Abstract
Plain language summary
Ovarian cancer stands as the deadliest malignant tumor within the female reproductive tract. As a result of the absence of effective diagnostic and monitoring markers, 75% of ovarian cancer cases are diagnosed at a late stage, leading to a mere 50% survival rate within five years. Nearly 80% of advanced stages have a poor prognosis or recurrence within five years. Ovarian cancer is linked to a grim long-term prognosis attributable to its elevated mortality and recurrence rates. The advancement of molecular biology and diagnostic methods is essential for accurate diagnosis and treatment of ovarian cancer. Liquid biopsy is an innovative method of detecting malignant tumors that has gained increasing attention over the past few years. Cell-free DNA assay-based liquid biopsies show potential in delineating tumor status heterogeneity and tracking tumor recurrence. DNA methylation represents a prevalent epigenetic modification. DNA methylation influences a multitude of biological functions and diseases, especially during the initial phases of cancer. The cell-free DNA methylation profiling system has emerged as a sensitive and non-invasive technique for identifying and detecting the biological origins of cancer. This review assesses recent progress and obstacles linked to cell-free DNA methylation analysis for diagnosing, prognostic monitoring, and evaluating therapeutic responses in managing ovarian cancers.
Keywords
Introduction
Ovarian cancer (OC) comprises a diverse range of malignancies originating from ovarian tissue and differ in their clinicopathology, molecular characteristics, and histological origin. 1 The Global Cancer Statistics from the International Agency for Research on Cancer (IARC) by the World Health Organization indicate that ovarian malignancies persist as among the most fatal cancers among women globally, accounting for roughly 313,959 new cases and 207,252 deaths in 2020. 70% of these cases are in developing countries (https://gco.iarc.fr/). 2
Early detection and screening for OC are ineffective due to its insidious nature. 3 As ovarian epithelial cancer tumor markers, CA125 and HE4 are useful for diagnosing, monitoring efficacy, and monitoring recurrence. However, their levels can also be influenced by tumor stage, histological type, age, and menopausal status, along with observed elevation in some benign and malignant conditions related to the female reproductive system. Early-stage OC positive rates range from 43.5-65.7%. Additional tumor markers like CA199 and CEA may also show elevated levels but lack specificity to OC, often associated with gastrointestinal malignancies. 4 While imaging can be used to detect ovarian masses, it cannot make a definitive diagnosis or differentiate benign from malignant lesions, and the lesion cannot be detected until the size reaches a significant level. 5 There remains a deficiency in specificity concerning tumor marker testing alone, transvaginal ultrasound screening alone, or their combined use, which results in unsatisfactory rates of early detection. Multigene testing is useful for identifying individuals with OC at high risk, predicting their prognosis, and selecting therapeutic agents. 6 As ovarian cancers are mostly sporadic, with only 15% being hereditary, and genetic testing is currently costly, the test is not widely available as a screening tool. There is a need to explore more screening methods for the public.
Currently, OC patients are managed with surgery, radiotherapy, targeted therapy, hormone therapy, immune therapy, and Chinese herbal medicine. 7 Following the development of targeted therapy and immunotherapy technologies, polyadenosine diphosphate ribose polymerase inhibitors (PARPi) were introduced and applied to OC maintenance therapy, which completely altered the dilemma of treating OC. 8 Clinical remission is attained in the majority of individuals diagnosed with ovarian carcinoma after their initial treatment. However, 70% of patients have recurrences after treatment, quickly develop platinum resistance, and exhibit a five-year survival rate of merely 46%. 9 Developing new effective biomarkers is crucial for detecting ovarian malignancies earlier, monitoring drug response, personalizing treatment, and improving patient outcomes.
During the past decade, liquid biopsy has emerged as a potent tool enhancing precision in oncology. In contrast to tissue biopsy, liquid biopsy is a rapid, convenient, non-invasive procedure that detects the source of tissue and reflects tumor heterogeneity by analyzing bodily fluids, encompassing blood, urine, stool, saliva, and cerebrospinal fluid. 10 Currently, liquid biopsy analytes include circulating tumor cells (CTCs), cell-free DNA (cfDNA), circulating tumor DNA (ctDNA), cell-free RNA (cfRNA), microRNA, exosomes, tumor-educated platelets (TEPs), proteins, and metabolites, among others. It’s crucial in identifying tumor patients in early screening, precise treatment, monitoring efficacy, and predicting their prognosis. 11 Peripheral cell-free DNA serves as one of the most widely used analytes. Normal cell renewal produces cfDNA in a healthy individual, while cancer patients produce high amounts of cfDNA.
