Title
Author
DOI
Article Type
Special Issue
Volume
Issue
Prediction of long-term survival in recurrent gynaecological malignancies treated with pelvic exenteration
1Department of General and Cancer Surgery, Surgery Clinic, North Estonian Medical Centre, 10111 Tallinn, Estonia
2Department of Surgical Oncology and Gynecology, Clinicum of University of Tartu, 50406 Tartu, Estonia
3Institute of Biomedicine and Translational Medicine, University of Tartu, 50406 Tartu, Estonia
4Department of Hematology and Oncology, Clinicum of University of Tartu, 50406 Tartu, Estonia
5Tartu Health Care College, 50406 Tartu, Estonia
*Corresponding Author(s):olav.tammik@kliinikum.ee (Olav Tammik)
| History | Submitted: 20 August 2025 | Accepted: 09 October 2025 | Published: 15 December 2025 |
| Copyright: | ©2025 The Author(s). Published by MRE Press. |
Background: The desire to ensure prolonged survival for patients with advanced gynaecological cancer is a key concern for every doctor. This study presents an analysis of outcomes from an Estonian regional cancer center, evaluating prognostic factors associated with locally advanced gynecological cancer and the potential for long-term survival following pelvic exenteration. Methods: The study included patients with resistant or recurrent gynecological cancer. The analysis present data on all 34 cases in which pelvic exenteration was performed at the Tartu University Hospital. The majority of patients (88.2%) had tumour recurrence. Thirteen patients (38%) had previously undergone surgery for gynaecological cancer. Results: Of the 34 patients, 21 (62%) underwent total pelvic exenteration and 13 (38%) exenterations with vulvectomy. The median age of the study cohort was 59 years (range, 34–80). The median duration of surgery was 4.4 hours (range, 121–530 minutes). The median blood loss was 707 mL (range 100–2500 mL). The median tumor diameter was 7 cm (range 2–25 cm). The mean follow-up was 52 months. The overall postoperative complication rate was 41%. One patient died postoperatively (2.9%). Median survival for ovarian cancer was 48.3 months, cervical cancer 31.6 months, vaginal or urethral cancer 29.2 months and for endometrial cancer 7.7 months. Overall survival rates were 62% at 1 year, 44% at 3 years, 28% at 5 years, and 15% at 15 years. Multivariable analysis showed that distant metastasis (p = 0.0002), endometrial cancer (p = 0.0096), resection line (p = 0.017) were the most important factors affecting long-term survival. Conclusions: In the present study, we found that patients with relapsed or persistent locally advanced cervical, vaginal, or ovarian cancer who underwent pelvic exenteration demonstrated substantially better survival outcomes than patients with endometrial cancer, particularly those with aggressive histologic subtypes and multiple unfavorable prognostic markers.
Cite this article
Olav Tammik, Aavo Lang, Heti Pisarev, Martti Laan, Katrin Lang, Hele Everaus, Karin Tammik. Prediction of long-term survival in recurrent gynaecological malignancies treated with pelvic exenteration. European Journal of Gynaecological Oncology. 2025; 46(12): 37-45. doi: 10.22514/ejgo.2025.144
1. Background
The treatment of locally widespread, drug-resistant, or relapsed gynaecological tumours poses major challenges for therapy. In the management of locally advanced pelvic tumors, securing negative microscopic resection margins represents the most decisive prognostic factor [1]. In cases of recurrent malignant tumors in the pelvic region, it is often difficult to achieve radical resection using conventional surgical methods, as these tumors are usually closely connected to neighboring structures or grow into them. Therefore, en bloc resection of the tumour and surrounding organs—i.e., total pelvic exenteration (TPE)—is required [2].
Total pelvic exenteration is a method that originated from the urgent need to help patients for whom conventional treatments did not provide relief. The method was first described by surgeon Brunswig, who performed this operation on a patient with recurrent cervical cancer. Initially called a “desperate Procedure”, the operation posed a great risk to the patient’s life, as the postoperative mortality rate was as high as 23% [3, 4]. Since its inception, pelvic exenteration has undergone many significant developments in terms of surgical techniques, diagnostics, and perioperative care approaches [5, 6, 7, 8]. Morbidity remains considerable, with 15–68% of patients experiencing complications following TPE [3, 7]. Postoperative mortality has also been reported to reach up to 14% [8].
