Endoscopic calabash technique for gastric mesenchymal tumours: A low hanging fruit or a novel endoscopic technique?

Abstract

The term subepithelial lesions encompasses a wide array of pathology of which numerous benign and malignant pathologies are grouped. A subset of these lesions are termed gastric mesenchymal tumours of which some have innate malignant potential. Currently there is various guidance on the recommended approach to the investigation and management of these lesions and there exists multiple methods of resection. Lin et al have developed and proposed a new method of resection of these gastric mesenchymal tumours within the field of endoscopy, a procedure they have termed endoscopic calabash ligation and resection. This editorial aims to outlay the current landscape for gastric mesenchymal tumours with regards to the various guidelines and resection techniques while comparing Lin et al’s new technique to those that are already established in the field of endoscopy. Advancements in endoscopy that maintain or improve patient outcomes compared to the gold standard approach are exciting developments. Lin et al’s study suggests that their technique is comparable in regard to patient outcomes while simultaneously being more efficient in its use of hospital resources including procedural time. Whilst the data and analysis proposed in the study is promising, there are areas that need to be addressed before advocating the procedure for widespread use. However, with further studies and analysis this may be foreseeable in the future.

Key Words: Gastric mesenchymal tumours; Subepithelial lesions; Gastric stromal tumours; Endoscopic resection; Endoscopic submucosal resection; Endoscopic calabash ligation resection

Core Tip: The current management of gastric stromal tumours under 20 mm can be via active surveillance or resection. Endoscopic calabash ligation and resection technique offers an exciting new way of resecting endoscopically compared to standard excision techniques already in use. Whilst showing some promising benefits in data outcomes, further studies are warranted to address the limitations in the current study.

INTRODUCTION

Gastric mesenchymal tumours (GMT)’s are a group of subepithelial Lesions (SELs) that arise from mesenchymal tissue. The subset gastric stromal tumours (GST) have significant malignant potential while gastric leiomyomas (GL) and gastric schwannomas (GS) are less common and carry a lesser but still present malignant risk. It is important to note that while these three tumour types represent the more common variants of an otherwise rare diagnosis, they do not encompass all types of GMT, of which there are numerous, and various other SELs exist. This includes benign lesions such as lipomas and inflammatory fibroid polyps to malignant lesions such as leiomyosarcomas and synovial sarcomas[1].

SELs, inclusive of GMTs, are most frequently identified during endoscopies for unrelated reasons. This is because smaller lesions are usually asymptomatic and only invite investigation when bleeding occurs, resulting in iron deficiency anaemia, or in larger lesions that result in variant degrees of obstruction when involving the gastric inlet or outlet[2]. Autopsy reports indicate that up to 20% of individuals have GSTs less than 20 mm in diameter while larger more overtly malignant GSTs are noted in 1.1 per million[3,4].

CURRENT GUIDELINES

Current European Society of Gastrointestinal Endoscopy (ESGE) guidelines recommend the use of endoscopic ultrasound (EUS) to help characterise SELs. Pathognomonic features for lipomas and varices can be seen on EUS but the accuracy for distinguishing between other forms of SELs is less than adequate. Features suggestive of GST are hypoechoic and heterogenous findings during EUS[5]. Contrast-enhanced EUS is useful for visualising microvasculature which can aid characterisation and determining the malignant potential of GSTs but should not replace tissue acquisition[4,6]. Guidelines throughout the world differ regarding which patients require tissue diagnosis. The ESGE suggests acquiring tissue diagnosis for SELs with features suggestive of GST if they are greater than 20 mm in diameter, have high risk stigmata (cystic spaces or echogenic foci, irregular borders), or require surgical resection or oncological treatment. Other societies including European Society of Medical Oncology (ESMO) and Chinese Society of Clinical Oncology (CSCO) recommend tissue sampling to obtain immunohistology for all lesions suggestive of GST by EUS[7,8]. For tissue diagnosis of lesions smaller than 20 mm in size mucosal incision assisted biopsy is preferential to EUS guided fine-needle biopsy but for larger lesions either option is recommended by ESGE.

