Transcript Genetic correlates of phenotypic hypersusceptibility to efavirenz

#___
Genetic Correlates of Phenotypic Hypersusceptibility to Efavirenz (EFV)
Among 444 Baseline Isolates from Five ACTG Studies
N Shulman*, R Bosch, J Mellors, M Albrecht, and D Katzenstein DACS 217 Study Team, NIAID Sponsored AIDS Clinical Trials Group, Bethesda, MD
BACKGROUND
Increased phenotypic susceptibility
(hypersusceptibility, HS) to NNRTIs is
observed in ~30% of viral isolates with NRTIresistance mutations1 and has been
associated with improved virologic response
to NNRTI-based therapy in several studies2-5.
Although preliminary analyses have shown an
association between nucleoside resistance
mutations and NNRTI HS1,2, the genetic basis
for NNRTI HS has not been thoroughly
defined. Elucidating the genetic basis for
NNRTI HS will improve interpretation of
genotypes to optimize use of NNRTIcontaining treatment regimens.
OBJECTIVES
In this analysis, we sought to identify specific
mutations in the reverse transcriptase
associated with efavirenz (EFV) HS and their
relative importance.
METHODS
Paired baseline genotypes (ABI) or (VGI) and
phenotypes (ViroLogic) were obtained from
444 subjects entering one of five completed
ACTG studies: 290, 359, 364, 370, or 398.
All subjects were NRTI-experienced but
NNRTI naïve at study entry. Two subjects
included at the time of the earlier abstract
analysis were excluded due to discovery that
they were NNRTI experienced.
Fisher’s exact tests, recursive partitioning
(Classification and Regression Trees, CART),
and logistic regression were used to identify
reverse transcriptase (RT) mutations
associated with EFV HS and evaluate their
relative importance. Analyses were
performed using S and SAS software.
EFV hypersusceptibility (HS) was defined as
< 0.4 fold-change in susceptibility relative to
the wild type control. A mutation was defined
as any amino acid that differed from the Los
Alamos HIV-1 consensus B sequence. A
mixture of mutant and wild-type was
classified as mutant, even if the mutant was
present in a minority population.
*Nancy Shulman, MD
Stanford University Medical Center
Division of Infectious Diseases
S-156
300 Pasteur Drive
Stanford, CA 94305
Email: [email protected]
RESULTS
Stepwise Logistic Regression
Prevalence
The prevalence of EFV HS was similar across
contributing ACTG studies:
Study
290
359
364
370
398
Total
EFV HS/n
2/26
50/136
48/129
25/78
28/75
153/444
%
8
37
37
32
37
34
For each subject, each RT position 1-440 was compared
to HIV-1 clade B consensus and a binary indicator of
consensus vs. non-consensus was determined. Each
indicator was then compared to EFV HS or non-HS in
2x2 tables. There were 30 mutations plus the 69 insert
that had p-values of <0.1. In order of significance:
Codon
P-value
Codon
P-value
69
121
201
178
297
172
245
103
102
219
324
60
317
162
6
0.0080
0.014
0.014
0.034
0.035
0.040
0.043
0.053
0.054
0.054
0.072
0.078
0.079
0.087
0.092
2151
4.9 x 10-15
41
7.1 x 10-13
210
2.5 x 10-10
118
2.2 x 10-7
208
3.2 x 10-7
67
5.4 x 10-7
43
4.7 x 10-6
44
9.8 x 10-5
211 1.2 x 10-3 (0.0012)
203
0.0012
138
0.0019
228
0.0026
69 ins
0.0030
3222
0.0033
223
0.0045
196
0.0061
1215Y
The above 26 RT positions were evaluated in a
stepwise logistic regression analysis, using a
threshold of p<0.01 to include the position in the
model. Three codons, 215, 208, and 118 were
independently associated with EFV HS.
RT Codon
215
208
118
Univariate – Fisher’s Exact
p=1.7 x 10-11, 215F p=0.13
2The
five codons in red were excluded from further
multivariate analyses due to incomplete data (>200
missing data points for each).
