Our discovery success in CMV treatment
CMV (human cytomegalovirus) is a large, complex virus, with over 200 proteins, and is a major disease-causing agent in transplant and other immunocompromised patients, and newborns. About 40,000 CMV-infected infants are born every year in the US. Of these, ~8000 are born with neurological symptoms of varying severity and up to 8000 more will later develop progressive hearing loss. Only about half of pregnant mothers have adequate immunity naturally. To date, none of the multiple vaccine efforts have yielded a product with better than 50 percent reduction in transmission to the fetus. This result is particularly frustrating in light of recent clinical evidence establishing high efficacy for passive immunization with CMV-specific gamma globulin from naturally immune donors. This landmark study was conducted by a team of academic investigators in the US and Italy, including Trellis collaborators Stuart Adler and Michael McVoy at Virginia Commonwealth University School of Medicine. The Institute of Medicine has ranked CMV prophylaxis development at its highest priority level.
There has been no effective vaccine to CMV. Antibodies to the viral surface protein gB are known to neutralize CMV infectivity and thus gB has been a common vaccine target. However, these vaccines have not been particularly effective and may reflect a failure to induce antibodies with sufficiently high affinity. It has been suggested by Trellis collaborator Lenore Pereira at UC San Francisco that weak affinity antibodies fail to mediate virus phagocytosis effectively and appear to actually promote transmission across the human placenta.
Previous efforts to clone human antibodies against gB have been successful but limited in scope, and none has met the requisite affinity threshold. Using our proprietary CellSpot™ platform for discovery of native human monoclonal antibodies from naturally immune people, Trellis has surveyed ~50 million antibody producing human B cells from ~75 donors. Neutralizing clones against the invariant AD-2 (site 1) epitope on the viral gB protein are extremely rare. Our lead antibody, CMV345, specific for this epitope, has been characterized with regard to virus neutralization in vitro, and has a potency of <1 µg/mL across a wide assortment of cell types, including the ubiquitous fibroblast that can serve as a viral reservoir, and the primary cytotrophoblast (the key placental cell type in the natural maternal-fetal pathogenesis pathway).
The gB (AD-2.1) epitope is also of great interest with regard to CMV morbidity in the transplant setting. Transplant patients with higher natural titer to this epitope tend to have improved clinical outcomes than patients with little or no antibody response to AD2.1. Thus these patients make an attractive target population to receive an effective passive immunotherapy such as CMV345. Trellis’ CMV program will enter the licensing phase in Q1 2012.