The histological classification of TETs is made according to the World Health Organization (WHO) criteria, which rely on the morphology of cancer cells, the degree of atypia, and the quantity of intratumoral thymocytes [
4,
5]. Histological classification alone is, however, insufficient to distinguish biologically indolent from aggressive TETs. Treatment outcomes and recurrence results appear to have a stronger correlation with the invasive and infiltrative characteristics of tumor cells [
3]. Consequently, histological benign thymomas may exhibit aggressive behavior [
3]. Survival rates vary, at 12.5% for invasive and 47% for non-invasive TETs over a 15-year period [
1]. Among the several histological subtypes of TETs, TC is identified as the most aggressive variant, displaying an increased tendency towards metastatic dissemination [
4,
5]. Given the challenges in differentiating TETs among cases with potentially benign or malignant behavior, researchers try to find more precise molecular tools for this purpose. In this context, exploring mutations and signaling pathways becomes pivotal in enhancing the overall understanding of tumor biology in TETs.
2. Genetic Alterations in Thymic Epithelial Tumors
2.1. Next Generation Sequencing
Numerous NGS investigations have recently tried to comprehensively explore the genetic foundation of TETs [
6,
7,
8,
9,
10,
11,
12,
13,
14,
15,
16,
17,
18,
19,
20,
21,
22,
23]. Due to the rarity of TETs, only limited studies have collected high-throughput sequencing datasets. Nonetheless, advancements in various molecular tests, especially NGS methods, have unveiled that TETs demonstrate the least tumor-mutation burden among adult cancers [
6]. Several NGS studies emphasize this low mutation burden in TETs. Yamaguchi et al., for example, did not find any mutations in 19 of 24 investigated cases, with only three cases demonstrating
KRAS and non-synonymous
HRAS mutations and two having low
DNMT3A mutations [
7].
The key tumor-suppressor genes,
P53 (
TP53) and
EGFR, commonly mutated in human cancers, have also been scrutinized in thymomas. A recent study by Syahruddin et al. identified
EGFR Exon 18 mutation (E709K) and a nonsense mutation in a small subset of 7.4% of cases [
8].
TP53 Exon 6 mutations, including both missense and nonsense mutations, were also detected only in 7.4% of cases. Both results suggest a limited role of
EGFR and
TP53 in the pathogenesis of thymomas [
8].
TP53 mutations, however, seem to be prominent in TCs, especially in their highly aggressive forms [
9].
TP53, often co-mutated with
BCOR, dominates mutations in TCs, while
ASXL1 (p.E657fs) and
DNMT3A (p.G728D) mutations were observed in Type B3 thymomas [
10]. NGS, in particular Whole Exome Sequencing (WES) and Targeted Sequencing (TS), in TETs has revealed a distinctive missense mutation in
GTF2I, which is a general transcription factor [
11]. This mutation, particularly p.Leu404His, has been reported to be specific for TETs, mainly in Type A and AB thymomas, with a presence in 76–83% of cases, and which occurs less frequently in other subtypes, notably only in 8% of TCs [
12]. Hsieh et al. also report the presence of another point
GTF2I mutation, namely p.Leu424His, in multinodular thymoma with lymphoid stroma (MN-T) and suggest that this mutation consistently characterizes this histological subtype [
13]. However, there are regional variations; studies in Japanese patients identified
GTF2I mutations in all thymoma types except for Type B3, which may, in a few cases, show
SMARCB1 and
STK11 gene mutations [
11]. In these studies,
GTF2I mutation was accompanied with
HRAS and
NRAS mutations, suggesting a potential exclusivity to indolent thymomas [
6,
11]. Further studies, including a recently meta-analysis, confirmed that
GTF2I mutations, along with
TP53 and
HRAS, are prevalent in thymomas, contributing to disease onset and progression [
14]. Furthermore, the persistence of specific
GTF2I mutations in certain subtypes could suggest a strong genetic link between them, which is exemplified in MN-T and Type A/AB thymomas [
13,
14].
