Adoption of vaginally assisted natural orifice transluminal endoscopic surgery for hysterectomy: A single tertiary experience.

The median age of the first 20 participants was 51.5 years (47–57 years) and the median BMI was 33.5 kg/m 2 (27.8–38.3 kg/m 2 ). The predominant indication for hysterectomy was complex hyperplasia with atypia (12/20, 60%) followed by abnormal uterine bleeding (5/20, 25%) and only one case (5%) was for prolapse with a concurrent ovarian mass (Table  1 ). The median ASA score was 2 (1–3) and median parity was two (1–3) with four patients who were nulliparous. Sequential attempted VANH case demographics and perioperative outcomes Abbreviations: AUB, abnormal uterine bleeding; BS, bilateral salpingectomy; BSO, bilateral salpingo‐oophorectomy; CAH, complex hyperplasia with atypia; CIN, cervical intraepithelial neoplasia; EBL, estimated blood loss; LOS, length of stay; VANH, vaginally assisted natural orifice transluminal endoscopic surgery for hysterectomy. The median blood loss was 125 mL (100–200 mL) with an operative time of 149 min (138–198 min). The median weight of the specimen was 181.5 g (66.5–219 g). The mean length of stay was 1.4 days (1–2 days). Only 60% of cases (12/20) had preoperative assessment for suitability for vaginal hysterectomy documented by other clinicians; of these, 50% (6/12) had been reported as not suitable for vaginal hysterectomy. One case, underwent VANH bilateral salpingo‐oophorectomy (case 7) for complex hyperplasia with atypia, had endometroid adenocarcinoma (Stage 1A) requiring only clinical follow up. The procedures were converted and completed laparoscopically in five cases (25%). The majority occurred within the first ten cases (cases 2, 4, 7, 10 and 18) performed. Reasons for conversion included suboptimal vision due to large bowel compromising visibility and access (cases 2, 4, 10), unexpected endometriosis with obliterated side wall with adhesions to adnexa and sigmoid colon (case 7), large bowel adhesions to adnexa and vascular pedicle in a patient with previous laparotomy for a benign ovarian cyst on the same side (case 18). Four of the five converted cases had a BMI in the morbidly obese range (35–40 kg/m 2 ) while the other was converted (BMI 28 kg/m 2 ) due to sidewall and adnexal adhesions. Of the five converted cases, three did not have suitability for vaginal hysterectomy recorded by other clinicians and two cases were recorded to have no descent. Four of the six (67%) cases who were deemed unsuitable for vaginal hysterectomy were successfully completed as VANH. There was one case (case 17) of obliterated POD from unexpected endometriosis in a post‐menopausal patient requiring hysterectomy for complex hyperplasia with atypia which was completed using the VANH approach. In cases 16 and 20 there was evidence of unexpected superficial endometriosis and significant adhesions involving the adnexa; however, the cases were completed utilising the VANH technique. In all cases where unexpected endometriosis was detected, adhesions were divided to restore normal anatomy to facilitate hysterectomy with some ergonomic challenges due to the single site nature of the procedure. In all procedures the uterine specimen was removed en bloc with adnexa. Of these, 65% (13/20) had concurrent bilateral adnexectomy and 35% (7/20) had concurrent bilateral salpingectomy. There was only one (CD type 1) intra‐operative complication (Case 7) where suture material was visible underneath the bladder mucosa at the time of cystoscopy. While not intravesical, the vault sutures were removed, and the vault re‐repaired with a catheter left in for seven days as a precaution. Subsequent retrograde cystogram was normal. There was one post‐operative complication (CD type 1) where the patient (Case 9) presented 12 days postoperatively to another tertiary hospital with per vaginal bleeding. Investigations including imaging were unremarkable and she was managed conservatively as an outpatient.

