Abstract
Background/Aim: The traditional melanoma staging system is not ideal for predicting patients’ individual risk of disease recurrence and death. Subsequently, suboptimal adjuvant treatment may be offered. Hence, identification of biomarkers to optimize risk stratification is warranted. Cyclins and cyclin-dependent kinase inhibitors, key players in the cell cycle regulation, are candidate prognostic biomarkers. Herein, their expression and prognostic value in melanoma are studied. Patients and Methods: The expression of Cyclin D1, Cyclin E, p16, p21, and p27 were assessed using immunohistochemistry in samples from 59 patients with melanoma and correlated with clinicopathological parameters, as well as relapse-free survival (RFS) and overall survival (OS). Results: Cyclin E expression in the nucleus, observed in 19% of patients, was correlated with Breslow thickness (p=0.017), whereas p16 loss in the cytoplasm and nucleus, detected in 68% and 61% of patients respectively, was correlated with Breslow thickness (p=0.001) and ulceration (p=0.035). The high expression of Cyclin E in the nucleus (p<0.001) and the loss of p16 in the cytoplasm (p=0.012) and nucleus (p=0.002) were associated with shorter OS. Cyclin E in the nucleus was a prognostic factor for RFS and OS at least as strong as Breslow thickness. Conclusion: Expression of Cyclin E and loss of p16 appeared to be significant prognostic biomarkers, with Cyclin E being prognostically at least as strong as Breslow thickness regarding RFS and OS. Both biomarkers may potentially be used to improve stratification of individual patient’s risk of recurrent disease and survival, and subsequently optimize adjuvant treatment.
The increasing incidence rate of cutaneous melanoma (CM) worldwide, combined with the observed higher mortality risk compared with other skin cancers, render the accurate diagnosis, prognosis and treatment of this malignancy of outmost importance (1). Based on recent data, it is estimated that in the United States over 100,000 people are annually diagnosed with CM and 8,290 individuals die of this disease (2). Recent global statistics of 185 countries demonstrate that melanoma ranks 17th among the most frequent malignancies with more than 330,000 new cases per year and 22nd regarding mortality due to malignant disease with almost 60,000 deaths per year (3).
A variety of genetic and environmental factors have been incriminated for the malignant transformation of melanocytes and the development of melanoma (4). But, despite the novel advances in molecular analysis, genomics and cancer biology technologies, we are far away from completely elucidating the molecular mechanisms, responsible for inhibition of apoptosis and tumor suppressor inactivation, which lead to melanoma pathogenesis and progression (5). As a consequence, melanoma still remains an unpredictable disease, and the classical methods have evidenced accuracy failure in individual patient categorization and prognostication of mortality risk (6). More precisely, the American Joint Committee on Cancer (AJCC) staging system, which is used worldwide for the stratification of patients with CM in risk categories and is based on evaluation of Breslow thickness, ulceration, lymph node status and presence of distant metastases, is not ideal for predicting patients’ individual risk of disease recurrence and death from melanoma (7, 8). Large studies have demonstrated that patients suffering from CMs of the same stage, according to the AJCC classification, may present a totally different disease course (9). For example, in the subgroup of stage I patients (invasive melanoma with Breslow thickness ≤1 mm, without metastases), which comprise nowadays the majority of melanoma patients and for which wide local excision alone is regarded curative, 5% will develop metastatic disease and eventually die from that malignancy within 10 years of diagnosis (6, 9-11). Among patients with initially localized melanoma, patients with thin melanoma (stage I) comprise 34% of all melanoma-specific deaths (12). Hence, identification of individual patients with early-stage disease but at high-risk for metastatic recurrence is important to provide them adjuvant systemic treatment, which they should not have received according to their early disease stage.
Thus, it is extremely important to develop additional prognostic tools in order to predict better the likely course and probable outcomes of the melanoma disease in individual patients (13, 14). These prognostic melanoma biomarkers would be useful for the optimal stratification of patients into risk subgroups and for earlier identification of individuals at high risk for metastasis or recurrence (6, 9, 14). Through that process, personalized treatment of patients with CM could be offered, resulting in improved outcome and diminished healthcare costs (14).
Indeed, recent new laboratory technologies have produced an impressive increment of new possible prognostic biomarkers in CM (5). Mutations in specific gene loci, melanoma-associated antigens, circulating cell-free nucleic acids, adhesion molecules and soluble proteins are among the biomarkers that are under current investigation (9). Particularly, proteins with fundamental roles in regulating cell proliferation, cell growth and apoptosis have drawn the attention of many experts, and many studies are conducted for validation and standardization (13). Cyclins and cyclin-dependent kinase inhibitors, which are key players in the cell cycle regulation, have attracted great interest in the field of melanomagenesis (15).
