The article by Champlin and colleagues summarizes exciting recent clinical results in allogeneic bone marrow transplantation (BMT) for leukemia and lymphoma achieved by reengineering the process to take maximum advantage of the powerful antileukemic effect of the allograft itself. These results are presented in the context of an elegant review of 2 decades of international competitive and collaborative translational research within the BMT community aimed at separating this graft-vs-leukemia (GVL) effect from the devastating complication of graft-vs-host disease (GVHD).
A Mixed Paradigm
Historically, successful allogeneic transplantation for malignancy has depended on a mixed paradigm. The preparative regimen had to be sufficiently immunosuppressive to permit engraftment while having sufficient anti-
neoplastic activity to eradicate the malignancy. The graft itself had to contain enough immunocompetent cells to overcome rejection and establish a functioning immune system graft, as well as sufficient hematopoietic stem cells to permanently restore hematopoiesis. The window of opportunity for dose escalation provided by the allografts ability to restore hematopoiesis had to be large enough to justify the extraordinary toxicities of the escalated dose of the combined antineoplastic and immunoablative preparative regimen, as well as the risk of severe or fatal GVHD.
The idea that the GVL effect alone might be sufficient to eradicate a substantial residual volume of malignancy and that the preparative regimen need only debulk the malignancy and permit transient engraftment is truly revolutionary. The further demonstration that the graft itself can be engineered to permit pre- and even post-graft modulation of GVHD and GVL effects opens up the possibility that the therapeutic index of this approach can be significantly enhanced.
Magnitude of Benefits Depends on the Malignancy Being Treated
The observation that the benefit of such mini-transplants varies widely depending on the malignancy being treated is sobering but provocative. On the one hand, the fact that clinically significant benefit has been demonstrated only for chronic myelocytic leukemia should limit enthusiasm for immediate widespread phase II trials. On the other hand, the fact that a less clinically useful GVL effect can still be demonstrated to occur in other malignancies invites efforts to better understand the mechanisms of this phenomenon to permit rational patient selection, as well as more clinical trials aimed at more effective application.
Is it possible that this GVL effect is not an allogeneic effect at all, but rather, represents a restoration or an amplification of a failed mechanism for immune surveillance against malignancy? Two observations make the latter explanation seem unlikely: (1) The GVL effect appears to be of significantly lower magnitude in syngeneic (identical twin) transplants; and (2) this effect is of intermediate clinical benefit in HLA-matched sibling transplants when GVHD does not occur. However, a simple allogeneic effect does not appear to explain the disease-specific variation in the magnitude of the phenomenon.
Disease Biology and Susceptibility to the GVL Effect
The correlation of disease biology with the susceptibility to the GVL effect deserves comment. In addition to chronic myelocytic leukemia, the other diseases in which clinical benefit seems likely to be demonstrated are the indolent lymphomas, multiple myeloma, and some myelodysplastic syndromes. There is a very limited, although still minimally demonstrable, effect in acute lymphocytic leukemia. Acute myelocytic leukemia appears to demonstrate intermediate sensitivity, which may simply reflect the substantial heterogeneity of this group of diseases.
In general, the susceptible diseases are those that have proven most susceptible to biological therapies other than allogeneic marrow transplantation. These include interferon for chronic myelocytic leukemia, monoclonal antibody therapy for indolent lymphoma, and thalidomide(Drug information on thalidomide) for multiple myeloma.
Other similarities, however, are the indolent nature of the susceptible diseases and their tendency to demonstrate a predominantly mature morphologic phenotype, despite indication of transformation of a very primitive cell, perhaps one not bearing HLA class II antigens, such as the hematopoietic stem cell itself. One might speculate that this process invokes a mechanism for protecting the host against the most dangerous transformation of all, that of the hematopoietic stem cell, or at least for controlling any differentiating, rapidly accumulating progeny of such a transformation.
In summary, these elegant clinical trials of marrow transplantation demonstrate the successful exploitation of the GVL effect first described 20 years ago, utilizing new knowledge of the cellular mechanisms of allogeneic reactivity. These trials may truly represent a reinvention of BMT, but at the very least, they stimulate us to think about the necessity for the treatment to fit the disease.
Currently, BMT can be viewed as the hammer within our antineoplastic tool set. As research and clinical strategic thinking along the lines described above progress, this hammer may, indeed, evolve into a variety of tools within that tool set. As we develop this broader array of tools, we also need to continue research to better define our malignant disease targets, which, in turn, will help us select the proper tool from the set.
It is perhaps fortuitous for current studies that the susceptible diseases occur in older individuals, who are also less likely to tolerate the rigors of maximum-intensity preparative regimens. It is also fortuitous that the susceptible diseases are those generally considered incurable from the time of diagnosis, making such mini intensive approaches doubly attractive.