Cd34 Dose And Chronic Graft Versus Host Disease (Cgvhd) Affect Survival In Allogeneic Peripheral Blood Stem Cell Transplantation (Allopbsct) Following Non-Myeloablative (Nm) Conditioning: The Vanderbilt University/Nashville Va Sct Program Experience

AlloPBSCT utilizing NM conditioning is hypothesized to minimize the toxicity of myeloablative regimens while harnessing a potent graft vs malignancy effect. We sought to identify factors predicting survival in a retrospective analysis of 60 patients (pts) undergoing alloPBSCT from HLA-matched related donors between 8/00 and 8/05 at our program. All pts received 90 mg/m2 fludarabine and 200 cGy TBI. GVHD prophylaxis consisted of CSA/MMF. The median age was 55 years (range 41-66). Male: female was 51:9. All transplants were performed for hematologic malignancies. The median number of treatments prior to transplant was 4 (range 0-8). The mean cell doses infused were 8.0 (range 3.5-16.3) x 106 CD34+/kg and 28.4 (range 0.9-65.9) x 107 CD3+/kg. Among 33 (55%) pts who became neutropenic, the median time to ANC > 500 was 21 days (range 14-49). Primary graft failure occurred in 2 patients. Of evaluable pts with chimerism data, median time to 100% donor chimerism was day +28 for CD33+ and day +180 for CD3+. AGVHD occurred in 42/60 (70%) pts (grade I in 14 and grades II-IV in 28), and cGVHD occurred in 28/52 (54%); 12 pts developed extensive cGVHD. 14 pts required donor lymphocyte infusion for disease progression. 100-day NRM was 10% (all from aGVHD); another 2 patients died of disease progression. After day +100, most pts died from disease progression or infection related to cGVHD. Kaplan-Meier probabilities for overall survival (OS) and progression free survival (PFS) at one year were 61% and 48%, respectively with median follow up of 505 days (range 31-2029). OS at 2 and 3 years was 45% and 36%. In univariate analysis, the development of cGVHD significantly correlated with improved OS (p = 0.02) and PFS (p = 0.01). Although CD34+ and CD3+ doses showed no significant association with development of cGVHD, higher doses of both were associated with development of aGVHD (p = 0.01 and 0.04 respectively) and higher CD34+ doses (6.8 vs 9.4×106/kg) were associated with development of higher grades of aGVHD (grades II-IV versus grades 0-I) (p < 0.01). Incremental increases in CD34 dose were associated with an increased hazard ratio for death; however, a dose above 8 ×106 CD34+/kg was associated with lower probability of relapse (p = 0.04). NM conditioning has a role in alloPBSCT and further investigation to optimize disease indications and CD34+ dosing is needed.