Transcript luis felipe gonzalez

Impact on drug resistance of non-B HIV-1
subtypes circulating in Brazil
AIDS Conference 2008, Mexico City
Rodrigo de Moraes Brindeiro, PhD
[email protected]
Laboratory of Molecular Virology
Federal Univ. of Rio de Janeiro –UFRJ
Brazil
HIV Subtype and Resistance
• Interactions between antiretroviral drugs and
their HIV target is dependent on its primary
sequence
• Consequently, differences in genetic sequence
may impact on these interactions
• Therefore HIV subtype variability has potential to
produce or influence drug resistance
• We need to consider the magnitude of this
phenomenon and its clinical relevance for each
drug and subtype
Global Infection with Different Subtypes
• Subtype C is the more prevalent
HIV variant in the globe.
•Subtype C
•Subtype A
•Subtype B
•Subtype G
•CRF01_AE
•CRF02_AG
•Subtype D
AIDS. 2006 Oct 24;20(16):W13-23
Subtype C spreading in Brazil is a monophyletic
group separated from worldwide C variants
Subtype Impact on Drug Resistance
Genetic Variability:
Different codon/amino acids at
various positions in RT and
protease
Selection of different
resistance pathways for
identical antiretroviral
drugs
Selection of different amino
acid substitutions at key
resistance positions
Subtype Impact on Drug Resistance
Genetic Variability:
Different codon/amino acids at
various positions in RT and
protease
Selection of different
resistance pathways for
identical antiretroviral
drugs
Selection of different amino
acid substitutions at key
resistance positions
≠ susceptibility to ARV in drug-naïve isolates
(different IC50% when compared to B prototype)
Subtype C hypersusceptibility to LPV
Subtype C hypersusceptibility to LPV
Protease resistance mutation at first failure
following nelfinavir or indinavir treatment
NFV
IDV
B
C
B
C
D30N
38%
6%
0
0
M36I
27%
98%
25%
100%
M46I
46%
6%
19%
5%
A71V
36%
0
19%
2%
V82A
8%
2%
25%
7%
V82I
0
10%
0
10%
I84V
0
2%
0
0%
L90M
9%
16%
6%
15%
AAC 2004 Jun;48(6):2159-65
Protease resistance mutation at first failure
following nelfinavir or indinavir treatment
NFV
IDV
B
C
B
C
D30N
38%
6%
0
0
M36I
27%
98%
25%
100%
M46I
46%
6%
19%
5%
A71V
36%
0
19%
2%
V82A
8%
2%
25%
7%
V82I
0
10%
0
10%
I84V
0
2%
0
0%
L90M
9%
16%
6%
15%
AAC 2004 Jun;48(6):2159-65
Subtype Impact on Drug Resistance
Genetic Variability:
Different codon/amino acids at
various positions in RT and
protease
Selection of different
resistance pathways for
identical antiretroviral
drugs
Replicative capacity
(RC)
Selection of different amino
acid substitutions at key
resistance positions
Genetic barrier: # of mutations
to aa substitution
HIV Protease
Mutations Positions 36, 89, 30
L89M can make a structural alteration together
with D30N, leading the virus carrying 89M/30N
not viable
30
36
89
HIV Drug Resistance Database, Los Alamos National Laboratory
http://resdb.lanl.gov/Resist_DB
Polymorphism at Protease Position 89 and
Selection of Resistance Pathway to Nelfinavir in
Subtypes B and C
Replication Capacity of in vitro
mutated protease
Subtype B
89L
Subtype C
89M
L90M
91.9%
78.9%
D30N
86.5%
No Growth
AAC. 2004 Sep;48(9):3552-5
1200
BWT
BL90M
BD30N+N88D
BD30N
1000
B
[P24](ng/ml)
800
600
400
200
0
0
2
4
6
8
6
8
culture days
1200
CWT
C L90M
CD30N+N88D
CD30N+M89L
CD30N+N83T
1000
C
[P24](ng/ml)
800
600
400
200
0
0
2
4
culture days
(B or C) 90M X (B or C+83T) 30N
CD30N (+N83T) X BD30N
100,000
BL90M X BD30N
s = 0,1745
BD30N X BWT
Rn/Ro ratio
10,000
CD30N (+N83T) X CWT
s = 0,1599
CL90M X CD30N (+N83T)
1,000
s = -0,0764
0,100
s = -0,3157
0,010
s = -0,3564
0,001
0
5
10
15
days post infection
20
(B or C+83T) 30N X (B or C) WT
In C, F and G HIV Protease, Position 89M
posses a structural barrier for the accumulation of
L90M
36
There is no
space for 2
Met residues.
89M/90M is very
rare
Can this 89M
play similar
role of 90M?
