Poster Board I-958
Several reports have identified soluble-form IL-2 receptor á (sIL-2Rá) as a significant prognostic factor in patients with non-Hodgkin lymphoma treated using chemotherapy, particularly in rituximab-containing regimens. However, the clinical significance of sIL-2R is not fully understood, as only small populations have been studied to date. The rationale for increasing of serum level of sIL-2Rá in non-Hodgkin lymphoma is also unclear. Patients and Methods: We analyzed 409 patients newly diagnosed with diffuse large B-cell lymphoma (DLBCL) between January 2001 and July 2008. Treatment comprised CHOP-like regimen with (R-CHOP-like) or without rituximab. Levels of sIL-2R were evaluated with enzyme-linked immunosorbent assay at diagnosis. Overall survival (OS) and progression-free survival (PFS, death from any cause, relapse and refractory disease) were analyzed using Kaplan-Meier methods and survival was compared using log-rank tests. To estimate the survival impact of several factors, including sIL-2Rá level, PS, LDH, B symptoms, extranodal sites ≥2 and age, we performed multivariate analysis using Cox proportional hazards. In 166 of 409 patients, CD25 (IL-2Rá) expression on tumor cells was evaluated using a lymphoma sample from the lymph node, bone marrow, blood or other extranodal organ by flow cytometry. To estimate CD25 expression of tumor cell, CD45 bright cells (mature lymphocyte gate) were gated and considered positive if positivity was seen in >20% of the population excluding CD4-positive cells) using three-color flow cytometry. Results: Median age was 68 years (range, 17-91 years), males/females 1.18, and 28.9% of patients were treated with CHOP-like regimen and 60.2% with R-CHOP-like regimen. Clinical stage was I in 24.4%, II in 24.2%, III in 13.1%, and IV in 38.8%. International Prognostic Index (IPI) was Low in 33.5%, LI in 23.5%, HI in 18.7% and H in 24.3%. Median follow-up for CHOP-like and R-CHOP-like groups was 924 days (range, 16-2878 days) and 799 days (range, 29-2688 days), respectively. Median sIL-2Rá value was 1360 U/L (range, 170-59,500 U/L). For the entire population, CR rate was 71.9%, 3-year OS was 67.6% and PFS was 58.8%. OS differed significantly between sIL-2R >1000 U/L and ≤1000 U/L, between >2000 U/L and ≤2000 U/L and between >3500 U/L and ≤3500 U/L, (p<0.001, <0.001, <0.001, respectively). PFS also differed at each sIL-2Rá point (p<0.001, respectively). The sIL-2Rá value correlated moderately or well with other prognostic factors, such as LDH, PS ≥2, B symptoms, ≥2 extranodal lesions, age and clinical stage by Spearman correlation analysis (r=0.579, 0.258, 0.404, 0.474, respectively). Multivariate analysis showed sIL-2Rá as a significant prognostic factor, in addition to several factors. In a group treated with R-CHOP-like regimen, 3-year OS was 74.5% and PFS was 68.8%. OS again differed significantly between sIL-2Rá >1000 U/L and ≤1000 U/L, between >2000 U/L and ≤2000 U/L and, between >3500 U/L and ≤3500 U/L (p<0.001, respectively). PFS was also significant at each sIL-2Rá value. The higher the level of IL-2R, the worse the 3-year OS at each sIL-2R value (63.6%, 60.1%, 53.2%, respectively). However, we could not identify statistical significance of sIL-2 level by multivariate analysis. IL-2Rá usually functions as a cytokine receptor on cell surface, called CD25. To show the importance of CD25 expression in lymphoma cells on serum level of soluble form IL-2Rá, we compared sIL-2Rá levels in CD25-positive and -negative cases. CD25-positive cases showed significantly higher sIL-2Rá level than CD25-negative cases among the overall population. After defining two group according to clinical stage (I+II and III+IV), sIL-2Rá level was higher in the CD25-positive group than in the CD25-negative group for the stage III+IV group (p=0.001), but this difference was not seen for the stage I+II group (p=0.390). This trend was also seen in the case of IPI, L+LI (p=0.642) and HI+H (p=0.0016)). These results suggest that one rationale for increasing level of sIL-2Rá in DLBCL is removing from tumor cell like other cytokine receptor. Conclusion: In terms of survival and relapse, sIL-2R remains an important risk factor of DLBCL, not only in CHOP-like regimens, but also in the R-CHOP era. The survival rate of patients with sIL-2Rá >3500 U/L is extremely poor even if treated with R-CHOP (53.2%). We showed that one rationale for increasing level of serum sIL-2Rá level in DLBCL is to remove from the tumor cell surface.
No relevant conflicts of interest to declare.
Asterisk with author names denotes non-ASH members.