ctDNA carries genetic mutations and can be identified in matched tumors from individuals which are specific to the tumor. ctDNA is a component of circulating free DNA (cfDNA) that enters the bloodstream of cancer patients, exhibiting a half-life of less than two hours (74). Comprising short DNA fragments typically ranging from 150 to 200 base pairs (75). Epigenetic or genetic changes, such as tumor-specific methylation markers, rearrangements, copy number variations, and somatic point mutations, allow for the distinction of ctDNA from normal cfDNA fragments (76). The modifications observed in the genetic and epigenetic profiles of ctDNA can mirror the genomic or epigenomic traits of the cells from which they are derived. This characteristic enhances their application in assessing prognosis, identifying residual disease, selecting treatment strategies, and predicting the risks of potential recurrence (77).Studies indicate that the ctDNA content within total cfDNA detected in patient plasma can be as low as 0.01% (78). Consequently, methods employed to detect and characterize ctDNA in cfDNA must possess extremely high sensitivity. Furthermore, the quantity of ctDNA in patient serum varies not only between individuals but also according to tumor type.

Ovarian cancer is frequently diagnosed at an advanced stage due to its subtle symptoms, with approximately 70% of patients receiving a diagnosis only in the later stages, which contributes to a poor prognosis (117). Thus, early and accurate diagnosis is essential for effective treatment of ovarian cancer. Currently, the FDA has approved only two tumor markers, CA-125 and human epididymis protein 4 (HE-4), for diagnostic purposes; however, their sensitivity and specificity are relatively low, rendering them inadequate for early screening. With the ongoing advancements in molecular biology technologies, research on the application of ctDNA in the early diagnosis of ovarian cancer has been on the rise.

In ovarian malignant tumors, ctDNA serves as a crucial biomarker for tumor growth, metastasis, treatment response, and recurrence (118). Research indicates that tracking the temporal heterogeneity of ctDNA allows for the determination of evolutionary pathways in ovarian cancer, even at resolutions below the exome scale (119). The current research focuses on multiple ctDNA analysis strategies. During tumorigenesis and progression, hypermethylation of CpG islands within gene promoters can suppress transcription, silence tumor suppressor genes, and activate oncogenes, ultimately promoting malignant tumor transformation (120). DNA methylation represents an early event in tumorigenesis, and the analysis of ctDNA methylation patterns has emerged as a promising method for diagnosing early-stage ovarian cancer (7). However, the clinical utility of individual studies must be evaluated in the context of their validation status and technical reproducibility.

Widschwendter et al. (8) investigated the application of ctDNA in the early diagnosis of ovarian cancer. By monitoring the methylation status of COL23A1, C2CD4D, and WNT6, the researchers found that the diagnostic rate for early-stage ovarian cancer was 23%, compared to 15.4% for the traditional CA125 testing method, with a specificity of 97%. Singh et al. (9) employed the MethyLight method to analyze 120 tissue samples (85 tumor samples and 35 healthy controls) and 70 matched serum samples (45 tumor samples and 25 healthy controls). The study found that promoter methylation of the tumor suppressor genes HOXA9 and HIC1 in ctDNA effectively distinguished epithelial ovarian cancer from healthy controls, yielding an area under the curve (AUC) value of 0.95, a sensitivity of 88.9%, and a specificity of 100%. In contrast, Dong et al. (10) included 36 ovarian cancer patients and 25 healthy controls in their study, utilizing PCR technology to detect the methylation status of the SLIT2 promoter in ctDNA from peripheral blood in both groups. The results indicated that the incidence of SLIT2 promoter methylation was 75% in ovarian cancer patients, while no SLIT2 promoter methylation was detected in the control group, highlighting challenges in ​​analytical consistency​​.

Additionally, Li et al. (11) screened 493 CpG sites from over 3.3 million cfDNA methylation markers based on methylation differences and P-values, ultimately constructing the MethylBERT-EOC diagnostic model. This model demonstrated significant diagnostic efficacy for early-stage (I-II) epithelial ovarian cancer, achieving an AUC of 0.95, a sensitivity of 79.45%, and a specificity of 94.34%. Notably, when combined with CA125, the model’s diagnostic sensitivity for early-stage epithelial ovarian cancer increased from 82.51% to 89.62%, while maintaining a specificity of 93.55%. This represents a significant advancement in computational complexity and potential accuracy. However, the high development costs and the need for sophisticated bioinformatics infrastructure, in comparison to simpler PCR-based methylation assays, may restrict its immediate widespread application in clinical settings.

