What Makes It Hard to Scale cGMP Manufacturing for Cell Therapies?

Cell therapy cGMP manufacturing scales differently from small molecule or biologics. Four constraints determine whether a program scales smoothly from Phase 1 to commercial: donor pool depth, equipment and consumable sourcing, regulatory variance across sites, and analytical method scale-up. Sponsors who treat scaling as “buy more bioreactors” miss the upstream constraints that actually break programs.

The four scaling bottlenecks

1. Donor pool depth A Phase 1 program might use 5–10 donors. A Phase 2 might use 50. A commercial program might need 500+. Most cell sourcing vendors do not have donor pools deep enough to support this without donor-to-donor variability that breaks process consistency. Phase 1 looks great. Phase 3 looks like a different drug. Fix: source from vendors with 1,000+ qualified donor pools and recall capability. The pool depth has to be there before you sign the MSA, not promised for later. 2. Equipment and consumable sourcing Cell therapy equipment is specialty. Bioreactors, closed-system processing devices, automated cryopreservation, controlled-rate freezers. Lead times for specialty equipment can be 12–18 months. Consumables (single-use bags, isolation reagents, media) have their own supply chain risks. A program scaling from Phase 1 to Phase 3 typically discovers equipment sourcing constraints 6 months too late. Fix: equipment scaling plan written at Phase 1, not Phase 2. 3. Regulatory variance across manufacturing sites A program manufactured at one US site may be fine for a US-only Phase 1 trial. Adding EU sites for Phase 2 introduces EMA-specific variations in donor eligibility, GTP, GMP. Adding Asia sites adds more. Each site needs its own regulatory comparability and inspection readiness. Fix: regulatory site selection scoped at Phase 1 with the global trial roadmap in view. 4. Analytical method scale-up QC methods that work at Phase 1 with N=10 samples per month do not work at Phase 3 with N=500 per month. Flow cytometry capacity, sterility testing turnaround, mycoplasma testing throughput, potency assay scale all become bottlenecks. Fix: analytical method scale and capacity modeling at Phase 1.

The constraint sponsors most often miss

Of the four bottlenecks, donor pool depth is the one most often missed because it does not show up until later phases. A Phase 1 program with 8 donors looks indistinguishable across donors. The phenotype is consistent, the cell yield is consistent, the post-thaw viability is consistent. The sponsor concludes the process is robust. At Phase 2, with 50 donors, donor-to-donor variability emerges. Some donors give 10 billion cells per leukopak; others give 6 billion. Some donors have higher CMV status; some have specific HLA distributions that affect downstream T cell expansion. Fixing this at Phase 2 means re-running comparability and updating the BLA-supporting comparability package. Most sponsors do this. It costs time. At Phase 3 or commercial, with 500+ donors needed, the underlying donor pool depth at the vendor matters more than the process. Vendors with shallow donor pools force the sponsor to either rotate vendors (regulatory comparability nightmare) or accept variability that breaks the spec. Fix: source from a vendor with 1,000+ qualified, recallable donors from Phase 1 onward.

What this means for vendor selection at Phase 1

The cell sourcing vendor selection at Phase 1 has more downstream consequences than sponsors typically appreciate. The vendor you choose at IND-enabling determines whether donor pool depth, RUO-to-cGMP continuity, and recall capability are present when Phase 3 needs them. Vendors with single donor pools serving both RUO and cGMP under one quality system avoid the donor requalification cascade. Vendors with FDA-registered apheresis subsidiaries (rather than partnered apheresis) have direct control over collection scaling. Vendors with multiple processing sites have geographic flexibility as your trial footprint expands.