Gene most commonly seen in Africans greatly increases efavirenz levels

Hard on the heels of the bad news regarding nevirapine monotherapy to prevent mother-to-child transmission, reported here, three studies presented on the last day of the Eleventh Conference on Retroviruses and Opportunistic Infections in San Francisco could have major global implications regarding the use of the non-nucleoside class as a whole.

STOP study: efavirenz levels surprisingly high in African women

First, Dr Stephen Taylor of Birmingham's Heartlands Hospital presented data from the STOP study, which was originally designed to discover the ideal time to stop efavirenz when discontinuing highly active antiretroviral therapy (HAART). Whilst the answer to that question was up to two weeks (a week longer than currently recommended in the 2003 BHIVA guidelines, although switching to another drug with a short half-life two weeks prior to stopping seems a safer alternative), the most important finding was that efavirenz levels in African women can persist up to five times longer and reach peak levels almost three times higher than average.

This small pharmacokinetic (PK) study enrolled six Caucasian males and four African women who were on efavirenz-based HAART (with Combivir the most common nucleoside component) and measured efavirenz concentrations at baseline, and four, seven, 14, and 21 days after stopping efavirenz. Viral load measurements and resistance testing was also performed at each time point.

Five of the ten participants – all Caucasian males – had efavirenz half-lives within the expected range of 40 - 55 hours. However, the other five had half-lives in excess of 100 hours (ranging from 116 - 228 hours); four of these were the African women participants, the other was a Caucasian man aged 64. Surprisingly, three of the African women were found to still have therapeutic levels of efavirenz two weeks after stopping the drug. All the African women also had very high peak levels of efavirenz prior to stopping the drug.

The next two presentations – both sub-studies of the notorious ACTG 5095 trial, which compared efavirenz-based HAART with the triple nucleoside combination of AZT (zidovudine)/3TC (lamivudine)/abacavir – provided some explanation as to why Taylor had observed such high and long-lasting efavirenz levels in the African women his study.

A5097s: ethnicity and weight associated with efavirenz levels

First, Harvard’s Heather Ribaudo presented results from A5097s, the 5095 sub-study that evaluated central nervous system (CNS) side-effects in those taking efavirenz over the first 24 weeks of treatment and looked for associations between efavirenz clearance and gender, ethnicity and weight.

Of the 202 people eligible, samples from 190 were available for analysis: 19% were women, 53% were Caucasian, 32% were African-American, 12% were Hispanic and 3% were not given an ethnicity. The median weight of the cohort was 75kg (range: 46 - 186kg).

Ribaudo found that ethnicity (which she termed “race”) was very strongly associated with efavirenz clearance: African-Americans and Hispanics had, on average, a 32% slower clearance of efavirenz than Caucasians (95% confidence interval: 15%, 51%; p 0.26).

Although 16% of participants discontinued efavirenz during the study - primarily due to CNS symptoms (n = 5), rash (n = 5), other toxicity (n = 4), subject/clinician decision (n = 6), and non-compliance (n = 6) - there was only a moderate association between an increasing rate of efavirenz discontinuation with decreasing clearance (p = 0.052) and increasing C_max (p = 0.048). However, there was no apparent association between EFV clearance and rates of first CNS toxicity (p = 0.99) or viral load of

NWCS214: gene found mainly in African-Americans associated with slower efavirenz clearance

The second sub-study, presented by David Haas of Vanderbilt University in Nashville, examined the relationships between CNS toxicity, pharmacokinetics, and genetic polymorphisms (i.e. common DNA variations). Efavirenz is metabolised in the liver by an enzyme known as CYP2B6; variations in this gene include the G/G, G/T and T/T genotypes.

Haas found that the T/T genotype at position 516 was more common in African-Americans (20%) than Caucasians (3%) and was associated with slower clearance and higher plasma efavirenz levels in all subjects (p = 0.001). The same association was seen when the T/T genotype was assessed separately among the different ethnicities.

When 24-hour median EFV levels (AUC₂₄) were analysed according to G/G, G/T, and T/T genotype, it was found that individuals with the T/T genotype had more than double the drug exposure at 130µg.h/l (n = 14) compared with 44 and 60 in those with G/G and G/T genotypes respectively. Additionally, the CYP2B (G⁵¹⁶T) T/T genotype was associated with adverse CNS symptoms (p = 0.04). None of the genotypes studied showed significant associations with initial or short-term changes in viral load or CD4 cell counts.

Haas concluded that although “racial differences may be largely explained by this T/T genotype,” it is also likely that other, undiscovered genetic polymorphisms could play an even more important role in the metabolism of efavirenz. These polymorphisms were also likely to affect the metabolism of nevirapine and other classes of antiretrovirals.

What does this all mean?

There are several important implications from the results of these three studies. If ethnic differences in levels and tolerance of efavirenz are due to genetics, then, in resource-rich countries, it should be possible to individualise therapy to reduce toxicity, minimise resistance and increase efficacy, through therapeutic drug monitoring – and, eventually, genetic testing.

There are implications for the scaling-up of efavirenz- or nevirapine-based HAART by the World Health Organization's 3x5 programme. Is it possible that the standard dose of efavirenz will be less well tolerated in Africans, who are more likely to have the T/T polymorphism, but possibly also weigh less than their Western counterparts? Will these differences affect the long-term success of therapy?

Finally, there are also questions that need answering about the use of efavirenz (and nevirapine) in Latin America and Asia. While it is possible that the Hispanics who had higher EFV levels in Ribaudo’s study came from countries where the slave trade was once active and have some African genes, no data have yet been presented on the effect of Asian ethnicity/genetics on NNRTI metabolism.

However, this is just speculation based on intriguing - but preliminary - data. The authors of the latter two papers told a press conference following their presentations that more research was needed before any recommendations, or changes to current therapies, could be made. However, since current British HIV Association guidelines recommend that therapeutic drug monitoring "may have a role in reducing toxicity in individual patients and in adjusting doses in those with ...extremes of body weight", perhaps this technology could be used in UK clinics whenever NNRTI-based regimes are prescribed to Africans and/or people with low body weight. Additionally, when stopping an NNRTI, perhaps drug-switching strategies should be preferred to stopping the NNRTI seven or more days earlier.

Further information on this website

Genes affect drug levels, but genetic tests far away - news story

Efavirenz blood levels: monitoring useful - news story