Biology and Molecular Pathogenesis of Mature T-Cell Lymphomas

Abstract

Peripheral T-cell lymphomas (PTCLs) constitute a highly heterogeneous group of hematological diseases with complex clinical and molecular features consistent with the diversity of the T-cell type from which they originate. In the past several years, the systematic implementation of high-throughput genomic technologies for the analysis of T-cell malignancies has supported an exponential progress in our understanding of the genetic drivers of oncogenesis and unraveled the molecular complexity of these diseases. Recent findings have helped redefine the classification of T-cell malignancies and provided novel biomarkers to improve diagnosis accuracy and analyze the response to therapy. In addition, multiple novel targeted therapies including small-molecule inhibitors, antibody-based approaches, and immunotherapy have shown promising results in early clinical analysis and have the potential to completely change the way T-cell malignancies have been treated traditionally.

Peripheral T-cell lymphomas (PTCLs) are a heterogeneous group of highly aggressive malignant hematologic tumors that arise from clonal proliferation of mature post-thymic T cells (Gaulard and de Leval 2014). PTCLs account for ∼15% of all non-Hodgkin lymphomas and cover almost 30 different entities (Swerdlow et al. 2016). Compared with B-cell lymphomas, PTCLs are less common, are more difficult to diagnose and classify, are more aggressive, and have inferior outcomes with current treatments. Indeed, with rare exceptions, the prognosis for PTCL is dismal, with survival rates in the 20%–30% range at 5 years (Savage 2008). T-cell lymphomas are also less understood in terms of their normal T-cell counterparts, and until recently, most subtypes lacked defined drivers that could potentially be targeted for the development of precision therapies. Interestingly, the incidence of PTCL, which generally affects the elderly population, has been consistently increasing over the past decades, but unfortunately, this increase has not been accompanied by improvements in survival (Weisenburger et al. 2011).

However, over the past few years, genomic and transcriptomic studies have substantially advanced our understanding of the molecular subtyping of PTCL and facilitated the identification of molecular vulnerabilities in these tumors, opening the field for the development of tailored therapies targeting specific driver genetic alterations that will hopefully be translated into better clinical outcomes.

CLASSIFICATION AND MOLECULAR ETIOLOGY OF T-CELL LYMPHOMA

The fourth edition of the World Health Organization (WHO) classification established the guidelines for classification of hematopoietic and lymphoid tumors in 2008, with a major revision published in 2016 primarily to address several major advances in the field with significant clinical and biological implications in patient management (Swerdlow et al. 2016). Many of these changes are the result of genomic studies, including analysis of Gene Expression Profile (GEP) and characterization of the genetic landscape of T-cell and natural killer (NK)-cell neoplasms. This revised classification of the lymphoid neoplasms included the addition of a number of provisional entities highlighted in Table 1. In this review, we will describe the most frequent subtypes of PTCL, as well as the most common genetic aberrations identified for each subgroup.

Table 1.

Classification of mature T and NK neoplasms according to WHO 2016 revision

Angioimmunoblastic T-cell lymphoma

Peripheral T-cell lymphoma, NOS

Follicular T-cell lymphoma*

Nodal peripheral T-cell lymphoma with TFH phenotype*

Anaplastic large-cell lymphoma, ALK+

Anaplastic large-cell lymphoma, ALK−*

Breast implant–associated anaplastic large-cell lymphoma*

T-cell prolymphocytic leukemia

T-cell large granular lymphocytic leukemia

Chronic lymphoproliferative disorder of NK cells

Aggressive NK-cell leukemia

Systemic EBV+ T-cell lymphoma of childhood*

Hydroa vacciniforme–like lymphoproliferative disorder*

Adult T-cell leukemia/lymphoma

Extranodal NK/T-cell lymphoma, nasal type

Enteropathy-associated T-cell lymphoma

Monomorphic epitheliotropic intestinal T-cell lymphoma*

Indolent T-cell lymphoproliferative disorder of the GI tract*

Hepatosplenic T-cell lymphoma

Subcutaneous panniculitis-like T-cell lymphoma

Mycosis fungoides

Sézary syndrome

Primary cutaneous CD30+ T-cell lymphoproliferative disorders

Lymphomatoid papulosis

Primary cutaneous anaplastic large cell lymphoma

Primary cutaneous γδ T-cell lymphoma

Primary cutaneous CD8+ aggressive epidermotropic cytotoxic T-cell lymphoma

Primary cutaneous acral CD8+ T-cell lymphoma*

Primary cutaneous CD4+ small/medium T-cell lymphoproliferative disorder*

Data adapted from Swerdlow et al. 2016.

Provisional entities are listed in italics and changes from the 2008 classification are indicated by asterisks (*). Entities that will be discussed in this review are shown in bold.

(NK) Natural killer, (NOS) not otherwise specified, (TFH) T-follicular helper, (ALK) anaplastic lymphoma kinase, (EBV) Epstein–Barr virus, (GI) gastrointestinal.

ANGIOIMMUNOBLASTIC T-CELL LYMPHOMA

Angioimmunoblastic T-cell lymphoma (AITL) is the second most common PTCL subtype worldwide and the most frequent subtype in the Western world, where it represents >30% of PTCL cases (Federico et al. 2013). AITL develops with a unique pathobiology and clinical presentation including a chronically activated immune system with frequent hypergammaglobulinemia and autoimmune features (Lunning and Vose 2017). AITL is characterized by the presence of clonal CD4⁺ neoplastic T cells, a highly reactive tumor microenvironment, and clonal expansions affecting the B-cell compartment, frequently positive for Epstein–Barr virus (EBV) (Zhou et al. 2007). Initial gene expression profiling studies contributed to establish T follicular helper (TFH) cells as the cell of origin in AITL (de Leval et al. 2007), based on a characteristic expression of TFH markers in the malignant T cells. Indeed, the differential diagnosis criteria for AITL requires the concurrent presence of at least two TFH markers, including PD-1, CXCR5, ICOS, BCL6, CXCL13, CD10, and SAP, to confirm the diagnosis of AITL.

