A retrospective study on the safety of small and large volume paracentesis and risk factors for adverse events in patients with gastrointestinal cancer with malignant ascites
Highlight box
Key findings
• This retrospective study demonstrated that large-volume drainage (LVD) for malignant ascites in gastrointestinal cancer patients was associated with a higher incidence of adverse events, particularly post-paracentesis circulatory dysfunction, compared to small-volume drainage (SVD). Additionally, a lower baseline serum albumin level was identified as a significant risk factor for the development of paracentesis-related adverse events.
What is known and what is new?
• While volume management in paracentesis is well-established for liver cirrhosis, clinical evidence specifically targeting gastrointestinal cancer patients with malignant ascites has been limited.
• Our findings provide clear evidence that LVD increases the risk of complications in this oncological population and underscore the clinical impact of pre-procedural hypoalbuminemia on patient safety.
What is the implication, and what should change now?
• When performing paracentesis for patients with advanced gastrointestinal cancer, clinicians should carefully consider the drainage volume and closely monitor patients with low serum albumin levels. Implementing stricter volume controls (e.g., prioritizing SVD or sequential drainage) and considering appropriate albumin supplementation for high-risk patients may mitigate the risk of severe adverse events and improve the safety of palliative ascites management.
Introduction
Malignant ascites (MA), defined as the accumulation of fluid in the peritoneal cavity in patients with malignant tumors, frequently occurs in the terminal stages of cancer. Ovarian cancer accounts for the majority of MA cases; however, gastric, uterine, breast, colorectal, and pancreatic cancers collectively contribute to approximately 80% of cases (1-3). Most cases of MA arise from peritoneal carcinomatosis, although other etiologies include portal hypertension secondary to extensive liver metastases, cirrhosis associated with hepatocellular carcinoma, and chylous ascites due to malignant lymphoma (4). The presence of MA has been identified as a poor prognostic factor, with reported median survival times ranging from 1 to 4 months following diagnosis (5). Even in patients admitted to palliative care units, patients with MA had a shorter survival than those without MA (6). Approximately 60% of patients with MA experience various distressing symptoms, such as abdominal swelling, abdominal pain, nausea, anorexia, vomiting, and dyspnea (2). For patients with adequate performance status, anti-neoplastic therapy may provide symptomatic relief by controlling MA. However, in patients who are either ineligible for such treatment or have malignancies resistant to anti-neoplastic interventions, effective symptom control remains a considerable challenge. Evidence guiding palliative management of MA-related symptoms is limited. Current treatment options typically include diuretics, temporary abdominal paracentesis, indwelling peritoneal catheters, cell-free and concentrated ascites reinfusion therapy, and peritoneovenous shunting.
Abdominal paracentesis is one of the most common interventions to palliate the symptoms of MA. It can offer rapid relief for most patients. Prospective studies have indicated that approximately 80% of MA patients who undergo paracentesis report improvements in abdominal swelling, abdominal pain, and nausea (7,8). Despite these benefits, fluid often reaccumulates, necessitating repeated procedures for continued symptomatic relief.
The optimal drainage volume of paracentesis in MA has not been established. While European guidelines allow for drainage of up to 5 L (9), small-volume drainage (SVD) of less than 3 L for patients with MA is common in Japan. It is because Japanese experts suggest SVD in frail patients and those with a limited prognosis, given the potential risk of increased adverse events (AEs) with large-volume drainage (LVD) (10). The rate of AEs associated with paracentesis in patients with MA is relatively low at 5.5–20%, reported AEs include hypotension, infection, renal dysfunction, and electrolyte abnormalities (11). However, few studies have directly compared the incidence of AEs between LVD and SVD and their risk factors. Identifying risk factors of abdominal paracentesis may help guide clinical decisions on the drainage volume in patients with MA.
Accordingly, the present study aimed to compare the safety of SVD and LVD in patients with MA and to identify risk factors of AEs. We present this article in accordance with the STROBE reporting checklist (available at https://apm.amegroups.com/article/view/10.21037/apm-2026-1-0002/rc).