With regards to the management of SELs there are three options: No surveillance, surveillance, and resection. ESGE currently recommends no surveillance of asymptomatic GL, GS, lipomas, heterotopic pancreas, granular cell tumours and glomus tumours, provided the diagnosis is clear. GSTs less than 20 mm in size have a low risk of malignancy and studies have shown surveillance is a safe option with intervening excision with tumour growth[9]. ESGE suggests either surveillance or resection is appropriate and should be discussed in a multidisciplinary meeting, while ESMO and CSCO advocate resection[7,8]. For undifferentiated lesions ESGE suggests surveillance at different intervals dependent on size. For surveillance EUS is advised as certain high-risk features may only be identified in this fashion. Clinicians need also be aware that low patient compliance with long-term follow-up can be expected[10]. Unknown SELs will naturally contain a proportion of unknown GSTs with inherent higher malignant risk. This enacts the need to find a balance between intensity of follow-up and patient compliance or proceeding to biopsy/resection for definitive diagnosis. Once again, these decisions are best discussed at a multidisciplinary meeting.

The current resection techniques can be via open surgery, laparoscopic wedge resections or endoscopic resection (ER). ER can be further categorised into endoscopic submucosal dissection (ESD)/endoscopic submucosal excavation (ESE), endoscopic full thickness resection (EFTR) of which there is exposing and non-exposing variants (i.e., intraperitoneal cavity exposed to gastric contents) and submucosal tunnelling ER (STER). The goal is to achieve R0 resection in lesions that have not invaded regional lymph nodes. Laparoscopic excision is considered the gold standard by institutions such as the National Comprehensive Cancer Network and ESMO for GSTs[11,12]. Unfortunately, there are no RCTs comparing the different resection techniques. Recent meta-analysis however shows similar outcomes between EFTR and laparoscopic resection[13]. Meta-Analysis of efficacy of ER of GSTs less than 20 mm showed a R0 resection rate of 95% (95%CI: 95%-99%)[14]. Meta-Analysis of STER of GSTs showed R0 resection rate of 97.9% (95%CI: 93.6%-99.3%)[15]. Studies show ESE to have a R0 resection rate of 90%-99% for GSTs less than 20 mm in diameter[16,17]. Retrospective studies have shown ESE to be equal in effectiveness to STER, particularly for GSTs less than 15 mm[17]. As previously alluded, undifferentiated SELs have a risk of being undiagnosed GST - and thus carry their inherent risk for malignancy. For this reason, they may be excised using any of the aforementioned techniques following discussion with the patient on an individual basis with each method being suitable for smaller GMTs.

The procedure for ESE involves identifying the submucosal lesion and marking the overlying mucosa via electrocoagulation. Submucosal injection of colloid/crystalloid solution infused with dye is then performed. The mucosa is then incised to access the submucosa and a circumferential mucosal incision is made. The submucosa is then dissected of the tumour capsule in a stepwise fashion to the tumour base to allow specimen retrieval before wound closure is performed with clips. The procedure can benefit from traction to aid exposure of the submucosa. EFTR is like ESE but involves a full thickness resection of the gastric wall. This creates a perforation in exposed techniques which is subsequently closed. The non-exposed technique involves a secure apposition of the serosa prior to excision. STER is a tunnelling technique in which a full muscularis propria can be excised without creating a full thickness perforation due to the remaining overlying mucosa flap. The approach for this technique requires accessing the submucosa usually a distance of 4 cm away from the lesion and burrowing through the submucosa to reach the intended target. Benefits of this procedure include reduced risk of exposure of peritoneal cavity to gastric contents and the lack of need for traction[18].

ENDOSCOPIC CALABASH LIGATION RESECTION TECHNIQUE

The retrospective study by Lin et al[19] proposes the use of their novel endoscopic therapeutic technique for resection of small GMT. The ‘endoscopic calabash ligation resection’ technique (ECLR). So named for the close resemblance the tumour and submucosa have to the family of fruit in the moment right before excision. The study proposes that this technique is equally efficacious in clinical outcomes and more efficient in relation to cost and procedural time compared to ESE.

The ECLR technique as described in Lin et als’ study[19] begins like ESE. The submucosal lesion is identified, overlying mucosa marked, submucosal injection, mucosal incision, exposure, and dissection around the tumour. Their technique then uses negative pressure to aspirate the exposed head of tumour into the endoscope tip before placing a nylon ligation at the tumour base and a further nylon ligation inferior to this forming a stalk like structure, hence mimicking the appearance of a calabash. The tumour was then excised by cutting between the two nylon ligations leaving the submucosa adhered together by the already placed inferior nylon ligation.

