Analysis and suggestions on the complications in 2000 cases of transvaginal natural orifice transluminal endoscopic surgery: Can it be a conventional surgery?

To reflect on the complications of transvaginal natural orifice transluminal endoscopic surgery (vNOTES), identify the corresponding risk factors, and provide caution to surgeons when performing this novel surgery.


| INTRODUC TI ON
Advancement in medical field and cosmetic requirements of patients have prompted doctors to constantly innovate in the transformation of minimally invasive surgery.The gynecologic surgical approach has advanced from open surgery to laparoscopy, natural orifice transluminal endoscopic surgery (NOTES), and transvaginal NOTES (vNOTES).
8][19] Moreover, the possibility of underreporting complications cannot be denied when a researcher evaluates his own surgical technique.5][26][27] Therefore, the present study aimed to fully and objectively reflect on the complications of vNOTES, identify the corresponding risk factors, and provide caution to surgeons when performing these novel surgeries.

| Study design and participants
The present study is part of the Longitudinal Vaginal Natural Orifice Transluminal Endoscopic Surgery Study (LovNOTESS), an ongoing study on gynecologic minimally invasive surgeries conducted in Chengdu, aiming to determine their short-and long-term complications, as well as the potential effects of vNOTES on patients'

| Data collection
The perioperative data of patients were retrospectively collected by researchers through inpatient and outpatient electronic medical record systems.The patient information was collected from the hospital database by gynecologists other than these surgeons.
The data included patient's age, maximum diameter and location of the myoma, body mass index (calculated as weight in kilograms divided by the square of height in meters), operative position, history of previous gravidity and abdominal surgery, blood loss (assessed by subjective visual quantification), total operation time (from skin incision to closure), intraoperative complications (injuries to the bladder, bowel, and vessels), simultaneous surgery, postoperative serum hemoglobin drop, conversion to another surgical procedure, postoperative fever (oral temperature ≥ 38.0°C in 24 or more hours postoperatively), hospital stay, time of first postoperative flatus, and postoperative complication (scaled using the modified Clavien-Dindo complication rating) during the 1-month period after surgery.These data ranged from those of adnexal surgery to uterine surgery, benign to malignant tumors, and even pelvic floor reconstruction.
Perioperative complications were stratified in chronologic order and specifically divided into those developed during the establishment process of the vNOTES approach, during surgery, postoperatively, and during 1-month follow up.To check for postoperative recovery and complete the clinical data, all patients underwent an outpatient review 1 month after surgery.

| Modified classifications of Clavien-Dindo complications for vNOTES
Previous Clavien-Dindo complication ratings focused only on postoperative complications.The scoring criteria for surgical complications have been modified and described in detail in our hospital, which included some intraoperative complications and modified classification of some postoperative complications.If the intraoperative blood loss was more than 500 mL, regardless of blood transfusion, considering the increased risk of infection, low postoperative immunity, and increased antibiotic dosage and duration, we classified it as a grade II complication.Patients with postoperative lower extremity venous thrombosis were classified as having grade III complications because of the possibility of serious complications such as hemorrhagic shock and pulmonary embolism (Table 1).

| Standardized surgical approach
Before surgery, a sodium phosphate oral bowel preparation was used considering adhesions or the high risk of intestinal injury.Except for a few patients who underwent myomectomy in the prone position (described in detail in our previous study), 28 the remaining patients were in the Trendelenburg position.Cefazolin (1 g) was routinely administered intravenously half an hour before surgery to prevent infection.The perineum and vagina were disinfected with iodophor and a catheter was placed before the operation.
In patients who underwent hysterectomy, the vaginal wall was cut in a circular manner along the cervicovaginal junction to the cephalic side approximately 0.5 cm.The peritoneum was then opened at the pouch of Douglas to enter the pelvic cavity.The bilateral uterosacral ligaments were severed.The peritoneum at the bladder reflex was opened after the bladder was pushed up from the anterior fornix.A special port for vNOTES was installed.Other procedures only require opening the posterior vaginal vault or anterior fornix.
After entering the abdominal cavity, the surgical procedure was the same as that used for laparoscopy.

| Statistical analysis
All statistical analyses were performed using SPSS version 25.0 (IBM Corp., Armonk, NY, USA).Mean ± standard deviation was used for enumeration data.The measurement data are expressed as centiles.All statistical significance was set at P less than 0.05 with two-tailed tests.
The Clavien-Dindo surgical complication classification system is shown in detail in Table 1: if the intraoperative blood loss was TA B L E 1 Modified classifications of Clavien-Dindo complications for vNOTES in our hospital.

