Abacavir hypersensitivity (HLA-B*57:01)

Carrying the HLA-B*57:01 immune type makes a severe allergic reaction to the HIV drug abacavir very likely, so abacavir is avoided in people who screen positive. This marker tags that immune type. It only matters if abacavir is ever considered.

What this marker stands in for

This variant is not itself the cause of anything. It sits in HCP5, next to the HLA region, and travels on the same ancestral stretch of chromosome as HLA-B*57:01, the immune-system allele that actually causes abacavir hypersensitivity.

Abacavir binds inside the HLA-B*57:01 molecule and changes which self-peptides it presents to T cells, which the immune system reads as foreign. The reaction, appearing in the first six weeks as fever, rash, gastrointestinal and respiratory symptoms, can be fatal on re-exposure.

This is a genuine screening success story

The PREDICT-1 trial established that screening for HLA-B*57:01 before starting abacavir essentially eliminates immunologically confirmed hypersensitivity reactions, and screening became standard of care worldwide. CPIC rates the association level A and recommends avoiding abacavir in anyone who carries the allele.

It is one of the clearest demonstrations that a pharmacogenetic test can prevent harm rather than merely refine a dose, and it changed practice within a few years.

Why a consumer result is not the test

The distinction between a tag and the allele itself matters here more than anywhere else on this site. In European-ancestry populations the correlation is close to complete, and a Spanish study reported 100% sensitivity and specificity against direct typing. In other populations the tag is less reliable, and studies in Italian and Mexican samples concluded that tag variants do not substitute for HLA-B*57:01 typing for treatment eligibility.

Clinical practice therefore uses direct HLA typing, not this marker. Never start, stop or refuse abacavir on the basis of a consumer genotype: if abacavir is being considered, ask for the clinical test, which is routine and inexpensive.

How common it is

In 1000 Genomes reference data, about 7% of European-ancestry individuals and 9% of South Asian-ancestry individuals carry one copy of the tag allele, against roughly 2% in African-ancestry and 1% in East Asian samples. HLA-B*57:01 frequency follows a broadly similar pattern.

What each rs2395029 genotype means

rs2395029 has three possible genotypes: TT, GT and GG.

HLA-B*57:01 not detected (rs2395029 TT). You don't carry the marker for HLA-B*57:01, so this tag makes the immune type that causes severe abacavir allergy unlikely for you. If abacavir is ever needed, standard guidance applies (a clinic may still confirm with a direct HLA test).

HLA-B*57:01 likely (avoid abacavir) (rs2395029 GT). You carry one copy of the marker for HLA-B*57:01, the immune type linked to a severe, sometimes dangerous allergic reaction to the HIV drug abacavir. Guidelines say to avoid abacavir in people who are positive. This only matters if abacavir is ever considered, and a direct HLA-B*57:01 test should confirm it.

HLA-B*57:01 likely (avoid abacavir) (rs2395029 GG). You carry two copies of the marker for HLA-B*57:01, strongly suggesting you have this immune type, which makes a severe allergic reaction to the HIV drug abacavir very likely. Abacavir should be avoided, confirmed by a direct HLA-B*57:01 test. It has no effect unless abacavir is being considered.

Evidence & sources

Across global populations, fewer than 1% of people carry two copies of this variant, and about 94% carry none.

CPIC provides dosing guidance for abacavir based on this gene.

Common questions

Does this result mean I am allergic to abacavir? It suggests you may carry HLA-B*57:01, which would mean abacavir should be avoided. It is not the clinical test, and in non-European populations the correlation is imperfect. If abacavir is ever proposed, direct HLA-B*57:01 typing is the standard step and is what should decide.

Does this affect any other drug? HLA-B*57:01 is also associated with liver injury from flucloxacillin, though far less predictably than the abacavir reaction and not at a level that guides prescribing in the same way.

Related

Drug acetylation speed (NAT2) · Efavirenz / drug metabolism (CYP2B6*6) · Gilbert syndrome (UGT1A1 gene) · Hepatitis C treatment response (IL28B) · Opioid receptor response (OPRM1 A118G) · Thiopurine metabolism (TPMT and NUDT15 genes) · Warfarin dose sensitivity