DNA methylation, beyond its role as a pivotal factor in gene regulation, stands as one of the extensively explored epigenetic mechanisms. Anomalies in DNA methylation are closely linked to both tumorigenesis and cancer progression, thereby establishing DNA methylation analysis as a potent approach for detecting cancer.
12
The potential feasibility and reliability of employing cfDNA methylation patterns as cancer biomarkers stem from their alignment with the methylation profiles of originating cells or tissues.
13
The FDA’s approval of a colorectal cancer screening test in 2016, reliant on the methylation status of the fecal SEPT9 promoter region, marks a pivotal advancement in cfDNA methylation.
14
Nevertheless, the low percentage of methylated cfDNA fragments poses challenges to detection methods for coverage, cost, and sensitivity.
15
This review examines cfDNA methylation as a promising liquid biopsy biomarker for diagnosing, prognosticating, and treating ovarian cancer (Figure 1). Future directions and challenges in the field are discussed. The liquid biopsies are non-invasive method and conducted by isolating and extracting tumor-derived components from body fluids, including CTCs, cfDNA, ctDNA, cfRNA, and et al. Tumors present with altered DNA methylation status, which is characterized by reduced genome-wide methylation levels and elevated sequence-specific methylation levels at CpG islands. The clinical significance of DNA methylation in the context of OC.
Cell-Free DNA Biology in Liquid Biopsies
In traditional medicine, histopathology and immunohistochemistry are used to diagnose tumors, but a biopsy of just one specimen is not comprehensive and cannot account for the entire tumor. In addition, pathologists have limited access to specimens and focus primarily on advanced tumors, which could lead to them overlooking progressed and invasive tumors. Furthermore, the heterogeneous expression of immune markers in different tissues may not reflect the true nature of the tumor. Lastly, invasive tests pose a significant challenge when monitoring drug therapy response and the likelihood of tumor recurrence. 16 The precision oncology field is slowly shifting toward liquid biopsy, an easier-to-reproduce, less invasive testing method with reduced invasiveness, accessibility, and reproducibility. Moreover, it possesses the capability to detect tumor heterogeneity and dynamically track tumor evolution. This ability proves particularly valuable in the early detection of cancer, evaluation of treatment resistance, and surveillance of residual lesions and recurrences.17,18
The discovery of cfDNA within the bodily fluids of healthy individuals and those afflicted with diverse diseases was initially documented by Mandel and Metais in 1948. 19 cfDNA is a term used to describe free fragments of double-stranded DNA within body fluids that are not contained within intact cell structures. The pieces of cfDNA can be found in normal and abnormal cells (for example, tumor cells), as well as viral DNA in virus carriers’ body fluids, and fetal DNA can be found in peripheral blood of expectant mothers. 20 In addition to apoptotic and necrotic cells, DNA fragments can also be released through cellular secretions. The fragmentation of cfDNA in healthy human body fluids is caused by apoptotic cells, while the production of cfDNA through cell necrosis is controversial.21,22 cfDNA fragments in the bodily fluids of individuals with malignant tumors originate from apoptotic cells as well as actively released DNA by tumor cells and fragments from tumor cell necrosis. 23 cfDNA fragments and their epigenetic modifications encapsulate genetic and molecular traits specific to the originating tissue, enabling the inference of tissue origin and consequently reducing tumor heterogeneity. 24 Its molecular weight is lower than genomic DNA, between 150-200 bp, and its half-life is about 15 minutes to 2 hours, so cfDNA responds “instantly” to disease. Healthy individuals typically exhibit low cfDNA concentrations, usually below 20 ng/mL. 21 Nevertheless, cfDNA concentrations in patients with malignancies can reach 1000 ng/mL in body fluids. Occasionally, cfDNA levels may increase in view of trauma, autoimmune diseases, or compression of body tissues. 25 As a result, it is still difficult to rely solely on the content of cfDNA to identify cancer biomarkers. Tumor cell-released cfDNA transforms into ctDNA, commonly comprising fragments exceeding 200 bp and occasionally surpassing 1000 bp in length. 26 Tumor load, tumor site, body metabolic rate, and anti-tumor therapy affect the ctDNA content, resulting in a range of .05% to 93% for ctDNA in cfDNA. 27 The level of ctDNA in a tumor reflects the level of tumor burden, and the measurements of ctDNA are minimally invasive and reproducible for monitoring tumors dynamically.28,29 The primary technical hurdle lies in distinguishing ctDNA from cfDNA.