ECLR STUDY RESULTS

The results analysed the outcomes between ESE and ECLR with 260 patients in each cohort. Both procedures resulted in intact tumour capsules and R0 excision in all patients. No patients in either cohort developed major intraoperative bleeding. 28 patients (3.08%) in the ESE cohort had intraoperative perforations compared to 3 (1.15%) in the ECLR cohort. All perforations were repaired endoscopically without the need for surgical conversion. There were statistically different post-operative pain scores with ESE 0.25 ± 0.56 and ECLR 0.16 ± 0.43, although given overall low level of pain, this was not considered clinically significant. One patient (0.38%) developed postoperatively bleeding in the ESE cohort with no postoperative bleeding in the ELCR cohort. There were no post-operative perforations in either cohort. One case of intraabdominal infection (0.38%) occurred in the ESE cohort with no cases in ECLR cohort. Of 3 cases (1.15%) of postoperative respiratory infections occurred in the ESE cohort and 1 (0.38%) in the ECLR cohort. A statistically significant difference in postoperative electrocoagulation syndrome was noted with 8 patients (3.08%) in the ESE cohort affected and only 1 (0.38%) in the ECLR cohort. Operative duration was significantly shorter in the ECLR cohort with mean time 33.40 ± 8.46 minutes compared to 54 ± 9.47 minutes in the ESE cohort. Length of hospital stay was comparable with average of 5.04 ± 1.46 in the ESE cohort and 5.24 ± 1.45 in the ECLR cohort. Medical costs were significantly cheaper in the ECLR cohort with 16111.40 ± 3709.71 China yuan (CNY) compared to 14435.25 ± 3458.43 CNY in the ESE group with material costs being significantly less in the ECLR cohort 3571.96 ± 1507.32 CNY compared to 4115.81 ± 1486.70 CNY. During follow-up no tumour reoccurrence, metastasis or death was noted in either cohort and length of follow up was statistically similar.

A strength of Lin et als’ study[19] is its large sample size, including a total of 520 subjects with cases spanning over a long period of time from November 2021 to March 2023. By including a substantial number of patients, the study offers a wide dataset for interpreting and comparing the outcomes associated with ESE and ECLR procedures. The study highlights a critical advantage of the ECLR procedure - its cost-effectiveness in comparison to ESE. ECLR has shown significant cost reductions in not only material costs but also medical costs when compared to the control group. With reduced material and equipment expenses, ECLR could potentially offer a more economical alternative to endoscopic GMT resections. This cost reduction could potentially have an impact on the future of GMT resections if this data was found to be consistently maintained across different countries and healthcare systems. Another key strength of the study is the reduced procedural time associated with ECLR compared to the control group. Shorter operative durations are advantageous as they minimize patient exposure to anaesthesia and thus associated risk of complications from prolonged exposure and leads to cost-effective use of clinical space and personnel. This could lead to an increase in potential number of cases that can be performed per day, reducing wait times and pressure on healthcare systems. The study demonstrates a reduced rate of post-operation electrocoagulation syndrome in the ECLR cohort, with the prevalence rate in the study group markedly lower relative to the control group. This reduction is deemed to be attributable to both the decreased procedural time and the use of a nylon loop for haemostasis instead of traditional electrocoagulation techniques. This finding suggests that ECLR may offer a safer alternative with fewer post-operative complications related to electrocoagulation. Although the reduction in pain levels was not statistically significant, the study reports a lower pain score in the ECLR cohort compared to the control group. This trend towards reduced pain shows the potential benefits of ECLR in enhancing patient comfort post-operation, which is an important consideration in the overall evaluation of surgical techniques and patient satisfaction.

LIMITATIONS

While the recent study on ECLR offers valuable insights into new techniques for resecting GSTs, it is important to consider several limitations from the study that may impact the interpretation and generalisability of its findings. Firstly, the study exclusively compares ESE with ECLR, without including comparison to other resection techniques. This narrow focus limits the ability to assess how this new procedure stands in comparison to alternative existing methods. Secondly, the study’s inclusion criteria specify that only patients who completed treatment through ESE or ECLR were considered. Consequently, this means patients who experienced technically unsuccessful resections were excluded from the study group and follow up. This exclusion means that the technical success rate of the procedures is not fully known, potentially skewing the results by not accounting for resections that may not have been successful. From this selection bias we are left in the unknowing position of how many patients who underwent failed ECLR that had what would be considered adverse outcomes. Within this excluded cohort we are unable to compare the prevalence of conversion to surgical management, severe intraoperative bleeding and inadequate resection or tumour capsule rupture. The design of the study in this way sets out to compare largely uncomplicated successful procedures but negates the ability to compare procedures in the real world scenario were procedures are often complicated or unsuccessful.

