1. Emerging patterns of drug resistance and viral tropism in cART-naïve and failing patients infected with HIV-1 subtype C
Thumbi Ndung’u, BVM, PhD
Associate Professor
Director, HIV Pathogenesis Programme
Doris Duke Medical Research Institute
Nelson R. Mandela School of Medicine
University of KwaZulu-Natal

3. HIV-1 Phylogeny
47.2%
12.3%
27.0%

4. Phenotypic Classification of HIV-1
Slow/low versus rapid/high
Syntitium-inducing (SI) versus NSI
Slow/Low = NSI (Early, slow progression)
Rapid/High = SI (Late, rapid progression)

5. HIV-1 coreceptor usage and viral tropism
Virus Variants
M-tropic
Dual tropic
T-tropic
CCR5
CXCR4
Macrophage
CCR5
Primary
T cell
CXCR4
T-cell line
CD4
CD4
CD4
Target Cell Types

6. >25 years of HIV/AIDS > 33
For every 2 people put on treatment, 5 others are infected

7. Selection of Resistant strains Selection of resistant quasispecies
Treatment begins
Drug-susceptible quasispecies
Drug-resistant quasispecies
Selection of resistant quasispecies
Viral load
Incomplete suppression
Inadequate potency
Inadequate drug levels
Inadequate adherence
Pre-existing resistance
Time

8. Study rationale Background:
Relatively limited information on coreceptor usage by HIV-1 subtype C isolates, particularly in children. However, most studies suggest very rare CXCR4 usage
Some reports suggest increasing X4 usage (in adults) eg. Johnston et. al. (n=28), 50% using X4 among ART experienced viremic patients
Previously used methods may be biased because they involved first generating viral isolates by co-culture

9. Study rationale
ART may boost T-cell immune responses which have been shown to preferentially suppress X4 viruses. Thus partially effective therapy may select against X4 viruses (Deeks et al, JID 2004; Harouse et al, PNAS 2003)
ART reduces CCR5 expression on T cells (due to reduction in T cell activation) potentially selecting for X4 viruses (Brumme et al, JID 2005; Anderson et al, AIDS 1998)
Suboptimal drug metabolism (such as AZT) in the cellular reservoirs for X4 viruses has been suggested and could lead to selection for X4 viruses (Boucher et al, AIDS 1992)

10. Aims
Specific Aims:
1) To determine the prevalence of major drug mutations in ART-naïve and failing children and adults
2) Determine overall prevalence of X4 tropism among children and adults initiating and failing HAART
3) Compare prevalence of X4-utilizing viruses between ART-naïve and ART-experienced subjects with detectable viremia
4) Explore factors associated with viral tropism in HIV-1C infection

11. HIV-1 Genotyping Assay plasma Blood cells T T C T C G A T C G RNA
centrifugation
Blood cells
RNA
RT-PCR
cDNA
PCR
DNA
PCR
Dye terminators
A T G C T T C T C G A T C G
Software analysis
ATAGACCAG : consensus sequence
I Q Q
ATCGACCTG : patient sequence
I Q *L

12. + Trofile assay summary- for coreceptor usage pcDNA-env gag pol env
5’LTR
gag
pol
env
vif
vpr
vpu
tat
rev
Luc
3’LTR +
pcDNA-env
CMV pA
Env
Luciferase assay
0.2µ
filter
0.2µ
filter
293T cells
CCR5 cells
CXCR4 cells

13. Table 1: Children Demographic and Clinical Characteristics
NOTE. Data are no. (%) of children unless otherwise indicated. For cases where the data is incomplete, the n value is indicated.
Statistical tests: a Mann-Whitney U test and b Fisher’s exact test (for WHO stage analysis, stages I, II and III were grouped together).

14. Table 1: Patient Demographic and Clinical Characteristics Cont.
NOTE. Data are no. (%) of children unless otherwise indicated. For cases where the data is incomplete, the n value is indicated.
Prior treatment indicated with underlined drug/s changed ● d4T, 3TC, ritonavir (n=1); * unknown; ○ d4T, 3TC, EFV (n=1) and AZT, 3TC, NVP (n=1);  d4T, 3TC, kaletra;  d4T, 3TC, EFV.
Statistical tests: a Mann-Whitney U test and b Fisher’s exact test

15. Frequency of drug resistance mutations and levels of resistance in HAART-failing children to the NRTIs (a) and NNRTIs (b)
58.5% had TAMs
39% had ≥3 TAMs

16. Average no. of major mutation in patients failing standard first line treatment (n=30)
d4T/3TC/EFV (n=25)
3 patients have no DRMs (VLs are 617; 79,400; 228,000)
20 NRTI DRM
2 NNRTI DRM
(one patient had a PI DRM)
d4T/3TC/kaletra (n=5)
3 patients have no DRMs (VLs are 143,000; 198,000; 4,410,000)
1 patient has 1 NRTI DRM (M184V) only
1 patient has 1 NRTI (M184V) and 1 NNRTI DRM (Y181C)

