Transvaginal Natural Orifice Transluminal Endoscopic Surgery (vNOTES) for Early-Stage Ovarian Cancer and Borderline Ovarian Tumors: A Case Series

1. Introduction

Surgery remains the cornerstone of ovarian cancer treatment, with the primary goal being the complete resection of the tumor. The quality of surgery and the surgeon’s exper-tise are critical to patient outcomes and survival. With advances in surgical techniques and a growing focus on improving patient perioperative outcomes, minimally invasive surgery (MIS) has become increasingly important in managing gynecological malignan-cies. However, its application in the treatment of ovarian cancer remains a subject of on-going debate.

Current guidelines recommend performing cytoreductive surgery for ovarian cancer via midline laparotomy, even in its early stages [1]. However, to date, no randomized con-trolled trials have directly compared MIS with open surgery for the treatment of early-stage ovarian cancer and borderline ovarian tumors (BOTs) [2], and minimally invasive ap-proaches are increasingly applied in their treatment with promising results [3,4]. Several studies suggest the feasibility and safety of MIS approaches for the management of ear-ly-stage ovarian cancer, appearing to be non-inferior to laparotomy [5,6] and presenting with lower rates of surgical complications [7,8].

Transvaginal Natural Orifice Transluminal Endoscopic Surgery (vNOTES) is an in-novative, minimally invasive approach that combines laparoscopy and vaginal surgery [9]. This approach has proven its feasibility and safety for treating several benign gyneco-logical conditions, being a valuable option for performing hysterectomies, myomectomies, adnexal procedures, and pelvic organ prolapse treatments with a short learning curve [3,9,10,11,12,13]. In addition, vNOTES has shown promising results in managing early-stage en-dometrial cancer, allowing complete surgical staging, including sentinel lymph node bi-opsies, lymphadenectomies, and omentectomies [11,14,15,16,17]. However, although increasing evidence supports the use of vNOTES approaches to manage early-stage endometrial can-cer and to perform benign adnexal surgeries, little is known about the feasibility and safety of performing vNOTES oncological staging for tubo-ovarian malignancies [15,18,19,20].

Hereby, we report our initial experience performing vNOTES surgical staging for ear-ly-stage ovarian cancer and BOTs.

2. Materials and Methods

2.1. Patient Selection, Data Collection, and Methods

vNOTES was implemented in our institution in May 2020. Since January 2022, we have collected retrospectively and prospectively data concerning patients who underwent vNOTES procedures to create an institutional database using the Research Electronic Data Capture (REDCap) software. The project received approval from the local ethical committee (CER-VD), with registration number 2021-02346, and all patients gave written informed consent.

From this database, we retrospectively identified and analyzed data from all patients with BOTs or early-stage ovarian cancer diagnosed between May 2020 and August 2024. Demographic features, as well as clinical and perioperative information, were collected and analyzed. Intraoperative parameters included total operative time (from catheteriza-tion of the bladder to vaginal closure), vNOTES port insertion time (from incision to in-trabdominal CO2 insufflation), estimated blood loss, intraoperative complications (in-cluding transfusion-requiring bleeding or iatrogenic organ injury), and the necessity for conversion to conventional laparoscopy or laparotomy. Postoperative assessments com-prised pain evaluation using the visual analog scale graded from 0 to 10 at 12-, 24-, and 48-hours post-surgery, opioid analgesic use, duration of hospital stay, and postoperative complications within 8 postoperative weeks, graded according to the Clavien-Dindo clas-sification (CD) [21]. In addition, we recorded histopathological results, the timing and type of any adjuvant therapies, and the patient status at the last follow-up.

Continuous variables were expressed as median and range, while dichotomous var-iables were represented as absolute numbers and percentages (%). No statistical inter-group comparisons were undertaken. Statistical analyses were performed using IBM SPSS version 29.0.2.0.

2.2. Surgical Technique

All interventions were performed by the same oncogynecological surgeon (DH). Pa-tients received a single dose of clindamycin vaginal cream 2% (5 g of cream with 100 mg of clindamycin) the day before the surgery, and 2-4 hours before the intervention, in addition to cefuroxime 1.5 g (3 g for patients weighing more than 80 kg) and metronidazole 500 mg intravenously at induction of anesthesia. Under general anesthesia and muscular relaxa-tion, patients were positioned in a horizontal dorsal lithotomy position, and a bladder catheter was placed.

