The aim of this study was to compare the indications, operative details and clinical outcomes of nonexposed endoscopic wall-inversion surgery (NEWS) with endoscopic-navigated laparoscopic wedge resection (LWR) for submucosal gastric tumours.
Laparoscopic wedge resection is currently the method of choice for the resection of benign and semi-malignant gastric tumours. It is superior to open resections in that it is associated with a shorter convalescence time, shorter operation length, and smaller intraoperative blood loss [
Laparoscopic-endoscopic hybrid wedge resections differ from LWRs in that the endoscopist has a more active role. In these techniques, both the endoscopist and the laparoscopist are involved in the resection of the tumour. In this study, we used nonexposed endoscopic wall-inversion surgery, which is based on the “close first, cut later” principle that avoids creating communication between the gastric lumen and peritoneal cavity [
Medical data for all patients undergoing NEWS at the Department of Surgery of the Faculty Hospital Kralovske Vinohrady in Prague were recorded in a prospective database. This database was reviewed to find all submucosal gastric tumours between February 2016 and October 2017. Patients with endoluminally growing submucosal tumours and early gastric cancer were indicated to undergo NEWS. For the purpose of the study, to make two comparable groups, we selected only submucosal tumours; patients with early gastric cancer were excluded from the study. Medical records from the Department of Visceral Surgery of Kepler University Hospital in Linz between August 2009 and October 2017 were retrospectively reviewed to find all patients who underwent LWR of submucosal gastric tumours. Data collected from each patient consisted of demographic details (age, sex, BMI, and clinical presentation), tumour characteristics (location, size, and histology), and surgical and perioperative details (type and length of operation, complications and length of postoperative hospital stay). Gastrointestinal stromal tumours (GISTs) were classified on the basis of their mitotic index and size according to the criteria of the NIH consensus statement [
All procedures were performed by a surgeon and an endoscopist. Intraoperative photographs are shown in Figure
Nonexposed endoscopic wall-inversion surgery: intraoperative photos. (a) Endoscopic view of a gastrointestinal stromal tumour of the stomach. (b) Laparoscopic seromuscular incision around the lesion. (c) Inversion of the tumour into the stomach. (d) Laparoscopic suture. (e) Endoscopic muco-sub-mucosal incision. (f) Endoscopic muco-sub-mucosal dissection. (g) Resected specimen inside the stomach. (h) Resected specimen after endoscopic extraction.
Nonexposed endoscopic wall-inversion surgery: illustration of the procedure. (a) Submucosal injection. (b) Seromuscular incision and laparoscopic suture. (c) Completion of the laparoscopic suture and inversion of the tumour in the gastric lumen. (d) Submucosal dissection.
Intraoperative photographs of the main steps are shown in Figure
Endoscopic-navigated laparoscopic wedge resections: intraoperative photos. (a) Endoscopic and laparoscopic confirmation of the location of the tumour in the stomach. (b and c) Wedge resection performed with a stapler. (d) The resected specimen inserted into an endosack. (e and f) Suture of the staple line.
The study group consisted of 11 patients who underwent NEWS, and the control group consisted of 12 patients who underwent LWR for gastric tumours. Results are summarised in Tables
Patient characteristics.
NEWS | LWR | |
---|---|---|
Age (average, range) | 65, 44-80 | 65, 31-77 |
Sex ratio (male : female) | 7 : 4 | 7 : 5 |
BMI (average, range) | 27.8, 17.9-38.7 | 25.6, 19.0-30.4 |
Presentation (symptomatic : asymptomatic) | 2 : 9 | 7 : 5 |
Tumour characteristics.
