Graft Purging in Autologous Bone Marrow Transplantation: A Promise Not Quite Fulfilled

In the early 1990s, a few European physicians met to design a trial assessing the value of high-dose therapy followed by autologous stem cell transplantation in relapsed follicular non-Hodgkin's lymphoma (NHL). Extensive discussions were also devoted to the topic of purging the stem cell graft. Some of us felt it did not make sense to reinfuse tumor cells after the application of high-dose therapy including total-body irradiation: "The tumor cells were clearly visible in the bone marrow." Others felt there may be a difference between monoclonal cells visible in the marrow and clonogenic cells that may give rise to a relapse. It was the era of the first positive data on purging in AML[1] and NHL.[2,3] In those discussions, it was apparent that there were believers and nonbelievers. Finally, it was decided to investigate this issue as a separate question in a new study that is currently known as the CUP trial (Chemotherapy vs high-dose therapy followed by Unpurged stem cell transplantation vs high-dose therapy followed by Purged stem cell transplantation). The data from this trial were recently published.[ 4] We did not see an advantage for purging; however, we were not able to solve the purging question definitively, due to poor accrual. Why, in this era of evidence-based medicine, are we unable to assess the appropriate place of purging? Apparently there is a lack of incentive. The Search for Solid Evidence
With current data, there are convincing arguments that purging can remove several logs of tumor cells- even to the point of polymerase chain reaction (PCR) negativity[2]-without severely delaying engraftment. Purging, however, is labor-intensive and, therefore, a costly procedure. With future good manufacturing practice guidelines, this will be an even more significant issue. Therefore, solid evidence in favor of purging is very welcome. With the lack of direct evidence, circumstantial evidence using data from twin transplants may be helpful. This setting is considered conducive to the optimal clean graft, without the potential immunologic effects of allogeneic transplants. Two recent studies using twin data may shed light on the purging issue. Bierman et al[5] analyzed International Bone Marrow Transplant Registry-European Group for Blood and Marrow Transplantation (IBMTR-EBMT) registry data to compare the results of syngeneic, allogeneic, and autologous transplants for NHL patients. They concluded that recipients of syngeneic grafts had a lower likelihood of relapse compared with autologous transplants (either purged or unpurged), especially in low-grade lymphoma. In a comparable study in myeloma patients using EBMT registry data, Gahrton et al[6] also came to the conclusion that relapse rates after syngeneic transplantation were lower than after autologous transplantation. This suggests that reinfusion of tumor cells may have a significant impact. However, the Bierman study[5] also raises some questions.[7] These investigators were unable to demonstrate a graft-vs-lymphoma effect, although at least one case report very convincingly demonstrates this effect.[ 8] This may suggest a confounding factor in patient selection, as may be expected from a registry study. An alternative hypothesis may be that the better syngeneic results can also be explained by a much-debated immunologic effect in syngeneic transplantation.[ 9,10] Although twin data are helpful in this context, many is- sues remain. Therefore, again, solid evidence in favor of purging is very welcome. Conclusions
It is rather disappointing that we have not solved the purging issue in the past 2 decades. The believers continue to work in this field but fail to do the proper randomized trials they generally recommend (to others?); the nonbelievers also continue in their approach. The final result is that we fail to advise our patients correctly because we lack the data.