Follicular helper T-cell–related lymphomas

Stefano A. Pileri

In this issue of Blood, Wang et al describe the occurrence and pathogenetic relevance of IDH2R172 mutations in angioimmunoblastic T-cell lymphoma (AITL).1

The images referred to the same case at (A) disease onset, (B) first relapse, and (C) second relapse. The tumor consists of cells characterized by a clear rim of cytoplasm and TFH phenotype as defined by the expression of PD1, CXCL13, BCL6, CD10, and ICOS (not shown); in panel A, it has a “follicular” growth pattern (hematoxylin and eosin, ×20), and in panel B, it has the typical “angioimmunoblastic” appearance (please note the hyperplastic HEV and the inflammatory component [hematoxylin and eosin, ×400]). In panel C, the “angioimmunoblastic” characteristics are lost and neoplastic cells acquire at least in part a more blastic appearance (hematoxylin and eosin, ×400).

A follicular helper T-cell (TFH)-related gene signature was for the first time reported in AITL by de Leval et al in 2007.2 However, these authors found that a small group of peripheral T-cell lymphoma, not otherwise specified (PTCL/NOS) bore a similar signature.2 Later, an increasing number of antibodies detected TFH-associated molecules (ie, CD10, B-cell lymphoma 6 [BCL6], programmed cell death-1 [PD1], CXC chemokine ligand 13 [CXCL13], CC chemokine receptor 5, signaling lymphocytic activation molecule–associated protein [SAP], and inducible T-cell costimulator [ICOS]), their use possibly surrogating for gene expression profiling (GEP).2,3 It was agreed that at least 3 of these antibodies should simultaneously be positive to suggest the derivation of a given T-cell neoplasm from TFH cells because a single marker might occur due to cell plasticity.4 In 2008, the World Health Organization (WHO) Classification of Tumors of Hematopoietic and Lymphoid Tissues, identified a variant of PTCL/NOS, characterized by a follicular growth pattern and sustained by TFH-related cells.5 It was not lumped with AITL because it lacked hyperplasia of follicular dendritic cells (FDCs) and high endothelial venules (HEVs), that is, the morphologic hallmarks of AITL.5 In addition, in about 20% of cases, t(5;9) is present, causing the formation of the hybrid gene ITK/SYK, found only in these tumors.5 In 2011, Agostinelli et al found that a wide spectrum of T-cell neoplasms carried the TFH phenotype, morphologically corresponding to AITL and PTCLs/NOS with and without follicular growth pattern.6 As opposed to the follicular variant, the latter showed a diffuse effacement of the lymph node structure. When compared with AITL, these cases lacked FDC and HEV hyperplasia and did not show a residual open marginal sinus but contained variable amounts of B blasts (either Epstein-Barr virus positive or negative [EBV+ or EBV]) intermingled with neoplastic T cells. In some cases with serial biopsies, transition was observed from 1 pattern to the other (eg, follicular at disease onset, AITL at the time of the first relapse, and PTCL/NOS at the second relapse [see figure]). Such observations, along with the distinctive morphology of neoplastic cells (small-medium size, slightly irregular nuclear profile, and wide rim of clear cytoplasm) and the common phenotype, led to the idea that AITL and TFH-related PTCLs/NOS with and without follicular growth pattern might represent different aspects of the same family of neoplasms.6 This concept was strengthened by the detection of the same gene mutations in AITL and PTCLs/NOS, although prevalence is different.7-9 These mutations affected TET2, IDH2, DNMT3A, and RHOA among others.7-9 Last but not least, ITK/SYK was also detected in AITL.10 This array of mutations is relevant for 2 reasons. First, they can play an important role in the process of lymphomagenesis, although, as correctly pointed out by Wang et al,1 some mutations like those affecting TET2 and DNMT3A may represent early events comparable to BCL2 rearrangement in the setting of follicular lymphoma. Second, they may have therapeutic implications.1

In this issue, Wang et al report on a study comparing TET2, IDH2, and DNMT3A mutations with GEP in 90 PTCLs of the AITL, NOS, and anaplastic large-cell lymphoma (ALCL) types.1 Interestingly, whereas TET2 and DNMT3A mutations assessed by targeted resequencing were observed in all 3 categories, although with different frequencies, IDH2 detected by Sanger sequencing occurred most frequently in AITL (32.8% vs 5% of NOS and 0% of ALCL). Moreover, the previously reported occurrence of IDH2 mutations at R172 was confirmed, as was the common co-occurrence with TET2 mutations. These findings are at variance with acute myeloid leukemia and glioblastoma, in which IDH2 mutations are at R140 and are mutually exclusive of the TET2 mutations. In particular, the IDH2R172-mutated cases showed a distinct GEP among AITLs and the IDH2/TET2 double-mutant cases carried upregulation of TFH-associated genes and downregulation of genes associated with T helper 1 (Th1), Th2, and Th17 phenotypes. These double-mutant AITLs were highly enriched for the signature of CD4+ T cells stimulated by interleukin-12, suggesting a more polarized TFH phenotype. Finally, yet importantly, Wang et al provide experimental evidence that IDHR172 mutations produce a significant increase in H3K27me3 and DNA hypermethylation of genes involved in T-cell receptor signaling and T-cell differentiation that likely contribute to lymphomagenesis in AITL. In addition, these findings give a strong rationale for the usage of hypomethylating agents in AITL treatment.

The Wang et al article highlights how sequencing analyses can help dissect apparently homogenous neoplasms using conventional techniques. This study provides new insights on the pathogenesis and subclassification of these tumors as well as on the detection of novel and hopefully more effective therapeutic targets.1 The latter represent a real need because most PTCLs have a dismal prognosis when treated with current therapies.5


  • Conflict-of-interest disclosure: The author declares no competing financial interests.