Cohen et al. (12) conducted a study involving 32 patients with high-grade serous ovarian cancer and 32 healthy controls for ctDNA sequencing. The findings revealed that genomic copy number variations were detected in 40.6% of ovarian cancer patients, with 38% of these being early-stage patients; conversely, only 6.25% of the healthy control group exhibited detectable plasma ctDNA copy number variations. The overall sensitivity of this detection method was approximately 41%, with a specificity of around 94%. Further analysis indicated a sensitivity of 33% for stage I/II patients and 44% for stage III/IV patients. Moreover, recent studies have shown that combining ctDNA concentration with integrity metrics significantly enhances the early diagnostic efficacy of ovarian cancer (13). This method is technically mature and cost-effective, making it suitable for targeted therapy selection. However, its utility in population screening is limited due to suboptimal sensitivity.

In addition, the multi-component integration strategy also shows great promise. The CancerSEEK method (14) facilitates early screening for eight common tumor types, including ovarian cancer, by detecting mutations in 16 genes within ctDNA and integrating this with an assessment of eight cancer-related proteins present in the patient’s blood. This method exhibits the highest sensitivity for ovarian cancer detection, reaching 98%, with a specificity exceeding 99%, thus significantly mitigating the risk of false-positive results. However, further research is necessary to evaluate the clinical applicability of this method and similar approaches. However, the translation of these findings into clinically available testing methods still requires large-scale prospective validation in both high-risk and general populations. Furthermore, ctDNA, as a novel biomarker, demonstrates commendable sensitivity and specificity in the diagnosis of ovarian cancer. When utilized alongside traditional biomarkers, it not only enhances diagnostic sensitivity but also preserves specificity, thereby facilitating the early detection of tumors. Existing evidence indicates that methylation-based detection achieves the best balance between sensitivity, specificity, and technological maturity, making it the most promising direction for early diagnosis at present. Future challenges lie in the large-scale prospective validation to transition it from a research tool to a standardized clinically available test, and to demonstrate its cost-effectiveness for screening in both high-risk and general populations.

The assessment of treatment response in ovarian cancer has traditionally relied on imaging and CA125 levels; however, these methods often exhibit delays and limitations.In ovarian cancer, changes in ctDNA concentration can dynamically reflect disease progression and have been shown to outperform traditional CA125 measurements (35). Furthermore, research indicates that ctDNA dynamics are associated with ovarian cancer patients’ responses to adjuvant chemotherapy and may predict treatment responses or disease progression (121).Research has shown that the dynamic changes in ctDNA levels are sensitive indicators for assessing treatment response. Over 96% of patients with high-grade serous ovarian cancer (HGSOC) exhibit alterations in the TP53 gene, which is recognized as one of the most prevalent mutations in cancer (17, 18). Neoadjuvant chemotherapy (NACT), a treatment modality for cancer, impacts genes involved in the TP53 signaling pathway (19). These mutations can serve as customizable markers for monitoring tumor growth and early changes, as well as for predicting response time and time to progression (TTP) (20). Analysis of ctDNA indicates that TP53 mutations are valuable biomarkers for predicting responses to first-line therapy and for dynamically monitoring treatment responses in patients with HGSOC (21). Kim et al. (22) monitored TP53 gene mutations in ctDNA before and after surgery in 28 patients with high-grade serous ovarian cancer. The results demonstrated that in patients who achieved satisfactory tumor debulking during surgery, TP53 allele mutations decreased by 93%, whereas in patients with residual disease post-surgery, the decrease was only 74%. In patients with HGSOC, there is a baseline correlation between ctDNA levels and disease burden. However, a shorter TTP is associated with a decrease in TP53 mutation allele frequency (MAF) of ≤60% following a single treatment cycle (23). This indicates that patients who may respond poorly to treatment can be identified very early in the therapeutic process.

High-sensitivity detection technology is the cornerstone for achieving precise monitoring. dPCR has become the ideal choice for longitudinal monitoring of known biomarkers due to its high specificity and sensitivity for specific mutations, such as TP53. On the other hand, NGS allows for a broader genomic analysis. For instance, the study conducted by Oikkonen et al. not only monitored the reduction in mutation frequency but also identified an increase in the frequency of TP53 mutations in the ctDNA of relapsed patients, effectively capturing the emergence of resistant clones (24).

These results highlight the significant role of plasma ctDNA testing in evaluating the efficacy of treatment for ovarian cancer. For patients with ovarian cancer, timely assessments of treatment efficacy are crucial for selecting appropriate subsequent treatment regimens. Despite its immense potential, the core obstacle to integrating ctDNA monitoring into routine clinical practice lies in the lack of standardized molecular response thresholds.