Outcomes of surgical therapy are influenced both by the surgeon’s expertise and by institutional surgical volume [9]. Treatment outcomes still vary widely, with five-year survival rates ranging from 15% to 77.6% in patients treated for the first line and from 0% to 23% in patients with recurrent disease [10, 11, 12]. Furthermore, the benefits of adjuvant therapy have not been reported in several types of uterine sarcomas. According to the National Cancer Database 1884 cases, chemotherapy had no effect on survival and post-operative radiotherapy even had a negative effect on survival [13]. Nevertheless, it is critically important to select patients who are suitable for TPE [14]. Thus, the aim of the present study was to evaluate the morbidity and outcomes of patients with locally advanced gynaecological cancer undergoing TPE at the Department of Surgical Oncology, University of Tartu, and to identify possible reasons for the prolonged survival.
Given the high incidence of gynaecological tumors in Estonia, this is an important research topic. In 2021, among cancer cases prevalent in Estonian women, uterine cancer ranked as the second most common, cervical cancer as the fourth, and ovarian cancer as the eighth most frequent malignancy of the female genital tract. At the time of diagnosis, 50.8% of cervical cancer cases were locoregional and 8.2% had distant metastases. For uterine cancer, 7.1% and 7.9% and for ovarian cancer, 43.4% and 23.3%, respectively. Up to 10% of uterine cancers remain at an unknown stage at diagnosis [15].
2. Patients and methods
2.1 Study type
The present work was designed as an observational study to identify prognostic factors for long-term survival in patients with advanced gynaecological cancer treated after cancer recurrence or due to persistent cancer after radical radiochemotherapy. We included all patients diagnosed with gynecological cancer who underwent pelvic exenteration during the study period, regardless of their precise diagnosis, prior treatments, comorbidities, gender, or age. No patients diagnosed with gynecological cancer who underwent pelvic exenteration in the region were excluded from the study.
Between January 2001 and February 2021, a total of 34 total pelvic exenteration (TPE) procedures were performed for gynecological malignancies in the Department of Surgical Oncology at the University of Tartu, Estonia (population approximately 1.3 million). Of these patients, 22 had cervical cancer, 5 endometrial cancer, 4 vaginal cancer, 2 ovarian cancer and urethral cancer. Thirteen patients (38%) had previously undergone surgery for cervical or vaginal cancer, subsequently received radiochemotherapy after recurrence, and were later operated on for persistent disease. Six patients (18%) underwent emergency surgery due to tumor-related bleeding. Fifteen patients (44%) presented with persistent tumor following radical radiotherapy and chemotherapy. Some of the cervical cancer cases presented have been previously analyzed in our preliminary analysis [16]. Preoperative assessment was performed using computed tomography scan (CT scan), magnetic resonance imaging (MRI) or both modalities.
All patients received perioperative prophylactic antibiotics, and those with positive urine cultures were treated according to the antibiogram. Since 2002 we use enhanced recovery protocol after surgery (ERAS) [17] for all operated patients.
All procedures were conducted by a single dedicated surgical oncology team.
2.2 Pathology
Staging was conducted according to the American Joint Committee on Cancer (AJCC) Tumour, Node, Metastasis (TNM) classification, 8th edition, and pathology reports were used to determine histological subtype, margin status, lymphovascular invasion, nodal status, and completeness of excision.
2.3 Evaluation of morbidity and mortality
Patient data were obtained from medical records and charts. Collected variables included demographic information, diagnosis, TNM classification, previous treatments, surgical parameters (technique, duration, blood loss, bed rest), histopathological findings, follow-up, and overall survival. Complications were classified according to the Clavien-Dindo system and categorized as mild or severe, with the latter requiring reoperation. Early complications were defined as events occurring within 30 days postoperatively. Follow-up continued until death or 01 June 2025.
2.4 Statistical analysis of survival
Categorical variables were presented as frequencies and percentages, whereas continuous variables were expressed as medians with corresponding ranges. Overall survival was defined as the interval from the date of tumor resection to the date of last documented follow-up or death from any cause. The data cutoff was set at 01 June 2025, after which patients were censored. Survival according to risk factors was estimated at 5 and 10 years.
