The development of precision medicine approaches for DLBCL (Diffuse Large B Cell Lymphoma) is complicated by its genetic, phenotypic and clinical heterogeneity. Current classification methods do not fully explain the observed differences in clinical outcomes to current chemotherapy and targeted therapy. Therefore, better analytical methods to classify and predict DLBCL patients’ treatment response are needed.
Investigators at the NCI have developed LymphGen; a statistical algorithm that classifies a patient’s DLBCL as belonging to one or more types of genetic subtypes. The algorithm uses a Bayesian model to quantify the probability that a patient’s biopsy belongs to one of six DLBCL subtypes. In this way, LymphGen surpasses current clustering-based algorithms which are unsuitable for patient classification due to their sensitivity to other cohort samples. This classification provides biological insight into the oncogenic mechanisms driving each lymphoma subtype. Investigations showed that LymphGen genetic subtypes were prognostic of patients’ treatment responses to current options – such as R-CHOP chemotherapy, and targeted therapy – such as the BTK inhibitor ibrutinib. Patient samples can be derived from numerous sources, including fresh frozen tissue, formalin-fixed paraffin-embedded tissue (FFPE), or data from a subset of human genes present in a targeted sequencing platform.
In summary, NCI investigators developed a statistical framework that surpasses current algorithms in classifying DLBCL into genetic subtypes to better predict treatment response. The investigators are seeking licensing and/or collaborative research partners for further clinical development.