Both Cyclin D1 and Cyclin E promote, through interaction with their catalytic partner cyclin-dependent kinases (CDKs) 4, 6 and 2, cell proliferation (16). These complexes facilitate the cellular transition beyond the G1/S restriction point via phosphorylation of the protein retinoblastoma (Rb), producing the functional inactivation of Rb, the release of E2F transcription factors and the initiation of DNA replication (16, 17). To the contrary, p16, member of the INK4a family of kinase inhibitors, is involved in the regulation of the cell cycle mainly by preventing the phosphorylation of Rb at the G1 phase through inhibition of CDK4 and CDK6 (18). As a result, the progression to the S phase of the cell cycle becomes postponed (18). Similarly, p21 and p27, two members of the CIP/KIP family of kinase inhibitors, are considered tumor suppressors as they mediate the block of cell proliferation by arresting the cell cycle progression through inhibition of CDK-cyclin complexes (19, 20). Nevertheless, both these proteins possess a contradicting role. More precisely, p21 regulates the transcription of several genes, including those encoding c-myc, E2F and MITF, promoting repair of damaged DNA and protecting via that action cells from apoptosis (19). Moreover, the translocation of p27 from the nucleus to cytoplasm insures the cellular motility and viability (20).
The purpose of this study was to investigate the expression of Cyclin D1, Cyclin E, p16, p21 and p27 in primary invasive CM and to assess their correlation with clinicopathological variables. Additionally, the impact of these proteins on the progression of melanoma and clinical outcomes was specifically analyzed, in order to evaluate their potential role as prognostic biomarkers.
Patients and Methods
Patient selection and data collection. The study cohort consisted of 59 consecutive patients diagnosed with invasive CM after diagnostic excisional biopsy at a tertiary hospital in Greece. Patients with distant metastases at diagnosis, those with a follow-up period of less than five years and those referred with a diagnostic biopsy performed elsewhere were not included in the study. Subsequently, wide local excision, sentinel lymph node biopsy and/or regional lymph node dissection and adjuvant treatment were applied according to the international guidelines that were then used. As a result, 14 patients received adjuvant immunotherapy after the surgical intervention, mainly nowadays obsolete interferon-α. The disease stage was classified according to the 8th edition of AJCC staging system. Follow-up and treatment of recurrent disease were practiced according to the international guidelines. All necessary epidemiological, clinical and imaging data were retrospectively obtained from medical records. The formalin-fixed, paraffin-embedded tissue blocks of the respective primary tumors were retrieved from the archive of the hospital’s pathology department. The study was approved by the hospital’s Ethical Committee (protocol number 7804).
Immunohistochemistry and immunoreactivity scoring. The representative 3-μm-thick sections of the lesions were deparaffinized in xylene for 60 min and rehydrated using descending graded concentrations of ethanol. The process of antigen retrieval was performed in a steamer containing ethylene diamine tetraacetic acid (EDTA) for 45 min and the activity of endogenous peroxidase was blocked using hydrogen peroxide for 15 min. Tris buffered saline (TBS) was the preferred buffer for reducing the nonspecific background staining, and afterwards the tissue samples were incubated with the antibodies following manufacturers’ instructions. The sections were washed in TBS buffer and 3, 3′-diaminobenzidine (DAB) was applied, yielding a dark brown reaction product. Before evaluation of immunoreactivity, the slides were rewashed, counterstained with hematoxylin and dehydrated through a graded ethanol series. Finally, Entellan (Agilent Technologies, Santa Clara, CA, USA) was used for the permanent mounting and storage of specimens. The clones, the sources and the dilutions of antibodies used in this study are depicted in Table I. The immunostaining of sections was analyzed using a semiquantitative process based on proportion of tumor cells showing unequivocal positive reaction as proposed by Sanki et al. (21-23). More precisely, a single blinded pathologist (MT) evaluated the expression of proteins by grading the percentage of positive cells and using a 0 to 3 scoring system: 0=absent (0% stained cells); 1=weak (<10% stained cells); 2=low-moderate (10-49% stained cells) and 3=strong (equal or more than 50% stained cells). In the cases of Cyclin D1, p21 and p27 only nuclear staining was reported (Figure 1A-E). To the contrary, nuclear and cytoplasmic staining was graded separately for p16 and Cyclin E (Figure 1A-E). Previous studies have suggested that cytoplasmic expression of p16 and Cyclin E is not merely nonspecific binding and likely represents cytoplasmic localization of the proteins with distinct and specific biological functions (21, 24, 25). Moreover, the reproducibility of the process was guaranteed by defining clear cut-off criteria for each marker based on the endogenous expression of proteins in non-tumoral cells (16, 24-26) (Table I).
