Harnessing ovarian cancer ascites for translational science: models, biomarkers, and therapeutics.

The development of ascites often causes symptoms in ovarian cancer patients, including abdominal distention, pain, shortness of breath and early satiety. Historically, management included diuretics, salt restriction, radioactive isotopes, and intraperitoneal administration of sclerosing agents [ 213 , 214 ]. However, these treatments were unsuccessful and caused significant toxicity, falling out of favor [ 213 ]. Currently, symptomatic relief is provided by performing paracentesis when required, or by placing an abdominal pleurX, a small tube from the skin tunneled to the peritoneal cavity, allowing frequent drainage at home. Risks of pleurX include infection (cellulitis or peritonitis) and hypoalbuminemia from frequent drainage. The need to better manage malignant ascites has prompted innovative treatment approaches. As ascites formation depends on tumor vascularity and permeability, VEGF-specific monoclonal antibodies have markedly decreased ascites formation in preclinical studies This led to using bevacizumab in patients with recurrent ovarian cancer and ascites requiring repeat paracentesis, demonstrating ascites resolution in case reports and series [ 215 – 218 ]. Bevacizumab, combined with platinum-based chemotherapy and as maintenance therapy, remains important for newly diagnosed advanced ovarian cancer, decreasing ascites formation and prolonging PFS [ 4 , 5 , 219 ]. Additionally, bevacizumab increases objective response rate and prolongs PFS when combined with standard-of-care chemotherapy in platinum-sensitive and resistant settings [ 220 , 221 ]. Similarly, Aflibercept, or VEGF trap has been studied in treating malignant ascites. A double-blind, placebo-controlled, phase II study on 55 patients with advanced chemoresistant ovarian cancer and recurrent ascites received either aflibercept or placebo. The mean time to repeat paracentesis was 31 days longer with aflibercept treatment. However, the frequency of fatal gastrointestinal events is higher with aflibercept, and this medication has not been routinely incorporated into clinical practice [ 222 ]. Catumaxomab, a trifunctional monoclonal antibody binding to EpCAM and CD3 antigens, has also been studied. A randomized phase II/III study compared catumaxomab plus paracentesis alone for ascites treatment [ 223 ]. Puncture-free survival was significantly longer in the catumaxomab group than in the control group (46 vs. 11 days). This medication was approved by the European Medicines Agency in 2009 for the treatment of malignant ascites in EpCAM-positive tumors [ 213 , 223 , 224 ]. However, Catumaxomab was voluntarily withdrawn in the United States in 2013 and for commercial reasons by the European Commission in 2017. However, the European Commission approved catumaxomab again in 2025 for the treatment of malignant ascites in EpCAM-positive tumors. Overall, a few targeted therapies have been studied for the management of malignant ascites, and most have not been incorporated into clinical practice. Another novel agent under investigation for malignant ascites treatment is M701, a bispecific antibody targeting EpCAM and CD3. Interim results from a randomized, controlled phase II trial compared paracentesis with intraperitoneal M701 infusions on days 1, 4, 11, and 18 (additional infusions allowed every 2 weeks) to paracentesis as needed days 1 to 18. Both groups received investigator-chosen systemic therapy (84 patients participated, 30% with ovarian cancer) with longer puncture-free survival in the experimental arm, median 54 versus 24 days, Hazard Ratio: 0.39, 95% Confidence Interval: 0.21–0.72, p  = 0.001 [ 225 ]. While the above treatment approaches may reduce or mitigate the discomforts associated with the development of ascites, the predominant strategy for the most successful management of ascites is effective anticancer therapy, as ascites resolves when cancer responds to treatment [ 226 ].

Ovarian cancer is a highly heterogeneous disease in terms of genetic mutations and disease phenotypes and is frequently associated with the production of malignant ascites. Ascites contains both acellular and cellular components that facilitate tumor progression and immune evasion. Furthermore, acellular factors, such as cytokines and growth factors (e.g., VEGF, IL-6, and IL-8), promote tumor cell migration, invasion, metastasis, and immune disruption. Cellular factors present in ascites modulate innate and adaptive immune responses and are linked to disease progression and chemoresistance. To study the cellular components of ascites, malignant cells can be preserved in cell blocks for IHC testing or frozen for further molecular analysis [ 227 , 228 ]. Furthermore, the functional analysis of ascitic cells has been facilitated by technological advancements that can help isolate and culture cells derived from the ascitic fluid of patients with advanced and/or recurrent ovarian cancer for in vitro and in vivo testing. This provides opportunities for identifying and validating therapeutic targets and predictive biomarkers. Current therapies used in the management of ascites include VEGF inhibitors, but they have limited clinical success. Therefore, the development of effective therapeutic options for the treatment of persistent malignant ascites is required, and understanding the tumorigenic role of ascites is a crucial step towards achieving this goal.

An important first consideration in the approach to utilize ascites for translational research is the processes for the collection, biobanking, and annotation of ascites samples. Ascites can be collected from ovarian cancer patients either at the time of surgical debulking or from paracentesis. Many cancer centers have established biorepositories or biospecimen cores that are specialized in collecting, processing, and storing cancer biospecimens. However, to our and others'experience [ 140 ], collection and biobanking of ascites specimens, for the purpose of in vitro drug testing, biomarker screening, development of patient-derived cell lines and xenografts, organoids, cfDNA, or for clinical trials, requires a highly coordinated inter-departmental collaboration and specialized expertise in consenting, collection, tissue processing, and data extraction support for annotation of clinical data. Harmonizing protocols for processing of ascites is also important to ensure that the data generated using this complex specimen is consistent and robust. Ascites can be processed using a temperature-controlled centrifuge to obtain a cell pellet. The acellular ascites fluid supernatant can be frozen and stored at −80 °C, which can be used for cytokine/growth factor analysis or in vitro experiments. Cell pellets can be treated with lysis buffer to remove red blood cells, and then ascites cells can be plated in tissue-culture plates for use in various experiments [ 141 – 143 ]. These efforts provide access to patient specimens to various basic/translational and clinical research groups, thus fostering collaboration to understand disease progression and develop novel treatment strategies to improve patient outcomes. Ovarian cancer spheroids isolated from ascites are emerging as promising experimental systems for developing new classes of anticancer therapeutics and for implementing functional precision medicine approaches [ 101 , 143 – 150 ]. Multicellular ovarian cancer spheroids isolated directly from ascites possess both tumor-like three-dimensionality and the pathological and genetic profiles of clinical cancer [ 143 , 145 ]. Therefore, ascites-isolated spheroids may be superior for screening and evaluating ovarian cancer therapeutics compared to the increasingly disfavored cancer cell line models cultured in either monolayer or as spheroids in vitro. Two-dimensional cancer cell cultures cannot model the cellular uptake of nutrients and drug molecules inherent to three-dimensional tumor biology. Furthermore, established ovarian cancer cell lines selected for growth in artificial tissue culture media do not retain their original genetic and biological characteristics or recapitulate the characteristics of in vivo cancer biology, especially in individual patients [ 143 , 145 , 147 , 150 ]. Therefore, ovarian cancer spheroids isolated from ascites represent a promising and increasingly utilized experimental system for drug development (Fig.  3 ). Fig. 3 Potential applications of ascites in translational and clinical research. Presence of tumor cells in malignant ascites can help with cancer staging. Ovarian cancer spheroids isolated from ascites can serve as a clinically relevant model system to identify and validate novel drug targets for drug development. These patient-derived ascites organoids can also be implanted in mice to test in vivo efficacy of therapeutic molecules. Ascitic fluid can be a promising tool for the discovery of novel diagnostic and/or prognostic biomarkers. (Created with BioRender.com) Potential applications of ascites in translational and clinical research. Presence of tumor cells in malignant ascites can help with cancer staging. Ovarian cancer spheroids isolated from ascites can serve as a clinically relevant model system to identify and validate novel drug targets for drug development. These patient-derived ascites organoids can also be implanted in mice to test in vivo efficacy of therapeutic molecules. Ascitic fluid can be a promising tool for the discovery of novel diagnostic and/or prognostic biomarkers. (Created with BioRender.com) Recently, several studies have established cellular experimental systems using patient-isolated ascites for anticancer drug screening [ 101 , 144 , 151 ]. In principle, ascites-derived cell models can be used to provide functional precision treatment to guide the clinical management of patients with ovarian cancer using currently approved therapeutics [ 144 , 148 , 149 ]. Additionally, these systems can be used to identify or validate novel therapeutic molecules for clinical development [ 144 , 145 , 152 , 153 ]. Reports of ascites-derived cellular systems differ in the degree of differentiation and purification of heterogeneous ascites fluid before establishing the cellular system. One approach separates cancer spheroids from other cells present in a complex mixture of cells in ascites culture and then establishes ex vivo patient-derived tumor organoid (PTDO) cultures from spheroids [ 144 , 145 ]. PTDO is typically cultured in complex growth factor-rich media and extracellular matrix. Ex vivo ovarian cancer PTDOs are amenable to efficient drug screening, which can guide the precise treatment of individual patients [ 144 , 145 , 152 ]. For example, the OVAREX study sought to establish ex vivo PDTOs isolated from ascites fluid for diagnostic and clinical predictive value in managing ovarian cancer [ 149 ]. The second approach is to directly culture ascites fluid, which is presumably a complex mixture of cancer spheroids, individual