This is the first reported feasibility study at an Australian tertiary unit demonstrating a novel proctoring model for adoption of VANH. Our experience is similar to that of other international units apart from the longer operative time and much higher rate of conversion to laparoscopy, especially in the initial ten cases. 15 , 16 , 17 This may reflect the early part of the learning curve of the primary surgeon or case selection as most patients were older with high BMI and less uterine descent than generally considered adequate for vaginal hysterectomy. A single surgeon study on the learning curve for adopting VANH ( n  = 79) demonstrates reducing operative time over the four unique phases (phase I, 113.8 min to phase IV, 82.8 min) and suggested 20 cases to achieve competence. 18 This trend is also noted in this case series where the median operative time of the first ten cases was 170 min which decreased to 145 min in the subsequent ten cases. There is limited data on VANH in obese women, with one study ( n  = 200) reporting increased mean operative times noted in the obese cohort (> = 30 kg/m 2 , 115 min) versus the non‐obese cohort (<30 kg/m 2 , 79.8 min). 11 While there was no formal guidance on patient selection and adoption when VANH was first offered (2019), there is now an expert‐based consensus paper which recommends the initial VANH procedures be on patients with a small uterus (<11 cm), parous with no previous caesarean sections and low BMI (< 30 kg/m 2 ). 7 The higher‐than‐expected conversions may be due to patient complexity and the known learning curve. There were notably fewer conversions in the subsequent ten cases, potentially due to increased confidence in optimising visibility during the procedure, managing unexpected adhesions intraoperatively and overcoming ergonomic challenges that are expected due to the single site nature of the procedure. The availability of the GelPOINT V‐Path (Applied Medical) in the subsequent cases which is appropriate for vaginal surgery, may have eased technical challenges due to a suboptimal port. The absence of a real‐time proctor may have also contributed to a lower threshold to convert and complete the procedure by laparoscopy which is the primary surgeon's default approach to hysterectomy. There was an increase in use of sponge packs following port insertion to help retract the bowel and improve visibility during the endoscopic phase of the procedure, which helped overcome previously noted limitations in visibility (Figs.  1 and 2 ). The use of a transcervical instrument to provide uterine manipulation, freeing up an arm to provide additional retraction was described in a video case article and may be beneficial if widely adopted in the future. 19 While VANH can overcome some of the perceived challenges encountered with traditional vaginal surgery, it is still subject to the limitations associated with endoscopy (inability to tolerate Trendelenburg, suboptimal pneumoperitoneum/pneumovagina, suboptimal vision due to bowel). Vaginally assisted natural orifice transluminal endoscopic surgery for hysterectomy for a 55 year old (body mass index 49) without sponge packs. C, cervix; O, ovary; S, sigmoid colon; U, uterus. Vaginally assisted natural orifice transluminal endoscopic surgery for hysterectomy for 55 year old (body mass index 49) with sponge packs to retract bowel. U, uterus; O, ovary; OL, ovarian ligament; P, pack. There were two cases of (grade 1 CD) intra‐operative and post‐operative complications. However, the small numbers do not allow for meaningful conclusions to be made on the safety profile of VANH. Re‐siting a submucosal bladder suture does not generally require further management; however, in the setting of introducing a new technique, this was thought to be wise. Routine vaginal and rectal examination pre‐operatively is not widely advocated; however, it is recommended to help avoid unexpected POD adhesions. 13 While in this case series one case had an unexpected obliterated POD, in the largest single surgeon case series of VANH ( N  = 750), there was a 0.4% conversion rate where three of the four cases were due to pelvic adhesions and endometriosis. 8 Such a low conversion rate suggests that history and clinical examination are sensitive in picking up most relative contraindications to vNOTES procedures. There is a clinical trial ( NCT05623514 ) which is currently recruiting patients to evaluate the use of ultrasound preoperative imaging to improve patient selection for vNOTES. The case of VANH with unexpected stage 4 endometriosis was completed, as VANH and has been described in other case reports. However, until robust evidence is available, endometriosis is considered a relative contraindication when the primary indication is for excision of endometriosis. 