Patient Characteristics and Evolution of Surgical Management of Gastric/ Gastro-esophageal Junction (GOJ) Adenocarcinoma From 1998 to 2018: Experience From a Single Gastrointestinal Surgical Unit in a Resource Limited Setting


 Introduction: Gastric/gastro-oesopageal junctional adenocarcinoma/ (GC/GOJ CA) is the fourth commonly diagnosed cancer and the second commonest cancer related cause of death worldwide. Curative therapy for GC involves surgical resection with an accompanying lymphadenectomy. Yet, most of the GC/GOJ tumours are inoperable at presentation and has a poor prognosis in our setup probably due to the absence of screening programme. The objective of this study was to describe the operability and curative resection rate and outcomes of gastric/GOJ adenocarcinoma in Sri Lankan setting from 1998 to 2018.Methodology- In a retrospective analysis of prospectively maintained data from, surgery for GCs in a tertiary care center over a period of 21 years. All patients underwent surgery for GC/GOJ adenocarcinoma by single gastrointestinal surgeon. Results: There were 153 patients with mean age of 57.4(22 to 84) years. The majority (n=109) were Males (Male: Female was 2.3 : 1) . There were 88(56.4%) distal, 60(39.2%) proximal and 5(3.2%) diffuse cancers. Only 82 (53.6 %) patients underwent curative resection. Only 3 patients underwent palliative major resection (total gastrectomy(n=1), distal gastrectomy(n=2) Among curative resections,51(62.2%) had distal gastrectomy and 17(20.7%) had total gastrectomy.R0 resection was performed in 75 (91.5%), and 6 (7.3%) had microscopic positive margins (R1). Resectability of the proximal GCs (48.3%, n=29) was lower (63.4%, n=54) than the distal GCs (p= 0.117). Further, resectability among patients with GCs has not increased significantly over the past 21 years.Conclusions: Majority of the patients especially with proximal GC are inoperable at presentation. Despite the availability of endoscopes the operability has not changed significantly over the years. Given the low incidence, screening for GC is not seemingly feasible in Sri Lanka. Therefore an attempt has to be made to identify those with high risk factors during diagnostic endoscopic procedures.


Introduction
Gastric/gastrooesopahgeal junctional adenocarcinoma(GC/GOJ) is the second leading cause of cancerrelated deaths as well as the fourth commonly diagnosed cancer worldwide [1,2]. Over 70% of new gastric cancer related deaths occur in developing countries, and its 5-year survival rate is less than 20% [3].The incidence of GC is low in South and Central Asia. The incidence of GC is reported in Chennai, India as 9.1 per 100,000 and in Karachi as 4.5 per 100,000 [4,5]. According to the National Cancer Registry data of 2011, Sri Lanka has lower incidence of GCA in comparison with global and regional countries, with an incidence of 1.2 per 100,000 population, and an age-adjusted mortality rate of 6.7 [6].
The only curative treatment for GC is the complete resection (Endoscopic or surgical) [7,8] Curative surgery for GC consists of en-bloc resection of the primary (total or subtotal gastrectomy) along with accompanying lymphadenectomy [9,10] and resection of adjacent involved organs selectively [11] Survival of GC patients is improved by adjuvant chemo-radiotherapy. On an average, in Western countries, operability rate is 73.3% and curative resection rate is 30% [12]. Locally, the operability rate is reported to be 66.7% and curative resection rate 33.3% [12]. According to a recent study, only around 40% of patients with GCA qualify for curative treatment, which consists of surgical resection with or without adjuvant chemotherapy [12].Unfortunately, the 5-year overall survival (OS) even after curative treatment remains poor [12]. At present there are no published data regarding the surgical outcomes and resectability trends over a longer period of time for GAC for the Sri Lankan setting. The aim of this study was to describe clinical pro le of patients underwent surgery, determine the resectability and curative resection rate and to describe the surgical approach, outcome and of GC in Sri Lankan setting.

