Predictors of Virologic Failure and HIV Drug Resistance Among Patients Receiving Fixed Dose Combination Stavudine/Lamivudine/Nevirapine in Northern Tanzania.

Predictors of Virologic Failure and HIV Drug Resistance Among Patients Receiving Fixed Dose Combination Stavudine/Lamivudine/Nevirapine in

Northern Tanzania

Duke University-KCMC Collaboration

Kilimanjaro Christian Medical Centre

Moshi

TANZANIA

Habib O. Ramadhani,* Nathan M. Thielman, Feng Gao, Jennifer Kirchherr, Rekha Shah, Keren Z. Landman, Evelyn M. Ndosi,

Humphrey J. Shao, Susan C. Morpeth, Jonathan McNeill, Venance P. Maro, John F. Shao, John A. Bartlett,

John A. Crump

Background

• ART access expanding rapidly in Africa– Free ART in Tanzania from September 2004– Before and during transition some patients paid– Focus on numbers of persons on therapy

• Few data are available on prevalence and risk factors for virologic failure and resistance

• Needed to optimize ART programs

Aims

1. Determine extent and risk factors for virologic failure

2. Describe HIV genotypic resistance mutations

Methods

• Retrospective cohort study– 150 HIV-infected adult patients, June-August 2005– FDC D4T/3TC/NVP ≥ 6 months– Consecutive patients presenting for follow up

• Standardized questionnaire– Sociodemographics– Economic conditions– HIV and ART knowledge, beliefs, disclosure– Adherence– Access to care

• Plasma– HIV RNA quantitation– HIV genotypic resistance testing

Methods

• Virologic failure– HIV RNA ≥400 copies/mL

• Drug resistance– ≥1 major NRTI or NNRTI mutation– International AIDS Society (USA) November 2005

Revision

Results• 94 (63%) female

• Median age 41 years (range 19-69)

• Median CD4 ART initiation 113 cells/mm3 (range 1-628)

• Median duration ART 12 months (range 6-27)

• 23 (16%) adherence less than 100%

Results

All subjects 150 (100%)

Virologic failure ≥ 400 copies/mL 48 (32%)

Not virologic failure <400 copies/mL 102 (68%)

Risk Factors for Virologic FailureBivariable Analysis

OR 95% CI p-value

Proportion on months on ART self-funded > median 2.7 1.3 5.6 0.006

Primary or no education 1.8 0.9 3.8 0.092

Male gender 1.3 0.7 2.6 0.452

Minutes walking to clinic 1.0 1.0 1.0 0.084

Age 1.0 1.0 1.0 0.537

Number of months on SVD/LMV/NVP 1.0 1.0 1.0 0.488

Rufiji Asset Score 1.0 1.0 1.1 0.469

Disclosure of HIV serostatus to persons other than clinic staff

0.17 0.3 0.9 0.040

Risk Factors for Virologic FailureMultivariable Analysis*

OR 95% CI p-value

Proportion on months on ART self-funded > median 4.2 1.7 10.5 0.002

Anyone beside clinic staff know HIV serostatus 0.11 0.2 0.7 0.023

*Model controlled for age and gender and included variables with p<0.1 from biavriable analysis

Self-Funded ART and Virologic Failure

• Persons paying for ART were more likely to be maladherent

r=0.54, p<0.001

Results

Genotyping ≥1,000 copies/mL 35 (73%)

No genotyping <1,000 copies/mL 13 (27%)

Sequenced 27 (77%)

Not sequenced 8 (23%)

≥1 resistance mutation 15 (56%)

No resistance mutation 12 (44%)

All subjects 150 (100%)

Virologic failure ≥ 400 copies/mL 48 (32%)

Not virologic failure <400 copies/mL 102 (68%)

Genotypic Resistance

No. mutations n (%) Type Mutation (n)

1 2 (13) NRTI na

NNRTI K103N (1), Y181C (1)

2 7 (47) NRTI M184V (7)

NNRTI K103N (3), Y181C (1), G190A (3)

≥3 6 (40) NRTI Q151M (1), M184V (4), M184I (1) TAMs: M41L (1), K65R (1), L210W (1), T215Y (1)

NNRTI K103N (3), V108I (1), Y181C (4), G190A (2)

Risk Factors for Resistance MutationsBivariable Analysis

OR 95% CI p-value

CD4 count <median at treatment initiation 4.8 1.5, 21.9 0.009

Conclusions

• One third of adult patients taking ART for ≥6 months are experiencing virologic failure

• Patients who paid for ART are at risk for virologic failure mediated by maladherence

• Disclosure of HIV status is protective against virologic failure

• Presence of resistance mutations is associated with low CD4 count at initiation

Recommendations

• Provide free ART and promote social coping to enhance adherence and to limit virologic failure

