Malignant ascites in ovarian cancer is compatible with long-term (10 year) survival with associations to clinicopathological features.

Average overall survival after onset of malignant ascites in ovarian cancer was over 6.5 years with nearly one out of ten surviving more than 10 years. Young age, low-grade histology and debulking surgery, high serum albumin and total protein, and low platelet count also predicted long survival, while patients with malignant ascites as first presentation or recurrence did not differ in outcome.

A total of 277 cases of metastatic ovarian cancer with malignant ascites were retrieved. The patients had an average age of 56.7 at the time of presentation with malignant ascites, with a mean overall survival of 69.3 months (55.0–83.5) (Fig.  1 ). The most common histological diagnosis was high-grade serous carcinoma ( n  = 42, 15.2%), followed by clear cell carcinoma ( n  = 24, 8.7%) and endometrioid carcinoma ( n  = 13, 4.7%) (Table  1 ). There were 143 cases of recurrent disease, including 130 recurrences after surgery, of which 53 also received adjuvant chemotherapy, and 13 recurrent cases after primary chemotherapy. Fig. 1 Significant histological and blood parameters on Kaplan Meier survival analysis. ( a ) Age, red– age > 50 years, blue– age ≤ 50 years, p   10 g/dL, blue– hemoglobin ≤ 10 g/dL, p  = 0.01. ( c ) Activated partial thromboplastin time, red– APTT > 30 s, blue– APTT ≤ 30 s, p   40 g/dL, blue– albumin ≤ 40 g/dL; p   40 g/dL, blue– albumin ≤ 40 g/dL, p   50 years, blue– age ≤ 50 years, p   10 g/dL, blue– hemoglobin ≤ 10 g/dL, p  = 0.01. ( c ) Activated partial thromboplastin time, red– APTT > 30 s, blue– APTT ≤ 30 s, p   40 g/dL, blue– albumin ≤ 40 g/dL; p   40 g/dL, blue– albumin ≤ 40 g/dL, p  < 0.01 Table 1 Demographics of the cohorts Number of patients 277 Age (average) 56.7 (10–90) Overall survival (months, 95% C.I.) 69.3 (55.0–83.5) Histotype  High grade serous carcinoma 42 (15.2%)  Low grade serous carcinoma 10 (3.6%)  Clear cell carcinoma 24 (8.7%)  Mucinous 10 (3.6%)  Endometrioid 13 (4.7%)  Malignant germ cell tumor 2 (0.7%)  Others* 176 (63.5%) * Includes those diagnosed as carcinoma, not specified, poorly differentiated carcinoma and radiological lesions without histological confirmation Demographics of the cohorts * Includes those diagnosed as carcinoma, not specified, poorly differentiated carcinoma and radiological lesions without histological confirmation On univariate analysis, older age ( p  < 0.01), high-grade histology ( p  = 0.03), lower hemoglobin level ( p  < 0.01), lower serum albumin ( p  < 0.01) and total protein ( p  < 0.01), shorter activated partial thromboplastin time (APTT) ( p  = 0.03), higher Eastern Cooperative Oncology Group performance status scale score (ECOG score) ( p  < 0.01) and prior chemotherapy treatment ( p  < 0.01) were associated with mortality. Administration of chemotherapy after the diagnosis of malignant ascites ( p  < 0.01) were negatively correlated with mortality (Table  2 ). Kaplan Meier analysis demonstrated significant cut-offs for age (≤ 50 years, p   10 g/dL, p  = 0.01), APTT (≤ 30 s, p   40 g/dL, p   70 g/L, p  < 0.01) (Fig.  1 ) and associations with histologic grade and other clinical parameters similar to univariate analysis (Supplementary Table 1). APTT (HR: 0.93 (0.88–0.98), p  = 0.01), total serum protein (HR: 0.89 (0.82–0.96), p  < 0.01), ECOG score ≥ 3 (HR: 10.97 (2.56–47.07), p  < 0.01), and prior chemotherapy treatment before diagnosis of malignant ascites (HR: 6.62 (2.02–21.69), p  < 0.01) were independent predictors on multivariable analysis (Table  2 ). Table 2 Correlation of clinicopathologic parameters with overall survival Univariate analysis Adjusted hazards ratio (95% C.I.) p  value Adjusted hazards ratio (95% C.I.) p  value Age 1.02 (1.01, 1.03) < 0.01 Clotting profile  Neutrophil to lymphocyte ratio 1.01 (1,1.03) 0.17 Histotype  Prothrombin time 1.01 (0.96,1.06) 0.68  High grade serous carcinoma 1.03 (0.68, 1.54) 0.90  Activated partial thromboplastin time 0.98 (0.96,1) 0.03  Low grade serous carcinoma 0 (0, ---) 0.99  International normalized ratio 1.03 (1,1.05) 0.08  Clear cell carcinoma 0.80 (0.45, 1.42) 0.45  Mucinous 1.42 (0.69,2.91) 0.34 Tumor markers  Endometrioid 0.58 (0.27,1.26) 0.17  Carcinoembryonic antigen 1.03 (0.96,1.09) 0.43  High grade 1.73 (1.05,2.85) 0.03  Cancer antigen 125 1 (1,1) 0.3 Blood count ECOG performance status scale  Hemoglobin 0.86 (0.78,0.94) < 