Background:  The rate at which autologous neutralizing antibody responses are mounted following HIV infection varies greatly among individuals. We hypothesized that infection with more neutralization-sensitive virus may result in stronger neutralizing antibody responses.

Methods:  Plasma samples were obtained from 38 individuals (37 male, 1 female; median age 38.5 years) with recent HIV infection (median estimated duration of infection, 85 days) who elected to remain off therapy for as long as a year following enrollment. Neutralizing antibody responses were measured between baseline virus and plasma from subsequent time-points during 12 months of follow-up using a recombinant virus assay, and quantified as the reciprocal of the dilution of plasma resulting in a 50% reduction in replication (limit of detection, 20). Viral set-point and the slope of log-transformed neutralizing antibody titers over time were estimated using linear random effects models. Neutralization sensitivity of baseline virus was measured using broadly cross-neutralizing control sera (N16). To determine the dependence of heterologous neutralization sensitivity on the choice of positive control serum, we screened a panel of primary isolates with N16 and broadly reactive sera obtained from 2 HIV-infected long-term non-progressors.

Results:  Autologous neutralizing antibody titers against baseline virus increased following infection (median 10.3-fold per year), with high variation between individuals (range 0.84- to 516-fold per year). Heterologous neutralization of baseline virus by N16 was highly variable among individuals (median neutralizing antibody titer, 48, range 20 to 490). Individuals infected with virus which could be neutralized by N16 exhibited a higher rate of increase of neutralizing antibody responses (13.5-fold vs 4.3-fold, p <0.05) and a 0.55 log₁₀ copies/mL higher viral set-point, which approached statistical significance (p = 0.08). There was excellent agreement between neutralization by N16 and sera from long-term non-progressors.

Conclusions:  Our results support a model in which infection with more neutralization-sensitive virus elicits more powerful neutralizing antibody responses, providing a biological rationale for the use of neutralization sensitive strains as vaccine candidates. More sensitive virus may also have a higher replication rate in vivo.