Abstract
Background/Aim: Non-invasive circulating tumor biomarkers in liquid biopsy, such as microRNAs (miRNA), provide for better personalization of treatment strategies. The aim of our study was to assess the prognosis of patients with melanoma undergoing tumor resection with curative intent based on analysis of selected circulating miRNAs. Patients and Methods: A total of 22 patients with stage I to III melanoma were enrolled into this prospective study. Plasma samples were obtained pre-surgery and early post-surgery from peripheral blood draws. A panel of 23 candidate miRNAs was designed and expression of miRNAs were analyzed by reverse transcription-quantitative polymerase chain reaction with exogenous reference control cel-miR-39-3p. Results: Higher preoperative expression levels of miR-99a (p=0.008), miR-320 (p=0.009), miR-1908 (p=0.001), miR-494 (p=0.018) and miR-4487 (p=0.048) were associated with a shorter disease-free interval. Similarly, higher preoperative plasma levels of miR-99a (p=0.017), miR-221 (p=0.026), miR-320 (p=0.016), miR-494 (p=0.009), miR-1260 (p=0.026) and miR-1908 (p=0.024) were associated with worse overall survival. No significant differences between pre- and postoperative plasma miRNA levels were observed. Conclusion: Liquid biopsy is a minimally-invasive approach which can lead to a better understanding of cancer behavior and offers the possibility of precise patient prognosis, allowing selection of the most appropriate treatment. Our study showed that preoperative plasma levels of miR-99a, miR-221, miR-320, miR-494, miR-1908 and miR-4487 were associated with disease-free interval and overall survival of patients with early-stage melanoma. This approach may help in decision-making about the appropriateness of modern adjuvant treatment administration in patients with resectable melanoma.
Melanoma is the deadliest form of skin cancer (1), associated with aggressive tumor behavior and therapy resistance in advanced disease in a significant number of patients (2). With an incidence rate of one in 50 individuals, in the Western population, melanoma has been noted as one of the rapidly increasing types of cancer, however, the mortality rate remains stable based on recent treatment advancement (3, 4). It is the fifth and the seventh most common cancer in males and females, respectively, being most frequent in the sixth decade of life (5). Melanoma may occur in different body parts. It is majorly recognized as a skin cancer originating from epidermal melanocytes (6, 7). Cutaneous melanoma, mucosal melanoma, and the ocular disease are the main three subtypes of this cancer (8) associated with various risk factors including UV exposure, pre-existing nevi, sunburn injuries, fair skin, familial history of the disease, and clinical conditions including familial atypical multiple mole-melanoma syndrome, and melanoma-astrocytoma syndrome (7, 9, 10).
Despite advances in adjuvant treatment [targeted therapy by B-Raf proto-oncogene serine/threonine kinase (BRAF) and mitogen-activated protein kinase (MEK) inhibitors], as well as immunotherapy with monoclonal antibodies against programmed cell death protein 1 (PD1) and its ligand PDL1 for resectable melanoma, there are still no relevant prognostic and predictive biomarkers. Biomarkers which are identified based on a liquid biopsy are a promising way to better select patients appropriate for these costly treatment strategies with potentially serious adverse effects (11-13).
Liquid biopsy-based analysis of nucleic acids (DNA, RNA) (14-19) in various body fluids has revolutionized cancer therapeutic approaches and has led to more accurate treatment strategies (20-22). The analysis of circulating tumor DNA (ctDNA) appears to be promising for the prediction of targeted treatment efficacy (18, 23). However, due to intertumoral heterogeneity, mutation profiles vary among patients (24). One of the ways to overcome this challenge is the assessment of prognosis based on a ‘melanoma specific’ microRNA (miRNA) profile (25).
miRNAs are significant post-transcriptional gene regulators, with an average length of 22 nucleotides, targeting specific messenger RNAs, which may play an important role in cancer formation by regulation of the expression of oncogenes and tumor-suppressor genes (26, 27). nilRNAs regulate vital cell behaviors, such as survival, apoptosis, and proliferation, through various mechanisms. One of the main mechanisms through which miRNAs regulate gene expressions is through targeting the 3′ untranslated region of messenger RNA (28, 29). Studying patterns of miRNA alteration in cancer may provide significant biomarkers of cancer initiation and therapy responses (30); and assist in better understanding of the molecular pathways in tumor progression, while proposing novel therapeutic targets (8).