Cell-free DNA Methylation in Oncology
DNA methylation is one of the initial and extensively explored epigenetic regulatory agents governing DNA. 30 DNA methylation involves a chemical modification procedure wherein distinct bases on the deoxyribonucleic acid sequences undergo methylation through the covalent attachment of a methyl group facilitated by DNA methyltransferases (DNMT1, DNMT3a, DNMT3b), utilizing S-adenosyl methionine (SAM) as the methyl donor. 12 Occurrence of methylation modification at the 5th carbon atom of cytosine within the CpG island dinucleotide located in the gene promoter, creating 5-methylcytosine (5-MC).
Mutations in methylation may be a significant cause of tumorigenesis and an early molecular event in tumorigenesis. 31 Tumors present with altered DNA methylation status, which is characterized by reduced genome-wide methylation levels and elevated sequence-specific methylation levels at CpG islands. 32 Hypermethylation suppresses gene expression and hypomethylation leads to genomic instability. The hypomethylation of promoter regions and non-coding repeats of proto-oncogenes causes gene activation and destabilizes the genome structure, resulting in abnormal cell proliferation. 33 In the past, DNA methylation was assumed to be irreversible. Nevertheless, studies conducted over the past decade have revealed that gene expression and chromosome stability depend on a dynamic balance between methylation initiation, methylation maintenance, and demethylation.15,34 By catalyzing the conversion of 5-methyl cytosine into its derivatives like 5-hydroxymethyl cytosine (5hmC), 5-carboxyl cytosine (5caC), and 5-formyl cytosine (5fC), TET family proteins alter the recognition of DNMT at the conversion site, facilitating the process of demethylation. 35
Genetic mutations are the drivers of cancer and the underlying cause of resistance to targeted therapies. 36 Plasma cfDNA methylation profiles are highly correlated with tumor DNA methylation profiles. In specific genetic regions, aberrant methylation of cfDNA is highly consistent. Different types of tumors exhibit specific DNA methylation sites. 37 It is more promising to use cfDNA methylation testing for tumor diagnosis, monitoring, treatment effectiveness, and prognosis determination than tumor-specific mutation detection. 36 The human genome contains approximately 28 million CpG methylation loci, providing valuable information to cfDNA to differentiate tumor origins. 38 Compared to cfDNA quantitative or mutation site testing, cfDNA methylation testing has the following advantages: (1) cfDNA methylation modifications occur much earlier in the tumor. (2)The methylation of cfDNA can reveal tumor traceability at primary foci for unidentified cancers.39,40 (3)Homogeneous and stable in cancer, promising to solve tumor heterogeneity. 41 (4) The brief half-life of cfDNA and the reversible nature of its methylation modifications allow for real-time monitoring of tumor progression and treatment response. 42
Research on the methylation of cfDNA currently focuses more on the diagnosis, monitoring, and prognosis of breast, colon, liver, and lung cancers.43,44 The clinical application value in OC is less clear. This review systematically describes the definite potential of cfDNA methylation testing in OC patients, the currently available technical support, and the future challenges facing the field.
Cell-Free DNA Methylation as a Biomarker for the Diagnosis of OC
Diagnosis of Methylated cfDNA/ctDNA in patients With Ovarian cancer (OC).
BC-OC ovarian cancer follows breast cancer, CC clear cell carcinoma, ddPCR droplet digital PCR, DMRs Differentially Methylated Region, EOC Epithelial ovarian carcinoma, ENDO endometroid carcinoma, HGSC high-grade serous carcinoma, LMP low malignant potential carcinoma, LGSC low-grade serous carcinoma, MSP methylation-specific PCR, M-MSP multiplex methylation-specific PCR, MeDIP-seq methylated DNA immunoprecipitation sequencing, MIX mixed type carcinoma, MUC mucinous carcinoma, MRE-seq methylation-sensitive restriction enzyme sequencing, N Normal population, NR Not report, OC ovary cancer, PS papillary serous carcinoma, PP primary peritoneal origin carcinoma, qPCR quantitative PCR, RRBS reduced representation bisulfite sequencing, TC* transitional cell carcinoma, TC Training cohort, UNDIFF undifferentiated carcinoma, VC Validation cohort, * methylation of either or both genes
Single-Gene Methylation Biomarkers
RASSF1A and BRCA1 are essential oncogenes. The BRCA1 gene is crucial in regulating cell replication, DNA repair, and cell growth. RASSF1A regulates the cell cycle, stabilizes microtubules, governs cellular adhesion and movement, and induces apoptosis through activating RAS proteins. According to Ibanez de Caceres, 54 methylation-specific PCR (MSP) analyses were performed on 50 OC patients to determine BRCA1 and RASSF1A methylation levels in tumor tissue, serum, and plasma. Forty-one out of fifty samples (82%) of the corresponding serum or plasma DNA showed the same gene hypermethylation pattern. The tumor stage was not statistically associated with positive serum detection (P = .47). Combining BRCA1 and RASSF1A enhances diagnostic coverage, and their sensitivity exceeds that of BRCA1 and RASSF1A individually (68% vs 24% and 50%). In contrast, 20 normal, age-matched women did not show genomic hypermethylation (100% specificity). Further, Wu et al 46 reported a 36.2% methylation rate of RASSF2A in epithelial ovarian cancer (EOC) patient plasma, with no methylation in benign conditions and healthy controls, suggesting a significant link between serum and tissue methylation profiles in EOC. Sandeep et al 47 presented a methylation frequency for BRCA1 of 62% and RASSF1a of 38% in the serum of disorders with OC. Statistical analysis indicated a significant correlation between BRCA1 methylation status and FIGO staging (P = .001), CA-125 level (P = .001), histological type (P = .042), and CEA level (P = .001).
