Abstract
Background/Aim: Splenectomy is performed in ovarian cytoreductive surgery when pre-operative imaging and intra-operative examination suggest disease involvement. This study aimed to evaluate the incidence of splenectomy, the diagnostic accuracy of pre-operative imaging and intra-operative assessment, and the associated short- and long-term peri-operative complications.
Patients and Methods: We conducted a single institution retrospective study to assess rates of splenectomy, accuracy of imaging and intra-operative assessment, peri-operative complications and patient compliance with prophylactic antimicrobial therapy and immunizations.
Results: Over a period of 8 years (2014-2022), 469 cytoreductive surgeries for International Federation of Obstetrics and Gynaecology (FIGO) stage III-IV epithelial ovarian cancer were undertaken. A splenectomy was performed in 61 (13%) patients. Complete cytoreduction was achieved in 50 cases (82%) undergoing a splenectomy. On pre-operative imaging assessment, splenic disease was suspected in 36 patients (7.7%) and a further 23 (4.9%) patients had malignancy suspected at intraoperative assessment. In five patients (1%), the spleen was removed due to omental disease inseparable from the spleen. In three (0.6%) cases a splenectomy was performed due to intraoperative trauma. Postoperative pathology confirmed splenic disease in 49 (80.3%) patients. A biochemical leak of no clinical significance was identified in 15/61 (24.6%) splenectomy cases, and one case (1.6%) of post-operative pancreatic fistula (POPF) was identified. Remaining living patients were audited for adherence to recommended post-operative management. Seventeen of the eighteen patients (94%) had received recommended vaccinations between 2-6 weeks post-surgery, and 17/18 (94%) consistently continued ongoing vaccinations. Additionally, 16 (88.9%) adhered to prescribed daily prophylactic antibiotics.
Conclusion: Splenectomy, integral to ovarian cytoreduction, facilitates high rates of complete cytoreduction. Preoperative and intraoperative assessments can predict the existence of malignancy in the spleen, confirmed by pathology. Traumatic splenectomy was rare. Emphasising adherence to a post-splenectomy protocol is crucial to mitigate post-splenectomy complications such as POPF, which can significantly delay the initiation of chemotherapy.
Introduction
Ovarian cancer is the sixth most common malignancy affecting the female population in the United Kingdom, with approximately 7,500 new cases recorded annually (1). It poses diagnostic challenges due to its insidious and nonspecific symptomatology, often resulting in advanced-stage disease detection, and a 5-year survival rate as low as 43%, with prognosis worsening with advancing disease (2).
The standard therapeutic protocol for advanced stages of this disease consists of cytoreductive surgery combined with chemotherapy (3). The primary objective of surgery is to achieve no visible residual disease, as its macroscopic absence is a well-established and significant prognostic factor (4, 5). Achieving complete tumour cytoreduction in cases with advanced stage disease often necessitates extensive surgical interventions, targeting also the upper abdominal region (6). Splenic metastases reportedly occur in approximately 2.3-7.1% of the cases, and resection requires either general or upper gastrointestinal surgeons, or gynaecological oncologists with experience in managing splenic involvement (7).
The primary indication for splenectomy is suspected splenic metastasis based on pre-operative imaging or during intra-operative assessment, while intra-operative trauma is a rare cause. Typically, disease involvement manifests predominantly at the splenic hilum, with parenchymal involvement being less frequent (8). It is noteworthy that not all cases of suspected spleen involvement are confirmed upon final pathology analysis (9). In one retrospective study, there was a positive correlation of 58.2% between clinical suspicion and histology confirmation (10).
The anatomical proximity of the spleen to the tail of the pancreas rarely requires distal pancreatectomy to achieve complete cytoreduction (11). However, this combined procedure is associated with a higher risk of postoperative complications. Of particular concern is the occurrence of postoperative pancreatic fistula (POPF), which presents as one of the most severe complications, leading to increased morbidity, prolonged hospitalisation, and potential delays in initiating adjuvant chemotherapy (12).
In association with rising rates of complete cytoreduction, the incidence of splenectomy is rising, but there remains a paucity of evidence and dedicated studies focusing on the incidence and optimal management of complications post-splenectomy (13-16). Furthermore, there is variation in the use of the updated 2016 International Study Group of Pancreatic Surgery (ISGPS) definition and classification of POPF (17).
