Home HLA and Immune Genetics HLA-B*15:02 Test: Carbamazepine Skin Reaction Risk and Results

HLA-B*15:02 Test: Carbamazepine Skin Reaction Risk and Results

By
Vita Library
-
4
Understand HLA-B*15:02 testing before carbamazepine, including who needs it, positive and negative results, SJS/TEN risk, medication choices, and urgent rash signs.

The HLA-B15:02 test checks whether a person carries an immune-system allele strongly linked to carbamazepine-induced Stevens–Johnson syndrome and toxic epidermal necrolysis. These rare reactions damage the skin and mucous membranes and can become life-threatening. Testing is most useful before the first dose of carbamazepine or oxcarbazepine, especially for people whose ancestry includes populations in which HLA-B15:02 is more common. A positive result usually leads clinicians to choose a different medicine unless the benefits clearly outweigh the risk and no suitable alternative exists. A negative result greatly lowers the specific HLA-B*15:02-related risk, but it does not eliminate all serious skin reactions or predict other carbamazepine side effects. The result is lifelong and normally needs to be obtained only once, provided the laboratory directly tested the allele with a validated method. Any new rash, blistering, mouth sores, facial swelling, fever, or eye pain after starting an antiseizure medicine needs urgent medical assessment.

  • A positive HLA-B*15:02 result means at least one copy of the risk allele was detected and carbamazepine or oxcarbazepine is usually avoided in treatment-naïve patients.
  • A negative result reduces the risk of HLA-B*15:02-associated SJS/TEN, but does not make carbamazepine risk-free.
  • Testing should be completed before the first dose whenever possible; no fasting or medication pause is usually required.
  • Ancestry guides testing because allele frequency varies, but self-identified race alone can miss mixed or unknown ancestry.
  • The test predicts a specific severe skin-reaction risk, not drug effectiveness, dose needs, or common side effects such as dizziness or low sodium.
  • A rash with blisters, skin pain, mouth or eye sores, or fever is an emergency warning, regardless of genotype.

Table of Contents

What HLA-B*15:02 Testing Measures

HLA-B15:02 is a specific allele of the HLA-B gene. HLA molecules sit on cell surfaces and present small protein fragments to T cells. This system helps the immune system distinguish ordinary body proteins from infected or abnormal cells. Certain medicines can interact with particular HLA molecules in a way that triggers a harmful T-cell response. With carbamazepine, HLA-B15:02 is strongly associated with Stevens–Johnson syndrome, or SJS, and toxic epidermal necrolysis, or TEN, particularly in several Asian populations.

SJS and TEN are part of the same disease spectrum. SJS involves less skin detachment, while TEN involves more extensive detachment. Both can affect the mouth, eyes, genitals, airways, and internal organs. The association with HLA-B15:02 is strongest for SJS/TEN rather than for every type of carbamazepine rash. Other HLA alleles, including HLA-A31:01, are linked to a broader range of carbamazepine hypersensitivity reactions in some populations.

The notation has a precise meaning. “HLA-B” identifies the gene. “15” identifies an allele family, and “02” identifies a specific protein-level allele. A report written as HLA-B15:02 positive means the laboratory detected at least one copy. Because people inherit one HLA-B allele from each parent, the report may show one or two HLA-B15:02 copies, but one copy is enough to classify the person as positive for medication guidance.

This is a focused pharmacogenetic test. It does not assess the full HLA region, seizure type, carbamazepine blood concentration, liver metabolism, or whether the medicine will control symptoms. It answers a narrower question: does this person carry an allele that substantially increases the risk of carbamazepine- or oxcarbazepine-associated SJS/TEN?

The result is generally stable for life because inherited DNA does not change. A valid prior result can often be reused, but the clinician should verify that the earlier test specifically resolved HLA-B15:02. Some older or low-resolution HLA tests report only HLA-B15, which includes many alleles and cannot reliably substitute for direct HLA-B15:02 typing.

Who Should Be Tested Before Carbamazepine

Testing is most important before carbamazepine or oxcarbazepine is started in people with ancestry from populations where HLA-B*15:02 occurs at meaningful frequency. The allele is more common in parts of Southeast Asia, South Asia, and some East Asian and Oceanian populations. Frequency varies greatly within countries and ethnic groups, so broad labels such as “Asian” are an imperfect shortcut.

