Abstract
Abstract 4064
Chronic graft-versus-host disease (cGVHD) is one of the main obstacles for long-term survival after allogeneic hematopoietic stem cell transplantation (HSCT). Although the important role of antigen-presenting cells in cGVHD and the association between cGVHD and increased monocytes after HSCT were reported, it is unclear whether the number of peripheral blood (PB) monocytes influences the outcome after HSCT.
Patients and methods: We retrospectively reviewed 198 consecutive patients who underwent HSCT for hematologic malignancies and bone marrow failure syndrome in our institution between 1997 and 2009 and survived at least 1 year without relapse. The association between the absolute number of PB monocytes at 1 year after HSCT and transplant outcome was assessed.
Median follow-up period from HSCT was 61.9 months (range: 12.3–180.6 months). Patients with cGVHD at 1 year after HSCT showed increased PB monocytes compared with those without cGVHD (median 0.54×109/L vs 0.40×109/L, P =0.0008). While relapse rate was not significantly different according to PB monocyte count, patients with PB monocyte count less than 0.50×109/L at 1 year after HSCT had significantly better disease free survival (DFS), better overall survival (OS) and lower non-relapse mortality (NRM) rate compared with those with more than 0.50×109/L (10-year DFS 86.0 vs 63.6%; P =0.0070, 10-year OS 88.8 vs 63.8%; P =0.0031, and 10-year NRM rate 5.8 vs 21.4%; P =0.011). Eighteen of 198 patients (9.1%) died from non-relapse causes after 1 year post-HSCT; 14 of 91 in the high PB monocyte count group (15.3%) and 4 of 107 in the low PB monocyte count group (3.7%) (P =0.0057). Among the 18 patients, 12 died of cGVHD (9 in the high monocyte count group and 3 in the low monocyte count group).
By Kaplan-Maier estimates, risk factors of higher NRM rate were age older than 35 (P =0.043), female donor to male recipient (P =0.020), high disease risk (P =0.0033), unrelated donor (P =0.044), human leukocyte antigen (HLA) mismatch (P =0.00023), steroid use at 1 year after HSCT (P =0.00013), immunosuppressant use at 1 year after HSCT (P =0.010), and PB monocyte count more than 0.50 × 109/L at 1 year after HSCT (P =0.0081). The factors with P ≤0.1 in the univariate analyses, i.e. grade 2–4 acute GVHD (P =0.078), extensive type cGVHD at 1 year after HSCT (P =0.057) and PB platelet count less than 1.0×1011/L at 1 year after HSCT (P =0.054) were also included in the following multivariate analysis. Multivariate analysis with competing risk regression model revealed that NRM rate was associated with age older than 35 (HR=3.99, P =0.010), HLA mismatch (HR=7.96, P =0.040), steroid use at 1 year after HSCT (HR=5.55, P =0.010), and PB monocyte count more than 0.50×109/L at 1 year after HSCT (HR=3.65, P =0.023). Cox multivariate analysis for DFS also showed that PB monocyte count more than 0.50×109/L at 1 year after HSCT was an independent prognostic factor (HR=2.22, P =0.028). Other risk factors for DFS were high disease risk (HR=2.14, P =0.032) and extensive type cGVHD at 1 year after HSCT (HR=2.09, P =0.040).
Increased PB monocytes also predict future onset of cGVHD. Among patients without cGVHD at 1 year after HSCT (n=68), those with increased PB monocytes had higher rates of future cGVHD onset (10-year 6.5 vs 42.6%; P =0.013) and NRM (10-year 0 vs 21.6%; P =0.0086).
Increased PB monocytes after HSCT were associated with cGVHD, future cGVHD onset, higher NRM rate and poor survival. Our data suggest the importance of monocytes in the etiology of cGVHD.
No relevant conflicts of interest to declare.
Author notes
Asterisk with author names denotes non-ASH members.