Grades Definition
Grade I Any deviation from the normal postoperative course without the need for pharmacologic treatment or surgical, endoscopic and radiologic interventions.Acceptable therapeutic regimens are drugs as antiemetics, antipyretics, analgesics, diuretics and electrolytes, and physiotherapy.

Grade II
Requiring pharmacologic treatment with drugs other than such allowed for grade I complications.
Poor wound healing treated at the bedside, re-catheterization due to urinary retention, gastrointestinal decompression, intraoperative bleeding volume > 500 mL, blood transfusions, and parenteral nutrition are also included.
Grade III Requiring surgical, endoscopic, or radiologic intervention.Intraoperative organ injury, postoperative venous thrombosis.
Grade III-a Intervention not under general anesthesia.Intraoperative organ injury, postoperative venous thrombosis.
Grade III-b Intervention under general anesthesia.
Grade IV Life-threatening complication (including central nervous system complications) a requiring intermediate care/ intensive care unit management.
Grade IV-a Single organ dysfunction (including dialysis).
Grade IV-b Multi-organ dysfunction.
Grade V Death of a patient.
Suffix "d" If the patient suffers from a complication at the time of discharge.The suffix "d" (for "disability") is added to the respective grade of complication.This label indicates the need for a follow up to fully evaluate the complication.
a Brain hemorrhage, ischemic stroke, subarachnoidal bleeding, but excluding transient ischemic attacks.

F I G U R E 1
The overview of the proportion of complications.
greater than 500 mL, regardless of blood transfusion, the com- Information regarding the operation conversion is provided in Table S2.Eight patients underwent conversion, including five cases of rectal injury repair.One patient's ectopic pregnancy lesion was found in the uterine horn, and the operation was completed using trans-umbilical laparoscopy because of difficulties in hemostasis. of the surgical approach is not essential.The five cases of intestinal injury, in our study, were repaired under vNOTES, and the postoperative healing was good.

| DISCUSS ION
Surgeons are constantly being trained in surgical skills and innovating instruments to effectively treat diseases and minimize trauma to patients.However, any new procedure, regardless of its advantages, is difficult to generalize if it has widespread and severe complications.The overall complication rate of gynecologic vNOTES was 4.4% in the present study, which was comparable with that of existing minimally invasive surgical techniques, such as laparoscopic 29,30 and robotic 31,32 surgeries.The differences in study design and sample size 28,29,33,34 or the author's subjective bias in the reporting process may have caused this difference among the previous studies.
However, the complication rate reported in the present study was still very low compared with that of previous studies under more stringent rating criteria.
Our modified Clavien-Dindo classifications were consistent with the findings of Katayama et al. 35 A more detailed grading standard would be beneficial for unifying the grades of intraoperative and postoperative complications.It would be convenient for follow-up studies to compare different surgical methods under a unified standard, making the study more objective and accurate.
Compared with LESS, the most distinctive feature of vNOTES is its innovative approach.Therefore, we divided complications during the operation into those during the approaching period (establishment of the vNOTES platform) and the intraoperative stage (after entering the abdominal cavity).Notably, only five cases (0.25%) of complications occurred during the approach in our study, but all were grade III rectal injuries.7][38][39] Deep endometriosis was the cause in all five cases.This is consistent with the 2021 vNOTES Consensus 40 recommendation that the presence of deep endometriosis, previous severe pelvic inflammatory disease, and a history of pelvic radiation therapy are contraindications for vNOTES. 24,41,424][45] In addition, preoperative transvaginal ultrasound assessment of the "sliding sign" is consistent with the auxiliary examination recommended by Zhang et al. 27 The last line of defense is to pull the vaginal wall back and forth at the posterior fornix before the incision.
If adhesion exists, there is a concave sign.In other words, the local vaginal wall appears pulled and unable to move.We believe that the combined use of these methods can effectively avoid complications caused by errors in patient selection.These complications are primarily associated with posterior vault adhesions.Our concept of anterior fornix adhesion during hysterectomy was consistent with that reported by Guan et al. [46][47][48] As long as the posterior fornix can be entered smoothly, the anterior fornix adhesion can be "aided" by the lateral approach.
The most common intraoperative complications were rectal injury in two cases (0.10%), as the result of severe adhesions, and hemorrhage in 28 cases (1.40%).No hidden damage was caused by the instruments in the blind area.In previous studies, intestinal injury, not observed during surgery, was a serious complication that led to patient death. 36,49The advantage of vNOTES is that the intestines are pushed out of the pelvis before the operation; hence, the instrument does not touch the intestines during the process.
Therefore, patients exhibit a faster recovery of intestinal function.
Urinary tract injury often occurs during traditional gynecologic laparoscopic surgery.0][51] However, none of our patients experienced ureteral injury.In addition to the great skill of the surgeons, there is a close relationship with the unique perspective of vNOTES.In vNOTES, the ureter is closer to the posterior cervical fornix and operative area, the visual field and ureter are parallel structures, and the fallopian tube is clearer in the visual field.Moreover, the relationship between the ureter and operating area changes from the anteroposterior relationship in LESS to the left and right relationship in vNOTES.Hence, the surgeon can more easily identify the position of the ureter when viewed from the caudal side.
In our study, 33 patients (1.65%) had hemorrhage, 28 (1.40%) had intraoperative bleeding greater than 500 mL, and 5 (0.25%) had postoperative pelvic hematoma.The hematoma may have been caused by inadequate suturing.This is a common characteristic of laparoscopic surgery.Tarik et al. 52 reported 0.5% bleeding complications during gynecologic laparoscopic surgery.No major vascular injuries were observed in any of our patients.It is worth pointing out that in one case of ectopic pregnancy, the lesion was in the isthmus of the fallopian tube.Owing to failure to reach the uterine fundus, the patient underwent trans-umbilical single-port laparoscopy for hemostasis.This reflects a limitation of vNOTES.It is difficult to manage lesions in the uterine fundus regardless of whether they are uterine fibroids, fallopian tube pregnancy in the isthmus, or interstitial. 9Although vNOTES has anatomical limitations, there were no puncture holes in the abdominal wall.Vascular puncture injury or subcutaneous emphysema caused by the trocars did not occur.
This is important because vascular injury is an extremely urgent and dangerous complication of laparoscopic surgery that could lead to death.
The most common postoperative complication was varying degrees of infection.There were 10 cases (0.5%) of fever, 11 (0.6%) of pelvic inflammatory disease requiring antibiotic treatment, 2 (0.1%) of pelvic abscess requiring incision or drainage, and 1 (0.05%) of septic shock.In addition, two cases (0.1%) of poor vaginal wound healing were considered to be caused by infection.The overall infection TA B L E 4 Suggestions for vNOTES.