Postoperative survival was estimated using the Kaplan-Meier method [2], and all survival outcomes were reported with 95% confidence intervals. Univariate survival comparisons were performed with the log-rank test. Multivariate analysis of prognostic factors for overall survival was conducted using the Cox proportional hazards model [1]. A two-sided p-value of < 0.05 was considered statistically significant. All statistical analyses were performed using R software, version 4.3.2 (R Foundation for Statistical Computing, Vienna, Austria).
3. Results
All 34 patients underwent a comprehensive and standardized preoperative preparation protocol designed to optimize surgical outcomes and minimize perioperative. This included a thorough anesthesiologist evaluation with risk assessment according to the American Society of Anesthesiologists (ASA) classification, along with a detailed assessment of potential perioperative complications and verification of an adequate blood reserve to manage intraoperative hemorrhage. Bowel preparation was performed when indicated to reduce the risk of contamination, and pharmacologic prophylaxis with low-molecular-weight heparin (LMWH) was administered to mitigate the risk of thromboembolic events. In addition, preoperative skin preparation and shaving were conducted to lower the risk of surgical site infection, informed consent was obtained from all patients to ensure ethical compliance, and prophylactic antibiotics were administered 30 minutes prior to skin incision in accordance with established surgical guidelines.
3.1 Surgery
Study characteristics are presented in Table 1. Median patient age was 59 years (range, 34–80). The median operation time was 4.4 hours (263 minutes, range 121–530 minutes). For 3 patients (9%), operation time was 3 hours or less (range 121–180 minutes), for 24 patients (70%), operative time was 3 to 6 hours (range 181–360 minutes), and for 7 patients (21%), operative time was 5 to nearly 9 hours (range 301–530 minutes).
| Patient characteristics | Estimate, range | Percent, % | |
| Age (yr) | |||
| Median | 59 | ||
| Range | 34–80 | ||
| Tumor type | |||
| Cervical | 22 | 64 | |
| Endometrial | 5 | 15 | |
| Vaginal | 4 | 12 | |
| Ovarian | 2 | 6 | |
| Urethral | 1 | 3 | |
| Operation time (min) | |||
| Median | 263 | ||
| Range | 121–530 | ||
| Blood loss (mL) | |||
| Median | 707 | ||
| Range | 100–2500 | ||
| Operation type | |||
| Total | 21 | 62 | |
| Anterior | 13 | 38 | |
| With vulvectomy | 13 | 38 | |
| Rectal anastomosis | 5 | 24 | |
| With omentoplasty | 2 | 15 | |
| Urinary derivation | |||
| Bricker conduit | 32 | 94 | |
| Lund pouch | 2 | 6 | |
| Tumor diameter (cm) | |||
| Median | 7 | ||
| Range | 2–25 | ||
| Histological subtype | |||
| Squamous cell carcinoma | 23 | 67 | |
| Adenocarcinoma | 3 | 9 | |
| Papillary serous carcinoma | 2 | 6 | |
| Adenosquamous carcinoma | 1 | 3 | |
| Papillary carcinoma | 1 | 3 | |
| Primitive neuroectodermal tumor | 1 | 3 | |
| Liposarcoma | 1 | 3 | |
| Serous carcinoma | 1 | 3 | |
| Mucinous adenocarcinoma | 1 | 3 | |
| Tumor stage | |||
| pT0 | 2 | 6 | |
| pT1 | 2 | 6 | |
| pT2/2b | 6 | 18 | |
| pT3/3b | 7 | 20 | |
| pT4 | 17 | 50 | |
| Lymph node metastases N1/2 | 13 | 38 | |
| R1 | 4 | 12 | |
| Distant metastases M1 | 9 | 26 | |
| Hospitalization (d) | |||
| Median | 16 | ||
| Range | 9–58 | ||
| Postoperative complications | 14 | 41 | |
| Operated due to complication | 3 | 9 | |
| Postoperative mortality | 1 | 3 | |
| Follow up (mon) | |||
| Median | 25 | ||
| Range | 0.4–268 |
The median blood loss was 707 mL (range 100–2500 mL). For 17 patients (50%), blood loss was 100–700 mL, for 17 patients (50%), blood loss was 701–2500 mL.