Methods
Patients
This single institutional retrospective study included patients with gastrointestinal cancer who received abdominal paracentesis between 2014 and 2023. The main selection criteria were as follows: (I) age ≥20 years; (II) diagnosed with unresectable, recurrent, or metastatic gastrointestinal cancer; (III) received abdominal paracentesis for symptom relief for the first time; and (IV) systolic blood pressure ≥90 mmHg before abdominal paracentesis. Patients with hepatocellular carcinoma or liver cirrhosis, or who received cell-free and concentrated ascites reinfusion therapy, were excluded. The study was approved by the Institutional Review Board of Osaka Medical and Pharmaceutical University (No. 2023-154) and conducted in accordance with the Declaration of Helsinki and its subsequent amendments. Informed consent was not required because of the retrospective observational nature of the study. However, patient consent was obtained using the opt-out method.
Measurements
The following data were collected when patients received abdominal paracentesis: age, sex, body mass index (weight in kilograms divided by height in meters squared), Eastern Cooperative Oncology Group performance status (ECOG PS), primary cancer site, opioid use, diuretic agents use, during chemotherapy or best supportive care, drainage volume, and concurrent administration of albumin. All patients had their blood pressure measured before and immediately after abdominal paracentesis. Data on serum albumin, sodium, and creatinine were collected within 14 days before and 14 days after abdominal paracentesis.
Statistical analysis
The primary outcome was the proportion of paracentesis-related AEs, which included hypotension defined as systolic blood pressure <90 mmHg after abdominal paracentesis, ascites leakage, infection, and renal dysfunction defined as elevation of creatinine ≥0.3 mg/dL (12) or hyponatremia defined as a decrease of sodium ≥5 mEq/L (13). Patients were divided between SVD group, defined as less than 3 L, and LVD group, defined as 3 L or more. Patients characteristics were compared using chi-squared test, or Fisher’s exact test, as appropriate. Overall survival (OS) was compared between two groups using the Kaplan-Meier method and the log-rank test. OS was defined as the time from first paracentesis to death. The interval to subsequent paracentesis was defined as the time from the first paracentesis to next paracentesis in patients who received the next paracentesis. Univariate and multivariate analysis using logistic regression models were used to determine risk factors for paracentesis-related AEs. Covariates included in the multivariate analysis were age (median), sex, ECOG PS (0–2 vs. 3–4), body mass index (median), serum albumin (median), serum creatinine (median), serum sodium (<135 vs. ≥135 mEq/L) before paracentesis and paracentesis volume (SVD vs. LVD). A two-tailed P value <0.05 was considered significant. All analyses were performed with EZR ver1.50.
Results
Patient characteristics
A total of 220 patients with gastrointestinal cancer received paracentesis in our hospital between 2014 and 2023. Finally, we analyzed 89 patients consisting of 58 patients with SVD (65.2%) and 31 patients with LVD (34.8%). The most common reasons for exclusion were hepatocellular carcinoma (n=54), followed by not unresectable, recurrent or metastatic status (n=16; Figure S1). Baseline characteristics of all eligible patients and patients in SVD and LVD groups are shown in Table 1. The median age was 66 years and there were more male (62.9%) than female patients (37.1%). The ECOG PS was 0 in 3 patients (3.4%), 1 in 29 (32.6%), 2 in 29 (32.6%), 3 in 24 (27.0%), and 4 in 4 (4.5%).