Clinical guidance commonly emphasizes testing for people with ancestry connected to areas such as:

  • Thailand, Vietnam, Malaysia, Singapore, Indonesia, or the Philippines
  • southern China and some other Chinese populations
  • India and other parts of South Asia
  • regions where family ancestry is mixed, uncertain, or not fully known but may include higher-frequency populations

The allele is generally uncommon in many European, Japanese, Korean, African, and Indigenous American populations, but uncommon does not mean impossible. Migration, mixed ancestry, adoption, and incomplete family history can make ancestry-based screening unreliable. Some health systems therefore use broader or universal preemptive testing when rapid, affordable testing is available.

Testing is normally ordered before the first exposure. The highest SJS/TEN risk is early in treatment, especially during the first several weeks to three months. Waiting until a rash appears defeats the preventive purpose. A result obtained after treatment has begun may still be useful, but it should not delay evaluation of symptoms.

People who have taken carbamazepine continuously for more than three months without a skin reaction are at much lower risk of developing HLA-B*15:02-associated SJS/TEN from continued use. Guidelines may allow ongoing therapy in an established, tolerant patient even when a newly obtained result is positive, after careful clinical review. This is not the same as restarting carbamazepine after an interruption or after a prior hypersensitivity reaction. The prescriber must consider treatment history, duration, and alternatives.

Testing is also relevant when oxcarbazepine is being considered because the same allele is associated with oxcarbazepine-induced SJS/TEN. Evidence for related aromatic antiseizure medicines is less uniform. HLA-B*15:02 is associated with severe skin reactions from phenytoin in several populations, and some guidance extends caution to fosphenytoin. Associations with lamotrigine, eslicarbazepine, and other drugs are less consistent or supported by lower-quality evidence. A clinician should not assume that every structurally related drug is automatically safe—or automatically contraindicated—based on one result.

A person with a personal history of SJS/TEN from carbamazepine generally should not be re-exposed, regardless of whether later genetic testing is negative. The clinical reaction itself outweighs a negative marker result.

How the Test Is Done

HLA-B*15:02 testing uses DNA from blood, a cheek swab, or saliva, depending on the laboratory. No fasting is required, and ordinary medicines do not alter the inherited genotype. The sample can often be collected in a clinic, hospital, pharmacy-based program, or home kit, but treatment decisions should use a test from a qualified clinical laboratory.

Laboratories may use targeted genotyping, sequence-specific PCR, real-time PCR, Sanger sequencing, or next-generation sequencing. A targeted assay answers whether HLA-B*15:02 is present. Broader HLA-B genetic testing may report both HLA-B alleles at high resolution and can sometimes identify other clinically relevant variants.

Turnaround time ranges from same-day or a few days for rapid assays to one or two weeks for send-out testing. When carbamazepine is not urgently required, it is safer to wait for the result before prescribing. When treatment cannot wait, the clinician may choose a medicine not dependent on the HLA-B*15:02 result.

A complete report should state the tested allele and the result clearly. Common formats include:

  • HLA-B*15:02 detected / not detected
  • positive / negative
  • HLA-B allele pair, such as HLA-B*15:02 plus another HLA-B allele
  • an uncommon “indeterminate,” “inconclusive,” or “no call” result if the assay cannot resolve the genotype

The specimen source matters after an allogeneic bone marrow or stem cell transplant. Blood DNA may reflect the donor’s genotype rather than the recipient’s inherited genotype. In that situation, the laboratory may need a pre-transplant specimen or a non-blood tissue source. Recent blood transfusion usually does not alter routine DNA testing in the same way, but the laboratory should still be told about transfusions, transplant, or unusual sample circumstances.

Direct-to-consumer genotyping may use nearby markers or imputation rather than high-resolution HLA typing. Because treatment decisions can prevent a life-threatening reaction, an unconfirmed consumer result should not be treated as equivalent to a validated clinical test.

Positive, Negative, and Indeterminate Results

Positive result

A positive result means HLA-B*15:02 was detected. For a person who has never taken carbamazepine, clinical guidelines generally recommend avoiding carbamazepine because the increased SJS/TEN risk outweighs the benefit when suitable alternatives are available. Oxcarbazepine is also generally avoided in a treatment-naïve positive patient.

A positive result does not mean the person will definitely develop SJS/TEN. Many carriers could take the drug without the reaction. The problem is that the reaction can be severe, rapid, and difficult to predict, while alternative medicines are often available. The allele therefore has strong preventive value even though its positive predictive value is far below 100%.