We looked for specific polymorphisms in the NNRTI
binding pocket that may affect EFV susceptibility:
NNRTI polymorphisms associated with EFV HS:
K103R
p=0.053
V179I
p=0.029
No difference: A98S, K101Q/R, V106I
Absent in EFV HS (but low frequencies overall):
A98G, L100I, K101E, V179D.
CART Tree
Odds Ratio
3.94
5.30
2.40
95% CI
(2.48,6.37)
(2.04,16.6)
(1.38, 4.23)
p-value
<0.001
<0.001
0.002
Interaction terms between these 3 codons were not
significant (p>0.19), consistent with the interpretation
that a mutations at one of these codons confers
similar risk of EFV HS, regardless of whether the
other two codons are mutant or not.
Because odds ratios tend to overestimate risk ratios
for non-rare events6, the same model was evaluated
in a log-binomial regression model which yields
relative risk estimates:
RT Codon
215
208
118
Risk Ratio
2.77
1.52
1.35
95% CI
(1.99, 3.98)
(1.16, 1.97)
(1.04, 1.74)
p-value
<0.001
0.004
0.024
Covariate space of 215, 208, 118:
215
0
0
0
Yes
0
Yes
Yes
Yes
208
0
0
Yes
0
Yes
0
Yes
Yes
118
0
Yes
0
0
Yes
Yes
0
Yes
n
199
11
1
152
0
55
17
9
% EFV HS for significant mutations alone and in combination
78% (119/153) of the EFV HS isolates had a mutation at 215.
Mutation(s)
215
41
210
118
208
215 & 118
215 & 211
215 & 208
EFV HS/n % HS
119/233
51%
101/189
53%
70/119
59%
46/75
61%
22/27
81%
44/64
69%
89/148
60%
22/26
85%
CONCLUSIONS
• Efavirenz hypersusceptibility was present in 34% of isolates and had a consistent prevalence across studies.
• Hypersusceptibility was most associated with the presence of amino acid changes at RT codons 215, 208, and 118. These
codons were among the most significant in the univariate analysis, were the first three branches of the CART and were
independently associated with EFV hypersusceptibility in the stepwise model.
• Of 26 isolates containing mutations at both 215 and 208, 22 (85%) were HS to EFV.
• Site-directed mutagenesis studies are underway to validate the importance of these mutations in EFV HS.
• These results can be used to identify the biochemical and structural basis for EFV HS and to predict its presence in clinical
samples.
REFERENCES
1. Whitcomb J, Huang W, Limoli K, et al. Hypersusceptibility to non-nucleoside reverse transcriptase inhibitors in HIV-1: clinical, phenotypic and genotypic correlates. AIDS 2002; 16:F41-7.
2. Shulman NS, Zolopa AR, Shafer RW, et al. Phenotypic Hyper-susceptibility of Non-Nucleoside Reverse Transcriptase Inhibitors in Treatment Experienced HIV-infected Patients: Impact on
Virological Response to Efavirenz-Based Therapy. AIDS 2001; 15:1125-1132.
3. Haubrich R, Kemper C, Hellmann N, et al. The clinical relevance of non-nucleoside reverse transcriptase inhibitor hypersusceptibility: a prospective cohort analysis. AIDS 2002; 16:F33-40.
4. Hammer S, Vaida F, Bennett K, et al. Dual vs. single protease inhibitor therapy following antiretroviral treatment failure: a randomized trial. JAMA 2002; 288:169-80.
5. Katzenstein DA, Bosch RJ, Hellmann N et al. Phenotypic susceptibility and virologic outcome in nucleoside-experienced patients receiving three or four antiretroviral drugs. AIDS, 2003;17:821830.
6. Skov T, Deddens J, Petersen M, Endahl L. Prevalence proportion ratios: estimation and hypothesis testing. Int J Epidemiol 1998; 27:91-5.