Table 1
| Variables | All patients (n=89) | SVD (n=58) | LVD (n=31) | P |
|---|---|---|---|---|
| Age, years | 66 [32–89] | 67 [42–89] | 66 [32–81] | 0.21 |
| Sex, female | 33 (37.1) | 21 (36.2) | 12 (38.7) | 0.82 |
| ECOG PS | 0.47 | |||
| 0–2 | 61 (68.5) | 38 (65.5) | 23 (74.2) | |
| 3–4 | 28 (31.5) | 20 (34.5) | 8 (25.8) | |
| Body mass index, kg/m2 | 21.5 [14.8–31.0] | 21.2 [14.8–30.0] | 21.9 [16.1–31.0] | 0.28 |
| Primary cancer site | 0.90 | |||
| Esophagus | 1 (1.1) | 1 (1.7) | 0 (0.0) | |
| Stomach | 46 (51.7) | 28 (48.3) | 18 (58.1) | |
| Colorectum | 13 (14.6) | 9 (15.5) | 4 (12.9) | |
| Pancreas | 28 (31.5) | 19 (32.8) | 9 (29.0) | |
| Ileum | 1 (1.1) | 1 (1.7) | 0 (0.0) | |
| Albumin use | 3 (3.4) | 1 (1.7) | 2 (6.5) | 0.27 |
| Opioid use | 29 (32.6) | 18 (31.0) | 11 (35.5) | 0.81 |
| Diuretics use | 42 (47.2) | 27 (46.6) | 15 (48.4) | >0.99 |
| During chemotherapy | 40 (44.9) | 24 (41.4) | 16 (51.6) | 0.38 |
| Liver metastasis | 41 (46.1) | 28 (48.3) | 13 (41.9) | 0.66 |
| Blood pressure, mmHg | 121 [91–173] | 116 [91–173] | 122 [94–153] | 0.86 |
| Serum albumin, g/dL | 2.5 [1.4–4.2] | 2.4 [1.4–4.2] | 2.5 [1.8–3.4] | 0.47 |
| Serum creatinine, mg/dL | 0.82 [0.36–3.52] | 0.81 [0.36–3.52] | 0.85 [0.39–1.71] | 0.76 |
| Serum sodium, mEq/L | 137 [123–145] | 136.5 [124–144] | 138 [123–145] | 0.26 |
Data are presented as n (%) or median [range]. ECOG PS, Eastern Cooperative Oncology Group performance status; LVD, large-volume drainage; SVD, small-volume drainage.
The most common primary cancer site was stomach (n=46, 51.7%), followed by pancreas (n=28, 31.5%), colorectum (n=13, 14.6%), esophagus (n=1, 1.1%) and small bowel (n=1, 1.1%). Baseline characteristics between the SVD group and the LVD group did not differ significantly.
Safety of abdominal paracentesis
Median drainage volume in the SVD group and the LVD group was 2,000 mL (range, 600–2,900 mL) and 3,000 mL (range, 3,000–4,000 mL), respectively (P<0.01).
All AEs were significantly higher in the LVD group than the SVD group (20.7% vs. 41.9%, P=0.04). Regarding each AE, hypotension was significantly higher in the LVD group than the SVD group (1.7% vs. 12.9%, P=0.04). There was no significant difference between two groups in renal dysfunction (12.1% vs. 16.1%, P=0.74), hyponatremia (12.1% vs. 9.7%, P>0.99), infection (0% vs. 6.5%, P=0.11) and leakage of ascites (1.7% vs. 3.2%, P>0.99) (Table 2). There was no other severe AE in both groups.
Table 2
| Variables | All patients (n=89), n (%) | SVD (n=58), n (%) | LVD (n=31), n (%) | P |
|---|---|---|---|---|
| Any adverse events | 25 (28.1) | 12 (20.7) | 13 (41.9) | 0.04 |
| Serum creatinine ≥0.3 mg/dL increase | 12 (13.5) | 7 (12.1) | 5 (16.1) | 0.74 |
| Serum sodium ≥5 mg/dL decrease | 10 (11.2) | 7 (12.1) | 3 (9.7) | >0.99 |
| Hypotension | 5 (5.6) | 1 (1.7) | 4 (12.9) | 0.04 |
| Infection | 2 (2.2) | 0 (0.0) | 2 (6.5) | 0.11 |
| Leakage of ascites | 2 (2.2) | 1 (1.7) | 1 (3.2) | >0.99 |
LVD, large-volume drainage; SVD, small-volume drainage.
