How detection of epigenetic alterations of blood-borne DNA could improve melanoma diagnosis

Abstract

Detection in blood of the genetic and epigenetic changes present in metastatic cancers is opening up new possibilities in molecular diagnostics. A number of methodological and clinical issues await resolution before serum epigenetic biomarkers can be considered a routine part of the management of melanoma patients following primary excision. However, there is every possibility that blood testing for the presence of methylated DNA will become an integral part of the clinical follow-up of such patients. The ability to identify patients with subclinical (asymptomatic) metastatic melanoma, combined with new, highly active targeted and immunomodulatory agents, may lead to further improvements in outcomes for this patient population.

Keywords::

• blood borne
• epigenetics
• melanoma
• metastasis
• serum

Malignant melanoma is an aggressive skin cancer, which, in its advanced forms, has (until recently) been associated with a uniformly poor outcome. Although thin primary cutaneous melanomas have an excellent prognosis when surgically excised, patients with deeper lesions have an incrementally increasing risk of recurrence, metastasis, and death. In addition to those patients in whom metastasis occurs after apparent surgical cure, others may present acutely with disseminated disease. For both groups, cytotoxic chemotherapeutic agents that are active in other solid tumors show little to no efficacy in melanoma, reflecting the intrinsic chemorefractory phenotype so typical of melanoma. However, the therapeutic landscape for patients with advanced melanoma has been transformed in the last few years by remarkable new developments in targeted and immunotherapeutic approaches [1]. First, cancer genomics has identified several key driver mutations that encode proteins fundamental to melanoma growth and are amenable to therapeutic inhibition. Foremost among these is mutation in BRAF which is present in approximately 40% of cases. NRAS mutation is also common and reported in 15–20% of cases (typically lacking BRAF mutations). In both cases, the mutation results in constitutive activation of MAPK signaling. A third common mutation is in c-KIT, this occurs in mucosal melanomas. Each mutation predicts sensitivity to pharmacological inhibition by specific agents: the BRAF inhibitors vemurafenib and dabrafenib are indicated in cases with BRAF mutations. Addition of the mitogen-activated protein/extracellular signal-regulated kinase kinase (MEK) inhibitor trametinib significantly increases the clinical efficacy of BRAF inhibition [2]. MEK inhibition may also be efficacious in cases with NRAS mutation and imatinib in mucosal melanomas with mutation in c-KIT. Second, in addition to these targeted therapies, there is now compelling evidence that antibodies directed against targets such as CTLA-4 and PD-1/PD-L1 show efficacy in advanced disease and can produce extended clinical benefit in such patients [3].

Why do we need melanoma biomarkers?

Diagnostic and prognostic assessment of surgically excised primary melanomas in routine clinical practice continues to rely on traditional histopathological parameters. The most important of these parameters for risk stratification is the Breslow thickness, an indication of the vertical penetration of the melanoma. Other prognostically useful variables include the presence of ulceration, lymphovascular invasion, host immune response (implied by the presence of infiltrating lymphocytes) and the mitotic index. Molecular genetic biomarkers (aside from the treatment-guiding BRAF and NRAS) are not routinely used in the prognostic assessment at this time. In many ways, melanoma epitomizes the axiom frequently applied to cancer that early diagnosis results in better clinical outcomes. For example, surgical excision of thin melanomas (with a Breslow thickness less than 1 mm) results in cure in the vast majority of cases. In contrast, higher Breslow thickness lesions have progressively less favorable outcomes. With emerging new and effective therapies, it is becoming apparent that early detection may also improve outcomes in metastatic melanoma. Evidence from early trials of ipilimumab made the important observation that patients with stage IVA and stage IVB (lung metastases) disease had better clinical outcomes than those with stage IVC (disseminated) melanoma [4]. This suggests that the use of immunotherapy in low-volume, early metastatic disease will result in superior efficacy and better clinical outcomes than when therapy is initiated in more extensive disease. Similar considerations may well apply to targeted therapies, although clinical confirmation of this is awaited. Therefore, the implication clearly is that biomarkers that can detect recurrent and/or metastatic melanoma in its earliest (typically asymptomatic) stage will have clinical utility in informing the use of immunotherapeutic and perhaps also targeted agents.

