Personalized Medicine for HLA-associated Drug-hypersensitivity Reactions

Mandvi Bharadwaj; Patricia Illing; Lyudmila Kostenko


Personalized Medicine. 2010;7(5):495-516. 

In This Article

HLA Screening & Personalized Treatment in HLA-associated Drug Hypersensitivity

It is evident from this article that there are many hurdles that must be overcome in determining the utility of HLA screening in the prevention of drug hypersensitivity, from the discovery and determination of associations, to the implementation of screening. Furthermore, there is a clear necessity for each association to be considered individually.

At the outset it is imperative that the hypersensitivity has a clearly defined phenotype that can be diagnosed by clinicians. For example, true AHS involves a panel of symptoms, including fever, rash, nausea, vomiting, abdominal pain, lethargy and malaise, as well as a positive patch test result.[33] Without the final stipulation the strength of the association of AHS with HLA-B*5701 was initially hidden, due to over-diagnosis. It is also by differentiating between the different carbamazepine-induced hypersensitivities, that we see the strength of the association between HLA-B*1502 and carbamazepine–SJS/TEN,[20] and it may be that only by very clear delineation of nevirapine hypersensitivities that any sense can be made of the plethora of HLA associations suggested. As phenotypically different hypersensitivity reactions may involve different causative mechanisms, until they can be clearly distinguished, the discovery of the conditions necessary for disease manifestation, genetic or otherwise, will be impaired.

In the next phase, it is then necessary to estimate both the positive and negative predictive value of the test across large sample sizes and different genetic backgrounds. Population studies such as the SHAPE study are key to this process, providing data not only on the sensitivity of screening, but also the specificity, thereby providing an indication of the cases of hypersensitivity prevented compared with those inappropriately denied treatment.[33,103] Without such data it is impossible to judge the proportion of individuals positive for the associated HLA, that could tolerate the drug, and hence, difficult to assess how many individuals would be inappropriately denied the drug. For example, if implementation of testing were to be judged on the prevalence of the HLA allele in the hypersensitive cohort alone, inappropriate denial of drugs such as flucloxacillin would potentially occur a thousand-times for every incidence of hypersensitivity prevented.[60] As such, low incidence of hypersensitivity in individuals with the risk-associated allele may render testing useful as a diagnostic tool alone.

Furthermore, if these studies crossover several populations, with varied genetic background and allele frequencies, the widespread applicability of the test can be judged. This also provides insight into the involvement of the underlying genetic background, which appears to be of importance in carbamazepine–SJS/TEN.[46]

Once these factors have been determined it becomes a balancing act. Questions must then be asked such as; does inappropriate denial of the drug from certain individuals impose a lesser toll than the hypersensitivity reactions prevented by the screening? And is screening for an allele that is weakly associated with the hypersensitivity worthwhile? The answer to these questions will depend on the seriousness of the initial disease being treated and the availability of alternative treatments, compared with the seriousness and incidence rate of the hypersensitivity. Further questions about the nature of the cost of testing versus the cost of managing the hypersensitivity reactions must also be asked, and judgements on the populations in which testing might be useful must be made.

Only after all of these questions have been answered can it truly be judged whether HLA typing is a good candidate as a pharmacogenetic test for prevention of a specific ADR.