Surgical and Survival Outcomes of Operable Gastric Cancer—Experience from a Tertiary Care Center in South India

Abstract

Gastric cancer is one of the leading cancers in Southern India. Data regarding the gastric cancers among the Indian population is sparse. Most patients in our country have locally advanced gastric cancers due to delayed presentation. In this article, we present our data regarding the presentation patterns, epidemiological demographics, surgical outcomes, and survival patterns from a tertiary care center in South India. This is a retrospective analysis of gastric cancer patients who underwent gastrectomy in our institution between January 2015 and November 2021 (n = 102). The data regarding patient characteristics, histopathology, and perioperative outcomes were analysed from medical records. The adjuvant treatment received and survival details were collected from the follow-up records and by telephonic interviews. A total of 128 patients were assessable, 102 patients underwent gastrectomy in a period of 6 years. The median age of presentation was 60 years and males were more commonly affected (70.6%). Most common presentation was pain abdomen followed by gastric outlet obstruction. Adenocarcinoma NOS (93%) was the most common histological type. Most of the Patients had antropyloric growths (79.4%) and subtotal gastrectomy with D2 lymphadenectomy was the most common surgery performed. Majority of the tumors were T4 tumors (55.9%) and nodal metastases were detected in 74% of the specimens. Predominant morbidity was wound infection (6.1%) followed by anastomotic leak (5.9%) with a combined overall morbidity of 16.7% and 30-day mortality of 2.9%. Seventy five (80.5%) patients were able to complete all planned 6 cycles of adjuvant chemotherapy. The median time of survival calculated by Kaplan–Meier method was 23 months with 2-year and 3-year overall survival rates of 31% and 22%, respectively. Lymphovascular invasion (LVSI) and lymph nodal burden were the risk factors associated with recurrences and deaths. The patient characteristics, histological factors, and perioperative outcomes revealed most of our patients presented in locally advanced stages with poor risk histological types and increased nodal burden contributing to the lower survival in our population. Inferior survival outcomes suggest the need to explore perioperative and neoadjuvant chemotherapy options in our population.

Introduction

Stomach cancer remains an important cancer worldwide, ranking fifth in incidence and fourth for mortality globally [1]. Gastric cancer is the sixth leading cause of cancer and deaths in 2020 in India and the third most common among males [2]. The global 5-year survival is around 20%, with the exceptions of 65% in Japan [3] and 71.5% in South Korea [4], where population screening has led to the effective diagnosis of tumors at early stages [5]. The overall 5-year survival in India has been estimated to be between 4 and 15% only [6][6].

Randomized clinical trials have shown that combined modality therapy is effective in prolonging survival for patients with nonmetastatic gastric and gastroesophageal adenocarcinoma. Standard treatment approach until recently included D2 lymphadenectomy followed by adjuvant chemotherapy. Currently, this trend is changing and guidelines list perioperative chemotherapy or postoperative chemotherapy plus chemoradiation or postoperative chemotherapy as effective adjuvant treatment options after an adequate lymph node dissection [8]. Large database studies, such as the National Cancer Database, demonstrate an increase in the application of preoperative therapy, but one-third to one-fourth of patients still undergo a surgery first approach [9]. This scenario is particularly true in India. Data available regarding operable gastric cancers in India is sparse [10]. India is a sub-continent with different ethnic population hence gastric cancer presentation and survival cannot be compared with American, European, and Asian population. We present our experience in treating operable gastric cancer of 102 patients during a period of 6 years 2015–2021.

Materials and Methods

This is a retrospective analysis of gastric cancer patients who were treated in surgical oncology department in our institute between January 2015 and November 2021 (n = 102). Mandatory investigations done were upper GI endoscopy with biopsy for histopathological confirmation, contrast enhanced computed tomography of abdomen with gastric protocol, and chest X-ray. All clinically and radiologically nonmetastatic resectable cases were taken up for surgery. Hospital medical records were used to collect data on clinical parameters, type of surgery, and histopathological staging. Follow-up and survival data were collected by telephonic interviews and from patients’ medical records. All patients were discussed in multidisciplinary tumor board before surgery.

Depending on the site of primary tumor, radical total or subtotal gastrectomy was performed. If the tumor included an adjacent organ, multivisceral resection was considered provided R0 resection was achievable. Majority of our patients underwent D1 + or D2 lymphadenectomy. Histopathological staging was done using tumor, node, metastasis (TNM) classification as defined by American Joint Committee on Cancer (AJCC), 7th and 8th edition (after 2018) [11] [12]. Following surgery, all cases were rediscussed in a multidisciplinary tumor board meeting and adjuvant therapy was decided. The patients were followed up in the surgical oncology outpatient and the medical oncology outpatient clinics on a regular basis.