Access was gained with a posterior 2 cm colpotomy through Douglas’s pouch to per-form interventions limited to the adnexa. If hysterectomies were performed, access to the abdominal cavity was achieved through anterior and posterior colpotomies, with the transvaginal uterosacral ligaments section when developing the posterior access. A vNOTES port (GelPoint vPath, Applied Medical, Rancho Santa Margarita, CA, USA) with an adapted diameter (7 cm for adnexectomies and 9.5 cm for hysterectomies) was placed in the abdominal cavity through the anterior and/or posterior colpotomies. Carbon dioxide was insufflated to create a pneumoperitoneum with an intraperitoneal pressure of 8-15 mmHg. Three 10 mm trocars were used to insert a 10-mm rigid 30° scope, 5-mm instru-ments such as Johan and bipolar graspers, and sealing devices. If necessary, a 4th 15 mm supplementary trocar was added.

Surgical staging included peritoneal washing, uni- or bilateral salpin-go-oophorectomy, abdominal cavity inspection, peritoneal biopsies, infracolic omentecto-my, and total hysterectomy. In selected cases, fertility-sparing approaches with unilateral salpingo-oophorectomy or cystectomy and uterus preservation were performed. To per-form hysterectomies, the uterine vessels, broad ligaments, and round ligaments were sealed and cut from caudal to cranial. Salpingo-oophorectomies were always performed after correctly visualizing the ureters, the fallopian tubes, and the infundibulopelvic liga-ments, with utmost care to avoid spillage.

All specimens have been extracted vaginally. To avoid intraabdominal spillage, large adnexal lesions were retrieved into an Inzii Endobag Retrieval System of 10 or 15 cm of diameter, or Alexis Contained Extraction System of 14 or 17 cm (Applied Medical, Rancho Santa Margarita, CA, USA). Intraoperative frozen section analysis was performed in cases with suspicious adnexal masses. Omentectomies were performed with an articulating sealing device, as we previously described [15].

At the end of the procedure, the colpotomy was closed under direct visualization us-ing a running suture with Vicryl 0. Postoperatively, patients received a single dose of clindamycin vaginal cream 2% (5 g of cream with 100 mg of clindamycin) once a day dur-ing the first 7 postoperative days.

3. Results

From October 2021 to August 2024, 7 patients with early-stage tubal or ovarian cancer and 4 patients with BOTs underwent surgical staging by vNOTES at Valais Hospital (Sion, Switzerland).

The median age was 47 (27 - 81) years, with a median body mass index of 28.1 (22.4 – 39.2) Kg/m2. Seven patients (63.6%) were classified as American Society of Anesthesiolo-gists score (ASA) II and 4 (36.4%) as ASA III. No patients showed any evidence of ad-vanced ovarian oncological disease at the preoperative workup, which included a pelvic ultrasound and a thoracoabdominal computed tomography. A pelvic magnetic resonance imaging was also performed when further ovarian lesion characterization was necessary according to ESGO recommendation [22]. Table 1 provides an overview of patient char-acteristics and their perioperative outcomes.

Bilateral salpingo-oophorectomy was performed in 5 patients (45.5%), while 6 pa-tients (54.5%) underwent fertility-sparing surgery with servation of at least one ovary and the uterus. Table 2 summarizes the surgical procedures performed to complete surgical staging. The median operating time was 70 (35 – 138) minutes, with a median blood loss of 50 (10 – 100) ml. No conversion to conventional laparoscopy or laparotomy was necessary, and all procedures were performed as planned. In a patient with suspicious pelvic im-plants, a hybrid approach was used to explore the utero-vesical peritoneum. All surgical material was extracted vaginally with an endobag. No intraoperative complications were reported, except for one case involving a minimal pelvic ovarian spillage during extraction in the retrieval system (9.1%) (Table 2).

Post-operative complications were reported in 3 patients (27.3%). These included one surgical site infection (9.1%) and 2 cases of cystitis (18.2%). All 3 postoperative complica-tions were graded as CD grade 2 and treated with antibiotics. The median hospital stay was 48 (24 – 96) hours. After the final histological results, 4 patients underwent a second intervention to complete the surgical staging, 3 by vNOTES and one by conventional lap-aroscopy.

