Liquid Biopsy to Catch the Epigenetic Changes in Endometrial Cancer

The liquid biopsy (LB) concept was introduced and coined for the first time in 2010 (1) for circulating tumor cells and rapidly extended to other circulating biomarkers such as circulating tumor DNA (ctDNA), circulating cell-free RNA (noncoding and messenger RNA), extracellular vesicles (exosomes), circulating microRNA, and tumor-educated platelets, as well as the immune cells (2). Moreover, this concept has been expanded to other body fluids, including urine, cerebrospinal fluid, bone marrow, saliva, or sputum (2).

A LB blood test is minimally invasive and complementary to needle biopsies of tissue, and it has the benefit to be more sensitive than conventional imaging (3). Moreover, collecting blood is a quick procedure that can be easily repeated in close intervals for any patients with cancer. The blood is a pool of tumor cells and products released into circulation, and LB can thus detect cells and cellular products shed from different sites (primary tumor and different metastatic sites). Thus, LB blood tests can monitor real-time disease progression and drug efficacy in individual patients with cancer (2).

Endometrial cancer (EC) is one of the most common invasive malignancies of the female genital tract, and despite the increasing incidence of EC worldwide and the poor overall survival of patients (4), there is no reliable blood-based biomarker to detect and track EC recurrence during routine follow-up. LB will play a key role in managing patients with EC in the coming years, mostly in the early diagnosis of this aggressive cancer, therapeutic targets and resistance mechanisms identification for the right therapy selection over time, minimal residual cancer detection, and drug efficacy assessment in real-time. Muinelo-Romay et al. provided an overview of the different circulating biomarkers suitable in patients with EC as opportunities toward precision medicine (e.g., circulating tumor cells, endothelial progenitor cells, ctDNA, exosomes, circulating microRNA) (5). There are many articles published on LB and EC, but in this editorial, we focus on recent data detecting ctDNA to assess its clinical relevance in this cancer type.

Using targeted next-generation sequencing, Moss et al. led a pilot study on a small cohort of 13 patients with EC to investigate the clinical relevance of ctDNA for detecting and monitoring EC recurrence and progression (6). The authors showed that the analysis of longitudinal plasma DNA revealed earlier detection of EC recurrence and dynamic kinetics of ctDNA reflected treatment response. However, caution is warranted for this preliminary study since a validation study in a larger cohort is needed to determine the lead time and its ability to identify local as well as distant recurrent disease. Shintani et al. detected tumor-related ctDNA before surgery and showed that it was associated with poorer clinical outcome on univariate analysis in patients with EC harboring PIK3CA or KRAS mutations (7).

More recently, Feng et al. analyzed the individual tumor genomes of high-risk EC and evaluated the sensitivity and specificity of this approach in long-term and dynamic follow-up (8). Using droplet digital PCR (ddPCR) assays specific to patient-specific mutations, ctDNA at baseline and in sequential plasma samples collected after surgery were tracked. The authors concluded that ctDNA was valuable in monitoring high-risk EC relapse during postoperative follow-up as a prognostic marker with better performance than traditional serum tumor markers. However, there were also limitations in this study: (i) ctDNA was not detected in all cases (mostly in early stage—International Federation of Gynecology and Obstetrics stage I to II), suggesting that ctDNA was not specific enough in cases of confined lesions but more suitable in cases of high tumor load; (ii) the sample size was very small, with a short follow-up time; and (iii) the genes included in the next-generation sequencing panel were not custom-made for EC, and thus some of the key genes related to EC might have been missed, potentially restricting the design of the ddPCR. As an alternative approach, Grassi et al demonstrated in their pilot study the feasibility of using personalized tumor-specific junction panels for detecting ctDNA in the plasma of EC patients (9).

In this issue of Clinical Chemistry, Beinse et al. describe the highly specific ddPCR detection of universally methylated circulating tumor DNA in EC (10). As previously mentioned, few previous studies showed that ctDNA could be detected in plasma from patients with EC. These studies were mainly based on mutation detection and thus limited by the mutational heterogeneity of EC and technical constraints that limited analytical sensitivity. In contrast, the approach of Beinse et al. is based on the analysis of cancer-associated DNA methylation patterns to overcome the mutational heterogeneity of EC. Highly sensitive ddPCR has emerged as a major tool to maximize ctDNA detection. The authors optimized a bioinformatic and experimental workflow that allowed the identification of DNA positions specifically methylated in almost all patients with EC, leading to a sensitive ddPCR assay for the robust detection and dynamic quantification of ctDNA in patients with EC.

