Current Advances of Molecular Biomarkers and Liquid Biopsies Techniques in Ovarian Cancer

Jingxia Zhang: methodology, writing – original draft. Qiaoping Xu: project administration. Weiyan Shan: software. Jie Yu: visualization. Huan Shi: data curation. Hongkai Shang: writing – review and editing.

Despite the remarkable clinical potential of liquid biopsy technologies in the early diagnosis, therapeutic response monitoring, and recurrence surveillance of OC, the field is plagued by several unresolved methodological challenges that may compromise the accuracy and reliability of technology translation. At the study design level, small sample sizes (mostly < 100 cases [ 69 ], spectrum bias caused by over‐enrollment of advanced‐stage patients [ 184 , 185 ]), and the lack of long‐term follow‐up in retrospective studies tend to overestimate the diagnostic performance of assays and hinder the validation of their long‐term predictive value [ 69 ]. During analytical validation, most diagnostic models have not undergone external independent cohort validation [ 186 ], and some studies suffer from statistical over‐optimism due to uncorrected multiple comparisons, resulting in poor generalizability. In sample preprocessing, the confounding effects of blood collection tube types, transportation temperatures, processing timeframes, and patient comorbidities (e.g., infections and inflammation, which can elevate circulating free DNA levels) [ 186 ] are often overlooked. Furthermore, existing cohorts limited data from asymptomatic high‐risk individuals and the general population. To promote the standardized development and clinical translation of OC liquid biopsy technologies, future research should focus on the following directions: ① Adopt prospective cohort designs with expanded sample sizes (≥ 1000 cases), covering populations across different risk strata, and extend follow‐up duration to ≥ 5 years to fully capture long‐term clinical outcomes; ② Strengthen the analytical validation system, requiring all diagnostic models to be validated in at least one external independent cohort, with clear reporting of blinded study design and biomarker threshold setting methods; ③ Standardize sample preprocessing protocols, unify operating specifications for blood collection, transportation, and processing, and provide detailed disclosures in study reports; ④ Conduct head‐to‐head benchmarking studies to systematically compare the diagnostic efficacy and long‐term cost‐effectiveness of different liquid biopsy technologies against conventional detection methods; ⑤ Evaluate model stability using methods such as cross‐validation and Bootstrap resampling, and strictly control Type I errors in multiple comparisons to avoid statistical over‐optimism; ⑥ Enhance the representativeness of study cohorts by including individuals of diverse ethnicities, ages, and comorbidity profiles to improve the clinical applicability of research findings.

The authors have nothing to report.

This study was funded by the National Natural Science Foundation of China (grant no. 82171659); the Construction Fund for Key Medical Disciplines of Hangzhou (grant no. 2025HZZD07); the Zhejiang Provincial Medical and Health Science and Technology Program (grant no. 2023KY933); the Zhejiang Province Traditional Chinese Medicine Science and Technology Project (grant nos. 2023ZL565, 2022ZA139); and the Zhejiang Provincial Natural Science Foundation of China (grant no. LY21H040001).