cancer cells, immune cells, fibroblasts, and other cell types [ 12 , 145 ]. A prominent example of this non-differentiated ascites culture approach is the recent publication describing the Drug Efficacy Testing in 3D Cultures (DET3Ct) method [ 144 ]. The DET3Ct method involves culturing a complex ascites cellular mixture in ultralow attachment plates to promote aggregate/spheroid growth, followed by drug screening for anticancer efficacy [ 12 , 144 ]. Alternatively, ovarian cancer organoids can be isolated from ascites samples collected after paracentesis or upfront surgery from discarded peritoneal fluid following informed consent. Ascites sample can be centrifuged to obtain a cell pellet, which can be resuspended in media and strained through a 70 µm cell strainer. The organoids collected from the strainer can then be transferred to ultra-low attachment flat-bottom plates to use for experiments without freezing [ 154 ]. In another strategy, the cell pellet derived from the ascites of patients can be resuspended in specialized organoid growth media with basement membrane extract and cultured. Further characterization of these patient ascites-derived organoids has shown that they retain the tumor histological features and transcriptomic profiles consistent with the acquisition of platinum-resistance [ 155 ]. Ascites-derived organoids can also be established following mechanical and enzymatic digestion to form a single-cell suspension, which can then be resuspended in basement membrane extract supplemented with engineered media. Organoids are usually established within 7–10 days and can be cultured for up to 5 months with regular passaging [ 156 , 157 ]. Despite the different methodologies, these models can be reliably and reproducibly used to study disease progression and test novel treatment strategies. Importantly, a potential limitation of these approaches is the cellular heterogeneity outside of and within the ovarian cancer spheroids present in the ascites [ 12 , 144 ]. Ovarian cancer spheroids present in ascites fluid are not uniform in size and cellular properties [ 12 , 151 , 158 , 159 ]. Ascites-derived spheroid cultures are not cultured from individual spheroids but instead arise from all spheroids present in the ascites, which could have varied properties and drug sensitivities. New spheroid/organoid sorting and dispensing instruments can be used to sort individual cancer spheroids of uniform size and morphology from ascites samples before establishing PTDOs or spheroid cultures for drug screening [ 160 ]. Patient-derived ascites cells can be utilized to establish patient-derived xenograft (PDX) models that have features similar to the primary tumor. Subcutaneous implantation of fragmented primary tumor and tumor cells isolated from peritoneal and pleural effusions of ovarian cancer patients showed similar pathological and immunohistochemical features. Genomic profiling revealed that the PDXs using both primary tumor and body fluids tended to have gene mutations in the serine/theronine kinase 11 ( STK11 ), neuroblastoma RAS viral oncogene homolog ( NRAS ), smoothened receptor ( SMO ), and phosphatase and TENsin homolog ( PTEN ) which were absent in the primary tumor. This suggests that patient-derived ascites cells could be an alternative to establish PDXs in the absence of a primary tumor surgical specimen [ 161 ]. Further, luciferized PDX models have been established using patient-derived ascites cells stably transduced with luciferase plasmid to monitor tumor growth kinetics and disease progression by taking advantage of bioluminescence imaging. In addition, the transduction of ascites cells with the luciferase gene did not influence the molecular and/or histological features of the PDX models [ 162 ]. The use of humanized PDX models provides an opportunity to study disease progression as influenced by the immune microenvironment. The use of huNBSGW mice to implant ascites-derived tumor cells successfully mimics the immune microenvironment of ovarian cancer patients [ 163 ]. This model demonstrates the ability of humanized PDXs to recruit macrophages and T-cells to solid tumors with a similar cytokine profile as the ascites fluid of the patient, thereby providing clinically relevant models to test immunotherapies [ 164 ]. PDXs established from ascites cells can be highly relevant for predicting therapeutic response and identifying innovative therapeutic strategies [ 149 ]. For example, a PDX model established using intraperitoneal implantation of ascites cells has been used to evaluate the efficacy of the novel combination of ubiquitin-specific protease 1 with PARP inhibitor in the chemo-resistant setting [ 165 ]. In another study, an ascites-derived PDX model was used to demonstrate the efficacy of ITGA6-specific monoclonal antibody in significantly reducing metastatic lesions of the diaphragm and lungs [ 100 ]. Ascitic fluid has a diverse range of molecules and cell types, along with considerable inter-patient variability. This complexity presents a promising opportunity for the discovery of novel diagnostic and prognostic biomarkers. Studies have identified specific proteins [ 166 – 168 ], mRNAs [ 169 , 170 ], and miRNAs [ 171 ] in ascitic fluid, sometimes at higher levels than in plasma. Several publications detail new technical approaches and datasets for analyzing individual factors and profiling various omics in ascites fluid [ 73 – 80 ]. These targeted and omics-based approaches have identified biomarkers that can distinguish patients with ovarian cancer from those with benign conditions or predict ovarian cancer status or patient outcomes (Table  1 ). Apart from the contents of ascites, the ascites volume can also be of prognostic value. A retrospective study has identified that even a low to intermediate volume of ascites (~ 100 mL) was associated with a shorter overall survival in epithelial ovarian cancer. Further, the prognostic value of ascites accumulation varies with histological subtypes and was strongly correlative for endometrioid and mucinous subtypes. This indicates that ascites is not merely a marker of advanced disease but also actively promotes disease progression [ 172 ]. Table 1 Significant associations were identified between biomarkers in ovarian cancer ascites fluid and malignancy or patient outcome. OVID search on 05/20/2025 1946 to May 20, 2025–239 articles, abstracts reviewed, and 88 articles were identified for discussion of biomarkers evaluated in ovarian cancer ascites fluid. Articles in 41 of these abstracts that reported statistically significant findings were reviewed, summarized, and discussed in this table Biomarker Statistically significant findings Ref Profiles  Autoantibodies Autoantibodies in ascites from patients with ovarian cancer recognize CREB3, MRPL46, EXOSC10, BCOR, HMGN2, HIP1R, OLFM4 and KIAA1755 in association with response to first line chemotherapy [ 173 ]  Glycome The n-glycome profiles of ascitic fluid from patients with ovarian are distinct from matching serum from these patients and healthy controls [ 174 ]  Peptidome Thousands of peptides were specifically identified in ascites from patients with ovarian cancer compared to ascites and serum from non-cancer patients [ 175 ]  Sialome 13 sialoglycopeptides were present in ascites and other fluids from patients with ovarian cancer, that were not present in fluids from patients with benign disease [ 176 ] Immune Cells  B cells IL-10 + B-cells are enriched in ovarian cancer ascites in association with lower IFNγ +, CD8 + T cells and higher Foxp3 +, CD4 + T cells and are capable of suppressing IFNγ production by CD8 + T cells, through a mechanism involving IL10 production [ 177 ]  Dendritic cells Ascites dendritic cells were not associated with the survival of patients with high grade serous ovarian cancer [ 83 ]  Macrophages 1) Tumor associated macrophages in ovarian cancer ascites that express upregulation of interferon signaling pathways are associated with longer survival compared to those with high expression of immunosuppressive markers [ 83 ] 2) Siglec-9 + tumor-associated macrophages [ 178 ]  Monocytes Intermediate blood monocytes (CD14 high  CD16 low ) expansion in ascites is associated with higher tumor burden, decreased classical monocytes, decreased effector/regulatory T cell ratio, decreased pro-immune cytokines and increased soluble factors involved in immune tolerance and tumor progression [ 179 ]  NK Cells NK cells were impaired in ovarian cancer cell-positive ascites, but not in cell-free ascites collected from ovarian cancer patients [ 180 ]  Tsen CD8 + T cells Compared to health controls, ovarian cancer patients had increased Tsen CD8 + T cells in ascites, which were associated with lymph node metastases, chemotherapy resistance and shorter PFS, [ 181 ]  PD-1/PDL-1 30% of ascites T-cells expressed PD-1 and 50% of ascites non-immune cells expressed PDL-1 [ 83 ]  T cells 1) A trend toward low percentages of CD4 + T cells was observed with improved outcome of patients with high grade serous ovarian cancer [ 83 ] 2) Cell-free ascites compared to ovarian cancer cell-positive ascites collected from ovarian cancer patients had fewer regulatory T cells [ 180 ] Immune/Angiogenic Factors  CYR61 CYR61 in ascites from patients with ovarian serous cystadenocarcinoma was higher than in ascites from patients with serous cystadenoma and was associated with ascites IL-6, FIGO stage, initial tumor size > 10 cm and residual tumor burden [ 182 ]  B7-H4 High levels of B7-H4 checkpoint protein is strongly associated with advanced disease, lymphatic metastasis, and platinum resistance [ 183 ]  Factor H Increased factor H, a negative regulator of complement activation, in ascites fluid was associated with improved overall survival [ 184 ]  HLA-G Soluble HLA-G in ascites from patients with ovarian carcinomatosis is associated with peritoneal tumor burden, immunosuppressive factors and a decreased effector/regulatory T-cell ratio and is absorbed by the peritoneal tissues [ 185 ]  IDO1 IDO1 metabolic activity (L-tryptophan metabolism) and its metabolite L-kynurenine were higher in ovarian cancer ascites compared to plasma. Arginase, inducible nitric oxide synthase or glutaminase 1 were not elevated [ 168 ]  IFNγ Ascites IFNγ was associated with advanced disease, tumor grade, sub-optimal debulking surgery, DFS and OS [ 117 ]  IL4I1 IL4I1 metabolic activity (L-tryptophan to indole-3-pyruvic acid; L-phenylalanine to phenylpyruvic acid; or L-tyrosine to 4-hydroxyphenylpyruvic acid) were higher in ovarian cancer ascites compared to plasma, and higher in FIGO stage IV compared to III [ 168 ]  IL-6 1) Elevated IL-6 in ascites from ovarian cancer patients compared to benign cases had 84% sensitivity and 74% sensitivity, which was improved over the ROC curve for elevated IL-6 in plasma [ 186 ] 2) Elevated IL-6 in ascites from patients with ovarian cancer compared to controls was associated with worse PFS [ 187 ] 3) IL-6 was elevated in ovarian cancer