V Miller, A CozziLepri, K Hertogs,

1
Association between phenotypic resistance and
virologic response to mega-HAART regimen in
patients from the Frankfurt HIV Cohort

• V Miller, A Cozzi-Lepri, K Hertogs,
• P Gute, B Larder, S Bloor,
• C Rottmann, AN Phillips, S Staszewski

2
Introduction

• The Frankfurt HIV Cohort consists of all patients
  attending the clinic who have presented at least
  once since Jan 1, 1995
• data pertaining to antiretroviral history, immune
  status, HIV-1 viral load, HIV-associated clinical
  events and demographics are collected
  prospectively
• patients with multiple treatment failures were
  treated with salvage regimen consisting of 6 or
  more antiretroviral drugs
• phenotypic resistance tests were performed from
  stored plasma samples

3
Patients Methods

• 50 patients met the definition criteria and the
  DAP requirements
• viral load Amplicor
• resistance Antivirogram (recombinant virus based
  phenotypic assay)
• virologic failure gt 400 copies/week 24 window
• statistical analysis as described per DAP

4
Study Population Baseline characteristics

• Characteristic Median Range
• HIV RNA (log10 copies/ml) 5.52 3.40 -
  6.70
• CD4 cell count (cells/mm3) 95 2 - 587
• Time on HAART (months) 18.3 1.7-53.5

5
Summary of previous ART
6
Susceptibility Status at Baseline
7
On - Treatment Drugs
8
Univariate Analysis Odds ratio of virological
failure
DAF DAC
NRTI sens/10
new drugs
NRTI sens/4
PI sens 4
PI sens/10
sens/10
sens/4
RNA
9
Multivariate Analysis Odds ratio of virological
failure
DAF DAC
Multivariate 2
Multivariate 1
RNA New Drugs Sens/4
RNA New Drugs Sens/10
10
Multivariate Analysis Odds ratio of virological
failure
DAF DAC
Multivariate 4
Multivariate 3
RNA New Drugs NRTI Sens/4
PI Sens/4
RNA New Drugs NRTI Sens/10 PI Sens
10
11
Conclusion

• in this retrospective analysis based on a small
  group of extensively experienced patients,
  phenotypic resistance was a significant predictor
  of virologic response, independent of baseline
  viral load and previous antiretroviral history
• either a 4-fold or a 10-fold cut-off provided
  significant independent information
• in models looking at resistance within individual
  drug classes, resistance to PI was more
  predictive than resistance to NRTI