Apart from
GTF2I, a very small subset of genes exhibits recurrent mutations in TETs, occurring in at least 3% of the cases [
12]. These genes include—in addition to
HRAS and
TP53—
CYLD,
PCLO, and
HDAC4 [
12].
HRAS mutations have been reported to be prevalent in Type A and AB thymomas, being the second most mutated gene in these tumors with a frequency of 7% [
12].
PCLO and
HDAC4 mutations occurred in various thymoma subtypes and TCs, each with a frequency of 3% [
12]. In summary, commonalities and differences in the identified genes of various WHO subtypes of TETs have been observed [
12]. As shown in this table, common genes found in multiple subtypes include GTF2I,
HRAS,
TP53, HDAC4,
PTPRB, and
NOD1. Types A, AB, and B2 thymomas share many of these genes, with additional occurrences of
PAX7 and
CSF1R in Type A thymoma and
ZMYM3 in Type AB thymoma. Individual genes specific to certain subtypes include
BCOR in Type A thymoma,
PBRM1 in Type AB thymoma, and
BRD4 in Type B2 thymoma [
12]. TCs exhibit an expanded genetic diversity, incorporating many genes which are also present in thymomas. However, TCs additionally feature
CYLD,
FGF3,
BRD4, and several recurring somatic mutations such as
TET2,
SETD2,
FBXW7, and
RB1 [
12]. The absence of a
GTF2I mutation in TCs, as noted by Saito et al., highlights a specific genetic distinction [
15].
Another interesting finding is the observation of
POLE mutations in two of the five investigated cases of an exceedingly rare thymoma subtype, namely metaplastic thymoma (MT), which histologically sometimes resembles Type A thymoma [
22].
Various investigations have also concentrated on examining samples of tumor tissue from individuals who have previously received chemotherapy [
23]. The analysis of pre-treated TET patients reveals somatic mutations in various genes, contributing to our understanding of the genetic landscape associated with disease progression. In this context, Wang et al. examined 197 cancer-associated genes in pre-treated TET patients, detecting somatic mutations in 39 genes [
23]. The prevalence was 62% in TCs and 13% in thymomas, with recurrent mutations in
BAP1,
BRCA2,
CDKN2A,
CYLD,
DNMT3A,
HRAS,
KIT,
SETD2,
SMARCA4,
TET2, and
TP53 [
23].
2.2. RNA Sequencing
Wang et al. utilized Ribonucleic Acid (RNA) Sequencing (RNA-Seq) and Fluorescence In Situ Hybridization (FISH) to explore gene fusions in thymomas. In 80% of MT cases, Wang et al. identified a
YAP1–
MAML2 rearrangement, which was not present in Type A thymomas [
22,
24] and which could help the differential diagnoses between these entities.
A similar observation from Vivero et al. [
25] suggested that MTs can be distinguished by the presence of
YAP1–
MAML2 fusions and the absence of
GTF2I mutations in contrast to Type A and AB thymomas; therefore, they represent a distinct and clinically benign entity [
24]. Moreover, MTs reportedly exhibit a chromosome 11 inversion [
22,
24].
Ji et al. conducted a further investigation on fusion genes in thymomas, identifying 21 fusions in 25 patients [
26]. Among them,
KMT2A–
MAML2,
HADHB–
REEP1,
COQ3–
CGA,
MCM4–
SNTB1, and
IFT140–
ACTN4 exhibited a significant expression, which was unique to thymoma samples [
26].
The study of Ji et al. also revealed abnormal long non-coding RNAs (lncRNAs) and microRNAs (miRNAs) in the regulatory network [
26]. The authors highlighted 65 distinct expression patterns of lncRNAs in thymomas, such as AFAP1–AS1, LINC00324, ADAMTS9–AS1, VLDLR–AS1, LINC00968, and NEAT1 and 1695 overexpressed miRNAs, validating their expression patterns through confirmation with “The Cancer Genome Atlas” (TCGA) database [
26]. Moreover, elevated expression of AFAP1–AS1 and reduced expression of LINC00324 significantly influenced patients’ survival outcomes [
26].