Antibodies used for immunohistochemistry.
Images depicting the immunohistochemical staining pattern of each protein marker. A) 10% of tumor cells showing positive nuclear expression of Cyclin D1 (×400). B) Extensive nuclear and cytoplasmic expression of Cyclin E (×200). C) No expression of p27 in melanocytes (×200). D) Positive nuclear staining of p21 (×100). E) Extensive nuclear and cytoplasmic expression of p16 (×200).
Statistical methods. The percentage of stained cells reflecting the expression of proteins was graded using a continuous variable ranged from 0% to 100%. Two types of recoded schema were applied. The first type was based on the cut-off values of proteins’ expression providing two groups (no expression <cut-off value, expression >cut-off value). The second schema used the system of immunoreactivity scoring and two distinct groups were created: <50% (low expression) and ≥50% (high expression). Death, local relapse, in transit metastases, lymph node metastases and distant metastases are expressed as binary outcomes (0=No, 1=Yes). Overall survival (OS) and relapse free survival (RFS) are expressed in months. Independent samples t-test and the corresponding non parametric Mann–Whitney test were used for two group’s comparison. When more than two groups were studied one-way ANOVA followed by the non-parametric Kruskal–Wallis test were applied. Pearson’s chi-square test was assessed to examine possible associations between two discrete variables. Box plots and Kaplan–Meier plots were applied for the graphical representation of the data. The log-rank test (Mantel–Cox) was assessed to compare the survival distribution between different groups of protein expression. Moreover, a Cox proportional hazards regression analysis was applied to determine the impact of different variables on survival time. IBM SPSS Statistics 24.0 (IBM, Armonk, NY, USA) was used for statistical analysis and a=0.05 limit was set as a level of acceptance.
Results
Cohort and tumor characteristics. The study population consisted of 59 newly diagnosed cases with clinically locoregional CM. Thirty-three of patients (55.9%) were male and the mean age was 57.3±16.5 years ranging from 19 to 93 years old. The clinical and pathological features of the cohort are presented in Table II and Table III. The mean Breslow thickness was 2.4±2.3 mm with a median of 1.4 mm ranging from 0.3-13.0 mm, and most of the patients had a 2-4 mm melanoma thickness (16, 27.1%). Based on TNM classification, 23 of cases (39.0%) were categorized in T1 stage and eight of them (13.6%) in T4 stage. Moreover, the vast majority of patients (42, 71.2%) were classified in N0 stage, as no lymph node involvement was detected. During the median follow-up period of 77 months (10 to 128 months), 16 patients (27.1%) relapsed and all ultimately died due to distant metastases.
Pathological characteristics of 59 cases with cutaneous melanoma.
Development of recurrent disease and death in the 59 patients with cutaneous melanoma.
Expression of proteins. The expression of each protein, corresponding to the percentage of stained cells, is depicted in Figure 2A-G. The most frequent (mode) expression of Cyclin D1 in the nucleus was observed at 80% (20 cases, 33.9%), whereas the mean expression was estimated at 58.8±23.5% with a median of 60%. Respectively, the expression of p27 in the nucleus showed a mean of 56.3±22.4% with a median of 60.0%. There was only one case (1.7%) without expression of p27 and the mode of expression was found at 70% (15 cases, 25.4%). As regards Cyclin E, nuclear and cytoplasmic staining was graded separately. The most common expression pattern of Cyclin E in the cytoplasm was observed in 43 cases (72.9%) with a mean value of 73.4±16.5% and a median value of 80%. To the contrary, 48 patients (81.4%) exhibited negative, under the cut-off limits, expression of Cyclin E in the nucleus. More precisely, the mean value of expression in positive samples was 5.1±13.8% and the distribution ranged from 10 to 60%. Regarding p21, 13 cases (22.0%) did not have any expression in the nucleus. Among p21 expressing samples, the most frequent expression was observed at 50% (9 cases, 15.3%), whereas the mean expression was measured at 32.2%±24.7% with a median of 30%. Similarly to Cyclin E, the expression of p16 in the nucleus and cytoplasm was estimated separately. There was no expression of p16 in the cytoplasm in 40 patients (67.8%). The mean expression was calculated at 20.9%±31.0% with a median of 0.0, but when only positive samples were included in the analysis, the most frequent expression was observed at 70% (7 cases, 11.9%). Furthermore, 36 patients (61.0%) did not present expression of p16 in the nucleus. The cases with expression over the threshold possessed a mean expression of 32.5±28.7% with a median of 20% and the most frequent expression was observed at 70% (9 cases, 15.3%).