The study reported the inaugural discovery of ESR1 methylation within the plasma ctDNA of patients diagnosed with high-grade serous ovarian carcinoma (HGSC). 52 There was a 38% sensitivity for methylated ESR1 in diagnosing HGSC. Of the paired specimens, 36 out of 48 (75.0%) exhibited substantial concordance (P = .004), achieving a Cohen’s kappa of .429.
Several research studies have shown that detecting OPCML methylation in serum helps detect OC. In Zhou’s study, 53 OPCML hypermethylation rates at FIGO stages I, II, III, and IV were calculated and found to be 42.9%, 66.7%, 85.7%, and 100%, respectively (P < .05). Similarly, Wang et al. 55 reported that serum OPCML methylation detection exhibited 87% sensitivity and 91% specificity for identifying early-stage OC.
Liquid biopsies can benefit from this improved analysis method due to its increased sensitivity. Faaborg et al 49 identified methylated homeobox A9(HOXA9) ctDNA, capable of targeting both sense and antisense DNA strands, as a potent OC marker. This dual-strand approach, utilizing droplet digital PCR, revealed methylated HOXA9 in 60% of patients, demonstrating a notable sensitivity increase, particularly in early-stage diagnoses.
Methylation Panels
cfDNA may be more easily detected by nested PCR, potentially leading to more accurate detection of OC when multiple methylation markers are tested simultaneously. A study by Melnikov 51 used the methylation of 5 genes (BRCA1, HIC1, PAX5, PGR (prox), THBS1) as an indicator of cancer risk, and their combination had an 85% sensitivity and 61% specificity. Similarly, Zhang et al 45 employed a multiplex-MSP assay targeting seven genes (RUNX3, OPCML, SFRP5, RASSF1A, APC, CDH1, and TFPI2), achieving significantly higher sensitivity (86%) and specificity (91%) compared to the single CA125 test during early-stage disease (P = .0036). Wang et al 59 developed a multiplex nested MSP system targeting RUNX3, TFPI2, and OPCML, which demonstrated 84% sensitivity and 91% specificity for stage I-II EOC. Su et al 48 analyzed six genes (LMX1A, SFRP2, PAX1, SFRP5, SFRP1, and SOX1), revealing higher methylation rates in OC patients compared to those with borderline malignancies or benign tumors (P < .001). Notably, methylation analysis of cfDNA in tissues and serum showed a strong association (kappa values .332 to .598) with high sensitivity and specificity (73% and 75%) for screening purposes. Widschwendter et al 57 reported a sensitivity of 41% and specificity of 91% for HGSO diagnosis using a combined panel of COL23A1, C2CD4D, and WNT6.
Singh’s development of a TaqMan-based qPCR assay, tailored for a novel set of methylation-specific genes, further refines diagnostic accuracy. 50 Evaluating HOXA9 and HIC1 promoter methylation separately, and then in combination, resulted in a substantial sensitivity boost to 88.9% (AUC = .95), underscoring the assay’s efficacy in noninvasive OC detection. Singh’s study report evaluates a high-sensitivity, high-specificity qPCR-based noninvasive epigenetic biomarker test designed for OC detection. 60 A proficient multiplex assay was developed for these genes, discerning cancer patients from healthy individuals (AUC: HOXA9 + SOX1 = .85, HIC1 + SOX1 = .93 and HOXA9 + HIC1 = .95). Liggett et al 58 identified cfDNA methylation patterns employing a microarray-based testing named MethDet 56. Methylation levels in promoters (EP300, CALCA, and RASSF1A) showed significant differences between OC and HC, exhibiting 90% sensitivity and 87% specificity. Miller et al 61 proposed EpiClass, a binary classification system for methylation density, aimed at optimizing methylation biomarker efficacy, especially in heterogeneous samples like liquid biopsies. EpiClass’s effectiveness was confirmed through its high predictive accuracy for early OC detection, emphasizing its utility across diverse cancer diagnostics.
