Reassurance on Live-Virus Shingles Vax With Anti-TNFs

ATLANTA -- If you've held off on recommending the live-virus varicella zoster vaccine (Zostavax) for patients on tumor necrosis factor (TNF) inhibitors, perhaps a randomized, double-blind trial reported here will change your mind.

With 317 patients on anti-TNF agents receiving the vaccine and 310 getting placebo shots, no one in either group developed varicella (chicken pox) infections within 6 weeks, none had shingles attacks, and rates of disease flare didn't differ between groups, reported Jeffrey Curtis, MD, MPH, of the University of Alabama at Birmingham, at the American College of Rheumatology (ACR) annual meeting.

Moreover, the vaccine appeared to be effective during the limited time period analyzed, with increases in IgG titers (by 31% more than placebo on average, 95% CI 11%-55%) and interferon-gamma (30%, 95% CI -10% to 90%) consistent with some degree of protection against shingles, Curtis said.

These immune responses were "reasonable" but "not amazing," he said at an ACR press conference, noting that they were weaker than typically seen in otherwise healthy older adults.

Overall, he said, the findings suggest that "theoretical concerns" about infections developing from live-virus vaccinations -- enshrined in rheumatology organizations' guidelines and, in some cases, product labels -- aren't supported in practice.

Those concerns arose from several considerations that weren't without merit, Curtis explained. Although live-virus vaccines such as Zostavax use attenuated strains, they are still capable of replication and their safety is based on testing in people with intact immune systems.

Patients with autoimmune diseases such as rheumatoid arthritis (RA) are known to be at increased risk for zoster virus reactivation, and certain of the treatments they frequently take, such as corticosteroids and JAK inhibitors, increase that risk even further. Curtis said anti-TNFs probably do not have that effect -- but nevertheless the concerns about runaway infection with vaccine strains are out there.

To bring evidence to bear, the National Institute of Arthritis and Musculoskeletal and Skin Diseases (NIAMS) agreed to fund a randomized trial, called VERVE. Patients had to be at least age 50 and receiving TNF inhibitors for any type of autoimmune/inflammatory disease. They had to have been born in the U.S. or lived in the country for at least 30 years, the assumption being that all such individuals have latent zoster virus infection. Patients who previously received Zostavax or had active malignancy were excluded.

Among the 627 enrolled, mean age was 63 and two-thirds were women. About 60% had RA, one-quarter had psoriatic arthritis, and the remainder were spread among other conditions such as inflammatory bowel disease and ankylosing spondylitis. Anti-TNFs in use were primarily adalimumab (Humira), infliximab (Remicade), and etanercept (Enbrel).

Six weeks was chosen as the follow-up time for varicella infection (the primary outcome) because it normally occurs quickly on exposure to the virus, Curtis said.

Twelve participants, including seven in the Zostavax group and five getting placebo, developed symptoms that might have been varicella infection. However, in the eight for whom PCR testing was performed (four in each arm), neither the vaccine strain nor the wild-type varicella virus was detected.

The only notable adverse effect differing between groups was injection site reactions, which Curtis said might have compromised blinding. Overall, 18 patients developed serious adverse events during 12 months after vaccination, of which 15 occurred prior to month 6; these were equally balanced between study arms.

Curtis noted that IgG titers and interferon-gamma levels continued to be monitored through 12 months in the Zostavax group. In 98 participants with data, the antibody titers remained elevated from baseline, whereas interferon levels had dropped below baseline in the majority.

Going forward, he added, shingles attack incidence will be tracked through administrative data (e.g., Medicare claims, physician records) in all participants.