7 , 20 , 21 Vaginal hysterectomy is the least invasive, cost‐effective and optimal approach to hysterectomy, yet the current generation of gynaecologists and trainees have declining exposure and therefore confidence in offering VH to patients despite the obvious clinical benefits. 1 , 2 , 22 , 23 , 24 A survey of 376 gynaecologists preferred VH (55.5%) as a mode of hysterectomy for themselves or their spouse, citing technical difficulty, potential for complications and caseload of VH as potential barriers to more widespread implementation. 2 The adoption of VANH potentially may increase the exposure of trainees to VH as seen in a study of 1458 cases where there was an increase in vaginal route for hysterectomy from 40.1% to 94.3% ( P  < 0.001) and resident participation with VH (58 to 131 cases per year). 25 It may also enable concurrent adnexal procedures to be addressed, although many experienced vaginal surgeons would be able to achieve adnexectomy and or salpingectomy vaginally. 26 , 27 , 28 VANH may also allow en bloc removal of larger uteri than can be removed via VH. This is particularly relevant in situations where the patient declines morcellation and where debulking of the specimen is inappropriate, as is the case when hysterectomy is performed for complex hyperplasia with atypia. This was particularly relevant for our institution, given the majority of the first 20 cases had a known or suspected premalignant or malignant pathology potential. VANH may allow easier access to transect the uterine blood supply earlier on in the procedure than LH where it can be challenging navigating around a large uterus. 910,29 VANH may also be a useful approach in instances where significant abdominal adhesions or alterations to normal abdominal anatomy are expected from previous surgery. Further prospective research is required to determine the exact clinical benefit and role of VANH as compared to other mainstream modes of hysterectomy from a patient and clinician perspective. Until such sound large‐scale data are available, iNOTESs recommends the use of the vNOTES surgical registry to prospectively capture the outcomes of VANH and promptly identify emerging trends as and when they occur. As VANH is more widely used, a specific surgical code will be necessary to track the rate of uptake and monitor for outcomes at a national level. VANH is a feasible option for hysterectomy for benign gynaecological indications where a surgeon is competent and experienced in both vaginal and laparoscopic surgery. There is a learning curve to achieve competence in this technique; therefore, adequate training and mentoring with careful case selection is recommended. Until further robust data are available to determine the clinical benefit and safety profile of VANH, patients should be carefully counselled and the decision on mode of hysterectomy be individualised.

Vaginal hysterectomy (VH) has long been considered the optimal minimally invasive option for hysterectomy and is associated with reduced operative time and length of stay compared with laparoscopic hysterectomy (LH). 1 However, there has been a notable decline in the rates of VH over the last few decades, possibly due to increasing patient comorbidities and complexities which limit the utility of VH, the popularity of laparoscopic techniques and the subsequent reduced exposure and lack of training. 2 , 3 , 4 Vaginal natural orifice transluminal endoscopic surgery (vNOTES) combines the advantages of vaginal approach together with the magnification and visibility from endoscopy. Vaginal assisted natural orifice transluminal endoscopic surgery hysterectomy (VANH) was first described in 2012 in a series of 16 non‐obese patients where feasibility of this novel technique was demonstrated. The first single surgeon non‐inferiority randomised controlled trial (RCT) comparing VANH ( n  = 35) to LH ( n  = 35) demonstrated that VANH was not inferior and also associated with shorter surgery, reduced length of stay and analgesia requirements in the first seven days. 5 The same benefits of VANH ( n  = 288) over LH ( n  = 430) have been reflected in a recent systematic review based on pooled analysis of one RCT and five retrospective cohort studies. 6 Given the absence of substantial robust long‐term evidence, an international expert consensus was put together to help guide clinicians adopting vNOTES into clinical practice. 7 The largest case series of VANH ( n  = 750) reported by a single surgeon demonstrated comparable perioperative outcomes to LH as described in the literature with 1.1% rate of cystotomy. 8 VANH has been reported in women with higher body mass index (BMI), larger uterine size, with lack of good uterine descent which are considered challenges for routine VH. 5 , 9 , 10 , 11 The aim of this study is to report our early experience and feasibility of adoption of VANH at a single tertiary Australian hospital.

