Is genetic screening for abacavir hypersensitivity risk appropriate for black Africans?

Routine testing for the HLA B*5701 gene – which is thought to be associated with an increased risk of abacavir hypersensitivity reaction (HSR) – is unhelpful in black African populations and may be inappropriately denying abacavir to them, according to a commentary by two London HIV physicians which was published in STI Online last month. In the absence of ethnic-specific data regarding the reliability and utility of B*5701 testing, the commentary suggests that clinicians should prescribe abacavir to their black African patients without testing for B*5701, whilst remaining vigilant for symptoms suggestive of abacavir HSR.

Recent studies from Australia and Brighton have suggested that screening for the B*5701 gene may be a useful tool to help prevent the hypersensitivity reaction to the nucleoside analogue drug, abacavir (Ziagen, also found in Kivexa/Epzicom and Trizivir).

However, as lead author Dr Tariq Sadiq – Senior Lecturer HIV/GU Medicine at St George’s Hospital in London – points out, most patients screened in these non-randomised prospective studies were Caucasians, and “associations between B*5701 and HSR have not been adequately investigated in black Africans who represent a substantial proportion of HIV-1 infected patients in the UK.”

The commentary argues that there are several problems associated with screening black African patients for B*5701. “The term ‘black African’ itself is unhelpful,” Dr Sadiq tells aidsmap, because generally speaking, genetic diversity is at its greatest between African populations – i.e. different non-African populations are closer to each other, genetically, compared with differences between various black African ones.

“This,” he explains, “has the potential to greatly undermine the appropriateness of some genetic tests in various black African populations if these populations are not involved in the studies that lead to genetic discovery.”

Screening for B*5701 may end up being an example of this, he argues, noting that the association between B*5701 with HSR in black African populations is uncertain, and may not be accurate.

The commentary provides some calculations to illustrate the issues facing UK clinicians who have a substantial proportion of black African patients from Uganda or Zimbabwe. For example, 1371 Shona Zimbabweans would need to be screened to correctly identify one Shona Zimbabwean at risk of abacavir HSR while incorrectly identifying – and inappropriately denying abacavir use in – 4.5 Shona Zimbabwean patients.

Although their calculations “suggest possible utility in some [Kampalan] Ugandans (and then only if the association between B*5701 and HSR holds),” notes the commentary, “testing Ugandans as a whole may still be inappropriate given low B*5701 carriage among Nilotic and Sudanic populations in Kenya, ethnic groups found in significant proportions among Ugandans at our clinic.”

Additionally, “more sobering estimates are arrived at” when calculating the effectiveness of B*5701 screening of the same populations if the genetic test was less accurate in people of black African ethnicity – as it was found to be in African Americans in GSK’s CNA30032 study. It would require 168 Kampalan Ugandans and over 8000 Shona Zimbabweans to be screened in order to correctly identify one patient at risk of abacavir HSR, and four Kampalan Ugandans and 31 Shona Zimbabweans would be inappropriately denied abacavir due to a false positive result.

“Denying abacavir to the individual becomes important if the alternatives might be unsafe or non-effective,” Dr Sadiq tells aidsmap. “The numbers incorrectly denied abacavir in a screening programme for black Africans may seem relatively small but at the individual level the chance of a positive test being accurate for susceptibility to HSR may be very low. In such a situation where abacavir might be particularly desirable (for example, the avoidance of serious long term side-effects with AZT or exacerbation of underlying renal disease with tenofovir), we would argue that abacavir should be prescribed with appropriate vigilance without testing for B*5701.”

He argues that using the B*5701 test when it is not yet proven to effectively screen for abacavir HSR amongst black African patients “is really not helpful. Interpreting a negative B*5701 test as meaning there is no possibility of HSR may be potentially disastrous, since there may be other undiscovered factors involved, particularly if there is truly a different form of hypersensitivity in black Africans as has been suggested elsewhere. More science on this manifestation in black Africans is obviously required.”

Dr Sadiq also points out a further issue with the utility of the B*5701 test as used in some UK clinics. The test is, in fact, not specific enough to rule out similar polymorphisms – in particular B*5703 – that are relatively more common in black Africans, but which are not associated with abacavir HSR. Although the Brighton study team originally reported that four black Africans tested positive for B*5701 – resulting in a surprisingly high 5% prevalence – subsequent retesting using tests that are B*5701-specific found that three of the four individuals had the B*5703 polymorphism, and not B*5701. This resulted in a more realistic 1% prevalence.

He tells aidsmap that B*5701 testing utilising sequence specific primers (SSP) that are specific for B*5701 are “now available and being used by some providers in the UK, and these would certainly be cheaper than previously confirming non-specific tests by full sequencing. In addition cell sorter analysis for B*57 (like a CD4 count) has been developed and may even be developed specifically for B*5701 in the future.”

The issue of cheap and accurate testing, as well as including all relevant ethnicities in clinical trials to ascertain the reliability and utility of these tests, is particularly pertinent to the scaling-up of use of abacavir as a component of second-line therapy in sub-Saharan Africa. Although a large study sponsored by GlaxoSmithKline (GSK) is assessing the clinical utility of B*5701 screening amongst Europeans, and another GSK-sponsored study will help better characterise abacavir HSR signs, symptoms and prevalence in African Americans, this may not be relevant to genetically diverse sub-Sarahan African populations.

“The crux of the problem is that there simply are not enough data at this stage to use the test appropriately in black Africans, and introducing screening in sub-Saharan Africa may cause more harm than good if introduced prematurely,” says Dr Sadiq.

“It is very possible we will have a situation where in some African populations the test is useful and in others it is not. However,” he notes, “this need not be that challenging to medicine if we are able to accept and use genetically more appropriate definitions of race.”