SNOWMASS, COLO. — An effective vaccine against Epstein-Barr virus could conceivably turn systemic lupus erythematosus into a disease of historical interest within a couple of decades—but that's far easier said than done.
“There have been three trials of capsid-based EBV vaccines. All have failed because the vaccines didn't protect,” said Dr. John B. Harley at a symposium sponsored by the American College of Rheumatology. Most people may be infected with five to nine substrains of the virus, he added. “We go through life possibly being reinfected by other strains as life progresses. Even the defenses that prevent us from getting mononucleosis over and over again are not sufficient to prevent reinfection. So finding a way to prevent the viral infection is going to be very complicated.”
For decades, Dr. Harley has been developing the hypothesis that EBV (in conjunction with genetic predisposition) causes SLE. The hypothesis initially went nowhere, but it has gained considerable traction as a result of mounting evidence that has converged from epidemiologic, immunologic, and genetic studies conducted in many centers.
When he first zeroed in on EBV as likely having a causative role in lupus, Dr. Harley understood that he would face scientific skepticism. For decades, it seemed that whenever researchers could not explain the pathogenesis of various poorly understood diseases, they'd try to pin it on EBV, which “has been blamed for everything, and yet very little has been established as being causative,” observed Dr. Harley, chair of the arthritis and immunology research program at the Oklahoma Medical Research Foundation in Oklahoma City.
Back in the 1980s, he laid the groundwork for his future studies of EBV when he recognized the unique research potential of the Department of Defense Serum Repository, which contains frozen serial specimens from more than 5 million armed forces personnel. He and his coworkers identified 130 individuals who developed SLE and for whom they found serum samples dating back to years before disease onset. This enabled the investigators to characterize—for the first time—the cascade of autoantibody production that often begins many years before SLE diagnosis (N. Engl. J. Med. 2003;349:1526-33).
The SLE patients tended to mount their immune response in a characteristic way. The first antibody to appear was directed against the viral protein EBV nuclear antigen–1 (EBNA-1). This antibody cross-reacted with the lupus-associated autoantigens Ro and Sm in lupus patients. Then, through molecular mimicry with self-antigens and the process of B-cell epitope spreading, the cross-reactive antibodies targeted non–cross-reactive autoepitopes and spread to a widening array of autoantigens, with generation of pathogenic autoimmunity. (Healthy individuals mount a far more limited immune response to EBV and EBNA-1 and do not produce cross-reactive antibodies.)
EBNA-1 is both immunogenic and antigenic. The researchers showed that nearly all military personnel who had SLE were seropositive for EBNA-1, whereas 12% of the military controls were not. This is consistent with the notion that, to lay the groundwork for SLE, an individual not only has to be infected with EBV but must also mount an immune response to EBNA-1.
“The host response to EBNA-1 is critical in the pathogenesis of SLE,” stressed Dr. Harley, who is also the George Lynn Cross Research Professor at the University of Oklahoma Health Sciences Center. EBV is a strong candidate for having a pathogenic role in SLE because it is a ubiquitous infection, with 95% or more of adults in the general population being seropositive. The virus persists in the host for life as a latent infection with a viral reservoir in B-lymphocytes. Low levels of lytic virus emerging in latency provoke persistent immune stimulation.
“It's remarkable that about 7% of our T-cell repertoire is directed against EBV,” the immunologist continued.
Because EBV infection is so common in adults, Dr. Harley and coworkers studied seropositivity rates in a series of children and teens. The investigators demonstrated that 116 of 117 individuals with SLE (average age, 15.6 years) were seropositive against EBV, compared with two-thirds of controls. Indeed, EBV seropositivity was associated with a 50-fold increased probability of lupus (J. Clin. Invest. 1997;100:3019-26). Other investigators have independently replicated this work in more than half a dozen cohorts.
There are additional differences between SLE patients and healthy controls in terms of response to EBV infection. Patients with SLE have 15- to 40-fold higher EBV loads, 10-100 times more EBV-infected B cells, and defective CD8 T-cell responses against EBV.
In an interview, Dr. Harley predicted that as EBV's role in the pathogenesis of SLE becomes more fully understood, “there will be new opportunities for treatment.” For example, once the specific T-cell responses to the virus in SLE are better grasped, clinical studies of anti–T-cell therapies directed at those mechanisms will be appropriate.