often the first to see such a change, because owners describe it as a behavior problem. Asking what the dog has been given recently costs one sentence and occasionally redirects an entire case.

9. Summary at a Glance

A 4-base-pair deletion is the main clinically established variant — It produces a frameshift and truncated, nonfunctional P-glycoprotein (Mealey et al., 2001).

The original ivermectin phenotype was strongest in homozygous mutants — Heterozygous dogs did not show increased ivermectin sensitivity in that original study, although carrier status matters for some other substrates (Mealey et al., 2001; Mealey et al., 2023).

The strongest pharmacokinetic evidence is not identical at every organ — Blood-brain-barrier protection and biliary excretion are well supported; intestinal effects vary by substrate, and renal contribution in dogs remains incompletely characterized (Mealey et al., 2023).

The biology extends beyond drug transport — Homozygous mutant dogs show altered cortisol metabolism and HPA-axis regulation; possible clinical consequences during severe illness are now being discussed (Gramer et al., 2022; Mealey, 2026).

It is not a Collie-only variant, but prevalence is population-specific — Testing cohorts and general-practice populations can produce radically different frequency estimates (Mealey & Meurs, 2008; Beckers et al., 2022).

Substrate does not mean prohibited — Some drugs are dangerous, some need adjustment or monitoring, and some products have direct safety data in homozygous mutant dogs (Mealey et al., 2023; Riggs et al., 2025).

The test is done once — A cheek swab or blood sample gives a lifelong inherited-genotype result.

9.1 What to Ask a Veterinarian

Useful questions are specific: Is this drug known to interact with canine P-glycoprotein? Is the concern substrate transport, inhibition, or both? Are there genotype-specific dose or product-safety data? What monitoring is appropriate? Those questions are more informative than asking whether a medication appears on a generic internet list.

9.2 The Evidence Base Is Uneven, Not Simply "Thin"

The field includes a strong molecular mechanism, pharmacokinetic studies, case reports, carrier data and prospective target-animal safety studies in genotyped dogs. The evidence is still uneven across individual drugs, which is why recommendations should be drug-specific rather than described globally as either "trial proven" or "consensus only" (Mealey et al., 2023; Riggs et al., 2025).

9.3 What Would Be Worth Studying

Priority questions now include the clinical importance of chronic HPA-axis suppression during severe illness, better characterization of heterozygous dogs across common substrates, the role of intestinal and renal P-glycoprotein for specific drugs, and prospective evaluation of clinically important drug-drug interactions.

10. Research Gaps and Critical Appraisal

The carrier phenotype remains drug-specific and incompletely mapped. Heterozygous dogs can show intermediate transporter function, but the clinical consequence depends on the substrate and dose (Mealey et al., 2023).

Intestinal and renal effects need cleaner canine data. Intestinal bioavailability findings are inconsistent across substrates, and the contribution of renal P-glycoprotein to drug clearance in dogs is not well quantified (Mealey et al., 2023).

HPA-axis findings are biologically convincing but clinically young. Altered cortisol metabolism is demonstrated, while the frequency and magnitude of clinically important adrenal insufficiency during illness remain to be established prospectively (Gramer et al., 2022; Mealey, 2026).

Prevalence data are population-dependent. Testing-laboratory samples can vastly overestimate the frequency encountered in routine general practice (Beckers et al., 2022).

Not every ABCB1 variant is clinically established. The common four-base-pair deletion has a clear pharmacogenetic phenotype. By contrast, the widely distributed c.-6-180T>G polymorphism still has uncertain functional and clinical significance (Palócz et al., 2026).

Drug-drug interactions deserve more in-vivo work. Canine in-vitro studies show that psychotherapeutic agents such as fluoxetine and clomipramine can inhibit P-glycoprotein, but the clinical magnitude of many such interactions remains unresolved (Schrickx & Fink-Gremmels, 2014).

Dosing evidence varies by medication. Some recommendations are based largely on mechanism and clinical experience; others are supported by prospective pharmacokinetic or target-animal safety studies. A single statement about the quality of "MDR1 dosing guidance" obscures this heterogeneity.

Communication across care settings is still a practical weak point. The pharmacogenetic result only helps when the people prescribing or administering a drug know it. How reliably genotype information follows dogs between clinics, emergency services, boarding facilities and households remains poorly studied.

11. Conclusion

The canine MDR1/ABCB1 phenotype is a model case for precision veterinary medicine. A four-base-pair deletion can abolish functional P-glycoprotein, with the clearest clinical consequence being altered exposure to certain drugs, particularly at the blood-brain barrier and in hepatobiliary disposition (Mealey et al., 2001; Mealey et al., 2023). The genotype does not turn every P-glycoprotein substrate into a contraindication: risk depends on the exact drug, dose, formulation, therapeutic margin and genotype, and prospective safety data exist for some modern products even in homozygous mutant Collies (Riggs et al., 2025).

