Table 3.
Application of liquid biopsy in other samples.
| Category of Liquid Biopsy | Biomarkers | Detection Method | No. of Participants (EC/Control) |
Clinical Significance/Findings/Accuracy | Author and Year |
|---|---|---|---|---|---|
| Uterine lavage fluid/uterine aspirates | |||||
| cfDNA, CTCs | PTEN, PIK3CA, TP53, CTNNB1, KRAS, etc. | NGS, ddPCR, CellSearch system | 60 EC | Genetic alterations were detected in 93% of EC through UAs. ctDNA was associated with high-risk tumors and disease progression. | Casas-Arozamena et al., 2020 [159] |
| cfDNA | BAT26, BAT25, NR24, NR21, Mono27 | ddPCR | 90 EC | A high concordance (96.67%) between MSI determinations in cfDNA and the standard of care was confirmed. | Casas-Arozamena et al., 2023 [160] |
| cfRNA | miR-146a-5p, miR-183-5p, miR-429 | Real-time PCR | 42/40 | miR-146a-5p, miR-183-5p, and miR-429 were significantly upregulated in EC. AUC: miR-183-5p: 0.675, miR-429: 0.709, miR-146a-5p: 0.685 |
Yang et al., 2023 [161] |
| Cervicovaginal fluid/cervicovaginal lavage | |||||
| Metabonomics | Phosphocholine, malate, asparagine | NMR spectroscopy | 21/33 | Metabolomic biomarkers in CVF for non-invasive detection of EC were identified and validated using ML algorithms. AUC: [training: 0.88–0.92; test: 0.75–0.80]; sensitivity (95% CI): forests: 0.75 (0.19–0.99); specificity (95% CI): forests: 0.80 (0.28–1.00) |
Cheng et al., 2019 [162] |
| Cytology | Malignant endometrial cells | Cytological analysis | 103/113 | Vaginal cytology demonstrated higher sensitivity (90.2%) compared to urine cytology (72.0%) but lower specificity. Sensitivity: [vaginal: 90.2%, urine: 72.0%, combined: 91.7%]; specificity: [vaginal: 88.7%, urine: 94.9%, combined: 88.8%] |
O’Flynn et al., 2021 [163] |
| Proteomics | 72 proteins (TIM-3, VEGF, TGF-α, IL-10, CA19–9, CA125, etc.) | Multiplex immunoassays | 66/126 | Identified lavage proteins could discriminate EC from benign conditions. AUC (95% CI): combined: 0.91 (0.78–0.97) Sensitivity: 86.1% (combined); specificity: 87.9% (combined) |
Łaniewski et al., 2022 [164] |
| Metabolomics and proteomics | Amino acid and nucleotide metabolism biomarkers | LC-MS/MS | 44/43 | Urine/intrauterine brushing metabolites correlate with tissue pathways (amino acid/nucleotide metabolism). AUC: 0.808 (urine) 0.847 (intrauterine brushing); Sensitivity: urine: 74.7% (top 5 metabolites) |
Yi et al., 2022 [165] |
| Somatic mutations | 47 genes panel (POLE, TP53, PTEN, etc.) | NGS | 139/107 | POLE mutations indicated excellent prognosis; TP53 mutations were associated with significant DFS differences among molecular subtypes. AUC: 0.83 (self-collected); sensitivity: 73% (clinician and self-collected); specificity: [80% (clinician-collected), 90% (self-collected)] |
Pelegrina et al., 2023 [166] |
| DNA methylation | ZSCAN12, GYPC | WID-qEC | 12/375 | WID-qEC test demonstrated superior diagnostic accuracy compared to transvaginal ultrasound in detecting uterine cancers. AUC (95% CI): 0.943 (0.847–1.000); sensitivity (95% CI):90.9% (62.3–98.4); specificity (95% CI): 92.1% (88.9–94.4) |
Evans et al., 2023 [167] |
| Proteomics | SERPINH1, VIM, TAGLN, PPIA, CSE1L, CTNNB1 | MS | 22/19 | Six protein biomarkers in cervical fluids were identified to distinguish women with abnormal uterine bleeding who are EC and those who are non-EC. AUC: [UF: > 0.71, LDHA, ENO1, PKM: > 0.9; M1: up to 0.83 (SERPINH1); M3: up to 0.84 (TAGLN)]; sensitivity: [M1: up to 83%; M3: up to 89%]; specificity: [M1: up to 81%; M3: up to 78%] |
Martinez-Garcia et al., 2023 [168] |
| DNA methylation | ZSCAN12, GYPC | WID-qEC | 28/74 | The WID-qEC test reliably detected uterine cancers (endometrial and cervical) across sampling devices and collection methods (gyn. vs. patient self-sampling). AUC (95% CI): 0.96 (0.91–1.00); sensitivity: 92.9% (gyn), 75.0% (self); specificity: 98.6% (gyn), 100.0% (self) |
Illah et al., 2024 [169] |
| DNA methylation | CDO1m, CELF4m | qMSP | 21/275 | Dual-gene methylation showed high sensitivity (85.7%) and specificity (87.6%) for EC screening. AUC (95% CI): 0.867 (0.788–0.946) for dual methylation; sensitivity (95% CI): 85.7% (0.707–1.000); specificity: 87.6% (0.837–0.915) |