BRCA1/2 germline mutations are the most significant genetic risk factors for epithelial ovarian cancer, occurring in 6–15% of women diagnosed with this disease. The treatment of ovarian cancer has entered the era of precision medicine, particularly with the application of poly (ADP-ribose) polymerase (PARP) inhibitors, which has profoundly changed the treatment landscape for patients with BRCA mutations. In clinical practice, PARP inhibitors are employed as maintenance therapy for ovarian cancer patients harboring BRCA gene mutations (122). However, the occurrence of reverse mutations in BRCA genes may diminish the efficacy of PARP inhibitors and potentially lead to resistance against these drugs (123, 124). Previous studies have predominantly concentrated on reverse mutations in circulating ctDNA from ovarian cancer patients, aiming to enhance the guidance for related drug therapies in clinical settings.

In HGSOC at stages III and IV, dynamic analysis of BRCA1/2 ctDNA can detect reversal mutations or mutations in BRCA1 and TP53, thereby predicting patients’ responses to PARP inhibitor therapy (26). Additionally, somatic reversal mutations or intragenic deletions that restore BRCA1 or BRCA2 function can confer resistance to PARP inhibitors or platinum-based chemotherapy in patients harboring germline BRCA1 or BRCA2 mutations (27). ctDNA sequencing analysis can identify potential BRCA1/2 reversal mutations in ovarian cancer patients (28), facilitating the screening of patients who are suitable for PARP inhibitor therapy (29, 30). Christie et al. (31) identified five cases of BRCA gene reversal mutations in tumor DNA from 30 patients with high-grade serous ovarian cancer through NGS analysis of TP53, BRCA1, and BRCA2. Among these, three patients exhibited the same reversal mutation in their ctDNA, with a detection sensitivity of 60%. These three patients developed resistance to platinum-based drugs or PARP inhibitors but demonstrated favorable responses to gemcitabine and bevacizumab therapy. Lin et al. (32) conducted ctDNA BRCA1 and BRCA2 reversal mutation testing in 78 patients with epithelial ovarian cancer prior to PARP inhibitor therapy, identifying eight patients with reversal mutations, whose progression-free survival (PFS) was significantly lower than that of patients without such mutations. In another study, ctDNA BRCA1 and BRCA2 reversal mutations were detected in 112 ovarian cancer patients treated with PARP inhibitors, with eight patients acquiring reversal mutations. Among these, four patients experienced disease progression after a median follow-up of 3.5 months, while the remaining four had reversal mutations detected at the time of disease progression. An additional analysis of 54 ovarian cancer patients identified three cases of BRCA reversal and demonstrated that mutation heterogeneity improved progression after PARP inhibitors, suggesting non-exclusive acquired resistance mechanisms such as upregulation of survival pathways (41%), replication fork stability (34%), homologous recombination repair restoration (28%), loss of targets (10%), and drug efflux (3%) (33). Additionally, a study collected 69 plasma samples from 135 ovarian cancer patients before and after systemic therapy. The results revealed that 60% of patients exhibited high methylation of the BRCA1 promoter during treatment, while 24% reverted from this high methylation state. This finding suggests that ctDNA testing plays a significant role in identifying treatment-resistant clones (34).

These results indicate that the analysis of ctDNA highlights various acquired resistance mechanisms, which will assist in treatment selection and resistance monitoring. The current challenge lies in how to translate the detection results into specific clinical actions, particularly in determining the optimal treatment plan after identifying resistant mutations. Furthermore, the sensitivity of ctDNA testing is not 100%, and negative results do not completely rule out resistance. Future prospective trials are needed to clarify how to best integrate ctDNA into the clinical decision-making pathways for resistant ovarian cancer.

Early detection of recurrence is crucial for guiding treatment decisions and reducing patient turnaround time, ultimately improving clinical outcomes. The detection of MRD after surgery or radical treatment is one of the most commonly used methods for prognostic assessment, in which ctDNA demonstrates superior performance compared to traditional methods.

Research indicates that ctDNA analysis exhibits higher sensitivity and specificity than CA-125 in detecting recurrence, typically identifying recurrences an average of 10 months earlier than imaging studies (35). Additionally, recent findings suggest that elevated levels of somatic mutations in ctDNA may correlate with decreased clinical treatment benefits and poorer PFS or overall survival (OS) (36, 37). Variations in ctDNA methylation levels during or after treatment generally serve as predictive indicators of patient prognosis. Advanced-stage patients are more likely to exhibit detectable ctDNA mutations prior to surgery and adjuvant chemotherapy, with these mutations being associated with poorer OS post-surgery, thereby establishing ctDNA as an independent predictor of PFS and OS (38).

ctDNA can achieve quantitative risk stratification. The study further confirms that ovarian cancer patients who underwent neoadjuvant surgery and exhibited higher ctDNA mutation copy numbers in plasma (>10 mutation copies/mL) had significantly lower OS compared to those with lower copy numbers (<10 mutation copies/mL) (125).