For 21 (62%) patients, the operation type was total exenteration, for 13 (38%) patients anterior exenteration; additionally, there were 5 (15%) total exenterations with rectal anastomosis, 2 (6%) total infralevator exenterations with omentoplasty and 13 (38%) exenterations with vulvectomy.
For urinary diversion, the Bricker conduit was employed in the majority of patients, as it represents a widely accepted and reliable standard technique. Ureteroenteric continuity was restored using the Wallace method of anastomosis. In two selected patients, however, a continent cutaneous diversion was performed by constructing a Lundiana pouch from a detubularized segment of the right colon, thereby providing an alternative approach aimed at maintaining continence.
Postoperative complications are summarized in Table 2. The overall complication rate was 41% (n = 14), including both minor and major events. Major complications accounted for 9% (n = 3) of cases and required reoperation. One patient died due to complication postoperatively (2.9%). In our series, most complications were classified as grade I–II according to the Clavien-Dindo classification (33%), grade IIIb complications accounted for 3 (9%), and one death was due to Gram-negative sepsis following jejunal perforation, grade V (3%).
The median length of stay following surgery was 16 days (range 9–58).
| Complication | Count | Percentage, % | Operation due to complication |
| Urinary tract infection | 5 | 15 | |
| Laparotomy wound infection | 3 | 9 | |
| Perineal wound infection | 1 | 3 | |
| Partial femoral nerve paresis | 1 | 3 | |
| Bowel obstruction, adhesions | 1 | 3 | Ileotransversostoma |
| Rectal anastomosis dehiscence | 1 | 3 | Colostoma |
| Pelvic abscess | 1 | 3 | |
| Ileus | 1 | 3 | Resectio coli, colostoma |
| Total | 14 | 41 |
3.2 Pathology
The median tumor diameter was 7 cm (range, 2–25 cm), and 23 patients (68%) were diagnosed with squamous cell carcinoma as the primary histological subtype. In addition, adenocarcinoma was identified in 3 patients, papillary serous carcinoma in 2 patients, and single cases were observed of adenosquamous carcinoma, papillary carcinoma, serous carcinoma, primitive neuroectodermal tumor, liposarcoma, and mucinous adenocarcinoma. In four cases, the tumor reached the resection line (12%). Thirteen patients (38%) had metastatic regional lymph nodes in the specimen, and nine patients (26%) had distant metastases.
3.3 Survival
Cumulative overall survival was 62% at 1 year, 50% at 2 years, 44% at 3 years, 28% at 5 years, 20% at 10 years, and 15% at 18 years (Fig. 1).
Fig. 1.Overall survival. CI: confidence interval.
Median survival for cervical cancer was 31.6 months, vaginal or urethral cancer 29.2 months, endometrial cancer 7.7 months and ovarian cancer 48.3 months after operation. Overall survival by localization of tumor is presented in Fig. 2.
Fig. 2.Survival by localization of tumor. CI: confidence interval.
The median survival was estimated at 1.7 years (95%, confidence interval (CI), 0.8–4.4). Using the reverse Kaplan-Meier method for censored data, the median duration of follow-up was 2.1 years (range, 0.03–22.4 years).
3.4 Prognostic factors
In addition to endometrial cancer (p = 0.0096), resection line (p = 0.017) and distant metastases (p = 0.0002), were statistically significant predictors of 5-year survival (Fig. 3). Younger age of patients (p = 0.21), regional lymph node metastases (p = 0.338), post-operative complications (p = 0.495), blood loss (p = 0.558), and tumour diameter (p = 0.559) did not have a statistically significant effect but showed a trend towards improved overall survival.
Fig. 3.Survival prognosis by patient characteristics. HR: Hazard ratio.
Operation time (p = 0.931) was not associated with survival. When stratified into two groups (<5 hours and >5 hours), no statistically significant differences in survival were observed between the groups (Fig. 3). The median survival was 64.7 months in patients younger than 50 years, 17.7 months in those aged 51–69 years, and 7.7 months in those aged 70–84 years. Overall survival by patient characteristics is shown in Fig. 3.