Box plots demonstrating the expression of proteins in the cytoplasm and nucleus. Expression of (A) Cyclin D1 in the nucleus, (B) p27 in the nucleus, (C) cyclin E in the cytoplasm, (D) Cyclin E in the nucleus, (E) p21 in the nucleus, (F) p16 in the cytoplasm and (G) p16 in the nucleus.
Correlation of proteins’ expression with clinicopathological features. Age and sex were analysed in order to clarify their effect on protein expression. Neither of these factors possessed significant impact in any of the measured proteins. On the contrary, Breslow thickness was significantly correlated with the expression of Cyclin E in the nucleus (p=0.017), p16 in the cytoplasm (p=0.004) and p16 in the nucleus (p=0.001). More precisely, there was positive association between the concentration of Cyclin E in the nucleus and tumor thickness. An opposite pattern of interaction was found for p16. The level of p16 expression was inversely proportional with Breslow thickness (Figure 3A and B). Similarly, the existence of ulceration was inversely correlated with the expression of p16 in the nucleus (p=0.035) (Figure 4).
Boxplots of p16 expression between two categories of Breslow thickness. Expression of p16 (A) in the cytoplasm and (B) in the nucleus.
Boxplot depicting the relationship between ulceration and expression of p16 in the nucleus.
Survival evaluation. As regards death of disease, the statistical analysis revealed significant association among that parameter and expression of Cyclin E in the nucleus (p=0.001), p16 in the cytoplasm (p=0.046) and p16 in the nucleus (p=0.025) (Table IV). Indeed, patients who died during the observational period exhibited a higher mean level of Cyclin E in the nucleus (16.3±23.1%) versus those who were alive in the same period (0.9±2.9%). In contrast, a pattern of decreased expression of p16 in the cytoplasm (p=0.046) and p16 in the nucleus (p=0.025) was found in patients who died from the disease. The mean expression in alive patients was 29.5±32.4% for p16 in the cytoplasm and 17.5±16.9% for p16 in the nucleus, significantly higher than those in cases with relapse of melanoma and death (10.6±17.7% and 17.5±16.9% correspondingly).
Correlation of proteins’ expression and death of patients.
According to Kaplan–Meier analysis, shorter RFS and OS were significantly related to loss of p16 and elevated expression of Cyclin E in the nucleus (Figure 5A-G and Figure 6A-G). More comprehensively, expression of Cyclin E in the nucleus equal or over 50% (high expression) was correlated with low mean OS (17.3, 95%CI=10.6-23.9 months). To the contrary, survival time was elevated in low (<50%) levels of protein (107.4, 95%CI=96.8-117.9 months), and that difference was statistically significant [log-rank χ2 (1)=39,214, p<0.001]. The same pattern was observed with RFS. Mean RFS was 104.6 (95%CI=92.8-116.5 months) in patients with low levels of Cyclin E in the nucleus versus 8.8 (95%CI=6.6-10.9 months) in cases with high expression of the protein [log-rank χ2 (1)=29,744, p<0.001]. An opposite pattern was observed for p16 [log-rank χ2 (1)=6,239, p=0.012 for cytoplasm and log-rank χ2 (1)=9.381, p=0.002 for nucleus]. Significantly higher mean OS was found in cases with p16 expression equal or more than 50% of in the cytoplasm (122.7, 95%CI=112.5-132.8 months) and in the nucleus (123.6, 95%CI=115.2-132.0 months) compared to those with absence of protein (90.5, 95%CI=75.5-105.4 months and 86.4, 95%CI=70.3-102.5 months respectively). The same results were detected regarding RFS [log-rank χ2 (1)=6,440, p=0.011 for p16 in the cytoplasm and log-rank χ2 (1)=9,577, p=0.002 for p16 in the nucleus].
