The Graft That Has Never Met an Antigen: Cord Blood Stem Cells and the Arithmetic of Engraftment

Key Takeaways:

• Cord blood T cells are overwhelmingly CD45RA-positive naive cells that lack whole adult effector populations outright, including the skin-homing CLA-positive subset — which is why a deliberately mismatched cord unit provokes less graft-versus-host disease than a fully matched adult marrow graft.
• In a 682-patient registry comparison of unrelated-donor transplants in adults with acute leukemia, ninety-four percent of cord blood grafts were HLA-mismatched while every marrow graft was matched, yet the risk of grade II to IV acute graft-versus-host disease was lower with cord blood, at a relative risk of 0.57.
• The limiting variable is dose, not match: cord units delivered a median of 23 million nucleated cells per kilogram of recipient weight against 290 million per kilogram for marrow, and neutrophil recovery was correspondingly delayed at a relative risk of 0.49.

The reason cord blood stem cells occupy a niche no adult donor can fill has less to do with the stem cells themselves than with the company they keep. A unit collected after delivery contains CD34-positive hematopoietic stem and progenitor cells — the self-renewing population that rebuilds an entire blood and immune system after myeloablative conditioning — but it also carries a lymphocyte compartment unlike anything circulating in an adult. Cord blood T cells are predominantly CD45RA-positive, CD62L-positive recent thymic emigrants: antigen-naive cells that secrete less cytokine and kill less efficiently than their adult counterparts, that display fewer activation markers, and that are missing entire effector lineages altogether, including the skin-homing CLA-positive population whose absence matters enormously given that skin is the signature target organ of acute graft-versus-host disease. That immunologic inexperience is what makes a deliberately imperfect match survivable. In the European registry comparison of 682 adults transplanted for acute leukemia, ninety-four percent of cord blood grafts carried an HLA mismatch while every marrow graft was fully matched, and the cord recipients still had a lower risk of grade II to IV acute graft-versus-host disease, with a relative risk of 0.57.

The price of that tolerance is arithmetic, and the arithmetic does not negotiate. A marrow or peripheral blood donor can be asked for more cells, or asked again months later; a cord unit is whatever the placenta and umbilical vessels yielded in the minutes after birth, cryopreserved once and never replenished. In that same registry comparison the median infused dose was 23 million nucleated cells per kilogram of recipient weight for cord blood against 290 million per kilogram for marrow — better than a tenfold gap — and because the graft is dosed per kilogram of the person receiving it, the constraint tightens as the recipient gets heavier, which is exactly why the technique matured first in children. Fewer progenitors means a longer climb out of aplasia. Neutrophil recovery after cord blood transplantation was significantly delayed, at a relative risk of 0.49, and platelet recovery lags further still. Every additional day of profound neutropenia is another day of exposure to invasive bacterial and fungal infection before the new marrow can defend its host, which is why so much of the field’s engineering effort — infusing two units instead of one to augment the graft dose, refining conditioning, sharpening supportive care — is aimed at that single number.

Who do cord blood stem cells actually help most?

Not the patient with a matched sibling, and not the patient with a 10/10 unrelated donor already sitting in a registry. The answer is the patient for whom that donor does not exist. HLA haplotypes are inherited as intact blocks and their frequencies diverge sharply between ancestral populations, so registries assembled disproportionately from donors of European descent return matches unevenly. In a prospective evaluation of 553 patients searching for a graft, a 10/10 HLA-matched unrelated adult donor was found for fifty-three percent of those with European ancestry but only twenty-one percent of those with non-European origins — and yet the majority of both groups had a suitable 5-6/6 cord blood unit available, because a cord graft simply does not demand that stringency. The consequence showed up in who actually reached transplant: patients of non-European ancestry were twenty-three percent of adult-donor recipients but fifty-six percent of cord blood recipients, and of the patients who found no stem cell source at all, seventy-three percent had non-European ancestry. That is the honest case for cord blood, and it is a case for public donation rather than private storage. Professional societies are consistent here: a privately banked autologous unit is rarely usable for the child it came from, because an inherited disorder or a preleukemic clone is often already present in the stored blood, whereas a unit given to a public bank enters a searchable inventory where it can match a stranger whose ancestry the registry has historically underserved.

References:

1. Szabolcs, P., Park, K. D., Reese, M., Marti, L., Broadwater, G., & Kurtzberg, J. (2003). Coexistent naive phenotype and higher cycling rate of cord blood T cells as compared to adult peripheral blood. Experimental Hematology, 31(8), 708-714.
2. Rocha, V., Labopin, M., Sanz, G., Arcese, W., Schwerdtfeger, R., Bosi, A., et al. (2004). Transplants of umbilical-cord blood or bone marrow from unrelated donors in adults with acute leukemia. The New England Journal of Medicine, 351(22), 2276-2285.
3. Barker, J. N., Byam, C. E., Kernan, N. A., Lee, S. S., Hawke, R. M., Doshi, K. A., et al. (2010). Availability of cord blood extends allogeneic hematopoietic stem cell transplant access to racial and ethnic minorities. Biology of Blood and Marrow Transplantation, 16(11), 1541-1548.