Abstract
Aim: To determine the impact of survival of peritoneal versus splenic metastasis in cases submitted to splenectomy as part of cytoreductive surgery in recurrent epithelial ovarian cancer. Patients and Methods: Between January 2002 and May 2014, 28 patients were submitted to splenectomy as part of secondary, tertiary and beyond tertiary cytoreduction at the Dan Setlacec Center of Gastrointestinal Disease and Liver Transplantation, Fundeni Clinical Institute, Bucharest. Results: Splenectomy was performed as follows: at secondary cytoreduction in 21 cases, at tertiary cytoreduction in six cases, and beyond tertiary cytoreduction in one case. An R0 resection was attempted in all cases; however, in two cases submitted to splenectomy as part of tertiary cytoreduction, R1 and R2 resection, were performed, respectively. Histopathological studies revealed the presence of peritoneal seeding in 11 cases at secondary cytoreduction and in four cases submitted to splenectomy as part of tertiary cytoreduction. Parenchymatous lesions were described in nine cases submitted to splenectomy as part of secondary cytoreduction and in two cases at tertiary cytoreduction. In a single case in which splenectomy was performed in the context of secondary cytoreduction, hilar involvement was found. Peritoneal seeding was described in the patient for whom splenectomy was performed at quaternary cytoreduction. Early postoperative mortality for the entire cohort (within 30 days) was 7.1% (death occurred in two cases submitted to splenectomy during the secondary cytoreduction). The median overall survival in patients with splenic involvement via peritoneal route was 35 months, while in cases with hematogenous splenic lesions, it was 12 months (p=0.2) at secondary cytoreduction. In the sub-group of patients submitted to splenectomy as part of tertiary cytoreduction, the median overall survival in patients with splenic involvement via peritoneal route was 21 months, while in cases with hematogenous splenic lesions it was 4 months (p=0.08). The patient submitted to quaternary cytoreduction died of disease 20 months later. Conclusion: splenectomy as part of secondary, tertiary and quaternary cytoreduction can be performed safely, with acceptable rates of morbidity and mortality. The maximal survival benefit seems to be obtained for patients with splenic involvement via peritoneal route, while those with hematogenous spread live a shorter period; further study is required in order to assess if resection in such cases is preferable to palliative chemotherapy. Maximal survival benefit occurs in the setting of secondary cytoreduction, although in selected cases, even quaternary cytoreduction can be followed by long-term survival.
Ovarian cancer is responsible for the highest tumor-related mortality of all gynecological malignancies, with an estimated number of 42704 deaths in Europe in 2012 (1, 2). This is mainly related to the fact that most patients are diagnosed with an advanced stage of the disease (3). However, even if a complete clinical remission is achievable in up to 80% of these cases due to association of cytoreductive surgery and adjuvant chemotherapy, the typical clinical course of the disease is characterized by multiple relapses, even in cases in which a complete surgical resection had been initially achieved (4-6). Multiple strategies involving surgery, chemotherapy and even radiotherapy have been proposed in order to treat recurrent disease (7); however, it seems that ovarian cancer relapse is best controlled by an aggressive surgical approach based on the same principles of maximal debulking effort and minimum residual disease. Even in cases submitted to an R0 resection for recurrent ovarian cancer, various outcomes have been observed, leading to the conclusion that there are also other prognostic factors which might influence the outcome (8-12). One of these factors seems to be the location of the tumoral burden, patients with upper abdominal involvement experiencing a poorer outcome when compared to those with pelvic-confined recurrent disease (13).
The aim of our study was to observe outcome differences with respect to different mechanisms of splenic involvement (parenchymatous versus peritoneal) in the setting of secondary, tertiary and beyond tertiary cytoreduction for ovarian cancer relapse.
Patients and Methods
After obtaining the Ethics Committee approval (no. 143/2015), we retrospectively reviewed data of patients submitted to splenectomy as part of secondary, tertiary and beyond tertiary cytoreduction at the Dan Setlacec Center of Gastrointestinal Disease and Liver Transplantation, Fundeni Clinical Institute, Bucharest, between January 2002 and May 2014. We found 28 patients eligible for the study; splenectomy was performed at the time of secondary cytoreduction in 21 cases, at tertiary cytoreduction in six cases and beyond tertiary cytoreduction in one case. In all cases, the splenic involvement was confirmed by the histopathological studies and was classified as peritoneal seeding (lesions developed on the peritoneal route with limited parenchimatous invasion), parenchymatous lesions (lesions with hematogenous origin, entirely surrounded by normal splenic parenchyma) and hilar lesions (tumoral burden in the splenic hilum). Dates of death were obtained from the National Register of Population. Survival was measured from the moment of performing splenectomy until the time of death. The differences between different subgroups were analyzed by the log-rank test and considered significant when p<0.05. Statistics and graphics including Kaplan–Meyer survival curves were performed using R Program-version 3.1.2 (http://www.r-project.org).