Lu’s 2022 study showcased the potential of cfMeDIP-seq-derived cfDNA methylomes in identifying OC-specific biomarkers, especially those linked to early-onset OC. 56 They identified differentially methylated regions (DMRs) and evaluated discrimination performance through iterative testing and training. Using the DMRs identified for cfDNA methylomes, it is possible to distinguish tumor groups from non-tumor groups with reasonable accuracy (AUC:0.86-.98). In a parallel study, Marinelli et al 62 employed Target Enrichment Long-probe Quantitative Amplified Signal (TELQAS) assays to examine methylated DNA markers in plasma samples obtained from both newly diagnosed OC patients and healthy women. An 11-gene panel (GYPC, CELF2, SRC, GPRIN1, BCAT1, CDO1, CAPN2, SIM2, RIPPLY3, FAIM2, AGRN) has demonstrated an excellent ability to differentiate OC from controls (96% specificity, 79% sensitivity, and .91AUC).
Ovarian Cancer Detection Model
Liang et al 63 developed a model for OC detection, identifying 18 methylation markers in cfDNA sequences via ELSA-seq and employing LASSO regression for the OC-D model construction. This model demonstrated superior diagnostic performance with 95% sensitivity and 89% specificity, outperforming CA125 (AUC: .97 vs .91, P = .03). Concurrently, Dvorská 64 research on methylation levels of tumor suppressor genes (PAX1, CDH1, PTEN, RASSF1) in plasma from individuals with healthy, benign, and cancerous ovarian conditions revealed a plasma-based model with 91% sensitivity and 56% specificity, showcasing its potential clinical applicability with an AUC of .82 and a 95%CI between .67 and .98.
Cell-Free DNA Methylation as a Biomarker for OC Prognosis
Prognosis and Prediction of Methylated cfDNA/ctDNA in patients With Ovarian cancer (OC).
ddPCR digital droplet PCR, MSP methylation-specific PCR, HGSOC high-grade serous ovarian cancer, NR not report, N normal population, NG not given, OS overall survival, OC ovarian cancer, PFS progression free survival, PARPi poly (ADP-ribose) polymerase inhibitor, RT-qPCR quantitative reverse transcription PCR,TC Training cohort, VC Validation cohort
Detection of Minimal Residual Disease (MRD)
Analysis of cfDNA methylation offers a non-invasive approach for assessing tumor burden and detecting MRD. Examination of cfDNA in bodily fluids, with a particular focus on methylation patterns, enables the identification and characterization of subclinical residual tumors. Consequently, this facilitates a more precise prognostication and assessment of recurrence risk in patients.
Zhu et al 70 observed that ctDNA was identifiable in the postoperative plasma of most patients deemed by surgical assessment to be free of visible residual disease following primary tumor debulking surgery. Lee et al 71 have elucidated that the detection of TP53 variants in cfDNA following neoadjuvant chemotherapy is instrumental in identifying MRD within tumors. Forshew et al 72 assessed disease recurrence following primary cytoreductive surgery in 23 OC patients through sequential sampling of cfDNA, revealing a sensitivity exceeding 90% for cfDNA in predicting tumor recurrence. Moreover, the investigation unveiled that cfDNA anticipated tumor recurrence approximately seven months earlier than computed tomography-assisted examination. 73
Association With Resistance to Treatment
Gifford et al 68 initially highlighted DNA methylation alterations in plasma, paving the way for targeted epigenetic therapy based on patient-specific methylation profiles during treatment. Analysis of plasma DNA from Phase III SCOTROC1 study participants with EOC involved evaluating hMLH1 CpG island methylation before carboplatin/taxoid chemotherapy and upon recurrence. A significant correlation was observed between post-progression survival in EOC patients and hMLH1 CpG island methylation (n = 131, HR: 1.83, P = .017), negatively impacting OS regardless of progression timing or age (HR: 1.99, P = .007). At relapse, hMLH1 methylation, associated with microsatellite instability in plasma DNA, serves as an independent marker for DNA mismatch repair (MMR) pathway functionality. The findings suggest that DNA MMR loss and hMLH1 methylation are two essential mechanisms of acquired drug resistance within EOC.
