Molecular Alterations in Cutaneous Squamous Cell Carcinoma in Immunocompetent and Immunosuppressed Hosts—A Systematic Review

Simple Summary Cutaneous squamous cell carcinoma (cSCC) is the second most common form of skin cancer worldwide. Due to its high mutational burden, cSCC remains poorly understood at the molecular level. A considerable number of profiling studies have previously been performed on cSCC, but progress in the field has been slow due to the lack of consensus among these studies. Immunosuppressed patients (e.g. organ transplant recipients) are at greater risk of developing cSCC and this population experiences greater morbidity from this disease. In this study, we aim to review the molecular profile of cSCC among immunocompetent patients (ICPs) and immunosuppressed patients (ISPs) and to identify novel biomarkers of this disease. The molecular characterization of cSCC will shed new light on dysregulated pathways and potentially identify new key drivers of the disease, which may guide the direction of future targeted therapy in cSCC. Abstract The characterization of cutaneous squamous cell carcinoma (cSCC) at the molecular level is lacking in the current literature due to the high mutational burden of this disease. Immunosuppressed patients afflicted with cSCC experience considerable morbidity and mortality. In this article, we review the molecular profile of cSCC among the immunosuppressed and immunocompetent populations at the genetic, epigenetic, transcriptomic, and proteometabolomic levels, as well as describing key differences in the tumor immune microenvironment between these two populations. We feature novel biomarkers from the recent literature which may serve as potential targets for therapy.


Introduction
Cutaneous squamous cell carcinoma (cSCC) is the second most common form of skin cancer worldwide. The global trends in cSCC are worrying. According to the Global Burden of Disease Study (GBD) 2017 database, which assessed worldwide trends in skin cancer across 195 countries from 1990 to 2017, the incidence of cSCC increased by 310% during this period. Statistically, cSCC ranks the highest among any cancer tracked by the GBD [1]. In Singapore, cSCC constitutes 28.3% of all skin cancers diagnosed, with the mean age of cancer diagnosis occurring at 72.7 years. The age-standardized incidence rates for cSCC across all ethnic groups locally ranged from 0.8 to 3.5 tumors per 100,000 person-years from 1968 to 2016 [2]. Although this figure is relatively low compared to those of Western countries, the rising incidence of cSCC worldwide poses a considerable public health threat.
Due to its high mutational burden, cSCC remains poorly understood at the molecular level. A considerable number of profiling studies have previously been performed on cSCC, but progress in the field has been hampered by the lack of consensus among these countries, the rising incidence of cSCC worldwide poses a considerable public health threat.
Due to its high mutational burden, cSCC remains poorly understood at the molecular level. A considerable number of profiling studies have previously been performed on cSCC, but progress in the field has been hampered by the lack of consensus among these studies. Immunosuppressed patients are at greater risk of developing cSCC and this population experiences greater morbidity from the disease. The molecular characterization of cSCC sheds light on dysregulated pathways and identifies key drivers of the disease, thus heavily influencing the direction of future therapy in cSCC. In this study, we aimed to review the molecular profile of cSCC among immunocompetent patients (ICPs) and immunosuppressed patients (ISPs) and to identify novel biomarkers of this deadly disease.

Materials and Methods
The Preferred Reporting Items for Systematic Reviews and Meta-Analysis protocols (PRISMA-P) guided the methodology of this systematic review (Figure 1). A review protocol was entered into the PROSPERO database (Registry number: CRD42023394796). Online databases were queried for articles written in English from 1 January 2017 to 25 May 2022. The following databases were searched: PubMed, Medline, and Embase, using a combination of terms and their synonyms. Four primary search domains were employed, which were combined with the Boolean operator AND, whereas search terms contained within each domain were combined with the Boolean operator OR ( Figure 2). The search was not restricted by study design.   The retrieved literature was screened by two independent authors (DAT, OCC) using titles and abstracts for inclusion. In situations in which the suitability of an article was uncertain, an assessment of the full text was carried out and discrepancies were resolved through a vote of consensus. Articles were selected based on the following inclusion criteria: (1) being written in the English language, (2) having an emphasis on cSCC, and (3) containing data focusing on the molecular profiling of cSCC. Articles were excluded for the following reasons: (1) not reporting original data, (2) having no data about molecular alterations, (3) having data which did not focus on cSCC, (4) the full text was not available, or (5) reporting on isolated cases. Human cell lines and representative animal models were not excluded. Additional articles found after the completion of the full-text review of the selected articles were added to our review if they met the eligibility criteria, did not meet the exclusion criteria, and were not duplicates.

Results
Our literature search enabled us to retrieve 1056 articles. An additional 16 articles were identified during the review and were also included. A total of 841 articles remained after duplicates were removed. After reviewing title and abstract, a further 703 articles were excluded. Full-text screening was performed on 138 articles, of which 80 were included for the final qualitative synthesis (Figure 1).

Molecular Alterations in Immunocompetent Hosts
A total of 68 articles discussed molecular alterations associated with cSCC in the immunocompetent population. Changes in genetic expression were reported in 22 articles and epigenetic alterations were described in six articles. A further 21 articles discussed transcriptomic changes in cSCC, whereas 15 studies examined proteometabolic expression. There were nine articles that analyzed the immune microenvironment of cSCC.