The biology is also broader than the older phrase "healthy until a substrate is given." Homozygous mutant dogs show altered cortisol metabolism and HPA-axis regulation, and a 2026 review discusses potential relevance during severe illness (Gramer et al., 2022; Mealey, 2026). What has not been established is a characteristic temperament, behavioral syndrome or universal shortened lifespan. For everyday practice the most important action remains simple: know the dog's genotype when risk ancestry or medication choice makes it relevant, document it, and let current drug-specific veterinary evidence guide prescribing rather than a static internet blacklist.

Key Insights (Takeaways)

• ABCB1-1Δ/nt230(del4) truncates P-glycoprotein and can profoundly alter the handling of selected drugs.

• Homozygous mutant dogs carry the highest established risk, while heterozygous effects depend on the substrate and dose.

• Blood-brain-barrier and biliary effects are well supported; intestinal effects are variable and renal contribution in dogs remains less well characterized.

• The mutation also alters cortisol/HPA-axis physiology, but this does not establish a behavioral phenotype.

• "P-glycoprotein substrate" does not automatically mean "do not use." Product- and dose-specific evidence matters.

• Modern parasite prevention should not be withheld automatically: some licensed products have direct safety data in homozygous mutant Collies.

• Loperamide and concentrated horse/livestock antiparasitics remain important avoidable exposure risks.

• The genotype is tested once; its practical value depends on making the result available wherever medication may be given.

References

Beckers, E., Casselman, I., Soudant, E., Daminet, S., Paepe, D., Peelman, L., & Broeckx, B. J. G. (2022). The prevalence of the ABCB1-1Δ variant in a clinical veterinary setting: The risk of not genotyping. PLoS ONE, 17(8), e0273706. https://doi.org/10.1371/journal.pone.0273706

Gramer, I., Karakus, E., Hartmann, M. F., Wudy, S. A., Bauer, N., Moritz, A., Aktürk, Z., & Geyer, J. (2022). Urinary cortisol metabolites are reduced in MDR1 mutant dogs in a pilot targeted GC-MS urinary steroid hormone metabolome analysis. Journal of Veterinary Pharmacology and Therapeutics, 45(3), 265–272. https://doi.org/10.1111/jvp.13050

Gramer, I., Leidolf, R., Döring, B., Klintzsch, S., Krämer, E.-M., Yalcin, E., Petzinger, E., & Geyer, J. (2011). Breed distribution of the nt230(del4) MDR1 mutation in dogs. The Veterinary Journal, 189(1), 67–71. https://doi.org/10.1016/j.tvjl.2010.06.012

Mackin, A. J., Riggs, C., Beatty, T., Vangundy, Z., Mochel, J. P., & Archer, T. M. (2020). Excessive cyclosporine-associated immunosuppression in a dog heterozygous for the MDR1 (ABCB1-1Δ) mutation. Journal of the American Animal Hospital Association, 56(3), 190–195. https://doi.org/10.5326/JAAHA-MS-7004

Mealey, K. L. (2026). A mini-review of chronic suppression of the hypothalamic-pituitary-adrenal axis in animals with P-glycoprotein deficiency. Journal of Small Animal Practice. Advance online publication. https://doi.org/10.1111/jsap.70098

Mealey, K. L., & Meurs, K. M. (2008). Breed distribution of the ABCB1-1Δ (multidrug sensitivity) polymorphism among dogs undergoing ABCB1 genotyping. Journal of the American Veterinary Medical Association, 233(6), 921–924. https://doi.org/10.2460/javma.233.6.921

Mealey, K. L., Bentjen, S. A., Gay, J. M., & Cantor, G. H. (2001). Ivermectin sensitivity in collies is associated with a deletion mutation of the mdr1 gene. Pharmacogenetics, 11(8), 727–733. https://doi.org/10.1097/00008571-200111000-00012

Mealey, K. L., Owens, J. G., & Freeman, E. (2023). Canine and feline P-glycoprotein deficiency: What we know and where we need to go. Journal of Veterinary Pharmacology and Therapeutics, 46(1), 1–16. https://doi.org/10.1111/jvp.13102

Palócz, O., Wágner, R., & Csikó, G. (2026). Association between the nt230(del4) mutation and c.-6-180T>G polymorphism in the canine ABCB1 gene. Veterinary Medicine International, 2026, 1075604. https://doi.org/10.1155/vmi/1075604

Riggs, K. L., Wang, X., & Wiseman, S. (2025). Safety of Credelio Quattro™ (lotilaner, moxidectin, praziquantel, and pyrantel chewable tablets) in homozygous MDR1-mutant collie dogs. Parasites & Vectors, 18(1). https://doi.org/10.1186/s13071-025-06795-y

Sartor, L. L., Bentjen, S. A., Trepanier, L., & Mealey, K. L. (2004). Loperamide toxicity in a collie with the MDR1 mutation associated with ivermectin sensitivity. Journal of Veterinary Internal Medicine, 18(1), 117–118. https://doi.org/10.1111/j.1939-1676.2004.tb00145.x

Schrickx, J. A., & Fink-Gremmels, J. (2014). Inhibition of P-glycoprotein by psychotherapeutic drugs in a canine cell model. Journal of Veterinary Pharmacology and Therapeutics, 37(5), 515–517. https://doi.org/10.1111/jvp.12111