compared to benign ascites and, when combined with serum CA-125 could discriminate between ovarian cancer and benign ascites [ 188 ]  IL-10 IL-10 was elevated in ovarian cancer compared to benign ascites but had low discriminating potential [ 188 ]  Leptin 1) Serum CA-125/ascites leptin was higher in ovarian cancer patients with chemotherapy-resistant compared to -sensitive disease and was predictive of worse OS [ 189 ] 2) Combination of serum CA-125 and ascites leptin levels had the highest discriminating potential between ovarian cancer and benign cases in comparison to serum CA-125 combinations with other ascites immune factors [ 188 ]  OPG OPG was elevated in ovarian cancer compared to benign ascites but had low discriminating potential [ 188 ]  TNFα TNFα was higher in cancer cell-positive ascites compared to cell-free ascites collected from ovarian cancer patients [ 180 ]  VEGF 1) VEGF is present at higher levels in ascites from patients with ovarian cancer compared to patients with ovarian hyperstimulation syndrome [ 190 ] 2) Elevated VEGF-A in ascites from patients with ovarian cancer compared to controls was associated with worse PFS [ 187 ] 3) VEGF is present at higher levels in ascites from patients with ovarian cancer compared to other cancers or benign conditions and is negatively associated with OS [ 191 ] 4) VEGF-C was present at higher levels in ascites from ovarian cancer patients compared to benign controls, and was correlated with tumor FIGO stage and grade, presence of lymph node metastases, and OS [ 192 ] EVs  CDCP1 +  High extracellular vesicle derived CDCP1 + in ovarian cancer compared to benign ascites [ 96 ]  MMP9 +  Higher levels of MMP9 were present on EVs in ascites from patients with ovarian cancer compared to portal hypertension [ 95 ] Proteins  ADAM17 ADAM17 metalloprotease was detected at higher levels in ovarian cancer ascites fluid compared to serum. Serum ADAM17 could be detected at early FIGO1/2 stages [ 193 ]  CA XII CA XII was expressed at higher levels in cancer cells from ascites from patients with ovarian cancer compared to patients with borderline ovarian tumors [ 194 ]  Ceruloplasmin Ceroplasmin was present at higher levels in ascites from patients with chemoresistant, compared to chemosensitive ovarian cancer [ 195 ]  EMMPRIN Higher EMMPRIN correlated with shorter PFI in NACT patients [ 196 ]  EpCAM EpCAM positive EVs in ascites is associated with shorter progression free survival [ 197 ]  Fibronectin Fibronectin is present in, and can be purified from, ascites from patients with ovarian cancer [ 166 ]  Mortalin Higher mortalin in ovarian cancer compared to benign ascites and blood in association with tumor stage, grade and histology. Mortalin levels in tumor, but not ascites, was prognostic [ 198 ]  MMP-2 Higher MMP-2 correlated with shorter PFI in NACT patients [ 196 ]  MMP-3 Higher MMP-3 correlated with shorter PFI in PDS patients [ 196 ]  LINE-1 ORFp Line-1 ORFp is detectable in human ovarian cancer ascites fluid [ 167 ]  PEDF PEDF expressed at higher levels in ascites from patients with ovarian cancer compared to patients with borderline ovarian tumors [ 199 ]  uPA Higher uPA correlated with shorter PFI in PDS patients [ 196 ]  Vitronectin Vitronectin is present in, and can be purified from, ascites from patients with ovarian cancer [ 166 ] DNA/RNA  cfDNA cfDNA and ctDNA were detected at higher levels, and ec-cfDNA was detected at lower levels in ascites compared to plasma. High ascites ctDNA levels were associated with a shorter paracentesis-free interval. Plasma ctDNA was associated with unfavorable survival and CA-125. No associations between cfDNA, ctDNA or ec-cfDNA and response were observed [ 200 ]  cfDNA GIS for HRD on cfDNA, which included ctDNA, were high for patients with BRCA mutations and superimposable with GIS scores on matching tumors [ 201 ]  mRNA’s 1) FXYD2 mRNA can be detected in ascites ovarian cancer cells [ 169 ] 2) Nectin-4, ADAM10 and ADAM17 mRNA was detected in ovarian cancer cells in ascites from patients with serous ovarian cancer [ 170 ]  miRNAs miRNAs are expressed at higher levels in ovarian cancer ascites compared to plasma from healthy controls. These include miR-21, the miR-200 family, miR-205, miR-10a and miR-346 [ 171 ] Cancer Cells  PAX8 PAX8 + combined with calretinin-negative staining could differentiate ascites cells that are ovarian cancer compared to other cancers or benign conditions [ 202 ]  Claudin-7 Ascites ovarian cancer cells, but not immune cells expressed high levels of Claudin-7 [ 203 ] BCOR BCL6 corepressor, cfDNA cell free DNA, CREB3 cAMP-response element3, CYR61 cysteine-rich angiogenic inducer 61, ec-cfDNA endothelial cell cfDNA, CA XII carbonic anhydrase II, ctDNA tumor-derived cfDNA, CDCP1 CUB domain-containing protein 1, DFS disease-free survival, EMMPRIN extracellular matrix metalloproteinase inducer/CD147, EVs extracellular vesicles, EXOSC10 Exosome Component 10, FXYD2 + FXYD domain containing ion transport regulator 2, GIS genomic instability scores, HIP1R huntingtin interacting protein 1 related, HLA human leukocyte antigen, HMGN2 high mobility group nucleosomal binding domain 2, HRD homologous recombination deficiency, IDO1 indoleamine 2,3-dioxygenase 1, IFNγ interferon gamma, IL4I1 interleukin 4 induced 1, KIAA1755 KIAA1755, LINE-1 ORFp Long interspersed element 1 (LINE-1) open reading frame 1 protein (ORFp), MMP matrix metalloproteinase, MRPL46 mitochondrial ribosomal protein L46, NACT neoadjuvant chemotherapy; NK cells = natural killer cells, OLFM4 olfactomedin 4, OPG osteoprotegerin, OS overall survival, PDS primary debulking surgery, PEDF Pigment epithelium-derived factor, PFI progression free interval, PFS progression free survival, Sialome sialic acid containing glycoproteins, Tsen CD8  +  T cells senescent (CD28- CD57 +) CD8 + T cells, uPA Urokinase-type plasminogen activator, VEGF vascular endothelial growth factor Significant associations were identified between biomarkers in ovarian cancer ascites fluid and malignancy or patient outcome. OVID search on 05/20/2025 1946 to May 20, 2025–239 articles, abstracts reviewed, and 88 articles were identified for discussion of biomarkers evaluated in ovarian cancer ascites fluid. Articles in 41 of these abstracts that reported statistically significant findings were reviewed, summarized, and discussed in this table BCOR BCL6 corepressor, cfDNA cell free DNA, CREB3 cAMP-response element3, CYR61 cysteine-rich angiogenic inducer 61, ec-cfDNA endothelial cell cfDNA, CA XII carbonic anhydrase II, ctDNA tumor-derived cfDNA, CDCP1 CUB domain-containing protein 1, DFS disease-free survival, EMMPRIN extracellular matrix metalloproteinase inducer/CD147, EVs extracellular vesicles, EXOSC10 Exosome Component 10, FXYD2 + FXYD domain containing ion transport regulator 2, GIS genomic instability scores, HIP1R huntingtin interacting protein 1 related, HLA human leukocyte antigen, HMGN2 high mobility group nucleosomal binding domain 2, HRD homologous recombination deficiency, IDO1 indoleamine 2,3-dioxygenase 1, IFNγ interferon gamma, IL4I1 interleukin 4 induced 1, KIAA1755 KIAA1755, LINE-1 ORFp Long interspersed element 1 (LINE-1) open reading frame 1 protein (ORFp), MMP matrix metalloproteinase, MRPL46 mitochondrial ribosomal protein L46, NACT neoadjuvant chemotherapy; NK cells = natural killer cells, OLFM4 olfactomedin 4, OPG osteoprotegerin, OS overall survival, PDS primary debulking surgery, PEDF Pigment epithelium-derived factor, PFI progression free interval, PFS progression free survival, Sialome sialic acid containing glycoproteins, Tsen CD8  +  T cells senescent (CD28- CD57 +) CD8 + T cells, uPA Urokinase-type plasminogen activator, VEGF vascular endothelial growth factor TAMs in ascites with upregulated interferon signaling pathways are associated with longer survival than those with high expression of immunosuppressive markers. The efficacy of anti-PD-1 and programmed cell death ligand 1 (PD-L1) drugs in the treatment of ovarian cancer may be affected by the observation that 30% of ascites T-cells express PD-1 and 50% of ascites non-immune cells express PD-L1 [ 83 ]. Higher levels of Siglec-9 + TAMs were observed in non-responders to PD-1 therapy [ 204 ]. Furthermore, elevated senescent CD8 + T cells in the ascites of patients with ovarian cancer compared with healthy controls are associated with shorter PFS in patients with lymph node metastases [ 181 ]. High IFNγ levels in ascites are associated with advanced disease, tumor grade, suboptimal debulking surgery, and lower DFS and OS [ 117 ]. Serum cancer antigen-125 (CA-125) to ascites leptin ratios were higher in chemotherapy-resistant patients, predicting a worse OS [ 205 ]. The combination of serum CA-125 and ascites leptin levels showed the highest potential for distinguishing between ovarian cancer and benign cases [ 188 ]. It has been reported that IL-6 levels in ascites alone and in combination with serum CA-125 levels can discriminate between ovarian cancer cases and benign controls with good sensitivity [ 188 ]. High VEGF levels were observed in the ascites of patients with ovarian cancer than in ascites of patients with other benign conditions [ 187 ] and are associated with The International Federation of Gynecology and Obstetrics (FIGO) tumor stage and grade, presence of lymph node metastases [ 192 ], and worse PFS [ 187 ] or OS [ 191 , 192 ]. Angiogenic factor, cysteine-rich angiogenic inducer 61 (CYR61), levels in ascites from patients with ovarian serous cystadenocarcinoma were higher than those in peritoneal lavage fluid from patients with serous cystadenoma and were associated with ascites IL-6, FIGO stage, initial tumor size > 10 cm, and residual tumor burden [ 182 ]. Soluble B7-H4, a checkpoint protein, levels in ascites were strongly associated with advanced disease, lymphatic metastasis, and platinum resistance, and can be a promising biomarker for epithelial ovarian cancer in conjunction with CA-125 [ 183 ]. Higher EpCAM-positive EV concentration in ascites was associated with shorter PFS in ovarian cancer patients and has prognostic value in predicting disease recurrence and development of chemoresistance [ 197 ]. Significant differences in metabolic protein and metabolite expression have been noted in ascitic fluid compared to plasma, associated with disease status and patient outcomes. Indoleamine 2,3-dioxygenase 1 (IDO1) metabolic activity (L-tryptophan metabolism) and its metabolite L-kynurenine were higher in ascites than plasma, unlike arginase, inducible nitric oxide synthase, or glutaminase 1. IL4I1 metabolic activity (L-tryptophan to indole-3-pyruvic acid, L-phenylalanine to phenyl pyruvic acid, or L-tyrosine to 4-hydroxyphenyl pyruvic acid) was higher in ascites than plasma and higher in FIGO stage IV than III [ 168 ]. Multiple metalloproteases, including A Disintegrin And Metalloproteinase (ADAM)−17 [ 206 ], have been detected at higher levels in ascites fluid than in the serum. Higher ascites levels of matrix metalloprotease (MMP)−2 or extracellular matrix metalloproteinase inducer/CD147 (EMMPRIN) correlate with progression-free interval (PFI) in patients treated with neoadjuvant chemotherapy, whereas higher ascites MMP-3 or uPA levels correlate with shorter PFI in patients undergoing primary debulking surgery [ 207 ]. Other proteins identified as diagnostic when expressed at high levels in ascites include pigment epithelium-derived factor (PEDF)−36 and carbonic anhydrase II [ 208 ]. The expression of ceruloplasmin at higher levels in ascites from chemo-resistant patients, compared to chemo-sensitive [ 195 ], may be a consequence of liver damage in chemo-resistant patients. Technological advances have improved the detection and measurement of cfDNA and tumor-derived cfDNA (ctDNA) in biological fluids to evaluate their potential as diagnostic and response biomarkers. The REZOLVE clinical trial investigated the effect of administering bevacizumab via the intraperitoneal (IP) route to reduce the accumulation rate of refractory ascites in patients with platinum-resistant ovarian cancer deemed unsuitable for further chemotherapy. cfDNA and ctDNA were detected at higher levels in the ascites than in the plasma, whereas endothelium-derived cfDNA was detected at lower levels. High ascites ctDNA levels are associated with a shorter paracentesis-free interval. Plasma ctDNA, but not cfDNA, ctDNA, or endothelial cells-cfDNA, is associated with unfavorable survival and CA-125 [ 200 ]. The similarities between ovarian cancer cells in ascites fluid and solid tumor tissue from the same patient suggest ascitic fluid's utility for diagnosis. However, significant differences were observed between cells from the two sites. Ovarian cancer cells in ascites exhibit chromosome instability [ 209 ]; and 52% concordance in ex vivo homologous recombination deficiency status compared to matching solid tumors [ 210 ]. Biomarkers indicating ERK/AKT pathway activation have been observed in solid tumor cells, but not in ascites cells [ 211 ]. Tumor mortalin levels are prognostic of patient outcomes, with blood mortalin levels associated with tumor stage, grade, and histology [ 212 ]. However, this correlation was not observed in ovarian cancer patients. Higher mortalin levels were found in ovarian cancer ascites than benign ascites but were not a prognostic factor. Thus, using ascites as a surrogate for ovarian cancer tissue evaluation is limited due to significant differences in protein expression between the two specimen types. Identifying and validating similarities and differences in biomarker expression in ovarian cancer cells from ascites versus tumor tissues can provide insights into the disease's biology and progression.

Tumor cells in malignant ascites can exist as single cells or multicellular aggregates (spheroids) with a heterotypic cellular composition [ 35 ]. Spheroids can act as ‘metastatic units’ that are critical for establishing metastases at distant sites [ 36 ] and the development of chemoresistance and disease recurrence, by adapting to their changing microenvironment and communicating and attaching to other cells/tissues in the peritoneal cavity. For example, these spheroids proliferate slowly and express high levels of transglutaminase 1 (TGM1), which contributes to tumor dissemination and ascites production [ 37 ]. Multicellular aggregates found in ascites have been shown to be composed of ovarian cancer cells expressing nuclear PAX8 protein. These cells also expressed mesenchymal and epithelial markers like alpha-smooth muscle actin and EpCAM, respectively, indicating that the hybrid phenotype assists in maintaining cellular plasticity that enables peritoneal implantation to establish metastasis [ 38 ]. A recent study has shown that ascites-derived EpCAM-positive tumor cells are dependent on platelet derived growth factor receptor beta (PDGFRβ) NUAK family SNF1-like kinase (NUAK1)-fibronectin induced tumorsphere formation [ 39 ]. Ascites-derived tumor cells may also express high levels of nectin-4 and nectin-1 on the surface, which promotes cell–cell adhesion, anchorage-independent viability, migration, and drug sensitivity [ 40 ]. Several proteins expressed in the tumor cells present in malignant ascites are associated with the development of drug resistance. These include members of the heat shock protein 70 kDa (Hsp70) family [ 41 ], multidrug-resistant (MDR) pumps, such as ATP-binding cassette (ABC) transporters, P-glycoprotein, multidrug resistance-associated proteins [ 42 , 43 ], and lung resistance-related protein (Lrp1, or major vault protein (MVP)), which are involved in the intracellular transport of substrates, including cisplatin, and are associated with poor prognosis and disease recurrence [ 44 , 45 ]. Malignant ascites also contains cancer cells with ‘stem-like’ properties. These cells often express drug transporters like ABCG2/BCRP1 (breast cancer-resistance protein-1) that render them resistant to many chemotherapeutic agents, a process that can be measured by cell extrusion of Hoechst dye, a common characteristic of stem cells. This dye-excluding side population expresses the embryonic stem cell markers Nanog homeobox (NANOG), POU class 5 homeobox 1 (POU5F1, known as OCT4), and developmental pluripotency-associated 3 (DPPA3, also known as STELLAR), and is tumorigenic and chemoresistant [ 46 ]. Furthermore, in syngeneic mouse models of ovarian cancer soluble factors in ascites can also help maintain ovarian cancer stem cells, e.g., alpha-2-macroglobulin, murinoglobulin [ 47 ]. Metabolic shifts in malignant cells in the ovarian cancer ascites fluid can support tumor proliferation and dissemination and contribute to drug resistance. For example, RNA sequencing analysis of tumor spheroids isolated from ascites showed a significant upregulation of oxidative phosphorylation relative to solid tumor tissue samples from patients with ovarian cancer. This metabolic switch was shown to confer a survival benefit and enhance the invasiveness of ascites-derived tumor spheroids [ 48 ]. Cancer cells in ascites can acquire traits enabling increased cell motility and migration that can ultimately promote peritoneal metastases. The death-domain-associated protein (DAXX) was shown to promote ascites cell proliferation, migration, and colony-forming capabilities in vitro via activation of the extracellular signal-regulated kinase (ERK) signaling pathway and tumor growth and ascites formation in vivo [ 49 ]. Further, transcriptomic profiling comparing primary ovarian tumor tissue, ascites, and peritoneal metastasis revealed that ascitic cancer cells downregulated genes associated with metabolism, cell cycle, DNA replication, and selectively upregulated programs involved in cell migration, chemokine signaling, cytokine-cytokine receptor interaction. This indicates that the ascitic microenvironment can induce a state of quiescence in the cancer cells associated with low metabolic activity, and once the cancer cells reach the preferential metastatic site, the cells are reactivated and proliferate [ 50 ]. Ascites also influences pre-metastatic niche formation by weakening the mesothelial barrier. It can destabilize the mesothelium by modifying cell junctions and reorganizing the actin cytoskeleton, and densification and reorganization of extracellular matrix proteins, thereby promoting integrin-dependent adhesion of ovarian cancer spheroids [ 51 ]. Mesothelial cells form a thin layer on the surface of the peritoneum. Ovarian cancer cells breach this protective mesothelial layer to establish metastasis. Mesothelial cells can also transform into cancer-associated mesothelial (CAM) cells that act as cancer allies [ 52 ]. Viable mesothelial cells released into the ascites may form complexes with disseminated ovarian cancer cells and promote their survival during adhesion to the peritoneum, leading to barrier disruption. Transforming growth factor-beta 1 (TGF-β1) in malignant ascites fluid induces mesenchymal transition into CAMs. CAMs develop decreased microvilli and increase intercellular gaps, which allows cancer cells to adhere to their surfaces, and allows the CAMs to become components of the tumor stroma [ 52 , 53 ]. Mesothelial cells secrete factors that promote tumor growth such as dipeptidyl peptidase IV (DPP-4), lysophosphatidic acid (LPA), and VEGF. LPA has been shown to improve tumor cell adhesion, migration, and invasion [ 54 , 55 ]. A study has shown that the ascites from ovarian cancer patients are also enriched in desmin (DES + ) mesothelial cells that can remold the ascites microenvironment through recruiting T cells and macrophages via the chemokine (C-X-C motif) ligand 12 (CXCL12) –CXC chemokine receptors 4 (CXCR 4) axis [ 56 , 57 ]. Ascites from ovarian cancer patients also contains ascites-derived stromal cells, known as ‘hospicells’, that are closely associated with tumor cells. These cells share homology with human mesenchymal stem cells (MSCs) derived from bone marrow or adipose tissue and are characterized by the presence of different markers, including CD9, CD10, CD146, CD166, CD29, and HLA-1. These cells have been shown to promote tumor growth by enhancing angiogenesis [ 34 ]. These MSCs in ascites can attract and activate macrophages, resulting in increased secretion of pro-angiogenic cytokines e.g., IL-6, IL-8, and VEGF, that promote tumor progression and development of chemoresistance [ 58 ]. In addition, hospicells can also inhibit the proliferation of human CD4 and CD8 T-cells, thus helping cancer cells evade immune surveillance[ 59 ]. In another study, the presence of these mesenchymal progenitor cells with ‘stem-like’ characteristics in ascites was associated with high expression of EGFR and Flt-4 (VEGF receptor) that are critical for promoting metastasis and angiogenesis [ 60 ]. Hospicells can promote endothelial cell proliferation, migration, and differentiation. Ascites-derived tumor-associated MSCs have also been shown to confer thermo-protective effects on tumor cells via CXCL12. This can influence the effectiveness of hyperthermic intraperitoneal chemotherapy in ovarian cancer patients [ 61 ]. Crosstalk between cancer stem cells and tumor-associated MSCs has been shown to promote chemoresistance, stemness, and metastasis through the PDGF signaling pathway [ 62 ]. Another study has demonstrated that KLF9 is significantly downregulated in MSCs, which enhances malignant potential, epithelial-to-mesenchymal transition, and facilitates stemness in ovarian cancer [ 63 ]. Thus, these ‘hospicells’ play a pivotal role in promoting ovarian cancer disease progression and metastasis. In ovarian cancer, cancer-associated fibroblasts (CAFs) can originate from various sources, including normal fibroblasts; epithelial, endothelial, and mesenchymal stem cells (MSCs); and adipocytes. The role of CAFs in promoting cell migration, invasion, and chemoresistance has been well-studied. Tumor spheroids in ascitic fluid express CAF markers, such as alpha-smooth muscle actin (α-SMA), the