Molecular pathogenesis of cutaneous lymphoma—Future directions

The pathogenesis of cutaneous T‐cell lymphomas is not clear. In recent years, the genetic changes in CTCL were explored. The detected mutations showed a great deal of heterogeneity between individual patients. The studies documented various copy number variations (CNV) and single nucleotide variations (SNV) in multiple genes involved in multiple signalling pathways. Recurrently mutated signalling pathways include JAK‐STAT, MAPK, T‐cell receptor, TNF receptor and NFκB signalling. In the period between 2018 and today, additional studies towards the genetic changes in CTCL were carried out. Genetic changes in gamma delta T‐cell lymphoma are also shown in genes of the JAK‐STAT, MAPK, MYC and chromatin signalling pathways. These studies might indicate a shift away from targeted sequencing approaches towards whole‐genome sequencing. This approach demands additional resources in terms of funding but has the advantage of finding mutations in non‐coding regions. These mutations were neglected for a long time, but as shown in contemporary research these regions harbour highly recurrent mutations affecting gene expression and regulation.

and involvement of the blood, lymph nodes and visceral organs. This progression is generally associated with poor prognosis. [4,5] The pathogenesis of cutaneous T-cell lymphomas is not clear.
Overall, the disease seems to be based on a poly aetiological process. This is supported by the following findings: antigenic stimulation in predisposed patients has been discussed and investigated for a long time [6] ; however, it has not yet been sustainably confirmed.
Brazzelli's research group found indications that specific HLA alleles may predispose to the development of MF and that certain alleles may be prognostically relevant. In particular, HLA-DQB1*05 seems to be associated with a worse prognosis. [7,8] Further work has focused on molecular metabolic pathways.

| COMPLE X MUTATIONAL L ANDSC APE IN C TCL
In recent years, the genetic changes in CTCL were explored. [9][10][11][12][13][14][15][16][17][18] The majority of the work focused on the genetic changes in SS patients (85% of the sequenced patients), while MF patients represented only a minority in the cohort (around 11%). The reason for this stark contrast between disease incidence and representation in the sequencing cohorts might be due to technical constraints. The current data situation of genetic changes in CTCL shows a great deal of heterogeneity between individual patients (Table 1). [13] Various copy number variations (CNV) and single nucleotide variations (SNV) in multiple genes involved in multiple signalling pathways were documented. Recurrently mutated signalling pathways include JAK-STAT, MAPK, T-cell receptor, TNF receptor and NFκB signalling. Further mutated cellular processes are chromatin remodelling and modification (eg DNMT3A) as well as cell cycle checkpoints (eg CDKN2A) and genomic integrity processes (eg TP53). Interesting observations from the studies were that the recurrence of individual mutation or even mutated genes was quite low. For example, the gene most commonly affected by SNVs in SS is TP53 with an incidence of "only" 12%. [13] The other interesting fact is the high abundance of CNVs (10x more abundant than SNVs, 9). These mutations are also partly more recurrent than SNVs. For example, the loss of regions on chromosome 17 containing TP53 is present in around 40% of the cohort from. [9] Supplementing the genetic changes in CTCL a plethora of changes of the transcriptome or chromatin was detected. This includes the overexpression of TOX (thymocyte selection associated HMG-box) in SS and MF, which could be used to discern these diseases from benign inflammatory diseases . On a genome scale, a clear distinction between malignant and normal CD4 + cells in terms of chromatin accessibility and patterns of transcriptions factor activation were shown. [19] These studies were summarized by R. Stadler and R. Stranzenbach to give a precise view of the current knowledge in the understanding of the molecular pathogenesis of CTCL. [20] These genetic changes lead to T cells that show apoptosis deficits and a consequent continuous expansion. During this process, the malignant cells show increased differentiation into T helper type 2 cells with corresponding cytokine production of interleukin 4 and 13 and resistance to normal cell control mechanisms. [21] All these experimental and scientific efforts are intended to discover novel biomarkers and meaningful therapeutic targets to be addressed in the treatment of cutaneous T-cell lymphoma and, of course, to elucidate the complexity of the pathophysiology of CTCL. This fish assay detected GCNAs in 92% of the patients that met B2 blood involvement criteria by the International Society of Cutaneous Lymphoma (ISCL). [22] One therapeutic approach is the inhibition of miR-155. This micro-RNA is upregulated in tumor MF compared to early stage MF and might be a trigger for MF progression into advanced stages. An oligonucleotide inhibitor of miR-155, cobomarsen, coordinately regulates multiple survival pathways to reduce cellular proliferation is already in a phase 1 trial in cutaneous T-cell Lymphoma. [23] Another therapeutic approach is proposed by Nicolay et al Here, the hypothesis is that combined inhibition of the antiapoptotic mediator B-cell lymphoma 2 (BcL2) and the NFκB pathway can overcome cell death resistance. As the hyper-activated NFκB pathway