Abnormal miRNA expression influences immune cell functions and tumor progression. Wang et al. identified miR-130b-5p, miR-1307-3p, and miR-425-5p as predictive parameters for TET patients [
28]. Enkner et al. found significant differences in the miRNA repertoire between Type A thymomas and TCs, with contrasting expression of C19MC and C14MC miRNA clusters [
9].
TCs exhibit also an altered expression in non-clustered miRNAs, including upregulation of miR-21, miR-9-3, and miR-375, and reduced expression of miR-34b, miR-34c, miR-130a, and miR-195 [
9].
Most studies on tumorigenesis have focused on examining genetic alterations at the level of individual genes, with limited investigations in this area in TETs. Yu et al. examined 31 thymoma samples and identified 292 genes with overexpression of more than twofold, including six previously identified pivotal oncogenes (
FANCI,
NCAPD3,
NCAPG,
OXCT1,
EPHA1, and
MCM2) [
30]. Among the highly upregulated genes were
CCL25,
HIST1H1B,
SH2D1A,
DNTT,
PASK,
CENPF,
HIST1H2BD,
S100A14, and
NPTX1 [
30].
2.3. Quantitative Real-Time PCR
Quantitative Real-Time PCR (qRT-) PCR has also been employed to confirm alterations in gene expression in TETs. Yu et al. in a messenger RNA (mRNA) study demonstrated increased expression of
E2F2,
EPHA1,
CCL25, and
MCM2 as well as decreased expression of
MYOC,
FABP4,
IL6, and
CD36 [
30]. Interestingly,
EPHA1 showed a significant mRNA overexpression in 71.0%, and
MCM2 in 61.3% of the cases [
30].
Vodička et al. utilized reverse-transcription–quantitative PCR (RT-qPCR) to analyze the mRNA expression levels of
CTNNB1,
CCND1,
MYC,
AXIN2, and
CDH1 [
33], which were overexpressed in thymomas compared to control samples, with the exception of
AXIN2 in Type B thymomas [
33]. In this case, mRNA expression exhibited a gradual increase from Type B1 to Type B3 thymoma [
33]. Notably, thymomas Type A showed an exclusive significant increase in
AXIN2 mRNA expression [
33].
Xi et al. explored the differential gene expression in thymoma-associated myasthenia gravis (TAMG) and non-myasthenia gravis thymoma (NMG), revealing 169 genes with distinct expression levels; 6 of them were overexpressed in T cells, namely
ATM,
SFTPB,
ANKRD55,
BTLA,
CCR7,
TNFRSF25 [
34]. Additionally, in the comparison of the transcriptomes between TAMG and NMG, Yu et al. observed an upregulation of 5 genes (
PNISR,
CCL25,
NBPF14,
PIK3IP1, and
RTCA) by more than twofold, while over 30 genes were downregulated by more than a twofold decrease in TAMG [
30].
GADD45B,
SERTAD1,
TNFSF12,
MYC, and
ADPRHL1 were the most downregulated; these results were confirmed using qRT-PCR [
30]. Another interesting study found a reduced expression of miR-20b in TAMG [
35], which was identified to target
NFAT5 and
CAMTA1. The presence of miR-20b in cultured cells led to the suppression of
NFAT5 and
CAMTA1 expression, whereas an inverse correlation between miR-20b and
NFAT5/
CAMTA1 expression levels in patients with TAMG have been observed [
35].
