A Variety of T-Cell Subsets Contribute to CD4+ T-Cell Infiltration in Diffuse Large B-Cell Lymphoma and Both Total CD4+ and CD4+FoxP3+ T-Cell Numbers Predict Clinical Outcome,

CD4+ T-cells can be distinguished into subsets on the basis of surface marker expression and growth factor production. Follicular helper T-cells (Tfh cells) are characterized by the co-expression of surface markers (CD4, ICOS, PD1 and CXCR5) and nuclear BCL6. Normal germinal centre formation requires Tfh cells but is repressed by another CD4+ T-cell subset, Tregs, (demonstrating CD4 and CD25 expression with nuclear FoxP3). The numbers and architecture of infiltrating T-cells predict clinical outcome in follicular lymphoma but although T-cells are a component of diffuse large B cell lymphoma (DLBCL), the relative numbers of CD4+ T-cells and their Tfh and Treg subsets or their association with clinical outcome is not known. We used immunohistochemistry to investigate infiltration by total CD4+, Treg and Tfh cells in cases (n=23) from one centre. The male:female was 1.3:1.0, the age range was 30 to 78 years (median 65 years) and the anticipated association between overall survival and LDH (logrank test, P=0.02) was observed. Patients were treated with R-CHOP with a 21-day cycle. Histological sections were stained with anti-CD4, anti-PD1 and anti-FoxP3 antibodies. For each antibody the area of staining was measured using ImageJ software from 10 high power fields from the same area of each histological section. Tfh cells were identified by strong surface expression of PD1 and Tregs by nuclear expression of FoxP3. CD4+ T-cell infiltration varied by ∼50-fold, and could be diffuse or focal. In 13 cases (57%) the majority of CD4+ T-cells were neither FoxP3+ nor PD1+. Total CD4+ T-cell numbers were positively correlated with FoxP3 (P=0.04) (Figure 1) and with PD1 (P=0.009) (Figure 2) expressing cells suggesting that these subsets were expanded as part of a reaction to the lymphoma capable of stimulating several CD4+ T-cell subsets. High CD4+ (Figure 3) and PD1+ staining predicted good clinical outcome (logrank test, P=0.08) with median survival not being reached at 5 years, but the amount of FoxP3+ staining appeared to be a superior prognostic marker (logrank test, P=0.0069) (Figure 4). There was no association between the cell of origin classification of DLBCL (GCB or ABC) as defined immunohistochemically, and CD4, FoxP3 or PD1 expression. In summary, we have shown that numbers of infiltrating CD4+ T-cells vary between cases of DLBCL and comprises several T-cell subsets including Treg and Tfh cells. No consensus has been reached on the clinical significance of FoxP3+ cell infiltration in DLBCL. Whilst some workers have shown FoxP3 to be associated with a good clinical outcome (Tzankov A., et al. 2008; Lee N., et al. 2008), others have not found a relationship to prognosis (Hasselblom S. et al., 2007). Our data shows that the FoxP3+ Treg cell subset is associated with good clinical outcome but surprisingly we found that both increased total CD4+ T-cells and PD1+ Tfh cells also carry a good prognosis.