I was fascinated by a recent talk given by Irwin Bernstein that I attended and a clinical study done by his group, which is exploring the possibilities of cord blood hematopoietic stem cell (HSC) expansion, published in Nature Medicine.
Authors were not only able to expand cord blood HSC drastically, but also performed phase I of clinical trial, which appear to be safe and preliminary data demonstrated earlier rapid myeloid engraftment in 10 patients.
HSC expansion – from lab to the clinic
If you follow the progress in the field you know that many HSC expansion protocols were published, but only a few of them attempted to enter to clinical trials. I suspect it happened because the vast majority of them didn’t pass all of validation tests, such as functional assessment of HSC function in xenotransplant models after ex vivo expansion. But some protocols entered into the clinical trials phase. Most of them were done on bone marrow or mobilized blood HSC, but some with cord blood (CB). Some trials didn’t move from phase I (safety and feasibility) to phase II (efficacy).
There is a huge demand for HSC expansion from cord blood particularly, because this is the main disadvantage for wide CB transplantation as an alternative to bone marrow and mobilized blood. Currently, a few CB HSC expansion protocols are undergoing clinical trials. But Bernstain’s study we can consider as more solid and successful at this point. Companies are also actively involved in developing CB HSC expansion protocols and supervise the clinical trials. For example, Fate Therapeutics licensed the “small molecule – HSC expander” from Children’s Hospital Boston and started a clinical trial.
Is it possible to keep expanded HSC in undifferentiated state?
Unfortunately, more likely the answer is NO. Even if you can detect the difference of expansion within long-term HSC subset by flow cytometry, these guys already primed to be actively dividing but not quiescent after the transplant. It could sound like a speculation, but nobody has proven the opposite so far. Almost all studies calculate HSC expansion difference based on CD34+ total expression or CD34+/CD38-. But even the last mentioned population does not represent HSC, because most of CD34+/CD38- are progenitors and you need to go way further (Lin-/CD34+/CD38-/CD45RA-/CD90+) in order to conclude something about long-term HSC. Most of the in vivo studies do not assess multilineage engraftent long enough to conclude confidently about superior repopulation of expanded HSC.
That’s why earlier trials didn’t show advantage in neutrophil and platelet recovery time – they expand something withing CD34+ total population, but not that we need.
Citation from the review:
Initial efforts to expand UCB progenitors ex vivo have resulted in expansion of mature rather than immature HSC, confounded by the inability to accurately and reliably measure long-term reconstituting cells. Ex vivo expansion of UCB HSC has failed to improve engraftment because of resulting defects that promote apoptosis, disrupt marrow homing and initiate cell cycling.
Double unit CB competition for engraftment
As you can notice from Bernstein’s study, they didn’t transplant just expanded CB sample but rather mixed with competing second – unmanipulated unit. For the phase I of clinical trial this sounds reasonable and safe. As we know, in double unit CB transplant only the one sample will win engraftment eventually, but another sample will act as a helper. In the case of this trial, expanded CB sample was a helper and unmanipulated CB unit was a winner. Why so? Because virtually all expanded CD34+/CD38- cells were multipotent and myeloid primed progenitors with very very few true HSC. Those very few expanded HSC, the authors were not able to detect a few months after transplant in 8 out of 10 patients. 2 patients with some signs of expanded long-term HSC engraftment were under observation for 240 and 180 days. But in one of them they were not able to detect signs of multilineage engraftment at the time point of 1 year and second patient didn’t possess engraftment in T-cells.
So, even though they transplanted 6 millions per kg of expanded CB CD34+ cells with only 0.24 millions (25 times less!) per kg unmanipulated CD34+, the second unit won engraftment in all 10 patients by 1 year of observation period. Why so?
2-7 millions per kg of CD34+ cells we need for clinically significant engraftment in adults. Magically, in pediatrics, 0.1-0.3 millions per kg of CD34+ CB cells is enough for sufficient blood lineage recovery. Average CB sample can provide the dosage only below 1 million per kg for adults. So we need to at least expand CD34+ cells twice to make the patients happy. Why are researchers shooting 50-150 fold expansion then? Because in reality almost none of long-term HSC expanded, but primed to progenitors in vitro, as I mentioned above. So, when we are talking about twice more expansion, we should achieve it naturally and proportionally in all HSC and progenitor subsets, without “cell culture artifacts”, such as “progenitors skewing”.
Do we really need HSC expansion for clinical success?
It seems like expansion of common myeloid (CMP) or multipotent progenitors (MPP) is good enough to achieve earlier and rapid neutrophil recovery. Maybe we even don’t need to chase robust expansion of long-term quiescent HSC in experiment and pre-clinical setup. I would not, just because if I see 30-100 times expansion of long-term HSC my “cancer caution” alarm starts beeping. Total number of CD34+ or CD34+/CD38- game could be efficient for good clinical outcome, as this study showed. I wonder what would happen if transplant was just expanded CB sample alone? No, better not to take such a risk, but keep mixing with unmanipulated unit.