In a prospective study, researchers conducted serial ctDNA testing on 23 ovarian cancer patients to evaluate its potential in predicting disease recurrence after primary tumor resection. The results indicated that ctDNA achieved a sensitivity of 91% in predicting tumor recurrence (126). Additionally, studies have shown that detectable levels of ctDNA in plasma can predict disease recurrence an average of 7 months before CT scans (70).

Vitale et al. analyzed serum samples from 20 patients with high-grade serous ovarian cancer following their initial debulking surgery to investigate TP53-mutated ctDNA sequences. They found that patients with suboptimal debulking (67%) exhibited a higher proportion of TP53-mutated ctDNA compared to those with optimal debulking (45%) (127). Research presented at the 2024 ASCO Annual Meeting confirms that ctDNA is a powerful prognostic tool for guiding MRD monitoring in advanced ovarian cancer (128). Its sensitivity for detecting residual lesions (21.2%) is significantly higher than that of CA125 (1.3%). Patients with negative ctDNA results have an extremely low risk of recurrence (negative predictive value of 98.3%), while those with positive ctDNA results have a recurrence risk 57.5 times greater than that of negative patients, which is markedly higher than the 6.3-fold increase in risk associated with positive CA125 results. Consequently, monitoring postoperative MRD status and continuous ctDNA assessment during maintenance therapy are optimal precision management tools for predicting recurrence risk and guiding clinical treatment decisions, providing critical evidence for the precision treatment of ovarian cancer. The core challenge lies in determining the ctDNA threshold that triggers clinical intervention.

In summary, dynamic monitoring of ctDNA at the molecular level holds significant promise in assessing the prognosis of ovarian cancer patients. Future prospective studies are warranted to further validate whether ctDNA can serve as a novel biomarker for the diagnosis and prognosis of ovarian cancer.

Cervical cancer is primarily caused by persistent infection with high-risk human papillomavirus (HPV), with subtypes 16 and 18 accounting for approximately 70% of cases (129). Currently, the most commonly used methods for cervical cancer screening and diagnosis include liquid-based cytology and HPV-DNA testing. ctDNA analysis provides a new dimension for screening, particularly based on HPV ctDNA analysis. Studies have shown that in cervical cancer, the HPV viral genome can integrate into the tumor cell genome or free DNA (130), and can be detected in the plasma of HPV-associated cervical cancer patients using different PCR techniques (131).

HPV ctDNA has a natural high specificity due to its viral specificity. Gu et al. (39) established that HPV ctDNA serves as a specific diagnostic marker for cervical cancer, exhibiting moderate sensitivity (0.27, 95% CI: 0.24–0.30) and high specificity (0.94, 95% CI: 0.92–0.96), reflecting its high diagnostic accuracy. Multiple studies (40, 41) have identified that HPV ctDNA can be detected in blood samples from HPV-positive cervical cancer patients prior to treatment, while it is rarely found in HPV-positive cervical precancerous lesions or in patients without cervical lesions. Mittelstadt et al. (42) investigated HPV ctDNA in the plasma of cervical cancer patients and found a significant correlation between tumor burden and HPV ctDNA levels. Additionally, a quantitative study utilizing ddPCR technology revealed that HPV ctDNA levels in cervical cancer patients are associated with clinical stage and tumor size. Even in the subclinical stage, patients with HPV-related invasive cancer can have detectable HPV ctDNA in their plasma, with levels closely correlated to tumor dynamics (43).

Currently, ctDNA testing has been incorporated into the 2022 NCCN guidelines for cervical cancer, indicating that ctDNA can be considered for whole-genome analysis when tumor tissue samples are unavailable. Its future role may involve triaging women who test positive in primary HPV screening, identifying high-risk populations with integrated and transformative infections, thereby optimizing the referral strategy for colposcopy and improving screening efficiency.

Despite significant differences in ctDNA among different patients, the dynamic changes of ctDNA over time in individual patients are closely related to treatment response and disease burden. Therefore, ctDNA can serve as a specific biomarker for assessing treatment efficacy. The monitoring of treatment effects in cervical cancer can also benefit from the dynamic analysis of ctDNA, as its changes can reflect the tumor’s response to radiotherapy and chemotherapy in real-time.

