In Vivo Gene Therapy Delivery: Current Platform Architectures and Engineering Constraints

Search results lack papers or pilots from the past week. No granular data emerges on tropism prediction, dosing simulations, immune clearance kinetics, or six-month validation cycles. What we have are company announcements and reviews, months to a year old, teasing preclinical outcomes without the physics of delivery.

Available Preclinical Data

Azalea Therapeutics TRAC-CAR platform generates CAR T cells in vivo in primates. Dual vectors pair T cell-targeted Enveloped Delivery Vehicles with Cas9 RNP and AAV carrying a promoterless CAR template. Editing hits target, CAR T cells expand, B cells deplete deeply. No numbers on payload survival or clearance rates surface.

AAV vectors dominate muscle, lung, CNS delivery, capped at 5 kb cargo. CRISPR Therapeutics splits Cas9 and guide RNA across AAVs for heart disease genes.

LNPs push as non-viral alternatives, handling mRNA, DNA, siRNA, proteins without size limits, liver-focused now.

Critical Gaps in Available Data

No quantitative immune clearance shows up. Payloads often die before target, 80% lost to clearance in wrong setups, but sources stay silent. No tropism models. No dosing workflows. No validation timelines. Viral versus non-viral kinetics? Absent. Computational design slams into in vivo reality without integration paths detailed.

Senior engineers know the stall: you model tropism in silico, predict dosing, run mice. Six months later, immune sweep guts 80% of payload in non-human primates. Iteration resets. Failure looks like endless vector tweaks chasing ghosts, teams burning cycles on immunogenicity surprises instead of scalable stacks.

The Systems Integration Question

In vivo diverges hard from ex vivo. Ex vivo owns controlled bioreactors and QC gates. In vivo fights tissue chaos, immune spikes, variable pharmacokinetics, no mid-course corrections. Sources nod at this split, but no decision OS rises to map it. Ex vivo stacks fail here: too rigid for biological noise.

Recent ASGCT 2025 reports hint at cell therapy abstracts, viral vectors in play, yet details stay gated.

Data drought persists. For rigor, pull ASGCT proceedings, dosing studies, clearance kinetics.

What stacks have you rigged to bridge compute and in vivo? Notes from the trenches welcome.