To date, there is no established gold standard for OC screening, and the field remains fraught with controversies. Therefore, the clinical translation of liquid biopsy technologies for OC requires the construction of a structured screening system centered on risk stratification. High‐risk populations include carriers of BRCA1/2 mutations, individuals with a first‐degree relative diagnosed with OC, and carriers of other pathogenic mutations such as those underlying Lynch syndrome. Specifically, the cumulative lifetime risk of OC reaches up to 39%–45% for BRCA1 mutation carriers [ 174 ], 11%–16% for BRCA2 mutation carriers [ 175 ], 2%–7.5% for individuals with one affected first‐degree relative [ 176 ], and 8% for carriers of other mutations including Lynch syndrome [ 177 ]. Although the efficacy of screening in this high‐risk cohort is still under investigation, combined screening with protease activity assays plus ctDNA testing is recommended every 6 months. For the general population without high‐risk factors, the cumulative lifetime risk of OC is ≤ 2% [ 177 ]. Most international guidelines, including those from China and the National Comprehensive Cancer Network (NCCN), do not recommend routine screening [ 178 ]. Instead, annual exploratory assessment with serum CA‐125 combined with transvaginal sonography (TVS) is suggested, while ctDNA or protease activity assays are not recommended for routine application in this population. For high‐risk populations, the pre‐test probability is set at ≥ 5%, indicating that prior to biomarker‐based testing, the estimated likelihood of developing OC in this group is 5% or higher. The positive threshold for ctDNA is defined as a variant allele frequency (VAF) of TP53 mutations ≥ 0.05% [ 72 ]; for protease activity, the cutoff value is set at matrix metalloproteinase 3 (MMP3) activity ≥ 1.2 times the mean level in healthy individuals [ 148 ]. A positive screening result is determined if either of the two combined assays meets the respective threshold. For the general population, the pre‐test probability is set at ≤ 1.5%, reflecting a relatively low baseline cancer risk. The classic cutoff value of serum CA‐125 (35 U/mL) is retained [ 179 ], combined with TVS‐detected imaging abnormalities (ovarian mass with a diameter ≥ 2 cm accompanied by solid components) [ 180 ] as the positive criteria, to minimize unnecessary interventions arising from false‐positive results. Notable disparities exist in the cost and turnaround time across different detection modalities. Serum CA‐125 testing costs $50–80 per assay [ 181 ] with a turnaround time of 1–2 working days, making it suitable for initial screening of the average‐risk population. Targeted ctDNA sequencing costs approximately $1500–3000 per test and requires 7–10 working days for results [ 182 ], which is indicated for screening high‐risk populations and monitoring tumor recurrence. Protease activity assays using the G‐NBSs platform typically cost $800–1200 per test, with a turnaround time of 3–5 working days, but require fresh sample processing within 4 h [ 148 ], rendering it appropriate for combined screening in high‐risk groups. TVS costs $200–300 per examination and can be performed on‐site, serving as an imaging adjunct for all populations. Combined testing with multiple modalities costs $2500–4500 per session and is reserved exclusively for precision screening of high‐risk populations. As an emerging optical biosensing technology, the G‐NBSs protease activity assay platform has demonstrated potential for early OC screening in research settings. However, it has not yet obtained regulatory approval from the U.S. Food and Drug Administration (FDA) or the European Medicines Agency (EMA). To address regulatory hurdles, multi‐center consistency validation and standardization of fresh sample processing protocols (within 4 h) are imperative. Although ctDNA testing has been approved as a companion diagnostic, its current clinical application is primarily restricted to guiding targeted therapies (e.g., EGFR mutation‐directed treatments). For early OC screening, additional clinical validation data focusing on core target mutations (e.g., BRCA1/2 alterations) are urgently needed [ 32 ]. Both technologies face challenges including high costs (not covered by medical insurance) and insufficient ethnicity‐specific data, particularly for Asian populations, who account for approximately 51.8% of the global OC burden [ 183 ]. At present, these modalities remain in the clinical validation phase, and large‐scale multi‐center studies are required to accelerate their clinical translation.

