AITL and PTCL-U, the two most common forms of T-cell lymphomas in western countries, usually present as nodal disease and pursue an aggressive clinical course. AITL is commonly associated with a constellation of clinical symptoms and distinct pathological features. Conversely, PTCL-U lacks precise diagnostic criteria, and by default comprises cases not fulfilling criteria for other entities, including tumors with borderline features to ALCL and AITL. The genetic alterations and pathogenic mechanisms underlying AITL and PTCL-U are largely unknown. To determine whether the molecular signature of AITL and PTCL-U could help in distinguishing both entities and in understanding ther ontogeny, we performed gene expression profile (GEP) analysis of 15 PTCL-U tissue samples (6 CD30+ and 9 CD30−) and 19 AITL samples (including 2 sorted tumor cell suspensions) using Affymetrix HG-U133A Plus2.0 pan-genomic oligonucleotide microarrays, with comparison to that of previously published normal T-cell subsets (

J Immunol 173:68
;
J Immunol 175: 7837
; Blood 104: 1952). Principle component analysis (PCA, accumulated variance 95%) of all 33 tissue samples yielded three groups of tumors: one group of 12 AITLs, one group of 10 PTCLs-U and one mixed group comprising 5 AITLs (some with features borderline to PTCL-U) and 6 PTCLs-U (including 5 of 6 CD30+ tumors). The AITL molecular signature consisted of 442 genes with increased levels of expression in AITL compared to PTCL-U (t test, p<0.002), including genes encoding cell adhesion molecules, immune receptors, extracellular matrix components and several chemokines, B-cell-related and follicular dendritic cell-related genes, genes involved in endothelial and vascular biology, and several genes reported to belong to the gene expression signature of normal TFH cells (CXCL13, BCL6, PDCD1, CD40L, CD200). To specifically address the question of a molecular link beween AITL and TFH cells, we performed gene set enrichment analysis (GSEA) of our dataset using published gene sets specific of distinct normal T-cell subsets (TFH, TH1, TH2). Compared to that of PTCL-U, the molecular signature of AITL was significantly enriched in TFH-specific genes, and the enrichment was even higher for sorted AITL cells compared to AITL tissues. GSEA failed to disclose a molecular link between PTCL-U and known T-cell subsets (TH1, TH2, TFH). Compared to CD30− PTCL-U, CD30+ PTCL-U had lower expression of genes involved in TCR signalling (t test, p<0.002), and showed molecular similarities with ALK-negative ALCL. In conclusion, GEP of non-anaplastic nodal PTCL (1) segregates AITL and PTCL-U, supporting the basis for histotyping; (2) shows molecular analogies between AITL and TFH cells, strongly supporting the hypothesis of a histogenetic link; (3) suggests molecular analogies between CD30+ PTCL-U and ALK-negative ALCL.

Disclosure: No relevant conflicts of interest to declare.

Author notes

*

Corresponding author

Sign in via your Institution