The post-paracentesis systolic blood pressure in patients who developed hypotension did not significantly differ between the SVD group (median: 89 mmHg; range, 89–89 mmHg) and the LVD group (median: 88 mmHg; range, 77–89 mmHg; P=0.26). Similarly, changes in serum creatinine and serum sodium levels among patients with renal dysfunction and hyponatremia, respectively, were comparable between the SVD and LVD groups: 1.12 mg/dL (range, 0.35–3.01 mg/dL) vs. 0.78 mg/dL (range, 0.37–1.07 mg/dL) (P=0.29) for creatinine, and 6 mEq/L (range, 5–11 mEq/L) vs. 7 mEq/L (range, 5–9 mEq/L) (P=0.55) for sodium.
Risk factors of AEs
In univariate analysis, only LVD was significantly associated with AEs [odds ratio (OR) 2.85; 95% confidence interval (CI): 1.11–7.29, P=0.03]. Other factors, including age ≥66 years, female sex, ECOG PS 3–4, body mass index ≥21.5, serum albumin <2.5 mg/dL, serum creatinine ≥0.82 mg/dL, and serum sodium ≥135 mEq/L, were not significantly correlated with AEs (Table 3). In multivariate analysis, serum albumin <2.5 mg/dL (OR 3.20; 95% CI: 1.06–9.64, P=0.03) and drainage volume ≥3,000 mL (OR 3.45; 95% CI: 1.20–9.93, P=0.02) were the independent risk factors for abdominal paracentesis (Table 4).
Table 3
| Covariate | Odds ratio (95% CI) | P value |
|---|---|---|
| Age (≥66 vs. <66 years) | 1.17 (0.47–2.91) | 0.72 |
| Sex (female vs. male) | 1.00 (0.39–2.54) | >0.99 |
| ECOG PS (3–4 vs. 0–2) | 0.52 (0.19–1.41) | 0.20 |
| Body mass index (≥21.5 vs. <21.5 kg/m2) | 1.15 (0.47–2.84) | 0.76 |
| Serum albumin (<2.5 vs. ≥2.5 mg/dL) | 0.45 (0.18–1.15) | 0.10 |
| Serum creatinine (≥0.82 vs. <0.82 mg/dL) | 2.06 (0.82–5.22) | 0.13 |
| Serum sodium (≥135 vs. <135 mEq/L) | 1.93 (0.71–5.24) | 0.20 |
| Drainage volume (LVD vs. SVD) | 2.85 (1.11–7.29) | 0.03 |
CI, confidence interval; ECOG PS, Eastern Cooperative Oncology Group performance status; LVD, large-volume drainage; SVD, small-volume drainage.
Table 4
| Covariate | Odds ratio (95% CI) | P value |
|---|---|---|
| Age (≥66 vs. <66 years) | 0.82 (0.28–2.41) | 0.72 |
| Sex (female vs. male) | 1.19 (0.38–3.75) | 0.77 |
| ECOG PS (3–4 vs. 0–2) | 1.11 (0.35–3.58) | 0.86 |
| Body mass index (≥21.5 vs. <21.5 kg/m2) | 1.51 (0.53–4.30) | 0.44 |
| Serum albumin (<2.5 vs. ≥2.5 mg/dL) | 3.20 (1.06–9.64) | 0.03 |
| Serum creatinine (≥0.82 vs. <0.82 mg/dL) | 2.74 (0.88–8.58) | 0.08 |
| Serum sodium (≥135 vs. <135 mEq/L) | 2.07 (0.69–6.21) | 0.20 |
| Drainage volume (LVD vs. SVD) | 3.45 (1.20–9.93) | 0.02 |
CI, confidence interval; ECOG PS, Eastern Cooperative Oncology Group performance status; LVD, large-volume drainage; SVD, small-volume drainage.
Survival and interval to subsequent paracentesis
OS was not different between the two groups [median OS, 33 days (95% CI: 29–51) vs. 29 days (95% CI: 22–74); P=0.86] (Figure 1).