2.4. Chromatin Immunoprecipitation Sequencing
Moreover, an additional inventive strategy entails utilizing Chromatin Immunoprecipitation Sequencing (ChIP-seq) to discern distinct interactions between proteins and Deoxyribonucleic Acid (DNA). Chromatin-associated proteins are crucial in coordinating the spatial and temporal control of gene expression [
39]. It is imperative to ascertain the genomic locations where these proteins establish binding, as they are integral to unraveling gene regulation intricacies and facilitate the exploration of pathways pertinent to tumor progression [
39]. ChIP-seq precisely pinpoints genomic regions to which specific proteins such as transcription factors attach, thereby playing a pivotal role in comprehending the regulation of gene expression and the associated chromatin alterations [
39].
2.5. Methylation Analysis
The methylation analysis facilitates an in-depth examination of epigenetic changes, providing insights into the regulation of gene expression [
41]. Prompting alterations in gene expression, epigenetic changes involve adjustments in DNA methylation, post-translational modifications on histone tails, and disrupted expression of non-coding RNA molecules (ncRNAs) [
41]. These modifications are progressively acknowledged as factors in tumorigenesis, contributing to genome instability, chromosome abnormalities, activation of transposable elements, elevated expression of proto-oncogenes, and inhibition of tumor suppressor genes [
41]. There is indeed a growing interest in identifying epigenetic biomarkers for cancer, with potential applications in clinical settings for diagnostic or prognostic purposes, or as innovative targets for therapeutic interventions [
41].
TCs with
TET2 mutations seem to show more hypermethylated genes, correlating with downregulated gene expression. In a recent study, nine out of thirty nine mutated genes (23%) were involved in epigenetic regulation, with 34% of TCs exhibiting recurrent mutations in seven of them (
BAP1,
ASXL1,
SETD2,
SMARCA4,
DNMT3A,
TET2, and
WT1), an observation which was not present for thymomas [
15,
23]. TETs have also been documented to exhibit the inactivation of tumor suppressor genes, such as
FHIT,
MLH1, and
E-cad, through promoter hypermethylation [
40].
An investigation of Chen et al. on global methylation levels and the promoter methylation status of tumor suppressor genes (TSG)
hMLH1,
MGMT,
p-16INK4a,
RASSF1A,
FHIT,
APC1A,
RARB,
DAPK, and
E-cadherin in 65 TET samples indicated hypermethylation and reduced TSG expression in types B1 or higher thymomas [
45]. Compared to early-stage TET, there was a decrease in global DNA methylation levels, while the expression of
DNMT1,
DNMT3a, and
DNMT3b increased in advanced-stage thymomas [
45].
Muguruma et al. also conducted a comprehensive examination of the methylation status of the two cancer-related genes,
MT1A and
NPTX2, in a total of 48 thymic tumor samples (31 thymomas, 17 TCs) and 22 paired normal tissue samples [
46]. The methylation levels of the
MT1A gene were significantly elevated in TC compared to thymoma (26.4% versus 9.5%), demonstrating exceptional sensitivity and specificity in distinguishing between TCs and thymomas [
46]. Similarly, the
NPTX2 gene exhibited markedly higher methylation levels in TC compared to thymoma (38.0% vs. 17.5%), with notable sensitivity and specificity in terms of discriminating between TCs and thymomas [
46]. Despite a significant increase in DNA methylation in TC compared to normal thymus, no significant distinction was observed between the DNA methylation levels of thymoma and normal thymus. Furthermore, no variations in the DNA methylation of
MT1A and
NPTX2 were identified among the several histological types of thymoma based on the WHO histologic classification [
46].
The methylation analysis could also be employed for diagnostic purposes for the distinction of the various WHO subtypes [
48]. To improve precision, particularly in morphological borderline cases, Gaiser et al. explored a methylation pattern-based classification [
48]. In their study, an array-based DNA methylation analysis of 113 thymomas with detailed histological annotation was conducted. Unsupervised clustering and t-Distributed Stochastic Neighbor Embedding (t-SNE) analysis aligned thymoma samples with the current WHO classification. However, methylation analyses revealed a nuanced distinction within histological subgroups AB and B2, yielding two methylation classes: mono-/bi-phasic AB-thymomas and conventional/“B1-like” B2-thymomas. This emphasizes the potential of methylation-based classifications to refine diagnostic criteria, improve reproducibility, and impact treatment decisions [
48].
