Cell-Derived Nanoparticles Could Speed and Rescue CAR T Cell Manufacturing

Why this matters: CAR T therapy has transformed treatment of some cancers and autoimmune diseases, but manufacturing is a bottleneck. The authors cite 5–30% of patients who cannot receive therapy because their T cells fail to expand, or because their disease progresses during the long vein-to-vein interval. Expansion during manufacturing has also been linked to clinical outcomes. The standard CD3/28 magnetic beads deliver only two signals, use antibodies instead of natural ligands, and must be removed before infusion, which costs cells and limits rapid protocols.

What was studied: The team engineered K562-based artificial antigen-presenting cells (aAPCs). They first replaced the cumbersome antibody-loaded CD64 system with a 'ready-to-use' anti-CD3 (OKT3-derived) scFv displayed on the cell surface. They then added CD86, 4-1BBL, IL-7 and IL-15 with IL-15Rα, which lets IL-15 be transpresented to T cells. To avoid using intact irradiated tumor cells, they generated cell-derived nanoparticles (CDNPs) by two rounds of nitrogen cavitation followed by purification and concentration. For this they switched IL-7 to a membrane-bound form (K.a3.86.BBL.m7.15).

Key results (from the portion of the text available): The scFv aAPCs expanded primary T cells as well as antibody-loaded aAPCs after 9 days, and both beat CD3/28 beads. Multi-signal aAPCs expanded T cells further. Initial division rates were similar across substrates, which suggests the advantage comes from later proliferation or survival. CAR transduction efficiency was similar across substrates. The full-signal aAPC preserved more highly polyfunctional CD4 and CD8 CAR T cells, defined as producing at least three cytokines. CDNPs carrying membrane-bound IL-7 or IL-15Rα/IL-15 activated STAT3 and STAT5 in T cells, while wild-type K562 CDNPs and beads did not, which shows the cytokines stayed functional. CDNPs drove faster entry into the cell cycle than CD3/28 beads plus IL-7/IL-15: about 20% of CD4 and 40% of CD8 T cells had divided once by day 2, versus a negligible fraction with beads. This allowed Day 0 transduction, with activation and lentiviral transduction done together. For CDNP-activated cells this gave higher transduction than Day 1 transduction, the reverse of what was seen with beads.

Implications: According to the abstract, the remaining sections report that CDNPs expanded T cells from patients whose products could not be made with standard approaches. CDNP-derived CAR T cells controlled tumors in humanized mouse models. In the phase I trial NCT04684563 in CD19+ malignancies, cGMP-compatible CDNPs supported 3-day manufacturing of IL-18–expressing CD19 CAR T cells, with higher cell recovery, durable clinical responses and no unexpected toxicities. Together these point to CDNPs as a scalable, off-the-shelf alternative to beads and cell-based aAPCs.

Caveats: This is a non-peer-reviewed preprint. The text supplied ends partway through the second results section, so the details of the patient-sample rescue, the mouse experiments and the clinical trial come only from the abstract. The trial is early phase and small, so efficacy and safety claims need confirmation in larger, controlled studies.