12,400 Sessions, Zero Adverse Events
The registry behind the headline number — what we measured, how we adjudicated, and what we still do not know.
Zero is a load-bearing number. It is also the kind of number that, in clinical medicine, deserves immediate suspicion. So before we put it on the front page of our protocol documentation, we want to be very clear about what we counted, what we did not, and what the limits of the claim actually are.
1. What we counted
Every clinical session run under the BSerum protocol between Q3 2023 and Q4 2025 — 12,400 in total, across 1,847 patients in three Geneva clinics and one partner site in Zurich. Each session was monitored under the Guardian telemetry stack, and every adverse event candidate was adjudicated within 48 hours by a physician who was blinded to which arm of the protocol the patient was on.
2. How we adjudicated
We used the World Allergy Organization grading system for hypersensitivity (Grade 1-4). A trained adjudicator reviewed the full session telemetry, the clinician's notes, and the patient's 72-hour post-session diary. Disagreements between adjudicators were resolved by a third reviewer; we required unanimous Grade 0 for a session to count as event-free.
| Observation | WAO Grade | Counted as event? |
|---|---|---|
| Injection-site erythema > 4 h | 0 | No — pre-disclosed |
| Transient lightheadedness < 30 min | 0 | No — pre-disclosed |
| Localised pruritus | 0 | No |
| Generalised urticaria | 1 | Yes — none observed |
| Angioedema | 2 | Yes — none observed |
| Bronchospasm or hypotension | 3 | Yes — none observed |
| Anaphylaxis | 4 | Yes — none observed |
3. What we did count as observations
Injection-site erythema lasting more than four hours occurred in 8.2% of sessions. Transient (sub-30-minute) lightheadedness in 1.9%. Mild localised pruritus in 3.4%. These are expected and pre-disclosed; none rose to Grade 1 hypersensitivity under WAO criteria, but we report them because the patient lived them.
4. What we still do not know
Three things. First: long-term (>5-year) immunological consequences in patients who have completed multiple full protocols. We are following the cohort. Second: how the protocol performs in patients over 75, an underrepresented group in the registry. Third: interaction with biologics — particularly IL-4Rα antagonists — which we currently exclude as concomitant therapy.
"A zero is only as honest as the adjudication behind it. We publish ours."
5. The registry is open
Patients in the BSerum protocol have access to their own session telemetry. Partner clinics receive de-identified aggregate dashboards. Academic groups can apply for query-level access under a data use agreement. We believe this is the only way a number like zero earns its keep.