Reviewed by Sarah Alter, Ph.D. — Scientific Affairs, OrganaBio. 15 years of immunology research spanning autoimmunity, cancer, and infectious disease. University of Miami Miller School of Medicine. Registered Patent Agent.
The Starting Material Problem Allogeneic Programs Discover Too Late
Allogeneic cell therapy programs — off-the-shelf CAR-T, NK cell therapies, Treg products, and others using donor-derived rather than patient-derived cells — carry a starting material challenge that autologous programs avoid entirely: every donor cell that goes into a patient is, by definition, foreign. Alloreactivity isn’t a side effect to manage. It’s the central immunological problem.
The starting material decisions that determine alloreactivity risk get made early in development, when the donor pool is selected, when HLA matching criteria are set, and when the manufacturing process that determines how donor immune cells are characterized and processed gets locked in. Programs that treat these as late-stage clinical questions consistently reach Phase I with alloreactivity profiles they didn’t anticipate from their preclinical data.
This piece covers the HLA architecture considerations that determine allogeneic starting material strategy, the role of mixed lymphocyte reactions (MLR) in donor qualification, and how OrganaBio’s HLA-typed donor pools integrate into allogeneic development programs.
HLA Architecture: What You’re Actually Managing
Alloreactivity — a recipient’s immune system recognizing and rejecting donor cells — is driven primarily by HLA mismatches. The recipient’s T cells detect donor HLA molecules as foreign and mount a response. This is the rejection mechanism that organ transplant medicine has managed for decades, and allogeneic cell therapy faces the same fundamental biology at smaller scale.
The HLA system has three relevant layers for allogeneic cell therapy:
HLA Class I (A, B, C): Expressed on nearly all nucleated cells. Donor HLA Class I molecules present peptides to recipient CD8+ cytotoxic T cells. Mismatched Class I alleles are the primary driver of cytotoxic T cell-mediated rejection. For off-the-shelf products, avoiding major Class I mismatches with recipient populations requires understanding the HLA frequency distribution in your target patient population.
HLA Class II (DR, DQ, DP): Expressed on antigen-presenting cells. Donor HLA Class II mismatches drive CD4+ helper T cell alloreactive responses that support B cell antibody production (donor-specific antibodies) and provide help for CD8+ cytotoxic responses. Many autoimmune indications — rheumatoid arthritis, multiple sclerosis, celiac disease, Type 1 diabetes — have strong HLA Class II associations. The patient population you’re targeting has a non-random HLA Class II distribution, and your donor HLA profile needs to account for it.
KIR-HLA Interactions: NK cells use KIR receptors to detect HLA-C expression on target cells. Missing self — donor cells that lack recipient HLA-C ligands — triggers NK cell-mediated killing. For NK cell therapy programs, KIR-HLA compatibility between donor NK cells and recipient cells determines both efficacy (KIR mismatch may be exploited for GvL effect) and safety (NK cells must not be eliminated by recipient immune surveillance before they can act).
The Mixed Lymphocyte Reaction as Donor Qualification Tool
The mixed lymphocyte reaction (MLR) is the standard assay for measuring alloreactivity between donor and recipient immune cells. In a one-way MLR, recipient PBMCs are cultured with irradiated (or otherwise inactivated) donor PBMCs. The degree of proliferation and cytokine production by recipient T cells measures the alloreactive response to the donor HLA profile.
For allogeneic cell therapy donor qualification:
Low-alloreactivity MLR response is a qualifying criterion for some programs, particularly Treg products where host rejection of the regulatory cell product before it can suppress inflammation would eliminate the therapeutic effect. Donor selection that minimizes MLR stimulation of target patient populations reduces the probability that the therapeutic product is cleared before it acts.
High-alloreactivity MLR may be exploited in oncology contexts. For NK cell therapies targeting hematological malignancies, KIR-HLA mismatch that drives NK cell activation against recipient cells — including leukemia cells — is the intended mechanism. Here, the same alloreactive response that causes rejection risk in autoimmune programs is the therapeutic lever in oncology.