levels of which correlate with clinical stage IV disease [ 64 ]. Fibroblasts identified in ascites show high expression of epithelial-to-mesenchymal (EMT) transition markers, suggesting their role in promoting metastasis [ 65 ]. Fibroblasts secrete various tumor-promoting factors, including collagen type 1 alpha 1, which facilitate extracellular matrix remodeling, migration, and invasion of ovarian cancer cells [ 66 ]. CAFs with an immunomodulatory phenotype also secrete multiple cytokines, such as interleukin (IL) IL-6, IL-8, CXCL16, chemokine ligands (CCL) CCL2, and CXCL1. The expression of these ligands can modulate their communication with other cells in the tumor microenvironment (TME), thereby influencing disease progression, metastasis, and response to therapy [ 67 ]. Ovarian cancer is characterized by immunosuppressive TME wherein dysregulated immune cells significantly influence disease progression and chemoresistance. Different immune cell populations identified in ascites include macrophages, monocytes, dendritic cells (DC), T cells, B cells, and natural killer (NK) cells [ 68 – 70 ]. Studies have shown that macrophages are the most abundant (55.5%) among different immune cells, followed by T cells (23.9%), DC (4.2%), and B cells (2.6%). However, the clinical stage may also influence immune cell populations. Single-cell RNA-seq (scRNA-seq) of malignant ascites from five patients (three chemo-naïve and two recurrent) revealed a lower abundance of macrophages (44.96%), T cells (9.53%), and DCs (2.91%) and a higher abundance of B cells (9.08%). A scRNA-seq study performed using primary ovarian tumor, omental metastatic lesion, pelvic lymph node, malignant ascites, and peripheral blood of ovarian cancer patients with differential responses to chemotherapy highlighted the complexity of the TME [ 56 ]. These differences in immune cell populations may influence the immune response to tumorigenic growth and therapeutic efficacy. Macrophages play an important role in innate immunity, and depending on their phenotype, they can be anti- or pro-tumorigenic. In malignant ascites, tumor-associated macrophages (TAMs) can be either pro-inflammatory M1 macrophages, predominant in naïve samples, or anti-inflammatory M2 macrophages, primarily observed in recurrent samples [ 71 ]. Among the M2 macrophages, the M2D are the most abundant immunosuppressive cells (M2D-polarized cells that are IL-10 high , IL-12 low , ILT3 high , CD86 low ) that originate from blood monocytes. The presence of high concentrations of leukemia inhibitory factor and IL-6 in ascites can enhance TAM-like cell induction [ 72 , 73 ]. A study has shown functional differences between tumor-enriched and ascites-enriched macrophages. The former recruits T cells and shapes the immunosuppressive TME, while the latter exhibits a high expression of the S100A family of proteins and a low level of human leukocyte antigen class II (HLA-II) genes. These ascites-enriched macrophages also upregulate genes associated with the leukocyte migration pathways, contributing to a pro-tumorigenic environment in ascites [ 56 ]. A scRNA-seq profiling study identified a novel subset of perilipin-2 high macrophages that correlate with ovarian cancer progression. These ‘ascites-associated macrophages’ show high lipid droplet accumulation and promote ascites development and cancer metastasis via the hypoxia-inducible factor 1 α (HIF1α)/secreted phosphoprotein 1 (SPP1) signaling axis [ 74 ]. Thus, establishing a link between lipid metabolism and hypoxia in the ascites microenvironment underlying the critical role of ascites in transcoelomic metastasis. Tumor-infiltrating lymphocytes (TILs) are often observed within and around tumors, including ovarian carcinomas. Elevated levels of cluster of differentiation (CD) 3 (CD3) positive TILs in ovarian cancers are associated with a better prognosis than tumors with little to no CD3 + TILs [ 75 ], and the distinct profiles of TIL subtypes may promote or inhibit immune evasion and tumor progression, especially those present in ascites, as they act as a medium for tumor dissemination. For example, regulatory T cells (Tregs) suppress tumor-specific T-cell immunity, contribute to tumor growth, and reduce survival [ 76 ]. Also, low levels of programmed cell death 1 (PD-1) + CD8 + TILs and high levels of CD57 + CD8 + tumor ascites lymphocytes are indicative of poor prognosis and a suppressive immune microenvironment [ 77 ]. Functional Tregs (CD4 + CD25 + CD3 + T-cells) constitute approximately 10–17% of all T-cells in ascites and are more abundant in stage II and IV patients than in their blood or non-malignant ascites. Additionally, a higher abundance of Tregs in ovarian cancer tissues is also associated with late-stage disease. Furthermore, a higher tumor grade was associated with reduced CD3 + CD56 + NK T cells, which play a crucial role in cancer surveillance and control, and an increased ascites-to-blood Treg (CD4 + CD25 + ) ratio. Platinum-resistant patients had a reduced ascites-to-blood NK ratio compared to platinum-sensitive patients, indicating that selective accumulation of the NK cell population in ascites before first-line chemotherapy may be a predictive factor for platinum resistance [ 78 ]. A recent study has demonstrated that a large proportion of NK cells and CD8 + T cells present in ovarian cancer ascites express tissue-residency markers that are consistent with intraepithelial type 1 innate lymphoid cells (ieILC1) cells with cytolytic potential. This subset also had high expression of inhibitory receptor natural killer group 2 A (NKG2A) and showed sensitivity towards ovarian cancer cells. Blocking NKG2A in combination with immunotherapies targeting T cells could be beneficial in ovarian cancer patients [ 79 ]. A study of the NK cell lipidome in ascites revealed that the polar lipids in ascites can induce NK cell dysfunction and suppress cytotoxic T cells. Further, inhibiting the scavenger receptor class B type 1 (SR-B1)-mediated lipid uptake can restore NK cell function, suggesting a potential immunotherapy target [ 80 ]. Another study has reported that TGF-β1 is also an inhibitory factor abundant in ascites contributing to NK cell dysfunction. Thus, blocking TGF-β1 signaling using galunisertib can augment NK cell response in ovarian cancer patients [ 81 ]. Recently, a higher frequency of effector-type Tregs in HGSOC has been reported than in other histotypes [ 82 ], which has also been correlated with PD-1 expression on CD8 + T cells. Ascites is also explicitly enriched in memory T-cells and can act as an important pool of TILs for omental metastasis, suggesting the presence of multiple immune-inhibitory networks in malignant ascites and its role in influencing the ovarian TME [ 56 , 79 ]. Further, mucosal-associated invariant T (MAIT) cells were also enriched in the ascites and peripheral blood of patients responsive to chemotherapy. These MAIT cells in chemosensitive patients overexpressed genes associated with T cell activation (e.g., ZFP36, JUN, DUSP1, NCR3, and KLRB ), whereas chemoresistant patients overexpressed genes associated with immunosuppression (e.g., LAG3 and IFITM3 ). This indicates that activated MAIT cell levels in the peripheral blood could be a potential predictor of chemotherapy response [ 56 ]. DCs are antigen-presenting cells that act as a bridge between the innate and adaptive immune systems by collecting and presenting antigens to T cells. The percentage of DC subset cells has been described in malignant ascites, with 1.8% Brain Development Complementary Antigen (BDCA) BDCA1 + myeloid DCs (mDCs), 0.9% BDCA-3 + mDCs, and 2.8% D16 + mDCs, 2.1% plasmacytoid DCs (pDCs). When compared to the percentage of T cells (45% CD4 + T helper cells and Tregs, and 33% for CD8 + cytotoxic T cells), the percentage of pDCs was positively correlated with CD4 + T cells. A trend towards improved outcomes in patients with low percentages of CD4 + T cells was observed, but the percentage of DC and T cell subsets in ascites did not correlate with the survival of patients with HGSOC [ 83 ]. A study has shown that conventional DCs in ascites could be a potential source of lysosomal-associated membrane glycoprotein (LAMP) LAMP + DCs, characterized by enhanced expression of CD40 (associated with interaction between T cells and myeloid cells) and IL12B (involved in T helper cell development), in tumor tissue. These LAMP + DCs could potentially promote T helper cell infiltration and differentiation in ovarian tumors [ 56 ]. Monocytes also play a role in the immune response to cancer. Subsets of monocytes have been characterized in ascitic fluid and blood samples from chemotherapy-naïve patients. Specifically, CD14 high CD16 low intermediate blood monocyte levels (IBM) in the blood were positively correlated with Treg, CD4 + T cell, and B-cell levels. Conversely, a significant inverse correlation between the proportion of IBMs and cytotoxic CD8 + T cells, NK cells, and the CD8 + /Treg cell ratio was observed [ 84 ]. Taken together, these results suggest that circulating IBM expansion is associated with a decrease in the effector/Treg cell ratio in tumor ascites fluid. Studies have demonstrated that monocytes can suppress the anti-tumor T cell function. Their presence in the ovarian TME correlated with poor survival and resistance to immune checkpoint blockade [ 85 ]. Platelets play an important role in cancer progression and metastasis. They can extravasate from circulation into malignant ascites, where they promote tumor progression [ 86 ]. An abnormally high platelet count (thrombocytosis) is an independent predictor of poor PFS in ovarian cancer, and a high pre- and post-treatment platelet-to-lymphocyte ratio is associated with malignant ascites [ 87 ]. Platelets have a high capacity to bind to collagen in the peritoneum, making it challenging to quantify them accurately in ascites. However, studies have reported the presence of platelets in both resting and activated states, characterized by the presence of platelet factor 4 (PF4) and glycoprotein (GP) IIb/IIIa [ 88 ]. In ascites, platelets can promote metastasis and tissue factor markers, favoring a disease-promoting phenotype [ 89 ]. Platelets are a major source of circulating TGF-β in the bloodstream, and TGF-β increases ascites formation and volume in preclinical models of ovarian cancer [ 90 ]. In orthotopic animal models, interference in the TGF-β signaling pathway using soluble TGF-β receptor II (TGFBR II) reduced tumor weight and VEGF, the major contributor to ascites formation, and improved lymphatic drainage, thereby reducing ascites formation. The role of platelets in ascites suggests that targeting platelets might be an indirect way to prevent ascites formation and ascites-mediated metastasis in ovarian cancer. Taken together, these studies highlight the complex immune landscape within ascites fluid contributing to the immune-suppressed environment and suggest potential targets for immunotherapeutic intervention.