Genetic Expression
Perineural invasion (PNI) is defined as tumor cells invading the perineural space. This feature has been established as one of the high-risk factors of cSCC, and its occurrence portends higher rates of local recurrence, metastases, and poor survival [3]. Three studies reported genetic alterations observed in cSCC with PNI (Table 1). All three articles examined cSCC of the head and neck (cSCCHN). Expression profiling of autophagy-related genes by Zheng et al. identified 239 differentially expressed genes (DEGs). These included upregulated genes in cSCC with PNI, such as MAPK8, which had the highest node degree (41), followed by ERBB2 (31) and HIF1A (30), as well as downregulated genes such as TNF, with the highest node degree of 44, followed by MYC (42), BCL2L1 (36), MTOR (34), and PPARγ (32). In addition, RAB23 gene expression was positively correlated with HIF1A (p = 0.001, r = 0.690), MAPK8 (p = 0.007, r = 0.583), and ARFGAP1 (p = 0.000, r = 0.655) but negatively associated with MTOR (p = 0.002, r = −0.748) and BCL2L1 (p = 0.015, r = The retrieved literature was screened by two independent authors (DAT, OCC) using titles and abstracts for inclusion. In situations in which the suitability of an article was uncertain, an assessment of the full text was carried out and discrepancies were resolved through a vote of consensus. Articles were selected based on the following inclusion criteria: (1) being written in the English language, (2) having an emphasis on cSCC, and (3) containing data focusing on the molecular profiling of cSCC. Articles were excluded for the following reasons: (1) not reporting original data, (2) having no data about molecular alterations, (3) having data which did not focus on cSCC, (4) the full text was not available, or (5) reporting on isolated cases. Human cell lines and representative animal models were not excluded. Additional articles found after the completion of the full-text review of the selected articles were added to our review if they met the eligibility criteria, did not meet the exclusion criteria, and were not duplicates.

Results
Our literature search enabled us to retrieve 1056 articles. An additional 16 articles were identified during the review and were also included. A total of 841 articles remained after duplicates were removed. After reviewing title and abstract, a further 703 articles were excluded. Full-text screening was performed on 138 articles, of which 80 were included for the final qualitative synthesis ( Figure 1).

Molecular Alterations in Immunocompetent Hosts
A total of 68 articles discussed molecular alterations associated with cSCC in the immunocompetent population. Changes in genetic expression were reported in 22 articles and epigenetic alterations were described in six articles. A further 21 articles discussed transcriptomic changes in cSCC, whereas 15 studies examined proteometabolic expression. There were nine articles that analyzed the immune microenvironment of cSCC.

ERBB4
Seen only in localized cSCC (20%). ERBB4 and STK11 were found to be significantly co-occurring in localized high-risk SCC (pair wise Fisher's exact test p < 0.05)

KDR
One of the most frequently mutated genes in localized (40%) and metastatic cSCC (30%)

KIT
One of the most frequently mutated genes in localized (10%) and metastatic cSCC (20%)

PTEN
One of the most frequently mutated genes in localized (10%) and metastatic cSCC (20%)

STK11
Seen only in localized cSCC (30%). ERBB4 and STK11 were found to be significantly co-occurring in localized high-risk SCC (pair wise Fisher's exact test p < 0.05)

TP53
One of the most frequently mutated genes in localized (70%) and metastatic cSCC (70%) Genes with tumor-suppressive roles. Seven studies reported genes with tumorsuppressive roles in cSCC development (Table 3). Using immunocompetent mice, Alameda et al. showed that a moderate increase in CYLD expression levels reduced NF-kB activation, which favored the differentiation of tumor epidermal cells and inhibited its proliferation and decreased tumor angiogenesis [11]. In another study utilizing mouse models, the authors identified RIPK-PKP1 signaling as a novel axis involved in skin stratification and tumorigenesis, whereby the phosphorylation of PKP1's N-terminal domain by RIPK4 is necessary for epidermal differentiation [12]. Partial loss-of-function mutations in the genes encoding subunits of RNase H2 compromised ribonucleotide excision repair in mouse epidermis, which led to spontaneous DNA damage, a type I interferon response, skin inflammation, and cSCC development [13]. Using a murine skin carcinogenesis model, Sunkara et al. showed that SFRP1 knockout resulted in the upregulation of genes involved in epithelial-to-mesenchymal transition, stemness, proliferation, and metastasis [14]. Moreover, Zhou et al. identified HOXA9, a direct target of onco-miR-365, to be significantly downregulated in human cSCC tumors and cell lines. The absence of HOXA9 positively regulates HIF-1 and its downstream glycolytic regulators, which enhances glycolysis necessary for cSCC development, proliferation, migration, and invasion [15]. Using whole-genome sequencing, Thind et al. demonstrated significant recurrent copy number loss in the tumor suppressor genes KANSL1 and PTPRD [16]. Likewise, KMT2C, CREBBP, and NCOA2 were identified to demonstrate tumor-suppressive roles in the initiation and progression of human cSCC [17]. Genes with oncogenic roles. There were 27 unique genes which demonstrated oncogenic roles in the development of cSCC in our review ( Table 4). The expression of CDC258 and CDC25C was increased in cSCC compared to normal skin (p ≤ 0.001) and led to the suppression of apoptosis in cSCC cells by stimulating Pl3K/Akt signaling and survivin expression [18]. FGFR2 activation and protein expression was increased in cSCC cells and found to be low in premalignant lesions and normal skin [19]. In a large study which sought to identify genetic variants in the HLA class II region associated with the risk of cSCC, the authors found that cSCC risk was associated with rs28535317 (OR = 1.20, p = 9.88 × 10 −11 ), which corresponded to an amino-acid change from phenylalanine to leucine at codon 26 of HLA-DRB1 (OR = 1.17, p = 2.48 × 10 −10 ). An independent association was observed for a threonine-to-isoleucine change at codon 107 of HLA-DQA1 (OR = 1.14, p = 2.34 × 10 −9 ). Additional independent cSCC associations with DQA1*05:01 and DQA1*05:05 were also uncovered. Among the classical HLA alleles, cSCC was associated with DRB1*01 (OR = 1.18, p = 5.86 × 10 −10 ) [20]. The expression of SERPINE1 was upregulated in all tumor cohorts compared to controls in a gene expression profiling study [7]. Thind et al. also observed that the 3 UTR regions of EVC (48%), PPP1R1A (48%), and LUM (16%) were significantly functionally altered (Q-value < 0.05) in the non-coding genome of cSCC. Moreover, significant recurrent copy number gains in CALR, CCND1, and FGF3 were observed for coding regions [16]. In a recent study by Yan et al., differential expression analysis demonstrated that many members belonging to the S100 gene family, the SPRR gene family, and FABP5 were significantly upregulated in cSCC cells. The top ten upregulated genes comprised of S100A9, S100A8, SFN, S100A7, S100A2, SPRR2A, FABP5, ISG15, KRT6B, and KRT16. Furthermore, seven keratin-encoded genes (KRT5, KRT6A, KRT6B, KRT6C, KRT14, KRT16, and KRT17), six genes from the S100 family (S100A2, S100A7, S100A7A, S100A8, and S100A9), and five genes from the SPRR family (SPRR2A, SPRR2B, SPRR2D, SPRR2F, and SPRR1B) were significantly upregulated in cSCC cells [21].
Five new genes, HEPHL1, FBN2, SULF1, SULF2, and TCN1, were recently discovered in cSCC, which were significantly upregulated compared to normal skin (p < 0.001) and actinic keratosis (AK) (p < 0.01) [22]. Novel somatic mutations in MLH1 (Q407*, Q426*, R423*) were also observed in an analysis of ten cases of high-risk cSCCHN [6]. In another study, EPHA6 and EPHA7 were identified as targets within the Eph-ephrin pathway, which is responsible for mitigating decreased cell viability in cSCC [23]. The authors also reported that RAC1 was the largest hub within the Eph-ephrin signaling pathway (degree = 14). Novel mutations in BPI were identified in the same study (HGVS DNA reference: g.36938975G > A, g.36954682C > T) [23].
The expression of METTL3 was observed to be upregulated in cSCC samples. Mechanistically, METTL3-mediated m 6 A modification regulated the expression of ∆Np63 in cSCC and influenced its stem-like properties, including its colony-forming ability and tumorigenicity [24]. Quan et al. profiled CD133+ cSCC cells, observing the increased expression of multiple components of the NOTCH and NF-κB signaling pathways, including the key components NOTCH1, IKKα (CHUK), RELA, and RELB, which contributed to the maintenance of its stem-like phenotypic features [25].

