Breadth of Neutralizing Antibodies in HIV Serum: Clustering Analysis and Association with Clinical V

1
Breadth of Neutralizing Antibodies in HIV Serum
Clustering Analysis and Association with
Clinical Variables

• Mark Connors, M.D.
• HIV-Specific Immunity Section
• National Institute of Allergy and Infectious
  Diseases

2
Introduction

• Induction of broadly cross-reactive neutralizing
  antibodies (NAb) is a major goal for HIV vaccines
• Current vaccine candidates elicit weak,
  narrowly-directed NAb, however
• Some HIV-infected patients make very broadly
  cross-reactive NAb

3
Purpose of Study

• Characterize neutralization breadth in a large
  diverse cohort gt100 patients
• Previous study neutralization of 5 isolates,
  qualitative
• This study neutralization of 20 isolates,
  quantitative
• Find clinical or demographic correlates of
  breadth
• Find patterns of serum reactivity to virus
  isolates

4
Methods

• Serum clinical, demographic data from patients
  at NIH
• 25 LTNP/EC, VL lt50, stable CD4
• Migueles et al 2002
• 78 viremic patients
• Neutralization assay
• TZM-bl assay
• 20-virus panel
• Primary isolate envs of A, B, C clades
• Range of neutralization sensitivities, Tier 2
• Clinical correlations
• Clustering analysis

5
Viremic patients have variable NAb
ID50
6
Breadth is not rare in viremic patients
ID50

• Half of sera neutralize at least 11/20 isolates

• 20 of sera neutralize at least 15/20 isolates

7
LTNP/EC
LTNP/EC

• Rarely have broad NAb
• Reflect low end of variation found in viremic
  patients

Viremic
8
Breadth correlates with viral load

• Modest positive correlation with VL
• No consistent correlation with
• CD4 count
• Years since dx
• HLA Class II alleles
• ARV experience
• Risk Group
• Age
• Gender
• Race/ethnicity

Viremic
LTNP/EC
Log Viral Load
Log Geomean ID50
Plt0.001, r0.67
9
Heatmap of neutralization data
Sera
Isolates
Log ID50
10
Statistical analysis allows precise cluster
assignments
Sera
Isolates
Log ID50
11
Serum clusters reflect reactivity with isolates
LTNP/EC
Sera
Isolates
Log ID50
12
Isolate clusters based on serum reactivity do not
match clades

• Cluster ? clade
• Related sequences are near each other but not
  always in same cluster

13
Sera and MAbs define different isolate clusters

• Isolate clustering based on sera or monoclonals
  did not match
• Isolate clustering based on sera did not
  correlate with reactivity to any one monoclonal
• Consistent with polyclonal nature of serum

IC50 (ug/ml)
14
Conclusions

• Sera with broad NAb are not uncommon in viremic
  patients
• Breadth is associated with exposure to antigen
• Clustering analysis reveals relationships between
  isolates based on patterns of serum
  neutralization reactivity
• Clusters distinct from clade
• Clusters distinct from sensitivity to MAb

15
Acknowledgements

• Clinical Research Section
• Nancy Cogliano-Schutta
• Gregg Roby
• Cathy Rehm
• Sara Stallings
• Los Alamos National Laboratory
• Bette Korber
• Marcus Daniels
• Alan Lapedes
• Tanmoy Bhattacharya
• Patients

• National Institute of Allergy and Infectious
  Diseases
• HIV-Specific Immunity Section
• Nicole Doria-Rose
• Rachel Klein
• Stephen Migueles
• Adjoa Smalls-Mantey
• Biostatistics Research Branch
• Martha Nason
• Vaccine Research Center
• John Mascola
• Sijy ODell
• Krisha Svehla
• Mark Louder
• Richard Wyatt