NEWS | LWR | |
---|---|---|
Location | ||
Subcardial region | ||
|
2 | 0 |
|
1 | 0 |
|
0 | 1 |
Body | ||
|
0 | 5 |
|
2 | 0 |
Fundus | ||
|
0 | 2 |
|
2 | 0 |
|
1 | 0 |
Prepyloric region | ||
|
1 | 4 |
|
1 | 0 |
|
1 | 0 |
Diameter of resected specimen in mm (average, range) | 40, 20-55 | 45, 35-95 |
Diameter of tumour in mm (average, range) | 27, 5-50 | 35, 15-80 |
Difference in diameter between resected specimen and tumour in mm (average, range) | 13, 5-23 | 30, 15-40 |
GIST tumours | 6 | 7 |
Very low malignant potential | 6 | 6 |
Low malignant potential | 0 | 1 |
Other tumours | 5 | 5 |
Surgical and perioperative details.
NEWS | LWR | |
---|---|---|
Length of operation (minutes) | 96, 70-120 | 62, 41-92 |
Intraoperative complications | 2 | 1 |
Postoperative complications | 2 | 0 |
Length of postoperative hospital stay (days) | 6.8, 5-10 | 6.5, 3-11 |
The average patient age was 65 years with a range of 44 to 80. The average BMI was 27.8, and the male to female ratio was 7 : 4. Nine patients were asymptomatic; the tumours were discovered incidentally. One patient presented with anaemia and one with abdominal pain.
The average patient age was 65 years with a range of 31 to 77. The average BMI was 25.6, and the male to female ratio was 7 : 5. Five tumours were detected incidentally. Three presented with gastrointestinal bleeding, two with gastroesophageal reflux, one with dysphagia, and one with dyspeptic symptoms.
Three tumours were located in the subcardial region, two in the body, three in the fundus, and three in the prepyloric region. The average largest diameter of the tumours was 27 mm (range: 5 to 50 mm). The average difference in size between the largest diameter of the resected specimen and tumour was 13 mm. R0 resection margins were achieved in all cases.
Histological examination of the resected specimen revealed six GISTs, one submucosal lipoma, one leiomyoma, one endocrine tumour, one Vanek’s tumour, and one case of ectopic pancreatic tissue. The GISTs were classified according to the NIH consensus statement. All had mitotic indices of less than 5 per 50 HPF and diameters of less than 50 mm and were thus classified as having very low malignant potential.
Five tumours were located in the body, four in the prepyloric region, two in the fundus, and one in the subcardial region. The average diameter of the tumours was 35 mm (range: 15 to 80 mm). The average difference in size between the largest diameter of the resected specimen and tumour was 30 mm. R0 resections margins were achieved in eleven of the 12 cases; in one case, tumour tissue was detected microscopically at the resection margin.
Histological examination revealed seven GISTs, one leiomyoma, two cases of ectopic pancreatic tissue, one endometriosis, and one hyperplasiogenic polyp. All seven GISTs had mitotic indices of less than 5 per 50 HPF. One had a diameter of 53 mm and was classified as having low malignant potential. The other six had diameters between 20 and 50 mm and were thus classified as having very low malignant potential.
The average operating time was 96 minutes (range: 70 to 120 ), and the average length of hospitalisation was 6.8 days (range: 5 to 10).
Two intraoperative complications occurred. In one case, bleeding occurred at the resection line after endoscopic submucosal dissection, which was successfully treated by endoclips (Figure
Two postoperative complications occurred: one case of suture line bleeding and one subcapsular liver hematoma. The suture line bleeding presented as hematemesis. The patient underwent acute gastroscopy, which revealed resection line bleeding and was treated with argon photocoagulation and hemoclips. No significant drop in hemoglobin concentration occurred, and the patient remained stable throughout the postoperative period.
The subcapsular liver hematoma occurred in a patient with hepatic steatosis. The hematoma was presumably caused intraoperatively by the laparosopic retractor. It had a size of
The average length of operation was 62 minutes (range: 41 to 92). One case of suture line bleeding occurred, which was treated by the application of the hemostatic agent PerClot. No postoperative complications occurred. The average length of hospitalisation was 6.5 days (range: 3 to 11).