Adjuvant chemotherapy was administered in 4 patients (36.4%), one of whom re-ceived palliative chemotherapy for relapsed pancreatic disease. The median time from surgery to adjuvant therapy was 23 (19 – 31) days. In this series, no evidence of recurrence was observed, with a median follow-up time of 11.9 (4 – 37.4) months. One patient (9.1%) died of metastatic pancreatic cancer one year after the surgery. The final histopathological diagnoses are summarized for each patient in Table 1.

4. Discussion

The role of MIS in gynecological oncology has undergone progressive development. This has involved introducing both conventional and robot-assisted laparoscopic tech-niques, which have demonstrated their feasibility and efficacy in staging and treating uterus-confined endometrial cancer [14,23,24]. In the case of early-stage ovarian cancer, the latest international guidelines maintain that the standard procedure for the treatment and staging of ovarian cancer is midline laparotomy. The rationale behind this is that the open procedure offers more accurate abdominal exploration and a reduced risk of rupture of the primary tumor. Nevertheless, the laparoscopic approach is frequently used world-wide for BOTs and early-stage ovarian cancer, and some studies have shown better surgi-cal outcomes and no difference in recurrence rates or survival for those who received minimally invasive versus open surgical staging [25,26,27,28,29]. However, the oncologic out-comes remain a topic of debate, lacking sufficient high-quality evidence to change current guidelines [2,6,29,30]. To date, only a few publications report a vNOTES approach in the management of ovarian cancer [15,18,19].

According to the current guidelines of the European Society of Gynaecological On-cology (ESGO), surgical management of Stage I to II ovarian cancer must include a total hysterectomy and bilateral salpingo-oophorectomy or fertility-sparing surgery (unilateral salpingo-oophorectomy) in selected patients desiring fertility. Peritoneal washings or cy-tology, taken before manipulation of the tumor, and peritoneal biopsies with at least in-fracolic omentectomy are also recommended [31]. Since omentectomy via vNOTES has been proven to be feasible [15,19], in the case of intraoperative diagnosis of BOTs or ear-ly-stage ovarian cancer, surgical staging through the same vaginal incision is possible. In our series, 4 out of 11 patients have an intraoperative BOT diagnosis, and complete peri-toneal staging was successfully performed by vNOTES.

A further challenge of the MIS approaches in early-stage ovarian cancers is the ability to perform a complete pelvic and paraaortic lymphadenectomy. However, many ear-ly-stage ovarian cancer diagnoses are made postoperatively on lesions initially presumed benign [32]. The accuracy of frozen section varies between 82 and 88% for BOTs and ma-lignant tumor with most discordancy encountered for younger, premenopausal women, early-stage ovarian malignancies and mucinous histology [33]. In our series, we identified one case of serous tubal intraepithelial carcinoma (STIC) during surgery for endometrioid endometrial carcinoma, along with four borderline ovarian tumors (BOTs) identified in-traoperatively through frozen section analysis. The remaining six cases were diagnosed postoperatively, with one diagnosis made prior to referral to our institution. Among the seven ovarian cancer patients, three had non-epithelial histologies, including immature teratoma and adult granulosa cell tumor, while one patient presented with invasive mu-cinous histology. For clinical stage I and low-risk invasive ovarian tumors—such as mu-cinous, malignant germ cell, and sex cord-stromal tumors—as well as for BOTs and STIC, systematic lymphadenectomy is not recommended [34]. The survival benefit of complete staging with lymphadenectomy in early-stage epithelial ovarian cancer has not been con-firmed in prospective trials [32], though it is known that 10-15% of cases are upstaged due to positive nodal involvement [35]. For the last two patients the final histology showed a stage IIA high-grade tubal carcinoma and a stage IIB low-grade serous ovarian carcinoma and the multidisciplinary tumor board decided against lymphadenectomy as no enlarged nodes were present and there was not impact on the adjuvant treatments allocation. We performed a subsequent retroperitoneal pelvic and paraaortic lymphadenectomy for a stage IA mucinous invasive carcinoma due to a minimal infiltrative invasive component in a majoritarian expansile tumor.