4. Discussion
When TPE, as a novel technique, was first introduced, the main use of total pelvic exenteration was to relieve advanced malignant tumors of the pelvis [5, 8, 18, 19, 20]. Currently, the main indication is the treatment of locally advanced or recurrent malignant tumors in the pelvic region for curative purposes [11]. Complete removal of the tumor is the only way to ensure the longest possible survival. High 5-year survival rates (41–48%) with excellent local control after TPE have been reported [14, 18, 20, 21]. Traditionally, this procedure has been associated with high mortality and morbidity rates [22], as mortality rates reached up to 23% in the early days of the method [5]. By now, TPE has become considerably safer and more effective [19, 23].
The main objective of this study was to evaluate the effectiveness of pelvic exenterations performed at our center and to identify possible prognostic factors that would ensure long-term survival even in cases of locally advanced gynecological tumors. In our study, most patients had recurrent gynecological tumors and there were no good alternative treatment options, which helps to plan further possible approaches on this basis. Due to the heterogeneity of patients, we cannot speculate on the reason for cancer recurrence. The main limitation of this study was the small and heterogeneous study group. Consequently, some potential prognostic markers—such as patient age, regional lymph node metastases, postoperative complications, blood loss, and tumor diameter—showed a trend toward improved survival; however, this trend was not statistically significant. The definitive importance of these markers, therefore, requires confirmation in larger study cohorts.
Negative resection margin rates have the greatest impact on long-term survival prognosis [24]. This study shows that with a positive resection margins have a median survival of 5.5 months at five years, compared to 39.9 months for those with clear margins.
The presence of distant metastases at the time of surgery also has a serious impact on long-term survival. In our study, the median survival of patients with distant metastases was 7.7 months compared to 29.2 months for patients without distant metastases.
The reported complication rate following pelvic exenteration (PE) varies widely in the literature. Some studies document rates as high as 78–81% [13, 24, 25], while others report that up to 62% of patients remain free of complications [26]. Most authors agree that urinary tract and wound infections are the most common postoperative complications [14, 26]. In the present series, 41% of patients experienced a complication, including urinary tract infections in 15% and wound infections in 12%, which are comparable to rates reported previously [13, 25, 26]. Subsequent studies have shown that patients with urinary outflow obstruction also had mild renal insufficiency, and infection was the most common problem in the continent patient group. As other publications have shown, long-term morbidity associated with urological problems is the most common phenomenon [27]. The reoperation rate in our series was 9%, also in line with recent studies [28, 29]. Based on the results of our study, it can be said that even smaller hospitals with specialized cancer surgery departments are able to perform PE procedures with an acceptable rate of complications and reoperations. In our cohort, the postoperative mortality rate was 2.9%, which is consistent with data reported by other groups (0–5%) [25, 26, 27].
The 5-year survival analysis showed significantly better outcomes for patients younger than 50 years (median survival 64.7 months). On the other hand, patients older than 70 years experienced worse outcomes (7.7 months). Several authors have shown that the rate of perioperative complications increases with age, but this does not seem to affect survival after pelvic exenteration [25, 26, 27]. Similarly, to the present study, however, it has been shown that age below 63 years of age was a significant factor with positive influence on survival [1].
Blood loss was a factor affecting 5-year survival in our study, although not statistically so. When dividing the study group into two groups (17/17) based on mean blood loss, we found that patients with a blood loss of less than 700 mL had a median survival of 39.97 months, and while those with a higher blood loss had a median survival of 17.77 months (Fig. 3). Although operation time greater than six hours has been identified as a prognostic factor in other recent studies, our findings suggest that blood loss may be of greater importance.
When we analysed the relationship between tumour diameter and long-term survival, we found that tumours with diameters of 2–6 and 7–12 cm showed a trend for much better median survival than tumours with diameters of larger than 13 cm. The corresponding figures were 51.4, 20.4 and 8.2 months, respectively.
Positive lymph nodes after pelvic exenteration have been reported in 12–57% of cases in the literature [30]; in our cohort, the rate was 38%. The presence of regional metastases also had a clear, although not statistically significant, negative effect on survival. Median survival was 20.1 months in patients with positive lymph nodes and 29.2 months in patients without lymph node metastases.
