This presentation by Rick Pesano, who just joined Virologic Inc., provided a nice overview of HIV resistance assays. Dr. Pesano also discussed issues related to genotype-phenotype discordance, and he provided some updated information about new mutations that may become important, as well as definitions for the loss of susceptibility to atazanavir (ATV, Reyataz) and tipranavir (TPV, Aptivus).

Dr. Pesano began his presentation by reviewing the technology that is required to produce a genotype report (i.e., producing a DNA sequence of the protease [PR] gene and the first two thirds of the reverse transcriptase [RT] gene, and interpreting that sequence using rules-based algorithms).

He then briefly reviewed the technology behind the PhenoSense recombinant viral assay, which uses the patient's PR and RT genes in a stable background HIV virus in order to measure the susceptibility of the constructed virus to HIV medications. As of July 2005, Virologic had in its database 162,000 phenotypes -- more than 90,000 sequences, of which more than 54,000 are linked genotype-phenotypes.

In addition, Virologic also has 37,000 replication capacity test results in its database. The replication capacity assay measures the constructed patient virus's ability to grow in a controlled environment in the absence of antiretroviral drugs. Replication capacity is usually measured as the percentage of wild-type control virus that is also grown at the same time.

Dr. Pesano also brought up the problem of assay discordance, which is simply when a phenotype test indicates that a virus was still susceptible to an antiretroviral drug, while a genotype test scores the virus as resistant to that drug. This situation can often be the result of a patient carrying a mixture of viral strains, in which case a genotype might be somewhat more sensitive in detecting the mixture, whereas a phenotype cannot. Other times, mixtures can be absent, but discordance between assay types can still exist. This could be the result of resensitization (i.e., RT mutations like K65R, L74V, L100I, Y181C and M184V, which all cause primary drug resistance to one drug but may cause other drugs that were resistant to regain some of their activity).

A reverse discordance can also appear. That is when the phenotype scores a drug as resistant while the genotype scores the drug as susceptible. This could be a result of cross resistance. For example, in the case of lamivudine (3TC, Epivir) and abacavir (ABC, Ziagen) cross resistance, a phenotype would score both as resistant, but, depending on the rules, a genotype might score one as susceptible -- incomplete rules that might result in under-calling genotypic resistance (as can be the case with tenofovir [TDF, Viread], didanosine [ddI, Videx] or amprenavir [APV, Agenerase]). Reverse discordance could also be a result of new substitutions at known codons that would not necessarily be scored as resistant by a genotype algorithm (i.e., RT- K65N, M184T and PR-V82C).

Another important observation is that some protease inhibitors (PIs) may demonstrate reduced susceptibility in a phenotype assay, although no primary PR mutations are present. How could this happen? This could be the result of new mutations in the PR gene (I13V, M36IV, I93L), which can decrease nelfinavir (NFV, Viracept) susceptibility or mutations in the gag gene region (418ER, 436R, 437V). An example mentioned by Dr. Pesano involved a 93L mutation and a 431V mutation (in the absence of other primary PI mutations) that demonstrated reduced nelfinavir susceptibility.

Dr. Pesano also discussed new susceptibility reporting findings for two of the newest HIV drugs, atazanavir and tipranavir. Based on data from Bristol-Myers Squibb (BMS) 043 (PI-naive) and 045 (PI-experienced) studies, and over 7,000 isolates tested, the atazanavir phenotype cutoff for unboosted atazanavir is 2.2 and for ritonavir-boosted atazanavir it is 5.2. So, fold-change values above these numbers would be considered a loss of susceptibility.

Dr. Pesano listed the 14 relevant PR mutations important for atazanavir resistance. If a virus has 3 or less mutations, it will likely be susceptible to unboosted atazanavir; if a virus has 5 or less mutations, it will likely be susceptible to boosted atazanavir. However, this is not a hard and fast rule. Depending on which specific mutations are present, more or less mutations may indicate loss of susceptibility or retained activity. There is a broad distribution of activity based on this mutational score and some discordant samples (loss of phenotypic activity despite 3 or less PR mutations) were found. Thus, Virologic is refining these rules to define loss of susceptibility.

In addition, there are new cutoffs for tipranavir. Based on more than 570 isolates, the median cutoff for unboosted tipranavir is 2.1 and for boosted tipranavir it is 4. Since this drug will never be given as an unboosted PI, a fold change cutoff of 4 is the one important value to remember.

Dr. Pesano then presented the same scheme, where it was found that 4 or less PR mutations correlated with a fold change of less than 4. Again, there was a broad distribution of activity based on this mutational score and some discordant samples (loss of phenotypic activity despite 4 or less PR mutations) were found.

In terms of the replication capacity assay, Dr. Pesano mentioned that even with wild-type virus there was a broad distribution of replication capacity. However, in general, a decreased replication capacity is associated with an increasing number of PR or RT mutations.

He pointed out that certain mutation combinations can result in profound reductions in replication capacity. The replication capacity assay may have some utility in newly diagnosed patients, for whom a decreased replication capacity may indicate a slower rate of clinical progression. In heavily treated patients with multi-drug resistant virus, a markedly reduced replication capacity might suggest that practitioners would not need to change regimens right a way. However, the exact clinical indication for replication capacity assay utilization remains to be defined.