Cooking Up Tolerance: Has a New Recipe Been Created?

In a recent issue of Science Translational Medicine, Leventhal et al. (1) describe a cohort of patients that underwent combined hematopoietic stem cell transplant (HSCT) and renal transplant. Of eight patients enrolled, five have developed full replacement with donor hematopoiesis, and each accepted their concomitantly placed renal allografts without the need for long-term immunosuppression.

Although this study is remarkable for many reasons, the induction of renal allograft tolerance in the setting of full replacement with donor hematopoiesis is not one of them. There are many examples of patients receiving prior HSCT, whose hematopoietic compartment is 100% donor, and are able to accept a kidney from the same donor without the need for immunosuppression (2). These patients exhibit an immunologic “switched identity” where organs from the same donor would not be recognized as immunologically foreign.

What is truly remarkable is the immunologic barriers crossed during HSCT, without widespread adverse events. Combined transplants that cross HLA barriers are not generally considered achievable in chemotherapy-naïve patients, especially with less toxic reduced-intensity pre-HSCT conditioning, since these transplants have historically been associated with high rates of both graft rejection and (in the patients who engraft): graft-versus-host disease (GvHD). In the Leventhal et al. study, although graft rejection did occur (with three of the first four patients experiencing rejection of the donor HSCT and failure to induce tolerance to the donor’s kidney), GvHD has apparently not occurred, even across significant HLA disparity. If this result becomes widely applicable, it would be a revolutionary step forward for HSCT, and for solid organ transplant patients who could benefit from a relatively low-toxicity approach that could ensure indefinite allograft survival without lifelong immunosuppression.

The question then becomes, what were the key ingredients in the tolerance recipe used? The answer to this is not yet defined. Although the majority of the individual components have been used previously (3), the combination used in this study was unique. The backbone included megadoses of purified CD34+ cells, previously shown to have potent graft-promoting and GvHD-protective capabilities (4,5). HSCT grafts also contained a moderate dose of conventional T cells, known to promote engraftment posttransplant, but large enough to confer a substantial risk of GvHD. Finally, there was something close to a “secret ingredient”: the authors included “facilitating cells (FCs)” in their stem cell product, a mixed product containing both plasmacytoid dendritic cells and other cell types (6). Unfortunately, the stem cell product cannot be independently produced for outside confirmation of their results, given the fact that the production strategy is proprietary, limiting the portability of this methodology. Finally, both pre- and posttransplant cyclophosphamide treatment was used, based upon growing evidence that post-HSCT cyclophosphamide may be efficacious in promoting engraftment and preventing GvHD by selectively deleting T cells activated in the early days after HSCT (7).

One of the important observations the authors made in this study was the high risk of graft rejection that accompanied their transplant strategy. Three of the first four patients experienced graft rejection early after HSCT. This rejection of donor hematopoietic stem cells was accompanied by a lack of tolerance to the renal allograft, supporting a model of linked tolerance, where solid organ tolerance requires ongoing donor hematopoiesis. The authors attributed the rejection to a combination of potential factors including (i) at least one of these recipients did not receive the full conditioning regimen, including the omission of the post-transplant cyclophosphamide and (ii) rejection occurred in two recipients that received lower CD34+, conventional T cell and FC doses. While the last four patients enrolled on the study did successfully engraft in the setting of higher doses of CD34+, conventional T cells and FCs, the rejection observed in three patients points out the “knifes edge” of success that may accompany reduced-intensity transplants in this patient population. If this regimen is going to be applied widely to patients not at risk of imminent death from their primary disease, the risks associated with failed tolerance-induction must be weighed carefully with the risks of ongoing dialysis or lifelong immunosuppression.

