The Sil Index Is a Useful Prognostic Indicator for Diffuse Large B-Cell Lymphoma

Background: Rituximab (R) plus CHOP (R-CHOP) is the standard of care for patients with diffuse large B-cell lymphoma (DLBCL). The International Prognostic Index (IPI) and revised IPI were reported as prognostic indicators for DLBCL in 1993 and 2007, respectively. Although they are widely accepted, the performance status (PS) factor is sometimes ambiguous or subjective. Therefore, we developed a new prognostic index, the SIL, that includes only three objective prognostic factors: the clinical stage (S), a soluble interleukin-2 receptor level >2,500 U/mL (I), and an elevated lactate dehydrogenase level (L) (Cancer Sci. 2012). This study was conducted to confirm the value of the SIL index in a larger cohort and in each risk stratification of patients and to validate the SIL index in an independent patient cohort.

Methods: Between 2003 and 2012, we registered and treated 781 consecutive patients with DLBCL, excluding those with mediastinal large B-cell lymphoma, intravascular large B-cell lymphoma, and primary effusion lymphoma. All the included patients were scheduled to undergo primary therapy with six cycles of full-dose R-CHOP. Patients in whom the initial therapy dose was reduced by >20% were excluded. Finally, 572 of 781 patients were retrospectively analyzed. Patients with partial remission (PR) after the initial four cycles underwent eight R-CHOP cycles in total, whereas those who did not achieve PR after the initial four R-CHOP cycles or those who exhibited disease progression at any given time received salvage therapy. If deemed necessary by the attending physician, additional local irradiation was performed in patients with PR or complete remission. Furthermore, we verified the value of the SIL index in an independent cohort of 89 DLBCL patients.

Results: The median age at diagnosis was 63 years (range, 18-89 years). The median number of therapy cycles was 6 (range, 1-8), and 90% of patients received >6 cycles. Sixty-one patients (11%) received radiation therapy as primary treatment, which was often used to treat sites of residual masses at the end of chemotherapy. The median observation time for survivors was 55 months (range, 1-131 months). For 572 patients, the 5-year progression-free survival (PFS) and 5-year overall survival (OS) rates were 70% and 81%, respectively. The 5-year PFS rate was significantly different as 86%, 73%, 63%, and 41% for 0, 1, 2, and 3 of SIL index, respectively (Fig 1; P < 0.0001). The 5-year OS rate was also significantly different as 92%, 87%, 78%, and 52% for 0, 1, 2, and 3 of SIL index, respectively (P < 0.0001). According to the SIL index, 367 (64%) and 205 patients (36%) were classified as having standard (SIL index: 0 or 1) and high (SIL index: 2 or 3) risks, respectively. In patients with a low/low-intermediate risk on the IPI, 84% were categorized as having standard risk according to the SIL index, whereas in patients with a high-intermediate/high risk on the IPI, 82% were categorized as having high risk according to the SIL index. Five-year PFS rates in the standard and high risk groups according to the SIL index were 79% and 53%, respectively (Fig 2; P < 0.0001). Five-year OS rates in the standard risk and high risk groups were 90% and 66%, respectively (P < 0.0001). Cox regression analysis of the SIL index, age (>60 years), PS (2-4), sites of extranodal involvement (>1), and sex showed that the SIL index (P <0.0001; hazard ratio [HR]: 2.38) and PS (P = 0.005; HR: 1.73) were independent risk factors for PFS. Similarly, the SIL index (P < 0.0001; HR: 2.62) and PS (P = 0.006; HR: 1.89) were independent risk factors for OS. When patients were divided into two groups by age (<60 years and >60 years), the SIL index was a good prognostic indicator for PFS and OS in both groups. When they were divided by the number of extranodal involvement sites (0-1 and >1), and sex, the SIL index was still a good prognostic indicator for PFS and OS in both groups. Lastly, when they were divided by the PS (0-1 and 2-4), the SIL index was effective in the good PS group. However, in the poor PS group, the SIL index showed a statistically significant difference in the OS, but not in the PFS. In the validation cohort analysis, 5-year PFS rates in the standard and high risk groups were 81% and 49%, respectively (Fig 3; P = 0.001). Five-year OS rates in the standard risk and high risk groups were 87% and 59%, respectively (P = 0.003).

Conclusion: The SIL index is a simple and objective prognostic indicator for DLBCL patients treated with R-CHOP.