Comparison of Less Sensitive HIV

1
Comparison of Less Sensitive HIV Incidence Tests
for the Serological Testing Algorithm for Recent
HIV Seroconversion (STARHS) J. Schiffer, B.
Byers, B. Branson, S. Ethridge, K. Delaney, J.
Mei
Centers for Disease Control and
Prevention Atlanta, GA 30341
2
Background

• STARHS uses a modified protocol of the bioMerieux
  Vironostika HIV-1 MicroElisa system (V-LS)- To be
  replaced in 2005
• We investigated the comparability of the
  replacement test, Vironostika HIV-1 Plus O
  (VPlusO-LS)
• Longitudinal specimens from the HIVNET Infected
  Participant Cohort (IPC) were tested using both
  kits

3
Methods

• IPC specimen set
• 1009 specimens from 105 recent seroconverters
• Known dates of last negative and first positive
  HIV test
• Tested in triplicate on both assays

4
Methods

• Window Period Calculation
• 324 specimens from 59 patients
• 3 or more specimens collected before starting
  antiretroviral therapy (Mean length of follow up
  504 Days)
• Mean window period definition
• Number of days between detection of HIV antibody
  with a standard EIA and a threshold SOD of 1.0
  with the LS-EIA

5
Results

• V-LS classified 459 samples as incident (SODlt1),
  and VplusO classified 472 as incident (90.2
  concordance).
• Specimen SODs diminished after patients started
  antiretroviral therapy.

6
(No Transcript)
7
Table 1 Window Period and Confidence Intervals
for LS Tests
            Vironostika-LS Vironostika-LS Vironostika-LS Vironostika Plus O-LS Vironostika Plus O-LS Vironostika Plus O-LS
SOD Cutoff Window Period 95 Confidence Interval N (Patients, Samples) Window Period 95 Confidence Interval N (Patients, Samples)
0.7 157 140-189 59, 314 142 127-157 59, 324
0.8 179 157-202 59, 314 163 149-184 59, 324
0.9 198 174-224 59, 314 183 167-218 59, 324
1 218 193-250 59, 314 203 185-243 59, 324
1.1 241 214-275 59, 314 225 203-264 59, 324
8
Two-Year Rule

• SODs tended to plateau or decline after 2 years
  from last negative specimen.
• May result in longer window periods
• New window periods calculated excluding specimens
  collected after 2 years from last negative test

9
Table 2
Window Period (SOD 1) 95 Confidence Interval N (Patients, Samples)
Vironostika-LS (IPC, 2 year cutoff) Vironostika-LS (IPC, 2 year cutoff) Vironostika-LS (IPC, 2 year cutoff) 212 187-244 59, 290
Vironostika Plus O-LS (IPC, 2 year cutoff) Vironostika Plus O-LS (IPC, 2 year cutoff) Vironostika Plus O-LS (IPC, 2 year cutoff) 196 177-228 59, 294
Vironostika-LS (SFMHS, 2003) Vironostika-LS (SFMHS, 2003) Vironostika-LS (SFMHS, 2003) 170 145-200 48, 227
Window periods and confidence intervals for V-LS
and VplusO-LS calculated with the two year cutoff
rule applied to the IPC sample set. The V-LS
window period previously published using the
SFMHS sample set (Kothe et al. 2003) is shown for
comparison.
10
Acknowledgements

• Chip Sheppard and Sean Watson of the California
  Department of Health Services for providing the
  HIVNET IPC specimens, without which this work
  would not have been possible.
• We would also like to thank Debra Kuehl, Silvina
  Masciotra, Hetal Patel, and Donna Rudolph for
  their help with organizing and aliquoting the IPC
  sample set