Faster first-in-human T-cell engager studies

Three ways to reach a biologically active dose sooner

Index of the project on speeding up Phase 1 and 1b T-cell engager studies: intrapatient dose escalation, a healthy volunteer substudy, and a MABEL starting dose, with the documents working each one out.

A first-in-human T-cell engager (TCE) study starts orders of magnitude below the biologically active dose and climbs toward it one cohort at a time. This project collects the ways of climbing faster, and works each one out far enough to say where it applies and what it saves.

Approaches

The three approaches, and how far each has been taken here

Approach	What it changes	Where it applies	Status here
1. Intrapatient dose escalation, with single-patient cohorts	Climbs the ladder within one patient rather than one cohort per rung	Any indication; the value rises with uncertainty about the active dose	Specified as a simulation, not built
2. Healthy volunteer escalation	Finds the active dose in volunteers before the first patient is enrolled	Autoimmune indications with a B-cell or plasma-cell target; closed in oncology	Registry search done, three studies found, none reported
3. MABEL starting dose	Starts higher, on a better prediction of the active dose	Any indication	Out of scope, since it is widely practised; entry 2 of the references is the pointer

The three are not exclusive. A better starting dose shrinks the gap the other two cross, and a volunteer study, where it is open, is a starting-dose method with human data behind it.

Documents

• Working specification: plasma-cell targets. Approach 1. The design, the PK and pharmacodynamic models, the endpoints, and the implementation sequence for the intrapatient escalation simulation, for an engager that kills plasma cells, with serum free light chains as the stopping marker. Its In Brief section summarizes the rest on one screen, and its Section 1 records the open decision between approaches 1 and 2.
• Specification 1: B-cell targets. The same design with peripheral B cells as the marker. Frozen on 2026-10-07; edits go in the working specification.
• T-cell engagers in healthy volunteers. Approach 2. The three registered volunteer studies, all autoimmune and all since December 2025, what a volunteer study can measure, the translation to patients, and a speed estimate under stated assumptions.
• References. What to read and in what order, for approaches 1 and 2, and a marker on every source recording whether its use here has been checked. Approach 3’s pointer is entry 2.
• MCLA-117 notes. The published Cycle 1 escalation scheme the specification cites as precedent, its dose tables, and four features of it that bear on the design.
• Simon 1997 reading notes. What the accelerated titration designs paper reports when intrapatient escalation is switched on and off, with every number sourced to a page.

Code

• make_escalation_figure.py draws mcla117-escalation.svg, the Cycle 1 escalation figure in Section 6 of the specification. Plain Python with no plotting library, so the site build never runs it. Run it after changing the dose table it carries.