Reply to: “Antibody affinity maturation and respiratory syncytial virus disease”

Delgado et al. reply:

Shaw et al.¹ state that our observations indicate that a protective RSV vaccine must elicit affinity-matured neutralizing antibodies for enhanced respiratory disease (ERD) not to occur². We agree with this statement. We also agree that, ideally, vaccines should present native respiratory syncytial virus (RSV) structures², although we think that modifications that do not affect the generation of neutralizing antibodies would be acceptable.

In addition, Shaw et al.¹ state that our paper blames ERD on the lack of “antibody avidity alone.” This statement simplifies our observations. A widely accepted paradigm to explain ERD for decades ascribed the disease solely to formalin disruption of protective epitopes^3,4,5. As we showed in our paper, these epitopes are still recognized by a formalin-inactivated RSV (FI-RSV)-elicited germline antibody². Yet we subsequently showed that eliciting maturation of FI-RSV antibody would shift the response from native toward formalin-modified epitopes². This process, which did not occur in the 1960s when the vaccine failed to elicit maturation, alters recognition of protective areas and explains our decision to conduct experiments using ultraviolet light–inactivated RSV instead of FI-RSV.

The common denominator for all nonreplicating vaccines that prime for ERD is the generation of low-affinity antibodies^6,7. Thus, it is crucial to emphasize that there will be no safe nonreplicating RSV vaccine for infants in the absence of appropriate adjuvants. Certain vaccines may pose additional challenges, and we did not—at any point in our study—support the use of FI-RSV. We merely challenged the widespread belief that the problem with ERD is just a matter of epitope disruption, which could be solved using other methods for virus inactivation².

Shaw et al.¹ cite a study in which FI-RSV was formulated with a Toll-like receptor 4 (TLR4) agonist and reduced ERD symptoms⁸ but fail to mention that RSV challenge shortly after immunization encountered transient protection (probably attributable to steric hindrance) in other nonreplicating RSV vaccine formulations^7,9. So we conclude that formalin should never be used again to inactivate RSV for vaccine development.

Finally, there is a small but key difference between the statement quoted in the correspondence from Shaw et al.¹ (“...subsequently led to severe disease”) and our statement in the abstract (“...did not protect the children and consequently led to severe disease”)². We showed that sera from mice immunized with ultraviolet light–inactivated RSV plus TLR agonists protected FI-RSV–immunized mice lacking antibodies from ERD after RSV challenge. In other words, as a consequence of antibody-mediated protection in these adoptive transfer experiments, and independent from TLR effects on T helper or other cells, ERD did not occur.

In summary, FI-RSV failed to protect primarily as a result of poor avidity, as germline antibodies continued to recognize protective epitopes. Moreover, specifically maturing FI-RSV–specific antibody would not have solved the problem. Last, no nonreplicating vaccine against RSV will be safe for infants if it fails to elicit affinity maturation.