However, survival differed significantly for gynaecological tumours with different localisation. Median survival for cervical, vaginal, and urothelial cancers was around 30 months. For ovarian cancer, which is a more sensitive to chemotherapy, median survival was over 48 months. In contrast, the median survival for uterine cancer was notably low at 7.67 months (p = 0.0096). Closer analysis of uterine tumours revealed a confluence of several poor prognostic factors: all operated uterine tumours were large (12–15 cm in diameter), with both distant and regional metastases, and high intraoperative blood loss (735–2500 mL). Also, the pathohistological tumour types of the uterine tumours were mostly rare (primitive neuroectodermal tumor, mucinous adenocarcinoma).
The most favorable treatment regimen for recurrent cervical and uterine cancer is radiotherapy or radiochemotherapy which offer as high as 45% overall survival rates. However, this is not the case for patients with a history of prior radiotherapy [28]. When local recurrence occurs, therapeutic options are severely limited, largely because many patients have already received pelvic irradiation as part of their primary cancer treatment. Reirradiation is generally considered contraindicated due to the risk of toxicity in previously treated tissues, and the efficacy of chemotherapy is often diminished, particularly for tumors located within the poorly vascularized, irradiated pelvic tissue [21, 29]. In this context, radical surgical approaches such as pelvic exenteration may offer the only potentially curative option for selected patients, despite the high complexity and associated morbidity of the procedure.
Among the patients included in this study, nearly all patients with gynecological malignancies had previously undergone radical radiotherapy and/or chemotherapy, which substantially limits subsequent treatment options. Consequently, for patients with recurrent or persistent cervical cancer, surgical intervention frequently represents the only potentially curative approach, even within previously irradiated fields [31]. Although palliative chemotherapy may provide modest benefit, with a median overall survival of approximately 10 months [32], pelvic exenteration can extend median survival up to threefold in carefully selected patients. Furthermore, radical surgery remains the only treatment capable of achieving five-year survival rates as high as 40% in this population, offering meaningful long-term outcomes when all other therapeutic modalities have been exhausted [25, 28, 33]. Despite the considerable complexity and associated morbidity of pelvic exenteration, careful patient selection and meticulous perioperative management can help optimize postoperative recovery and maintain quality of life for survivors.
Of the study, the main finding is evidence of the operative success of pelvic exenteration surgery in specialized centres with low morbidity and mortality in patients who have exhausted treatment options for gynaecological cancer and whose cancer is persistent or recurrent after radical chemoradiation [10, 34]. It is crucial to take into account possible prognostic factors when planning pelvic exenteration [5]. For optimal outcomes, surgery should be performed with clear resection margins. Also, it is advisable to consider the patient’s age, regional metastases, intraoperative blood loss and postoperative complications, and tumour diameter and pathohistology. In palliative and emergency settings, the presence of metastases should be taken into account when predicting survival.
5. Conclusions
Our study showed that pelvic excision is a feasible operation for advanced gynaecological malignancies and has the potential to ensure long-term survival for selected patients. The therapeutic potential of pelvic exenterative surgery is clearly observed in patients without risk factors that negatively affect long-term survival. Pelvic exenteration can be performed with low complication and mortality rates. Several patients in this series lived for 15 or more years.
In this study, we found that even patients with relapsed or persistent locally advanced cervical, vaginal, or ovarian cancer treated with pelvic exenteration had a much better survival prognosis than patients with endometrial cancer with aggressive histologic tissue type and various poor prognostic markers. To determine the significance of this result, it is necessary to wait for the results of larger multicenter studies.
Availability of data and materials
The data that support the findings of this study are available on request from the corresponding author. The data are not publicly available due to privacy or ethical restrictions.
Author contributions
OT—Conceptualization, data curation, writing. AL—Formal analysis, administration. HP—Analyzed bioinformatical results. ML—writing and editing. KL—Software and analysis. KT—Methodology, investigation. HE—Supervision, administration.
Ethics approval and consent to participate
The Research Ethics Committee of the University of Tartu approved the study on 18 May 2020, identified as: Surgical treatment results for pelvic organs malignancies, protocol nr: 315/T–1. Informed consent was obtained from all patients to ensure ethical compliance.
Acknowledgment
We would like to express my sincere gratitude to Karl Kull, whose dedicated guidance and support enabled me to start my career in surgical oncology.
Funding
This research received no external funding.
Conflict of interest
The authors declare no conflict of interest.