Epigenetic Alterations
The epigenetic landscape plays a key role in the tumor microenvironment. Through epigenomic profiling, Latil et al. demonstrated that the priming of the epithelial-mesenchymal transition (EMT) occurs in the cancer cell of origin. The authors showed that chromatin opening was mainly associated with gene activation, whereas chromatin closing was associated with gene repression during EMT. The transcription factor (TF) motifs that were found to be upregulated with the highest statistical significance during EMT and that were enriched in the open chromatin regions of tumor mesenchymal-like cells (TMCs) were Jun/AP1 (42%), NF1 (45%), Ets1 (10%), bHLH TFs (20-45%), Nfatc (27%), and Smad2 (37%). Motif enrichment analysis of the chromatin regions that opened during tumorigenesis revealed a strong enrichment of the binding sites of TFs such as Jun/AP1 (65%), Ets1 (37%), Runx (29%), Nf-kb (22%), and TEAD (25%) [26]. Aiderus et al. defined two mutually exclusive paralogous oncogenic drivers, Zmiz1 and Zmiz2, among the most recurrent drivers of cSCC development. Interestingly, all cSCC tumors with Zmiz1/2 insertions had inactivating insertions in at least one gene involved in chromatin remodeling, suggesting that alterations in epigenetic regulation are essential in cSCC development [17]. Actinlike protein 6A (ACTL6A, BAF53a) is a key protein subunit of the SWI/SNF epigenetic chromatin regulatory complex. ACTL6A knockdown was reported to reduce cSCC cell proliferation, spheroid formation, invasion, and migration. The authors proposed that ACTL6A suppresses p21Cip1 promoter activity to reduce p21Cip1 protein as a mechanism for maintaining the aggressive epidermal squamous cell carcinoma phenotype [27].

FGFR2
Human primary (SCC12A and SCC118) and metastatic cSCC cell lines (SCC7). Human normal skin samples (n = 9) AK (n = 9), cSCC (n = 28) and metastatic cSCC (n = 21) Strong expression of FGFR2 was observed in less than 5% of AK samples. In cSCC and metastatic cSCC, cytoplasmic and perinuclear FGFR2 was noted in tumor cells in the invasive margin and expression was predominantly strong (55% and 60%, respectively)

FBN2
Tissue samples of NNS (n = 6), NES (n = 6), AK (n = 6), and cSCC (n = 6)  DNA methylation is one of the various epigenetic mechanisms that control gene expression at the cellular level. The role of DNA methylation has been examined in three articles (Table 5). In a genome-wide DNA methylation profiling study, Hervás-Marín et al. found that invasive cSCC showed lower methylation levels than premalignant actinic keratosis. By contrast, high-risk non-metastatic and metastatic cSCC demonstrated higher methylation levels compared to low-risk cSCC (p < 0.001, two-sided t-test). Although overall, there was a risk-dependent change in DNA methylation patterns, mostly reflecting a gain of methylation, the authors proposed that a non-sequential and complex pattern of DNA methylation exists during cSCC progression [28]. Additionally, the expression of ID4 and UCHL1 was found to be significantly downregulated in cSCC tissues (p = 0.0111, p = 0.0205 respectively) and correlated with increased levels of promoter methylation (p = 0.00295, p = 0.0499, respectively) [29]. Hypermethylation of the FILIP1L locus was also observed in human cSCC and its expression was decreased in human cSCC cell lines [30].