The patients underwent follow-up gastroscopy at three and six months after surgery. At twelve months, they underwent gastroscopy with biopsy of the scar and endosonography. No tumour recurrences or gastric motility disorders were reported, and all patients remained alive and healthy throughout the follow-up period.
The patients underwent follow-up gastroscopy at three, six, and twelve months after surgery. Two deaths occurred (430 and 874 days after surgery), unrelated to the oncological disease. The one patient in whom R1 resection margins were reported remained recurrence-free throughout the follow-up period.
Cooperative endoscopic-laparoscopic surgery for the resection of gastric tumours has been performed at the Department of Surgery of the Faculty Hospital Kralovske Vinohrady since January 2016. Our initial experience with these techniques was detailed in a previously published study [
After two reports on porcine models, NEWS was introduced by Mitsui et al. in 2014 in human patients [
When NEWS was performed, dye was injected endoscopically into the submucosa, which could then be seen on the external surface of the stomach by the laparoscope. This allows for a more superior way of delineating the extent of the resection than relying solely on the endoscopic light shining through the gastric wall as is done in LWRs. We showed that the difference in size between the resected specimen and tumour was smaller and that negative resections were more frequently achieved in the NEWS patients. These more precise resections reduce the likelihood of stomach deformation from needless loss of excessive tissue, while still achieving negative resection margins. Furthermore, NEWS does not require opening the gastric wall, and therefore, gastric content (bacteria or tumour cells) does not contaminate the peritoneal cavity [
Our NEWS technique differs slightly from the method described by Goto et al. [
More precise resections achieved by NEWS may lead to a lower incidence of gastric motility disorders. At one year of follow-up no symptoms suggestive of gastric motility were reported in any of the patients who underwent NEWS. Similar results were achieved by Tsujimoto et al., who reported no evidence of gastric motility disorders in a cohort of 20 patients, and in a study by Waseda et al., who reported two cases of gastric motility disorders in a cohort of 22 patients [
Despite these advantages to the hybrid approach, as these techniques are still in their early days there are some associated disadvantages thought to be due to limited experience and the technical difficulty of these techniques. The hybrid resections were associated with longer operation times when compared with standard laparoscopic resections. Although no serious complications occurred in either group, more complications were associated with the NEWS operations. Again, this reflects the complexity and limited experience of these operations and we expect the complication rate to decline with increased experience.
Length of hospitalisation was slightly longer in patients undergoing NEWS. Our main goal was to safely implement a new technique. Reduction of hospital stay is a secondary goal. The average length of hospital stay was increased by the two patients who had postoperative complications. When these patients are not included, the average length of stay was reduced to 6.3 days. We believe that with increasing experience the length of stay will decrease, and now we aim to discharge patients by the fifth postoperative day.
Tumour location, size, and direction of growth are the key factors when choosing the operative approach. Concerning size, we should emphasise that tumours with diameters larger than 4 cm can be resected using the NEWS technique as long as the other two diameters are under 4 cm, so tumour can be extracted via the oesophagus. Endoluminally growing tumours are more easily accessible by the endoscope and are more appropriate for the NEWS technique. Tumours in cardiac and pyloric regions are difficult for the laparoscopic approach and often necessitate open resections. These tumours can be more easily approached endoscopically and thus can be resected using the NEWS technique. In this study, three tumours in the cardiac region and two tumours in the pyloric region were successfully resected with the NEWS technique. When considering the size of the tumour, smaller tumours are more preferably resected using the NEWS technique. For larger tumours, laparoscopic or open wedge resection is more convenient.
Nonexposed endoscopic wall-inversion surgeries allow for more precise resections and more frequent achievement of negative resection margins than LWRs. They may result in better preservation of the gastric function and reduction of peritoneal contamination and tumour seeding. This study should be followed by larger prospective and randomised trials to verify our observations.
The data used to support the findings of this study are included within the article.
The authors declare no conflict of interest.
This work was supported by the Charles University research program PROGRES Q 28 (Oncology).