Variables Suggestions a
Before operation 1. Medical history inquiry, if the patients have dysmenorrhea and previous pelvic adhesions, point 2 and point 3 assessments should be completed more carefully (Grade A) 2. Trimanual examination, if touching pain nodules of cervical posterior fornix and poor uterine activity, avoid performing vNOTES (Grade A) 3. Ultrasonic evaluation of posterior fornix activity, if the "sliding sign" is negative, avoid performing vNOTES (Grade B) rate was 1.3%.In a double-blind randomized controlled trial of gynecologic laparoscopic surgery, the infection rate at the surgical incision was 10.2%, and the infection rate at all surgical sites was 16.6% in 640 patients. 53Although the vNOTES approach involves a class II incision, after strict disinfection and prophylactic use of antibiotics, the complications of surgery-related infections do not increase. 54,55ter surgical subgroup analysis, we found that the majority of in- shown that for patients with a high risk of thrombosis, if the operation can be completed through the vagina in a short time, it is more conducive to reducing the risk of thrombosis. 56In the malignant tumor group, although there were only four cases of postoperative complications, all of which were bleeding without severe grade, vNOTES was feasible.However, the complication rate reached 16.67%, indicating that the safety of the aforementioned complex operations still requires a considerable amount of data for confirmation.Moreover, as our study was limited to 1 month of follow up after surgery, further follow up is needed for long-term complications, such as mesh exposure.
The rate of intraoperative conversion to other approaches is also one of the most important immediate judgment criteria for the safety and feasibility of novel procedures, such as vNOTES.
Transabdominal laparoscopy was performed in eight patients (0.4%) in our study.The main reason for this was the repair of intestinal injury in five cases (0.25%).Baekelandt et al. 23,57 reported a conversion rate of 4% in a study on vNOTES among 1000 cases of benign disease.Their data, similar to ours, confirmed that vNOTES is a safe and effective gynecologic procedure that can be performed with adequate preoperative evaluation.Our study included fewer patients with pelvic floor dysfunction and malignancy, which reduced the number of surgery-related complications.We believe that, with the subsequent development of complex surgical procedures, Objectively, this avoids the bias caused by anatomical differences between different surgical methods.Second, the study findings will help surgeons to focus on the effective performance and promotion of the novel technique; it may also help them to pay more attention to complications related to the establishment of the access platform.
In addition to using data to prove that the operation is safe and feasible, we shared our experience without reservation to provide effective help to surgeons who intend to perform this operation.Finally, owing to the large volume of case data, we can identify the problems in this operation and clarify the direction of future research; thus, promoting the long-term development of this new type of surgery.
The present study had several limitations.First, our data included a relatively limited number of patients with pelvic floor dysfunction and malignant tumors, reducing the number of related complications.We believe that with the continued development of complex surgical methods, more data will be obtained.Second, our study was retrospective, and we need to increase the demonstration of research data through prospective randomized controlled trials.
However, we believe that through continuous research, experience sexual function, pregnancy, and vaginal delivery (Chinese Clinical Trial Registry: ChiCTR2200059282 [https://www.chictr.org.cn/showp roj.aspx?proj=167900], approved by the Ethics Committee of the Chengdu Women's and Children's Central Hospital: No. 202130).The present study was conducted in accordance with the Declaration of Helsinki.During the study period, the total number of laparoscopic surgeries performed annually ranged from 3500 to 4000.In this subgroup study, clinical data from the database of the Chengdu Women and Children's Hospital between May 2019 and May 2022 were retrospectively analyzed.Written or oral informed consent was obtained from all participants.All patients who underwent gynecologic vNOTES during this period were included in the study, including those who were unable to complete the conversion to conventional laparoscopy.Patients with insufficient information and those who underwent trans-umbilical laparoscopic-assisted vaginal hysterectomy were excluded.The vNOTES procedures performed between April 2018 and April 2019 were excluded from the study because senior surgeons were on training to complete the learning curve during this period, and the operation had not reached a stable mature stage during this period.We analyzed the perioperative data of 2000 cases of vNOTES, which cover almost all types of gynecologic surgeries.

