DocLogic,
You are thinking about this the right way. The time advantage question is the correct question to ask, and you are right that the T cell expansion capability matters strategically. Where I think you can push your own analysis one step further is on where the time advantage actually lives.
A single-cell multiomics paper posted on bioRxiv May 2 (DePuyt, Mailliard, Kalinski et al., 149,535 cells from 7 participants) just showed that the alphaDC1 preparation the field has treated as homogeneous for twenty years is actually three distinct subpopulations. Only 0.5-2% of alphaDC1 cells produce IL-12p70 (Cluster E, IRF8-high). One subpopulation (Cluster D) actively recruits Tregs through CCL17/CCL22, working against the therapeutic intent inside every unsorted preparation ever used in a clinical trial. They identified surface markers to sort these populations at baseline using existing GMP protocols.
This changes the expansion math. Your assumption, which is reasonable, is that you need a partner like Ori for massive T cell expansion because the instruction event is inefficient. With an unsorted preparation where 98% of DCs are non-producing or counter-instructing, that is correct. You do need scale to compensate. But sort for Cluster E before the co-culture step, and every T cell contacts a functionally competent IL-12p70-producing DC. The yield of correctly educated T cells goes from a diluted fraction to essentially the entire output. When every cell carries the dual-key TCR/NKR recognition Kalinski described on April 29 (two-hundred-fold enhanced sensitivity through stress-ligand co-engagement), potency per cell rises and the dose requirement drops.
Published data show 1,000-fold T cell expansion is achievable within EDEN’s perfusion environment with preserved memory phenotype and antitumor capacity. The EDEN patent describes DC-T cell co-culture with chemokine gradient focusing and T cell harvest in the same closed system. The scale problem Ori would solve becomes dramatically smaller when you solve the potency problem at the source.
Here is the other piece to consider. T cell expansion is a commodity capability. Ori can do it. Miltenyi’s CliniMACS Prodigy can do it. Lonza’s Cocoon can do it. Any GMP suspension bioreactor can expand T cells. What none of them can do is the instruction event. That is exclusively NWBO’s through the Kalinski licenses (confirmed as paid NWBO consultant in the DePuyt paper’s COI statement). The competitive moat is not in expansion. It is in the educated T cell that comes out of the EDEN co-culture, carrying polyclonal multi-antigen specificity, dual-key recognition, and memory phenotype that no CAR-T or TIL platform replicates. Splitting the manufacturing chain at that point hands a third party access to the most valuable intermediate in the workflow.
You are absolutely right that Annalisa Jenkins is positioned to put this in proper perspective. The time-to-market question, whether EDEN’s co-culture chambers can be optimized for the Cluster E workflow fast enough versus partnering for expansion, is exactly the kind of strategic call her experience equips her to make. But the biology is now pointing toward keeping everything inside one closed system, because the DePuyt paper just showed how to solve the problem Ori would address before it arises.
Your instinct that NWBO’s IP in DC-mediated T cell activation is the core asset is exactly right. That is the moat. The question is whether you protect the moat by keeping the full workflow in-house or create a dependency on a partner for a commodity step. I think you know which way that points.
Best wishes.
Market Data 
Markets 
AI