Kaplan–Meier plots depicting the correlation of recurrence-free survival (RFS) with proteins. Correlation of RFS with (A) Cyclin D1 in the nucleus, (B) p27 in the nucleus, (C) cyclin E in the cytoplasm, (D) Cyclin E in the nucleus, (E) p21 in the nucleus, (F) p16 in the cytoplasm and (G) p16 in the nucleus.
Kaplan–Meier plots demonstrating the correlation of overall survival (OS) with proteins. Correlation of OS with (A) Cyclin D1 in the nucleus, (B) p27 in the nucleus, (C) cyclin E in the cytoplasm, (D) Cyclin E in the nucleus, (E) p21 in the nucleus, (F) p16 in the cytoplasm and (G) p16 in the nucleus.
Unfortunately, the number of patients was too small to perform subgroup analysis with regard to the additional prognostic value of p16 and Cyclin E within various T, N or stage groups in order to optimize risk classification. The Kaplan–Meier analysis was followed by a Cox proportional hazards regression model with forward selection in order to clarify the impact of Breslow thickness and expression of cycle regulators Cyclin D1, Cyclin E, p16, p21, and p27 on survival time. The results revealed a significant association of Breslow depth and high levels of Cyclin E in the nucleus with OS and RFS (Table V).
Cox proportional hazards regression model.
Discussion
The malignant transformation of melanocytes and development of CM is considered the result of interactions between genetic and environmental factors leading to the acquisition of sequential alterations in pathways controlling and regulating crucial cell functions (4). In this study, we evaluated the immunohistochemical expression of the cell cycle regulators Cyclin D1, Cyclin E, p16, p21, and p27 in 59 patients with primary invasive locoregional CM.
In agreement with previous studies, the results of our survey dictate that Cyclin E is highly expressed in primary melanomas and a positive correlation between the levels of the protein in the nucleus and Breslow thickness is apparent (27, 28). There is accumulation of this protein in the cytoplasm (80%) and, despite the fact that Cyclin E was no detected in the nucleus in 48 patients due to natural translocation to the cytoplasm, the level of its expression in the nucleus proved to be associated with higher Breslow depth and higher tumor stage. That interaction could be explained by the fact that Cyclin E over-expression leads to enhanced CDK2 activity, cell cycle transition beyond the G1/S restriction point and cellular proliferation (29). Moreover, deregulation of Cyclin E has been also observed in a broad spectrum of human malignancies, including breast cancer, colorectal carcinoma and osteosarcoma (30).
The tumor suppressor p16 has attracted great interest in the field of melanocytic pathology and several studies have analyzed its role in the development of CM (31). There is a consensus among scientists that the level of p16 is high in benign melanocytic lesions and becomes diminished or even eliminated in dysplastic nevi, in-situ melanomas and invasive primary tumors (32-34). In accordance with these studies, this protein was not detected in more than half of our patients with invasive melanoma. More precisely, 40 and 36 patients exhibited no expression of p16 in the cytoplasm and nucleus, respectively. Furthermore, even in the cases expressing p16, the levels of the protein proved to be inversely proportional with Breslow thickness. This suggests that the CDKN2A gene which encodes p16 may be mutated or disordered resulting in decrease in protein’s concentration, loss of normal p16 activity and progression of melanoma (35). The association between tumor invasiveness and down regulation of p16 can be also verified by the interaction of ulceration and expression of protein in the nucleus that was detected in our study. Straume et al. first reported correlation between deregulation of nuclear p16 and presence of ulceration in patients with vertical growth phase melanoma, an observation that was subsequently confirmed (36).