89
90
HIV Drug Resistance Database, Los Alamos National Laboratory
http://resdb.lanl.gov/Resist_DB
Role of L89M substitution in subtype F
protease
7
30
A
Fold Resistance
5
B
25
Fold Resistance
6
B
Bwt 89L90L
Bwt 89M90L
Bwt 89L90M
Bwt 89M90M
Bwt 89I90L
Bwt 89I90M
4
3
F13 89L90L
F13 89M90L
F13 89L90M
F13 89M90M
F13 89I90L
F13 89I90M
F
20
10
2
5
1
0
0
SQV
IDV
RTV
NFV
APV
LPV
SQV
IDV
RTV
PI`s
NFV
APV
LPV
PI's
-----
Biological FR cut-off
89M makes subtype F isolates more resistant to PI.
This is translated in a substantial drop in the EC50 when 89M is
reverted to 89L
Role of L89M substitution in subtype F
protease
7
30
A
Fold Resistance
5
B
25
Fold Resistance
6
B
Bwt 89L90L
Bwt 89M90L
Bwt 89L90M
Bwt 89M90M
Bwt 89I90L
Bwt 89I90M
3x FR
4
3
F13 89L90L
F13 89M90L
F13 89L90M
F13 89M90M
F13 89I90L
F13 89I90M
F
5x-7x FR
20
10
2
5
1
0
0
SQV
IDV
RTV
NFV
APV
LPV
SQV
IDV
RTV
PI`s
NFV
APV
LPV
PI's
-----
Biological FR cut-off
89M makes subtype F isolates more resistant to PI.
This is translated in a substantial drop in the EC50 when 89M is
reverted to 89L
Role of L89M substitution in
subtype F protease held true in
other field samples
F Subtype Series
8
F1
F2
6
Fold resistance
F13 89L90L
F13 89M90L
F18 89L90L
F18 89M90L
F17 89L90L
F17 89M90L
F22 89L90L
F22 89M90L
4
2
0
SQV
IDV
RTV
NFV
PI's
APV
LPV
Role of L89M substitution in subtype F
protease
10,00
Bw t 89L90L x F13 89L90L
B
Bw t 89L90L x F13 89M90L
Bw t 89L90L x F13 89L90M
Rn/Ro ratio
Bw t 89L90L x F13 89M90M
LM = 0.0165
1,00
ML = -0.0098
MM(2) = 0.0499
LL(1) = -0.0720
MM(1) = -0.1622
LL(2) = 0.0459
0,10
0
2
4
6
8
10
12
Time in days
14
16
18
20
Can the Polymorphism at Protease Position
36 Select for different Resistance Pathway?
• Patients failing first PI regimen
• Impact of baseline M36I polymorphism on selection
of L90M at failure
36I at baseline
No
Yes
P value
Adjusted odds
ratios of acquiring
90M at failure
1.00
13.5
0.009
J Infect Dis. 2004 Jun 1;189(11):1983-7
Subtype Impact on Drug Resistance
Genetic Variability:
Different codon/amino acids at
various positions in RT and
protease
Selection of different
resistance pathways for
identical antiretroviral
drugs
Replicative capacity
(RC)
Selection of different amino
acid substitutions at key
resistance positions
Genetic barrier: # of mutations
to aa substitution
codon usage in different subtypes change the acquisition of
DR mutations?
NRTI Mutation L210W in Subtype B and F
• Subtype F codon 210 usage for Leucine differs from
subtype B TTG versus CTG
• This may difficult the selection of L210W in Subtype F
Subtype B TTG to TGG = 1 change)
Subtype F- CTG to TGG = 2 changes)
• L210W is significantly more prevalent in patients
infected with subtype B failing AZT/D4T regimens when
compared to subtype F counterpart.
Further clinical data needed to determine relevance
Codon Usage Influencing the DR Pathway
NNRTI Resistance Mutation V106M
in Subtypes B and C
Subtype B
Subtype C
WT
Res
WT
Res
V106V
V106M
V106V
V106M
GTA
ATG
(2)
GTG
ATG
(1)
AIDS. 2003 Jan 3;17(1):F1-5
Prevalence of V106M in Subtypes B and C
Patients Failing Efavirenz
V106M
Subtype B
Subtype C
0.3%
24%
AIDS. 2004 Apr 9;18(6):909-15
NRTI Mutation K65R in Subtype B and C
• Subtype C codon 65 usage differs from subtype B (AAA
versus AAG)
• This may facilitate more rapid selection of K65R
Subtype C -(AAG to AGG = 1 change)
Subtype B- (AAA to AGG = 2 changes)
• K65R selected more rapidly in culture by Subtype C than
subtypes B, CRF1_AE, CRF2_AG or G
• 7/23 ddI experienced pts selected K65R in Botswana
AIDS. 2006 Jun 12;20(9):F9-13, AAC 2006 Dec;50(12):4182-5
Subtype Impact on Selection of Resistance Pathway
Subtypes
C
A, C, D
B, G
Drugs
Mutations
ZDV + ddI + NNRTI
67N + 70R + 215Y
Most common TAMs
NVP
K103N more common in C
and D than A
Protease inhibitors
82A more common in B
82M/T more common in G
14th CROI Los Angeles 2007 Ab 664, JAIDS 2006 Aug 15;42(5):610-3, Antivir Ther 2005; 10:S151,
JID 2004; 189:1232–8
Neural Network for non-B drug resistance
hybrid computational model for
selecting resistance- encoding
amino acids of HIV-1 protease.
