Abstract

Emergence of rituximab (R) has significantly changed the clinical outcome of patients with B-cell lymphoma, which necessitates investigators to reassess the roles of previously determined prognostic factors. We performed a retrospective study to clarify the prognostic significance of serum sIL2R levels among other factors in patients with DLBCL. Study group consisted of 208 consecutive untreated patients with DLBCL who had serum sIL2R levels examined prior to treatment, and were treated in our institution between January 1999 and December 2006. Patients were treated with CHOP based chemotherapy without (R-, n=112) or with (R+, n=96) rituximab. Median follow up duration of each group was 55 and 19 months, respectively. Age (18–92), performance status (PS, ≥2 or <2), B symptoms (present or absent), stage (1–4), number of extranodal involvement (≥2 or <2), bulky diseases (largest diameter of the disease ≥ 10cm, present or absent) serum LDH levels and sIL2R levels were available in all patients. Beta-2 microglobulin levels were available in 130 patients. Serum sIL2R levels ranged from 316 to 35100 U/mL (median 1080), while standard range in healthy individuals was <520 U/mL. Spearman correlation analyses revealed that Log sIL2R level (LsIL2R) was associated with age (p<0.001), PS (p<0.001), B symptoms (p<0.001), stage (p<0.001), number of extranodal involvement ≥2 (p<0.001), Log LDH level (LLDH) (p<0.001) and beta2 microlobulin (p<0.001). The impact of each factor on overall survival (OS) was analyzed by Cox proportional Hazard model. Univariate analyses revealed that age (Hazard ratio [HR]=1.034, p=0.010), stage (HR=1.357, p=0.021), B symptoms (HR=1.152, p<0.001), R− (HR=2.123, p=0.049), PS (HR=4.102, p<0.001), number of extranodal involvement ≥2 (HR=2.592, p<0.001), LsIL2R (HR=3.249, p<0.001), LLDH (HR=4.378, p<0.001) and beta2 microglobulin (HR=1.283, p<0.001) were associated with shorter OS. When analyses was performed separately in R− and R+ groups, HR of LsIL2R was 3.59 (p<0.001) and 2.92 (p=0.071), respectively. Multivariate analysis in the entire group revealed that age (HR=1.030, p=0.041), PS (HR=2.089, p=0.038), B symptoms (HR=2.307, p=0.022), R− (HR=2.721, p=0.038), and LsIL2R (HR=2.140, p=0.018) were independently associated with shorter OS. When multivariate analysis was performed in R− group alone, age (HR=1.035, p=0.035), B symptoms (HR=3.565, p=0.001), and LsIL2R (HR=2.704, p=0.004) independently correlated with shorter OS. In the group of R+, only PS was left as an independent prognostic factor (HR=5.304, p<0.001). When LsIL2R was put back into the multivariate model of R+ group, HR of PS and LsIL2R became 4.308 (p=0.035) and 2.383 (p=0.165), respectively. In conclusion, serum sIL2R levels are frequently elevated in untreated patients with DLBCL and the higher level is significantly associated with shorter OS, independent of other known prognostic factors. When analysis was limited to R+ group, known prognostic factors such as age were not deemed independently associated with OS possibly due to short follow up duration and excellent prognosis in this population (estimated 3-year OS rate of 88%); thus longer follow up of this large cohort is necessary. Updated analyses for OS and various different analyses, such as analyses for correlation with complete response or time to treatment failure will be presented to further clarify the value of sIL2R levels.

Author notes

Disclosure: No relevant conflicts of interest to declare.