Perhaps the most striking result in this series is the lack of GvHD reported, even in patients that were fully replaced with donor cells, including donor T cells. Of the five patients at risk for either acute of chronic GvHD (those that did not reject the donor stem cells) no GvHD was observed. Given the small numbers of patients reported, it is not yet possible to make global predictions from these data, other than to state that, quite contrary to expectations, in this highly HLA-mismatched cohort, the transplant regimen did not induce a huge GvHD risk. Di Ianni et al (4) describe a similar regimen, which used megadoses of CD34⁺ cells, a moderate dose of conventional T cells, along with the use of Tregs (although after myeloablative recipient conditioning) to promote engraftment and prevent GvHD. Two of 26 patients developed GvHD (and died of their disease). Notably, the rates of GvHD between these two series are not statistically different from each other (p = 0.72 by χ ² test). This may serve as an important reminder of the difficulty in making predictions from small studies. While the fact that Leventhal et al. observed no GvHD is indeed impressive, the small number of transplants reported does not yet allow one to conclude that when more broadly applied, their regimen will be entirely devoid of GvHD. The question must be addressed then: how much GvHD is acceptable for the solid organ transplant community?

While the authors suggest that it was the FCs that were most important in promoting engraftment without GvHD, this was not proven in their study, given that this was a single arm study that did not compare transplants with and without FC. While the authors note that their Data and Safety Monitoring Board thought it unwise in this high-risk patient population to eliminate a potential GvHD protective measure, if the transplant community is to rigorously determine the importance of the FCs to the success of this regimen, a direct comparison is required. One potential avenue by which to more safely do this would be to perform a similar trial in the HLA-matched setting: although the risk of GvHD in these transplants is lower than in the major histocompatibility complex-mismatched setting, these recipients still do have a significant risk of GvHD, and it would be arguably safer to use this patient group for a randomized trial design. This is especially important given the fact that in addition to FCs, the Leventhal et al. group employed two other potentially important GvHD protective maneuvers in their transplants: the combination of high-dose CD34 cells (known to have “veto activity”) along with posttransplant cyclophosphamide. Both strategies have been previously shown to significantly protect against GvHD even across multiple HLA disparities (4,7).

While hematopoietic engraftment, renal transplant acceptance and freedom from GvHD are certainly critical markers of success after combined HSCT/renal transplantation, there is one last, critical component of success: immunologic health. It is on this measure that there is some reason to be concerned about the patients reported by Leventhal et al. One developed a serious viral illness (requiring ICU admission, and intubation, resulting in sepsis-related loss of the renal allograft). Immune deficiency and the high risk of serious infectious complications, (most often from difficult-to-treat viral and fungal disease) has been the major drawback of highly CD34-selected HSCT (3). In patients with hematologic malignancies, the resulting immune deficiency and infectious morbidity and mortality has often made the overall outcome of these transplants not better than T cell replete transplants, despite the significant protection from GvHD that CD34 selection affords. There might be some cause for concern over the state of the protective immune response in the patients reported in the current study, given that the level of immune reconstitution appears to be significantly delayed compared to what is commonly seen, even after CD34-selected HSCT. As shown in their supplementary figure S1, the majority have very low CD4+ T cell counts even 18 months post-HSCT, with the reported counts <200 cell/μL for all three patients analyzed up to 18 months posttransplant. A CD4 count <200/μL has been documented to lead to significant risk of opportunistic infections in both the transplant and HIV literature (3,8,9). While the authors report a relative ability to respond to candida antigen in vitro compared to a recipient-directed response (the absolute response was not shown, precluding a quantitative interpretation of that data), several common measures of posttransplant immune competence were not reported, including cytomegalovirus and Epstein-Barr virus levels, known to cause significant post-transplant toxicities. The authors also have not reported other more specific measures of intact protective immunity in their patients, which might include tetramer analysis of virus-specific cells, the enumeration of recent thymic emigrants or response to posttransplant vaccination. In the long term, these measures of intact immunity will be as critical as prevention of GvHD to the health of these patients.

Leventhal et al. should be recognized for developing a bold approach, which, if successful in the long-run, could open up transplantation to patients with both end-stage organ disease and nonmalignant hematologic conditions who currently are ineligible for transplant because of the lack of a sufficiently HLA-matched donor. Their unique formula now needs to be tested in larger patient populations, and the importance of the FC “secret ingredient” rigorously evaluated. Finally, the patients on the current trial need to be closely monitored for their long-term immunologic health. It will be this final parameter that will ultimately determine whether FC-based therapies really are a recipe for success.