JAK inhibitors are effective drugs for rheumatoid arthritis — but they carry a known downside: a higher risk of shingles. That makes a practical question surprisingly important: should patients get the shingles vaccine before starting the drug, or start treatment right away?

A randomised trial, STOP-HZ, published in Annals of Internal Medicine, put that question to the test across 12 sites in Japan.

Why JAK inhibitors raise shingles risk

The connection is mechanistic rather than incidental.

Shingles is reactivation of the varicella-zoster virus, which persists in nerve cells for life after chickenpox. It is held dormant by continuous immune surveillance, principally by T cells, which is why reactivation rises with age as that surveillance weakens.

JAK inhibitors block signalling pathways used by many cytokines, including interferons, which are central to controlling viral infection. Suppressing that signalling reduces exactly the surveillance keeping the virus latent.

The effect is consistent enough across the drug class to be considered a class effect rather than a property of any one agent, and it is more pronounced than with most other RA treatments.

Why the vaccine can be given at all

The recombinant zoster vaccine is not a live vaccine, which is what makes the question a timing question rather than a prohibition.

Live vaccines are generally contraindicated in immunosuppressed patients, because a weakened but replicating virus can cause disease when defences are impaired. An older live shingles vaccine could not be given to people on these drugs.

The recombinant version contains a viral protein with an adjuvant and nothing capable of replicating, so it is safe under immunosuppression. What it may be is less effective — a vaccine works by provoking an immune response, and a suppressed immune system responds less strongly.

The trade-off it found

The results point to a genuine trade-off rather than a clear winner.

Delaying the JAK inhibitor until after completing the two-dose vaccine produced a stronger early antibody response. Starting the JAK inhibitor immediately gave patients faster relief from arthritis symptoms. By 12 weeks, antibody responses were similar between the strategies.

The convergence at 12 weeks is the most useful finding. It suggests the drug slows the response rather than preventing it — so the cost of starting treatment immediately is a period of weaker protection rather than a permanently inferior result.

What the untreated interval costs

The other side is easy to understate. Completing a two-dose vaccine schedule takes months, since the doses are separated by an interval.

For someone with active rheumatoid arthritis that is months of uncontrolled inflammation. Joint damage in RA accumulates during active disease and is irreversible, so delay has a permanent cost of its own — which is why modern practice emphasises early aggressive control.

What it means for patients

The takeaway is a judgment call rather than a rule: weigh how urgently the arthritis needs controlling against the desire for strongest early vaccine protection.

For someone with severe, active RA, faster symptom control may win. For someone at high shingles risk with milder disease, finishing the vaccine first may be preferable.

Important limits

This was a small, short, exploratory study — not powered to declare one approach best — and it did not track actual shingles cases, only antibody responses.

That gap matters. Antibody levels are a convenient surrogate and protection against shingles depends substantially on cell-mediated immunity, which antibodies do not measure. A difference in antibody titre does not straightforwardly translate into a difference in who gets shingles.

Why this question keeps recurring across immunology

Vaccine timing around immunosuppressive therapy is a general problem in rheumatology, and shingles with JAK inhibitors is one instance of a pattern.

The same tension applies to influenza and pneumococcal vaccination before starting biologics, to hepatitis B vaccination before B-cell depleting therapy, and to the general principle that vaccines given during immunosuppression produce weaker responses.

B-cell depletion is the sharpest case. Those drugs remove the cells that produce antibodies, and vaccine responses after treatment can be profoundly blunted for many months — making pre-treatment vaccination considerably more important there than the present trial suggests it is for JAK inhibitors.

Guidelines accordingly recommend reviewing and updating vaccination status before starting any immunosuppressive therapy where disease activity permits, which is the general principle this trial refines rather than establishes.

What it adds is quantification for one specific pairing: the disadvantage of not waiting appears to be temporary. That is more useful than a general recommendation, because it tells a clinician what is actually being traded away when the arthritis cannot wait.

It is a useful signal to inform a conversation, not a definitive protocol. General information, not medical advice.