Transcriptomic Changes MicroRNAs
MicroRNAs (miRNAs) are short noncoding RNAs that regulate the expression of protein-coding genes at the post-transcriptional level. We found seven studies that examined the expression of miRNAs in cSCC (Table S1). Wimmer et al. demonstrated that the overexpression of miR-10b conferred the stem cell-characteristic of a capacity for 3D-spheroid formation to keratinocytes. Analysis of the downstream effects of miR-10b identified the actin-and tubulin cytoskeleton-associated protein DIAPH2 as a novel putative target of miR-10b [31]. Several other miRNAs were found to have oncogenic roles in cSCC development. MiR-21 and miR-205 were upregulated in invasive cSCC compared to cSCC in situ (p < 0.05) [32]. MiR-31 demonstrated increased expression in cSCC compared to normal skin cell lines (p < 0.01) and enhanced cSCC cell viability (p < 0.01) [33]. MiR-186 overexpression led to significantly enhanced cell proliferation, invasion, and migration in cSCC cells compared with controls (p < 0.01) [34]. Likewise, miR-221 also showed in-creased expression in cSCC compared to noncancerous tissues (p < 0.05) and promoted cell proliferation [35]. By contrast, two miRNAs, miR-130a and miR-181a, demonstrated tumorsuppressive effects in cSCC development. MiR-130a expression was almost undetectable in cSCC samples, and its overexpression led to a significant reduction in tumor volume (p < 0.05 at week 4 and p < 0.01 at week 5) [36]. Similarly, miR-181a showed low abundance in cSCC samples compared to normal skin (p = 0.0088). MiR-181a overexpressing cells were observed to grow slower and reach termination criteria at later time points (p = 0.0001) [37].

ID4
8 pairs of matched human cSCC and adjacent normal skin tissues; sun-exposed normal skin in the head and neck region (n = 60) from normal patients and distal non-exposed normal skin from cSCC patients (n = 60) ID4 expression was downregulated in cSCC (p = 0.0111) and correlated with increased levels of promoter methylation (p = 0.00295) L. Li

Circular RNAs
Circular RNAs (circRNAs) are single-stranded, covalently closed RNA molecules which serve biological functions through various mechanisms such as by acting as transcriptional regulators, miRNA sponges, and protein templates [38]. Several circRNAs were identified among studies which examined transcriptomic alterations in cSCC (Table S2). Mahapatra et al. identified 55 circRNAs with significantly (p < 0.05) altered expression in cSCC. The majority of differentially expressed circRNAs (53 of 55) were downregulated and only two were upregulated in cSCC relative to healthy skin. Amongst the differentially expressed circRNAs, the most significantly downregulated circRNA was IFFO2, whereas the most upregulated was circ_EPSTI in terms of fold change [39]. In another study by Wei et al., 54 differentially expressed circRNAs were identified in cSCC. The top six upregulated circRNAs were hsa_circ_0068631, hsa_circ_0070-933, hsa_circ_0067772, hsa_circ_0003528, hsa_circ_0070934, and hsa_circ_0001955, and the top six downregulated circRNAs were hsa_circ_0022392, hsa_circ_0022383, hsa_circ_0005085, hsa_circ_0046449, hsa_circ_007-2279, and hsa_circ_0000375. Their parental genes were most enriched in the mitophagy, PPAR, HIF-1, and AMPK signaling pathways [40].

Transcription Factors
Transcription factors are involved in the conversion of genetic information from DNA into RNA. A large number of TFs with altered expression in cSCC were identified by Mahapatra et al. in a comprehensive analysis of transcriptomic changes in cSCC (Table S3).

Long Non-Coding RNAs
Long non-coding RNAs (lncRNAs) are RNAs longer than 200 nucleotides that are not translated into functional proteins. We identified 12 studies that examined the role of lncRNAs in cSCC development. In a recent comprehensive analysis of the lncRNA-mRNA co-expression network by Hu et al., a large number of lncRNAs were found to be differentially expressed between cSCC and healthy controls. The top seven differentially expressed lncRNAs are included in Table S4. Three lncRNAs (PVT1, CTD-2521M24.9, and AL353997.3) were upregulated in cSCC and four lncRNAs (MIR4720, BX004987.5, CTD-2619J13.13, and LINC00478) were downregulated in cSCC. The same study also identified six previously unstudied lncRNAs (GXYLT1P3, LINC00348, LOC101928131, A-33-p3340852, A-21-p0003442, and LOC644838) which could contribute to cSCC progression [42]. LncRNA RP11-493L12.5 was found to be most upregulated (46.77-fold), whereas KB-1410C5.3/lnc-GRHL2 (0.005-fold) was the most downregulated among the significantly altered lncRNAs in a transcriptomic analysis of cSCC [39]. Zhang et al. proposed a novel c-MYC-assisted MALAT1-KTN1-EGFR axis, which contributes to cSCC progression [43]. Another lncRNA, AK144841, demonstrated 40-fold greater expression in cSCC than in healthy skin. Moreover, AK144841 was found to inhibit gene expression, specifically downregulating the expression of genes of the Lce1 family, which is involved in epidermal terminal differentiation, and of anticancer genes (including Cgref1, Brsk1, Basp1, Dusp5, Btg2, Anpep, Dhrs9, Stfa2, Tpm1, SerpinB2, Cpa4, Crct1, Cryab, Il24, Csf2, and Rgs16), contributing to the dedifferentiation of tumor-forming keratinocytes, as well as molecular cascades involved in cSCC development [44]. LINC01048 was more greatly expressed in cSCC (p < 0.01) and recurrence tissues compared with adjacent normal and non-recurrence tissues. Mechanistically, LINC01048 was demonstrated to be transcriptionally activated by USF1, and the USF1-induced upregulation of LINC01048 promoted cell proliferation and apoptosis in cSCC by binding to TAF15 to transcriptionally activate YAP1 [45]. Li et al. reported the significant upregulation of LINC00319 in cSCC, which was associated with larger tumor size and lymphovascular invasion. Functional studies demonstrated that LINC00319 promoted CSCC cell proliferation, accelerated cell cycle progression, facilitated cell migration and invasion, and inhibited cell apoptosis. Mechanistic studies showed that LINC00319 exerts its oncogenic functions in CSCC through miR-1207-5p-mediated regulation of cyclin-dependent kinase 3 [46]. LncRNA EZR-AS1 expression was found to be significantly upregulated in cSCC tissues and cells compared with adjacent healthy tissues (p < 0.01), and its knockdown inhibited cSCC cell proliferation, migration, and invasion and promoted cell apoptosis [47]. LncRNA HCP5 exhibited the greatest upregulation in cSCC (logFC = 1.8) and the highest relative expression levels in cSCC compared to normal adjacent tissue (p < 0.001) in a recent study examining the competing endogenous RNA (ceRNA) network in cSCC [48]. The authors proposed that HCP5 may competitively bind to miR-138-5p to regulate EZH2 in cSCC cells, promoting autophagy and reducing apoptosis through the STAT3/VEGFR2 pathway.
By contrast, several other lncRNAs have demonstrated tumor-suppressive effects on cSCC development. Yu et al. proposed that lncRNA HOTAIR acted as a ceRNA to regulate PRAF2 expression through competitive binding to miR-326, contributing to cSCC development. HOTAIR overexpression significantly enhanced cSCC cell migration and proliferation and the EMT process, and its downregulation impeded these processes [49]. The expression of LINC00520, a novel lncRNA which has only been reported in a few tumors, was also reported to be downregulated in cSCC. The authors demonstrated that LINC00520 targeted EGFR, thus inhibiting the PI3K-AKT signaling pathway and suppressing cell proliferation and migration [50]. Other research showed that lncRNA TINCR overexpression promoted ALA-PDT-induced apoptosis and autophagy through the ERK1/2-SP3 pathway [51]. Finally, significant functional alterations were observed in the tumor-suppressing lncRNA LINC01003 (68% of specimens, Q-value: 0.0158) in a recent study analyzing cSCC samples from patients with regional metastases of the head and neck [16].

