Genetic marker predicts fat loss due to d4T in Thai patients

A genetic marker can predict with a high level of accuracy whether Thai patients receiving antiretroviral therapy that contains d4T (stavudine) will develop lipoatrophy (subcutaneous fat loss) as a result of treatment, researchers from Thailand and Japan report in Clinical Infectious Diseases this week.

Lipoatrophy – fat loss from the limbs, face and buttocks - is a frequent side-effect of antiretroviral therapy that contains d4T, and to a lesser extent AZT (zidovudine).

Around 40% of patients develop this side-effect after two to three years of treatment, and for this reason d4T has been dropped as a component of first-line treatment in Europe, North America and other well-resourced settings.

However d4T remains a component of first-line therapy for several million people in low and middle-income countries. Despite a WHO recommendation to move away from the use of d4T, some countries – including South Africa and Thailand – retain the drug as a component of first-line treatment due to its low cost.

Although d4T is widely acknowledged to be the primary cause of lipoatrophy, there is some evidence that the cytokine tumour necrosis factor and the genes that regulate its production may have some impact on the development of fat loss in response to ART.

Pharmacogenomic studies have recently shown linkages between particular genetic variants and the side-effects of abacavir and nevirapine. The study reported this week set out to examine whether any genetic variant is associated with the development of lipoatrophy.

Researchers at Mahidol University in Bangkok, and the Center for Genomic Medicine in Tokyo, who had previously conducted a study of genetic predictors for nevirapine-associated rash, were able to use genetic sequencing information from 103 patients receiving antiretroviral therapy.

All patients underwent evaluation for lipodystrophy, including patient and physician grading of symptoms, skinfold testing to measure fat thickness, bioelectrical impedance to measure fat mass and DEXA scan to measure fat distribution.

Fifty-five lipodystrophy cases were identified, and matched with a control group of ART patients without lipodystrophy. Apart from body fat distribution the only significant differences in characteristics between the two groups were in the proportion of patients with undetectable viral load (100% of the lipoatrophy group compared to 88% of the control group, p=0.013), and the proportion with a prior AIDS diagnosis 58% vs 37%, p=0.036).

Patients had been exposed to antiretroviral therapy for a median of 47 months in the lipodystrophy group and 44 months in the control group.

The lipodystrophy group had significantly lower body weight (53kg vs 56kg), lower body mass index (20.2 vs 23.1), lower midthigh, waist and hip circumference, and significantly lower skinfold thickness at the triceps, biceps and suprailiac measurements. DEXA scans showed lower percentages of fat in the legs, arms and trunk in the lipodystrophy group.

Analysis of genotypes in the HLA-A, HLA-B, HLA-C, HLA-DRB-1, HLA-DQB1 and HLA-DPB1 loci showed that the HLA-B*4001 allele was present at a significantly higher frequency in lipodystrophy cases, although its presence was by no means universal (present in 29% of cases).

Nevertheless multivariate analysis showed that carriage of this allele was strongly associated with lipodystrophy (OR 14.05, 95% confidence interval 2.57 – 76.59, p=0.02).

In comparison, each extra month of stavudine treatment increased the risk of lipodystrophy by 2% (OR 1.02, 95% CI 1.00 – 1.04, p= 0.02).

The authors say that the fact some patients did not develop lipodystrophy despite more than four years of stavudine treatment supports the view that some people are more genetically prone to lipodystrophy when they receive treatment with the drug.

Carriage of the allele had a positive predictive value for lipodystrophy of 88.9%, whereas absence of the allele was a much weaker predictor of the absence of lipodystrophy (negative predictive value of 54%).

The authors say that this genetic variant is known to be associated with HIV disease progression and to have a role in antigen presentation, but do not speculate about the possible mechanism that might account for the association between this gene and lipodystrophy.

They say that larger, prospective studies are needed to confirm the result, and to study the association in other racial groups. They also note the risk of sampling bias in their finding, because every patient was recruited to this study as a result of participation in a previous study of nevirapine rash.

“If these findings are confirmed in future prospective studies, then it will be possible to use this allele to avoid a substantial portion of stavudine-associated lipodystrophy,” they conclude.

“In settings where stavudine cannot be totally replaced with zidovudine or tenofovir because of limited resources, HLA-B*4001 may be used to select a subset of patients who are at higher risk of lipodystrophy and have priority to substitute stavudine with other NRTIs.”