Adjusted conditioning for allogeneic transplantation in a single center setting: mixed chimerism heralds relapse.

The role of mixed chimerism on subsequent relapse was prospectively evaluated in an allotransplant program. Sixty-six patients with median age of 54 and mainly high-risk hematologic disease and/or solid tumors had individually adjusted non-myeloablative conditioning. Thirty-nine donors were siblings and 27 unrelated. Frequent chimerism analyses supported immune manipulation including donor lymphocyte infusions. The need for transfusions, i.v. fluids, and antibiotics, and weight loss was less than in a control cohort. Most patients had immediate full and consistent donor chimerism; one-third required immune manipulation. Eight of ten evaluable CML patients were BCR/ABL-negative at days 14-58 post-transplant. Mixed chimerism frequently preceded relapse, and the relapse rate was 38% in 26 patients with mixed chimerism vs. 11% among 35 with consistent full donor chimerism (p = 0.015). The current transplant- and disease-related mortality were 11 and 9%, respectively, among 35 non-high-risk patients, and 35 and 10% for 29 high-risk patients with hematologic malignancy. With a median follow-up of 15 months the 2-year overall survival is 73% for non-high-risk, and 46% for high-risk patients. Adjusted conditioning reduces early toxicity and resource requirements without impairing tumor control, probably due to a rapid establishment of the graft-versus-cancer effect. Mixed chimerism heralded relapse, and tumor-related mortality is not greater with adjusted than with conventional conditioning.