Predicting clinical responses to ddI from genotypic resistance
Simon Collins, HIV i-Base
Shulman and colleagues addressed the difficulty of interpreting resistance to ddI, by analysing stored samples from 444 treatment-experienced patients in three ACTG studies (359, 368 and 398) to test the results against responses to ddI monotherapy in ACTG 175 and ACTG 309.
Paired genotype-phenotype samples were compared to consensus RT and compared with treatment response. 16/444 had >4 fold change (all with either Q151M or substitutions at 69) and these samples were excused from the analysis.
In the remaining 428 cases, 20 codons with P<0.01 in the univariate analysis were used in stepwise regression modeling. This resulted in nine codons in the final multivariate model: 184, 215, 43, 67, 74, 223, 210, and 75 were associated with increased resistance (all p<0.001) while 83 was associated with decreased resistance.
The predictability of these mutations was tested against 52 genotype and 46 phenotype samples from ACTG 175 and ACTG 307. The resulting resistance score (total score = 184 + 215 + 43 + 67 + 74 + 223 + 210 + 75 83) had a strong correlation between with ddI-fold sensitivity change (Spearman r=0.67 (P<0.0001).
This fitted more accurately than the Jaguar score (M41L + L74V + D67N + K219Q/E + T215Y/F + T69D K70R M184V) which is based on 4 week virological response to ddI in patients with a relatively high CD4 count, and correlated with fold-change, r=0.38, P<0.0001.
The low phenotypic change associated with loss of activity with ddI makes predicting response from genotypic results a difficult challenge, for which there is certainly a clinical need. An important suggestion from the audience, similar to other studies, was that cooperation between the researchers in ths study and those involved Jaguar should usefully now cross-validate each set of results.
Shulman N, Bosch R, Fiscus S et al. Mutations associated with didanosine resistance determined from 444 matched genotype-phenotype pairs. 14th International HIV Drug Resistance Workshop (14th IHDRW), 7-11 June 2005, Quebec City, Canada. Abstract 49. Antiviral Therapy 2005; 10:S54