The number of copies usually does not change the basic recommendation. One copy is sufficient to classify the patient as positive. The result may also have implications for biologic relatives, but routine family testing is usually driven by whether a relative is considering a relevant medicine rather than by disease inheritance. HLA alleles are inherited, so a parent, sibling, or child may also carry HLA-B*15:02.

Negative result

A negative result means the assay did not detect HLA-B15:02. This sharply reduces the likelihood of the specific HLA-B15:02-associated SJS/TEN reaction, especially in populations where the association has been well established. It does not reduce the risk to zero.

A negative patient can still develop:

  • SJS/TEN through another genetic or non-genetic mechanism
  • drug reaction with eosinophilia and systemic symptoms, or DRESS
  • a milder rash
  • liver injury, blood-cell abnormalities, low sodium, dizziness, or other carbamazepine effects
  • hypersensitivity associated with HLA-A*31:01 or another allele

The prescriber still needs ordinary clinical monitoring. HLA-B*15:02 testing should never replace prompt evaluation of a new rash or systemic symptoms.

Indeterminate or ambiguous result

An indeterminate result means the laboratory could not reliably call the allele. Reasons include low DNA quality, assay interference, a rare nearby allele, insufficient resolution, or sample mix-up. The result should not be treated as negative. Repeat collection or a different high-resolution method is usually appropriate before carbamazepine exposure.

A report may also identify HLA-B15 family alleles that are not HLA-B15:02. These are not interchangeable. Some alleles, such as HLA-B15:11 or HLA-B15:21, have been associated with severe reactions in certain populations, but recommendations and evidence differ. The ordering clinician or pharmacogenomics specialist should interpret the exact allele rather than relying on the first two digits.

Medication Decisions After the Result

The result informs drug selection; it does not select the replacement medicine by itself. Carbamazepine is used for focal seizures, trigeminal neuralgia, bipolar disorder, and several off-label indications. Alternatives depend on the condition, age, pregnancy potential, kidney and liver function, interacting medicines, and prior treatment response.

For epilepsy, possible alternatives may include levetiracetam, valproate, lacosamide, topiramate, or others, but each has its own indications and risks. Valproate is often avoided or tightly restricted in people who could become pregnant because of fetal and developmental risks. Lamotrigine also carries a risk of serious rash, particularly with rapid dose escalation or valproate coadministration. Medication substitution therefore requires an individualized plan rather than a simple “positive equals one specific drug” rule.

For trigeminal neuralgia, oxcarbazepine is commonly considered an alternative to carbamazepine, but an HLA-B*15:02-positive result affects both drugs. Other options may be less effective or less well tolerated. Neurology or pain specialists may need to balance pharmacogenetic risk against the severity of the condition.

For bipolar disorder, the clinician may consider lithium, valproate, lamotrigine, an atypical antipsychotic, or another regimen according to the clinical phase and patient factors. The HLA test does not compare their effectiveness.

When a positive patient has already tolerated carbamazepine for longer than three months, guideline recommendations are more nuanced. The severe reaction usually occurs early, so continuation may be considered with caution if treatment has been uninterrupted and effective. The patient should not stop carbamazepine abruptly without medical advice, because sudden withdrawal can provoke seizures or psychiatric destabilization. The clinician should document the genotype, exposure history, and reasoning.

A prior negative result should be entered into the medical record in a reusable form. The report should name the exact allele and laboratory. Recording only “pharmacogenetics normal” is not sufficient because other drug–gene pairs may remain untested.

Clinical teams also need a plan for results that arrive after a prescription has been written. If the patient has not taken the first dose, the prescriber can usually pause and review the genotype. If exposure has already started, the decision depends on the result, treatment duration, indication, and availability of a safe substitute. A positive result during the early high-risk period usually prompts urgent prescriber review rather than automatic unsupervised discontinuation. For epilepsy, the replacement medicine may need a loading strategy or cross-taper to prevent breakthrough seizures.

Electronic health records can preserve the value of testing by placing HLA-B*15:02 in both the pharmacogenomics section and the medication-alert system. The alert should distinguish a genetic risk from a proven prior allergy. It should also name carbamazepine and oxcarbazepine rather than using an overly broad class warning that blocks every antiseizure medicine. Patients can keep a copy of the report or a wallet-card summary, especially when they receive care from several health systems.

Skin-Reaction Signs and Timing

SJS/TEN most often begins within the first one to eight weeks after starting the culprit medicine, although the broader high-risk period extends through roughly the first three months. Early symptoms may resemble a viral illness before obvious skin detachment appears.