a
One case was converted to laparoscopy because of failure to expose the adnexa due to severe pelvic adhesions.The last patient could not complete the surgery because of the low position of the uterine myoma.Fibroids affected the establishment of the vNOTES platform.We divided our recommendations into three levels (A.Essential; B. Highly recommended, and C. Optional) before, during, and after surgery (Table4).Dysmenorrhea, previous history of pelvic adhesions, tender nodules, or poor uterine activity during gynecologic examination may indicate the presence of severe pelvic adhesions; therefore, these cases can be excluded for vNOTES.Preoperative ultrasound evaluation is also an important tool to exclude pelvic adhesions: first, the "sliding sign" between the posterior wall of the uterus and intestinal tube clearly distinguishes adhesions between the posterior wall of the uterus and the intestinal tube in the depression of Douglas; second, if the lesions are at the uterine fundus or close to the vaginal vault incision, vNOTES is not suitable because it causes difficulty in entering the pelvic cavity or is beyond the scope of surgical instruments.Antibiotics are recommended for routine preoperative prophylactic use because our previous study found that the risk of postoperative infection is higher with vNOTES than with LESS.Moreover, if adjacent organs such as the intestines are injured during the operation, conversion TA B L E 2 Description of the patients' characteristics.a Data are presented as mean ± standard deviation or as number (percentage).b Body mass index is calculated as weight in kilograms divided by the square of height in meters.

TA B L E 3
Description of the complication classifications by disease types.a Multiple complications in the same patient are classified as the highest grade.a Data are presented as number (percentage).