In addition to the aforementioned CTCs and circulating tumor DNA (ctDNA), emerging liquid biopsy modalities include enzyme Activity‐based Liquid Biopsy (EALB), tumor‐educated platelets (TEPs), circular RNA (circRNA), exosomes, and microRNAs (miRNAs). While these technologies have yet to be validated in large‐scale prospective clinical trials due to limited sample sizes, suboptimal study designs, and a lack of standardized protocols for sample collection, processing, and analysis, they have demonstrated significant potential in early diagnosis and prognostic evaluation [ 142 ], real‐time monitoring of therapeutic response [ 143 ], and providing comprehensive molecular portraits of disease [ 144 ]. Figure  3 shows the sample sources, molecular biomarkers and clinical applications of liquid biopsy in OC. Molecular biomarkers and liquid biopsies in Ovarian Cancer. Enzyme Activity‐based Liquid Biopsy (EALB) represents a type of functional biomarker detection technology with great potential for clinical translation. This technology accurately reflects the dynamic functional state of tumor proliferation and invasion by specifically capturing abnormally activated proteases (such as matrix metalloproteinases, cathepsins, etc.) in the tumor microenvironment. It stands in sharp contrast to biomarkers like tumor‐educated platelets (TEPs) and circular RNAs (circRNAs), which only provide static molecular snapshots [ 145 ]. Its core advantage lies in its high sensitivity to early‐stage, low‐burden tumors. Even when the ctDNA concentration is extremely low in the early stages of the tumor, if the tumor is invasively active, the activity level of the proteases it secretes can already increase significantly, providing crucial functional evidence for early diagnosis [ 146 , 147 ]. More importantly, the tumor invasion potential indicated by protease activity can be combined with the gene mutation profile provided by ctDNA and the metastasis risk evaluated by CTCs to achieve a three‐dimensional comprehensive assessment of the tumor's “functional state + genotype + phenotype,” strongly supporting individualized treatment decisions. A study published by Covarrubias‐Zambrano O et al. in the journal Cells in 2025 further verified the clinical application value of this technology. The optical nanobiosensor based on few‐layer exfoliated graphene (G‐NBSs) they developed can specifically detect the activity of proteases such as MMP3 in serum by linking fluorescently labeled protease‐specific sequences and tetracarboxyphenyl porphyrin (TCPP). The detection limit of this detection system reaches the sub‐attomolar level (< 10 −15  mol/L), and the detection accuracy for early‐stage OC is 94.5%, with both sensitivity and specificity being 0.94 [ 148 ]. Although this technology demonstrates significant advantages, it currently still faces key challenges such as insufficient standardization of the sensor array and stringent requirements for sample handling (fresh samples are required to avoid protease activity degradation). Overall, it is in the early stage of clinical translation, and its widespread clinical application still requires further large‐sample validation and technical optimization. TEPs are platelets “educated” through interactions with tumor cells. By absorbing and storing tumor‐associated molecules, including coding/noncoding RNAs and proteins, they acquire tumor‐specific molecular signatures [ 149 ]. TEPs contribute to the progression and dissemination of various solid tumors. Splicing‐derived TEP RNA signatures can provide specific information about cancer presence, location, and molecular characteristics [ 150 ]. In OC diagnosis, TEPs exhibit high accuracy. One study reported an 80% accuracy rate in distinguishing malignant from benign ovarian tumors using TEP‐based scoring [ 151 ]. In an international hospital‐based diagnostic study, 761 patients with histologically confirmed adnexal masses undergoing primary treatment (enrolled from September 2016 to May 2018 across nine medical centers) and 167 healthy controls were analyzed. In validation cohorts from China (VC1, VC2) and Europe (VC3), the performance of TEPs alone and combined with CA125 was as follows: TEPs alone: Area under the curve (AUC) values were 0.918 (combined cohorts), 0.923 (VC1), 0.918 (VC2), and 0.887 (VC3). TEPs+CA125: AUC values improved to 0.922 (combined), 0.955 (VC1), 0.939 (VC2), and 0.917 (VC3). Subgroup analyses further demonstrated the utility of TEPs in detecting early‐stage disease (AUC = 0.858), borderline tumors (AUC = 0.859), nonepithelial malignancies (AUC = 0.920), and differentiating OC from endometriosis (AUC = 0.899) [ 68 ]. Despite their significant potential, the clinical application of tumor‐educated platelets (TEPs) faces several challenges and limitations. First, the mechanisms underlying platelet “education” by tumor cells remain incompletely understood. Further research is required to elucidate how tumor cells reprogram platelets and the specific molecular pathways governing RNA splicing within TEPs [ 150 , 152 ]. Second, TEP‐derived signals may be contaminated by nontumor cells, compromising specificity. Studies have shown that transcripts detected in TEPs could originate from nonplatelet cells, necessitating computational noise‐filtering algorithms to improve signal accuracy [ 152 ]. Finally, TEP detection relies on specialized technologies and infrastructure, such as high‐throughput sequencing and sophisticated bioinformatics pipelines, whose limited accessibility and high costs may hinder widespread clinical adoption [ 153 ]. Circular RNAs (circRNAs) are a unique class of noncoding RNA molecules produced through back‐splicing of precursor mRNAs, forming a covalently closed loop structure [ 154 ]. Unlike linear