Warning signs include:

  • fever, marked fatigue, sore throat, or flu-like symptoms
  • painful or burning skin rather than a simple itchy rash
  • red or purple spreading spots, target-like lesions, or blisters
  • peeling or detachment of skin
  • painful mouth sores, lip crusting, or difficulty swallowing
  • red, painful, light-sensitive eyes or vision changes
  • genital or urinary mucosal sores
  • facial swelling, swollen lymph nodes, or breathing difficulty

A person taking carbamazepine or another high-risk medicine who develops these symptoms needs immediate medical assessment. The drug should not be restarted after a clinician suspects SJS/TEN. Because abrupt antiseizure-drug withdrawal can also be dangerous, patients should seek emergency guidance rather than trying to manage the situation alone.

Not every rash is SJS/TEN, but it is difficult to distinguish early severe reactions from mild eruptions without examination. DRESS may develop two to eight weeks after exposure and can include rash, fever, facial edema, eosinophilia, enlarged lymph nodes, and injury to the liver, kidneys, lungs, or heart. HLA-B*15:02 is not a reliable stand-alone predictor for DRESS.

Photos, a full medication timeline, laboratory tests, and sometimes skin biopsy help specialists assess causality. The reaction should be documented with the exact drug and clinical diagnosis. “Allergy to all anticonvulsants” is too broad and may unnecessarily restrict future therapy, while “carbamazepine-associated SJS/TEN” communicates the actual risk.

Limitations and Common Misunderstandings

HLA-B*15:02 testing is highly useful, but it has a defined scope. The following distinctions prevent common errors:

  • It is not a general allergy test. It predicts a genetically associated severe skin-reaction risk for specific drugs.
  • It is not a carbamazepine dose test. Drug concentration and dose depend on clinical response, interactions, liver metabolism, age, and other factors.
  • Negative does not mean “safe.” Serious reactions can occur through other pathways.
  • Positive does not diagnose a current reaction. A patient with symptoms needs clinical evaluation, not just genetic interpretation.
  • HLA-B15 is not the same as HLA-B*15:02. Low-resolution typing may be insufficient.
  • Ancestry is a screening aid, not a biologic guarantee. Mixed and unknown ancestry can hide risk.
  • One result does not cover every aromatic antiseizure drug. Evidence and recommendations differ by medicine.

HLA-A31:01 is a separate marker linked to carbamazepine hypersensitivity, including maculopapular eruption, DRESS, and SJS/TEN in several ancestry groups. A person may be HLA-B15:02 negative but HLA-A31:01 positive. Some clinical panels test both alleles. The separate HLA-A31:01 result should be interpreted according to its own evidence and guideline recommendations.

Rare genotyping errors, allele ambiguity, sample problems, and limited assay coverage can occur. A result that conflicts with ancestry, prior typing, or a serious reaction may warrant repeat high-resolution testing. Laboratories should use current HLA nomenclature and specify whether they detect rare alleles that can interfere with targeted assays.

The most effective use of testing combines accurate laboratory methods, pre-prescription timing, electronic medication alerts, clinician education, and rapid recognition of symptoms. Genotyping reduces preventable risk, but clinical vigilance remains necessary for every patient starting carbamazepine or oxcarbazepine.

A pharmacogenetic result should be interpreted before the first carbamazepine dose whenever feasible. Testing after several uneventful months has less preventive value because the highest risk period for severe cutaneous reactions is early, although the allele remains relevant if the medicine is stopped and later restarted.

Patients should know the early warning signs that require immediate medical advice: a spreading rash, blisters, painful skin, mouth or eye sores, facial swelling, fever, or flu-like illness. They should not take another dose while waiting for routine advice when a severe reaction is suspected. Emergency evaluation is appropriate for mucosal involvement, skin pain, breathing difficulty, or widespread blistering.

Alternative medicines have their own genetic and clinical risks. Oxcarbazepine carries a related HLA-B*15:02 concern, while aromatic antiseizure medicines can show cross-reactivity that is not fully predicted by this allele. The prescriber should choose an alternative based on seizure type, psychiatric indication, ancestry, pregnancy potential, organ function, interactions, and previous reactions rather than genetics alone.

References

Disclaimer

This article is educational and does not replace prescribing advice, genetic counseling, or emergency assessment. Do not start, stop, or restart carbamazepine or another antiseizure medicine based only on this information. A painful rash, blistering, mouth or eye sores, fever, or skin peeling after medication exposure requires urgent medical care.

RELATED ARTICLESMORE FROM AUTHOR