Protein Expression
There were 45 proteins of interest identified among 14 studies. Six proteins identified from two studies were downregulated in cSCC (Table S5). β-catenin expression was significantly reduced from normal in cSCC (93% to 69%, p < 0.001), and CK10 was found to be inversely correlated with cancer development (rs = −0.626, p < 0.001) [52]. Four other proteins, COL28A1, COL6A6, COL1A1, and TLN2, were also found to be decreased in cSCC. These proteins were mapped to two pathways (protein digestion and absorption and platelet activation), which were notably enriched in a study utilizing integrated proteomic and metabolomic analysis [53].
Several proteins were observed to contribute towards cSCC metastasis and differentiation (Table S7). A. Azimi et al. identified 5037 proteins across primary and metastatic cSCC samples, of which 19 proteins, including ISG15, APOA1, and MARCKS, which have roles in metastasis, were increased and 11 proteins, including DMKN, APCS, and CST6, were decreased in metastatic cSCC lesions relative to the primary phenotypes (adj. p-value < 0.05) [59]. In another study, RAC1 was observed to be upregulated in metastatic compared to low-risk cSCC [60]. The same study also revealed that PABPC1, LGALS3BP, MARCKS, and SND1 were significantly increased in high-risk cSCC compared to low-risk cSCC [60]. Recently, uPAR was identified as a biomarker in metastatic cSCC in a study which demonstrated that uPAR protein levels were significantly increased in metastatic cSCC. The same study showed that increased expression of the uPAR protein was correlated with the downregulation of hsa-miR-340-5p and hsa-miR-377-3p [7]. Another protein, ENTPD1, was found to demonstrate significantly higher expression in human cSCC that metastasized compared to tumors that were nonmetastatic ([+] Met, n = 54, [−] Met, n = 51, p < 0.001) [61]. Finally, poorly differentiated cSCC displayed significantly higher cytoplasmic and lower nuclear iASPP expression compared to well-differentiated tumors [62]. ∆Np63 and TAp63 are isoforms encoded by the TP63 gene and play important roles in cSCC development. ∆Np63 was reported to prime the cancer cell of origin toward well-differentiated tumors [26]. In another study, homozygous deletion of TAp63 not only increased the susceptibility of mice to cSCC (46.67% in TAp63−/− mice vs. 20% in wild-type mice), it also enhanced metastasis [63].

Metabolic Changes
The metabolic profile of cSCC was examined in one study (Table S8). W. Chen et al. showed that several standard amino acids were mapped to regulatory pathways involved in cSCC development. The levels of L-glutamate, L-aspartate, L-arginine, L-glutamine, and L-phenylalanine were increased, whereas the level of arachidonate was decreased in cSCC samples compared to matched noncancerous tissue samples [53].

Immune Landscape
Our review featured several studies that linked cSCC development with alterations within the innate immune system ( Table 6). A predominant protumor gene expression signature of tumor-associated neutrophils (TANs) was identified in cSCC in a study utilizing a mouse cSCC cell line. The study demonstrated a significant increase in the fraction of neutrophils within cSCC compared to surrounding skin (p < 0.01), with TANs accounting for 30-80% of tumor-infiltrating CD45+ cells [64]. In addition, analysis of the expression of inflammasome components in cSCC cell lines and normal human epidermal keratinocytes (NHEKs) demonstrated the upregulation of the expression of AIM2 in cSCC cells (p < 0.01). The authors found that AIM2 knockdown caused the downregulation of many genes encoding proteins regulating the cell cycle, including CDK1, CDC7, CCNA1, CCNB3, KIF11, and TTK, which led to a reduction in cell viability and invasion, the suppression of growth, and the onset of apoptosis in cSCC cells [65].
The role of the complement system in cSCC development was examined in four studies utilizing human cSCC cell lines. C1r was demonstrated to upregulate the production of invasion-associated matrix metalloproteinases (MMPs) MMP1, MMP13, MMP10, and MMP12 and to promote the invasion of cSCC cells [66]. C3 and CFB mRNA expression was increased in cSCC compared to NHEK samples (p < 0.05) [67]. The role of C3 in cSCC development was highlighted in another study, which showed that C3a upregulated cyclin D1, cyclin E, VEGF, pro-MMP1, and pro-MMP2 expression. Moreover, the expression of stemness factors Sox-2, Nanog, Oct-4, c-Myc, and CD-44 was also observed to be stimulated by C3a [68]. CFI overexpression was shown to increase the production of MMP-13 and MMP-2, ERK1/2 activation, and cell proliferation and to enhance cSCC invasion (p < 0.01) [69].
The T-lymphocyte profiles of cSCC have been described in two articles (Table 6). In a study comparing T-lymphocyte profiles between keratoacanthomas (KA) and invasive human cSCC, CD8+ cells were found to be increased, whereas CD4+ cells were found to be decreased in cSCC relative to KA. In addition, the infiltration of FOXp3+ T-regulatory cells was lower in invasive cSCC compared to KA. Bauer et al. also demonstrated that PD1+ and PD-L1+ cells were enriched in cSCC and PD-L1 was correlated with the differentiation state in cSCC. For invasive cSCC, increased PD-L1 expression was correlated with enhanced infiltration of CD4+ and CD8+, as well as FOXp3+ T-cells (p = 0.0049, p = 0.0069, and p = 0.0397, respectively) [70]. Non-progressing primary cSCCHN samples were characterized by greater CD8+ (p = 0.006) and CD4+ (p = 0.004) T-cell responses, with numerically enhanced regulatory T-cells compared to tumors which metastasized [71].