RNAs, circRNAs exhibit greater stability and tissue specificity [ 155 , 156 ]. They exist as single‐stranded molecules and are detectable in various bodily fluids, including blood, saliva, and urine, and particularly ascites—a biofluid uniquely abundant in OC patients [ 156 , 157 ]. Their high stability and accessibility in biofluids position circRNAs as promising tools for OC diagnosis, monitoring, prognosis, and therapeutic guidance [ 158 ]. Significant progress has been made in circRNA‐based liquid biopsy for OC. In plasma, several circRNAs have been identified as robust diagnostic biomarkers. For example, plasma circN4BP2L2 was significantly downregulated in EOC patients, with an area under the receiver operating characteristic curve (AUC) of 0.82 for distinguishing EOC from benign ovarian cysts and 0.90 for distinguishing EOC from healthy controls; notably, it could also differentiate early‐stage EOC from benign (AUC = 0.81) and healthy (AUC = 0.90) cohorts [ 159 ]. Another plasma circRNA, circBNC2, was also found to be downregulated in EOC, achieving an AUC of 0.879 for EOC versus benign tumors (sensitivity 96.4%, specificity 80.7%) and 0.923 for EOC versus healthy controls (sensitivity 95.2%, specificity 85.5%). Importantly, for early‐stage (FIGO I–II) EOC, circBNC2 yielded an AUC of 0.864 versus benign and 0.908 versus healthy controls, outperforming both CA125 and HE4 [ 160 ]. A plasma circRNA microarray profiling study further identified a two‐circRNA diagnostic model (circCOMBO: hsa_circ_0003972 and hsa_circ_0007288) with an AUC of 0.781, and when combined with CA125, the AUC improved to 0.923 [ 161 ]. A unique advantage of circRNA liquid biopsy in OC lies in the utilization of ascites fluid, which contains a high concentration of tumor‐derived molecules. A systematic review of RNA biomarkers in biofluids for early OC diagnosis highlighted ascites as a promising source for extracellular RNA‐based signatures, with multiple studies exploring the transcriptome of ascites from OC patients [ 157 ]. Despite their promising potential, challenges and opportunities remain for circRNA‐based liquid biopsy in OC. First, the sensitivity and specificity of circRNA detection methods require further refinement, particularly for low‐abundance circRNAs in peripheral blood [ 162 ]. Second, the mechanistic roles and biological functions of most dysregulated circRNAs in OC remain incompletely understood [ 163 ]. Additionally, translating circRNA research into clinical practice faces OC‐specific challenges: the high heterogeneity of ascites fluid across patients complicates standardized extraction protocols, and there is a lack of large‐scale prospective validation studies specifically in EOC populations [ 157 ]. Exosomes are nanoscale extracellular vesicles secreted by various cell types, with diameters typically ranging from 30 to 150 nm. They are present in diverse bodily fluids, including blood, urine, and ascites [ 164 , 165 ], and carry bioactive molecules such as proteins, nucleic acids (e.g., miRNAs and long noncoding RNAs), and lipids. Exosomes mediate intercellular communication and participate in numerous physiological and pathological processes [ 164 , 166 ]. Due to their high stability, low immunogenicity, minimal cytotoxicity, strong tissue penetration potential, and homing capability, exosomes are recognized as effective carriers for immunotherapeutic agents [ 164 , 167 ]. Elevated expression levels of miR‐1290 and miR‐1307 in plasma‐derived exosomes have been identified as potential biomarkers and therapeutic targets for OC [ 165 ], offering utility in diagnosis, treatment, and prognosis evaluation [ 164 ]. Beyond these, a growing list of exosomal cargoes has been investigated in OC, including miRNAs such as miR‐21, miR‐200b, and miR‐205 [ 168 ], as well as proteins like EpCAM and CD24 [ 169 ]. Accumulating evidence indicates that miR‐1290 functions as an oncogenic miRNA in OC: it is significantly upregulated in OC tissues, cell lines and patient serum exosomes, and its high diagnostic accuracy (AUC = 0.794) may reflect its role in OC progression, suggesting its potential as a prognostic indicator [ 170 , 171 ]. Mechanistically, miR‐1290 has been shown to promote OC cell proliferation and invasion, with studies reporting its targeting of OGN in cancer‐associated fibroblasts, leading to activation of the mTOR/Akt pathway [ 172 ]. Furthermore, miR‐1290 can inhibit SOCS4 expression in OC [ 170 ]. Since SOCS family proteins negatively regulate JAK/STAT signaling, the suppression of SOCS4 may lead to the activation of the JAK/STAT3 pathway [ 173 ]. These findings suggest that miR‐1290 may play a critical role in cancer progression and could serve as a diagnostic marker or therapeutic target. Despite their promise, the clinical application of exosomes in OC faces several challenges: First, the translation of exosome‐based concepts into clinical practice remains unrealized [ 164 ]. Second, the precise mechanisms and functional roles of exosomes require further investigation and validation. Third, the stability and detectability of exosomes across different biofluids need optimization to enhance reproducibility [ 165 ].