Epigenetics as a source of biomarkers

A biomarker is a characteristic that is objectively measured and evaluated as an indicator of normal biological processes, pathogenic processes or pharmacologic responses to therapeutic intervention [5]. Biomarkers have many valuable applications in disease detection and monitoring. In cancer, several biomarkers have been used to reflect the extent of tumor growth and metastasis or as tools for screening and monitoring of disease such as the measurement of prostate-specific antigen or Ca-125.

Epigenetics is the study of nonstructural but heritable and reversible chemical modifications in the genome affecting the chromatin architecture and, as a result, gene expression. An increasing number of epigenetic changes have been described in cancer, in particular global hypomethylation, hypermethylation of tumor suppressor genes and several changes in the histone modification pattern [6]. The most widely studied epigenetic modification is the cytosine methylation of DNA in CpG dinucleotides. This does not occur uniformly throughout the genome but either in clusters of large repetitive sequences (such as rDNA, satellite sequences or centromeric repeats) or in short CG-rich DNA regions, known as ‘CpG islands’ (CGI) and found preferentially in the promoter region of genes. CGIs are normally unmethylated, consistent with the ability of genes containing these islands within their promoter region to be transcribed in the presence of necessary transcription factors. Methylation of cytosine residues within CGIs is associated with the binding of methyl-binding domain proteins, recruitment of histone deacetylases and histone methyltransferases, chromatin condensation and transcriptional inactivation of the associated genes. Aberrant methylation in CGIs has been described in many genes in cancer and is associated with transcriptional silencing.

It has been recognized for a number of years that circulating DNA shed from tumor cells is present in the peripheral blood of cancer patients. Furthermore, the genetic and epigenetic changes present in the tumor tissue are represented and can be detected in the blood-borne DNA. For example, BRAF mutations can be detected in the peripheral blood of melanoma patients [7,8]. The frequent (but not invariable) association of CGI methylation with cancer, together with the relative stability of methylated genomic DNA in peripheral blood, makes it an attractive potential source of biomarkers and detection in the serum of methylated DNA from several genes has been reported. In some cases, an association of specific methylated CGI with metastasis has been described, implying that it may be possible to design assays to specifically identify patients with metastatic rather than local or lymph node-only relapse (clinical scenarios which can be effectively managed surgically). In melanoma, several methylated CGIs have been reported as candidate serum biomarkers [9–11].

How can blood-borne biomarkers improve clinical outcomes in melanoma?

Early metastatic melanoma is frequently asymptomatic and presentation in advanced cases is often with multifocal, large-volume disease. The aim of developing clinically useful serum biomarkers is to provide the physician with a relatively simple blood test that accurately and reliably detects the presence of subclinical, recurrent/metastatic melanoma in asymptomatic, apparently disease-free patients and thereby informs the deployment of systemic therapies at the earliest possible opportunity. Given the increasing evidence that new agents for melanoma are likely to be most effective in low-volume disease, such a test could prove extremely useful in guiding optimal use and timing of agents that have significant toxicities and expense. Another situation in which blood testing might be valuable is in the population of patients being considered for sentinel lymph node biopsy. The availability of a blood test indicative of the presence of metastatic disease in patients with negative sentinel lymph node biopsy and histology would be of obvious benefit. Third, blood testing could have clinical utility in stratifying patients for adjuvant therapy. No regimen has yet demonstrated clear adjuvant benefit following excision of primary melanoma, although interferon is frequently offered to patients in the USA. Nonetheless, the efficacy of targeted and immunotherapeutic agents in advanced disease gives hope that these will be active in the adjuvant setting. The considerable toxicities and expense associated with these drugs require that they are used in patients most likely to be at risk of recurrent disease, in other words, those in whom excision is actually incomplete despite histological evidence of clear resection margins and negative staging investigations. A positive blood test after apparently curative surgical excision could be useful in risk-stratifying patients and identifying those in whom adjuvant therapy is appropriate.

What do we need to do to make this happen?

Identification of leading candidate genes and ongoing discovery initiatives imply that we may, in the relatively near future, have at our disposal one or more validated biomarkers. However, a number of critical issues await resolution before the routine use of blood-borne DNA as a melanoma biomarker can be envisaged.