0.01  ECOG ≥ 3 3.98 (2.38,6.66) < 0.01  White blood cell count 0.99 (0.96,1.01) 0.31  Platelet count 1 (1,1) 0.19 Treatment received prior to ascites  Neutrophil 0.98 (0.94,1.03) 0.46  (Adjuvant) chemotherapy* 3.04 (2.22,4.16) < 0.01  Lymphocyte 0.99 (0.96,1.02) 0.38  Surgery 0.9 (0.67,1.2) 0.46 Liver/renal function Treatment received after diagnosis  Creatinine 1.01 (1,1.01) 0.08  Chemotherapy* 0.41 (0.3,0.57) < 0.01  Serum albumin 0.95 (0.93,0.97) < 0.01  Radiotherapy 0.7 (0.26,1.89) 0.48  Alkaline phosphatase 1 (1,1) 0.17  Surgery 0.36 (0.24,0.54) < 0.01  Alanine transaminase 1 (0.99,1.01) 0.7  Total serum protein 0.95 (0.94,0.97) < 0.01 Multivariable regression analysis Adjusted hazards ratio (95% C.I.) p  value  Age 1.04 (0.99,1.09) 0.09  High histologic grade 1.58 (0.5,4.97) 0.44  Haemoglobin 0.91 (0.72,1.16) 0.43  White blood cell count 1 (0.87,1.16) 0.98  Activated partial thromboplastin time 0.93 (0.88,0.98) 0.01  Albumin 0.99 (0.9,1.08) < 0.01  Total serum protein 0.89 (0.82,0.96) < 0.01  ECOG ≥ 3 10.97 (2.56,47.07)) < 0.01  Chemotherapy before diagnosis* 6.62 (2.02,21.69) < 0.01  Surgery after diagnosis 0.41 (0.12,1.43) 0.16 * Includes targeted therapy Correlation of clinicopathologic parameters with overall survival * Includes targeted therapy There were 27 (9.7%) long survivors identified in the cohort, with a mean survival of 188 months (15.6 years) (Table  3 ). Ten (37.0%) of the patients were deceased at the end of the follow-up period. The cause of death was documented for 8 patients and only one was attributed to disease, who was a patient with poorly differentiated carcinoma surviving for 139 months with disease (Table  3 , Supplementary Table 2). The remaining 17 (63.0%) patients were all alive without disease. The patient with the longest survival was 311 months, with a diagnosis of low-grade serous carcinoma and dying of acute coronary syndrome and without evidence of disease. Table 3 Clinicopathological features of long survivors Number of patients 27 Age at diagnosis (average, range) 50.4 (19–74) Overall survival in months (average, range) 188 (120–311) Deceased during follow-up period 10  Attributable to cancer 1  Not attributable to cancer 7  Cause not documented 2 Histotype  High grade serous carcinoma 3  Low grade serous carcinoma 5  Clear cell carcinoma 3  Mucinous carcinoma 2  Others 14 Clinicopathological features of long survivors The most common type of malignancy with long survival was low-grade serous carcinoma ( n  = 5/27, 18.5%, p  < 0.01), but high-grade carcinomas including high-grade serous ( n  = 3) and clear cell carcinoma ( n  = 3) were also compatible with long survival. Long survival was correlated with low-grade serous ( p  < 0.01) and endometrioid ( p  = 0.04) histology, younger age ( p  = 0.01), higher serum albumin ( p  = 0.01) and total protein ( p  = 0.01), lower platelet count ( p  = 0.02) and surgical treatment received after ascites ( p   0.05) (Table  4 ). Table 4 Clinicopathological parameters associated with long-term survival Long survivors Non-long survivor P -value Long survivors Non-long survivor P -value Histotype Age 50.41 57.39 0.01  High-grade serous carcinoma   Yes 3 39 Blood count   No 21 126 0.22  Hemoglobin 11.43 11.07 0.36  Low-grade serous carcinoma  White blood cell count 8.87 9.93 0.58   Yes 5 5  Platelet count 321.17 378.84 0.02   No 19 160 < 0.01  Neutrophil 8.06 7.45 0.69  Clear cell carcinoma  Lymphocyte 1.23 1.42 0.92   Yes 3 21  Neutrophil to lymphocyte ratio 7.82 10.34 0.51   No 21 144 0.98  Mucinous carcinoma Clotting profile   Yes 2 8  Prothrombin time 11.37 12.08 0.70   No 22 157 0.48  Activated partial thromboplastin time 35.90 31.35 0.38  Endometrioid carcinoma  International normalized ratio 1.08 2.24 0.72   Yes 4 9   No 20 156 0.04 Liver/renal function  High grade  Creatinine 66.00 64.59 0.80   Yes 11 72  Serum albumin 37.67 31.89 0.01   No 9 30 0.17  Alkaline phosphatase 82.08 78.57 0.83  Alanine transaminase 23.83 21.99 0.82 ECOG performance status scale  Total serum protein 76.58 69.53 0.01  ECOG ≥ 3 1 17  ECOG < 3 14 203 0.88 Tumor markers  Carcinoembryonic antigen 2.833 36.380 0.84 Treatment received prior to ascites  Cancer antigen 125 162.33 2232.22 0.39  Chemotherapy*   Yes 2 64   No 16 173 0.14 Surgery   Yes 11 119   No 7 118 0.37 Treatment received after diagnosis  Chemotherapy*   Yes 16 176   No 2 61 0.17 Radiotherapy   Yes 1 4   No 17 233 0.25  Surgery   Yes 10 45   No 18 192 < 0.01 * Includes targeted therapy Clinicopathological parameters associated with long-term survival * Includes targeted therapy