TA B L E 1 Park et al identified 55 genes in lymphomagenesis including a gain of function mutation in RLTPR that potentiates T-cell receptor signalling
Genomic analysis of 220 CTCL's [13] Chromatin remodelling BCOR, KDM6A, SMARCB1, TRRAP is a common survival factor in CTCL, this approach might be applicable for CTCL therapy. The used drug is dimethyl fumarate and inhibitor of the NFκB activation. [24,25] Further interest was directed on T-cell receptor sequencing.
This genetic feature poses a unique biomarker in T-cell lymphoma (or B-cell lymphoma in case of B-cell receptor sequencing). T-cell receptor clonality assessment is used for a long time in CTCL diagnosis [26] ; in recent years, the NGS-based clonality assessment [27,28] is increasingly replacing the gel based methods. To improve the results from T-cell receptor sequencing (eg minimization of amplification bias), multiple protocols were developed. In 2019, the euroclonality group published their new gold standard for NGS-based TCR clonality analyses, including spike-in DNA and controls to deal with bias and polyclonal background. [29,30] In addition to using T-cell receptor "only" for diagnostic purposes, the data might also be used as a prognostic marker. Adele de Masson et al measured the tumor clone frequency in lesional skin by high-throughput sequencing of the TCRB gene and identified the clone frequency as an independent prognostic factor in early stage patients of cutaneous T-cell lymphoma. [31]

| HOW HA S THE FIELD PUR SUED THIS LINE OF RE S E ARCH IN THE ME ANTIME?
Based on these results, there is a critically unmet need to identify putative molecular targets. The genetics for this special lymphoproliferative disease with these different manifestations remain obscure. Therefore, the development of a powerful targeted therapy is very difficult.
In the period between 2018 and today, additional studies towards the genetic changes in CTCL were carried out. This includes the analysis of additional seven Sézary syndrome cases by whole-exome sequencing. Over all their cases, they found 551 non-synonymous SNVs split across 525 genes. Several of these mutated genes are part of the pathways already known to be affected in CTCL like JAK-STAT and PI3K/AKT signalling. Additional mutations were found in peroxisome proliferator-activated receptors and fibroblast growth factor receptors adding up to the great complexity and heterogeneity of SS (Table 2B). [32] In Mycosis fungoides, nine cases were investigated by whole-genome sequencing. [33] Here, the most notable result was the recurrent deletion of HNRNPK and SOCS1, and both inhibitor of the JAK-STAT signalling pathway were detected. This result showed especially for MF the great importance of dysregulated JAK-STAT signalling on the MF pathogenesis. (Table 2A). [33] Furthermore, the genetic landscape of gamma delta T-cell lymphoma was investigated by performing whole-genome, RNA and T-cell receptor sequencing on 29 cases. This study showed that gamma delta T-cell lymphoma arise from Vδ1 or Vδ2 cells dependent on the tissue compartment from which the lymphomas are derived.
Genetic changes in gamma delta T-cell lymphoma are often in genes of the JAK-STAT, MAPK, MYC and chromatin signalling pathways. [34] Especially, the studies from Torres [33] and Daniels [34]  Type 4 monoclonal antibody) and IPH410lon (humanized that targets the immune receptor KIR3DL2/CD158k). [36][37][38][39][40][41] Targeting CD47 is expressed on all normal cells, targets SIRPα on the surface of myeloid cells and may be a new cancer therapeutic strategy also in cutaneous T-cell lymphoma.
Inhibition of CD47-SIRPα via anti-CD 47 antibodies activates the innate immunity, promoting cancer cell destruction by macrophages. [42] The phosphatidylinositol 3-kinase/AKT/mammalian target of rapamycin (PI3K/AKT/mTOR) signalling pathway regulates central aspects of cancer biology, such as metabolism, cellular growth and survival. In particular, a gain of function or hyperactivity of phos- stimulation. [43] An upregulation of this pathway has been reported in CTCL. [20] Duvelisib is a promising new oral inhibitor of phosphatidylinositol 3-kinase (PI3K)-δ/γ isoforms currently in clinical development.
PI3K-δ/γ inhibition may directly inhibit malignant T-cell growth, making duvelisib a promising candidate for patients with CTCL lymphoma. [44] Preliminary clinical data support a potential role of this drug in the treatment of CTCL. In a phase I open label study of duvelisib in F I G U R E 1 In this figure, the known mutations in different signalling pathways are illustrated with the red dots. The possible targeting therapeutics are mentioned AB against surface molecules: patients with T-cell lymphoma, 19 CTCL patients were included with a response rate of 31.6%, [45] Figure 1.