• Encourage earlier detection of persons with HIV infection

• Sentinel site HIV RNA quantitation useful to– Evaluate ART program – Understand how to optimize ART– Identify treatment failure early

Acknowledgements

• Kilimanjaro Christian Medical Centre– Mark E. Swai, MD

– Sr Gertrude I. Kessy

– Sr Janet U. Kimaro

– Sr Bona K. Shirima

– Francis P. Karia, MBA

– Ahazi T. Kulanga, MBA

• KIWAKKUKI– Dafrosa K. Itemba

– Anna Mgonja

– Rehema A. Kiwera

• Division of Infectious Diseases and International Health, Duke University – John D. Hamilton, MD– Christopher W. Woods, MD, MPH– L. Barth Reller, MD– Anne B. Morrissey

• University of California at San Francisco– David R. Bangsberg, MD, MPH

• US National Institutes of Health– Duke Center for AIDS Research– AITRP– K24– ACTU

Kilimanjaro Christian Medical Centre, Moshi, TANZANIA

Kilimanjaro Christian Medical College, Tumaini University, Moshi, TANZANIA

Backup Slides

Amplification of Partial pol Gene

• RNA purified from plasma samples using the Qiagen EZ1 Virus Mini Kit (QIAGEN) and subjected to cDNA synthesis using SuperScript III (Invitrogen)

• Viral cDNA samples were subjected to nested PCR amplification using the following primers with the QIA quick Gel Extraction kit (QIAGEN)

– gagout (5'- TAGAAGAAATGATGACAGCATGTCAGGGAGT -3')– polout (5'- CTAACTTCTGTATATCATTGACAGTCCA -3')– gagin (5'- CTGAGAGACAGGCTAATTTTTTAGGGA -3')– polin (5'- CATCCAAAGGAATGGAGGTTCTTTCTGATG -3')

• PCR cycling conditions included– 1 cycle at 94C for 2 min– 30 cycles of a denaturing step at 94C for 20s– An annealing step at 50C for 15s– An extension step at 68C for 2 min– 1 cycle of an additional extension at 68C for 10min– PCR products were visualized on 0.7% agarose gels and purified

Sequence Analysis

• Performed on purified PCR products by cycle-sequencing and dye terminator methods with an ABI PRISM® 3100 Genetic Analyzer (Applied Biosystems) as recommended by the manufacturer

• Sequences were determined for both strands of DNA– Individual sequence fragments of pol gene amplicon from

each virus were assembled and edited using the Sequencher program 4.2 (Gene Codes).

Phylogenetic Analysis

• Phylogenetic relationships of new viral sequences determined by comparing partial pol nucleotide sequences to group M subtype references– Nucleotide sequences aligned using CLUSTAL– Pairwise evolutionary distances were estimated by using

the Kimura’s two-parameter method to correct for superimposed substitutions

– Sites where there was a gap in any sequence and ambiguous areas within the alignment were excluded from all comparisons

– Phylogenetic trees were constructed by the neighbor-joining method

– Reliability of branching orders was assessed by bootstrap analysis using 1,000 replicates

NRTI Mutations of Patients with Virologic Failure, KCMC, 2005

Patient M-41 A-62 K-65 D-67 K-70 V-75 F-77 F-116 F-151 M-184 L-210 T-215 K-219

19 L I W

22 V

24

31 V Y

32 V

45 V

61 V

75 V

85 V

88 R

95 V

108

129 V

130 V

138 V

International AIDS Society (USA) November 2005 Revision: Major Mutations

NNRTI Mutations of Patients with Virologic Failure, KCMC, 2005

Patient L-100 K-103 V-106 V-108 Y-181 Y-188 G-190 M-230

19 N C

22 A

24 N

31 A

32 N

45 N

61 N

75 I C

85 A

88 C A

95 A

108 C

129 C

130 N C

138 N

International AIDS Society (USA) November 2005 Revision: Major Mutations

Subtype

• HIV-1– Subtype A 10 (37%)– Subtype C 7 (26%)– Subtype D 10 (37%)

H.VI997

H.90CF056J.SE9173

J.SE9280K.EQTB11C

K.MP535B.MNCGB.WEAU160

D.NDKD.94UG1141

TZ19TZ95

TZ24TZ38

TZ103TZ36

TZ88TZ85TZ75

TZ138F1.93BR020.1

F1.FIN9363G.HH8793

G.92NG083C.92BR025

TZ54TZ61

C.ETH2220TZ118

TZ26TZ130

TZ32TZ129

TZ39TZ31

A.KE.Q23-CXC-CGTZ45

TZ108TZ92

TZ27TZ22A.UG.92UG037

TZ20TZ68

TZ43

0.02

Subtype C

Subtype D

Subtype A

HIV-1 Subtypes of Patients with Virologic Failure, KCMC, 2005