Zhao et al., 2024 [170] |
| DNA methylation | CDO1, CELF4 | qPCR | 40/98 | Combined test specificity (95.9%) outperformed transvaginal ultrasound (ET) and CA125 and detected all Type II EC cases. AUC (95% CI): 0.917 (0.853–0.91) for combined test; sensitivity (95% CI): 87.5% (73.2–95.8); specificity: 95.9% (89.9–98.9) |
Cai et al., 2024 [171] |
| Proteomics | HPT, LG3BP, FGA, LY6D, IGHM | SWATH-MS | 53/65 | Cervico-vaginal fluid protein signatures showed superior accuracy over plasma in detecting Stage I EC and advanced tumors effectively. AUC (95% CI): [cervico-vaginal: 0.95 (0.91–0.98), plasma: 0.87 (0.81–0.93)]; sensitivity: [cervico-vaginal: 91% (83–98%), plasma: 75% (64–86%)]; specificity: [cervico-vaginal: 86% (78–95%), plasma: 84% (75–93%)] |
Njoku et al., 2024 [172] |
| Proteomics | Angiopoietin-2, endoglin, FAP, MIA, VEGF-A | Multiplex immunoassays | 66 EC/108 benign | Five key biomarkers were significantly elevated in EC. The multivariate model showed prognostic value for tumor grade, size, invasion, and MMR status. AUC: 0.918; sensitivity: 87.8%; specificity: 90.7% |
Harris et al., 2024 [173] |
| Metabolomics | Lipids, amino acids, and other metabolites | UPLC-MS | 66/108 | Metabolic dysregulation was linked to tumor characteristics (size, myometrial invasion); noninvasive detection and risk stratification improved; multivariate models achieved high diagnostic accuracy. AUC: 0.800–0.951 (25-feature model); sensitivity: 78.6% (for EC); specificity: 83.3% for EC, 79.6% for benign |
Lorentzen et al., 2024 [174] |
| Tampons | |||||
| DNA methylation | 28 Methylated DNA markers | qMSP | 100/92 | The sensitivity to detecting EC was high even when vaginal fluid samples were collected before endometrial sampling. AUC (95% CI): 0.91 (0.85–0.97); sensitivity (95% CI):82% (70–91%); specificity (95% CI): 96% (87–99%) |
Bakkum-Gamez et al., 2023 [175] |
| Cervical scrapings and vaginal swabs | |||||
| Genomic DNA | 100 EC-related genes | NGS | 39/11 | Cervical swab-based gDNA genomic data demonstrated enhanced detection ability and enabled patient classification. Sensitivity: 67%; specificity: 100% |
Kim et al., 2022 [176] |
| DNA methylation | BHLHE22, CDO1 | MPap | 494 EC | MPap test showed high sensitivity and specificity for EC detection. AUC (95% CI): [Stage 1: 0.91 (0.87–0.94), Stage 2: 0.90 (0.84–0.95)]; sensitivity (95% CI): [Stage 1: 92.9% (80.5–98.5%), Stage 2: 92.5% (82.9–100.0%)]; specificity (95% CI): [Stage 1: 71.5% (64.8–77.5%), Stage 2: 73.8% (67.6–79.4%)] |
Wen et al., 2022 [177] |
| DNA methylation | GYPC, ZSCAN12 | qPCR | 562 (various groups) | The WID-qEC test offered a non-invasive EC screening and triage with high sensitivity and specificity. AUC: 0.94 (Barcelona); sensitivity: [97.2% (FORECEE), 90.1% (Barcelona), 100% (PMB cohort)]; specificity: [75.8% (FORECEE), 86.7% (Barcelona), 89.1% (PMB cohort)] |
Herzog et al., 2022 [178] |
| DNA methylation | ADCYAP1, BHLHE22, CDH13, CDO1, GALR1, GHSR, HAND2, SST, ZIC1 | qMSP | 103/317 | DNA methylation marker analysis in urine, cervicovaginal self-samples, and clinician-taken cervical scrapes achieved high diagnostic accuracy for EC detection. AUC: [urine: 0.95, self-samples: 0.94, scrapes: 0.97]; sensitivity: [urine: 90%, self-samples: 89%, scrapes: 93%]; specificity: [urine: 90%, self-samples: 92%, scrapes: 90%] |
Wever et al., 2023 [179] |
| DNA methylation | RASSF1A, HIST1H4F | qPCR | 19/75 | Methylation levels of RASSF1A/HIST1H4F increased with endometrial lesion severity. AUC: RASSF1A: 0.938; HIST1H4F: 0.951 |
Wang et al., 2024 [180] |
| Other samples | |||||
| cfDNA | KRAS, PIK3CA | NGS, qPCR | 50/7 | KRAS/PIK3CA mutations were detected in 47.4% of peritoneal lavages and correlated with tumor tissue. | Mayo-de-las-Casas et al., 2020 [181] |
| Metabolomics and proteomics | SOAT1, CE | ELISA, colorimetric assay, RT-qPCR, IHC | 32/16 | SOAT1 and CE may be associated with malignancy, aggressiveness, and poor prognosis. AUC: peritoneal fluid SOAT1: 0.767; sensitivity: 80%; specificity: 67% |
Ayyagari et al., 2023 [182] |