The aim of our study was to assess the prognosis of patients with melanoma undergoing tumor resection with curative intent based on analysis of selected circulating miRNAs. The prediction of prognosis may help to better select appropriate patients for modern adjuvant therapy.
Patients and Methods
Patients. We analyzed pre- and postoperative plasma samples from 22 patients with stages I, II, and III melanoma who underwent tumor resection at the Department of plastic surgery, University Hospital Pilsen between the years 2017 to 2022. The study was approved by the Institutional Review Board and local Ethics Committee of the University Hospital in Pilsen (no. 04112019). Patients enrolled in this study provided a signed informed-consent form for using their blood samples for analyzing prognostic factors. The melanoma diagnosis was verified by the dermatopathologists according to the American Joint Committee of Cancer, Melanoma of the Skin staging (eighth edition) (31). Clinicopathological features of our study cohort are summarized in Table I.
Demographic information for the participants in our study (n=22).
Blood samples. After collecting specimens of peripheral blood from the median cubital vein in K3EDTA Vacutainer tubes (Greiner Bio-One, Kremsmünster, Austria), we centrifuged 6 ml of whole blood, at 1,370×g, for 10 minutes at 4°C. Then plasma samples were separated and any cell debris was removed, as a standard procedure used for cell-free miRNA assessment. Until performing the RNA isolation process, plasma samples were stored at −80°C. A total of 44 plasma samples from 22 patients were analyzed.
miRNA assessment. Based on previously performed studies, we identified 23 suitable miRNA candidates for this research that are summarized in the Table II. The quantitative analysis of these miRNAs was performed using reverse transcription-quantitative polymerase chain reaction (RT-qPCR) of plasma samples obtained before and immediately after surgery.
Candidate microRNAs were assessed in this study based on previously reported analyses.
miRNA isolation. RNA templates were isolated from 200 μl of plasma using Qiagen miRNeasy Serum/Plasma kit (Qiagen, Hilden, Germany), according to the manufacturer’s manual. Next, the total RNA concentrations were quantified using a Nanodrop ND1000 spectrophotometer (Thermo Fisher Scientific, Foster City, CA, USA), and samples with an RNA concentration of ≥10 μl/ml, were used for further assessments. Samples were stored at −80°C until amplification.
Quantitative assessment: RNA was reverse transcribed using T100™ Thermal Cycler apparatus (Bio Rad, Hercules, CA, USA) in a 15 μl reaction containing 11.5 μl of master mix (TaqMan™ MicroRNA Reverse Transcription Kit containing water, dNTP, inhibitor, TaqMan™ Universal Master Mix II, no uracyl N-glycosylase and reverse transcriptase), 2.5 μl of miRNA-specific primer (Applied Biosystems; Thermo Fisher Scientific, Inc., Waltham, MA, USA), followed by quantitative estimation of the selected 23 miRNAs (Table II), using RT-qPCR in TaqMan™ MicroRNA assays (Applied Biosystems; Thermo Fisher Scientific, Inc.) in technical duplicates on a Stratagene Mx3005P (Agilent Technologies, Santa Clara, CA, USA). To obtain plasma levels of the miRNAs of interest the deltaCt approach was used. The results of plasma miRNA expression are presented as relative expression values calculated as 2−(Ct of miRNA of interest − Ct of normalizer). From currently used approaches for normalization (52), exogenous spike-in control (reference) cel-miR-39-3p was used as a normalizer for miRNAs circulating in plasma (53, 54).