Patients who received the next paracentesis were 31 (53.4%) in the SVD group and 18 (58.1%) in the LVD group. Most of the reasons for patients who did not receive the next paracentesis were lost follow-up due to transfer to palliative care units. Among patients who received the next paracentesis, the median interval to subsequent paracentesis did not differ between the two groups (10 vs. 7 days, P=0.21) (Figure S2).
Discussion
In this retrospective study of 89 gastrointestinal cancer patients with MA, we found that LVD was associated with a higher incidence of AEs, although the absolute number of events was small, particularly hypotension, compared to SVD. Importantly, the interval to subsequent paracentesis and OS did not differ significantly between the groups. Multivariate analysis identified LVD and low serum albumin as independent risk factors for AEs, highlighting the importance of individualized fluid management in this vulnerable population.
Previous studies have reported the safety and efficacy of paracentesis in MA patients, but direct comparisons of drainage volume and their associated complications have been limited. McNamara et al. (14) demonstrated that the majority of symptom relief occurred within the first 2 hours of drainage, regardless of the total volume removed. Similarly, the EASED study showed no significant differences in symptom relief or safety between SVD and moderate-volume drainage (7). Our study adds to this literature by incorporating rigorous laboratory-based monitoring and identifying specific risk factors for paracentesis-related AEs. The higher incidence of AEs in our cohort may be attributable to a more sensitive and systematic post-procedural assessment, including changes in vital signs and biochemical markers.
The novelty of our study lies in its comprehensive evaluation of safety outcomes—including laboratory-based AEs—and the identification of risk factors. To our knowledge, this is one of the few studies to demonstrate that LVD increases AE risk without demonstrating clear clinical benefits such as prolonged time to re-accumulation or improved survival. Notably, our data showed that even in patients who required repeated paracentesis, drainage volume did not significantly affect the interval to subsequent procedures, supporting the clinical sufficiency of SVD.
These findings are particularly relevant in the palliative care setting, where minimizing procedural risks is a key therapeutic goal. Patients with hypoalbuminemia are at higher risk for circulatory collapse due to reduced oncotic pressure, and our data suggest that conservative fluid removal should be considered in this subgroup. While LVD may be appropriate in select cases with severe symptom burden, the overall balance of safety and efficacy appears to favor a small-volume approach.
The strengths of this study include detailed data collection on pre- and post-procedural laboratory and clinical parameters, as well as the identification of clinically actionable risk factors. However, several limitations should be acknowledged. First, the retrospective, single-center design may limit generalizability. Second, the median difference in drainage volume between the SVD and LVD groups was approximately 1,000 mL, which may underestimate the true impact of larger volume differences. Third, the interval to subsequent paracentesis may have been influenced by clinical decisions and institutional practices, rather than solely by patient need or symptom recurrence. Fourth, we did not assess symptom relief using validated patient-reported outcomes, which would provide additional insight into clinical benefit. Finally, our cohort did not include patients with ovarian cancer—a major contributor to MA—which limits the applicability of our findings to this population.
Conclusions
In conclusion, our findings suggest that small-volume paracentesis is a safe and effective option for managing MA in patients with gastrointestinal cancer. LVD was associated with a higher risk of AEs, especially in patients with hypoalbuminemia, without evidence of improved clinical efficacy. These results support a conservative and individualized approach to paracentesis volume, particularly in frail or end-of-life patients. Future prospective studies incorporating symptom scores and quality-of-life measures are warranted to validate these findings and optimize paracentesis strategies.
Acknowledgments
None.
Footnote
Reporting Checklist: The authors have completed the STROBE reporting checklist. Available at https://apm.amegroups.com/article/view/10.21037/apm-2026-1-0002/rc
Data Sharing Statement: Available at https://apm.amegroups.com/article/view/10.21037/apm-2026-1-0002/dss
Peer Review File: Available at https://apm.amegroups.com/article/view/10.21037/apm-2026-1-0002/prf
Funding: None.