Genetic alterations affecting epigenetic regulators, including TET2, DNMT3A, and IDH2, are one of the hallmarks of AITL (Cortés and Palomero 2016). The ten-eleven translocation 2 (TET2) encodes a 2-oxoglutarate/Fe2⁺–dependent oxygenase that participates in the epigenetic control of gene expression by catalyzing the oxidation of DNA 5-methylcytosine to 5-hydroxymethylcytosine (Tahiliani et al. 2009; Ito et al. 2010; Cimmino et al. 2011). First identified as a tumor suppressor in myeloid malignancies (Abdel-Wahab et al. 2009; Delhommeau et al. 2009), multiple studies have shown a high prevalence of loss-of-function mutations in TET2 in PTCLs particularly in AITL (∼50%) and PTCL-not otherwise specified (NOS) (∼40%), in which they are frequently identified as an initial event in the process of transformation (Couronné et al. 2012; Palomero et al. 2014; Sakata-Yanagimoto et al. 2014; Yoo et al. 2014) and are associated with worse outcome (Lemonnier et al. 2012). In a recent study, analysis of TET2 mutations and TCR rearrangements in AITL and PTCL-NOS cases demonstrated the presence of biclonal populations derived from a common TET2 mutant progenitor, suggestive of parallel tumor evolution from common hematopoietic progenitor pool (Yao et al. 2020). Loss of Tet2 has been associated with enhanced hematopoietic stem cell (HSC) self-renewal and an increased HSC pool (Moran-Crusio et al. 2011). Moreover, TET2 mutations are frequently found in the blood of healthy elderly individuals, a condition termed clonal hematopoiesis of indeterminate potential (CHIP), and are associated with increased risk for developing a hematologic malignancy (Busque et al. 2012; Steensma 2018; Steensma and Ebert 2020). The relative high frequency of clonal hematopoiesis in otherwise healthy individuals indicates that these initial TET2 mutations are not sufficient for transformation; however, mutant premalignant cells can fully transform into diverse malignancies upon acquisition of secondary genetic hits.

DNMT3A encodes a DNA methyltransferase that controls cytosine methylation and is recurrently affected by loss-of-function mutations in 10%–40% of AITL patients. Mutations in DNMT3A are considered an initial event in the process of transformation and interestingly frequently co-occur with TET2 mutations (Cortés and Palomero 2016). Indeed, cooperation between loss-of-function mutations in Dnmt3a and Tet2 has been shown to lead to malignant transformation in mouse models (Scourzic et al. 2016).

Finally, IDH2 is a metabolic mitochondrial enzyme involved in the generation of 2-oxoglutarate that is specifically mutated in 30%–40% of AITL patients (Cortés and Palomero 2016). Mutant IDH2 forms have a neomorphic enzymatic activity leading instead to the generation of 2-hydroxyglutarate, an oncometabolite that antagonizes the activity the TET family of enzymes (Cortés and Palomero 2016). AITL is the only PTCL subgroup in which IDH2 mutations are found, and, remarkably, they occur almost exclusively in position R172 (R172K, R172S) (Cairns et al. 2012), which is associated with increased production of 2-HG compared to other IDH2 mutant alleles. Although mutations in TET2 and IDH2 are also present in myeloid malignancies, where they are mutually exclusive, AITLs frequently present co-occurring mutations in IDH2 and TET2 (Cortés and Palomero 2016). Interestingly, IDH2 mutations in AITL are presumed to constitute a secondary hit and might refine the differentiation of the premalignant clones. Indeed, the co-occurrence of TET2 inactivation and IDH2 mutations is associated with a potential cooperative effect in the regulation of the expression of TFH specific genes, leading to a more polarized TFH signature than achieved by the presence of TET2 mutations alone (Wang et al. 2015a).

Gene expression profiling studies have proposed that AITL tumors may be driven by increased T-cell receptor signaling. Consistently, genomic studies have uncovered the presence of highly recurrent mutations affecting T-cell receptor–initiated signaling pathways, including the frequent mutations in the small GTPase RHOA, as well as genetic alterations affecting VAV1, PLCG1, CD28, and the tyrosine kinase FYN (Vallois et al. 2016).

A central role of RHOA in the pathogenesis of AITL is supported by the identification of a recurrent, highly prevalent heterozygous missense RHOA G17V mutation detected in up to 70% of the patients with AITLs (Fig. 1; Palomero et al. 2014; Sakata-Yanagimoto et al. 2014; Yoo et al. 2014). The presence of the RHOA G17V mutation is highly specific of AITL and has become a molecular diagnostic determinant for this PTCL subgroup. Clinical studies have even suggested that the presence of RHOA G17V could be associated with worse prognosis (Nagao et al. 2016; Ondrejka et al. 2016). Mutations in RHOA G17V generally co-occur with mutations in epigenetic regulators, being most frequently associated with loss-of-function mutations in TET2. Mutations in RHOA G17V act as dominant negatives interfering with the signaling initiated by wild-type RHOA (Palomero et al. 2014). Recent studies using engineered mouse models demonstrated that expression of Rhoa G17V in CD4⁺ T cells induces increased proliferation and TFH-cell specification (Cortés et al. 2018; Ng et al. 2018). Notably, the combination of Tet2 loss and expression of Rhoa G17V is sufficient to induce AITL development (Cortés et al. 2018). This result supports a multistep model for AITL development in which loss-of-function mutations in epigenetics factors in early hematopoietic progenitors will be an initiating event in the process of transformation, inducing the generation of multipotent premalignant clones. Upon acquisition of a second hit in RHOA G17V, premalignant cells differentiate toward TFH cells and induce development of AITL (Fig. 1). Both TFH lineage specification and AITL development downstream of RHOA G17V depend on the activation of the ICOS/PI3K/mTOR signaling pathway and are sensitive to PI3K inhibitors (Cortés et al. 2018).