Patients who have undergone splenectomy face a heightened susceptibility to severe and life-threatening infections, with a risk approximately 10 times higher than that of the general population (18). Overwhelming post-splenectomy infection (OPSI) carries a high mortality rate. The primary causative agent implicated in OPSI is Streptococcus pneumoniae, although other encapsulated bacteria also pose significant risks (19). Accordingly, individuals identified as at risk should adhere to established immunization protocols and maintain ongoing antibiotic prophylaxis to mitigate infection risk (20).
Given the limited evidence regarding the safety of splenectomy in advanced stage ovarian cancer, there is a need for further studies to establish peri and post operative risks. In the current study, the primary objective was to assess the incidence of splenectomy in cytoreduction surgeries for advanced stage ovarian cancer and to determine the incidence and consequences of POPF (i). Additionally, we aimed to evaluate the performance of pre-operative and intra-operative assessment of splenic disease, along with peri-operative outcomes and complications (ii). Furthermore, we sought to investigate patients’ compliance with prophylactic antimicrobial therapy and recommended immunizations within our institution (iii).
Patients and Methods
We conducted a retrospective analysis at the University College London Hospital, using hospital electronic records to identify patients diagnosed with stage III/IV epithelial ovarian cancer according to the International Federation of Obstetrics and Gynaecology (FIGO) criteria who underwent surgery. Ethical approval for audit purposes was obtained. All consecutive patients who underwent splenectomy with or without distal pancreatectomy during primary or interval cytoreductive surgery for advanced ovarian cancer, between October 2014 and December 2022, were included.
Patients undergoing aborted laparotomies (open/close surgeries) were excluded. Over this time, procedures were carried out by a team comprising seven gynaecological oncologists and eleven gastrointestinal surgeons at a single tertiary gynaecological oncology centre.
Patient demographic information including age at diagnosis and comorbidities were identified. All patients underwent contrast-enhanced computed tomography (CT) scans of the chest, abdomen, and pelvis, and pre-operative estimation of splenic involvement. Pre-operative chemotherapy and tumour marker results were recorded.
The operative notes were assessed for operating surgeon specialty, procedure details, completeness of cytoreduction and intraoperative blood loss. The completeness of cytoreductive surgery was evaluated based on the maximum dimension of the largest residual deposit. Complete cytoreduction was defined as the absence of visible macroscopic disease and optimal cytoreduction as residual disease measuring less than 1 cm. Cases with residual disease exceeding 1 cm were classified as suboptimal cytoreduction. Surgical complexity was assessed using the modified Aletti surgical complexity score (21).
Histological data, disease stage and subsequent clinical outcomes such as length of hospital stay, re-operation rates, short-term complications, follow-up data, and mortality were extracted from medical records. Complications were stratified according to the Clavien-Dindo grading system (22). Time to chemotherapy was defined as the duration in days from postoperative day 1 to the commencement of adjuvant chemotherapy. Long-term complications and patient adherence to prophylactic antimicrobial therapy and recommended immunizations were assessed via telephone consultations.
Postoperative pancreatic fistula: Definition and Classification. Postoperative pancreatic fistula was defined using the updated criteria established by the International Study Group of Pancreatic Surgery (ISGPS) in 2016 (17). It is characterized by a drain fluid amylase level exceeding three times the upper limit of the institution’s normal serum amylase value, accompanied by clinically significant sequelae directly attributable to the postoperative pancreatic fistula. Consequently, the former designation of “grade A” pancreatic fistula is now referred to as a “biochemical leak” due to its lack of clinical significance.
“Grade B” fistulas necessitate alterations in postoperative management, involving either prolonged drain placement exceeding three weeks or repositioning via endoscopic or percutaneous interventions. “Grade C” fistulas denote severe complications requiring reoperation or leading to single or multiple organ failure and/or mortality attributable to the pancreatic fistula.
Statistics. Statistical analyses were performed using MedCalc (version 23.2.1, 2025, MedCalc Software Ltd, Oostende, Belgium) Descriptive statistics – including means, medians, standard deviations, and frequency distributions – were used to summarise patient demographics and clinical outcomes. Incidence rates were reported as proportions with exact 95% confidence intervals, calculated using the binomial distribution method. Group comparisons were conducted using the Chi-square test or Fisher’s exact test, where appropriate. A p-value of <0.05 was considered statistically significant.
