Related Resources for this page:

Molecular prognostic models

Gene expression profiling

  • The biological heterogeneity of DLBCL emphasizes the need for a patient-specific and genetically based approach.36 The identification of molecular risk factors based on individual patients' diseases will help facilitate personalized treatments that may result in improved clinical outcomes6
  • DLBCL patients can be divided into clinical risk groups based on results from gene expression profiling and DLBCL prognostic signatures7,37
    (Figure 2)
    • Germinal center B cell (GCB) DLBCL has a gene expression profile that is characteristic of B cells of the germinal center of the lymph node. Patients with GCB-derived disease have been shown to have better survival outcome compared with patients with non-GCB–derived disease25
    • Activated B cell (ABC) DLBCL is derived from B cells that are in the process of differentiating from germinal center B cells to plasma cells.7,38 ABC patients have been found to have poorer outcomes than GCB patients25
    • Primary mediastinal large B-cell lymphomas (PMBL) express a gene signature similar to the malignant cells of Hodgkin's lymphoma. PMBL characteristically arises in the mediastinum and a thymic remnant is often found associated with tumor mass.38 This subgroup is a clinically favorable subgroup of DLBCL39

Figure 2. Prognosis based on cell of origin of DLBCL

Figure 2. Prognosis based on cell of origin of DLBCL.40

  • Many different studies applying gene expression profiling have shown that clinical outcomes can vary among identified patient-risk subgroups7,32,37,41
    • Recently, a 6-gene model has been shown to predict overall survival in DLBCL and shows potential as a prognostic marker for outcomes in both chemotherapy-only and chemoimmunotherapy settings41 (Figure 3)

Figure 3. Prediction of outcome with the 6-gene model among patients with DLBCL treated with chemotherapy or chemoimmunotherapy

Figure 3. Prediction of outcome with the 6-gene model among patients with DLBCL treated with chemotherapy or chemoimmunotherapy.41

Immunohistochemistry

  • The prognosis of patients with DLBCL also can be determined with immunohistochemistry. Unlike gene expression profiling, which is based on the presence of mRNA transcripts, immunohistochemistry is used to characterize protein biomarkers expressed by DLBCL cells42-44 (eg, Hans algorithm, Figure 4 and refined Choi algorithm, Figure 5)

Figure 4. GCB vs non-GCB subtypes of DLBCL: Immunophenotypic classification

Figure 4. GCB vs non-GCB subtypes of DLBCL: Immunophenotypic classification.42

Figure 5. The refined Choi algorithm, a new immunostain algorithm that classifies DLBCL into molecular subtypes with high accuracy

Figure 5. The refined Choi algorithm, a new immunostain algorithm that classifies DLBCL into molecular subtypes with high accuracy.44

  • The subclassification by gene expression for GCB and non-GCB DLBCLs can be correlated with immunostaining for protein expression.42 The expression of many individual proteins as detected by immunohistochemistry has been shown to have prognostic significance6 (Table 7)

Table 7. Examples of individual immunophenotype features with reported prognostic significance in DLBCL

Table 7. Examples of individual immunophenotype features with reported prognostic significance in DLBCL.6,31,45

Molecular prognostic tools: Risk stratification and treatment strategy

  • While the molecular classification of DLBCL is possible, its incorporation into routine clinical practice is not yet a reality28
    • Current data demonstrate that there is no consensus regarding the prognostic models to describe the heterogeneity of DLBCL
  • Initial gene expression profiling and immunohistochemical profiling studies have determined survival differences in patients with GCB and ABC subtypes of DLBCL, yet in the era of chemoimmunotherapy, the potential role of these prognostic tools in treatment decision making remains unclear6,34

View References from this site section