|
|
|
|
|
|
Session 70
Poster Presentations Resistance Testing: Methodology and Clinical Applications Session Day and Time: Tuesday 1:30 - 3:30 pm Room: Hall A |
Background: PhenoSense is an accurate and reproducible assay for
measurement of HIV drug susceptibility. For some drugs, especially stavudine,
didanosine, and tenofovir, reductions in susceptibility that are below 2-fold
are clinically relevant. Thus, appropriate interpretation of phenotypic test
results should take into consideration the natural variability in
susceptibility of wild-type (WT) viruses and inherent assay variability.
Methods: The ViroLogic phenotype-genotype database (n >12,000)
was queried for clinical samples lacking any drug-selected mutation in RT
(certain amino acids at positions 41, 65, 67, 69, 70, 74, 75, 100, 101, 103,
106, 151, 181, 184, 188, 190, 210, 215, 219, 225, 227, 236) or protease
(positions 23, 24, 30, 32, 33, 46, 47, 48, 50, 54, 82, 84, 88, 90). The upper
end of each wild-type virus fold change (FC) distribution was defined by the
mean plus 2 SD and the 99th percentiles of the distribution. To
assess assay variability for each drug, 10 reference viruses with varying drug
susceptibilities were tested 8 to 30 times each, and the average coefficient of
variation (CV) across samples was calculated for each drug.
Results: The
average CVs for the reference viruses ranged from 12%–23% for all drugs except
ZDV (36%). The log10-transformed FC data from WT clinical isolates
were normally distributed. Results are summarized below:
RTI
|
ZDV
|
3TC
|
ddI
|
ddC
|
d4T
|
ABC
|
TDF
|
NVP
|
DLV
|
EFV
|
|
Median |
0.89 |
1.01 |
0.99 |
0.97 |
0.95 |
0.90 |
0.86 |
0.89 |
1.25 |
0.86 |
|
Mean + 2SD |
1.80 |
1.59 |
1.35 |
1.40 |
1.34 |
1.36 |
1.19 |
3.18 |
4.69 |
2.25 |
|
99th
percentile |
2.08 |
1.74 |
1.50 |
1.56 |
1.48 |
1.53 |
1.26 |
4.58 |
7.03 |
2.73 |
|
Number |
2,262 |
2,262 |
2,256 |
1,991 |
2,262 |
2,166 |
1,420 |
2,261 |
2,227 |
2,254 |
|
|
|
|
|
|
|
|
|
|
|
|
PI
|
APV
|
IDV
|
NFV
|
RTV
|
SQV
|
LPV
|
ATV*
|
|
|
|
|
Median |
0.70 |
0.78 |
1.04 |
0.82 |
0.70 |
0.69 |
0.70 |
|
|
|
|
Mean + 2SD |
1.92 |
1.88 |
3.09 |
2.29 |
1.54 |
1.58 |
1.90 |
|
|
|
|
99th
percentile |
2.08 |
2.25 |
3.78 |
2.69 |
1.77 |
1.67 |
2.33 |
|
|
|
|
Number |
2175 |
2262 |
2261 |
2258 |
2260 |
1242 |
284 |
|
|
|
*ATV = atazanavir
Conclusions: The mean + 2SD FC value for all drugs in WT viruses is
lower than previously described for “biological” cut-offs using other
phenotypic assays. The larger variability in WT drug susceptibility for some
drugs (e.g., NFV and the NNRTIs) is not due to differences in assay variation,
since the CVs are low and not significantly different between drugs. These
observations indicate that the PhenoSense assay can be used to reliably
determine clinically relevant breakpoints at low fold-change values (<
2-fold) for stavudine, didanosine, and tenofovir. In addition, greater
precision in phenotypic data facilitates the development of more accurate
genotype interpretation algorithms.