There is no explanation of how patients were selected for the ESE or ECLR, raising concerns about potential selection bias. The patients may have been randomly assigned into each procedure but from the information available this is not apparent. The study criteria also states patients with ‘severe primary diseases affecting the heart, liver, kidney, or hematopoietic system’ and ‘patients with other conditions significantly increasing treatment risk and hospital stay duration’ were all excluded. As a result, 11 patients were excluded from the analysis. This raises questions about whether the outcomes observed from ECLR are applicable to higher-risk patients, potentially limiting the study’s application in the general population. It is reasonable to predict that a procedure with shorter procedural times would be of benefit to these more co-morbid patients.

The study methods state 297 patients underwent ESE procedures and 292 received ECLR procedure before propensity score matching (PSM) reduced the control and study group to 260 patients each. It is unclear why these extra patients were excluded from the study, thus this may have altered the research data and final conclusions drawn. There is some evidence to show that using PSM can have less than the desired effect leading to imbalance, model dependence and selection bias[20].

The study describes the endoscopic experience of the participating endoscopists as having a cumulative total of 500 cases but does not specify how many individual endoscopists were involved or their respective experience levels and individual caseloads. This ambiguity makes it difficult to assess the impact of individual expertise on the outcomes achieved.

As aforementioned, the management strategy for GMTs under 20 mm is controversial. The study does not address variations in management strategies recommended by different guidelines, such as those from the ESGE and other international bodies, which advocate for biopsy and/or surveillance for smaller GMTs. Lin et al’s study[19] was completed in China in Shenzhen People’s Hospital. Guidelines from the Chinese Clinical Oncology recommend surgical resection of a GMTs, even if under 20mm, thus making this study relevant in endoscopy in China where the study was performed[14]. Considering this, ECLR might not be as clinically useful worldwide in countries where these patients may be managed conservatively.

Another thing for consideration is the theoretical risk of tumour capsule rupture while using negative suction techniques, like that used during ECLR. How does this compare to surgical and other endoscopic techniques? Given the inclusion criteria of this study we can’t compare rates of tumour capsule rupture. It is well appreciated that tumour capsule rupture during surgical resection of GSTs is an independent risk factor for peritoneal metastasis. However, with ER techniques it is currently not well known if tumour rupture increases recurrence rate or increases the risk of peritoneal metastasis. There was no statistically significant difference in the rate of peritoneal metastasis in a small study that compared GST tumour rupture in 27 patients vs en-bloc resection in 168 patients at the two-year follow-up mark[21]. This study would suggest that even if rates of tumour rupture are different between techniques it’s clinical significance might be less than we would naturally expect in comparison to surgical tumour rupture.

As Lin et al’s study[19] mentions - further research with multi-centre prospective studies should be performed to further assess ECLR’s impact on GST resection. Patients should be followed up for longer periods of time to assess post-operative complications and tumour recurrence rates. The exclusion criteria for those with other medical conditions was strict for this ECLR study. It would be useful to assess how co-morbid higher risk patients tolerated ECLR, making it more clinically applicable to patient populations. Additionally, it would be beneficial for studies with ECLR to be inclusive of GSTs > 20 mm, to assess if it has favourable outcomes with the GSTs that are more likely to be resected.

CONCLUSION

In conclusion, there is some debate about the management of GMT. The decision to resect or not should be based at a multidisciplinary meeting to take into account the lesion’s characteristics, biopsy results, size, location, fitness of the patient, local expertise and patient preferences. The paper from Lin et al[19] describes an exciting new ER technique for the resection of these tumours. However, to reliably assess whether this technique has a role that warrants widespread adoption, more detailed information is required. For instance, a comparison with other established and emerging methods for resecting GSTs, such as ESD or the STER method, would be valuable. This highlights important areas for further research into ECLR’s potential applicability. Furthermore, further evaluation with prospective studies in multiple centres with greater patient inclusion criteria and longer follow up times to reliably compare outcomes. ECLR is shown to be more cost and time efficient compared to ESE in uncomplicated resections. These are premium aspects to consider in the medical world where resource abundance is often limited. If further studies can definitively conclude its safety profile, then it has the potential to not only have widespread use but also change management plans as the cost-benefit ratio may swing more in favour of intervention over surveillance in those regions than currently favour surveillance.