Abstract
One hundred and fifty-seven patients with leukemia (80 CML, 48 AML/MDS, 29 ALL) received a T cell depleted myeloablative allogeneic stem cell transplantation (SCT) from an HLA-matched sibling between 09/1993–09/2005. Conditioning consisted of TBI (12–13.6 Gy) + cyclophosphamide (96) or cyclophosphamide and fludarabine (61). The stem cell source was G-CSF mobilized peripheral blood stem cell (PBSC) in 129 and bone marrow in 28 patients. T cell dose with graft ranged from 0.2 – 2 × 105/kg CD3+ cells. GVHD prophylaxis was with low dose cyclosporine (level 100–200 ng/ml) in 103 and standard dose in 54. Patients without ≥ grade II acute GVHD received 1–2 donor lymphocyte infusion of 107 CD3+ cells/kg between days 45 and 100. Absolute lymphocytes on day 30 (LC30) was available in 154 patients (3 patients died before day 30) and 54 day +30 post-transplant samples were available for lymphocyte subset analysis. Median lymphocyte count on day +30 (LC30) was 400/μl, (range 10–3295) and 150/μl, (range 6–1005) for CD56+, CD16+ CD3− NK cells (NK30). Statistical analysis was performed on SPSS14 software. Median age of the group was 34 years (range 10–56). 78 patients had standard risk (SR) disease in first remission or first chronic phase of CML. The remaining 79 had high risk (HR) disease. At the time of analysis 85 patients are (51.5±4%) are alive with median follow-up of surviving patients 1392 days (range 147–4208). Only 9 patients (3 above median LC30) developed aGVHD before day+30. Patients with ≥ median LC30 had significantly better transplant outcome: survival 71±5 vs. 36 ±5.5%, p<0.0001; DFS 71±5 vs. 31±5 %, p<0.0001; NRM 10±3.5 vs. 38 ±6%, p<0.0001; relapse 22±5 vs. 51±7.5%, p=0.004; ≥ II aGVHD 34 ±5 vs. 51±6%, p=0.05. In multivariate analysis only disease risk and LC30 emerged as independent prognostic factors: LC30 above 400/μl was associated with improved survival (RR 4.3), DFS (RR 4.5), less relapse (RR 10.3), NRM (RR 3.3) and aGVHD (RR 2.3). LC30 impacted on outcome of both HR and SR disease groups (Figure). LC30 and NK30 were highly correlated (r2– 0.45, p<0.0001) and NK30 above 150/μl was also associated with improved transplant outcome: In multivariate analysis of this subset of 54 patients, SR disease and NK30 emerged as the only independent factors with better outcome for NK30 >150/μl: higher survival (RR 3), and DFS (RR 3), less relapse (RR 4.8), less NRM (RR 3) and less aGVHD (RR 5.3). This study does not define whether LC30 is a surrogate for NK cell count or whether both are a surrogate for some other undetected prognostic factor. However the inverse relationship between NK count and aGVHD suggests an NK-mediated effect through elimination of host antigen presenting cells as has been described in mismatched SCT but also in HLA identical SCT by
Figure
Disclosure: No relevant conflicts of interest to declare.