Results
Cohort characteristics. Between October 2014 and December 2022, a total of 469 cytoreductive surgeries were performed for FIGO stage III/IV ovarian cancer. Among these cases, splenectomy was undertaken in 61 (13%) cases. The median age of patients undergoing splenectomy was 63.6 [interquartile range (IQR)=55.8-71.5] (Table I). Histologically, the majority of cases exhibited high-grade serous carcinomas (n=51, 83.6%), with the remaining comprising low-grade serous (n=7; 11.5%), clear cell carcinoma (n=2; 3.3%) and endometrioid (n=1; 1.6%). FIGO staging distribution was as follows: III (n=32; 52.5%), IVA (n=4; 6.5%), and IVB (n=25; 41.0%).
Cohort demographics, clinico-pathological and surgical characteristics.
Assessment of disease. During pre-operative work-up, suspicion of splenic involvement was noted in 36 (7.7%) cases based on imaging findings, and in the majority of these patients, 35 (97.2%), splenic disease was visualised intra-operatively and the spleen was removed. In these patients, (n=35), histological analysis confirmed splenic disease in 32 patients, representing a confirmation rate of 91.4%. In one case, the intra-operative assessment did not confirm the pre-operative suspicion on imaging, but splenectomy was performed as a result of trauma.
Additionally, intra-operative assessment raised the suspicion of splenic disease in 23 cases (4.9%). Out of these 23 cases, splenic disease was confirmed on final histology in 73.9% cases (n=17). In a further two cases, (0.4%), splenectomy was performed due to trauma. During intra-operative assessment, suspicious areas for spleen disease were the capsule (n=33; 58.9%), hilum (n=19; 33.9%), or both capsule and hilum (n=4; 7.1%).
Overall, histological analysis confirmed splenic involvement in 49 (80.3%) patients. One spleen excised due to trauma was positive for disease, but the remaining two were negative. In the remaining nine cases (14.7%), splenectomy was performed with en-bloc resection of adherent omentum to spleen in five cases (8.2%), while in the remaining four cases (6.5%), splenectomy was performed based on intraoperative suspicion of splenic disease. The preoperative CT scan assessment demonstrated a low false positive rate of 2.7% (95%CI=0.1-16.2%). Specificity was also high at 99.7% (95%CI=98.4- 9.9%), but sensitivity was relatively low at 59.3% (95%CI=45.7-71.6%).
Surgery. Notably, a significant proportion of splenectomies (n=33; 54.1%) were conducted by three gynaecological oncology surgeons, with the remaining cases (n=28; 45.9%) being performed in collaboration with gastrointestinal surgeons. Among all splenectomy cases, primary cytoreductive surgery (pCRS) was performed in 20 cases (32.8%), while interval cytoreductive surgery (iCRS) was performed in 41 cases (67.2%) following neoadjuvant chemotherapy. In the entire cohort of 469 patients underwent cytoreductive surgery, 39.4% had primary cytoreductive surgery, while 60.6% underwent interval cytoreductive surgery. Only one case (1.5%) necessitated splenectomy together with distal pancreatectomy due to the location of the disease. Cytoreduction to no visible residual disease was achieved in 50 cases (82%), while optimal cytoreduction was achieved in nine cases (15%%). Suboptimal cytoreduction was observed in the remaining two cases (3%).
Among the cohort, 40 cases (65.6%) exhibited a high surgical complexity score (modified Aletti score >8), while in 20 cases (32.8%) there was an intermediate score ranging between four and seven. One patient had a low surgical complexity score of <4. The median estimated blood loss was 1,500 ml, with an IQR of 1,200-2,275 ml.
Perioperative outcomes. In our centre, among all 61 cases of splenectomy, there was only one (1.6%) instance of postoperative pancreatic fistula documented based on the new definition. This occurred in the patient who underwent a distal pancreatectomy.