17. d4T/3TC/EFV (n=25) → AZT/ddI/Kaletra Potential low-level (n=2)
How many major mutations compromise the standard second line treatment?
d4T/3TC/EFV (n=25) → AZT/ddI/Kaletra
3 patients susceptible to all drugs – no change needed
All patients susceptible to kaletra
3 patients susceptible to 3 drugs in standard second line tx.
AZT Resistance
ddI Resistance
Susceptible (n=2)
High-Level (n=2)
Low level (n=5)
Potential low-level (n=2)
Low-level (n=1)
Intermediate (n=2)
Intermediate (n=8)
Low-level (n=3)
High-level (n=1)
High-Level (n=4)

18. Overall, 13 of 25 (52%) patients will have some degree of resistance (low to high) to two of the three drugs in their new regimen (excluding potential low-level resistance)

19. d4T/3TC/kaletra (n=5) → AZT/ddI/(NVP/EFV)
4 of 5 patients are susceptible to all second line drugs
1 patient had intermediate resistance to EFV (3.7 yrs old)(Y181C)
Note: Overall better if not changed
All still susceptible to PIs and d4T with 3 patients still susceptible to 3TC [ high-level resistance to 3TC (M184V)]
Patients <3 years will be put on NVP
Patients >3 years will be put on EFV

20. Comparison of coreceptor usage in HAART-failing and HAART-naïve children

23. Evaluation of Several Genotypic Tools for the Prediction of CXCR4-usage
a A total of 52 pure subtype C isolates with both phenotypic and genotypic data were included in this analysis. bA false positive rate of 10% was used. c A combination of the first four genotypic tools were used where the majority prediction was considered as the final genotype prediction (n=47).

24. Adult patient information
Characteristic
HAART-Experienced Patients failing Treatment (n=45)
HAART-Naïve Patients (n=45)
p-value
Age, median years
(Q1-Q3)
36 (24-51)
36 (20-78)
0.65
Gender: Female
28 (65%)
27 (60%)
Black race
45 (100%)
CD4 count, median cells/mm3 (Q1-Q3)
Current
Nadir
174 (9-718)
57 (3-197)
123 (8-660)
0.036
0.0004
Vial load,
median copies/ml
6, 653
( ,010)
44,042
(1,702-1,167,759)
0.001
WHO stage at visit
I-III IV
32 (71 %)
13 (29 %)
9 (20 %)
36 (80%)

25. Patterns of drug
resistance

26. What is the outcome of patients failing if started on the standard second line of treatment without having genotypic data?

27. Average no. of major mutation in patients failing standard first line treatment (n=16)
d4T/3TC/ (EFV/NVP) (n=16) (Note: 2 on NVP)
No major PI mutations
1.75 NRTI DRM
1.69 NNRTI DRM

28. How many compromise the standard second line treatment?
d4T/3TC/ (EFV/NVP) (n=16) → AZT/ddI/LPV/r
All patients susceptible to kaletra (LPV/r)
6 patients susceptible to all 3 drugs in standard second line tx.
AZT Resistance
ddI Resistance
Susceptible (n=4)
Potential low-level (n=3)
High-Level (n=1)
Potential low-level (n=2)
Susceptible(n=1)
Low-level (n=1)
Low-level (n=2)
Intermediate (n=2)

29. 4 of 16 (25%) patients will have some degree of resistance (low to intermediate) to two of the three drugs in their new regimen (excluding potential low-level resistance).
6 of 16 (37.5%) will have some degree of resistance (low to high) to one of the three drugs in their new regimen (excluding potential low-level resistance).

30. High levels of CXCR4 viruses in patients failing therapy- limited salvage options

31. V3 loop-based methods for coreceptor usage prediction
% of sequences correctly predicted
% of R5 sequences correctly predicted
% of X4/D/M sequences correctly predicted
11/25 78 90 55
Overall net V3 charge 75 71 81
C-PSSM 85 72
Geno2Pheno 84 86 82
Combined algorithm* 87 80
*In the combined algorithm, concordant results from at least 3 of 4 methods
(i.e. the amino acids at positions 11 and/or 25, the overall net V3 charge, C-PSSM
prediction and Geno2Pheno prediction) were used.

32. Conclusions Virologic failure is mainly due to DRMs
High levels of TAMs is source of concern- suggests subpotimal adherence and need for intensive monitoring
Higher levels of CXCR4 using viruses among HAART experienced patients- need to explore CCR5 antagonists as part of first-line/early treatment
Collectively, these data highlight the need for intensified adherence counselling and better HAART monitoring to maximize benefits.

33. Acknowledgements UKZN McCord Hospital Monogram Biosciences
Taryn Green
Ashika Singh
Mohendran Archary
Michelle Gordon
Raziya Bobat
Hoosen Coovadia
McCord Hospital
Henry Sunpath
Richard Murphy
Monogram Biosciences
Jacqueline Reeves
Yolanda Lie
Elizabeth Anton
Harvard University
Daniel Kuritzkes
Bruce Walker
Funding
IMPAACT Network, NIH
Harvard University CFAR
South African DST/NRF
Hasso Plattner Foundation