HPV ctDNA is an ideal biomarker for monitoring the tumor burden associated with viral infections. Kim et al. used NGS to monitor the treatment response of 25 cervical cancer patients following radical radiotherapy, employing HPV16/18 and 67 cancer gene target capture libraries. They assessed the efficacy of HPV ctDNA in comparison to traditional treatment monitoring methods, such as tumor markers and magnetic resonance imaging (MRI). The tumor markers evaluated included squamous cell carcinoma antigen (SCC) and carcinoembryonic antigen (CEA). HPV ctDNA was quantified by calculating the ratio of HPV read counts to total read counts (human reads + HPV reads). The results indicated that, in most cases, both tumor marker levels and HPV ratios decreased as tumor volume diminished. Notably, the HPV ratio exhibited greater sensitivity in detecting temporal changes between the second and third patient visits. In two patients with distant metastasis, the HPV ratio was significantly elevated compared to tumor markers. Conversely, in patients without evidence of disease, the HPV ratio decreased while tumor marker levels increased. These findings suggest that the HPV ctDNA ratio, based on targeted NGS, may serve as an effective tool for treatment monitoring and prognostic assessment (44).

Conversely, Li et al. (45) prospectively recruited 16 patients with locally advanced cervical cancer who required neoadjuvant chemotherapy. Through next-generation sequencing (NGS) analysis of circulating tumor DNA (ctDNA) and tumor tissue DNA (ttDNA) from these patients, mutations in PBRM1, SETD2, and ROS1 were frequently detected in both ctDNA and ttDNA of non-responders. Among these mutations, PBRM1 mutations were linked to poorer survival outcomes in patients. In vitro experiments demonstrated that knocking down PBRM1 enhanced STAT3 signaling in cervical cancer cells, thereby promoting resistance to cisplatin. These findings suggest that PBRM1 mutations may serve as potential ctDNA biomarkers for drug resistance in patients undergoing neoadjuvant chemotherapy. This indicates that ctDNA analysis can not only monitor tumor burden but also reveal mechanisms of resistance, providing molecular evidence for timely adjustments to treatment regimens.

Therefore, in contrast to traditional imaging examinations and pathological biopsies, ctDNA can more effectively monitor treatment responses and evaluate treatment efficacy during cervical cancer therapy. Additionally, it allows for dynamic adjustments to treatment plans based on fluctuations in its concentration levels, thereby facilitating individualized management and precision treatment for patients.

Traditional treatment for cervical cancer primarily relies on surgery combined with radiotherapy and systemic chemotherapy (132). ctDNA offers the potential for precision therapy by identifying actionable genetic mutations. Numerous studies have demonstrated that monitoring treatment via ctDNA is a sensitive method compared to traditional tissue biopsy or imaging techniques (133).

Charo et al. (47) analyzed ctDNA sequencing results from 13 cervical cancer patients alongside other gynecological tumor patients. In cervical cancer patients, PIK3CA emerged as the most frequently mutated gene. The study results indicated that in multivariate analysis, a higher maximum mutated allele frequency of ctDNA was associated with poorer OS (HR = 1.91, P = 0.03). In contrast, patients who received treatment aligned with ctDNA alterations exhibited significantly improved OS (HR = 0.34, P = 0.007), while those who did not receive matched treatment did not show such an association. This study suggests that the detection of ctDNA levels can identify gene mutations with potential value for diagnosis, prognosis, and tumor monitoring, thereby guiding treatment decisions. Additionally, Zhang et al. (48) conducted NGS testing covering 1,020 genes on plasma samples from 126 cervical cancer patients in China. The results indicated that PIK3CA was the most commonly mutated gene in cervical cancer (approximately 30%), consistent with the detection results of PIK3CA in cervical cancer tissue. This finding further confirms the correlation between plasma ctDNA and mutated genes in tumor tissue, providing a basis for selecting targeted therapies based on ctDNA mutations.

Lalondrelle et al. (49) recently conducted a study using an NGS technique called “panHPV detection” to detect HPV ctDNA and evaluate treatment responses in patients with locally advanced cervical cancer. The test successfully identified recurrence in three patients, despite imaging studies indicating complete clearance, with HPV ctDNA being detected three months post-treatment. Furthermore, four patients exhibiting partial clearance on imaging, who did not show HPV ctDNA at the three-month follow-up, displayed no signs of recurrence. These findings suggest that panHPV ctDNA detection holds significant potential for assessing treatment response and predicting recurrence in cervical cancer patients following chemoradiotherapy.

Currently, targeted treatment options for cervical cancer remain relatively limited. The guiding role of ctDNA is more evident in identifying patients who could benefit from specific targeted clinical trials. In the future, as more targeted drugs are developed, the guiding value of ctDNA will further increase. However, further prospective studies are needed to validate the guidance on the timing and selection of treatment regimens.