Biochemical leaks were observed in 15 cases (24.5%), all of whom did not require further intervention. There were eight cases (13.1%) with a grade 3-4 Clavien-Dindo complication, none of which were related to the splenectomy. Out of these cases, four patients developed pleural effusion with associated chest infection, two patients had pelvic collections requiring drainage, one patient had an abdominal hematoma that necessitated drainage, and one patient experienced gastric perforation, requiring surgical reintervention. There were three (4.9%) deaths within 90 days. The causes of death in our cohort were cardiac arrest associated with ventilator-associated pneumonia (Pseudomonas aeruginosa), complications following rectal stump breakdown, and septic shock with multi-organ failure due to postoperative bowel perforation. None of these fatalities were directly attributable to splenectomy-specific complications.
The median length of inpatient stay was nine days, with an IQR of 7-14 days. Among patients requiring adjuvant chemotherapy post-surgery, the median time to chemotherapy (TTC) was 33.5 days (IQR=27.25-43 days), with 93.1% initiating treatment within the recommended 6-8 weeks post-surgery.
When comparing patients who experienced biochemical leak to those who did not, the median TTC was 5.0 weeks (35 days) in the leak cohort and 4.3 weeks (30 days) in the non-leak cohort, with no statistically significant difference between the groups (p=0.61). The patient who experienced a POPF had a delay in TTC exceeding 12 weeks.
Long-term follow-up. All patients appropriately received post-splenectomy instructions at the time of discharge. Among the patients alive at the time of censoring, 18 out of 22 consented to telephone consultations to audit compliance with the recommended post-treatment protocol. The median follow-up duration for this cohort was 26.6 months, ranging from 13.4-86.5 months from the date of surgery to the completion of the questionnaires.
Within this subgroup, 17 out of 18 patients (94%) reported receiving vaccinations post-splenectomy within 6-8 weeks, while 17 patients (94%) continued to receive the yearly vaccinations as advised. Regarding long-term prophylactic antibiotic usage, 16 patients (88.9%) affirmed ongoing adherence to recommendations.
Discussion
This study uses the new grading system from the International Study Group of Pancreatic Surgery to define and classify POPF after splenectomy for ovarian cancer and therefore provides an up-to-date and accurate description of the severity of this postoperative complication. Reassuringly, we have identified a very low rate of clinically significant leak of 1.6%.
In this study of 469 patients with advanced epithelial ovarian cancer, the incidence of splenectomy during cytoreductive surgery was 13% (95%CI=9.95-16.71%) and only one case was associated with distal pancreatectomy (0.21%, 95%CI=0.0054-1.1188%). This splenectomy rate aligns with the 4.1-13.8% reported in a 2011 systematic review by Hanprasertpong et al. (13) but is slightly less than the 19.4% reported by Bizzari et al. (12).
Distal pancreatectomy is a well-established risk factor for the development of POPF, increasing the likelihood of this complication (12, 16). The incidence of distal pancreatectomy in this study is significantly lower than rates reported by Bizzari et al., which noted a 9.7% incidence (n=13, p<0.00001), and Sozzi et al., which reported a rate of 4.6% (n=10, p<0.00001) (12, 13).
However, the incidence of biochemical leak in our cohort was 24.5% (n=15), which was higher than that reported by Rush et al. (8.5%, n=4, p=0.0482), Bizzarri et al. (8.2%, n=11, p=0.0037), and Sozzi et al. (10.2%, n=22, p=0.0069) (12, 23, 24). This may have been related to local practice where drains are usually placed and amylase is often routinely assessed post-operatively. Despite the higher leakage rate, all cases were managed conservatively and patient recovery and time to chemotherapy were not impacted.
A meta-analysis of 6,297 patients by Wang et al. identified that patients who underwent splenectomy had a higher overall incidence of post-operative complications, prolonged hospital stays and an increased time interval between surgery and the initiation of adjuvant chemotherapy (25). Our cohort experienced a 13% (n=8) rate of grade 3-4 Clavien-Dindo complications. These findings are comparable (p=0.6524) to those of a large meta-analysis of cytoreductive surgery in ovarian cancer by Kengsakul et al., involving 15,325 patients, where grade 3-4 postoperative complications were reported in 15% (n=2,357) of cases (26). This suggests that performing splenectomy as part of cytoreductive surgery for advanced ovarian cancer does not significantly increase the risk of major complications. The notably low incidence of major specific and non-specific complications, such as true POPF, further supports the safety of this surgical approach.