Flanagan et al 66 further demonstrated that intact MMR enhances platinum-induced methylation changes, with specific CpG methylation alterations in plasma at relapse correlating with survival outcomes independent of clinical factors (HR: 3.7, 95%CI: 1.8-7.6). There are the same changes in ovarian tumors at relapse, which are related to patient survival (HR:2.6; 95% CI:1.0-6.8, P = .048). Therefore, assessing DNA methylation in the circulation during relapse after chemotherapy, as opposed to its assessment at initial diagnosis, can provide valuable insights into the survival outcomes of OC patients.
Correlation With Progression-Free Survival (PFS) and Overall Survival (OS)
Elazezy et al 65 presented a PCR-based liquid biopsy assay that is highly sensitive and specific for methylation. According to their findings, 60% of cancer patients showed hypermethylation of the BRCA1 promoter before treatment, although 24% had it reverted as the treatment progressed. Multivariate survival analyses suggest that relapses operate as autonomous occurrences, with hypermethylation and methylation conversion, each showing independent associations with prolonged recurrence-free survival. This suggests a frequent reversal of BRCA1 promoter hypermethylation upon tumor recurrence, which associate with poorer clinical outcomes.
For the first time, Tserpeli et al 67 reported that the methylation status of SLFN11 in cfDNA within plasma was associated with decreased PFS in advanced HGSOC (P = .045). Liang et al 63 sequenced plasma-derived cfDNA using ELSA-seq from eligible participants. They explored how methylation markers, identified within the predictive model, might impact tumor progression, immune response, and homologous recombination repair signaling within the TCGAOC cohort using tissue methylomes. Results revealed that the high-risk group exhibited significantly shorter OS and PFS durations compared to the low-risk group (HR: .75, 95% CI: .60-.93, P = .01; HR: .82, 95% CI: .67-.99, P = .045, respectively). Research findings confirm that utilizing cfDNA methylation markers represents a credible and precise approach for OC detection and prognostication.
Rusan et al 69 examined the potential of HOXA9 promoter methylation as a biomarker in the treatment of platinum-resistant BRCA-mutated OC utilizing PARP inhibitors. They employed a ddPCR to analyze the HOXA9 methylation status in ctDNA at the initial assessment and preceding each therapeutic cycle. After three rounds of chemotherapy, patients exhibiting detectable methylation of ctDNA had a median PFS of 5.1 months, in contrast to 8.3 months for those without detectable meth-ctDNA (P = .0001). Their median OS was 9.5 months compared to 19.4 months in patients lacking detectable meth-ctDNA (P = .002).
Cell-free DNA Methylation as a Biomarker for Monitoring OC Response to Therapy
Methylated cfDNA/ctDNA Biomarkers are Used to Monitor Ovarian cancer (OC) patients’ Response to Treatment.
ddPCR digital droplet PCR, HGSCOC high-grade serous ovarian cancer, EOC epithelial ovarian cancer, NG not given, PFS progression-free survival, PARPi Poly (ADP-ribose) polymerase inhibitor, OS overall survival, qPCR quantitative PCR, RRBS reduced representation bisulfite sequencing, Real-Time MSP Real-Time methylation-specific PCR
Monitoring OC Response to Chemotherapy
Widschwendter et al 57 recruited 25 patients with OC who were receiving carboplatin-based chemotherapy. The study demonstrated a significant decrease in the levels of three specific markers in the pre-chemotherapy phase compared to after two cycles of therapy, enabling the distinction between non-responders and responders with accuracies of 86% and 78%, respectively (Fisher’s exact test, P = .04). This performance surpassed that of a CA125 cut-off of 35 IU/mL. Furthermore, 78% of responders and 100% of non-responders exhibited no residual disease post-interval debulking surgery (P = .007). Data from this study suggests that DNA methylation patterns in cfDNA may provide useful information for personalized treatment planning.
Bondurant ‘s study 76 supports the potential for monitoring RASSF1A methylation in serum specimens collected from OC patients. Real-time PCR utilized non-specific primers and probes containing sequences featuring a lone CpG site within bisulfite-modified DNA. This probe facilitates the determination of the methylation status of individual alleles within a specific specimen. Over the course of treatment, serial serum samples revealed fluctuations in the level of RASSF1A methylation that were associated with disease progression. They found that real-time assay provides a reliable means of monitoring RASSF1A methylation status in serum specimens. Statistical analysis of RASSF1A methylation could yield insights into disease progression.