Most cervical cancer patients require regular follow-up examinations within five years after completing treatment. These follow-up procedures primarily involve monitoring for recurrence and metastasis of cervical cancer, evaluating adverse reactions that may have occurred during treatment, and conducting comprehensive systemic examinations. Circulating tumor DNA testing has the advantage of being repeatable in terms of sample collection and processing, enabling dynamic monitoring of tumor progression in patients. Multiple studies have shown that ctDNA can be used to assess the efficacy of radiotherapy, chemotherapy, and other treatment regimens. Notably, the detection levels of HPV ctDNA not only serve to evaluate treatment outcomes but are also closely associated with cervical cancer prognosis; however, the results of these tests are influenced by various factors (134).

Multiple studies have confirmed the prognostic value of ctDNA. A study conducted by Jeannot et al. revealed that HPV ctDNA was detected in 63% of patients prior to treatment, with ctDNA levels positively correlated to higher FIGO stages and para-aortic lymph node involvement. Furthermore, complete clearance of HPV ctDNA at the conclusion of treatment was significantly associated with extended progression-free survival, suggesting that HPV ctDNA detection can serve as an effective biomarker for predicting cervical cancer recurrence (50). Cabel et al. (51) analyzed blood and tumor samples from 55 HPV-positive patients with locally advanced cervical cancer who underwent chemoradiotherapy, finding that positive HPV ctDNA detection was linked to lower DFS (P = 0.048) and OS (P = 0.0013). This indicates that residual HPV ctDNA at the end of chemoradiotherapy may assist in identifying patients at risk of subsequent recurrence. In a recent study, Han et al. (52) compared the accuracy of FDG-PET and ddPCR in predicting plasma HPV ctDNA levels in patients with locally advanced cervical cancer. Through serial sampling, post-treatment plasma HPV ctDNA levels were shown to predict metastasis. The study reported that patients with undetectable HPV ctDNA levels in plasma after treatment were associated with higher PFS compared to those with detectable HPV DNA levels. Collectively, these studies indicate that ctDNA possesses significant prognostic predictive value for cervical cancer.

Recently, Chung et al. (46) used ddPCR to assess the efficacy of PIK3CA mutation detection in cervical cancer patients. In plasma ctDNA samples collected prior to treatment from 177 patients with invasive cervical cancer, two PIK3CA mutations (p.E545K and p.E542K) were identified. The results indicated that 22.2% of patients had at least one PIK3CA mutation in their plasma ctDNA. Notably, the study found that the PIK3CA mutation status in plasma ctDNA was significantly associated with reduced median tumor size, disease-free survival, and overall survival. A retrospective clinical study by Tian et al. (53) found that ctDNA could detect recurrent metastatic cervical cancer patients with any mutations in PIK3CA, BRAF, GNA11, FBXW7, and CDH1, whose PFS and OS were significantly shorter than those without detectable mutations.

The recent CALLA study reveals that in patients with locally advanced cervical cancer (LACC), the application of ultra-high-sensitivity personalized ctDNA detection based on whole-genome sequencing (NeXT Personal) demonstrates exceptional sensitivity, evidenced by a baseline detection rate of 99%. The findings indicate that elevated baseline ctDNA levels correlate with an increased risk of disease progression. Notably, the detection of ctDNA during treatment (cycle 3) and post-treatment (6 months) serves as a robust predictor of significantly heightened recurrence risk, with hazard ratios (HR) as low as 0.04–0.05. Furthermore, ctDNA clearance is closely linked to improved PFS, with the median PFS not yet reached. Although the combination therapy with durvalumab (in contrast to radiotherapy and chemotherapy alone) enhances ctDNA clearance, particularly in the PD-L1-high expression subgroup, no significant difference in disease progression risk was observed between the two groups when evaluated based on ctDNA status. Consequently, this study primarily establishes the significance of ctDNA as an independent prognostic biomarker, thereby supporting its future application in guiding risk stratification and treatment decisions for LACC patients (54).

In conclusion, ctDNA monitoring is pivotal in prognostic assessment and recurrence surveillance during the treatment of cervical cancer patients. It functions not only as a residual tumor marker at the conclusion of treatment but also as a sensitive biological indicator for monitoring recurrence during follow-up.

Although the overall prognosis for endometrial cancer is generally favorable, approximately 30% of cases are diagnosed at an advanced stage with invasive metastasis, leading to a poorer prognosis. Early diagnosis can significantly improve patient outcomes (135). However, there is currently no established screening protocol for endometrial cancer, either in the general population or among specific high-risk groups (136). Therefore, identifying more sensitive and specific screening markers is crucial, and ctDNA has emerged as a promising biomarker for endometrial cancer.

