Expanding and expounding the genomic map of CTCL

Over the past 3 years, a veritable explosion of genomic data has begun to shed much needed light on the mutational landscape of cutaneous T-cell lymphoma (CTCL). In this issue of Blood, Park et al expand the boundaries and improve the resolution of the known mutational map of CTCL.¹ 

The term CTCL describes a heterogeneous family of incurable primary extranodal lymphomas of mature T cells that uniformly present in the skin, with a variable risk of systemic dissemination. The most common type of CTCL (∼60%) is mycosis fungoides (MF), a low-grade lymphoma of skin-homing CD4⁺ memory T cells. In approximately two-thirds of the patients, MF presents with a slow but progressive accumulation of atypical lymphocytes in the upper layers of the skin (epidermotropism), leading to the development of discrete cutaneous patches and plaques, usually affecting sunlight-protected areas. As the disease advances, the neoplastic T cells acquire an invasive pattern of vertical growth leading to the development of disfiguring exophytic or deep-seated tumor lesions and become able to migrate to lymph nodes, peripheral blood, and visceral organs. A very small subset of patients with CTCL (<5%) present with Sezary syndrome (SS), defined by the triad of leukemic involvement with atypical T cells, erythroderma, and lymphadenopathy. Whether MF and SS are clinically and biologically distinct malignancies remains controversial. Despite the introduction of several new systemic agents over the past 10 years, the median survival of patients with SS and advanced-stage MF remains <5 years, and MF/SS patients who develop nodal or visceral disease (stage IV) have very poor survival (<2 years). Thus, there is an urgent unmet need to understand the molecular pathways driving and sustaining the neoplastic phenotype in MF/SS, with the goal of identifying cancer dependencies that predict vulnerability to specific therapies.

Prior to the development and the application of next-generation sequencing (NGS) of MF/SS tumor cells, array comparative genomic hybridization (aCGH) studies had given a glimpse of the complex global genetic architecture of advanced-stage MF/SS, revealing multiple hot spots of genetic gains and losses at chromosomes 1, 8, 9, 10, and 17, in addition to many other recurrent aberrations that reflected a high degree of genomic instability. These studies also identified a number of candidate genes and signaling pathways, whose frequent amplification and constitutive activation, or loss, suggested a pathogenic role in MF/SS. Frequently amplified genes included cMYC, STAT3, STAT5A/B, TWIST1; frequently lost ones included TP53, PTEN, MTAP, and CDKN1A/B. In addition, recurrent genomic gains at loci for genes encoding cytokines and their receptors (interleukin-2 [IL-2], IL-2R, and IL-7) and loss of negative regulators of cytokine transcription (TCF8/ZEB1, DUSP5) highlighted the importance of aberrant cytokine signaling in the pathogenesis of MF/SS.

The recent reporting of whole-genome and whole-exome NGS data from well-characterized MF/SS samples, coupled with transcriptome analysis via RNA sequencing, has significantly advanced this initial bird view of the complex mutational landscape of CTCL.^2-8  These studies have now confirmed, expanded, and functionally validated many of the genetic aberrations detected by aCGH in MF/SS, a broad and diverse spectrum that include genes associated with T-cell receptor (TCR) signaling, activation of NF-κB, JAK/STAT signaling, chromatin remodeling, and DNA damage response. However, despite a significant degree of overlap, there was great variability in the identity and frequency of alterations at putative driver genes across studies. For example, some studies identified loss-of-function mutations or deletions of TCF8/ZEB1 in 56% to 65% of cases,^2,5,6  whereas others did not.⁴  By compiling the published NGS data from 220 MF/SS patients, and subjecting them to a uniform statistical analysis, Park et al were able to achieve a sample size adequate to identify 55 putative driver genes, including novel mutations at 17 genes, 5 of which had never been involved in cancer. Importantly, the functional analysis of 1 of these 5 genes (RGD, leucine-rich repeat, tropomodulin and proline-rich containing protein [RLTPR]) further highlights the involvement of the CD80/CD86-CD28 axis in MF/SS, already evident from previous studies, and illustrates the role of TCR signaling as a codriver of MF/SS development. Analogous to other T-cell malignancy models, where JAK/STAT mutations initiate tumor cell proliferation only in the presence of cytokine signaling,⁹ RLTPR mutations alone were not sufficient to induce transcriptional programs associated with T-cell activation but potentiated >30-fold the effects of TCR signaling. Thus, the functional effect of mutations affecting downstream components of TCR or cytokine signaling needs to be tested and interpreted in the context of the appropriate microenvironment.

This leads us to the observation that mutational landscape studies must, in the end, find validation in preclinical models, where the single or combined effect of specific mutations in the development and progression of T-cell lymphoma can be tested and where rationally designed therapeutic interventions can be explored. In this regard, in vivo data showing the importance of γ cytokine signaling in CTCL come from the stage-dependent overexpression of IL-15 in MF/SS patient samples and from the fact that IL-15 transgenic mice spontaneously develop CTCL.¹⁰  Furthermore, TCF8/ZEB1, a frequent target of loss-of-function mutations or deletions in MF/SS,^2,5,6  is epigenetically silenced in patient samples that overexpress IL-15,¹⁰  thus providing a mechanistic link between specific genetic aberrations and signaling pathways known to be important in CTCL pathogenesis.

The study by Park et al delivers a well-curated genetic map of CTCL. We now have to use that map to plot our journey toward a cure.