Résumé
B-lymphocytes infected by the Epstein-Barr virus (EBV) are strongly controlled by the immune system, with the persistence of low counts of EBV-infected B lymphocytes in 98% of healthy individuals. Both T- and Natural Killer (NK)-lymphocytes play a main role in immune surveillance of EBV infection. EBV-reactivation (R) frequently occurs in patients having allogeneic transplantation (AT).
We evaluated the impact of controlled EBV-R on survival of allografted adult patients with haematological malignancies and analyzed the different circulating lymphocyte populations.
190 consecutive adult patients were included in this study. Patients received fludarabine, busulfan, ATG for reduced intensity conditioning regimen or cyclophosphamide plus TBI for myeloablative conditioning regimen. Prophylaxis of acute GVHD was conventional. EBV viral load (EBV-VL) was monitored in PB, for all the patients after AT, weekly for the first 6 months, then monthly using standard PCR technique. EBV-R was defined by an EBV-VL ³500 copies/ml, increasing 1 week later. This cut-off level of EBV-VL was used to decide therapy, first reducing immunosuppression followed by rituximab if no EBV-VL decrease (375mg/m2, 4 injections weekly). Lymphocyte immunophenotyping (CD3,4,8,3+56+,3-56+,19,25,DR) was performed weekly for 3 months, then monthly using standard technique. In order to show interdependency between the parameters, a multivariate analysis was performed for OS and PFS including parameters like EBV-R, ATG usage. All analysis was performed with SAS 9.3.
There were 105 acute leukaemias (26 ALL, 71 AML) and myelodysplastic syndromes (8), 71 lymphoproliferative disorders (17 lymphomas, 8 CLL and 46 MM), 5 aplastic anaemia and 9 CML. There were 76 females and 114 males (median age: 51y, range 18-69). 86 (45%) patients presented EBV-R, with 74 patients having EBV-VL>1000 copies/mL. There was no statistical difference between the 2 groups of patients (EBV-R and no EBV-R), for age, sex, cancer type, conditioning regimen, particularly fludarabine, type of graft, and number of infused CD34+ hematopoietic progenitors. Patients with EBV-R received more frequently ATG (P =0.004). The median time of EBV-R detection was 73 days (IQR: 40-210) after transplantation. In case of EBV-R, the dose of immunosuppressive drugs was reduced before stopping. 69/86 patients received rituximab. Patients with no EBV-R did not receive rituximab. Within the 3 months after rituximab, GVHD intensity was decreased, particularly for acute GVHD, which was less frequent. The median follow up was 36.6 months (95%IC 31.5-45.7). Patients with EBV-R had a longer PFS than those with no EBV-R (at 2 years 51% vs. 69%, at 5 years 47% vs 38%) (P<.04). Similarly, OS was longer in patients with EBV-R than that observed for patients with no EBV-R (at 2 years 76% vs. 64% and at 5 years 63 % vs. 47%, P<.001).
OS and PFS were not different between patients according to the type of hematological malignancy (respectively, P=.46 and P=.49). The use of ATG was not a significant factor on OS (univariate (p=0.93) and multivariate analysis (p=0.81)). Similarly, it was not a significant factor on PFS (univariate (p=0.47) or multivariate analysis (p=0.68)).
Among the patients with EBV-R, OS and PFS were not significantly different for patients receiving or not rituximab (respectively p=0.67 and p=0.88).
The median duration of lymphocyte measurement was 94 days (IQR 59-248) after AT. B-lymphocytes and particularly T-lymphocytes, including CD4+ and CD8+ lymphocytes were not significantly different between the 2 groups, respectively P=.29 and P=.48. Only circulating NK cells were significantly different, with a median increased by 22% (P=.03) in patients with EBV-R. CMV co-infection did not impact NK cell counts (P=.058). NK cell count was not different according the use of rituximab (P=.22). The use of rituximab did not modify NK cell counts at 3 and 6 months after stopping treatment (both P=.99).
Controlled EBV-R in allografted cancer patients is associated with better OS and PFS, in association with a significant increase in circulating NK cells. A prospective study is ongoing for analyzing different subpopulations of NK cells.
No relevant conflicts of interest to declare.