Nearly 10% of patients achieved 10-year long survival with malignant ascites in ovarian cancer, and is correlated with age, histology, treatment and serological parameters.

The mean overall survival after onset of malignant ascites due to ovarian cancer was over 6.5 years (69.3 months), with nearly 10% of patients achieving long survival (10 year). Significant factors correlating long survival were young age, low-grade serous or endometrioid histology, surgery after onset of malignant ascites, higher serum albumin and total protein, and lower platelet count. Whether surgery was performed prior to the onset of malignant ascites was not predictive of survival, suggesting similar outcomes between patients with malignant ascites as first presentation and recurrence. Findings of the study confirm that age, histology, hematological and biochemical markers are useful in predicting outcomes in patients with ovarian cancer developing malignant ascites, and that long survival is compatible with advanced ovarian cancers.

Malignant ascites and abdominal/peritoneal involvement are common in ovarian cancers [ 1 ] and the majority of patients are with advanced (stage III/IV) disease at the time of diagnosis [ 10 ]. Symptoms associated with malignant ascites and abdominal/peritoneal deposits, such as distension, bloating, gastrointestinal disturbance, urinary urgency and pain are the main presenting complaints for ovarian cancer [ 2 , 11 ]. Similarly, development of these symptoms in patients with treated ovarian cancer heralds disease recurrence [ 4 ]. Due to the technical difficulty in obtaining core biopsies from the ovary, the risk of seeding malignant cells and rupturing cystic lesions [ 12 ], it is not uncommon for the primary and only tissue diagnosis of ovarian cancer to be peritoneal fluid cytology [ 13 ], and diagnosis of malignant ascites is dependent on detection of malignant cells in peritoneal fluid [ 14 ]. The prognosis of malignant ascites is very guarded, as it often equates to non-resectable and widespread abdominal and/or peritoneal tumor deposits [ 15 ]. Even with advanced disease, terminal patients with malignant ascites have shorter survival than those without [ 16 ]. However, malignant ascites developed as a result of ovarian carcinomas,, compared to types of malignancies, were reported to have better outcomes [ 11 ], supported by a reported rate of 18% for long-term (10 year) disease free survival rate of stage III ovarian cancer by Pitiyarachchi et al. [ 5 ]. Age, low histologic grade and poor ECOG performance status were associated with shorter survival, in line with established prognostic factors for ovarian cancer irrespective of stage [ 17 ]. Whether prior surgical treatment was performed did not affect overall survival (Table  2 ), suggesting that malignant ascites as an initial presentation does not differ from recurrent disease in terms of prognosis. However, a history of adjuvant chemotherapy was strongly associated with shorter survival, which can be attributed to an increased risk of treatment resistance in recurrence compared to treatment-naïve cases [ 18 , 19 ]. Patients amenable to chemotherapy, the mainstay being platinum based from the collection period of the cohort (Supplementary Table 2), and debulking surgery after the diagnosis of malignant ascites showed lower mortality, supporting the role of cytoreduction in advanced ovarian cancers [ 20 ]. Hematologic and biochemical parameters (lower APTT, higher hemoglobin, serum albumin and total protein) demonstrated correlation with overall survival. Associations between these factors and survival are reported not only in ovarian cancers [ 21 ], but also other malignancies such as breast, gastrointestinal, hematological and lung cancers [ 22 – 26 ]. It is postulated that serum proteins (including albumin and clotting factors) and hemoglobin levels are reflective the general nutrition status [ 27 ]. Malnutrition impedes wound healing, increases the risks of complications such as bleeding and infection, and negatively affects the response in systemic treatment [ 25 , 28 , 29 ]. In addition, abnormalities in blood counts, clotting profile and liver/renal functions may stem from unrecognized or subclinical medical conditions [ 30 ]. APTT, serum albumin and total protein remained strongly associated with overall