Herd Thresholds and the Trap of Antibody-Dependent Enhancement

From Ronald Ross's Reproduction Number to Dengue Hemorrhagic Fever and the 1966 RSV Trial

The quantitative foundation of population-level disease control rests on the mathematical relationship between pathogen transmissibility and the fraction of immune hosts required to break chains of infection. Formulated through Nobel laureate Ronald Ross's 1911 differential equations for malaria mosquito thresholds, formalized as the net reproductive rate by Alfred Lotka and Richard Böckh, and applied to infectious epidemiology by George Macdonald in the 1950s and Roy Anderson and Robert May in the 1970s, the basic reproduction number $R_0$ defines the average number of secondary infections generated by one infectious case in a completely susceptible population. If a fraction $p$ of the population possesses sterilizing immunity, each infectious individual successfully transmits the pathogen to only $R_{\text{eff}} = R_0 (1 - p)$ contacts. Setting $R_{\text{eff}} < 1$ yields the critical herd-immunity threshold $p_c$ above which epidemic chains spontaneously decay:

This equation assumes a monotonic benefit to humoral immunity: antibodies either neutralize a pathogen when present above a protective stoichiometric threshold or, as titers wane, simply allow normal susceptibility to return. Across the 1960s, however, two medical crises demonstrated that under specific structural and stoichiometric conditions, non-neutralizing or sub-neutralizing antibodies actively worsen viral disease.

The first pathology was uncovered in Bangkok, Thailand and across Southeast Asia during the 1960s and 1970s by physician-virologist Scott Halstead. Studying children hospitalized with Dengue Hemorrhagic Fever (DHF) and Dengue Shock Syndrome (DSS)—characterized by abrupt capillary plasma leakage, thrombocytopenia, hemoconcentration, and circulatory collapse—Halstead observed a striking bimodal epidemiological pattern. Severe hemorrhagic shock almost never struck older children or adults experiencing their very first dengue infection. Instead, DHF clustered sharply in two groups: children experiencing a secondary infection with a different one of the four dengue virus serotypes (DENV-1 through DENV-4), and infants between six and nine months of age experiencing a primary dengue infection just as maternal anti-dengue $\text{IgG}$ acquired across the placenta decayed to sub-neutralizing concentrations.

Halstead termed the underlying mechanism Antibody-Dependent Enhancement (ADE). Complete neutralization of a flavivirus particle requires high-affinity antibodies to occupy enough of the 180 envelope (E) glycoproteins on the virion surface to prevent endosomal membrane fusion. When a child who previously recovered from DENV-1 is later bitten by an Aedes mosquito carrying DENV-2, pre-existing cross-reactive memory antibodies bind the heterologous DENV-2 virion at a stoichiometry below the neutralization threshold. Worse, because dengue virus primarily infects myeloid cells—monocytes, macrophages, and dendritic cells—that display surface $\text{Fc}\gamma\text{RIIA}$ and $\text{Fc}\gamma\text{RIIB}$ receptors, the exposed $\text{Fc}$ stems of those non-neutralizing antibodies act as molecular grappling hooks. The antibody-coated virion binds $\text{Fc}\gamma$ receptors, triggers receptor-mediated endocytosis into the monocyte, and suppresses intracellular type I interferon signaling (extrinsic and intrinsic ADE). That facilitation multiplies systemic viral load ten- to one-hundred-fold and unleashes a vasoactive cytokine storm. When the chimeric tetravalent vaccine Dengvaxia was deployed across the Philippines in 2016–2017, follow-up revealed that immunizing dengue-naive children mimicked a silent primary infection, predisposing seronegative vaccinees to enhanced hospitalization upon their first natural dengue exposure.

A second, mechanically distinct form of vaccine-associated enhanced respiratory disease (VAERD) occurred in 1966 during an NIH-sponsored clinical trial of a formalin-inactivated, alum-precipitated vaccine against Respiratory Syncytial Virus (FI-RSV, Lot 100). Among 31 infants aged two to seven months who received FI-RSV in Washington, D.C. and subsequently contracted wild RSV in the community, 25 developed severe lower respiratory tract bronchiolitis and pneumonia. That represented an 80 percent hospitalization rate among infected vaccinees compared to 5 percent in the control group, and two vaccinated toddlers died. Decades of structural and immunological forensics led by Fernando Polack, Barney Graham, and Jason McLellan revealed why formalin inactivation failed so catastrophically for RSV where it had succeeded for poliovirus. Unlike rigid icosahedral poliovirus capsids, the RSV fusion (F) envelope glycoprotein is a spring-loaded class I viral fusion machine. Heating RSV in formalin for 72 hours at $37^\circ\text{C}$ stripped away its delicate prefusion neutralization epitopes and triggered the F protein to flip prematurely into its elongated postfusion rod conformation. Infants injected with FI-RSV produced high titers of low-avidity antibodies against denatured postfusion epitopes that failed to neutralize live virus. When wild RSV arrived in the lungs, those non-neutralizing antibodies formed complement-fixing immune complexes in the bronchioles while alum-primed $\text{CD4}^+$ $\text{Th2}$ cells secreted interleukin-4, interleukin-5, and interleukin-13, flooding the infant alveoli with eosinophils and neutrophils.