Combines:
genetic algorithm (GA) based on
changes of aa hydrophobicity for
each substitution (Holland,1960;
Kyte & Doolitte, 1999)
+
embedded Kernel Fisher
Discriminant (KFD)
non-linear classifier
for predictions a posteriori
(Bayesian).
Neural Network for non-B drug resistance
Proof of concept: subtype B sequences.
Example: NFV-experienced in last regimen:
(20 simulations; sensitivity =100%; 97.2%>specificity>61.2; 97.3%>accuracy>69.1)
Posição
Freq.(%)
Posição
Freq.(%)
Posição
Freq.(%)
L10
T12
I13
I15
L19
K20
L23
A28
D30
L33
E35
M36
S37
73,7
26,3
31,6
68,4
36,8
47,4
5,3
21,1
31,6
10,5
57,9
84,2
78,9
P39
R41
K45
M46
G48
G49
I50
G51
I54
R57
I62
L63
I64
10,5
63,2
15,8
36,8
10,5
5,3
5,3
10,5
36,8
31,6
36,8
78,9
68,4
E65
K70
A71
I72
G73
V74
V77
V82
I84
N88
L89
L90
I93
10,5
10,5
36,8
57,9
10,5
10,5
68,4
15,8
21,1
42,1
15,8
21,1
52,6
Clinical Impact of Subtype on Resistance
• The genetic variability between subtypes clearly has an
impact on resistance
• The importance depends on the specific drug and
subtype
• Magnitude and type of interaction will determine to what
degree there is substantial clinical relevance
• Since antiretroviral drugs were primarily designed to
inhibit subtype B, one might potentially encounter less
favorable interactions with other subtypes
Nevertheless…
• Current evidences suggest excellent results with
antiretroviral drugs in Non B subtypes
Short-Term Virological Success Rates with
HAART in Non-B Subtype Settings
Location
Subtypes
Drugs
Success Rate
Botswana
C
2 NRTI + NNRTI
92.3% 1 yr
88.8% 2 yr
A, CRF02_AG
2 NRTI + PI
2 NRTI + NNRTI
60 – 80% 3 yr
Cote d'Ivoire
CRF02_AG
2 NRTI + NFV
2 NRTI + EFV
66% 10 mn
Africa, Asia,
South America
Multiple
2 NRTI + NNRTI
Majority (70%)
76% 6 mn
Malawi
C
prevalent
D4T + 3TC +
NVP
84% 1 yr
Senegal
Lancet. 2006 Mar 11;367(9513):817-24, Ped Infect Dis J. 2005 Dec;24(12):1072-6,
14th CROI, Los Angeles 2007 Ab 507 JIADS. 2005 Jan 1;38(1):14-7 Lancet. 2006 Apr 22;367(9519):1335-42
Summary
• Most drugs work well in most subtypes studied, with similar
selection of resistance
• Small differences may be difficult to determine and often are
not the major factor in patient drug failure
• Some specific subtype related issues have been identified
and have clinical relevance
• Presently subtype does not seem to be the major factor impacting
success/failure
• We need to continue the studies the clinical response of
patients infected with non-B HIV isolates.
• Subtype knowledge should never be a
reason to withhold therapy from a patient
Acknowledgements
Molecular Virology Lab- UFRJ
Amilcar Tanuri
Luciana Jesus da Costa
Alexandre Calazans
Mônica Arruda
Angélica Martins
Helena Pereira
Orlando Ferreira Jr.
Marisa M Morgado – FIOCRUZ / RJ
Jose Carlos Fernandez – FIOCRUZ / RJ
Ricardo S Diaz – EPM / UFESP
Ester Sabino – Fund. Hemocentro / SP
Ana Teresa Dumans
- UniRio
Marcelo A Soares
- UFRJ
Luis Felipe Gonzalez
- Venezuela