MLR data from disease-state donors reflects the actual alloreactive challenge your therapy will face. If your allogeneic product will be administered to RA patients, the MLR stimulating population should include RA donor PBMCs with shared epitope HLA, not just healthy donors with average population HLA frequencies. The alloreactive T cell repertoire in RA patients includes clones specific for shared epitope HLA molecules — a preclinical model using only general population healthy donors will not capture this.
Off-the-Shelf Strategy and the HLA Frequency Problem
Off-the-shelf allogeneic products aim to be usable across a broad patient population from a single or limited donor pool. The fundamental tension: HLA diversity in human populations means no donor is HLA-matched with most patients. The strategies for managing this tension fall into three categories.
Universal donor engineering: HLA knockout via CRISPR (B2M disruption to eliminate Class I, CIITA disruption to eliminate Class II) to make donor cells invisible to recipient T cell recognition. These approaches eliminate direct alloreactivity at the cost of creating NK cell activation risk (missing self) and may require additional modifications (HLA-E overexpression, CD47 overexpression) to evade NK surveillance. Starting material for these programs needs to be processable through gene editing with high efficiency and viability, and the donor HLA profile at baseline matters for initial characterization even if it will be disrupted.
HLA-matching of donors to patient subpopulations: Selecting donor cells that match the HLA haplotypes most common in the target indication’s patient population. For autoimmune indications with skewed HLA distributions — DR4-enriched RA populations, DRB1*15:01-enriched MS populations, DQ2/DQ8-enriched celiac and T1D populations — this means intentionally selecting donors who carry those alleles to reduce Class II mismatch frequency. OrganaBio’s disease-state donor pools, characterized for HLA at collection, provide a practical resource for programs that need to select donors based on HLA compatibility with specific patient populations.
Regulatory T cell products with tolerance induction: Treg therapies that aim to induce antigen-specific or broader tolerance may not require strict HLA matching if the mechanism involves generating regulatory dominance over the alloreactive response rather than avoiding it. But this approach requires preclinical alloreactivity data demonstrating that the Treg product can suppress MLR responses in the relevant disease-state donor context.
Donor Pool Size and the Lot Coverage Problem
A practical constraint rarely discussed early enough: allogeneic manufacturing from a small donor pool creates lot coverage limitations. If the manufacturing process requires a specific donor or specific HLA profile, scalability depends on the depth of that donor pool. Single-donor dependency — common in early development when manufacturing from one characterized “super-donor” — creates a supply chain vulnerability that only surfaces when that donor is unavailable or their cells no longer pass quality specifications after repeated expansion cycles.
Building a characterized, HLA-typed donor pool from early development gives programs:
- Lot-to-lot comparability data across multiple donors — critical for regulatory submissions arguing that the product is independent of a specific donor’s biology
- Backup donors characterized to the same specifications for manufacturing continuity
- The ability to select donors by HLA profile as the indication-specific alloreactivity data accumulates
- Disease-state donor controls for MLR validation as described above
OrganaBio maintains recallable donors who are fully consented for repeat apheresis collection, HLA-typed at baseline and at each collection, and characterized with comprehensive immunophenotyping. For allogeneic programs that need a stable, documented donor pool rather than one-time procurement, this recallable model provides the supply consistency that process development requires.
What to Ask Your Starting Material Supplier
Before sourcing starting material for an allogeneic program, these questions determine whether the supplier can actually support development through IND:
One: What is the depth of HLA typing? High-resolution 4-digit typing for Class I and Class II, or just serological-level approximations? Allogeneic programs need high-resolution data for alloreactivity modeling.
Two: Are donors recallable for repeat collections? Starting material qualification requires lot-to-lot comparability data from multiple collections from the same donors.
Three: Can disease-state donors be provided for MLR stimulating populations? For autoimmune indications, MLR controls using general healthy donors are insufficient.
Four: What is the documentation chain for each lot — HLA typing, immunophenotype, infectious disease screening, and collection protocol — that would support an IND submission referencing this material?
OrganaBio’s CTDMO team can advise on HLA-typing depth, recallable donor program structure, and documentation packages for IND-supporting starting material sourcing. Contact us to discuss allogeneic program requirements.
