Development of Late over Early Full Donor Chimerism (FDC) Results in Improved Progression-Free and Overall Survival in Patients with Advanced Malignant Lymphomas Receiving Nonmyeloablative Allogeneic Hematopoietic Stem Cell Transplantation (HSCT).

Mixed chimerism (MC) can be induced in mice after non-myeloablative conditioning and allogeneic HSCT. Full donor hematopoiesis conversion from MC can occur when nontolerant donor lymphocyte infusions (DLI) are given weeks following the transplant as a result of lymphohematopoietic GVH reactions that eradicate leukemia/lymphoma without GVHD. Based on this murine model, we initiated a trial in patients with advanced hematologic malignancies of non-myeloablative therapy with HLA-matched related donor HSCT with intentional induction of MC followed by prophylactic DLI (pDLI) at day +35 post-HSCT in those without clinical GVHD. This analysis examines the influence of the kinetics of sustained donor lymphoid (CD3+ cells) chimerism on transplant outcomes. Fifty-four patients with advanced hematologic malignancies received nonmyeloablative preparative therapy (cyclophosphamide, equine antithymocyte globulin, thymic irradiation and a brief course of cyclosporine) and HLA-matched HSCT (bone marrow, BM, n=42; GCSF-mobilized peripheral blood stem cells, PBSC, n=12). Chimerism analyses were performed weekly until day +100 post-HSCT and then once monthly for a year. Twenty-nine patients with aggressive malignant lymphomas (NHL, n=21; HD, n=8) were selected out of 50 evaluable patients in an effort to analyze a relatively uniform group in terms of their disease and sustained chimerism; patients who lost the allograft were removed from this analysis. Twenty five patients attained stable FDC, defined as > or = 90% donor T cells, and 4 maintained stable MC. Using 42 days post-HSCT as the defining point for chimerism conversion 13 patients achieved early FDC and 16 patients achieved late FDC. Median time to achieve FDC in the early and late groups was 14 days (range, 7–42) and 70 days (range, 49–91) respectively. Separation produced fairly equal distribution by disease characteristics as follows (early to late): NHL, 10 to 11 and HD, 3 to 5. The early group included 10 PBSC recipients who did not receive pDLI because of the spontaneous conversion to FDC; in contrast, the majority of the late group received BMT and 13 of them received DLI (pDLI, n=10, therapeutic DLI, n=3). The incidence of acute GVHD (grades II–IV) was similar between the early and late groups with rates of 77% (n=10 of 13) and 69% (n=11 of 16) respectively; for chronic GVHD, there was a tendency toward an increased rate in the early group of 75% (3/4) compared to the late group rate of 63% (5/8). Overall disease response rates in the early and late groups were 62% and 69%, but there was a trend toward improved complete remission rates in the late group relative to the early group (50% vs 31%). Progression-free survival (PFS) and overall survival (OS) at 1 year were improved in the late compared to the early group [31% (5/16) vs 15% (2/13) and 50% (8/16) vs 38% (5/13), respectively]. At 5 years, there was a statistically significant improved PFS and OS rate in the late group over the early group [31% (5/16) vs 0% (0/13), p =0.037, for both PFS and OS]. In conclusion, these data support the observation from our mouse models that the most potent graft-vs-tumor effect occurs in mixed chimeras with late chimerism conversion, often after delayed pDLI.