| WHI CH MA JOR OPEN QUE S TI ON S REMAIN TODAY ?
One of the major basic questions is: How can an apparently innocent inflammatory lesion develop into a major life threatening disease as represented by advanced cutaneous T-cell lymphoma?
The complex interaction with initial mutagenic insult resulting in GCNAs but also SNVs, in selected genes involved with T-cell activation, epigenetic regulation and cell cycle regulation promote con- It is further mandatory to understand the apparent T-cell activation pathway. It is known that genes, for example EB1, are driving T H 2 differentiation and that this gene has been implicated in CTCL pathogenesis and suggesting a mechanism to escape from TGs beta-mediated inhibition of proliferation. However, it has been shown recently that by targeting the T H 2 signalling pathway via dupilumab, a fully human monoclonal antibody that binds the interleukin 4 receptor alpha subunit and inhibits signalling of interleukin-4 and interleukin-13 seems to be inappropriate for the use in cutaneous T-cell lymphoma patients. This is an excellent example for the complexity in understanding the immunological behaviour of CTCL. [21,46] One of the major questions is to elucidate recurrent mutations in mycosis fungoides (MF), concentrating on the most reported mutations in all published data or to elucidate recurrent patterns of mutated pathway in (MF).
Transcriptional programmes guiding lymphocyte differentiation depend on precise expression and timing of transcription factors (TFs). For example, BACH2 is a TF essential for T and B lymphocyte maturation and differentiation. Mutations in the Bach2 locus are associated with multiple autoimmune diseases. [47] There are no published data for cutaneous T-cell lymphoma.

| WHI CH CON CRE TE E XPERIMENTAL APPROACHE S S HOULD B E ADVO C ATED TO PROVIDE THE S E MISS ING AN SWER S IN THE NE AR FUTURE?
Further sequencing studies of MF cases, where the study design is targeted towards answering the major open questions, like "Are there highly recurrent mutations in MF" or "Are there highly recurrent altered/mutated pathways in MF." To get a reliable data set for answering or ruling out one of the questions mentioned above, the study design and sample addition must be carefully evaluated.
On the side of the samples to be included in such a study, technical parameters as tumor cell fraction of the sample but also biological parameters as potential genetic alterations by preceding therapies must be considered. Sample collection and storage must be adapted to the respective subsequent application.
The methodical approach of this study should be a combined whole-genome/transcriptome sequencing approach with subsequent validation of, for example hyper-activated pathways by immunohistochemistry. Using this approach on a sufficient large set of carefully selected samples, it might be possible to draw connections between the genetic changes and the phenotypic changes in MF cells.

| OUR PER SONAL VIS I ON ON HOW THIS LINE OF RE S E ARCH S HOULD B E DE VELOPED OVER THE NE X T DEC ADE AND WHY COULD THIS SUBS TANTIALLY CHANG E E XIS TING PAR ADIG MS IN CLINI C AL OR RE S E ARCH PR AC TICE?
The above-mentioned approach of whole-genome/transcriptome sequencing of a large cohort of MF cases is only feasible in the near future as a collaboration project. Primarily collaboration must define important sample parameter criteria and provide sufficient methods for measuring/deriving these parameters. Afterwards, details for the experimental procedure, data generation, data storage and data analysis must be devised. After establishing, these important foundations on data generation analysis can be carried out. We see multiple advantages of such a collaboration. In the case of finding high recurrent genetic events, this study would further enhance our understanding of the molecular processes involved in MF and it might bring us one step closer to define important biomarkers.
In case that the data from the proposed collaboration of different lymphoma centre would not detect constant patterns of specific molecular changes in different patients, this study might also be of advantage in two aspects: (a) to further proof that MF is a complex genetically heterogeneous disease and (b) creates a necessary infrastructure to continue to analyse large genomic regions of MF patients to obtain single case genetic information for enrolment into precise targeted therapy studies. There is a high need to make high-throughput sequencing (HTS) available for personalized diagnostic in clinic daylife.

ACK N OWLED G EM ENTS
Open access funding enabled and organized by Projekt DEAL.

CO N FLI C T O F I NTE R E S T
I herewith declare that I have no conflict of interest within the manuscript.

AUTH O R CO NTR I B UTI O N
RS (first author) wrote the paper and analysed the data. CH contributed to writing the paper analysed the data and the figure. CC Analysed the data and contributed the necessary tools. RS contributed to writing the paper analysed the data and the figure. All authors have read and approved the final manuscript.

DATA AVA I L A B I L I T Y S TAT E M E N T
Data sharing not applicable to this article as no datasets were generated or analysed during the current study.