A retrospective review was performed to determine the feasibility and perioperative outcomes of the first 20 cases of VANH. Ethics approval was provided by the Western Sydney Local Health District for a Quality Assurance project (2309‐10). The primary surgeon had extensive experience in LH, including training in laparoscopic single site surgery. Specific vNOTES training included two international courses (Belgium and Thailand). Initial experience involved observation of 12 VANH cases by expert international vNOTES surgeons with two cases performed under supervision. To acquire vNOTES certification as endorsed by the International NOTES Society (iNOTESs), the first ten cases had to be proctored by a vNOTES expert surgeon. Due to the lack of a local national expert, an international vNOTES expert surgeon (who had individually completed over 1000 vNOTES cases at the time) remotely proctored the primary surgeon. VANH procedures commenced in 2019. Given the introduction of a novel procedure, all cases had to be prospectively recorded and reported at a unit level for complications. Potential cases were planned in consultation with a local urogynaecologist experienced in vaginal surgery and both surgeons were present for the first 20 cases. The target of 20 cases was determined arbitrarily, deemed to be a more cautious approach than the minimum requirement of ten cases as required by iNOTESs for certification. At the time of implementation, there was no other evidence‐based guidance on the learning curve for integrating VANH into practice. The VANH procedural steps and perioperative care were in line with the ten‐step approach described and taught in the standardised vNOTES courses. 12 The steps include: circumcision of the cervix following infiltration with local anaesthetic and adrenaline posterior colpotomy anterior colpotomy transection of the uterosacral ligaments preparation and placement of vNOTES port identification of the ureter and transection of the parametrium (including uterine artery) transection of the infundibulopelvic or ovarian ligament haemostasis and port removal specimen removal vault closure circumcision of the cervix following infiltration with local anaesthetic and adrenaline posterior colpotomy anterior colpotomy transection of the uterosacral ligaments preparation and placement of vNOTES port identification of the ureter and transection of the parametrium (including uterine artery) transection of the infundibulopelvic or ovarian ligament haemostasis and port removal specimen removal vault closure Patients who required hysterectomy for benign gynaecological indications and were suitable for a laparoscopic assisted vaginal hysterectomy (LAVH) were selected, counselled and consented for VANH. The procedure was booked as a LAVH, and the patient consented to having the procedure completed vaginally where feasible. All patients had a pelvic ultrasound, preoperative vaginal and rectal examination to assess for obliteration or adhesions in the pouch of Douglas (POD) by the operating surgeon. 13 To consider a patient suitable for VANH, vaginal access to POD was considered a pre‐requisite by the operating surgeon. Patients with a large uterus requiring possible morcellation (as determined by pelvic ultrasound (US) and clinical examination to determine the likelihood for en bloc vaginal specimen removal) or any previous caesarean sections were excluded along with other reported relative contraindications to vNOTES procedures (rectovaginal endometriosis, prior rectal surgery, history of severe pelvic inflammatory disease, previous mesh sacrocolpopexy, pelvic radiotherapy). 7 Vaginal and endoscopic components were performed using typical instruments used for vaginal and laparoscopic hysterectomy respectively. A vaginal port was the only additional equipment required to perform VANH. The initial cases were performed using Applied Medical GelPOINT® Advanced and Mini Advanced access ports (off‐label) in the vagina, and patients were counselled accordingly. The subsequent cases were performed using GelPOINT® V‐Path medium (Applied Medical) which was approved by Therapeutic Goods Administration for use in Australia and available later in 2019. All patients had a cystoscopy on completion of the procedure along with a vaginal pack. Both were removed within 24 h. Following discharge, the patients were reviewed within six weeks post‐procedure. 7 Patients who underwent VANH were recorded prospectively according to surgical date due to department audit requirements for a new surgical technique. There were a few variables initially not collected that were subsequently deemed relevant and collected through a formal retrospective review of the 20 cases. The variables collected from electronic medical records included patient age, BMI, indications for surgery, parity, number of previous vaginal deliveries, American Society of Anaesthesiologists (ASA) score (1–4), preoperative assessment of descent (by other clinicians), duration of procedure (time from surgical prep to vaginal pack), conversion, estimated blood loss (EBL in mL), length of stay (days), specimen weight (post‐fixation with formalin in grams), and complications were defined according to Clavien–Dindo classification system (CD). 14 Patients were sequentially analysed in order of surgical date. Descriptive statistics are used to report patient demographics and baseline perioperative outcomes with continuous variables reported as median and interquartile range for non‐normally distributed data. Dichotomous and non‐continuous variables are reported as percentages.