survival in multivariable regression analysis ( p  ≤ 0.01), along with ECOG performance status and a history of adjuvant chemotherapy. The mean overall survival of the cohort was more than 6.5 years (69.3 months), and nearly 10% of patients in the cohort achieved long-term 10-year survival. These figures are superior to that of gastrointestinal and pancreatic carcinomas [ 16 , 31 ], the other leading causes of malignant ascites [ 14 ]. There was only one patient with active disease in the group of long survivors, which was a case of poorly differentiated carcinoma. The patient presented with advanced disease and received debulking surgery, followed by multiple cycles of chemotherapy without achieving sustained disease remission. It can be concluded that the risk of recurrence after 10 years for patients with malignant ascites is low. Examination of the survival curve also showed mortality plateauing at 9 to 10 years after onset of malignant ascites (Fig.  2 ). Although late recurrences have been reported in ovarian cancer [ 32 ], even in status post-hysterectomy, the possibility of a second primary disease arising from endometriosis or residual Mullerian structures cannot be excluded [ 33 ]. Notably, concurrent abdominal, peritoneal, nodal, and visceral organ metastasis were present in 40% ( n  = 11/27) long survivors, and complete remission was seen in all but the patient with poorly differentiated carcinoma who died of disease. Fig. 2 Survival curve of the cohort (months) Survival curve of the cohort (months) Young age, endometrioid and low-grade serous histology favored long survival, but aggressive histotypes (clear cell, high-grade serous and poorly differentiated carcinomas) are still occasionally compatible with long survival. However, in contrast with overall survival, ECOG score and treatment received before and after onset of malignant ascites were not different in the long survivor group, except for debulking surgery after development of malignant ascites. As ovarian cancers affect a younger demographic, a considerable proportion of patients have low ECOG scores and are fit for treatment [ 34 ], and differences may not be observable for comparison in the relatively small subgroup of long survivors without a greater number of cases. Higher serum albumin and total protein, and lower platelet count were seen in long survivors. It is unknown whether optimization of serum protein levels benefits survival or that it is merely a prognostic indicator. Current indications for albumin infusion in cancer patients are largely limited to the correction of hypoalbuminemia [ 35 ], but not supported for the improvement of general nutrition status [ 36 ]. Association between elevated platelet count and poor survival in cancers, including ovarian cancers, have been reported [ 37 , 38 ], and is postulated to be an effect of pro-inflammatory cytokines stimulating platelet production [ 37 ]. However, the associations between overall survival and platelet, neutrophil and lymphocyte counts were not demonstrable in the current cohort. There is modest evidence that elevated neutrophil and/or platelet counts are prognostically significant, not only in ovarian cancers [ 39 , 40 ] but also other solid human cancers [ 41 ]. The difference may be that only metastatic ovarian cancers were included in the cohort. Neutrophil and platelet are an indicator systemic inflammatory response and can be an indicator of disease dissemination [ 41 ]. The study was limited by a retrospective design. Periodical imaging was not performed, as such, disease remission cannot be confidently determined in cases without extended clinical follow-up, thus disease-free survival and recurrence was only analysed in long survivors. The prognostic value of biochemical, hematological, serological markers may not be limited to only at the time of diagnosis, but also in terms of serial changes in regard to treatment and recurrence. Interval monitoring particularly pertains to tumor markers, as carcinoembryonic antigen, and cancer antigen 125 levels at the time of diagnosis failed to demonstrate correlation with mortality nor long survival. Tumor markers are useful in initial diagnosis, monitoring of disease progression/recurrence and treatment response [ 42 , 43 ], but the true prognostic value of tumor markers lies the decrease in serum level and/or a low nadir level indicating a reduction in tumor burden or even remission [ 44 , 45 ].