Figure 1.

Schematic (A) representation of the RHOA mutations identified in AITL and model for process of malignant transformation leading to AITL (B). (A, Adapted from Palomero et al. 2014, courtesy of Springer/Nature © 2014; B, adapted from Cortés et al. 2018, with permission from Elsevier © 2018.)

Mutations in VAV1 have been identified in up to 10% of AITL and PTCL-NOS patients. VAV1 encodes a hematopoietic-specific Rho family-specific guanine exchange factor (Bustelo 2000). Genomic alterations in VAV1 in PTCL include a recurrent in-frame deletion mutation (VAV1 Δ778–786) generated by alternative splicing mechanism as well as multiple VAV1 gene fusions affecting the carboxy-terminal domain of VAV1 that lead to its constitutive activation (Abate et al. 2017). Alterations in RHOA and VAV1 are mutually exclusive, suggesting they could be affecting the same genetic pathway (Fujisawa et al. 2018).

Recurrent mutations in CD28, a major co-stimulatory molecule for the T-cell receptor (TCR), have also been described in AITL (Rohr et al. 2015). PTCL-associated CD28 mutations affect the D124 and T195 residues and increase ligand–receptor interaction and signal transduction. Notably, CD28-mutated AITL patients have inferior survival compared to nonmutated cases (Rohr et al. 2015). Finally, recurrent activating mutations in FYN, a SRC family kinase, have been found in ∼3% of PTCL-NOS and AITL cases (Palomero et al. 2014).

Similar genomic aberrations to the ones identified in AITL have been described for follicular T-cell lymphoma and nodal peripheral T-cell lymphoma with TFH phenotype (see Table 1). This has led to the grouping of these two provisional entities with AITL. All three groups share similar gene expression patterns, TFH-associated immunophenotypic markers, and genetic abnormalities; however, despite their common TFH phenotype, these diseases have differing clinical and pathological features and, as such, remain distinct entities within the classification (Swerdlow et al. 2016).

PTCL-NOS

PTCL-NOS includes mature T-cell lymphoma cases that cannot unequivocally be assigned to other subtypes, making this group highly heterogeneous and poorly understood. PTCL-NOS is the most common entity in mature T-cell neoplasms; however, the diagnosis of PTCL-NOS has decreased in recent years because of the introduction of molecular profiling that allows for better lymphoma classification.

The characterization of the molecular landscape of PTCL-NOS has identified the presence of frequent mutations epigenetic regulators, including KMT2C and KMT2D, KDM6A, ARID1B, SETD1B, CREBBP and ARID2 as well as TET2 and DNMT3A, which are likely associated with the TFH-like group (Palomero et al. 2014). Also, genes involved in signaling pathways are frequent targets of alterations in PTCL-NOS, including genes in the NF-κB and TCR signaling pathways such as TNFAIP3, APC, CHD8, ZAP70, NF1, TNFRSF14, TRAF3, and VAV1 (Palomero et al. 2014; Abate et al. 2017; Watatani et al. 2019). Recent analysis of a large PTCL-NOS cohort has identified additional altered genes, such as YTHDF2, a RNA methyl transferase, and PD-1 (Watatani et al. 2019). More interestingly, a previously undescribed molecular subtype characterized by TP53 and/or CDKN2A mutations and deletions, has been identified associated with worse prognosis, chromosomal instability, and altered immune evasion (Watatani et al. 2019), suggesting a driver role for TP53 in PTCL-NOS (Watatani et al. 2019).

Gene expression profiling in PTCL-NOS has identified two distinct molecular subgroups characterized by the differential high expression of TBX21 or GATA3, two key regulators of Th1 and Th2 cell differentiation (Iqbal et al. 2014; Wang et al. 2015a). Although the precise mechanisms that lead to the increased expression of these master regulators are unknown, they might be related to different cells of origin. More importantly, these groups are associated with different clinical behavior and response to therapy, with expression of GATA3 being linked to worse outcome (Iqbal et al. 2014).

ANAPLASTIC LARGE-CELL LYMPHOMAS

Anaplastic large cell lymphoma (ALCL) includes several subtypes of aggressive lymphoid neoplasms with similar histologic characteristics but distinct genomic features. The latest WHO classification contains two systemic ALCL entities: ALK⁻ ALCL and ALK⁺ ALCL (Swerdlow et al. 2016) in addition to a primarily cutaneous form of ALCL, (cALCL), more closely related to ALK⁻ ALCL, and a new provisional category, breast implant–associated ALCL (BIALCL) (Swerdlow et al. 2016). The neoplastic cells are typically large with abundant cytoplasm and characteristic horseshoe nuclei shapes and characterized by strong expression of CD30 and negative for TCR (Shustov and Soma 2019). Despite the lack of expression of surface TCR, ALCLs are classified as T-cell lymphoma because of their peripheral presentation and the presence of TCR rearrangements that are detectable in the tumor cells at the molecular level that are supportive of a T-cell origin. Interestingly, data from murine models suggests that transformation could be initiated in early thymocytes, before TCR β-rearrangement, and although TCR is still required for thymic egress and development of peripheral murine tumors, down-regulation of the TCR would be required to complete lymphomagenesis. This model is consistent with the characteristics of TCR rearrangements in human ALCL, which are often aberrant and would not be permissive for survival during thymic development (Malcolm et al. 2016)