Methods
This descriptive study included consecutive GC patients underwent surgery from 1998 to 2018 by a single gastrointestinal surgeon in two Teaching Hospitals Kandy and Colombo South). A total of 153patients with GC underwent surgery. Patients with other gastric tumours(i.e Gastrointestinal stromal tumours, sarcomas), metastatic disease and those who refused surgical treatment were excluded. Furthermore 22 patients with missing medical data were excluded from the study. Informed written consent was obtained from all patients for surgical procedures. Ethical approval was not necessary as this paper included a secondary analysis of already collected data without identi cation details.
All patients underwent upper gastrointestinal endoscopy and biopsy for diagnosis. Patients were staged preoperatively with USS abdomen or computed tomography of the abdomen and pelvis.

Collection of data
Data was collected from personally kept patient records and operation records. Pathology data were collected from the operation register maintained in the unit.

Surgical technique
All procedures were performed according to standard techniques. All procedure were carried out with a curative intent. The type of resection was based on the tumour location. Subtotal gastrectomy was performed for the distal or antral cancers whereas, total gastrectomy was performed for proximal cancers, linitis plastica and cancers involving the body of stomach Continuity was restored with Billroth-II gastrojejunostomy in cases of subtotal gastrectomy and end to side oesophagojujnostomy in cases of total gastrectomy. Whenever possible either D1 gastrectomy D2 resection was performed according to the Japanese (JSGE) guidelines. Palliative resection (subtotal or total) was performed in patients with metastatic disease who had bleeding not amenable to endoscopic treatment. Palliative gastrojejunostomy was performed in patients with inoperable distal gastric tumours with gastric outlet obstruction. In patients with inoperable tumours not suitable for gastric resection or bypass, only biopsy was obtained.

Analysis
The data were analyzed using Statistical Package for Social Sciences (SPSS) version 21.0. Descriptive statistics were used. In addition, association between resectability and relevant parameters were analyzed using the Chi square test at 5% signi cance level.

Results
There were 109 males and 44 females, with a male to female ratio of 2.3:1. The mean age at the diagnosis was 57.4 (range 22-84) years. The majority (n = 120,78.4%)of tumour were stage III or above. There were 88(56.4%) distal gastric tumour and locations of tumours are given in Table 1. The majority of the patients were presented with features of gastric outlet obstruction (n = 51,33.3% followed by dysphagia (n = 39,25.5%). Figure 1 shows the clinical presentation of gastric cancers in this cohort.     Figure 4 shows the resection status of the proximal and distal tumours underwent surgery with curative intent. Resectability of proximal GC (n = 29,48.3% was lower than distal GC (61.4%, n = 54), p = 0.161 Enbloc resection of adjacent organs was carried out in 4 (7%) cases of distal GC. These included 3 distal pancreatectomies with spleenectomy and 1 transverse colon resection.

Resection margin status in curative resections
Among these R0 resection was performed in 75 (91.4%) cases and 6 (7.4%) cases had microscopic positive margins. Status of resection margins were showed in Table 4.  Post-operative complications There were 33 patients (21.5%) who experienced post-operative complications. Two patients developed post-operative duodenal stump leak and one each developed anastomotic leak and pelvic abscess which were managed with reoperation and drainage. There were 12 (7.8%) post-operative deaths due to nonsurgical post-operative complications. The mean duration of postoperative hospital stay was 8.2 days (range 2-26). Table 6 shows the overall post-operative complications.