Statistical analysis. The statistical analysis was performed using SPSS 22 software (IBM Corp., Armonk, NY, USA). The statistical significance was set at p≤0.05, and Wilcoxon rank-sum test was used for identifying significant changes in preoperative and postoperative results; Kaplan-Meier methods and log-rank test were used to perform survival analysis. As a cut-off value, the median normalized expression of each miRNA was selected.
Results
Pre-and post-surgery miRNA evaluation. The values of miRNA expression were normalized based on the exogenous spike-in control cel-miR-39-3p. We compared the miRNA expression levels between plasma samples taken pre- and post surgery using the Wilcoxon rank-sum test. No significant differences were observed; detailed results are shown in Table III.
Median concentration of the targeted microRNAs (miRNAs) in pre- and postoperative blood samples. Results were normalized, based on the exogenous spike-in control cel-miR-39-3p.
Survival analysis. We analyzed the relationship of miRNA expression with the disease-free interval (DFI) and overall survival (OS). The median follow-up was 23.5 months in the case of DFI and 26 months in the case of OS. We found that there was significant relationship between DFI and preoperative plasma level of miR-320 (p=0.009), miR-494 (p=0.018), miR-1908 (p=0.001), miR-4487 (p=0.048), and miR-99a (p=0.008): higher expression was associated with unfavorable prognosis. Detailed results are presented in Figure 1 and Table IV.
Kaplan-Meier curves for the disease-free interval according to the preoperative plasma level of miRNAs reportedly associated with disease-free survival in patients with melanoma. Five miRNAs are presented (p<0.05), namely miR-320, miR-494, miR-1908, miR-4487, and miR-99a.
Association of preoperative microRNA (miRNA) levels with the disease-free interval (DFI).
Moreover, a higher preoperative plasma level of miR-221 (p=0.026), miR-320 (p=0.016), miR-494 (p=0.009), miR-1260 (p=0.026), miR-1908 (p=0.024), and miR-99a (p=0.017) were associated with unfavorable OS. Detailed results are presented in Figure 2 and Table V.
Kaplan-Meier survival curves for preoperative plasma level of six miRNAs associated with overall survival in melanoma. Six miRNAs are presented (p<0.05), namely miR-221, miR-320, miR-494, miR-1260, miR-1908, and miR-99a.
Association of preoperative microRNA (miRNA) levels with overall survival (OS).
Discussion
The aim of our study was to determine the prognostic value of 23 selected melanoma-specific miRNAs as biomarkers for potential decision-making in adjuvant treatment of melanoma. We used pre-and post surgery plasma samples, in a total of 22 patients with stage I-III melanoma (77% stages I and II). We observed a significant relationship between higher preoperative plasma levels of miR-1908, miR-99a, miR-320, miR-494, and miR-4487 and a shorter DFI. In addition, we observed an association between higher preoperative levels of miR-494, miR-320, miR-99a, miR-1908, miR-1260, and miR-221 and worse OS. We suggest that the preoperative plasma levels of these miRNAs may help predict disease outcome and identify patients with a higher risk of disease relapse; those patients may be appropriate candidates for modern adjuvant treatment.
We did not observe significant changes between pre- and postoperative miRNA concentrations. We assume that, since the degradation of miRNAs is typically slow, lasting over 10 hours on average (55), it could have a significant impact on the results, preventing achieving real-time meaningful differences in pre- and postoperative samples.
In the past two decades, the emergence of novel therapeutic strategies, such as targeted therapy using BRAF and MEK inhibitors, as well as immunotherapy with anti-PD1/PDL1, have significantly improved disease outcomes in patients with melanoma (22, 56). In the case of targeted therapy, dabrafenib and vemurafenib were designed to treat patients with melanoma associated with a mutated version of the BRAF gene (57, 58). Dabrafenib acts as an ATP competitor for BRAF oncoprotein, and can be prescribed both as a single agent and in combination with trametinib, an inhibitor of a downstream kinase MEK (58). Vemurafenib is a dose-dependent therapy agent which induces cell-death by blocking BRAF oncoprotein, with better response when used in combination with MEK inhibitors (58).