We hypothesize that both the vNOTES technique for pelvic lymphadenectomy and paraaortic lymphadenectomy can be successfully applied to early-stage ovarian malig-nancies. The vNOTES approach for pelvic lymphadenectomy was first described in 2014, with subsequent validation by other authors [36,37,38]. Additionally, in 2024, a hybrid tech-nique combining vNOTES with a single-port retroperitoneal approach for pelvic and in-frarenal paraaortic lymphadenectomy was reported [39]. For patients diagnosed in-traoperatively with early-stage invasive ovarian cancer requiring both pelvic and paraaor-tic lymphadenectomy, a vNOTES hybrid approach, may be an option [39,40,41]. If restaging is required, the absence of an abdominal peritoneal scar after retroperitoneal vNOTES can simplify the subsequent procedure.

One limitation of VNOTES staging is the restricted accessibility of certain anatomical regions, including the prevesical peritoneum, the posterior costodiaphragmatic recesses, the Morrison’s pouch, the lesser omentum, the omental foramen and bursa and the mesen-teric root. Nevertheless, some of these regions are challenging to examine also by conven-tional laparoscopy. Ghezzi et al. emphasized that isolated metastases in these specific ar-eas are extremely rare [42]. Despite these limitations, several studies have shown no sig-nificant differences in surgical outcomes, recurrence rates or survival between patients undergoing minimally invasive versus open surgical staging for patients with early-stage ovarian cancer [6,8,25].

The duration of vNOTES and standard laparoscopic procedures for early stages of adnexal malignancies seems equivalent. Data in the literature is heterogeneous, compar-ing the time of the open versus MIS approach with no clear advantage for one or another. The surgeons’ experience may be the main factor influencing the operating time [6,7,30,43]. The blood loss reported in our series is low and consistent with existing literature [6,7,30,43].

In our series, perioperative complication rates were low. No intraoperative complica-tions were noted, except for one case of minimal ovarian spillage during adnexal extrac-tion in endobag (9,1%). This is crucial as spillage can lower survival rates, and is associ-ated with an upstaging of the tumor [43,44,45]. Some studies suggest a higher risk of cyst rupture with laparoscopic cystectomy, which may be reduced if adnexectomy is performed rather than cystectomy [46]. Evidence on spillage risk with vNOTES is scarce but appears similar to the laparoscopic approach [3,47,48]. Tumor spillage might occur even in large laparotomies, raising the possibility that aggressive biology associated with more adher-ent and fragile tumors may be responsible for rupture more than the surgical ap-proach [32].

Lower rates of all types of complications have been reported with the vNOTES ap-proach in benign indications, ranging from 2.5% to 4.1% [49,50,51]. Laparoscopy has been demonstrated to significantly reduce the duration of hospitalization compared to lapa-rotomy [52], a finding consistent with our series, which had a median hospital stay of 48 (24 – 96) hours. Postoperative complications related to surgery were minimal in our cohort, with one case surgical site infection (9.1%) and 2 cases of cystitis (18.2%), all successfully treated with antibiotics.

In the vNOTES approach, the single vaginal scar can improve the rapid post-operative recovery. Fast recovery is particularly important for the management of ma-lignant cases, allowing for the earlier administration of adjuvant treatments. In our series, the median time from surgery to adjuvant therapy was 23 (19 – 31) days. Present recom-mendations endorse starting adjuvant treatments 28 to 42 days after the surgery [53,54].

Abdominal port site metastases has been an important concern in MIS for in-tra-abdominal malignancies. The vNOTES approach offers an advantage, particularly in patients requiring hysterectomy, by eliminating the need for additional abdominal inci-sions. For adnexal surgery, vNOTES limits this concern by the presence of a single vaginal incision. Furthermore, vNOTES allows extraction of masses of up to 6-7 cm without the need for morcelation or puncture. For bigger sizes, adnexal mass extraction after puncture in surgical bags is possible with the same approaches as with conventional laparoscopy [3]. Baekelandt et al. have recently described a technique for bagging a 20 cm BOTs via vNOTES without spillage [20].

We acknowledge limitations in this study, notably the small sample size, which limits comparative analysis with other methods. Additionally, case heterogeneity and short fol-low-up as well as single-center setting with only one oncogynecological surgeon reduce generalizability. However, our focus on early surgical outcomes supports the feasibility of vNOTES for highly selected early-stage ovarian cancer patients, with potential benefits in perioperative morbidity and quality of life.

5. Conclusions

Despite the limited cohort size, our findings indicate the technical feasibility of vNOTES for highly selected patients with early-stage ovarian cancer. Further research with larger cohorts and extended follow-up is needed to assess the long-term oncological outcomes and safety of this technique.