Sequencing a Zoster Vaccine Against an Urgent JAK Inhibitor Start
A rheumatoid arthritis flare needs a JAK inhibitor now, in a patient whose zoster history and drug class both raise real reactivation risk before a two-dose vaccine series can finish.
Helen T., a 58-year-old woman, spent thirty years as a bedside nurse before her rheumatoid arthritis made twelve-hour shifts on her feet impossible, and though she officially retired last year, she still volunteers twice a month doing intake at a free clinic near her house — work she describes as the part of nursing she never wanted to give up. Her RA has been managed on methotrexate and folic acid for the past twelve years, adequately for most of that time, but over the last four months her hands and both knees have flared badly enough that she can no longer manage a full clinic shift, and her rheumatologist has recommended moving to upadacitinib, a JAK inhibitor, given the methotrexate failure. A decade ago she had a single, uncomplicated episode of shingles across one thoracic dermatome, treated with valacyclovir at the time without complication.
That single prior episode matters more than it might seem to, on top of the drug class itself. JAK inhibitors carry one of the more consistently replicated safety signals in modern rheumatology — a consistently elevated risk of herpes zoster reactivation compared with TNF inhibitors. ORAL Surveillance, the randomized safety trial that produced the sharpest version of this finding, reported zoster hazard ratios near three against a TNF inhibitor; real-world claims cohorts land nearer a doubling. The trial is worth reading against her rather than around her: it enrolled patients aged fifty or older with at least one cardiovascular risk factor, and it studied tofacitinib, not upadacitinib. She is fifty-eight, so she sits inside its age criterion, and the zoster signal has since been reproduced across the JAK class; but the risk-factor entry criterion and the specific drug are both extrapolations, and her own strongest risk factor — a prior zoster episode — is one the trial did not select for at all. The recombinant zoster vaccine is a non-live subunit product, which means — unlike its discontinued live predecessor — it isn't contraindicated in someone about to be immunosuppressed; the open question is timing, not eligibility. Vaccine immunogenicity during active JAK-inhibitor therapy is somewhat reduced relative to giving it beforehand, and that reduction is the only thing genuinely at stake in the timing question. Nobody is deciding whether she gets vaccinated. They are deciding how much of the vaccine's effect she is going to forfeit, and there is no version of this month in which she forfeits none of it.
Fitting a two-dose vaccine series into a disease that won't wait
I want to start upadacitinib now, not after a vaccination series that itself takes months to complete. Her disease control has been slipping for four months, and RZV's own dosing schedule — two doses, typically two to six months apart — means even if we delayed starting the drug entirely, we wouldn't have full vaccine protection in place before the immunosuppression begins anyway. Waiting doesn't actually solve the sequencing problem the way it sounds like it should.
The arithmetic is yours and it holds: two doses, two to six months apart, against a drug she needs this month. The full series cannot finish first. I'm not proposing we pretend otherwise.
But there's a meaningful difference between 'the full series can't finish first' and 'none of it should happen first.' Getting just the first RZV dose in before upadacitinib starts, even by a week or two, gives that dose a real immunogenicity advantage over giving it during active JAK-inhibitor exposure, and she's not a low-risk bystander here — a prior zoster episode is itself one of the strongest individual risk factors for recurrence, on top of the class risk — ORAL Surveillance put the hazard ratio near three, claims data nearer two. A short, partial delay — first dose now, upadacitinib starting right behind it rather than months later — gets her most of the immunologic benefit without asking her joints to wait for a full series.
Agreed: first RZV dose given today, upadacitinib started one week later rather than delayed for the full two-dose series, and the second vaccine dose scheduled during ongoing JAK-inhibitor therapy per the standard interval, accepting somewhat reduced immunogenicity for that second dose rather than delaying treatment further to optimize it.
Not raised as a live disagreement but worth naming: neither voice suggested antiviral prophylaxis, such as low-dose valacyclovir, as an alternative or supplement to vaccination — an approach with less direct trial support in this specific population than in other iatrogenically immunosuppressed groups, but occasionally used in practice for patients judged especially high-risk. It wasn't raised for her, and the visit ended without either voice being asked why not.