In a recent study, Jakobsen et al 75 compared the response rates of ctDNA with objective response rates to evaluate their effectiveness as an alternative marker for OS in advanced cancer patients undergoing chemotherapy. Based on the included subtypes of tumors and associated therapies, a notable link existed between ctDNA response and median survival (R2 = .99), superior to both the first evaluation and overall response rates (R2 = .70 and .57, respectively). The ctDNA response holds promise as a potential alternative marker for OS.
Werner et al 74 conducted a study that indicated that cfDNA is traceable between consecutive samples. Across a 36-day timeframe, samples of ascites were collected from a solitary patient in the cohort. The patient, with recurrent disease and not undergoing chemotherapy at enrollment, received carboplatin during the collection period. Analysis revealed a significant increase in cfDNA levels across the seven samples (P = .017), consistently exhibiting methylation of IFFO1 and HOXA9 genes. Over 72 hours, ascites cfDNA fraction size, quantity, and tumor content remained stable. Based on the results of this investigation, ascites cfDNA appears to be an effective, readily available, and convenient source of DNA. As such, it may offer insight into OC pathogenesis and allow for customized treatment options.
Monitoring OC Response to Radiotherapy
Radiation therapy, a cornerstone in cancer treatment, lacks a personalized, radiobiological approach integrating tumor biology and patient-specific risk. Liquid biopsy aids in identifying oncogenic and epigenomic changes, crucial for assessing neoplastic susceptibility to radiotherapy. Specifically, mutations in DNA repair genes like those in the BRCA1/2 pathways may increase tumor sensitivity to radiotherapy, enhancing treatment efficacy. Blomain et al 77 conducted research to identify ctDNA variations correlating with radiotherapy responses, suggesting ctDNA-based adjustment of therapy intensity could optimize outcomes and reduce toxicity. Liquid biopsy, by analyzing ctDNA, offers a method for continuous monitoring of tumor genetic diversity and the evolution of radiation-resistant subclones. Additionally, Shukla et al 78 highlighted the potential of proteomics in identifying markers predictive of radiation resistance and toxicity, as well as in evaluating radiation therapy’s effectiveness.
Cifuentes et al 79 evaluated the effectiveness of genomic analysis of reproductive system cancers and cfDNA in enhancing radiotherapy approaches for early-stage cancer. Their longitudinal analysis of ctDNA variants before, during, and after radiation therapy established a link between pre-treatment ctDNA presence and improved PFS in early-stage cancer patients. The elucidation of tumor-specific genetic mutations enhances the capability for tracking clonal diversificationand acquired resistance, thereby offering a framework for stratifying patients based on projected treatment responses. 80
Monitoring OC Response to Immunotherapy
The effectiveness of immune checkpoint blockade (ICB) has been demonstrated across various tumor types and has gained widespread adoption in clinical settings. However, only a minority of treated patients exhibit prolonged clinical benefit, underscoring the imperative to identify biomarkers capable of prognosticating patient response to ICB therapy. Recent investigations have elucidated genomic hypomethylation status as a pivotal determinant influencing the efficacy of ICB treatment. Pioneering this avenue of inquiry, Kim et al 81 were the first to probe the feasibility of employing genomic methylation analysis of cfDNA in the domain of cancer immunotherapy. They devised an analytical methodology termed iMethyl, tailored explicitly for evaluating the extent of genomic hypomethylation in cfDNA specimens. Significantly, this technique remains unaffected by tumor purity in tissue samples and exhibits robust predictive capacity for patient response to ICB therapy. Notably, iMethyl facilitates real-time monitoring of treatment response during the early phases of therapy (within 3 to 6 weeks post-ICB initiation) and enables detection of hypomethylation dynamics accompanying tumor progression. Comparative analysis with conventional measures such as tumor mutation load and PD-L1 expression levels revealed iMethyl’s superior accuracy in prognosticating ICB treatment outcomes.
Rusan 69 suggests that monitoring HOXA9 meth-ctDNA could assist clinicians in selecting suitable treatments for platinum-resistant BRCA-mutated OC involving PARPi. The status of methylation and changes in methylation from baseline were associated with OS and PFS. Identifying HOXA9 meth-ctDNA during PARPi administration contributed to unfavorable clinical outcomes. Patients who tested positive for HOXA9 meth-ctDNA after three treatment cycles had a median PFS of 5.1 months, compared to 8.3 months for those without detectable levels (P < .0001). Similarly, their median OS was 9.5 months vs 19.4 months (P = .002). Notably, patients initially positive for HOXA9 methyl-ctDNA at baseline but subsequently became undetectable showed the most favorable clinical outcomes, followed by those maintaining consistently undetectable levels during treatment.