ALK⁺ ALCL is generally diagnosed in younger individuals of Western origin with slight male predominance (Ferreri et al. 2012). ALK⁺ ALCL constitutes also the most frequent mature aggressive lymphoma within the pediatric population. Moreover, it is also the PTCL subtype with the best overall outcome (Savage 2008). ALK⁺ ALCL is associated with the presence of chromosomal translocations that lead to the constitutive activation of ALK, a receptor tyrosine kinase, being the most frequent t(2;5) translocation (nucleophosmin [NPM]-ALK). ALK translocations lead to increased activation of ERK/MAPK, JAK/STAT, and PI3K signaling pathways. Moreover, ALK fusion-mediated activation of STAT3 has been shown to be required for the maintenance of the transformed phenotype in ALK⁺ ALCL (Chiarle et al. 2005). ALK fusions constitutively activate VAV1 and VAV3 in ALCL, increasing RAC1 and CDC42 activity, down-regulating the Wiskott–Aldrich syndrome protein (WASP) and WASP-interacting protein (WIP) tumor suppressors and promoting lymphoma survival (Choudhari et al. 2016; Menotti et al. 2019).

ALK⁻ ALCL patients are generally older and associated with a worse outcome (Ferreri et al. 2013). Molecular events affecting DUSP22 and TP63 have been reported to be mutually exclusive and predictors of survival outcomes, with expression of DUSP22 associated with a favorable prognosis and TP63 being an unfavorable marker. Recurrent translocations and loss of TP53 and PRDM1/BLIMP1 have also been proved to have a pathogenic role associated with less favorable outcomes (Boi et al. 2013; Parrilla Castellar et al. 2014). In addition, very frequent activating mutations in the JAK/STAT pathways, more specifically in JAK1 and STAT3, have been identified in almost 20% of ALK-ALCL (Crescenzo et al. 2015). Recurrent gene translocations linked to JAK/STAT pathway activation, including NCOR2-ROS1, NFkB2-ROS1, and different fusions affecting the TYK2 kinase, have also been identified (Crescenzo et al. 2015). This suggests that activation of STAT3 is a common mechanism of malignant transformation in ALCL, independently of ALK status, and provides a genetic rationale for the morphologic and phenotypic similarities between ALK⁺ and ALK⁻ ALCL (Fig. 2).

Figure 2.

Recurrent mutations and kinase fusions in anaplastic large cell lymphoma (ALCL) converge to induce oncogenic STAT3 activation in ALK⁺ and ALK⁻ ALCL subtypes. (Figure was created from data adapted from Crescenzo et al. 2015.)

Finally, there is a novel provisional category, BIALCL, arising in association with prolonged exposure to breast implants, for which the factors leading to progression still remain to be determined (Miranda et al. 2014; Laurent et al. 2016).

ADULT T-CELL LEUKEMIA/LYMPHOMA

Adult T-cell leukemia/lymphoma (ATLL) is a mature T-cell lymphoma derived from the transformation of regulatory T cells and associated with the infection by the human T-cell lymphotropic virus type 1 (HTLV-1) (Bangham and Toulza 2011). In ATLL pathogenesis, viral products such as Tax and HBZ have important roles in oncogenic transformation (Giam and Semmes 2016). However, although the Tax gene, a nonstructural gene on the HTLV-1 genome, appears to play a critical role in transformation initiation, it is not critical to sustain tumor cell growth, and it is frequently no longer expressed in transformed cells. HBZ is one of the few genes consistently overexpressed in most ATLL cases (Bazarbachi et al. 2011). Interestingly, <5% of HTLV1-infected individuals will eventually develop ATLL and only after a long latency (Giam and Semmes 2016); thus, other factors are required for transformation in a multistep process that might involve viral and epigenetic mechanisms in conjunction with acquired genomic alterations. ATLL is geographically restricted to those regions where HTLV-1 is endemic, including Japan, Central and South America, the Caribbean, and certain regions in western Africa. In this regard, the introduction of preventative approaches to reduce the spread of HTLV1 infection could have a significant impact on the incidence of ATLL.

Four different clinical presentations of ATLL have been described: acute, lymphomatous, chronic, and smoldering (Matutes 2007). The acute presentation is the most common and severe, being associated with the worst prognosis. Malignant cells have a classical nuclear morphology resembling a cloverleaf, and might express CD30. In most cases, ATLL cells express CD3, CD4, and CD25, and lack CD7, whereas in 10%–15% of cases, there is expression of CD4 and CD8. The cell of origin of ATLL cells is hypothesized to be regulatory T cells based on the expression of FoxP3 in some tumor cells; however, the assignment remains controversial (Malcolm et al. 2016). ATLL cells are also characterized by high expression of CCR4 and CCR7 chemokine receptors, which are involved in ATLL cell migration and infiltration of lymphoid organs and skin (Ohshima 2015).