A key strength of our study is that this cohort comprises real-world data in an unselected cohort of patients, and includes information about how patients adhere to recommendations postoperatively. All consecutive patients were included over an eight-year period at a gynaecological centre, with comprehensive documentation of complications and further treatment.
The main weakness of this study is the retrospective approach, assessing only patients alive and well, greater than a year after surgery. It may be that only the most engaged and fit patients were able to respond, and therefore introduces a potential bias in the rate of adherence to recommendations. We found that 95% (17/18) of patients consistently continued with the vaccination and immunisation scheme, which is a higher rate than that described by Rush et al., of 89.4% (42/47). However, assessing at >1 year has allowed us to identify whether patients had any long-term effects of splenectomy.
Our study indicates that while preoperative CT scans have high specificity for detecting splenic involvement in advanced epithelial ovarian cancer, their sensitivity is limited. Consequently, there is a likelihood of identifying splenic disease intraoperatively in approximately one out of 20 patients who initially had no suspicion of splenic metastasis on their CT scan. Given these findings, it is crucial to counsel patients about the potential need for splenectomy during cytoreductive surgery for advanced ovarian cancer. Patients should be made aware of the life-altering consequences and the associated risks of undergoing a splenectomy, even when pre-operative imaging does not identify disease. Additional pre-operative imaging does not appear to add benefit, as Berthelin et al. demonstrated that combining CT and MRI does not significantly enhance the preoperative diagnosis of splenic involvement in patients with epithelial ovarian cancer compared to using CT alone (27).
Our study is among the few that have assessed patient compliance with lifelong antibiotic prophylaxis and immunisation following splenectomy. Splenectomy is a life-altering intervention with potential consequences on the quality of life (QoL), particularly in oncological patients. There is a substantial need for further research exploring QoL outcomes and adherence to post-splenectomy care protocols in this patient population.
Conclusion
This study highlights the high accuracy of preoperative imaging in detecting splenic disease, although its limitations necessitate careful intraoperative assessment. Overall, the results of this study suggest that splenectomy can be safely performed as part of cytoreductive surgery for advanced ovarian cancer, without significantly increasing the risk of major complications. Notably, this is one of the few studies to assess and demonstrate high compliance with post-splenectomy prophylactic management, including immunisation and antibiotic prophylaxis. These findings contribute to the growing body of evidence supporting the safety and efficacy of this surgical approach, advocating for comprehensive preoperative counselling regarding the potential need for splenectomy and its implications for patient care.
Footnotes
Authors’ Contributions
Study concepts: Andrei Corha, Radha Graham, Dhivya Chandrasekaran, Nicola D MacDonald, Khaled Dawas, Ioannis C. Kotsopoulos; Study design: Andrei Corha, Ioannis C Kotsopoulos, Radha Graham; Data acquisition: Andrei Corha; Quality control of data and algorithms: Andrei Corha; Data analysis and interpretation: Andrei Corha, Ioannis C Kotsopoulos, Radha Graham; Statistical analysis: Andrei Corha; Manuscript preparation: Andrei Corha; Manuscript editing: Andrei Corha, Radha Graham, Dhivya Chandrasekaran, Nicola D MacDonald, Khaled Dawas, Ioannis C Kotsopoulos; Manuscript review: Andrei Corha, Radha Graham, Dhivya Chandrasekaran, Nicola D MacDonald, Khaled Dawas, Ioannis C Kotsopoulos.
Conflicts of Interest
The Authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.
Artificial Intelligence (AI) Disclosure
During the preparation of this manuscript, a large language model [ChatGPT (GPT-4, OpenAI] was used solely for language editing and stylistic improvements in select paragraphs. No sections involving the generation, analysis, or interpretation of research data were produced by generative AI. All scientific content was created and verified by the authors. Furthermore, no figures or visual data were generated or modified using generative AI or machine learning-based image enhancement tools.
- Received October 18, 2025.
- Revision received November 9, 2025.
- Accepted November 17, 2025.
- Copyright © 2026 The Author(s). Published by the International Institute of Anticancer Research.
This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY-NC-ND) 4.0 international license (https://creativecommons.org/licenses/by-nc-nd/4.0).
References
Jump to section
Related Articles
Cited By...
- No citing articles found.