Immunotherapy allows modulation of the anticancer immune response based on interaction with cell-surface proteins involved in immune checkpoint regulation, such as cytotoxic T-lymphocyte associated protein-4 (CD152) and PD1 (59). Immune checkpoint inhibitors have shown promising results in both non-metastatic and metastatic cancers, regardless of the mutated gene and with longer-lasting effects in some patients (60, 61). Ipilimumab was the first therapeutic agent in this group which improved OS by targeting cytotoxic T-lymphocyte associated protein-4 (62, 63) however, immune-related adverse events were reported (62). Nivolumab and pembrolizumab were designed specifically to target PD1 on the cell surface and increased OS to more than 60 months in patients with metastatic melanoma (7, 8, 43, 64). In individuals who demonstrated resistance to BRAF inhibitors, good responses were reported when nivolumab, pembrolizumab and ipilimumab/nivolumab combinational therapies were used (5, 9, 44, 65), although combination therapies may also increase the risk of adverse-events (45, 66).
Patients with resectable melanoma may also benefit from novel therapeutic approaches such as targeted therapy and immunotherapy (11-13). Biomarker assays based on liquid biopsy are a promising way to better select patients appropriate for these costly treatment strategies which have potentially serious adverse effects.
Liquid biopsy is highly beneficial both for the patients and the medical team, by overcoming the limitations and the potential hazards of tissue biopsies (6, 67-71). Due to some limitations/challenges facing the analysis of circulating tumor cells (such as their limited release into the bloodstream in early stages) and ctDNA (potential intertumor heterogeneity and low mutation concentration), analysis of the expression of circulating miRNAs may complement the assessment of prognosis of patients with early-stage melanoma (72-74). miRNAs play a significant role in melanoma (75, 76). Alterations in some miRNAs indicate cancer initiation as well as formation of a metastasis (77). There are results showing that miRNA assessments effectively estimated survival rates, more reliable compared to serum S100B and lactate dehydrogenase levels (p<0.001) (44). Moreover, studies indicate the association of miRNA with melanoma recurrence (78), disease prognosis, tumor invasion (75), and stage of the disease (79). miRNAs have also been shown to be potential therapeutic targets, and by down-regulating/up-regulating certain miRNAs, tumor development could be prevented and disease progression even be reversed (29, 80, 81).
Similarly to our results, Kanemaru et al. (82) reported that a higher level of circulating miR-221 in patients with melanoma was associated with worse prognosis. Li et al. (83) also showed that serum samples from patients with melanoma contained higher concentrations of miR-221 compared to the control group (p<0.0001), and was associated with lower 5-year OS (p=0.018) and an increased risk for disease relapse (p=0.008). These clinical findings reflect the pathophysiological effects of miR-221 in melanoma formation. miR-221 causes disease progression by down-regulating c-KIT receptor and p27KIP1, while targeting TRK-fused gene, which is a tumor-suppressor gene (84).
Despite our results indicating the association of higher preoperative miR-320 levels with worse OS and DFI in melanoma, other studies which commonly focused on other cancer types, indicated the tumor-suppressive role of this miRNA (85-87). It has been reported that miR-320 prevents cell migration, tumor invasion, and cell proliferation by down-regulating epithelial–mesenchymal transition and the phosphatidylinositol-4,5-bisphosphate 3-kinase (PI3K)/AKT serine/threonine kinase 1 (AKT) signaling pathway (87). Latchana et al. (32) showed that miR-494 was overexpressed in advanced melanoma; similarly, we showed the relationship of miR-494 to OS and DFI. One of the well-known oncogenic mechanisms of miR-494 was described as the activation of the PI3K/AKT signaling pathway (88).
In the case of miR-1260, Sand et al. (89) observed the up-regulation of this miRNA in melanoma tumor tissue compared to benign nevi, which our results are in accordance with.