Molecular Alterations in Immunosuppressed Hosts
There were 14 studies which focused on the immunosuppressed population. Among these studies, the majority focused on the general organ-transplant population, including renal transplant recipients (RTRs). A small number of articles studied populations with unspecified immunosuppression.

IL-1β
No difference in IL-1β expression levels in cSCC with pain vs. cSCC without pain in OTRs.

PGE2
cSCC with pain is associated with increased levels of PGE 2 compared with cSCC without pain (OR = 1.9, 95% CI = 1.1-3.4, p = 0.002), adjusted for age and sex.  The increased expression of PD-L1 on CD8+ T cells in metastatic cSCC among ACIS patients was absent among ISP

Gene Polymorphisms
Seven unique alleles were identified across three studies ( Table 8). The IRF4 rs12203592 T allele was associated with a significantly increased hazard ratio for the time to first cSCC (HR = 1.36, p = 0.02) using univariate analysis. This association was maintained when adjusted for age, gender, organ transplanted, and Fitzpatrick skin type (HR = 1.34, p = 0.04) [77]. By contrast, OTRs homozygous for the brown eye alleles rs916977 (GG) and rs12913832 (AA) exhibited significant delays in the time to first cSCC post-transplant relative to OTRs homozygous for blue eye alleles (HR = 0.34, p < 0.001 and HR = 0.54, p = 0.012, respectively) [78]. Likewise, the SLC45A2 rs16891982 C allele was associated with a decreased hazard for cSCC in univariate analysis (HR = 0.58, p = 0.04). This effect was comparable but not significant with the application of a multivariate model (HR = 0.74, p = 0.06) [77]. In a later study by Kuzmanov et al., a genome-wide association study (GWAS) identified the SNV rs34567942 to be significantly associated with cSCC in OTRs (p-value threshold of 5 × 10 −8 ) [79].

Genetic and Epigenetic Alterations
Peters et al. identified 16 differentially methylated regions in RTRs, including ZNF577 and FLOT1 [80] (Table 9). A subsequent study revealed that higher DNA methylation of SERPINB9 occurred in RTRs who developed cSCC than in those who did not. The median DNA methylation of SERPINB9 was 58.7% (range 32.5-81.3%) for region 1 and 54.4% (30.0-78.5%) for region 2 in patients with cSCC, and 50.2% (21.8-77.5%) for region 1 and 46.4% (22.1-74.0%) for region 2 in the non-cSCC patients (region 1: p = 0.004; region 2: p = 0.008) [81]. In another study examining the tissue and circulating expression of miRNAs in OTRs with and without cSCC, the authors found that mir-1246 and mir-1290 were associated with cSCC in OTRs (p = 0.013 and p = 0.037, respectively) [82] (Table 9). In addition, a previously unknown mutational signature, termed signature 32, was discovered in cSCC samples of OTRs during whole-exome sequencing and mutational signature analysis. Signature 32 describes predominately C > T mutations (75%) in combination with C > A, T > A, and T > C mutations. Its pattern was observed to be putatively associated with azathioprine treatment and distinct from any of the previously known mutational signatures [83]. An analysis of treatment times revealed a strong positive correlation with the estimated time of azathioprine exposure and the prevalence of signature 32 (Spearman's rank order correlation rs (26) = 0.679, p < 0.0001) [83]. Substantial work has been carried out in ICP populations, with noteworthy findings including the major roles of TP53, NOTCH, TGFβ, and CDKN2A in the development of cSCC in this population [85][86][87][88]. In the recent literature, novel genetic alterations in cSCC have emerged, with roles in autophagy, perineural invasion, and metastasis.
The dysregulation of autophagy contributes towards cancer development, and recent research has shown that autophagy might play a pivotal role in the pathogenesis of cSCC [89,90]. Zheng et al. identified several key autophagy-related DEGs, namely, HIF1A, MAPK8, mTOR, BCL2L1, and RAB23, which were involved in cSCCHN with clin-ical PNI [4]. A close correlation may exist between autophagy and cSCC outcomes, and autophagy-related genes are a promising treatment target for skin cancer.
Moreover, a number of genes, including PLAU, PLAUR, MMP1, MMP10, MMP13, ITGA5, and VEGFA, were recently revealed to be differentially upregulated in metastatic compared to non-metastatic cSCC [7]. These genes are involved in cellular pathways and functions that support matrix remodeling, epithelial-to-mesenchymal transition, cell survival, and migration, all of which play important roles in tumor metastasis [7]. The discovery of these genes could be useful in identifying primary cSCC tumors with metastatic potential. TIMP4 was previously reported to be downregulated in non-cutaneous head and neck SCCs [91], but its downregulation in metastatic cSCC compared to sun-exposed skin was only recently uncovered [7].
Novel somatic mutations in MLH1 (Q407*, Q426*, R423*) were observed in an analysis of 10 cases of high-risk cSCCHN [6]. These mutations led to premature truncation and loss of the C-terminal dimerization domain. Two other new mutations, FGFR2 A380D and D528N, were discovered in a cohort of patients with high-risk cSCCHN. These mutations resulted in changes within the transmembrane domain and the protein tyrosine kinase domain, respectively, and were exclusively seen in patients with histologically proven PNI. Five novel genes, HEPHL1, FBN2, SULF1, SULF2, and TCN1, were also observed to be significantly upregulated in cSCC compared to normal skin [22]. In a human-to-mouse comparison of cSCC tumors, the authors discovered that miR-30c-2* and miR-497 were under-expressed in TAp63-deficient cSCC. In the same study, a seven-gene signature was identified, including five putative targets of miR-30c-2* (FAT2, ITGA6, KIF18B, ORC1, and PKMYT1) and four predicted targets of miR-497 (AURKA, CDK6, KIF18B, and PKMYT1), which were frequently overexpressed in cSCC [63]. Further studies on the roles of these genes in cSCC development could further enhance diagnosis and risk stratification and broaden therapeutic options.
Four previously unstudied circRNAs, circ_IFFO2, circ_TNFRSF21, circ_KRT1, and circ_POF1B, which were identified by Mahapatra et al., showed differential expression in cSCC [39]. CircRNAs have recently emerged as a novel member of the noncoding cancer genome, with roles in controlling cancer gene expression through mechanisms such as decoys to sponge miRNAs and as regulators of transcription and alternative splicing [92]. The role of circRNAs in cSCC is being studied at an increasing pace, and these molecular biomarkers may introduce new therapeutic opportunities for cSCC in the foreseeable future.
LncRNAs have also garnered interest as novel regulators of gene expression in the recent literature. Hu et al. identified six previously unstudied lncRNAs, including GXYLT1P3, LINC00348, LOC101928131, A-33-p3340852, A-21-p0003442, and LOC644838 [42]. These lncRNAs demonstrated co-expression with the mRNAs ACY3, NR1D1, and MZB1, which could contribute considerably to cSCC progression by regulating apoptosis induced by endoplasmic reticulum stress, cellular signal transduction, and autophagy [42]. MALAT, another lncRNA, was highlighted as part of a novel c-MYC-assisted MALAT1-KTN1-EGFR axis responsible for cSCC progression [43]. Transcriptomic sequencing identified KTN1 as the key mediator through which MALAT1 positively regulated the expression of EGFR and contributed to cSCC development [43].

