Introduction

DLBCL is the most common non-Hodgkin lymphoma and a heterogeneous subtype from the biological and clinical point of view. 20-40% of the patients relapse so it is important to identify a high risk population that could benefit from different therapeutic approaches. For this purpose, several scores have been developed. The IPI has been the most widely used but lacks the ability to identify this very high risk population in Rituximab era. Low absolute lymphocyte count and high levels of blood monocytes have shown to be unfavorable risk factors. The red cell distribution width (RDW) has been associated with aging and active inflammatory processes and beta-2-microglobulin (B2M) with tumor load and proliferation as well as comorbidities such as kidney failure. All these variables are easily obtainable at the time of diagnosis.

Patients and methods

From the national database of GELTAMO-IPI Project, we selected those patients with DLBCL homogeneously treated with R-CHOP ± radiotherapy. The complete blood count (CBC) values were standardized using the normal reference values of each centre. The survival analysis was estimated with Kaplan-Meier method. The comparison between variables was performed through Log-Rank test and multivariate analysis with Cox regression. The CBC quantitative variables cut points were calculated through MAXSTAT. A new prognosis score was generated taking into account the results of multivariate analysis for progression free survival (PFS), spliting the sample in two cohorts (discovery and validation).

Results

Nine hundred and ninety-two patients with DLBCL were retrospectively analyzed with a median follow up of 55 months (12-185). The characteristic of the patients are summarized in table 1. In the multivariate analysis, age, ECOG, stage, bulky mass, B2M, RDW and lymphocytes/monocytes ratio (LMR) were significantly independent variables for PFS. A new prognosis score was generated with these variables including age categorized in 3 groups (18-64, 65-80 y >80) with 0, 1 y 2 points, ECOG>3-4 with 2 points, stage III-IV, bulky mass, high B2M, LMR<2.25 and RDW>0.96 with 1 point each for a maximum of 9. This score could improve the discrimination of a very high risk subgroup with a 5-year PFS of 17% (Figure 1) versus 45%, 52%, 32% or 29% of R-IPI, NCCN-IPI, R-TS and GELTAMO-IPI, respectively and 5-year overall survival (OS) of 20% (Figure 2) versus 59%, 46%, 45% and 40% for R-IPI, NCCN-IPI, R-TS and GELTAMO-IPI.

Conclusion

A new prognosis score including easily obtainable CBC variables (LMR and RDW) and B2M is presented. This score identifies a high risk population both for PFS and OS in comparison with other scores for DLBCL. The addition of other biological or image markers could improve these results.

Disclosures

Martín:Roche: Consultancy, Honoraria, Other: Travel expenses; Janssen: Honoraria, Other: Travel expenses; Celgene: Consultancy, Honoraria, Other: Travel expenses; Servier: Honoraria, Other: Travel expenses. Sancho:SERVIER: Honoraria; JANSSEN: Honoraria, Speakers Bureau; MUNDIPHARMA: Honoraria; SANOFI: Honoraria; GILEAD: Honoraria, Research Funding; KERN FHARMA: Honoraria, Speakers Bureau; CELGENE: Honoraria; ROCHE: Honoraria, Speakers Bureau.

Author notes

*

Asterisk with author names denotes non-ASH members.

Sign in via your Institution