Current studies have validated the feasibility of screening early-stage patients by detecting somatic mutations associated with endometrial cancer in ctDNA. Feng et al. (55) found that for high-risk endometrial cancer, the mutated genes detected in ctDNA were highly consistent with those in baseline tumor tissue. Bolivar et al. (56) identified four common genes associated with endometrial cancer (CTNNB1, KRAS, PTEN, PIK3CA) and detected at least one mutation in 94% of patients, with mutations found in the plasma of 18% of early-stage patients. Additionally, enhancing the efficacy of early screening can be achieved by detecting ctDNA fragment length, copy number, or by combining ctDNA detection with other protein biomarkers. A study reported by ASCO in 2018 also indicated that the sensitivity of ctDNA detection for patients with endometrial cancer is less than 10%, while its specificity reaches as high as 95% (57).

Furthermore, a study combined ctDNA profiling with tumor-induced platelets (TEPs) to test 71 genes (58). This method utilizes TEPs for preliminary diagnosis to distinguish between healthy women and endometrial cancer patients, followed by ctDNA analysis for histological diagnosis. This approach provides valuable support for endometrial cancer diagnosis, achieving a sensitivity of 77.8% and an accuracy rate of 68.7%. Recent research has further indicated that the detection of microsatellite instability (MSI) and ctDNA in endometrial aspirates can function as a diagnostic tool for minimally invasive subtypes of endometrial cancer and can be utilized to monitor patients’ responses to immune checkpoint inhibitors (59, 60).

In summary, ctDNA not only serves as a diagnostic biomarker for endometrial cancer, but research on molecular subtyping guided by it is also likely to become an important future research direction, laying the foundation for the precise management of endometrial cancer.

The heterogeneity of endometrial cancer can lead to both congenital and acquired drug resistance, resulting in poor responses to conventional chemotherapy and targeted therapies. In such cases, characterizing tumor heterogeneity and monitoring tumor response to drugs in real-time can facilitate more accurate predictions of specific clinical outcomes at diagnosis, thereby enabling the development of the most appropriate treatment plan.

The levels of ctDNA are correlated with tumor burden and treatment response. Studies have shown that ctDNA testing is primarily conducted in high-risk endometrial cancer patients (55, 137). Ashley et al. (61) found that ctDNA levels are associated with endometrial cancer staging, and continuous measurement can reflect treatment response, disease progression, and recurrence. In over 90% of cases, tumor somatic mutations were accurately identified in ctDNA, and the presence of baseline or postoperative ctDNA was significantly associated with reduced PFS. Carlos et al. (62) used ddPCR technology to measure ctDNA levels in 60 endometrial cancer patients. The results indicated that 56.3% of high-risk tumor patients were ctDNA-positive at the time of surgery, whereas only 15.8% of low-risk tumor patients were ctDNA-positive. Furthermore, endometrial cancer patients with higher ctDNA levels and deeper myometrial invasion exhibited a higher tumor burden, suggesting that ctDNA can be employed to monitor tumor burden.

Silveira et al. (63) reported that ctDNA detection results for endometrial cancer exhibited high consistency (93%) with the clinically established reference pentablot method, and its dynamic changes during treatment were closely related to clinical response. Some researchers have utilized ddPCR to detect PIK3CA or KRAS mutations in ctDNA to explore the association between ctDNA and endometrial cancer (64). The results demonstrated that ctDNA was significantly associated with the histopathological type, stage, and grade of endometrial cancer and served as an independent risk factor for lymph node metastasis. Collectively, these studies indicate that ctDNA is a promising dynamic biomarker for assessing tumor burden and evaluating treatment efficacy.

In summary, the absence of specific biomarkers for monitoring disease activity in endometrial cancer underscores the significant clinical value of endometrial cancer-specific ctDNA testing. This approach is particularly beneficial for monitoring recurrence and guiding future treatment decisions.