Data was analyzed using Medcalc 2.0 software. Statistical significance for individual variables was analyzed using Fisher’s exact test for categorical variables. Survival analysis was performed using Kaplan–Meier curves and compared using log rank (Mantel-Cox) test. Postoperative complications (within 30 days of surgery) were reported according to Clavien-Dindo classification. Death occurring during postoperative period in hospital or within 30 days of surgery was considered as postoperative mortality. Overall survival (OS) was calculated from the date of surgery to the date of death. Patients who were alive were censored at their last follow-up.

Results

Among 102 patients who underwent gastrectomy, majority underwent subtotal gastrectomy 79.4% (n = 81) and D2 dissection was performed in 91% (n = 93). All patients had R0 resection except in 2 patients in whom distal margin was microscopically positive. Multivisceral resection was performed in 4 (3.9%) patients which included distal pancreatosplenectomy, liver resection, jejunal resection, and transverse colon resection. Most of the patients had antropyloric tumors 79% (n = 81) and most common histological type was adenocarcinoma (93%) with poor (46%) or moderately (44%) differentiated grade. Three-fourth (73.5%) of the patients had nodal metastasis. Mean number of nodes harvested was 19. Lymphovascular invasion seen in 32% (n = 33) of patients. Adjuvant chemotherapy was considered for pT3 and node positive tumors (n = 93). Out of 93 patients, only 75 (80.5%) could complete all planned 6 cycles of chemotherapy and the rest 19.5% (n = 18) failed to complete chemotherapy due to adverse effects (n = 9), deaths (n = 3), and 6 patients were lost to follow-up. All the baseline characteristics are tabulated in Table 1.

Table 1.

Baseline characteristics

Variables	N (%)
Total patients	102
Median age	60 years (range 28–80)
Sex
Male	72 (70.6)
Female	30 (29.4)
Location of tumor
Antropyloric	81 (79.4)
Body	19 (18.6)
Cardia	2 (1.9)
Diffuse	0
Type of surgery
Subtotal gastrectomy	81 (79.4)
Total gastrectomy	21 (20.6)
Lymphadenectomy
D1	6 (5.9)
D1 +	3 (2.9)
D2	93 (91.1)
Resection
R0	100 (98)
R1	2 (2)
R2	0
Histopathological types
Mucinous	1 (0.9)
Signet ring	5 (4.9)
NOS	95 (93.1)
NHL	1 (0.9)
Grade
Well differentiated	10 (9.8)
Moderately differentiated	45 (44.1)
Poorly differentiated	47 (46)
T stage
T1	1 (0.9)
T2	21 (20.6)
T3	23 (22.5)
T4	57 (55.9)
N stage
N0	27 (26.5)
N1	33 (32.4)
N2	18 (17.6)
N3	24 (23.5)
Stage
I	9 (8.8)
II	33 (32.3)
III	48 (47)
IV	12 (11.8)
Mean nodes sampled	19.3
Lymphovascular invasion (LVSI)	33 (32.4)
Adjuvant therapy
No	9 (8.8)
Yes	93 (91.2)
Completed	75 (80.5)
Incomplete (not tolerate/death)	12 (13)
Defaulted/not known	6 (6.5)

Postoperative complications occurred in 17 (16.7%) patients, most common was wound infection (6.1%) followed by anastomotic leak (5.9%) and other the complications were chest infection (n = 3), bleeding (n = 1), mesenteric ischemia (n = 1), and CVA (n = 1). Relaporotomy was performed in 2 patients due to intraabdominal bleed and mesenteric ischemia. Average hospital stay was 8 days for subtotal gastrectomy and 11 days for total gastrectomy with total overall average hospital stay of 9.5 days. Thirty-day mortality rate was 2.9% (n = 3), causes of death being sepsis with multiorgan failure (Table 2).

Table 2.

Postoperative outcomes

Postoperative complications
Anastomotic leak
Subtotal gastrectomy	5 (6.1%)
Total gastrectomy	0
Bleeding (relaporotomy)	1
Mesenteric ischemia (relaporotomy)	1
Wound morbidity	6 (5.9%)
Chest infection	3
CVA	1
	Overall 16.7%
Average hospital stay (days)
Subtotal gastrectomy	8
Total gastrectomy	11 Average 9.5 days
30-day mortality
Subtotal gastrectomy	3 (3.7%)
Total gastrectomy	0
	Overall 2.9%

Among 102 patients, 48 (47%) patients had recurrences. Standard risk factors for recurrence like tumor grade, LVSI, tumor stage, and nodal stage were analyzed for significance. Only nodal stage (p = 0.0051) and LVSI (p ≤ 0.0001) were found to be the significant risk factors associated with recurrences. Tumor stage (p = 0.6363) and tumor grade (p = 0.1633) did not show significance for recurrence (Table 3). A total of 58 (56.8%) patients reached the end point of death. Most common cause of death was disease related in 79.3% (n = 46) followed by natural in 10.3% (n = 6) and the others were postoperative complications in 5.3% (n = 3) and chemo related in 5.3% (n = 3). