Residency · Residency · Medical Genetics Genomics

HLA Pharmacogenomics and Adverse Drug Reactions

Introduction

The human leukocyte antigen (HLA) system, the most polymorphic region of the human genome, plays a critical role in immune-mediated adverse drug reactions (ADRs). Specific HLA alleles confer dramatically increased risk for severe cutaneous adverse reactions (SCARs) and other hypersensitivity syndromes, making HLA pharmacogenomics one of the most clinically impactful areas of precision medicine.

HLA System Overview

Structure and Function

HLA genes are located in the major histocompatibility complex (MHC) on chromosome 6p21.3. Class I molecules (HLA-A, -B, -C) present intracellular peptides to CD8+ T cells and are expressed on virtually all nucleated cells. Class II molecules (HLA-DR, -DQ, -DP) present extracellular peptides to CD4+ T cells and are expressed on antigen-presenting cells. Extreme polymorphism characterizes the system, with over 30,000 HLA alleles identified across the population. HLA allele frequencies vary dramatically across ethnic populations, with direct implications for pharmacogenomic risk.

Nomenclature

The standard format is HLA-[gene][allele group]:[protein] (for example, HLA-B57:01). Two-field resolution (4-digit) is the minimum for pharmacogenomic applications. High-resolution typing distinguishes alleles that differ in the peptide-binding groove.

Mechanisms of HLA-Mediated Drug Reactions

Proposed Models

The hapten/prohapten model proposes that a drug or metabolite covalently binds to a protein, creating a neoantigen presented by HLA. The pharmacological interaction (p-i) model suggests the drug binds non-covalently directly to the HLA molecule or the HLA-peptide complex, stimulating T cells without antigen processing. The altered peptide repertoire model proposes that the drug binds within the HLA peptide-binding groove, changing the spectrum of self-peptides presented; this mechanism is best characterized for HLA-B*57:01 and abacavir.

Immune Pathophysiology

Drug-specific CD8+ cytotoxic T cells and/or CD4+ T cells and NK cells mediate tissue damage. Cytokines including TNF-alpha, IFN-gamma, granulysin, and perforin/granzyme contribute to keratinocyte apoptosis in SCARs. Granulysin has been identified as a key mediator of blister formation in SJS/TEN.

Clinically Validated HLA-Drug Associations

HLA-B*57:01 and Abacavir

Abacavir hypersensitivity syndrome (AHS) is a multi-organ reaction with fever, rash, GI symptoms, and respiratory symptoms; rechallenge can be fatal. HLA-B57:01 has approximately 100% negative predictive value, and testing before prescribing virtually eliminates AHS. CPIC and major HIV treatment guidelines mandate HLA-B57:01 testing before abacavir initiation. This represents the most successful example of pharmacogenomic implementation globally. The allele frequency is approximately 5-8% in Europeans, approximately 1% in East Asians, and variable in other populations.

HLA-B*15:02 and Carbamazepine

A strong association exists with Stevens-Johnson syndrome/toxic epidermal necrolysis (SJS/TEN). Risk is highest in populations of Southeast Asian descent where the allele frequency is 6-8%. The FDA boxed warning recommends testing in patients of Southeast Asian ancestry before prescribing carbamazepine. The association extends to other aromatic antiepileptics including oxcarbazepine, phenytoin, and lamotrigine (with weaker evidence). HLA-B*15:02 prevalence is very low (less than 1%) in European and African populations.

HLA-B*58:01 and Allopurinol

HLA-B58:01 is associated with allopurinol hypersensitivity syndrome including SJS/TEN and DRESS. The risk allele frequency is approximately 6-8% in East Asians, approximately 3.5% in African Americans, and approximately 1-2% in Europeans. ACR guidelines recommend HLA-B58:01 testing before allopurinol initiation in populations with elevated allele frequency. Febuxostat is an alternative urate-lowering therapy for carriers.

HLA-A*31:01 and Carbamazepine

This allele is associated with DRESS (Drug Reaction with Eosinophilia and Systemic Symptoms) and maculopapular exanthema, primarily in European and Japanese populations. It has a lower odds ratio than HLA-B*15:02 for SJS/TEN but a broader phenotypic spectrum. The allele frequency is approximately 2-5% in Europeans.

Additional Associations

HLA-B13:01 is associated with dapsone hypersensitivity syndrome in Southeast Asian populations. HLA-B59:01 is linked to methazolamide-induced SJS/TEN in East Asian populations. HLA-DRB107:01 is associated with lapatinib-induced hepatotoxicity. HLA-A33:01 is linked to ticlopidine-induced liver injury.

Severe Cutaneous Adverse Reactions

Classification

SJS (Stevens-Johnson Syndrome) involves less than 10% body surface area epidermal detachment with mortality of approximately 5-10%. TEN (Toxic Epidermal Necrolysis) involves greater than 30% BSA detachment with mortality of approximately 25-35%. SJS/TEN overlap involves 10-30% BSA detachment. DRESS presents with fever, rash, eosinophilia, and internal organ involvement (liver most commonly) with mortality of approximately 5-10%. AGEP (Acute Generalized Exanthematous Pustulosis) features rapid onset of sterile pustules and is generally self-limited.

Clinical Management Implications

Immediate drug discontinuation is the most critical intervention. Supportive care in a burn unit is required for extensive SJS/TEN. The SCORTEN score predicts mortality in SJS/TEN. The HLA allele and causative drug must be permanently documented in the medical record and alert systems.

Implementation Considerations

Testing Platforms

PCR-based methods include sequence-specific primer (SSP) and sequence-specific oligonucleotide (SSO) typing. Next-generation sequencing-based HLA typing provides the highest resolution and is increasingly available. Results should ideally be available before drug initiation, and some institutions maintain reflex testing protocols.

Cost-Effectiveness

Pre-prescription HLA-B57:01 testing for abacavir is cost-effective across all populations studied. HLA-B15:02 testing before carbamazepine is cost-effective in Southeast Asian populations but not in populations with very low allele frequency. HLA-B*58:01 testing for allopurinol is cost-effective in East Asian and African American populations.

Clinical Pearls

HLA-B57:01 testing before abacavir is the gold standard of pharmacogenomic implementation, with near-complete elimination of abacavir hypersensitivity in tested populations. Population ancestry determines which HLA-drug associations are most relevant; HLA-B15:02 testing for carbamazepine is critical in Southeast Asian populations but has limited utility in Europeans. A negative HLA test does not eliminate all risk of drug hypersensitivity; it reduces risk from the specific HLA-mediated mechanism. All HLA-associated severe drug reactions and the specific HLA allele should be permanently documented in the patient record as a drug allergy.

References

1. Mallal S, Phillips E, Carosi G, et al. HLA-B5701 screening for hypersensitivity to abacavir. New England Journal of Medicine*. 2008;358(6):568-579.
2. Chung WH, Hung SI, Hong HS, et al. Medical genetics: a marker for Stevens-Johnson syndrome. Nature. 2004;428(6982):486.
3. Stamp LK, Taylor WJ, Jones PB, et al. Starting dose is a risk factor for allopurinol hypersensitivity syndrome: a proposed safe starting dose of allopurinol. Arthritis & Rheumatism. 2012;64(8):2529-2536.
4. Illing PT, Vivian JP, Dudek NL, et al. Immune self-reactivity triggered by drug-modified HLA-peptide repertoire. Nature. 2012;486(7404):554-558.

Read this lecture as Markdown