The treatment decisions for endometrial cancer, particularly regarding the use of immunotherapy, are highly dependent on molecular subtyping.

ctDNA plays a crucial role in the treatment of endometrial cancer. A large-scale Phase II KEYNOTE-158 study (65) demonstrated that PD-1 inhibitors exhibit superior efficacy in patients with microsatellite instability-high (MSI-H) or mismatch repair-deficient (dMMR) tumors. Additionally, ctDNA can effectively identify MSI-H status in recurrent endometrial carcinoma (66). Clinical guidelines recommend determining the MSI status of endometrial cancer patients to guide the use of PD-1 inhibitors based on ctDNA (67). In the DUO-E study of advanced or recurrent endometrial cancer, dynamic ctDNA monitoring revealed that 80% of patients had detectable ctDNA at baseline, and its presence was associated with shorter PFS and was a poor prognostic factor; a decrease in ctDNA levels during treatment was strongly associated with significantly improved prognosis. The study specifically demonstrated that in patients with proficient mismatch repair (pMMR), the combination of olaparib and durvalumab maintenance therapy significantly enhanced ctDNA clearance rates during the maintenance phase (48% vs. 17% with single-agent maintenance), elucidating the potential mechanism behind the improved efficacy of this regimen in pMMR patients (68, 69).

In summary, ctDNA, as a real-time and sensitive biomarker, shows considerable promise in evaluating treatment response, predicting recurrence risk, and guiding personalized treatment decisions, thereby advancing precision therapy for endometrial cancer.

High-risk endometrial cancer is characterized by elevated rates of metastasis and recurrence. Patients who experience recurrence tend to have poorer prognoses. Consequently, monitoring for recurrence and assessing prognosis are critical for developing appropriate individualized treatment plans. This approach ensures that patients at high risk of recurrence receive timely adjuvant therapy while avoiding unnecessary treatments for low-risk patients.

Pereira et al. (70) hypothesized that ctDNA serves as an independent predictor of survival in patients with endometrial cancer. Their research demonstrated that regular monitoring of ctDNA levels allows for earlier diagnosis of disease recurrence compared to traditional CT scans. Furthermore, post-treatment monitoring of ctDNA levels provides novel indicators for survival prediction. In a study involving nine high-risk endometrial cancer patients (55), ctDNA was detectable in 67% of baseline plasma samples, with significantly elevated levels of both cfDNA and ctDNA observed in recurrent patients compared to non-recurrent patients. The study also evaluated the correlation between ctDNA and tumor recurrence, revealing that postoperative ctDNA levels exhibited a sensitivity of 100% and a specificity of 83.3% for predicting tumor recurrence, with a Kappa index of 0.769. These findings indicate a substantial consistency between ctDNA detection and postoperative tumor recurrence, suggesting that ctDNA is superior to CA125 or HE4 in detecting tumor recurrence (71).

ctDNA can provide a significant lead time. Additionally, a study conducted in China involving 46 patients with endometrial cancer found that recurrent patients exhibited negative ctDNA levels preoperatively, which turned positive during postoperative follow-up, with clinical recurrence confirmed one month later (3). A recent small-sample study (66) also indicated that ctDNA can detect recurrence or progression of endometrial cancer approximately 2.5 months (ranging from 1 to 8 months) earlier than traditional imaging techniques.

In patients diagnosed with stage I endometrial cancer who subsequently relapsed, elevated ctDNA levels signaled disease progression, which was confirmed by imaging techniques. Additionally, ctDNA can accurately reflect the response to radiotherapy in patients with recurrent endometrial cancer undergoing chemotherapy. A recent real-world study demonstrated that ctDNA testing holds significant prognostic value in patients with advanced endometrial cancer. Patients with ctDNA positivity, particularly those carrying TP53 mutations, exhibited significantly poorer survival outcomes (median real-world overall survival [rwOS]: 71.2 months vs. 108 months; hazard ratio [HR]: 0.75; 95% confidence interval [CI]: 0.59–0.96; P = 0.02). When TP53 co-mutates or co-amplifies with other genes, it further exacerbates survival risk (rwOS: 66.4 months vs. 78.1 months; HR: 0.66; 95% CI: 0.47–0.92; P = 0.016). Genomic profiling analysis based on ctDNA revealed key driver mutations, and longitudinal monitoring can dynamically reflect tumor evolution and early progression. These findings support the integration of ctDNA, particularly TP53 status, into the management of endometrial cancer for risk stratification, prognostic assessment, and treatment monitoring (72). Notably, TP53 mutations also possess therapeutic predictive value; subsequent analysis of the GOG-86P trial showed that in advanced or recurrent patients, the TP53 mutation subgroup (44%) receiving bevacizumab plus chemotherapy (paclitaxel/carboplatin) had significantly longer survival compared to the temsirolimus regimen (median PFS: 12.5 months vs. 8.2 months; median OS: 29.1 months vs. 22.0 months), while wild-type patients did not benefit from this combination (73).

In summary, ctDNA is closely related to the prognosis of endometrial cancer and can serve as a real-time monitoring biomarker for assessing prognosis and detecting recurrence. Integrating ctDNA, particularly the TP53 status, for risk stratification, prognostic evaluation, and treatment monitoring is an important direction for the future.