In their study, Beinse et al. analyzed different cohorts of patients: (i) patients with stage I to IV EC, an independent cohort for validation of DNA positions found in silico as universally and specifically hypermethylated in EC (retrospective study); (ii) a prospective cohort of patients (n = 33) treated in the gynecological surgery department; (iii) an external cohort of healthy volunteers (n = 20); and (iv) an external cohort of patients with benign gynecology diseases (n = 30) (10).

Carefully checking the data, 3 DNA methylated positions were first identified as potential candidate positions hypermethylated in EC samples. Unfortunately, among the 3 identified targets, 1 showed background methylation levels in blood DNA methylation analyses (nearest gene: SIM1) and was excluded. The other DNA methylated positions were identified as located on OXT and ZSCAN12 promoters. Thus, despite the wide heterogeneity of EC, the authors identified 2 hypermethylated DNA positions shared by nearly all tumors: OXT and ZSCAN12. Importantly, OXT and ZSCAN12 were unmethylated in several noncancer cells/tissues (e.g., white blood cell), whole blood, and solid organs that potentially contribute to the release of circulating cell-free DNA. ddPCR analysis confirmed the high performance of combining OXT/ZSCAN12 methylation levels to classify EC samples from tumor-adjacent tissues in the external validation cohort (area under the curve = 0.99, 98.0% sensitivity at 97.1% specificity). Subsequent detection of circulating DNA by this methylation-specific assay in a prospective pilot cohort of 33 patients revealed positive findings in 14 of 31 (45%) patients in whom the first plasma was collected before surgery or chemotherapy. The number of early stage patients (FIGO I/II) was 16, and ctDNA could be detected in only 2 patients, limiting the use of this assay for early cancer detection. In contrast, in patients with advanced or relapsed disease, ctDNA was detected in 9 of 13 patients, consistent with the general observation in most tumor entities that patients with advanced cancer usually harbor higher ctDNA concentrations.

All samples without ctDNA detected were considered negative at ddPCR analytical sensitivity of ≤0.2% (target/albumin) with a lower limit of detection estimated to be ≤17.5 pg/mL of ctDNA. Thus, the high diagnostic sensitivity in EC and the high specificity vs normal tissues releasing circulating cell-free DNA allowed the identification of tumor-associated hypermethylated targets on ctDNA in plasma without biological background noise.

Interestingly, beyond blood plasma DNA, the high diagnostic specificity of OXT/ZSCAN12 could find application in urine or cervical scrapings, particularly for the purpose of cancer screening. The current study, however, still has some limitations. For example, the detection of ctDNA can be impaired by the fragmentation of circulating cell-free DNA due to the conversion temperature and by the inherent fragmentation of ctDNA itself. Moreover, the total number of patients analyzed was rather low, especially in the prospective validation cohort.

Overall, Beinse et al. reported a bioinformatic and technical framework allowing the development and validation of a highly analytically sensitive and specific ddPCR assay for ctDNA methylation detection in plasma from patients with EC (10). This is a promising approach for quantitative and dynamic ctDNA detection in patients with EC, particularly for patients with stage III/IV EC. Larger prospective studies are now required to validate the utility of this LB approach for the clinical management of EC.

The emerging field of ctDNA research has opened new avenues for cancer diagnostics over the past 10 years with important clinical opportunities for precision medicine in oncology. Broad clinical usage of the LB will depend on standardization of both preanalytic and analytic procedures. This task is currently ongoing by the members of the European Liquid Biopsy Society consortium. Indeed, standard operating procedures should be proposed and broadly validated. Finally, the synergy of multiple circulating biomarkers can reveal the specifics of a cancer: it is quantitative and qualitative information including information on epigenetics (11). Besides ctDNA, additional LB-based information can be obtained in patients with EC from the analysis of other circulating markers such as circulating tumor cells (12, 13), exosomes (14, 15), and tumor-educated platelets (16). The development of an algorithm that can combine all of these data into a composite biomarker profile is urgently needed and supported by recent advancements in artificial intelligence.

Author Contributions

All authors confirmed they have contributed to the intellectual content of this paper and have met the following 4 requirements: (a) significant contributions to the conception and design, acquisition of data, or analysis and interpretation of data; (b) drafting or revising the article for intellectual content; (c) final approval of the published article; and (d) agreement to be accountable for all aspects of the article thus ensuring that questions related to the accuracy or integrity of any part of the article are appropriately investigated and resolved.

Authors' Disclosures or Potential Conflicts of Interest

No authors declared any potential conflicts of interest.

References