We revealed the relationship of miR-99a overexpression to OS and DFI in our study. Nevertheless, this miRNA has not been fully studied in patients with melanoma. In other cancer types, it has been mentioned that miR-99a acts as an oncogene by targeting CTD (carboxy-terminal domain, RNA polymerase II, polypeptide A) small phosphatase-like and Tribbles homolog 2 (Drosophila) tumor-suppressor genes, which can cause disease development (90). However, Zhang et al. described an opposite effect in that miR-99a was able to prevent tumor progression by inhibiting mammalian target of rapamycin complexes 1 and 2 (91).
Our results showed that miR-1908 overexpression was significantly associated with worse prognosis in patients with melanoma (DFI and OS). This miRNA is highly expressed in human embryonic stem cells and also abnormally elevated in cancer (92, 93). The regulatory effect of miR-1908 on various cell functions includes apoptosis, cell proliferation, and motility. That is due to its mediating role in important cell signaling pathways such as transforming growth factor-β, PI3K/AKT, and mitogen-activated protein kinase pathways (94). Pencheva et al. (42), based on an in-vivo study, described increased miR-1908 expression in melanoma cancer cells. Based on that study, miR-1908 targets heat-shock factor DNAJ heat-shock protein family member A4 and metabolic gene apolipoprotein E. Cancer-secreted apolipoprotein E suppresses invasion and metastatic endothelial recruitment by engaging melanoma cell low-density lipoprotein receptor-related proteins 1 and 8 (42). On the other hand, Xu et al. reported that miR-1908 expression levels were lower in melanoma tissue compared to adjacent tissue (43).
Regarding the role of miR-4487 in cancer progression, studies show the association of increased miR-4487 level and reduced apoptosis in non-small-cell lung cancer cells (95), although the mechanism involved has not yet been fully revealed. However, we also observed an oncogenic effect of this miRNA in patients with melanoma.
However, there were significant limitations that should be mentioned. These mainly include the limited number of patients in our study group and the lack of consensus in normalization of plasma miRNA expression. RT-qPCR is extremely useful in targeted analysis of the level of miRNAs, present in very low concentrations, while being cost-friendly and having a short turnaround time, although it is limited to targeted miRNAs only and cannot report the whole transcriptome profile. Moreover, the use of our cut-off values of plasma miRNA expression in other studies might be limited.
Conclusion
Liquid biopsy and assay of circulating biomarkers provides the possibility to assess patients by a minimally-invasive approach, which may lead to a better understanding of cancer behavior. Our study showed that assessment of preoperative plasma levels of miR-99a, miR-221, miR-320, miR-494, miR-1908 and miR-4487 were associated with DFI and OS of patients with early-stage melanoma. This approach may help in decision-making about the appropriateness of modern adjuvant treatment administration in patients with resectable melanoma.
Acknowledgements
This work was supported by LTAUSA19080 program, INTER-EXCELLENCE, INTER-ACTION, Ministry of Education, Youth and Sports of the Czech Republic; by Cooperation Programme, research area MED/DIAG, Charles University; by a grant of the Ministry of Health of the Czech Republic-Conceptual Development of Research Organization (Faculty Hospital in Pilsen-FNPl, 00669806).
Footnotes
Authors’ Contributions
Conceptualization: J.P. and M.P.; methodology: M.S.B., K.H., T.K., K.P., D.S., R.N., J.P. and M.P.; validation: J.P. and M.P.; formal analysis: M.S.B., J.P., M.S., R.N. and M.P.; investigation: M.S.B., J.P., I.T., V.W., T.F., K.P., K.H., T.K., D.S and M.P.; data curation: I.T., V.W., T.F. and K.P.; writing-original draft preparation: M.S.B, J.P. and M.P.; writing-review and editing: J.P. and M.P.; supervision: J.P. and M.P.; project administration: I.T. and J.P.; funding acquisition: J.P. All Authors read and agreed to the published version of the article.
Conflicts of Interest
The Authors declare no conflicts of interest.
- Received November 26, 2022.
- Revision received December 16, 2022.
- Accepted December 20, 2022.
- Copyright © 2023 International Institute of Anticancer Research (Dr. George J. Delinasios), All rights reserved.
References
In this issue
Jump to section
Related Articles
Cited By...
- No citing articles found.