In addition to the WHO criteria, the Masaoka-Koga Staging System plays a pivotal role in the prognosis of TETs. In a comprehensive study by Li et al., an intricate analysis encompassed TCGA 450 K methylation array data, transcriptome sequencing data, WHO histologic classification, and the Masaoka-Koga staging system [
42]. The primary objective was to discern differentially expressed methylation sites distinguishing thymoma from thymic carcinoma, as well as identifying DNA methylation sites linked to the overall survival of TET patients. Employing pyrosequencing, 100 patients with TETs had four specific methylation sites (cg05784862, cg07154254, cg02543462, and cg06288355) sequenced from their tumor tissues. Examination of the TCGA dataset unveiled 5155 hypermethylated and 6967 hypomethylated CpG sites in Type A–B3 and Type C groups, respectively, with 3600 located within gene promoter regions [
42]. Promoter hypermethylation led to the silencing of 134 genes, while 174 mRNAs experienced upregulation. Four specific genes (cg05784862/
KSR1, cg07154254/
ELF3, cg02543462/
ILRN, and cg06288355/
RAG1) emerged as independent prognostic factors for overall survival in TET patients [
42]. The prognostic model, comprising these genes, demonstrated superior accuracy in predicting 5-year overall survival compared to the Masaoka-Koga clinical staging [
42].
2.6. Comparative Genomic Hybridization
Copy number alterations (CNAs) in different chromosomes have been extensively documented in TETs, and their occurrence is more prevalent in histological subtypes associated with aggressive behavior [
43]. The reported chromosomal changes seem to allow the categorization of TETs into distinct groups, unveiling a connection between genetic discoveries and histology based on WHO classification [
43]. Notably, there are shared alterations between Type AB and B2 thymoma, B2 and B3 thymoma, as well as B3 thymomas and TCs [
43].
Another Comparative Genomic Hybridization (CGH) study conducted by Lee et al. involving 39 TET patients, revealed different chromosomal losses among the various WHO histological subtypes [
52]. In particular, Type A thymomas displayed losses in chromosomes 2, 4, 6q, and 13; Type B1 in chromosomes 1p, 2q, 3q, 4, 5, 6q, 8, 13, and 18; Type B2 in chromosomes 1p, 2q, 3q, 4, 5, 6q, 8, 13, and 18; and Type B3 in chromosomes 2q, 4, 5, 6, 8, 12q, 13, and 18 [
52]. Notably, Type A exhibited the least chromosomal abnormality, whereas Type B thymoma subtypes demonstrated overlapping loss patterns across various chromosomes [
52]. Chromosome 9q gain was exclusive to Type B1, and chromosome 1q gain was present solely in Type B3 [
52].
2.7. Fluorescence In Situ Hybridization
A tumor profiling study conducted by Enkner et al. using FISH reported a 38% deletion of the
CDKN2A gene, 32% deletion of the
TP53 gene, and 8% deletion of the
ATM gene [
9]. Conversely, only one Type B3 thymoma displayed
CDKN2A gene loss, and none of the Type A thymomas showed any deletion [
9].
Chromosomal abnormalities have been documented across all histological subtypes of TETs, encompassing translocation t(15;19) and deletions in the 6p22-p25 region. A notable instance of such translocations is t(15;19)(q13:p13.1), resulting in the formation of the fusion gene
BRD4–
NUT, identified in undifferentiated TCs [
54]. Among thymomas, one of the prevailing genetic alterations is situated in the chromosome 6p21.3, specifically at the major histocompatibility complex (MHC) locus, and extends to 6q25.2 to 25.3 [
54].