Donor lymphocyte infusions after reduced intensity conditioning allogeneic transplantation: what we need to know

Recently in Blood there were 2 papers that increased our knowledge about donor lymphocyte infusions (DLIs) given after reduced intensity conditioning transplantation.^1,2  We published a United Kingdom national study in 81 similar patients 19 months ago.³  This study had a number of limitations mainly because there were multiple different conditioning regimens and the reasons for giving DLIs varied from patient to patient. In concluding, we suggested that there should be prospective randomized studies to address the major questions. Both of the new published studies have the advantage of uniform conditioning regimens and more complete multilineage chimerism data, but many questions remain.

A recent example perhaps best illustrates the gaps in our knowledge. A colleague contacted us about an adolescent patient with acute myeloid leukemia who received a nonmyeloablative unrelated donor transplant, having relapsed after the first full intensity transplant from the same donor. At a few months after transplantation, she had only 70% to 85% donor chimerism in T cells and myeloid cells. Given that the patient was at such high risk of relapse, an interventional strategy seemed justified, but it was hard to advise about the need for DLI and the chance of conversion to full donor chimerism, the starting dose of T cells, the chance of lethal acute or chronic graft-versus-host disease (GVHD), and the likelihood of reducing relapse risk with DLI.

First, is it necessary to convert a state of stable mixed chimerism to full donor chimerism in the absence of measurable disease? While intuitively the answer is yes, particularly in cases considered to be at high risk of relapse, the recent studies do not provide data regarding the natural history of stable mixed chimerism in the absence of DLIs.^1-3  Mixed chimerism can be stable and can be associated with GVHD and a graft-versus-malignancy (GVM) effect.⁴  In some studies^1,3  there does appear to be an association between disease response and conversion to full donor chimerism following DLI, but this is not a universal finding.^2,4  These studies also demonstrate that conversion to complete donor chimerism following DLI cannot be guaranteed, occurring in 34% to 82% of patients.^1-3  This question becomes more critical in the unrelated donor setting where the incidence of GVHD following DLI may be higher²  and lower starting doses may be required.

If we believe that dose escalation of DLI is the optimal strategy, when should it commence and at what starting dose? The University College Hospital (UCH) study²  gave fixed escalating doses of T cells to recipients at 3-month intervals commencing 6 months following transplantation. This study provided valuable data about DLI-associated toxicity at fixed intervals after transplantation. The Seattle consortium¹  using a nonmyeloablative regimen gave DLIs mainly for persistent or relapsed disease (48 of 53 patients). However, there was no prescribed dose of T cells given, and the starting dose was 1 log greater than in the UCH study. The toxicity of DLIs in the sibling setting appears acceptable, with acute GVHD II to IV occurring in 17% to 33% and transient graft hypoplasia, in 0% to 25% of patients.^1-3  Of note, none of 23 patients receiving 1 × 10⁶ T cells/kg from a sibling donor at a median of 254 days after transplantation developed acute GVHD.²  However, this may not be the case with unrelated donors, where acute GVHD occurred in 60% of patients at a similar dose. In addition, largely as a consequence of small patient numbers, none of these studies was able to show a clear association between either initial cell dose or timing after transplantation and the incidence of GVHD. Another major unanswered question is when to give the next dose if the first dose is deemed not to have worked. In the Seattle consortium study, the median time before first and second dose was 5.5 weeks, but it ranged from 6 to more than 300 days. When can one safely say that a dose of DLI has not achieved its goal? Very late responses have been seen in a number of patients with chronic myeloid leukemia, and it is possible that some patients receive a second dose of DLI that is not necessary and only adds toxicity. The place of multilineage chimerism in making the decision about the timing of the next dose is also unclear.

A central tenet of reduced intensity allogeneic transplantation has been that the initial transplantation procedure should act as a platform for subsequent allogeneic GVM reactions. The available data seem to support this concept. Even in the absence of adjunctive chemo/radiotherapy disease, responses were seen in 24% to 52% of patients. There are particularly promising responses reported in follicular non-Hodgkin lymphoma,³  Hodgkin disease,²  and multiple myeloma,²  and these responses were associated with the development of GVHD, providing further evidence for a GVM effect in these diseases. It remains a sobering observation, however, that a substantial percentage of these patients ultimately succumbed to subsequent disease progression or the complications of GVHD. Further, disease responses appeared to be less in patients with aggressive leukemias and lymphomas. Considerable further investigation is required to build on these initial observations. Although difficult to launch, large-scale multicenter trials are urgently needed in this important area of transplantation medicine. Distinct protocols may be required depending on the indication for the DLI and the donor relationship, and adjunctive treatment strategies may need to be incorporated for more aggressive diseases. Further studies are needed in pediatric patients, and more data are needed in recipients of unrelated donor transplants. Nonmyeloablative transplantations are here to stay, but their precise role in patients with different diseases remains to be defined. DLIs will continue to be an important part of this transplantation strategy because mixed chimerism and disease persistence or relapse will continue to be common outcomes of these procedures. Many of the decisions transplant physicians make are not evidence based, but unless we perform systematic prospective studies DLIs will remain among our most uncertain interventions.