Ovarian cancer is the second leading cause of death from gynecologic malignancies in the United States, with an estimated 20,890 new cases and 12,890 deaths by 2025 [ 1 ]. Unfortunately, there are no effective screening modalities, and approximately 70% of patients are diagnosed at advanced stages, leading to poor survival outcomes with a five-year survival rate of 31.4% in patients with distant metastases [ 1 , 2 ]. The current standard of care includes a combination of cytoreductive surgery and platinum-based chemotherapy, with consideration for the addition of vascular endothelial growth factor (VEGF) therapy, bevacizumab, in patients with suboptimal debulking and stage IV disease [ 3 – 5 ]. Recent advances in maintenance therapy with bevacizumab alone and/or poly (ADP) ribose polymerase inhibitors, such as Olaparib, Niraparib, and Rucaparib, in newly diagnosed advanced ovarian cancer have been shown to provide significant progression-free survival (PFS) benefits in certain subgroups [ 4 – 9 ]. While remission is achieved in many patients, most experience recurrence and eventually succumb to their disease [ 10 , 11 ]. Advanced ovarian cancer often spreads through the transcoelomic route, which is more common than the hematogenous and lymphatic pathways 11 , 12 ]. Detached tumor cells move through the fluid that accumulates in the peritoneal cavity, known as ascites, to seed the peritoneal surfaces, omentum, and other abdominal organs, resulting in the development of diffuse carcinomatosis. This can prevent optimal surgical resection and is an unfavorable factor for survival in patients with advanced and recurrent disease [ 11 ]. The cause of ascites fluid accumulation in the peritoneal cavity is multifactorial and involves a combination of increased fluid production and impaired drainage. For example, inflammatory cytokines and growth factors, such as VEGF increase angiogenesis and vascular permeability leading to fluid leakage into the peritoneal cavity, meanwhile, tumor cells can obstruct peritoneal lymphatic vessels by reducing fluid clearance. Further, increased capillary permeability allows proteins to leak into the peritoneal cavity, raising oncotic pressure within the ascitic fluid, which in turn draws more fluid into the cavity [ 13 ]. As such, large-volume ascites is more common in advanced-stage patients. Tumor cells exist as spheroids, cell aggregates, and/or single cells in ascitic fluid, and participate in tumorigenesis, metastasis, and chemoresistance [ 11 , 14 ]. These factors are important for disease recurrence and negatively affect overall survival (OS) [ 11 , 12 ]. Ascites plays a significant role in promoting metastasis and can be a prognostic factor for ovarian cancer patients. A large-scale multicentric retrospective study has shown that the presence of tumor cells in ascites is associated with poor prognosis in patients with epithelial ovarian cancer but not in patients with mucinous ovarian cancer [ 15 ]. Furthermore, other clinicopathological factors like age, histology, platelet count, serum albumin, and total protein are also predictive of long-term survival in ovarian cancer patients presenting with malignant ascites [ 16 ]. The study of this biospecimen allows for a deeper understanding of the biological mechanisms of metastasis and responses to current treatments and could serve as a platform for developing improved therapies. The examination of ascites fluid samples to identify the presence of cancerous cells is a critical aspect of cancer staging. It is essential to differentiate malignant cells from reactive mesothelial cells because they can exhibit similar morphological features. To enhance the morphological characteristics of nuclear and cytoplasmic features, fluid monolayer slides, such as ThinPrep®, BD SurePath™, and Cytospin™, are used in clinical cytology laboratories [ 17 , 18 ]. In addition, in the initial workup for the evaluation of malignancy, ascites cells are processed as blocks, which are sectioned for immunohistochemical (IHC) analysis of cell differentiation markers to identify the origin of epithelioid cells. Because epithelial and mesenchymal cells sometimes express the same markers, a panel of antibodies against two or three mesothelial and epithelial markers is used to distinguish epithelial cells from mesothelial cells. Mesothelial markers include Calretinin, Wilms’ tumor 1 (WT-1), cytokeratin 5/6 (CK5/6), and podoplanin (D2-40), whereas epithelial markers include epithelial cell adhesion molecule (EpCAM) antibodies MOC-31 or Ber-EP4, Claudin-4, and tumor-associated glycoprotein 72 (TAG-72) B72.3 (Fig.  1 ). If atypical cells are positive for epithelial markers and exhibit morphological features characteristic of malignancy, the ascitic fluid is considered positive. Malignant cells must be distinguished from benign epithelial cells typically observed in conditions such as endometriosis and endosalpingiosis. In these cases, cells are usually sparse, feature terminal bars (dark, dot-like structures) with cilia, and exhibit wild-type p53 expression. If atypical cells test positive for mesothelial markers but show no reactivity to epithelial markers, the ascitic fluid is considered negative for malignancy. However, if the mesothelial cells are significantly enlarged (more than six times the normal size), demonstrate marked nuclear pleomorphism, are found in cohesive clusters of more than 20 cells, display brisk mitotic activity, or show atypical mitoses, the possibility of mesothelioma must be excluded [ 19 – 21 ]. Fig. 1 Ascites fluid can be used for pathological and molecular characterization. Cells of malignant potential from ascites fluid are prepared as a cell block. Sections of the cell block are stained with various markers to distinguish mesothelial cells from epithelial cells (calretinin, podoplanin/D2-40 antibody and Wilm’s Tumor-1). The ovarian histological subtypes can be distinguished using various markers alone or in combination including epithelial cell adhesion marker (EpCAM/MOC-31 antibody), paired box gene 8 (PAX-8), Napsin A and p53. Representative hematoxylin and eosin (H & E) and immunohistochemical (IHC) stained sections are shown. Cell-free DNA derived from tumor cells in ascites can be used for genetic testing and guiding treatment decisions. (Created with BioRender.com) Ascites fluid can be used for pathological and molecular characterization. Cells of malignant potential from ascites fluid are prepared as a cell block. Sections of the cell block are stained with various markers to distinguish mesothelial cells from epithelial cells (calretinin, podoplanin/D2-40 antibody and Wilm’s Tumor-1). The ovarian histological subtypes can be distinguished using various markers alone or in combination including epithelial cell adhesion marker (EpCAM/MOC-31 antibody), paired box gene 8 (PAX-8), Napsin A and p53. Representative hematoxylin and eosin (H & E) and immunohistochemical (IHC) stained sections are shown. Cell-free DNA derived from tumor cells in ascites can be used for genetic testing and guiding treatment decisions. (Created with BioRender.com) Specific subtyping of ovarian cancer is rarely performed on ascitic fluid preparations, because many tumors have overlapping IHC profiles. However, some cases have a classic morphology in which a definitive diagnosis can be made with high certainty. High-grade serous carcinoma (HGSOC) has high nuclear pleomorphism, high mitotic activity, papillary growth, and psammomatous calcification, and is positive for p53 (mutated, null, or cytoplasmic staining pattern), p16, WT-1, and paired box gene-8 (Pax-8) (Fig.  1 ) [ 22 ]. Low-grade serous ovarian carcinoma and serous borderline tumors exhibit bland nuclear characteristics, express wild-type p53, are morphologically similar to mesothelial cells, and must be distinguished by IHC staining, as stated above. Morphologically, clear cell carcinoma has mild to moderately pleomorphic cells with abundant vacuolated cytoplasm and is positive for Napsin A, negative for estrogen and progesterone receptors, and exhibits wild-type p53 expression [ 23 ]. Endometrioid and mucinous carcinomas typically have overlapping IHC profiles with malignancies at other sites and cannot be diagnosed using ascites alone [ 24 ]. Cytopathology can also inform on disease severity and prognosis. A recent retrospective feasibility study demonstrated that cytological and architectural characterization of tumor spheroids in ascites was highly associated with OS and PFS, independent of chemotherapy response [ 25 ]. Another study has shown that positive cytology in ascites is associated with poor prognosis and a shortened recurrence-free interval. Further, the presence of pathologically identified tumor cell aggregates is indicative of cellular-level peritoneal micro-metastasis. Thus, cytopathology can be helpful in stratifying patients with a high risk of disease recurrence [ 26 ]. The use of deep learning artificial intelligence (AI) models has also been reported for the detection of ovarian cancer using ascitic fluid cytological whole slide images. Among the different models tested, the clustering-constrained attention multiple-instance learning-SB (CLAM-SB) model demonstrated the best performance. Further, errors made by AI were correctly identified by humans and vice versa, indicating that the complementary strengths of AI with human diagnostics can potentially improve the accuracy and reproducibility of ovarian cancer diagnostics [ 27 ]. Owing to the diverse pathological nature of ovarian cancer, cytopathologic evaluation of ascitic fluid is essential for an accurate diagnosis and may inform on the severity of the disease and prognosis. Furthermore, this fluid can be sampled multiple times during treatment and disease progression, thereby providing a resource to enhance our