Analyses of large patient cohorts using next-generation sequencing approaches have delineated the molecular landscape of ATLL and identified recurrent somatic mutations and genetic aberrations affecting multiple elements of the TCR-NF-κB signaling pathways (Kataoka et al. 2015). Thus, gain-of-function mutations and focal amplifications leading to constitutive activation of this pathway have been identified in proximal (CD28, PLCG1, VAV1, RHOA and FYN) and distal (PRKCB and CARD11) elements within the TCR-NF-κB pathway. Fusions involving CD28 such as CTLA4-CD28 and ICOS-CD28 are also recurrent. Mutations or deletions involving p16 and p53 tumor suppressors are common and generally associated with disease progression (Yamada and Kamihira 2005). Frequent mutations affecting CCR4 and CCR7 receptors, mostly involving truncation of the carboxy-terminal domains have been described. These mutations interfere with ligand-induced receptor internalization, leading to enhanced activation of downstream PI3K/AKT signaling (Kataoka et al. 2015). Finally, missense mutations affecting the NOTCH1 receptor have also been described in 30% of ATLL patients. The mutations affect the PEST domain of the NOTCH1 protein preventing the FBXW7-mediated degradation of NOTCH (Pancewicz et al. 2010). Interestingly, FBXW7 is also found mutated in 25% of ATLL (Yeh et al. 2016)

Several key transcription factors in lymphocyte activation and homeostasis are also recurrently mutated in ATLL. Thus, focal amplifications and frequent mutations in IRF4, a major downstream target of NF-κB, confer gain-of-function activity to the mutated protein. GATA3, required for T-cell differentiation, is affected by nonsense and frameshift mutations distributed throughout the coding regions that result in dominant negative isoforms, whereas IKZF2 is frequently affected by intragenic deletions in more than one-third of ATLL patients, leading to abnormally spliced shorter forms that act as dominant negatives and have oncogenic activity in vivo.

Last, molecules associated with immune surveillance, such as HLA-A/B, CD58, and FAS, are recurrently affected by loss-of-function mutations. Structural variations that truncate the PD-L1 3′-untranslated region leading to its overexpression are also frequent (Kataoka et al. 2019). Interestingly, expression of PD-L1 in ATLL cells has been correlated with a poor prognosis (Miyoshi et al. 2016). This might be of relevance in the context of ATLL immunotherapy, considering recent reports describing rapid disease progression in ATLL patients receiving PD-1 inhibitor therapy (Ratner et al. 2018).

NK/T-CELL LYMPHOMA

Natural killer/T-cell lymphoma (NKTCL) is an aggressive type of non-Hodgkin lymphoma characterized by a clonal proliferation of NK T cells. It most commonly affects the upper respiratory and digestive tracts although sometimes it involves extranasal tissues such as the skin, gastrointestinal tract, soft tissues, and testis (de Mel et al. 2019b,a). NKTCL is most prevalent in Asians and the Native American populations of Mexico, Central America, and South America (Haverkos et al. 2016).

NKTCL is commonly associated with EBV viral infection (Suzuki 2014; Chaudhary et al. 2015; Komabayashi et al. 2017), although little is known about the genomic and transcriptomic profiles of EBV in NKTCL. Sequence variations of LMP1, a key latent protein with abilities to promote cell proliferation and inhibit cell apoptosis in NKTCL, have been demonstrated (Ito et al. 2014). It has been suggested that the role of LMP1 in NK and T cells is to increase the sensitivity of the infected cell to the growth-promoting effects of IL-2.

Activation of the JAK/STAT pathway plays a prominent role in the pathogenesis of NKTCL, and up-regulation of IL2, IL10, and IFNGR1 are frequent events in NKTCL. JAK3-activating mutations promoting cytokine-independent STAT activation and enhancing NKTCL cell proliferation have also been described (Koo et al. 2012; Sim et al. 2017). Increased JAK/STAT signaling induces up-regulation of MYC and its interacting proteins (Huang et al. 2018); as well as phosphorylation of EZH2, a component of the Polycomb repressive complex 2 (PRC2), resulting in the up-regulation of genes involved in cell cycle regulation, DNA replication, invasiveness, and stemness (Yan et al. 2016).

In addition to the multiple modalities of JAK/STAT dysregulation in NKTCL, mutations in genes related to epigenetic modification are also common in NKTCL, including genes associated with histone methylation (KMT2D), histone acetylation (EP300), histone deubiquitination (ASXL3), and chromatin remodeling (ARID1A) (Choi et al. 2016).

ENTEROPATHY-ASSOCIATED T-CELL LYMPHOMAS

Enteropathy-associated T-cell lymphomas derive from the transformation of small intestinal intraepithelial lymphocytes (Chander et al. 2018). The most recent PTCL classification distinguishes between enteropathy-associated lymphoma closely linked to celiac disease (EATL, previously named EATL, type I) and monomorphic epitheliotropic intestinal T-cell lymphoma (MEITL, previously named EATL type II) (Swerdlow et al. 2016).

EATL is a rare entity usually associated to elderly individuals of Northern European descent (Delabie et al. 2011) and is a known complication of celiac disease, an autoimmune disorder characterized by gluten intolerance (Chander et al. 2018). Most EATLs were believed to arise from the neoplastic transformation of thymus-derived cytotoxic intraepithelial TCRαβ⁺ T cells; however, recent reports have also suggested an origin of a subset of EATLs from intraepithelial TCRγδ⁺ T cells (Chander et al. 2018). Recurrent activating mutations in members of the JAK/STAT pathway, including JAK1 and STAT3, have been recently described in EATL (Roberti et al. 2016; Moffitt et al. 2017), which suggests deregulation of cytokine signaling to be an early event in lymphomagenesis.

MEITL is less common than EATL, shows no association with celiac disease, and occurs worldwide (Roberti et al. 2016). In the majority of the cases, MEITL neoplastic cells are derived from γδ T cells, although there are also exceptions (Chan et al. 2011; Swerdlow et al. 2016). Recurrent alterations in SETD2 are the hallmark of MEITL, being detected in >90% of patients (Roberti et al. 2016). Additional mutations in other epigenetic regulators including CREBBP, ARID1, EP300, and EZH2 were also found at lower frequencies. Also recurrent were mutations affecting the JAK/STAT signaling pathway, including STAT5 and JAK3 (Kucuk et al. 2015; Nairismägi et al. 2016), the MAPK pathway (including BRAF, KRAS, and NRAS) and TP53 (Roberti et al. 2016).