Proteomic Biomarkers
Ten proteins identified by Sun et al. (CK10, CK17, CD44, EZR, E-cadherin, b-catenin, Hsp75, Hs-p90-α, EXOSC10, and SOD2) showed disease-progression-specific significance in cSCC [52]. The chaperone proteins, Hsp75, Hsp90-α, and SOD2, were the only three proteins that demonstrated a correlation with differentiation stages of cSCC. The expression of chaperone proteins increased in parallel at each stage of cSCC development. In contrast, SOD2 had no effect on early skin damage, as its expression was comparable in normal and pre-cancerous samples [52]. Increased Hsp promotes cancer progression by participating in microenvironment conditioning [93], which makes it an appealing target in cSCC treatment. Another protein reported to demonstrate an association with cSCC differentiation status was iASPP [62]. This protein acts by inhibiting the p53 and NF-kB signaling pathways during cSCC development. The authors also described a previously unreported mechanism in cSCC by which increased iASPP levels bind and repress MITF expression, reducing TRPM1 and consequently miR-211 expression, which contributes to the increased stabilization of p63 observed in cSCC [62]. Epigenetic dysregulation of this autoregulatory feedback loop which promotes cSCC development could serve as a novel candidate for targeted therapy.
Proteins associated with metastasis may be useful in discriminating between primary and metastatic cSCC lesions. Twelve protein markers were identified by Azimi et al. that could be used for the early diagnosis and risk stratification of primary cSCC lesions. Of these, six proteins including ISG15, APOA1, MARCKS, EFHD2, STMN1, and ACBD3 were increased and six proteins including DMKN, APCS, CPA3, KRT79, CST6, and CMA1 were decreased in metastatic cSCC compared to the primary lesions [59]. The exact roles of most of these proteins have not been well established, although proteins with increased abundance in metastatic cSCC have been reported to increase tumor invasion and metastasis in other cancers. Another six proteins, APOA1, ALB, SERPINA1, HLA-B, HP, and TXNDC5, were differentially abundant in cSCC compared to AK [54]. These proteins are useful biomarkers for distinguishing cSCC from its precursors and may serve as potential molecular targets for selective treatment of these tumors.

Immune Biomarkers
Defining the immune phenotypes within the tumor microenvironment (TME) of cSCC has been challenging, and immunoregulatory mechanisms leading to cSCC development remain unclear. T-lymphocyte profiles differ between AK, KA, and invasive cSCC [70]. Ferguson et al. showed that effective responses from both CD8+ and CD4+ T-cells in the TME are necessary for immune control of primary cSCCHN [71]. That study provided insights into the early events in cSCC that dictate the immune responses in primary tumors and influence disease outcomes. The ability to predict metastatic potential at the time of diagnosis of primary cSCC could be used to offer personalized care, including disease surveillance strategies and recognizing patients who will benefit the most from adjuvant therapy.
The complement system is a vital part of innate immunity against pathogen invasion. A number of complement factors have been implicated in solid organ cancers. In cSCC, complement factor H expression has previously been reported to enhance the growth and migration of cSCC cell lines [94]. Although C3 has been shown to be highly expressed in cSCC, its role remains unclear. Fan et al. elucidated a novel correlation between complement anaphylatoxin C3a and cSCC stemness, which could provide insights into the role of C3 in cSCC tumorigenesis [68]. CFB [67], CFI [69], and most recently C1r [66] were also reported to contribute towards cSCC progression.
Khou et al. discovered a predominant protumor gene expression signature of TANs in cSCC compared to normal skin. The authors found that in vivo depletion of neutrophils impeded tumor growth and significantly enhanced the frequency of proliferating IFN-gamma-producing CD8+ T-cells [64]. Mechanisms that limited antitumor responses involved high arginase activity, the production of reactive oxygen species (ROS) and nitrite (NO), and the expression of programmed death-ligand 1 (PD-L1) on TANs, concomitantly with an induction of PD-1 on CD8+ T-cells, which was correlated with tumor size [64]. These findings raise the possibility of targeting neutrophils and PD-L1-PD-1 interaction in the treatment of cSCC.

