Abstract 4312
Background
It is known that diffuse large-B-cell lymphoma (DLBCL) is a clinically heterogeneous entity. The most important clinical predictor of survival is the International Prognostic Index, which does not provide information regarding the heterogeneous biology of tumors. Two major subtypes of DLBCL have been identified by gene expression profiling (GEP) and classified by cell of origin into germinal center B-cell–like (GCB) and activated B-cell–like (ABC). GEP has become a reliable method for predicting the outcome of patients with DLBCL treated with R-CHOP chemotherapy. However, that it´s not easily applicable in clinical practise. Several IHC algorithms have been developed to assign patients into GCB and non-GCB/ABC subtypes.
Methods
We retrospectively analyzed 142 patients diagnosed of de novo DLBCL from 1999 to 2017 at our Hospital treated with chemoimmunotherapy. DLBCL was classified using the Hans algorithm into GCB and non-GCB subtypes. The primary end point was progression-free survival (PFS) according to the Hans algorithm, that it was estimated by the Kaplan–Meier method.
Results
The percentage of GCB and non-GCB subtypes was 54% and 46%, respectively. After a median follow-up of 37 months, the median progression-free survival was 100 months in the global population. No significant differences were found in PFS, although there was a trend to favor CGB subtype (PFS at 24 months 70% in CGB group and 59% in non-CGB group, with a median of 60 months in non-CGG and not reached in CGB group, p = 0.177). Despite of being a retrospective study and the low median follow-up of patients, in CGB subtype there was a trend towards better overall survival (OS) (2-year OS: 72% vs. 68%), not statistically significant (p = 0.661).
Conclusions
In our study there is a lack of evidence supporting the use of the Hans algorithm for stratifying patients into distinct prognostic groups, probably due to the low median follow-up. Rather, GEP remains the preferred method for predicting prognosis. IHQ for subclassification of DLBCL is feasible and reproducible, but the harmonization of techniques and centralized consensus review is necessary.
Clinical trial identification
Legal entity responsible for the study
Laura Galvez Carvajal.
Funding
Has not received any funding.
Editorial Acknowledgement
Disclosure
All authors have declared no conflicts of interest.