As we all know, Phase III clinical trial, sponsored by Osiris Therapeutics and assessed efficacy of mesenchymal stem cells (MSC) (product “Prochymal”) for treatment of Graft-Versus-Host-Disease (GVHD) failed more than 3 years ago. The possible reasons of failure were not analyzed and discussed publicly by community. Contrary to Osiris trial, similar “academic trials” in Europe were quite successful in Phase II. Recently, for the first time, Jacques Galipeau presents Prochymal failure analysis, based on discrepancy between MSC-based product characteriazation and preparation for US industry-sponsored trial versus European academic trials.
He analyzed 4 variables of MCS product – donor variance, ex vivo expansion, immunogenicity and cryopreservation. I’d briefly summarize each of his points here.
Galipeau pointed out that MSC have a huge inter-donor variability in terms of their immunoregulatory function. This function is critical in GVHD, because it underlies the mechanism of therapeutic action. The caveat is that ISCT’s minimal criteria for MSC are ok for product identity, but have no relevance to immunoregulatory function and potency in GVHD:
These markers are remarkably uniform in their expression among different individual donors, and the assumption is that marrow MSCs are equipotent independently of donor fitness or biology. This assumption may be flawed (14-18), especially when interrogating the immune plasticity of MSCs derived from otherwise indistinguishable donor sources (19,20).
In particular, interferon-gamma responsiveness is mandatory for MSC immunosupppressive function in vivo. And here is the first caveat:
interferon-gamma responsiveness is not uniform among human subjects and that MSCs derived from low IDO inducers may be substantially less potent than cells derived from high inducers.
So, if donor’s MSC were not tested for interferon-gamma responsiveness, it could generate a “potency bias” and eventually lead to divergence of clinical outcomes. In Osiris-sponsored trial MSC were derived from one universal donor, but for European trials – from few (many?) donors. It is very much possible, that Prochymal derived from a donor with low interferon-gamma responsiveness.
The next caveat that Galipeau presents is ex vivo expansion difference. In Osiris-sponsored trials, 10 000 or more doses of Prochymal were generated from one donor. In sharp contrast, MSC from European trials underwent much less proliferative stress (typically 5, max 10 doses).
The question arises: are there any meaningful immune functional differences between minimally expanded marrow- derived MSCs and similarly sourced industrial MSCs with nearly 3 log more expansion?
It is very much possible. Indeed, Le Blanc has published data, which indicates a direct positive link between early passeges MSC and clinical outcome in GVHD.
Galipeau also discusses how significant “expansion stress” could lead to replicative senescence of MSC and function deterioration:
Although senescent MSCs have a virtually indistinguishable marker phenotype from non-senescent MSCs, their functional properties, immune in particular, may be significantly altered.
… the hypothesis that cells approaching replicative exhaustion are functionally distinct from manufactured MSCs devoid of such exhaustion may serve as a testable notion in analysis of comparative MSC performance (or lack thereof) and may provide a mechanistically based rationale justifying use of modestly expanded MSCs for GvHD.
Due to “intrinsic” and acquired (culture in xeno- supplements) immunogenicity, MSC could be rapidly rejected or cause alloimmunization and non-responsiveness. He hypotheses that some “non-responders” in Osiris-sponsored trial “may be due to pre-existing or acquired alloimmunization and rapid MSC clearance”. Pre-existing alloimmunization could appear from large exposure to transfusion products (typical for GVHD patients). Because Prochymal was given through the course of 4 weeks (twice a week), this time is enough for de novo development of HLA alloimmunization. Yet another possible contributor to immunogenicity issue is a culture of Prochymal with fetal calf serum.
I’d like to point out here that the same concerns could be applied to some European trials.
As Galipeau’s group has demonstrated before, cryopreservation could affect the function of MSC.
We also have evidence that in vivo persistence of thawed human MSCs in transfused mice is markedly shortened compared with non-cryopreserved counterparts (unpublished data). The deleterious effects of thawing are self-limited and reverse within 24 hours in vitro. However, it is unknown if such restoration is possible in vivo because it is possible that even “live” MSCs may undergo an accelerated clearance because of their dysfunctional metabolomics (58).
Even thought, this part of analysis does not address the divergence in US versus European trials results, any differences in cryopreservation protocols should be taken in consideration.
In conclusion he harshly criticizing the silence of community about Osiris-sponsored trial results and rush into new trials without rigorous analysis of failure:
Einstein’s definition of insanity is doing the same thing over and over again and expecting different results; thus, we must take pause and attempt to understand and analyze the outcome of study NCT00366145 before investing time and limited resources in duplicative clinical trial efforts, which may be a source of disappointment and, worse, loss of interest from philanthropic and governmental funding agencies as well as private parties instrumental in clinical translation of this cell therapy platform.
Finally, based on his analysis of this failure he provides some recommendations for trialists:
I propose that at least five MSC parameters are worthy of study and further translational development: (i) donor selection based on mechanistically defined potency assays, (ii) analysis of early versus late passage, (iii) analysis of MSC functionality, (iv) analysis of senescent cell content and (v) analysis of pre-existing and acquired alloimmunization to MSC products and impact on transfusion biology.
I was really really thrilled to find this analysis and read! This is one of the best examples of how the analysis of cell therapy trial failure should be done! I’d highly encourage you to read this analysis!