The authors have nothing to report.

In accordance with the position of authoritative bodies including the British Gynaecological Cancer Society (BGCS) [ 18 ], there is currently no fully validated, mature OC screening strategy suitable for asymptomatic general populations [ 19 ]. The most common methods studied for OC screening are transvaginal ultrasonography, bimanual palpation, and measurement of the serum tumor marker CA‐125 [ 19 ]. Among them, CA‐125 is recognized as an optimal and widely utilized biomarker for the early clinical detection of OC. Nevertheless, it has a low predictive value (5%) and lacks specificity and sensitivity. A meta‐analysis conducted by Ferraro et al. [ 20 ] reported a specificity of 78% for CA‐125 in detecting OC. Only 50% of early‐stage patients have elevated CA‐125 levels, resulting in a relatively low sensitivity (50%–62%) for the detection of early‐stage OC. Moreover, CA‐125 levels may increase in other diseases, such as endometriosis, breast cancer, and lung cancer [ 21 ]. In the PLCO trial, the positive predictive value (PPV) of CA‐125 (> 35 U/mL) was determined to be 2.6% [ 22 ]. TVS is utilized to examine the structures of the adnexa and uterus. Although it provides some utility in the diagnostic process, its sensitivity and specificity are limited. Small lesions on the ovaries and the fimbriae of the fallopian tubes cannot be adequately visualized by TVS. In the PLCO trial, the PPV of TVS was found to be as low as 1% [ 22 ]. The specificity limitations of CA‐125 need to be supplemented by combining it with imaging techniques, and transvaginal sonography (TVS) can assist in differential diagnosis through structural visualization. Additionally, surgical exploration followed by pathological examination under a microscope is regarded as the gold standard for the definitive diagnosis of OC. However, surgical exploration, which relies on visual and tactile assessment, carries the risk of missing subtle lesions due to its inherent limitations. Moreover, its invasive nature and associated complications restrict its applicability in certain clinical scenarios [ 23 ]. Similarly, the accuracy of pathological examination is highly dependent on the quality of the sampled tissue. Sampling errors and time delays may compromise the results. Furthermore, significant intratumoral heterogeneity often exists within ovarian tumors, and a single‐site biopsy sample may not fully represent the characteristics of the entire tumor. This approach lacks the ability to adequately capture tumor heterogeneity and the evolution of the cancer genome, potentially leading to incomplete or biased diagnostic conclusions [ 24 ]. These limitations underscore the need for more comprehensive and precise diagnostic strategies to improve the accuracy and reliability of OC diagnosis. These limitations of conventional detection modalities have propelled the application of liquid biopsy technologies in the detection and monitoring of OC. Figure  1 summarizes the advantages and challenges of two core liquid biopsy technologies: ctDNA and CTCs. In contrast, liquid biopsy has the potential to capture molecular information derived from the entire tumor tissue, thereby overcoming, in certain cases, the challenges posed by tumor heterogeneity (or multiple metastatic lesions) [ 26 ]. Additionally, it enables real‐time monitoring of disease progression and treatment response through repeated sampling over short intervals. Compared to traditional tissue biopsy, liquid biopsy offers a noninvasive, rapid, and comprehensive approach, holding promise to usher in a new era in clinical oncology [ 13 ]. Schematic of clinical examination and exploratory diagnostic approaches for ovarian cancer. (1) Routine clinical examinations: ① Imaging examination: As a primary screening modality, transvaginal ultrasound facilitates lesion localization. ② Pathology examination: Pathological assessment of tissues collected from surgery or biopsy is universally recognized as the gold standard for definitive diagnosis. ③ Serum tumor marker detection: Serum CA‐125, a nonspecific biomarker, is widely used for initial screening and longitudinal follow‐up of disease progression. (2) Exploratory diagnostic approaches: ④ Liquid biopsy: Liquid biopsy platforms, including CTCs, TEPs, exosomes and ctDNA, are mainly applied for post‐treatment surveillance and are under exploration for early diagnosis. At present, liquid biopsy has not been adopted in routine clinical diagnosis, as it remains in translational and preclinical research [ 25 ]. Additionally, the clinical utility of ctDNA has not been fully confirmed and is yet to be widely used in standard clinical settings.

