JAK Inhibitors in Alopecia Areata: Safety, Goals and Why Stopping Treatment Backfires

Only a small percentage of patients with alopecia areata maintain their response after stopping JAK inhibitor therapy — and that single data point drives nearly every management decision Dr. Ungar walks through in his Case 3 follow-up review for Dermatology Times.

The patient in question is a normal CYP2C9 metabolizer. Her baseline workup is clean: normal CBC, normal TSH, normal ferritin, negative tuberculosis screening, and immunizations up to date. She is, in other words, cleared to start. The question is what to start, what to promise, and how long to keep going.

What does a realistic treatment goal look like?

Ungar begins where the patient begins: a conversation about what she hopes for and what is realistic. Severe disease may take longer to regrow. Not everyone responds to the first drug, and regrowth takes months, not weeks.

His overall goal is a patient who feels better and carries a minimal psychosocial burden — ideally with full regrowth. That framing matters, because the data show ongoing therapy is what sustains response; only a small share of patients hold their gains off treatment.

Takeaway: set expectations at the first prescription, not after the first disappointment.

How should clinicians counsel after a first therapy fails?

Counseling should start at initiation, not at failure. Ungar recommends cautious optimism: explain upfront that regrowth takes months and that not everyone responds to the first agent.

The critical message, he says, is that a failed first treatment does not predict failure of the next. That hope helps patients stay positive through the waiting period.

Takeaway: one failed drug is not a failed course of treatment.

Is the safety signal a reason to hesitate?

The class-wide safety concerns, Ungar notes, come from tofacitinib — a drug studied in a different disease, not alopecia areata. Translating those signals wholesale to this population overstates the risk.

These are immunosuppressive medications, and they are not zero-risk. But alopecia areata data across thousands of patients suggest the risks are in line with background rates of myocardial infarction, deep vein thrombosis, and malignancy. Any increase appears minimal, so the risk-benefit balance is favorable.

The differences among the three available agents are small. Access matters most, Ungar argues — the drug a patient can actually get and stay on is usually the right one.

For this specific patient, her normal CYP2C9 status makes deuruxolitinib a reasonable starting choice.

Takeaway: real-world alopecia areata data, not tofacitinib's label, should anchor the safety conversation.

What does long-term management require?

Stopping after regrowth carries a high likelihood of lost response. That relapse risk is why Ungar treats continuation, not cessation, as the default plan.

His monitoring schedule is straightforward. He checks labs roughly every three months at first, then spaces them out as the patient stabilizes. Regular follow-up visits stay on the calendar, and he encourages patients to call between visits about flares.

That open door matters. Fear of relapse persists even after hair returns, and patients need a clear channel to report changes before a flare becomes a crisis.

Takeaway: monitoring cadence plus open access keeps durable response achievable.

What comes next?

The series continues with "Patient Perspective on Alopecia Areata: Sudden Hair Loss and Delayed Diagnosis," in which a patient with alopecia areata universalis recounts how her hair loss began and why diagnosis took months — a reminder, before any prescription is written, of how long these patients often wait before they reach a clinician who recognizes the disease.