Prospects and the Future
Molecular modifications in DNA methylation offer crucial insights for diagnosis and prognosis due to the subsequent factors: Epigenetic modifications occur during the initial stages of OC pathogenesis, DNA methylation pattern changes seem more prevalent among OC subtypes than mutations particular to OCs, and PCR-based analysis is viable owing to the stability of DNA methylation. 82
This review thoroughly examines current research on DNA methylation analysis from liquid biopsy samples for the early detection and prognostic assessment of OC. Liquid biopsy, being noninvasive, holds promise as a biomarker, enabling early screening, recurrence monitoring, and prognostic evaluation of cancer (Figure 2).
83
Before its full integration into clinical practice, comprehensive, objective research is still required to address many challenges and limitations. Standardizing the collection, DNA isolation, and quantification methods in liquid biopsy remains essential.
84
The domain of cfDNA methylation in ovarian carcinoma. cfDNA methylation, characterized by its noninvasive nature, presents a promising biomarker for the early screening, assessment of tumor burden, evaluation of drug resistance, detection of tumor relapse, identification of minimal residual disease, monitoring of treatment efficacy, and prognostic assessment in oncology.
There is a lack of standardized DNA methylation reporting, concerning analytical approaches and target genes, which impedes the reproducibility of cfDNA methylation data. 85 Therefore, establishing a uniform reporting standard is crucial given the discovery of numerous potential biomarkers in recent studies. Other elements of physiology may modify cfDNA epigenetics and biology, such as myocardial infarction, and genetic variations and tumor heterogeneity may also exist objectively. 86 Consequently, when creating an epigenetic-based liquid biopsy assay, these aspects should be thoroughly addressed to avoid any potential disruption to the findings.
The intricate nature of biological properties, the diverse range of cfDNA molecules, and the substantial genetic variability linked to cancer pose challenges to interpretation results. 87 Furthermore, considering that sometimes DNA concentrations in body liquid are extremely low. 82 We need highly sensitive and precise detection techniques to overcome this obstacle and identify novel epigenetic fingerprints. Advancements in technology offer the possibility of minimizing the diversity found in cfDNA and precisely identifying molecular characteristics specific to tissues and diseases by utilizing DNA methylation profiles. A spectrum of sophisticated technologies, including advanced statistical data analysis, methylation arrays, next-generation sequencing, methyl immunoprecipitation, bisulfite sequencing, and digital PCR, is currently at our disposal to address the challenges of developing non-invasive methods in personalized medicine. Enhanced optimization of library preparation and methylation enrichment techniques is necessary to advance studies reliant on cfDNA. 13
Varied disease stages potentially influence the methylation status of specific genes. 88 Combining information from DNA methylation status might provide insight into disease progression. Different stages of the disease may have different DNA methylation profiles within the same target region. 89 Various types of cancer transfer DNA into the bloodstream at different rates. The measurement of cancer burden involves shedding rates, which vary according to the type, stage, and clinical condition of the cancer. These are all difficulties we need to overcome.
A single gene cannot capture the biology of a disease. Combining multiple biomarkers in this manner would enhance prediction and facilitate early diagnosis. Various screening modes and marker combinations may be used to diagnose cancer in the future. Combining diverse parameters from various facets of liquid biopsy enhances result reliability, including protein, epigenetic, or mutation-based biomarkers. 90 Despite this, multiparameter analysis is an analytical and computational challenge. 11 The traditional method would take considerable time and resources, requiring large teams and analytical capabilities. Anticipated progress in fundamental and molecular biology, detection technologies, statistical analysis, and machine learning is poised to drive its future evolution.
Conclusion
In summary, DNA methylation profiling holds significant promise as a molecular biomarker for directing the clinical management of OC. These findings indicate that detecting gene methylation levels in cfDNA offers a less invasive method for OC diagnosis and patient prognosis stratification. While these findings are promising, they have not yet been validated in large multicenter cohort studies and randomized trials that will enable their implementation in medical practice. In forthcoming studies, it is pivotal to enhance cfDNA extraction protocols to yield superior quality cfDNA, refine methylation assays suitable for limited-input cfDNA, and explore advanced, more sensitive, and specific methodologies for identifying cfDNA-specific methylation. The complete utilization and anticipated impact of cfDNA methylation on genomics-driven oncology and clinical cancer management hinge on the demonstration of empirical validity and clinical utility.
Footnotes
Authors’ Note
All authors contributed to the study conception and design. Literature collection were performed by YG and NZ. YG drafted the manuscript. NZ revised and finalized the review. JL provided supervision.
Declaration of Conflicting Interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The author(s) received no financial support for the research, authorship, and/or publication of this article.