Discussion
In this study, clinico-demographic characteristics, surgical approach, surgical techniques and postoperative outcomes over period 20 years were analyzed for patients GC/GOJ adenocarcinoms. In addition, resectability for patients with distal vs proximal locations of GC were determined. This is the largest series of single unit experience of GC/GOJ cohort underwent surgery in Sri Lankan setting. Our results showed that signi cant proportion of patients selected for surgery was found to have an unrectable disease.
The average age of presentation was 57.4 years, which is about 12 years lower than reported from American cancer registries and a decade higher than reports from Asia [13][14][15]. Over 70% of patients (male: female, 2.3:1) in our series were male; the male preponderance is in line with data from the American cancer surveillance database, Surveillance, Epidemiology, and End Results (SEER) [16]. In 2014, the gender ratios of newly diagnosed GCA cases were 1.6:1 (male: female) in the USA and 2.2:1 in China [17][18][19][20]. Therefore, our male:female ratio is higher compared to USA and China.
According to the National Cancer Registry data of 2011, Sri Lanka has lower incidence of GCA in comparison with global and regional countries, with an incidence of 1.2 per 100,000 population, and an age adjusted mortality rate of 6.7 [21]. Therefore, GCA does not qualify for a mass screening program in our country according to the WHO criteria nor does Sr Lanka has adequate resources for such program. This explains why almost all patients in our series were symptomatic and presented late Commonest clinical presentation in distal GC was gastric outlet obstruction and that for proximal cancers were dysphagia. Therefore, it is not surprising to have advanced stage of the disease at presentation in majority of the patients. Similar trend can be observed in other regions such as India, China, America and Europe where population-based cancer screening is not available most likely due to the low incidence of GCA [22][23][24].
In our study 56.9% of tumours originated in the antro-pyloric region. Studies from India [15,25] and China [26] also suggest that distal tumours as more common than proximal tumours. However, in developed countries, there is a trend towards more proximal tumours [27]. After 2010 in our cohort there is a rising trend in resectability in GEJ cancers. This could be due to multiple reasons including increase in the incidence of especially GEJ cancers [28] increased use of neoadjuvant chemotherapy and increased availability of facilities for minimally invasive operative procedures.
In our study, 153 GC/GOJ patients underwent surgery with an overall resectability rate of 53.6 % (82/153) and curative R0 resection rate of 51.9 (68/74) %. These results are comparable with local study conducted in India and also with western studies [29][30][31] though it is lower than 70-90% curative resection rate reported in Japanese and Australian studies [32,33] In our cohort of patients, resectability of proximal cancers was lower than distal ones.
Currently there is a wide variation in the use of adjuvant therapy for GC between Eastern and Western countries. In the Western countries since non-screened patients are diagnosed with locally advanced/metastatic disease [34] neo adjuvant therapy is commonly used in North America. Chemoradiotherapy is administrated in a neo adjuvant setting while in the other Western countries chemotherapy alone is used as neoadjuvant treatment for locally advanced gastric cancer [34] In contrast over 50% of gastric cancers diagnosed in Korea and Japan are early stage and surgery is offered as the primary treatment irrespective of the stage of diagnosis. The overall survival also vastly different between the East and the West (-60% vs. − 20%, respectively) [34]. In our setting where there is no screening program available and most patients presented with advanced disease. Almost all patients with GCA underwent primary surgery while in the case of patient s with advanced GEJ tumours regular use of neoadjuvanct chemotherapy can be observed towards the latter part of the study. One limitation of this study was that despite the fact that study period spanned over 20 years where surgical technique, adjuvant therapy, and pathological de nitions have changed over the time weather it has translated to increased overall survival could not be determined due to smaller sample size and inadequate follow up data

Conclusions
Most patients with operable GCA present at an advanced stage in this series. Majority, especially with proximal GC were not resectable but during last decade there was a rising trend in resectability of GEJ tumours. Despite the availability of endoscopes over the time, the resectability has not changed signi cantly over the years. In contrast, given the low incidence of the disease, endoscopic screening is not feasible in gastric adenocarcinoma in Sri Lanka. Therefore, an attempt has to be made to identify those with high risk factors (patients with atrophic gastritis, intestinal metaplasia and Barrett's) among patients who undergo endoscopy.

Acknowledgements-
The authors thank all the patients included in this study and ward staff who involved in care of patients with gastric cancer. Funding-

None
Data Availability- The data sets generated during and/or analyzed during the current study are not publicly available due to contractual limitations regarding rights to use of the data but may be available from the corresponding author on reasonable request.
Compliance with Ethical Standards Con ict of Interest-DS and TGAP declare that they have no con ict of interest.
Research Involving Human Participants and/or Animals-This article does not contain any studies with human participants or animals per-formed by any of the authors.
Informed Consent For this type of study, formal consent is not required. For all surgical procedures informed written consent was taken.
Con ict of interest -None