4 .
Preventive use of antibiotics half an hour before surgery and strict vaginal disinfection (Grade A) 5. Strictly screen the surgical indications, if the lesion is at the bottom of the uterus or the position of the lesion is too high, avoid performing vNOTES (Grade A) During operation Access process 1. Cervical forceps used to pull the cervix to ensure good cervical and uterine activity (Grade B) 2. During tissue separation, look for peritoneum in the direction of uterus to avoid damaging rectum (Grade B) 3. Timely change of direction when encountering adipose tissue and muscle layer (Grade A) 4. Sterile cloth sheet should effectively cover the perianal and other contaminated areas (especially in the prone position) (Grade B) Surgical process 1.If the operation cannot be completed due to limited surgical instruments scope, change the surgical approach promptly (Grade A) 2. If adjacent organs are damaged, change the operation approach according to the situation (Grade C) 3. Intraoperative bleeding volume >800 mL, if the operation cannot be completed in a short time, it is recommended to change the surgical approach (Grade C) After operation 1. Preventive use of antibiotics for 48 h postoperatively (Grade B) 2. Ultrasound exploration of pelvic cavity 1 month after operation (Grade C) 3. Strictly prohibit sexual life for 2 months (Grade B) a Grade A, Essential; Grade B, Highly recommended; Grade C, Optional.
fections occurred in the myomectomy group.The possible reasons are as follows: first, the operation time of myomectomy is longer than that of other types of surgery; second, the removal of giant uterine fibroids may penetrate the myometrium of the uterus, and incomplete suture of the uterus may be conducive to bacterial reproduction and infection; third, some uterine posterior wall myomas are operated on in the prone position.Repeated friction in the anal area after hook slipping may increase the probability of infection.The two patients with pelvic abscesses requiring a re-operative incision were in a prone position for myomectomy.This may be because it was difficult to expose the pouch of Douglas in a prone position for myomectomy (lifting the levator ani muscle is difficult), which led to slipping of the retractor and repeated contact with the contaminated area around the anus.Further large-sample randomized controlled trials trials are needed to confirm this hypothesis.Lower limb vein thrombosis was observed in the pelvic floor dysfunction subgroup.The main reason for this is that the patient was older and had a high risk of developing thrombosis.The mean postoperative thrombus score was 3. Deep vein thrombosis still occurs despite effective physical prevention and is cured after drug treatment.No surgical treatment was performed and no secondary complications of severe thrombosis were noted.Some studies have complications may gradually increase.However, through the continuous summarization of experience and technical training, vNOTES will allow great achievements in the field of gynecologic surgery.The strengths of the present study are the large sample size and standard operating procedures of vNOTES.Our study is the largest case study of vNOTES complications so far.Moreover, as one of the largest and most professional women's and children's hospitals in Southwest China, we have as many as 17 senior doctors who have mastered this new surgical approach.The 2000 cases covered all diseases treated with gynecologic surgery.Subjectively, it overcomes the technical bias caused by operation by a single surgeon.
summaries, and technical training, vNOTES will flourish among gynecologic surgeries.To conclude, considering the security and effectiveness of vNOTES, it can be routinely used in various gynecologic operations.However, surgeons should focus on preoperative evaluation, strengthen preoperative disinfection and vaginal preparation, conform to timely conversion during surgery, call experienced doctors on-stage, and routinely use antibiotics to prevent postoperative infections.AUTH O R CO NTR I B UTI O N SYL and LH designed the research protocol; QH and XL conducted the study; XL and YL analyzed the data; QH and XL drafted the manuscript; LH, ZG, DF, QZ, DG, and TL critically revised the manuscript; YL and LH acquired funding.All authors have accepted responsibility for the entire content of this manuscript and have approved its submission.

Table 2 .
The average age and body mass index of the pa- plication was classified as a grade II complication.Postoperative pelvic inflammatory disease or abscess was classified as grade II if treated with antibiotics alone, grade III if requiring incision and drainage, and grade IV if septic shock occurred; poor vaginal wound healing was classified as grade II complications if reoperation was not required.However, patients requiring reoperation tients were 37.86 ± 10.77 years and 24.20 ± 3.54, respectively.The incidences of complications were 19 (0.95%), 57 (2.85%),11 (0.55%), 1 (0.05%), and 0 (0%) in grade I, II, III, IV, and V complications, respectively.Complications developed during the establishment of the approach, intraoperatively, postoperatively, and within 1 month after discharge in 5 (0.25%), 30 (1.50%), 50 (2.50%),and 3 (0.15%) patients, respectively.The data in Table3clearly display the proportion of complications for each type of operation.The main complication of myomectomy was anemia in seven cases (4.19%).In the myomectomy group, five patients (2.99%) had postoperative infection and one patient (0.60%) had sepsis due to rectal injury (the only case of grade IV complication).In the hysterectomy group, anemia was the main complication in 11 patients (7.91%); however, there were no complications above grade III.In the pelvic floor dysfunction group, there was only one grade III complication (postoperative venous thrombosis).In the malignant tumor group, the main complication was anemia (16.67%), and no grade III or higher complications occurred.In the aforementioned cases, if the patients had several complications simultaneously, we classified them as having the highest grade according to their severity.verted to transabdominal laparoscopic repair.Of the eight cases of rectal injury, five were converted to transabdominal laparoscopy and three were repaired under vNOTES.Moreover, two patients with pelvic abscess underwent re-operative incision and drainage, one of whom developed severe sepsis and septic shock (classified as grade IV).During pelvic dysfunction surgery, an older patient developed postoperative lower extremity venous thrombosis, which was treated with physical therapy combined with anticoagulant drugs.