Beyond the relationship with certain clinicopathological variables, it is important to investigate the impact of these cell cycle regulators on clinical outcomes of patients with CM and to evaluate their potential role as prognostic biomarkers. According to analysis of our data, shorter OS and RFS were significantly related to loss of p16 and elevated expression of Cyclin E in the nucleus. As regards p16, both Straume et al. and Bachman et al. reported that the loss of p16 expression in primary nodular melanomas could independently predict decreased patient survival and was correlated with the presence of vascular invasion (36, 37). Similarly, Alonso et al. who analyzed 165 melanomas with different histological progression phases, concluded that p16 was an independent predictor of survival in patients with vertical growth phase melanoma, as the risk of death was 8 times greater when there was loss of protein expression (24). Moreover, the low levels of p16 expression in primary melanomas were associated with an increased risk of metastasis to lymph nodes and relapse of the disease after surgical excision (25, 38). Furthermore, Lade-Keller et al. performed immunohistochemical analysis of 355 primary melanomas and reported that the absence of p16 expression could predict overall relapse-free and distant metastasis-free survival, independently of Breslow thickness, ulceration and tumor stage (39). However, there are research studies that question the usefulness of p16 as prognostic biomarker. Sanki et al. and Fauri et al. who examined cutaneous melanomas specimens reported that the loss of expression of p16 could not be correlated with OS, recurrence or metastasis, and for that reason it should not be used for prognosis (21, 40). In contrast to p16 and the debate among scientists regarding its usefulness as prognostic biomarker, there is only one study, to the best of our knowledge, which investigated the impact of Cyclin E on the survival of patients with CM. More specifically, Alonso et al. reported that shorter OS was significantly related to over-expression of Cyclin E (24). Similarly, in our study, high Cyclin E levels were associated with poor RFS and OS. Moreover, Cyclin E was a prognostic factor for both survival parameters as strong as Breslow thickness, which is known as the strongest prognostic factor in localized CM. Larger studies should confirm the definite prognostic value of Cyclin E and assess whether indeed this biomarker might be an even stronger prognostic parameter than Breslow thickness. Moreover, subgroup analysis in larger cohorts should assess the prognostic value of p16 and Cyclin E in various Breslow thickness and nodal stages, allowing for more optimal risk stratification than the current TNM system. Most recently, acknowledging the need for better disease staging and appreciating the role of molecular biology, FIGO staging of endometrial cancer, which is solely based on clinicopathological features, has incorporated molecular features as p53 expression and POLE mutations (41).
Future research could explore the potential prognostic value of combining these two proteins with other biological factors linked to the development and progression of CM. For instance, genetic variations in CDKN2A (42) and factors associated with poorer prognosis, such as the over-expression of glucose transporters (43), might provide valuable insights. Beyond assessing the prognostic roles of these cell cycle regulators individually, precision medicine approaches could be developed to target CM patients based on their expression levels. For CM patients with low p16 expression, therapeutic strategies aimed at addressing CDKN2A loss – such as CDK4/6 inhibitors or more advanced approaches – are under investigation (44). Similarly, targeting Cyclin E could present an appealing therapeutic option for CM cases characterized by high Cyclin E expression (45).
Study limitations. The study cohort consisted of 59 patients with available tissue samples and the patients were initially treated more than five years before, due to the need for sufficient follow-up for outcome analyses as well as immunochemistry measurements and analyses. During this time period, advances in adjuvant systemic treatment, treatment for metastatic disease and supportive care may have affected RFS and OS. Furthermore, the sample size of our research is small and even quite large differences may not reach statistical significance. Nevertheless, the number of subjects included is comparable to other similar published studies. New research projects including large number of patients are necessary in order to validate the results of this study and understand the development and progression of CM.
Conclusion
In this melanoma cohort study, the correlation of cell cycle regulators Cyclin D1, Cyclin E, p16, p21, and p27 with known prognostic parameters, RFS and OS was evaluated. The level of expression of Cyclin E and p16 were strongly correlated with standard clinicopathological features, such as Breslow thickness, melanoma ulceration and tumor stage. Moreover, the univariate survival analysis revealed the association of these molecules with OS and RFS. Furthermore, expression of Cyclin E in the nucleus proved to be a prognostic factor for both survival parameters as strong as Breslow thickness, which is known as the strongest prognostic factor in localized CM. Larger studies should confirm the definite prognostic value of p16 and Cyclin E in CM and assess whether indeed Cyclin E might be a prognostic biomarker even stronger than Breslow thickness. Moreover, subgroup analysis in larger cohorts should assess the prognostic value of p16 and Cyclin E in various subgroups, allowing for more adequate risk stratification and subsequently for optimal adjuvant treatment. These biomarkers that identify more aggressive disease may also be useful as targets for precision medicine in the future.
Footnotes
Authors’ Contributions
IG, MT, and EB designed the study. IG, GD, and EM performed all experiments. IG, MT, AA, and EB analyzed the results. IG, MT, and EB contributed to the interpretation of the results. IG and EB wrote the manuscript with contribution of GS, DM, SKK, and KK. All Authors have read and approved the final manuscript.
Conflicts of Interest
The Authors do not have any conflicts of interest to declare in relation to this article.
Funding
None.
- Received November 25, 2024.
- Revision received December 9, 2024.
- Accepted December 16, 2024.
- Copyright © 2025 International Institute of Anticancer Research (Dr. George J. Delinasios), All rights reserved.
This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY-NC-ND) 4.0 international license (https://creativecommons.org/licenses/by-nc-nd/4.0).