Conflicts of Interest: All authors have completed the ICMJE uniform disclosure form (available at https://apm.amegroups.com/article/view/10.21037/apm-2026-1-0002/coif). The authors have no conflicts of interest to declare.
Ethical Statement: The authors are accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved. The study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments. This study was approved by the Institutional Review Board of Osaka Medical and Pharmaceutical University (No. 2023-154). The need for informed consent was waived in view of the retrospective observational nature of the research. However, patient consent was secured via the opt-out method.
Open Access Statement: This is an Open Access article distributed in accordance with the Creative Commons Attribution-NonCommercial-NoDerivs 4.0 International License (CC BY-NC-ND 4.0), which permits the non-commercial replication and distribution of the article with the strict proviso that no changes or edits are made and the original work is properly cited (including links to both the formal publication through the relevant DOI and the license). See: https://creativecommons.org/licenses/by-nc-nd/4.0/.
References
- Saif MW, Siddiqui IA, Sohail MA. Management of ascites due to gastrointestinal malignancy. Ann Saudi Med 2009;29:369-77. [Crossref] [PubMed]
- Ayantunde AA, Parsons SL. Pattern and prognostic factors in patients with malignant ascites: a retrospective study. Ann Oncol 2007;18:945-9. [Crossref] [PubMed]
- Parsons SL, Watson SA, Steele RJ. Malignant ascites. Br J Surg 1996;83:6-14. [Crossref] [PubMed]
- Runyon BA, Hoefs JC, Morgan TR. Ascitic fluid analysis in malignancy-related ascites. Hepatology 1988;8:1104-9. [Crossref] [PubMed]
- Adam RA, Adam YG. Malignant ascites: past, present, and future. J Am Coll Surg 2004;198:999-1011. [Crossref] [PubMed]
- Kadono T, Ishiki H, Yokomichi N, et al. Malignancy-related ascites in palliative care units: prognostic factor analysis. BMJ Support Palliat Care 2024;13:e1292-9. [Crossref] [PubMed]
- Ito T, Yokomichi N, Ishiki H, et al. Optimal Paracentesis Volume for Terminally Ill Cancer Patients With Ascites. J Pain Symptom Manage 2021;62:968-77. [Crossref] [PubMed]
- Seah DS, Wilcock A, Chang S, et al. Paracentesis for cancer-related ascites in palliative care: An international, prospective cohort study. Palliat Med 2022;36:1408-17. [Crossref] [PubMed]
- Becker G, Galandi D, Blum HE. Malignant ascites: systematic review and guideline for treatment. Eur J Cancer 2006;42:589-97. [Crossref] [PubMed]
- Hisanaga T, Shinjo T, Imai K, et al. Clinical Guidelines for Management of Gastrointestinal Symptoms in Cancer Patients: The Japanese Society of Palliative Medicine Recommendations. J Palliat Med 2019;22:986-97. [Crossref] [PubMed]
- Stukan M. Drainage of malignant ascites: patient selection and perspectives. Cancer Manag Res 2017;9:115-30. [Crossref] [PubMed]
- Pereira M, Rodrigues N, Godinho I, et al. Acute kidney injury in patients with severe sepsis or septic shock: a comparison between the 'Risk, Injury, Failure, Loss of kidney function, End-stage kidney disease' (RIFLE), Acute Kidney Injury Network (AKIN) and Kidney Disease: Improving Global Outcomes (KDIGO) classifications. Clin Kidney J 2017;10:332-40. [Crossref] [PubMed]
- Alessandria C, Elia C, Mezzabotta L, et al. Prevention of paracentesis-induced circulatory dysfunction in cirrhosis: standard vs half albumin doses. A prospective, randomized, unblinded pilot study. Dig Liver Dis 2011;43:881-6.
- McNamara P. Paracentesis--an effective method of symptom control in the palliative care setting? Palliat Med 2000;14:62-4. [Crossref] [PubMed]