understanding of the biological development of disease and therapeutic resistance. Emerging tools, such as artificial intelligence, are beginning to show promise in enhancing diagnostic accuracy and may aid pathologists in the future. Ascites fluid is also a source of tumor-derived cell-free DNA (cfDNA), which can provide unique access to the tumor prior to neoadjuvant chemotherapy or interval debulking surgery. Liquid biopsy using cfDNA in ascites is a less invasive method to analyze genomic profiles and monitor clonal evolution. Studies have shown that cfDNA can be used as an alternative to tumor biopsy for testing homologous recombination deficiency (HRD). Mutational concordance between genomic DNA from tumor biopsy, and cfDNA from ascites, and plasma has been compared and showed that clinically useful information regarding somatic mutations can be obtained reliably from ascites cfDNA and corresponds to the tumor tissue [ 28 ]. Another study has confirmed that ascites is a source of high concentration cfDNA that yields higher variant allele frequency relative to DNA isolated from tumor tissue [ 29 ]. Further, cfDNA isolated from longitudinal ascites sampling over time can provide actionable information regarding treatment with disease progression. Targeted next-generation sequencing performed on cfDNA isolated from tumor cells in ascites revealed similar somatic cancer-driving variants (e.g., TP53 , BRCA1 , BRCA2 , NF1 , RB1 , etc.) as in the formalin-fixed paraffin-embedded tissues and clinical reports. HRD status could be identified from cfDNA at an earlier time which could guide decisions on the use of PARP inhibitors in a maintenance and/or neoadjuvant setting. In addition, ascites sampling for cfDNA at the time of relapse also revealed changes in the genome instability markers and copy numbers with disease progression [ 30 , 31 ]. Thus, profiling cfDNA from ascites even when available in small volumes is feasible and yields reliable information regarding the pathogenic mutation status and genomic instability, which can be used clinically to guide personalized medicine in ovarian cancer (Fig.  1 ). Emerging approaches for the detection of ovarian cancer cells in ascites include combining molecular profiling and cytological evaluation. For example, an immunomagnetic cancer cell enrichment step followed by a combinatorial molecular approach involving TP53 deep sequencing and PAX8 cytological straining demonstrated an improved sensitivity in the detection of the neoplastic fraction of cells in ascites samples collected from ambiguous cases [ 32 ].

Exosomes are a class of small extracellular vesicles (sEV) that act as messengers by releasing and transporting specific cellular components, such as bioactive lipids, metabolites, proteins, and nucleic acids. Because of their role in promoting ovarian cancer development and progression as well as their potential as a source for liquid biopsy, ascites-derived exosomes/EVs are of particular interest. However, EVs in most biofluids, including ascites, are highly heterogeneous, reflecting the diversity of the surrounding tissues and organs. One of the challenges in exosome/EV research is the identification of different subpopulations and their primary sources. To better understand this heterogeneity, peritoneal ascites EVs were captured using cellulose nanofiber sheets attached to various peritoneal cavity organs and ovarian tumors in a mouse model and on human patient tissues ex vivo and intravitally. Their analysis revealed cancer-associated microRNA (miRNA) profiles distinct from disease stage and tissue location, demonstrating the location-dependent heterogeneity of ascites EVs [ 91 ]. A recent study has shown that the EVs in blood and ascites differ not only in the levels of the miRNA cargo but also in size and the expression of the surface epitopes. Interestingly, the ascitic fluid EVs expressed high levels of EpCAM, CD133, CD24, and CD44, whereas serum EVs had high expression of CD62P and CD42a (markers for platelet activation) and CD86 [ 92 ]. The distinct RNA of EVs isolated from HGSOC ascites compared with normal or benign peritoneal fluids can serve as a biomarker for disease stage and progression [ 93 ]. A 7-gene panel of EVs derived from human serum, plasma, ascites, and a mouse model were linked to tumor progression and metastasis and overlapped with plasma-derived EVs at various stages [ 94 ]. Higher levels of matrix metalloproteinase 9 in annexin-V-binding EVs have been found in malignant ascites than in ascites from patients with cirrhosis [ 95 ]. Higher numbers of CUB domain-containing protein 1 EVs were detected in malignant ascites than in benign ascites [ 96 ]. The cargo of EVs in malignant ascites may play a crucial role in promoting angiogenesis and metastasis. Soluble E-cadherin, a protein abundant in malignant ascites, is associated with EVs and triggers β-catenin and nuclear factor-kappa B (NF-κB) signaling to activate endothelial cells and promote angiogenesis [ 97 ]. Additionally, the hypoxic TME may lead to alterations in EV content to include more oncogenic factors, such as signal transducer and activator of transcription 3 (STAT3) and Fas cell surface death receptor (FAS) proteins, which increase migration, invasion, and chemoresistance [ 98 ]. EVs from ovarian cancer cells and malignant ascites can modify the TME to facilitate omental metastasis by increasing migration, invasion, proliferation, and α-SMA and fibroblast activation protein (FAP) levels in adipose-derived stem cells. This study demonstrated that elevated miR-320a (highly abundant in EVs) promotes tumor cell growth and omental metastases in vivo [ 99 ]. Further, peritoneal adipose-derived stem cells secrete EVs rich in epidermal growth factor receptor (EGFR) signaling molecules in the ascites that can promote ovarian cancer cell proliferation at the metastatic site [ 50 ]. In a recent study, integrin subunit alpha 6 (ITGA6) was shown to be secreted by ovarian tumor cells as a biologically active protein in exosomes present in ascites. It can activate the insulin-like growth factor 1 receptor (IGF1R) pathway, contributing to the pre-metastatic niche formation by engaging laminins and/or mesothelial cells [ 100 ]. Comparison of paired ascites samples from ovarian cancer patients before and after neoadjuvant chemotherapy revealed an increased presence of stress granule markers (Ras-GTPase-activating protein-binding protein 1 (G3BP1), T-cell intracellular antigen 1 (TIA1), and Poly(A) binding protein cytoplasmic 1 (PABPC1)) post-chemotherapy, indicating that stress granules could be exported via EVs [ 101 ]. These findings underscore the importance of ascites-derived EV-mediated signaling in the acquisition of therapy resistance. The glycoproteome of EVs isolated from the ascites of ovarian cancer patients with advanced disease and benign gynecological tumors may also differ significantly. Glycopeptides carrying one fucose moiety were decreased considerably in EVs isolated from patient ascites with advanced disease. Also, α2,6 sialylation was a specific feature observed in these patients, which could be a potential novel detection marker for HGSOC [ 102 ]. Alpha-2 Heremans Schmid glycoprotein (AHSG) was shown to be downregulated in ascites-derived exosomes. These ascites-derived exosomes containing AHSG inhibit ovarian cancer cell motility and migration via the p53/focal-adhesion kinase (FAK)/Src signaling pathways [ 103 ]. The transfer of EVs between tumor cells, immune cells, and other cells in the TME can contribute to an immunosuppressive environment and promote malignant progression. For instance, T-cell activation can be impaired by exosomes derived from malignant ascites without affecting their viability. This inhibition is temporary and reversible, and T cells can be reactivated once exosomes are removed [ 104 ]. In contrast, exosomes from TAMs in malignant ascites can suppress endothelial cell migration by transferring miR-146b-5p and repressing NF-κB signaling [ 105 ]. The additional acellular components of ascites include cytokines, chemokines, growth factors, and hormones. These molecules create a pro-tumorigenic and immunosuppressive environment by affecting immune-cell function. Higher levels of IL-6, IL-8, IL-10, IL-15, interferon-gamma inducible protein 10 (IP-10), monocyte chemoattractant protein (MCP-1), macrophage inflammatory protein −1β (MIP-1β), and VEGF, and lower levels of IL-2, IL-5, IL-7, IL-17, and platelet-derived growth factor-BB (PDGF-BB), and regulated on activation, normal T cell expressed and secreted (RANTES) have been observed in ascites than in plasma [ 106 – 108 ]. Higher levels of ascites osteoprotegerin, IL-10, leptin, CCL-2, and VEGF were associated with shorter PFS, potentially due to the anti-apoptotic activity of IL-10 [ 109 , 110 ]. Higher levels of IL-6 in ascites are also associated with shorter PFS [ 111 ]. Increased IL-6 concentrations in serum and ascites of ovarian cancer patients have been associated with poor prognosis and chemoresistance [ 112 ]. A significant association between lower tumor necrosis factor-alpha (TNF-α) and IL-6 levels in ascitic fluids and multiple immune checkpoints in CD8 + T cells suggests their role in the immune TME [ 113 ]. IL-6 and IL-10 levels in ascites are correlated with CD163 expression in macrophages, and their levels are inversely associated with relapse-free survival [ 25 ]. Furthermore, a recent study has shown that the high level of IL-6 in patient-derived acellular ascites fluid can induce chemoresistance in ovarian cancer cell lines by activation of the Janus kinase (JAK)/STAT, phosphoinositide 3-kinase-protein kinase B pathway (PI3K-AKT), and mitogen-activated protein kinase (MEK) pathways [ 114 , 115 ]. This highlights the importance of the acellular compartment of ascites in modulating