Ovarian cancer (OC), a malignant tumor originating from ovarian tissue, represents the most lethal gynecological malignancy [ 1 ]. Worldwide, global statistics from 2022 reported 324,603 new cases and 206,956 deaths, with Asia bearing the highest burden (49.2% of incident cases and 56.1% of mortality worldwide) [ 2 ]. Projected data indicate that the United States will have 20,890 new OC cases and 12,730 OC‐related deaths in 2025 [ 3 ]. Histologically, OC is primarily classified into epithelial tumors, sex cord‐stromal tumors, and germ cell tumors, with epithelial ovarian cancer (EOC) accounting for 95% of malignant cases [ 4 ]—one of the most challenging and lethal types in gynecological oncology. The World Health Organization (WHO) further categorizes EOC into five major subtypes based on epithelial morphology: high‐grade serous carcinoma (HGSC, 70%), endometrioid carcinoma (EC, 10%), clear cell carcinoma (CCC,10%), mucinous carcinoma (MC, 3%), and low‐grade serous carcinoma (LGSC, 3%) [ 5 ]. Although 90% of OC patients can be cured if detected early [ 6 ], the disease remains difficult to diagnose at an early stage due to nonspecific initial symptoms (e.g., abdominal distension, pain) that mimic gastrointestinal disorders. Consequently, approximately 80% of patients present with metastatic disease at the time of diagnosis [ 7 ], a key factor contributing to its high mortality rate. According to the National Comprehensive Cancer Network (NCCN) Guidelines (Version 1.2024), OC management encompasses a multi‐modal therapeutic arsenal. Current standard‐of‐care options include cytoreductive surgery, platinum‐based combination chemotherapy with paclitaxel, hyperthermic intraperitoneal chemotherapy (HIPEC), targeted therapy with bevacizumab, and PARP inhibitors. Emerging immunotherapeutic agents are still under clinical evaluation and have not been established as standard of care. Notably, while many patients demonstrate favorable initial responses to treatment, the majority of advanced OC cases relapse within years, with the underlying cause lying in the development of acquired drug resistance to chemotherapy and targeted agents [ 8 ]. Intrinsic resistance, defined as the inherent ability of cancer cells to withstand initial therapy, typically leads directly to primary treatment failure and early tumor relapse. In contrast, acquired resistance, which develops gradually via genetic or epigenetic alterations under therapeutic pressure, is closely associated with late tumor recurrence [ 9 ]. These two resistance mechanisms together constitute another pivotal factor driving the disease's high recurrence and mortality rate [ 10 , 11 ]. Thus, early diagnosis and treatment monitoring represent critical strategies to reduce OC mortality. In this context, liquid biopsy, an emerging molecular analytical approach in oncology, offers a novel paradigm for OC management. Broadly defined, “liquid biopsy” refers to the sampling and analysis of various biological fluids‐predominantly blood, but also urine, ascites, or pleural effusion—for tumor‐derived components [ 12 ]. By identifying and analyzing DNA fragments released from tumor cells or apoptotic tumor cells into the bloodstream, liquid biopsy provides detailed insights into tumor characteristics, including progression, staging, heterogeneity, genetic mutations, and clonal evolution [ 13 , 14 ]. This technology not only demonstrates substantial potential in early tumor screening and targeted therapy but also serves as a pivotal tool for detecting MRD, predicting recurrence, and capturing real‐time drug resistance—thereby enabling early disease monitoring and personalized treatment optimization [ 15 ]. Key biomarkers analyzed via liquid biopsy include circulating tumor DNA (ctDNA), circulating tumor cells (CTCs), extracellular RNA (e.g., microRNA), and extracellular vesicles [ 16 ]. Among these, ctDNA and CTCs are regarded as the two primary components of liquid biopsy [ 17 ], forming the focus of this review. Notably, protease activity‐based liquid biopsy serves as a functional complement: it reflects tumor invasive potential by detecting the aberrant activation of tumor‐associated proteases, which differentiates it from conventional tumor burden‐related biomarkers such as ctDNA and CA‐125. Here, we summarize the clinical applications of liquid biopsy in OC, encompassing early diagnosis, prognostic evaluation, recurrence detection, treatment response monitoring, and recent technological advancements.

The authors declare no conflicts of interest.