virco®TYPE HIV-1 References
Latest Clinical Cut-Off Values: Virco's Clinical Database has grown considerably between 2006 and 2008, as is evidenced by the increase in the number of Development and Validation data-set Treatment Change Episodes (TCEs) in Figure 1 below, allowing us to capture the most up-to-date relationship between baseline resistance and virological outcome for each drug.
The latest CCOs are based on the Development data-set of 6550 TCEs and the Validation data-set of 2299 TCEs. A TCE is defined as a Patient Record Including:
- Baseline viral load (VL) and genotype before switch (Genotype used to conduct virco®TYPE HIV-1 resistance analysis)
- Post switch regimen
- Post switch VL at 8 weeks
For two of the most recently introduced antiretrovirals, Tipranavir and most recently, Etravirine, CCOs were developed on smaller data-sets of highly experienced patients. The virco®TYPE HIV-1 report contains a cautionary note that the relevance of these CCOs for patients different from these study participants has not been evaluated.
Data for the development of virco®TYPE HIV-1 Clinical Cut-Offs for tipranavir/r were obtained from the patients who participated in the RESIST studies, and for Etravirine from the patients who participated in the Phase IIb and DUET (Phase III) studies.
The table below lists all currently available vircoTYPE HIV-1 CCO values ( status Jan 2008)and the predicted Fold Change of Wild Type viruses. The CCO were validated by a number of different methodologies, including the use of unseen data and numerous rounds of bootstrapping (random re-sampling with replacement), which resulted in the listed 90% confidence intervals around the cut-offs
Handling of Genotypic Mixtures
Genotyping may detect the presence of either single amino acids or mixtures at any position. The handling of mixtures by the virtualPhenotyp-LM bioinformatics engine was recently revised. Previously (version 4.0), the RWF (Resistance Weight Factor) for a given position was calculated by weighting the RWF of each amino acid observed based on the number of residues detected in the mixture.
For example, if an L/T/F mixture were detected, the RWF would have been calculated by adding 1/3 of the RWF each for L, for T and for F. Under the revised method (version 4.2), the RWF of the mutation conferring the most resistance is taken as the value for that position. This method is more conservative and in theory should also be more predictive of treatment response.
The improved performance would occur because the amino acid which confers the most resistance to a drug is also the most likely to be selected from those present in the mixture if the drug is included in a subsequent treatment regimen.
The following real-life example for tipranavir illustrates how mixture are treated by the vircoTYPE HIV-1 system
The column "LM Mutations and Pairs" below shows the mutations and mutation pairs included in the Linear Model prediction of the tipranavir FoldChange (FC). Where there is a mixture, e.g. in the case of 73wt/A/S/T, only 73T is taken into account in predicting the FC, as it is the mutation at position 73 which is most associated with tipranavir resistance.
Mutations (or mutation pairs) with a negative RWF in the column "RWF (logFC)" below have a reducing or resensitizing effect on resistance to tipranavir, as in the case the (non mixture) of 41K. Mutation 54L has a Mixture Contribution of 0 because it's negative or resensitizing RWF means the wild-type actually has a Mixture contribution of 1 in this case; similarly, the combination of 54L and 73T has a Mixture Contribution of 0.