Results
Splenectomy as part of secondary cytoreduction. In the subgroup of patients submitted to splenectomy at secondary cytoreduction, the mean age at initial diagnosis was 54.7 years (range 34 to 75 years), while International Federation of Gynecology and Obstetrics (FIGO) stage at diagnosis was stage III or higher in 14 cases. Postoperatively, all patients were submitted to taxane- and platinum salt-based adjuvant chemotherapy. The mean disease-free survival between ending the adjuvant oncological treatment and re-appearance of symptoms was 31 months (range 13 to 37 months).
At secondary cytoreduction, the most frequent associated visceral resection was colic or rectosigmoidial resection, which was performed in 16 cases (76.1%). Upper abdominal tumor burden involved the diaphragmatic peritoneum in 12 cases (57.1%), the gall bladder in six (28.5%), the distal part of the pancreas in four (19%), the diaphragm in three (14.2%), the liver by peritoneal seeding in two (9.5%), hematogenous route in one case (4.76%) and the stomach in one case (4.76%). Complete resection, defined as no macroscopic residual disease was achieved in all cases.
Preoperative and intraoperative characteristics of the patients submitted to splenectomy as part of secondary cytoreduction.
The main pre- and intraoperative characteristics of the cohort are shown in Table I.
The histopathological studies revealed the presence of the serous type in 95.2% of cases while the mucinous type was present in a single case (4.8% of cases). The same study revealed the presence of peritoneal seeding in 52.3% of cases, while the hematogenous route was incriminated in 42.8%; hilum-limited tumoral burden was found in 4.7% of patients.
The median length of hospitalization was 14 days (range 2 to 89 days). Early postoperative complications (occurring within the first 30 postoperative days) were reported in six cases. In three cases re-operation was needed: the first patient necessitated early re-operation due to a high-flow enteral fistula, the second case had a pancreatic fistula unresponsive to conservative treatment, while in the third case, re-operation was required due to early postoperative hemoperitoneum. The overall mortality rate was 9.5%; death occurred in two cases, due to uncontrolled abdominal sepsis in one case and hemorrhagic shock associated with disseminated intravascular coagulopathy developing on the third postoperative day in the second case.
The influence on overall survival of patients with peritoneal versus those with parenchymatous splenic involvement.
The main postoperative complications classified according to Dindo-Clavien scale (14) are shown in Table II.
The median overall survival after secondary cytoreduction was 24 months. The median overall survival in patients with splenic involvement via peritoneal route was 35 months, while in cases with hematogenous splenic lesion, it was 12 months (p=0.2) (Figure 1).
Splenectomy as part of tertiary cytoreduction. At tertiary cytoreduction, splenectomy was performed in six cases. The mean age at initial diagnosis was 42.8 years (range 39 to 46 years), while the FIGO stage was IIC in four cases, IIIA in one case and IIIC in one case.
Tertiary cytoreduction was performed at a median interval of 50 months after primary cytoreduction. All patients were submitted to adjuvant chemotherapy with platinum-based salts and taxane after primary and secondary cytoreduction. Other upper abdominal resections included: diaphragmatic peritonectomy in two cases, cholecystectomy in three cases, distal pancreatectomy in one case and atypical hepatectomy for a parenchymatous lesion in one case.
The main preoperative and intraoperative characteristics are shown in Table III.
The main postoperative complications at the time of secondary cytoreduction.
Pre- and intraoperative characteristics of the patients submitted to splenectomy as part of tertiary cytoreduction.
Although an R0 resection was attempted in all cases, it was achieved in four out of the six patients; in the other two cases, R1 and R2 resection were performed. The histopathological studies revealed the presence of a serous sub-type in five cases, while the sixth case was diagnosed with a mucinous epithelial ovarian tumor. When it comes to the type of splenic involvement, the peritoneal route was incriminated in four cases, while the other two cases had hematogenous lesions, entirely surrounded by normal splenic parenchyma. Postoperatively, two patients developed complications: an abdominal abscess and a biliary fistula, respectively, both being managed by conservative methods. The median hospitalization was 11 days (range 6 to 21 days).
The median overall survival was 6.7 months. The median overall survival in patients with splenic involvement via peritoneal route was 21 months, while in cases with hematogenous splenic lesion, it was 4 months (p=0.08).
Splenectomy for relapsed epithelial ovarian cancer beyond tertiary cytoreduction. Splenectomy was performed at quaternary cytoreduction in a single patient. At initial diagnosis of stage IIC ovarian cancer, the patient was 48 years old. The quaternary cytoreduction was performed 102 months after the initial diagnosis and consisted of splenectomy, atypical hepatectomy, partial gastrectomy and partial cystectomy for a pelvic recurrence invading the urinary bladder, an upper abdominal recurrence invading the great curvature of the stomach and the spleen, and a unique liver lesion which proved to have hematogenous origin. The histopathological studies revealed the presence of a moderately differentiated serous epithelial ovarian tumor; the specimen of splenectomy presented disseminated peritoneal lesions. The postoperative course was uneventful, the patient being discharged on the 11th postoperative day. The patient died of disease 20 months after tertiary cytoreduction.