HEPATOSPLENIC T-CELL LYMPHOMA

Hepatosplenic T-cell lymphoma (HSTCL) derives from nonactivated cytotoxic T cells. This rare entity is found in younger male patients, sometimes associated with chronic immunosuppression (Shi and Wang 2015). HSTCL primarily affects the liver, spleen, and frequently the bone marrow and is characterized by the expansion of malignant medium-sized T cells expressing cytotoxic markers (including T1A1⁺ and granzyme M⁺) and killer immunoglobulin receptors (Kirs) (Yabe et al. 2018). The most common HSCTL type is derived from γ/δ T cells with a small percentage derived from α/β T cells (Shi and Wang 2015; Swerdlow et al. 2016), although both subtypes are otherwise clinically and genetically similar. The most frequent mutations identified in HSCTL affect chromatin-modifier genes such as the SETD2 tumor suppressor, INO80, and ARID1B, in addition to recurrent activating mutations affecting elements of the JAK/STAT pathway, predominantly STAT5B and less often STAT3, and PIK3CD (Nicolae et al. 2014; McKinney et al. 2017)

CUTANEOUS T-CELL LYMPHOMAS

Cutaneous T-cell lymphomas (CTCLs) are characterized by the presence of skin-homing CD4⁺ malignant T cells. Although current models place the origin of mature cutaneous T-cell lymphomas from memory T cells in the periphery, there have also been reports of quiescent cancer precursor cells residing in bone marrow, which secondarily migrate to the skin producing clinically apparent cutaneous lymphoma lesions (Gniadecki et al. 2003; Hamrouni et al. 2019). Mycosis fungoides (MF) and Sézary syndrome (SS) are two major subtypes of CTCL, constituting 75% of the cases diagnosed as CTCL (Devata and Wilcox 2016). MF cases with limited skin involvement have a generally good prognosis; however, advanced skin involvement, presence of characteristic Sézary cells in peripheral blood and/or lymph node, and organ infiltration are associated with dismal outcomes.

Multiple genomic studies to define the mutational landscape of CTCL have demonstrated the existence of a marked genomic heterogeneity (Vaqué et al. 2014; Choi et al. 2015; da Silva Almeida et al. 2015; Ungewickell et al. 2015; Wang et al. 2015b). Common driver genetic lesions include mutations and copy number alterations affecting cell cycle regulators (CDKN1B), chromatin-modifying genes (ARID1A, ARID5B, SMARCC1, DNMT3A, and TET2), DNA damage response genes, and immune evasion genes, as well as genes involved in T-cell signaling, migration, and differentiation (including a recurrent mutation in RHOA N117I of unknown significance). Frequent mutations and chromosomal deletions affecting TP53 have been identified in >50% of the patients with CTCL (Chang et al. 2018). Genomic analysis of CTCL found recurrent gain-of-function mutations and amplifications affecting the JAK/STAT pathway, including JAK1, JAK3, STAT3, and STAT5B, resulting in the hyperactivation of this signaling pathway. In addition, the NF-κB pathway is frequently targeted by somatic mutations in CTCL, including PLCG1, CARD11, TNFRSF1B, and recently RLTPR, a scaffolding protein that strongly up-regulates the NF-κB signaling (Park et al. 2017). The convergence of diverse mutations on this pathway, generally mutually exclusive, highlights the role of NF-κB signaling in CTCL lymphomagenesis and supports the potential therapeutic role for NF-κB inhibitors.

NOVEL THERAPEUTIC APPROACHES FOR THE TREATMENT OF MATURE T-CELL LYMPHOMAS

The current standard of care for mature T-cell lymphomas is based on traditional chemotherapy with an induction therapy using a combination of cyclophosphamide, doxorubicin, vincristine, and prednisone (CHOP) followed by stem cell transplant after first remission. Unfortunately, this therapeutic approach only achieves survival rates of 40%–50% and prognosis remains dismal for most of the PTCL subtypes (Lue et al. 2017). CHOP-based approaches are specifically ineffective in ATLL patients, who present low response rates with frequent relapses and short survival. In this specific group, antiretroviral treatment, combining zidovudine (AZT) and interferon alpha (IFN-α), appears to be an option for first-line therapy (Bazarbachi et al. 2010).

The increased understanding of the biology of PTCL achieved during the past few years has supported an exponential rise in the development of novel targeted therapies in PTCL that are generally better tolerated and in multiple instances have achieved very significant responses (Table 2; Fig. 3). Most of these agents have been tested as second-line salvage therapies in relapsed/refractory PTCL and their efficacy has been mostly analyzed as single agents although a variety of clinical trials focused on novel drug combinations are also ongoing (Ma et al. 2018).

Table 2.

Targeted therapies for PTCL: Small molecules, antibodies, and immune checkpoint inhibitors targeting molecular pathways

Small molecules	Inhibitors
Vorinostat	Histone deacetylases (HDACs)
Romidepsin
Belinostat
Chidamide
Pralatrexate	Dihydrofolate reductase
Decitabine (5-aza-2′-deoxycytidine)	DNA methyltransferases
5-aza-cytidine
Duvelisib	PI3 kinases
Copanlisib
Tenalisib
Bortezomib	Blocks IκB degradation and NF-κB translocation by targeting 26S proteasome
Carfilzomib	20S proteasome
Ixazomib	Proteasome subunit beta type-5 (PSMB5)
Everolimus	Targets FK506 binding protein-12 (FKBP-12) inhibiting mTORC1
Crizotinib	Tyrosine kinase receptors: anaplastic lymphoma kinase (ALK), hepatocyte growth factor receptor (HGFR)
Antibodies
Mogamulizumab	CCR4 chemokine receptor
Brentuximab vedotin	CD30 (TNFRSF8)
Hu5F9-G4	CD47
TTI-621
Immune checkpoint inhibitors
Nivolumab	PD-L1
Durvalumab
Pembrolizumab	PD-1

Figure 3.