Key Molecular Alterations in the Immunosuppressed
CSCC is the most common post-transplant malignancy in organ-transplant recipients (OTRs) [95]. OTRs are 65-100 times more likely to develop cSCC than the general population [96,97]. Transplant patients with cSCC experience considerable mortality of 4.94 per 100,000 person years [98] and a greater morbidity ratio of up to 250 times compared with ICPs [99]. These differences are thought to arise from a more aggressive phenotype, a higher probability of metastasis (approximately 7%), and a greater recurrence rate (7-45%) [100][101][102].

DNA Methylation-A Novel Risk Factor for CSCC Development
Epigenetic alterations, in the form of DNA hypomethylation and hypermethylation, have previously been associated with the development of skin cancer in ICPs [103,104], although less is known regarding its role in ISP. A comparison of RTRs who developed a future de novo post-transplant cSCC and those who did not identified 16 differentially methylated regions. Noteworthy genes included ZNF577 and FLOT1, which encode a zinc-finger protein and a protein involved in T-cell migration, respectively [80]. The high DNA methylation of SERPINB9 in circulating T-cells was demonstrated in RTRs before the clinical onset of cSCC, as well as during recurrent post-transplant cSCC [81]. The DNA methylation of these genes could serve as a novel risk factor for the development of de-novo and subsequent post-transplant cSCC and provide opportunities for pretransplant risk stratification and post-transplant surveillance of cSCC.

Genetic Polymorphisms May Confer Protection against CSCC Development
Previously, GWAS performed for cSCC revealed that an increased cSCC risk was associated with polymorphisms in six pigment-related loci and that these influenced the cSCC risk independently of pigment-related phenotypes [105]. Moreover, variants at the HERC2/OCA2 locus have been associated with pigmentation phenotypes and the risk of developing several types of skin cancer. Candidate genes and GWAS have linked blue-eyeassociated alleles at the HERC2/OCA2 locus with an increased risk of cSCC in ICPs, but the roles of these alleles in ISPs are less established [78].
Wei et al. demonstrated that the blue-eye-associated variants of rs12913832 and rs916977 at the HERC2/OCA2 locus are associated with decreased time to first cSCC posttransplant in solid organ transplant recipients. In contrast, a protective role was found for several alleles located within genes influencing pigmentation, including the brown eye allele at OCA1/HERC2 and the C allele for SLC45A2, a transporter protein that mediates melanin synthesis and contributes to skin pigmentation. OTRs homozygous for the brown eye alleles rs916977 (GG) and rs12913832 (AA) exhibited significant delays in the time to first cSCC post-transplant compared with individuals who were homozygous for the blue eye alleles, whereas the SLC45A2 rs16891982 C allele was associated with a decreased HR for cSCC in univariate analysis [78]. These results could direct future recommendations in terms of screening guidelines and risk models for cSCC in OTRs.

Differences in the Immune Landscape between ICPs and ISPs May Influence Responses to Targeted Therapy
The tumor immune microenvironment has gained interest in the recent literature. An understanding of the immune mechanisms involved in cSCC progression is necessary in order to guide therapeutic approaches. ISPs exhibit lower levels of CD8+ T-cells compared to their immunocompetent counterparts [74]. The role of cytotoxic T-cells in tumor progression has been well-established, with increased CD8+ T-cells being associated with good prognosis in many cancers [104][105][106][107][108][109]. It is therefore not surprising that the absence of an appropriate CD8+ T-cell response leads to poorer outcomes among ISPs. The increased expression of several other immune checkpoint receptors, including PD-L1, PD-L2, IDO, and TIM3, observed in metastatic cSCC among ICPs was also absent among ISPs [71]. This finding is of significance because immunotherapeutic agents targeting these receptors may be less useful among the immunosuppressed population.

Conclusions and Future Directions
We have provided an overview of the molecular alterations driving cSCC progression among immunocompetent and immunosuppressed populations based on the recent literature, providing a framework for future research.
Novel biomarkers in cSCC have emerged as promising therapeutic targets in this era of precision medicine and may alter the treatment paradigm of this deadly disease. Genes with roles in autophagy, perineural invasion, and metastasis could serve as new therapeutic targets for high-risk cSCC. MiRNA, circRNA, and lncRNA are novel regulators of gene expression that drive cSCC development at the transcriptomic level and present exciting therapeutic opportunities, which could be explored in future studies. Certain proteins demonstrate disease-progression-specific significance and an association with cSCC differentiation status and may act as molecular targets for selective tumor treatment.
The complement system and the T-lymphocyte response influence cSCC progression by altering the tumor immune microenvironment. Immunotherapeutic agents targeting these signaling pathways could be investigated as therapeutic options.
Specific molecular alterations place ISPs at greater risk of cSCC development, and differences in tumor immune landscapes may explain the poorer therapeutic outcomes observed among ISPs compared to their immunocompetent counterparts. The unique molecular profile of cSCC in ISPs may be useful in predicting treatment responses, guiding the risk stratification process, and enhancing surveillance in this population.

Conflicts of Interest:
The authors have no conflict of interest to disclose.