Discussion
Ovarian cancer remains an aggressive disease, with a high capacity to metastasize via peritoneal, lymphatic and hematogenous routes. Unfortunately, a significant number of patients will present at initial diagnosis or at relapse with upper abdominal involvement consisting of liver, pancreatic, gastric, lesser omentum or splenic metastases (15). Concerning the splenic involvement, histopathological studies revealed that its invasion might develop secondarily to any of these patterns of spread, resulting in peritoneal seeding, parenchymatous or hilar lesions (15). When it comes to the different patterns of splenic invasion in advanced-stage ovarian cancer, large studies have demonstrated significant differences in terms of survival for patients with peritoneal versus parenchymatous or hilar lesions (16-19). Most authors considered that these differences in survival are related to a more aggressive biological behavior (16-18); however, the benefits achieved in overall survival enabled them to consider that splenectomy at the moment of primary cytoreduction is perfectly justified in order to maximize the surgical debulking effort (17-19).
The aim of our study was to determine if the same differences in terms of survival for the three patterns of spread are still present at the time of secondary cytoreduction and beyond. We formally excluded all cases in which the histopathological studies revealed no tumoral invasion of the spleen. The final cohort included 28 patients submitted to splenectomy during secondary (21 cases), tertiary (6 cases) and quaternary (one case) cytoreduction. Peritoneal involvement was found in 11, four and one case, respectively, and was associated with increased median overall survival (32 months, 21 months and 20 months, respectively), while parenchymatous lesions were found in nine cases at secondary and two at tertiary cytoreduction and were associated with a poorer median overall survival (12 months and 4 months, respectively (or peritoneal versus parenchymatous lesions p=0.199 fat secondary cytoreduction and p=0.08 at tertiary cytoreduction).
Matigbay et al. conducted a similar study on cases submitted to splenectomy during primary and secondary cytoreduction. They reported a total of 46 patients who underwent splenectomy as part of secondary cytoreduction, with a median age of 59.5 years (15). However, in their study group they also included 15 cases in which splenectomy was performed due to intraoperative trauma to the spleen with bleeding that could not be managed by usual hemostatic techniques. Similarly to our study, visceral resections such as colic resections or urinary tract resections were performed by Matigbay et al. (48% had bowel resection, and 7% had partial resection of the urinary tract) in order to achieve an R0 resection. The authors reported an overall postoperative mortality rate of 5%, lower when compared to our results; however, in their cohort the mortality rate was calculated on the entire group involving both primary and secondary cytoreduction, while in our case, the overall mortality was 9.5% in the sub-group submitted to splenectomy at secondary cytoreduction. In the same study, the median overall survival for the secondary cytoreduction group was 20.3 months, with an estimated 2-year survival of 42.3%. The same study concluded that no significant difference existed in terms in survival in patients with peritoneal versus those with parenchymatous involvement, neither at primary nor secondary cytoreduction (15).
In our cohort, the postoperative mortality among patients submitted to secondary cytoreduction was 9.5%. In one case death occurred due to uncontrolled abdominal sepsis which evolved into multiple organ dysfunction; however, the source of abscess was related to an associated pancreatic resection which developed a postoperative grade C pancreatic fistula and secondary acute necrotizing pancreatitis. In the second case, death occurred due to postoperative disseminated intravascular coagulation; this fact might be also correlated with the inherently increased risk of patients with ovarian cancer for developing coagulation disorders (15).
When it comes to tertiary cytoreduction, data reporting of splenectomy in order to maximize the debulking efforts are even scarcer. In a study conducted by Tang et al. involving 83 patients submitted to tertiary cytoreduction, splenectomy was performed in two cases only. The same study demonstrated that patients presenting upper abdominal tumoral involvement had a significantly poorer outcome when compared to those with pelvic-confined disease (with a reported median survival of only 14.6 months versus 34.9 months, respectively, p=0.001). However, in multivariate analysis, the tumor site was not a survival determinant. The same study demonstrated the benefits of maximal debulking effort during tertiary cytoreduction when compared to a control group of patients with the same characteristics in which only chemotherapy was performed. The median overall survival for cases submitted to tertiary cytoreductive surgery was 26.9 months, while the median overall survival for the chemotherapy-treated sub-group was only 15 months (p=0.01) (13).
Conclusion
The presence of parenchymatous splenic involvement was associated with decreased median overall survival in both secondary and tertiary cytoreduction. However, splenectomy was safely performed in all cases, with acceptable rates of postoperative complications.
- Received April 21, 2015.
- Revision received May 24, 2015.
- Accepted May 26, 2015.
- Copyright© 2015 International Institute of Anticancer Research (Dr. John G. Delinassios), All rights reserved