Relevant molecular pathways in PTCL amenable to therapeutic intervention using targeted therapies.

In general, PTCL exhibits a marked epigenetic dysregulation that supports the use of epigenetic modifiers including inhibitors of histone deacetylases (HDACs) and hypomethylating agents for the treatment of this disease.

Two widely used hypomethylating agents are decitabine (5-aza-2′-deoxycytidine) and 5-aza-cytidine, cytidine antimetabolite analogs that incorporate into DNA and inhibit DNA methyltransferases, resulting in hypomethylation of DNA. Initial case reports demonstrated a positive outcome for AITL patients treated with 5-aza-cytidine (Cheminant et al. 2015; Saillard et al. 2016) and a clinical trial with relapsed PTCL patients resulted in 53% overall response rate (ORR) and 75% ORR when only AITL patients were analyzed (Saillard et al. 2016).

HDAC inhibitors constitute a class of small-molecule inhibitors that activate gene expression by inhibiting HDAC activity. Although initially being considered as “epigenetic modifiers” with a predominant role in chromatin remodeling via histone acetylation, HDACis have been shown to exert wider more pleiotropic effects in tumor cells associated with acetylation of nonhistone proteins and regulation of DNA repair (O'Connor et al. 2014). Interestingly, HDACis have been particularly effective for the treatment of PTCL, and several have been approved by the U.S. Federal Drug Administration (FDA) including vorinostat (Duvic et al. 2007; Olsen et al. 2007), romidepsin (Piekarz et al. 2011; Coiffier et al. 2012, 2014; Maruyama et al. 2017b), and belinostat (Foss et al. 2015) with an additional HDACi, chidamide, being approved in China (Shi et al. 2017). Challenges remain to better define the patients that will respond to HDACi, identify the mechanisms of resistance, and leverage the effectiveness of epigenetic modifiers alone and in combination. Indeed, a recent clinical trial combining oral 5-azacytidine and romidepsin exhibited marked activity in patients with PTCL in a phase I clinical trial, with overall and complete response rates of 73% and 55%, respectively, in patients with T-cell lymphoma (O'Connor et al. 2019).

Pralatrexate is an antifolate that works by inhibiting dihydrofolate reductase, leading to depletion of tetrahydrofolate with the subsequent inhibition of pyrimidines synthesis and tumor growth. Palatrexate was FDA-approved in 2009 after two independent clinical trials reported a 29%–45% response rate (O'Connor et al. 2011; Maruyama et al. 2017a).

The PI3 K/AKT signaling pathway is frequently activated in PTCL and pharmacologic inhibition of the PI3 K/AKT pathway has been extensively studied in preclinical models and has shown efficacy in T-cell malignancies (Cortés et al. 2018). Currently, several PI3 K inhibitors, including duvelisib (Horwitz et al. 2018), copanlisib (Dreyling et al. 2017), and tenalisib (Ito et al. 2019) with differential affinity for the different PI3 K subunits, are being tested in T-cell malignancies, where they are thought to mediate both cell autonomous effects on tumor cells and immune-mediated effects because of their activity on the tumor microenvironment (Horwitz et al. 2018).

Bortezomib is a proteasome inhibitor that blocks degradation of IκB and the proteolytic processing of p105, thereby inhibiting NF-κB nuclear translocation (Wang et al. 2014). In mature lymphomas, the use of bortezomib led to an ORR of 67% in 15 patients with CTCL when used as single agent. In a phase II study in combination with CHOP in newly diagnosed PTCL cases, the drug combination achieved an 76% ORR (Kim et al. 2012). Newer proteasome inhibitors, including carfilzomib and ixazomib, are currently being tested in clinical trials for the treatment of relapsed PTCL in the context of drug combinations.

Everolimus, a derivative of rapamycin that inhibits the mTOR pathways, achieved an ORR of 44% as a single agent in relapsed or refractory lymphomas (Witzig et al. 2015), and a more recent phase II study with a combination of everolimus plus CHOP demonstrated an increased efficacy over CHOP alone (Kim et al. 2016).

Considering the prevalence of mutations affecting the JAK/STAT pathway in multiple subtypes of PTCL, treatment with small-molecule inhibitors targeting this pathway could have potential relevance for the treatment of PTCL. Recently, the JAK3 inhibitor tofacitinib has achieved hematologic responses in a phase I trial in TGL, whereas the dual SYK/JAK inhibitor cerdulatinib has also been studied in a phase II trial on relapsed PTCL achieving a 35% overall response (Ng and Jacobsen 2019). Similarly, dasatinib, a multikinase inhibitor that targets members of the Src family of intracellular tyrosine kinases, has shown promising results for the treatment of AITL in a phase I trial, probably because of its inhibitory effect on the deregulated TCR signaling characteristic of this tumors (Nguyen et al. 2020)

Finally, crizotinib, a small-molecule inhibitor of the ALK tyrosine kinase, has been markedly successful in patients with ALK-positive ALCL, for whom treatment with crizotinib induced a 91% response rate (Gambacorti Passerini et al. 2014). Crizotinib is currently being evaluated in additional clinical trials directed both at pediatric and adult populations. Additional ALK inhibitors are being considered as next-line therapy, anticipating the potential development of resistances to crizotinib (Sharma et al. 2018).

