Background to drug resistance

The genes of living organisms contain the information necessary for their growth and survival. These genes are carried as sequences of chemical �bases� (the chemical �letters� of the genetic code) in strands of deoxyribonucleic acid (DNA) in the chromosomes.

Proteins are the workhorses of living systems, forming structural components of cells and tissues, or acting as enzymes (biological catalysts) to drive essential reactions in the body. Each protein is specified (�encoded�) by a gene, the information in which is read end to end by using the genetic code to join amino acids in a chain. The amino acid sequence of a protein determines its three-dimensional structure, for example to create the �active site� of an enzyme - the space into which the molecules processed by the enzyme will fit. Inhibitors can often be made which will fit into an active site and block the enzyme, allowing their use as drugs. Mutations in the gene (changes in one or more of the bases) may lead to the incorporation of wrong amino acids, distorting the active site and crippling the enzyme, or preventing the binding of an inhibitory drug while leaving the enzyme active, so that therapy fails.

Viruses are not living organisms, but depend on infection of cells for their survival and multiplication. Consequently viruses must play the game by the same rules, albeit sometimes with variations such as the use of ribonucleic acid (RNA) rather than DNA for their main genetic material. Human immunodeficiency virus (HIV) particles contain RNA, which on infection of a cell (in this case in the immune system) must be copied first of all to DNA in order to take over control of the cell�s processes for the production of new virus particles. This copying is performed by the virus enzyme reverse transcriptase (RT) which was the initial (and continuing) target of some anti-HIV drugs such as AZT. Roche, and later other companies, attacked a second target through the powerful inhibition of protease � a viral enzyme which cuts up newly-made viral proteins to allow the formation of infectious particles. Because of their different targets, it has become commonplace to use inhibitors of these two enzymes in combination therapy, resulting in much-improved suppression of the �virus load� (circulating concentration of virus) to levels below the current limits of detection.

Though it is a very vigorous enzyme, RT is unable to correct frequent errors in copying the RNA, creating large numbers of viral variants, some of which cannot survive. Those that do, however, generate the highly mixed population of subtypes(�quasi-species�) which are present in an infected person. This gives the virus a potential head start over any change in its environment, such as the introduction of antiviral drugs, since a variant may already be present which will survive the new conditions better than the typical virus (�wild type�) of the original population, and can take over under the �selective pressure� of the drug. As a result, drugs such as reverse transcriptase inhibitors and protease inhibitors will tend to select for variants in their respective target enzymes, which can be detected as mutations in either gene. In combination therapy, each gene may change independently. Resistance to an RT inhibitor will have no direct effect on a protease inhibitor and vice versa, though it will reduce the potency of a drug combination. Within each class of compound, the mutations which bring about resistance to a particular drug depend on the chemical shape of that drug and its interaction with the enzyme. To a greater or lesser extent, therefore, compounds will have a �genetic signature� of resistance mutations, and depending on the similarity of their shape some compounds (eg indinavir and ritonavir) will share mutations and therefore cross-resistance. Saquinavir has a distinctive signature, marked by changes at protease amino acids 48 (�G48V�) or 90 (�L90M�), the former being the less common of the two in the clinic, using either the original, hard gelatin formulation (Inverase^�) or the soft gelatin formulation (Fortovase^�).

Though they may enable some growth in the presence of a selecting drug (�resistance� to that drug), mutations may also lower the efficiency of the enzyme and slow the multiplication of the virus. This is true of G48V and L90M, which is why they are virtually absent from untreated populations and (particularly G48V and the double mutation) arise late and infrequently during saquinavir treatment. Following the appearance of mutations to drug resistance, enzymes (hence, their genes) will undergo further selection for changes which will correct the distortion and so improve their efficiency and viral �fitness� � for example, changes at amino acids 63 and 82 tend to associate with G48V. In the case of protease, changes to improve fitness may also occur at the specific points in the virus� structural (�gag� and �pol�) proteins which are cleaved by the protease, to tailor them to the altered enzyme.

Testing in the clinic

Resistance in viral isolates obtained during treatment may be defined by �genotyping� (looking for particular mutations associated with relevant drugs � i.e. �probe-based genotyping� - or analysing gene sequences for all possibly relevant changes � i.e. �sequence-based genotyping�) or by �phenotyping� (looking for what may be defined as a significant change in viral drug sensitivity in tissue-culture). These are still matters for further research and clarification, particularly as they apply in a clinical context. Unfortunately, �fitness� changes have often been labelled �resistance mutations�, and the definition of phenotypic change is variable, adding to the complexity of an already difficult area. This confusion has hindered the wider understanding of HIV drug resistance and its role in progressing disease, and most importantly decision-making on the best application of the drugs available to the treating physician � which drug(s) to use, and when and how to change therapy in the best interests of the subject and limitations in the drugs budget, especially for multiple therapy. A precedent has been set for the use of single mutations to �define� drug resistance as a basis for clinical management, without taking account of other mutations, patient history and ongoing circumstances (viral load, CD4 count, etc.).

Roche collaborative surveillance programme

Roche has embarked on a strategy to apply genotyping in conjunction with new, customisable software for the interpretation of genetic changes in viral isolates, which will allow powerful analysis of resistance trends and offer appropriate guidance on clinical management.

The GREAT (Genotypic Resistance Evaluation to Aid Therapy-switching) Study, in collaboration with Virology Networks of Utrecht and Perkin-Elmer Corporation of Norwalk, Connecticut, will test the prototype RetroGram� Decision Support Software in realistic settings which allow a variety of antiviral treatment regimens. RetroGram� employs an electronic rule-based algorithm to interpret genotype data and guide physicians and patients in treatment decisions, and will be constantly updated as new information emerges on the type and clinical significance of resistance mutations in HIV genes.

GREAT is a randomized, international, parallel group, open label, 48-week trial enrolling up to 360 patients who are currently failing their first-line protease inhibitor anti-HIV combination regimens � failure being defined by at least 24 consecutive weeks' experience with a PI-containing combination regimen and a viral load of at least 5,000 copies/ml (or 3.7 log¹⁰) at screening. Participants will be randomized to a new regimen by one of two methods: either best clinical judgment or best clinical judgment in conjunction with real-time HIV resistance genotyping. This will provide a clinically useful tool to physicians and patients, and will also be an advance in rational combination drug regimen design, intended to maximize individuals' therapeutic options over the long term.

While Roche is a major GREAT sponsor, the trial has been designed to include the entire armamentarium of approved HIV treatments from a wide array of manufacturers. Roche has purposely left data analysis in the hands of Virology Networks, the investigators and community advocates involved in the trial. This will ensure that the GREAT data are thorough, accurate and impartial and can be applied to the broadest range of patients in the real world.