There is a large genetic variability in HIV-1. To date, nine subtypes (A–D, F–H, J and K), 21 circulating recombinant forms (CRFs) and multiple unique recombinant forms (URFs) have been recognised [–3]. CRFs result from the recombination of two or more different subtypes. To be classified as a CRF, a virus strain must be detected in at least three unrelated individuals and capable of establishing an epidemic on its own; they are referred to by their number – defined according to the order of their discovery – and the subtypes involved (e.g. CRF02_AG or CRF03_AB) or by their number and the letters `cpx' (for complex), when more than two subtypes are involved (e.g. CRF04_cpx or CRF06_cpx). URFs are identified in a single individual or in a cluster of associated individuals and remain a localised phenomenon. New recombinant forms are continually being identified as a result of the co-circulation of multiple subtypes within a given epidemic (e.g. CRF05_BF in Brazil and Argentina, or CRF14_BG in Portugal).
One obvious consequence of the genetic diversity of HIV-1 is the potential impact on the efficacy of a future vaccine. Less obvious and largely controversial is the impact of genetic diversity on disease progression, vertical transmission, response to antiretroviral therapy and drug-resistance pathways.
Almost everything we know about HIV-1 is the result of studies on populations infected with subtype B; this is the commonest variant in the USA and Western Europe, but representing less than 15% of HIV-1 infections worldwide. This knowledge from studies on B subtype should not be directly extrapolated to all subtypes, given the significant sequence differences observed in both the structural and regulatory genes of HIV-1 subtypes, which can influence the biological properties of the virus. Unfortunately, research on subtypes other than B is generally under-funded, resulting in small, poorly designed and mainly observational and retrospective studies, showing controversial – and sometimes conflicting – results.
In some studies, all non-B subtypes have been pooled into one group, which is then compared with subtype B; however, `non-B subtypes' is not an entity, just a convenient designation. As a result, a `dilution effect' will occur: if there is a difference between subtype B and one of the non-B subtypes, it will be `diluted' within the overall results of the non-B group. Another common problem is related to the comparison of sequences from a non-B subtype to a reference subtype B consensus sequence, which is a frequent occurrence in studies on drug resistance and response to therapy. As a consequence, some mutations that occur almost exclusively in some non-B subtypes are not detected, driving these studies to misleading conclusions.
Copyright © 2006, Mediscript.