Development of immunotherapeutic approaches has become a major interest in the field of T-cell malignancies. Promising results have been obtained with the use of monoclonal antibodies, immune checkpoint inhibitors and CAR-T-cell therapies. Mogamulizumab, a humanized monoclonal antibody directed against the CCR4 chemokine receptor that is highly expressed in ATLL and CTCL subtypes (Ishida et al. 2003; Hristov et al. 2011), has demonstrated efficacy in patients with CCR4-positive ATLL, PTCL, and CTCL (Yamamoto et al. 2010; Ishida et al. 2012; Duvic et al. 2015; Kim et al. 2018). The results from the phase III clinical trial on CTCL patients demonstrated an overall response rate of 28% (Kim et al. 2018), leading to the approval of mogamulizumab by the FDA for the treatment of patients with ATLL and CTCL. Brentuximab vedotin is an antibody-drug conjugate that contains the microtubule-disrupting agent mono-methylauristatin E (MMAE) linked to an anti-CD30 monoclonal antibody. CD30 is highly expressed on activated T cells as well as ALCL and at variable levels in other PTCL subtypes (Onaindia et al. 2016). Initial studies with anti-CD30 unconjugated antibodies have failed to show significant clinical responses (Ansell et al. 2007; Duvic et al. 2009); however, results from a phase II clinical trial using the MMAE-conjugated form showed impressive responses in ALCL (Pro et al. 2012) with a ORR of 86% and achievement of long-term remission. Significant improvement was also observed in CD30⁺ CTCL (Prince et al. 2017). However, clinical responses in PTCL-NOS or AITL were modest (Horwitz et al. 2014; Lamarque et al. 2016) and likely due to the variability in CD30 expression (Bossard et al. 2014). A different approach to target CD30 has been demonstrated with the development of AFM13, a bispecific chimeric antibody containing an anti CD16A domain to recruit NK T cells linked to an anti CD30 domain, which achieved partial responses in the initial studies in refractory lymphoma (Rothe et al. 2015) and is currently being tested on patients with CD30⁺ lymphomas with cutaneous involvement.

Finally, different antibody molecules targeting CD47, a ubiquitous membrane protein that binds to SIRPα inhibiting macrophage phagocytosis and allowing escape from immune surveillance, are currently being studied in clinical trials. Inhibition of the CD47-SIRPα interaction results in activation of innate immunity and promotion of cancer cell destruction by macrophages. Hu5F9-G4, a humanized anti-CD47 blocking antibody (Liu et al. 2018), and TTI-621, a fusion protein containing the CD47 binding domain of human SIRPα and linked to the Fc region of IgG1 with enhanced antitumor activity, are two examples of CD47 targeting molecules under investigation. In an ongoing phase I study in CTCL patients, intratumoral injections of TTI-621 led to decreased tumor size (Folkes et al. 2018).

The past few years have witnessed an increased effort toward the implementation of immunotherapy approaches for the treatment of lymphoid malignancies. Blocking antibodies targeting CTLA-4, PD-L1, and PD-1 immune checkpoints have resulted in the development of highly effective antitumor responses in solid tumors; however, their indication for the treatment of mature T-cell malignancies remains controversial. Although multiple trials are testing the effect of checkpoint inhibitors in PTCL, including nivolumab (Lesokhin et al. 2016), pembrolizumab (Kwong et al. 2017a,b), and durvalumab (Heyman and Yang 2018), recent reports have also described a role for PD-1 as a tumor suppressor in T-cell lymphomagenesis and demonstrated that Pd-1 deletion in mouse models leads to highly aggressive lymphoma (Wartewig et al. 2017). Similarly, expansion of malignant clones and oncogenic activation of the TCR pathways have described associated to checkpoint inhibitor therapy in ATLL (Ratner et al. 2018). These results, together with the variable efficacy of checkpoint inhibitors in different PTCL subtypes, suggest that immunotherapy targeting of PD-1 needs to be carefully evaluated for disease stage, composition of the tumor microenvironment, PD-1 genetic deletions, and TCR oncogenic alterations prior to treatment.

Modulating the immune system to eliminate tumor cells using CAR-T-cell-based cellular therapies has proven successful for the treatment of lymphoid malignancies, especially for patients with refractory or relapsed B-cell malignancies. However, treating T-cell malignancies has proven to be challenging endeavor because of the shared expression of many targetable antigens between normal and malignant T cells. This shared antigenicity can cause fratricide in CAR-transduced T cells, inhibiting their proliferation and viability, and in the clinic may also result in eradication of normal peripheral T cells leading to severe immunodeficiency and other serious side effects. Despite these caveats, significant progress has been made on the identification of targetable antigens and development of cell-based therapies for T-cell leukemia and lymphoma. CD30 represents an attractive and validated target for antibody-based therapies and its expression is generally restricted to lymphoma cells. CAR-T cells targeting CD30 have shown potent antilymphoma activity in preclinical studies in various tumor models (Hombach et al. 1999). In addition, recent success in targeting a constant region of the TCR beta chain (TRBC1) could prove relevant for the treatment of PTCL (Maciocia et al. 2017).

CLOSING REMARKS

In the past few years, multiple studies have led to an increased understanding of the pathogenesis and molecular classification of PTCL. High-throughput studies have achieved a deeper understanding of the underlying disease biology and led to improved diagnostic accuracy based on the molecular classification of PTCL. The identification of novel therapeutic vulnerabilities has induced a significant shift in the clinical management of these diseases because of the introduction of targeted agents. The success of epigenetic drugs in PTCL, specifically of HDACi and hypomethylating agents, supports leveraging their use alone and in combination for the treatment of this disease. Many challenges and opportunities for improvement remain, including identification of biomarkers of response to targeted therapies, characterization of mechanisms of resistance, and development of successful combination therapies for the treatment of PTCL.

In addition, continuous progress toward the understanding of the mechanisms that drive T-cell oncogenesis, generation of animal models, and patient-derived xenografts for the experimental therapeutics and implementation of immunotherapy approaches will be instrumental to improve survival for mature T-cell malignancies.