Which methodology?

An important issue in translating preliminary data into routine clinical use is to define optimal methodology. Robust, clinically useful biomarkers require not only the sensitivity and specificity incumbent of any biomarker but also cost–effectiveness and suitability for automation. All common techniques used for analysis of methylated DNA in blood are based on the chemical modification of genomic DNA with sodium bisulfite, which converts cytosine to thymidine when the cytosine residue is unmethylated, but fails to effect such a conversion if the cytosine is methylated. This process allows the investigator to distinguish between methylated and unmethylated DNA using PCR-based methodologies. Two major methods are used to analyze the methylation of bisulfite-modified DNA, each of which has advantages and disadvantages. Pyrosequencing allows the quantification of methylation at individual CG dinucleotides within an amplified fragment of DNA. To allow the unbiased amplification of unmethylated and methylated sequences, primers for amplification are designed from regions which contain no CG dinucleotides. Pyrosequencing allows high-resolution analysis of methylation and detection of small changes in methylation at each CG. In situations where a homogeneous cell population is undergoing analysis (for example cell lines and microdissected tumors), pyrosequencing allows the quantification of the relative proportions of unmethylated and methylated DNA. In other situations, the presence of large amounts of ‘contaminating’ normal DNA may dilute the methylated DNA masking the methylated alleles that are present and resulting in an underestimation of the methylation signal. This may be a particular problem in the analysis of blood wherein large amounts of normal cell DNA frequently circulate. Quantitative methylation-specific PCR, for example, with systems such as MethyLight, is a second technique used to detect methylated DNA in serum. The advantages of quantitative methylation-specific PCR are technical simplicity; compatibility with existing ‘real-time’ qPCR infrastructure; the PCR reaction is specific for methylated DNA and hence the presence of background levels of ‘normal’ genomic DNA in blood do not dilute the positive methylation signal, thereby reducing the possibility of false-negative samples.

Single genes or panel of genes?

The sensitivity and specificity of individual candidate genes to identify subclinical metastasis and relapse must be validated prospectively in large, independent, well-annotated clinical series. It remains to be determined whether analysis of a single gene will be adequate or whether a panel of marker genes is required. It may be that an initial methylation screen of tissue from each patient’s primary melanoma is required in order to generate a methylation map of a panel of genes and to use this as the basis to select the appropriate genes to analyze in blood. A limitation of this approach is that the methylation status of genes is dynamic and may change between the primary and metastatic lesion. Alternatively, the analysis of a set of predefined genes may be more cost–effective.

How do we use validated biomarkers?

An important outstanding question is how exactly the detection of methylated genomic DNA in serum can be most usefully employed in the melanoma clinic. Similar uncertainties apply to non-DNA biomarkers used in other solid tumors. Should clinical decisions be made simply on the basis that levels of methylated DNA are above predesignated cut-offs established in prior studies? Alternatively, should levels of methylated DNA be required to increase in serial blood samples before a test is deemed ‘positive?’ These questions can only be answered by appropriate testing.

Clinical trialing

Assuming the above issues are addressed and resolved, it then remains to be definitively determined whether the clinical use of blood tests for methylated DNA can actually improve clinical outcomes in melanoma. This would entail comparing appropriate clinical endpoints in patients receiving first-line therapy for metastatic melanoma informed by conventional clinical assessments (examination, imaging and possibly tissue re-biopsy) with an experimental cohort in whom systemic therapy was initiated as a result of positive blood testing.

Conclusion

Detection in blood of the genetic and epigenetic changes present in metastatic cancers is opening up new possibilities in molecular diagnostics. A number of issues discussed above await resolution before serum epigenetic biomarkers can be considered a routine part of the clinical management of melanoma patients following primary excision. However, there is every possibility that blood testing for the presence of methylated DNA will become an integral part of the clinical follow-up of such patients. The ability to identify patients with subclinical (asymptomatic) metastatic disease combined with new therapeutic agents may lead to further improvements in outcomes for this patient population.

Financial & competing interests disclosure

The authors have no relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript. This includes employment, consultancies, honoraria, stock ownership or options, expert testimony, grants or patents received or pending, or royalties.

No writing assistance was utilized in the production of this manuscript.