Consecutive peritoneal fluid cytology reports from the year 1995 to 2022 in the department pathology archives were retrieved. The cytologic diagnosis, and in cases of indeterminate cases, the microscopy and immunocytochemical results, were reviewed. Cases were reclassified into five tiered diagnostic categories [ 9 ] of insufficient/inadequate (C1), benign (C2), atypia (C3), suspicious for malignancy (C4) and malignant (C5), and patients with at least one a malignant or suspicious for malignancy cytologic diagnosis were included. Case notes of the corresponding patients were reviewed, and only those with ovarian malignancy, as indicated by the hospital diagnosis code and relevant investigations, were recruited into the study. In addition to the cytologic diagnosis, radiological and/or histological (core biopsy, excisional biopsy, or resection) confirmation of ovarian malignancy was required, and cases without supporting evidence were excluded. Demographical, clinical, biochemical, hematological, serological, and pathological data were reviewed from the electronic patient records under the public hospital, with standardized reporting formats. Biochemical, hematological and serological test results from external laboratories were not included. The date of diagnosis of malignant ascites was defined by that of the first peritoneal fluid cytology specimen with a diagnosis of malignant or suspicious for malignancy. For blood tests, only those collected within one week of diagnosis of malignant ascites were included, with the exception of tumor markers in which tests up to within one month were accepted. Survival data of the patients were reviewed, and long survivors were defined as patients with an overall survival of more than 10 year from the diagnosis of malignant ascites. Statistical analysis was performed using SPSS (version 26.0) and R. The t-test and Chi-square test were used for comparing clinicopathological parameters between long survivors and non-long survivors as continuous and categorical variables. Survival analysis was performed using the Kaplan Meier curve and hazard ratio. Multivariate analysis was performed using the backward Wald method. This study was approved by The Joint Chinese University of Hong Kong– New Territories East Cluster Clinical Research Ethics Committee (reference number: 2020.289).

Malignant ascites is a significant complication of ovarian cancer [ 1 ]. Patients not uncommonly present with associated symptoms at initial diagnosis [ 2 , 3 ], or in the event of disease recurrence [ 4 ]. The prognosis of malignant ascites is poor but is reported to be more favorable for ovarian cancers than other malignancies such as gastrointestinal and pancreatic carcinomas [ 5 ]. However, literature on the specific outcomes of ovarian cancers with malignant ascites is lacking. Either no differentiation was made between different underlying primary cancers, or that tissue diagnosis was not obtained for exclusion other causes of ascites in advanced cancers, such as liver metastasis or organ failure [ 5 – 7 ]. In advanced ovarian cancers, the disease course remains unpredictable up to 5 years follow-up, with the decline of survival stabilizing only at 5 to 10 years after diagnosis [ 8 ]. In this study, a retrospective cohort of ovarian cancer patients with cytologically confirmed malignant ascites were reviewed over an extended period, and clinicopathological data, including a broad panel of biochemical, hematological, and serological markers were retrieved, for detailing the long-term (10 year) survival rate, and corresponding predictive factors. As malignant ascites in ovarian cancer appears unique in that may compatible with long-term survival and remission, this study aims to elucidate, the outcome of malignant ascites in ovarian cancers, and to identify factors that are predictive of long-term survival, and thus prognostication and treatment guidance.

Below is the link to the electronic supplementary material. Supplementary Material 1 Supplementary Material 1