treatment response. IL-8 levels in ovarian cancer patient ascites have been shown to correlate with poor prognosis and chemoresistance. IL-8-dependent chemoresistance is mediated by increased expression of MDR1, apoptosis-inhibitory proteins, and activation of PI3K/Akt and Ras/MEK/ERK signaling pathways [ 116 ]. High levels of interferon-gamma (IFNγ) in ascites are associated with advanced disease and shorter disease-free survival (DFS) and OS. Thus, ascites IFNγ levels can be used as prognostic markers and potential targets for immunotherapy [ 117 ]. A later study found lower expression levels of IL-17a and higher levels of IL-10 in ascites, suggesting that the combination of surgical status and IL-17a and IL-21 expression levels in ascites can be used as a practical risk-scoring system to predict patient outcomes [ 118 ]. Further, IL-12 is pivotal for the activation of T cells and NK cells. Combining IL-12 with dual immune checkpoint inhibitors (anti-PD1 and anti-cytotoxic T-lymphocyte associated protein 4 (anti-CTLA4)) in a preclinical syngeneic model of advanced ovarian cancer (ID8-VEGF model). Ascites derived from these mice treated with IL-12 and dual checkpoint inhibitors had a lower frequency of tumor cells and a broad change in the immune landscape, particularly decreased myeloid function within the ascites, resulting in a pro-immune phenotype [ 119 ]. IL27 is co-expressed with EBI3 in monocytes/macrophages present in ascites. A preclinical study has shown that intraperitoneal administration of β-glucan and IFNγ can clear ascites and promote disease regression by stimulating differentiation of monocytes into IL27 + anti-tumor macrophages [ 120 ]. The expression of pro-inflammatory cytokines in ascites is linked to aggressive tumor phenotypes, highlighting their impact on disease severity and metastatic potential. Higher levels of CCL18 in ascites from advanced HGSOC relative to benign conditions and other cancers could stimulate tumor cell migration [ 121 , 122 ]. CCL22, CXCL9, CXCL10, and CXCL12 levels are associated with CCR4 +, CCR5 +, CXCR3 +, and CXCR4 + T cells in ascites, suggesting that these chemokines influence T-cell migration and interactions [ 123 ]. Stress hormones in ascites are associated with inflammatory and immunosuppressive cytokines as well as decreased T-cell function [ 106 ]. Specifically, normetanephrine and metanephrine metabolites were correlated with increased pro-inflammatory factors (VEGF, IL-6, and MCP-1, MCP-3). In contrast, cortisol was negatively correlated with IFNγ, IP-10, granulocyte–macrophage colony-stimulating factor (GM-CSF), IL-2, and FMS-like tyrosine kinase 3 ligand (FLT-3L). Galectin-3, a protein known to interact with chemokines to modulate immune response, was significantly higher in ascites and ovarian cyst fluid collected from the primary tumor of ovarian cancer patients relative to their serum. Galectin-3-induced oxidative stress could potentially degranulate neutrophils in ascites, rendering them dysfunctional [ 124 ]. Ovarian cancer has a predilection for metastasis to the adipose tissue in the peritoneal cavity especially the omentum [ 125 ]. This adipocyte-rich tumor microenvironment is not only a source of nutrients but also a source of secretory proteins called adipokines/adipocytokines including leptin, adiponectin, MCP-1, and tissue inhibitor of metalloproteinase-1 (TIMP-1), IL-6, IL-8, TNFα, etc. These factors are abundantly secreted by the omental adipocytes into the ascites of ovarian cancer patients thus contributing to metastasis tumor growth, and angiogenesis [ 126 ]. In ascites of ovarian cancer patients, FIGO stage III and IV, significantly higher levels of adiponectin have been reported relative to stage I and II [ 127 ]. Growth factors such as VEGF, hepatocyte growth factor (HGF), and GRO-1 may promote metastasis and immune suppression. Ascites with high VEGF levels often have low CD3 + CD56 + cell counts, whereas high IL-12 levels lead to increased CD3 + CD56 + cell counts, and high TNF-α levels correlate with low CD4 + CD25 + cell counts [ 118 ]. HGF is found at higher levels in ovarian cancer ascites than in benign peritoneal fluid and can stimulate human peritoneal mesothelial cell migration by activating c-Met and the downstream ERK1/2 and Akt pathways [ 128 ]. Malignant ascitic fluids containing HGF and GRO-1 (also known as CXCL1) induce senescence in peritoneal mesothelial cells and promote a pro-carcinogenic phenotype by increasing hemagglutinin (HA), urokinase-type plasminogen activator (uPA), IL-8, and MCP-1 production [ 129 ]. Further, TGFβ present in the acellular fraction of the ascitic fluid has been shown to activate TGFβ-ERK/MEK signaling in ovarian cancer cells thereby promoting cell proliferation, reduced fatty acid oxidation, and a Warburg-type rearrangement of oxidative metabolism [ 130 ]. High levels of complement component 3 in malignant ascites of obese ovarian cancer patients were associated with promoting metastasis and therapy resistance [ 131 ]. Further, fibrinogen and fibrin-derived from ascites has been shown to promote ovarian cancer spheroid implantation via Rac Family Small GTPase 1 (Rac1)-dependent cytoskeletal reorganization of mesothelial cells and peritoneal metastasis formation via involvement of αV and α5β1 integrins [ 51 ]. A correlation analysis performed between cytokines/chemokines and microbiome-derived metabolites in ascites revealed positive correlations between IL-23 and D-glucurono-6,3-lactone. Anti-inflammatory cytokine IL-10, known to promote ovarian cancer cell proliferation and migration [ 132 ], positively correlated with glucosamine, caffeine, and D-tagatose and negatively correlated with benzamide and thymol, indicating how the metabolites can influence cytokine regulation. MCP-1 negatively correlated with mevalonic acid, and 4-pyridoxic acid, potentially reducing infiltration of immune cells [ 133 ]. This highlights that the cytokines/chemokines can also form a network of interactions with not only the different cell types in the TME but also with metabolites thereby influencing disease progression and clinical outcomes. Metabolites provide valuable information regarding ovarian cancer progression as they play an important role in communication, chemoresistance, fatty acid regulation, and spheroid formation that can metastasize through the peritoneum [ 134 , 135 ]. Furthermore, metabolites in ascites can serve as valuable prognostic indicators and aid in the design of therapeutics for ovarian cancer. Notably, certain metabolites (e.g., small peptides, organic acids, esters, and cholesterol) play crucial roles in cellular processes (e.g., apoptosis, angiogenesis, and proliferation) and drug resistance. Specific metabolites such as LPA, cholesterol, and ceramides are exclusively found in malignant ascites [ 107 , 136 ]. Elevated LPA levels in blood and ascites are involved in the initiation, development, and dissemination of ovarian cancer. LPA induces the expression of VEGF and TNF-α, promotes invasion and metastasis, and induces the EMT and CAFs. Elevated cholesterol levels in ascites correlate with ovarian carcinoma malignancy and increased resistance to chemotherapy by upregulating MDR1 [ 107 ]. Upregulated ceramide (18:1) in malignant ascites fluid acts as a secondary messenger in oncogenic signaling and sustains pro-survival autophagic conditions in cancer cells [ 137 ]. Proteomic analysis performed on acellular ascites fluid has revealed the upregulation of several enzymes involved in the cholesterol metabolism pathways. Additionally, acellular ascites fluid collected from patients receiving carboplatin/paclitaxel showed high expression of apolipoprotein C-II (regulator of lipid metabolism) and ubiquitin-like modifier-activating enzyme 1 (UBA1) (involved in cholesterol metabolism through ferroptosis). Further, transmembrane protein (TMEM) 132 A (TMEM132A) was upregulated in the ascites sample collected from a patient refractory to platinum therapy, indicating its role in chemoresistance [ 138 ]. High lipid levels in ascites have been shown to induce M2-like phenotype in TAMs via inhibition of the Ras homolog family member A (RhoA)-Yes-associated protein 1 (YAP1) signaling cascade, highlighting the role of ascites-derived metabolites in cellular reprogramming [ 139 ]. Specific metabolites, such as glycolate, glucose, furanose, and fructose, are downregulated in ascites. Low glucose levels correspond to elevated glucose-1-phosphate and the Warburg effect, where mitochondria are utilized not only for cellular energy production but also for cancer progression intermediates [ 134 ]. Metabolomic profiling comparing malignant ascites samples from ovarian and gastrointestinal (GI) cancer patients revealed that the 2 cancer ascites types have distinct metabolic signatures though the profiles were similar for stage IV ovarian cancer and GI cancer. Interestingly, the stage IV ovarian cancer patient ascites had elevated microbiome-derived metabolites e.g., 3-hydroxyanthranilic acid, 4-pyridoxic acid, butyryl-L-carnitine, indole, etc. Most of these metabolites were involved in lipid metabolism and inflammation. Further, the broad variations between microbiome-derived metabolites identified in ascites from ovarian cancer patients in stage II/III and stage IV indicate that the microbiome can potentially influence the tumor microenvironment and disease progression [ 133 ]. Thus, the intricate interplay between metabolites within ascites provides invaluable insights into ovarian cancer progression and offers promising avenues for prognostication and therapeutic interventions. In summary, the complex cellular and acellular components and interactions within ascites present both a challenge and an opportunity for the future development of improved and clinically relevant therapeutic approaches.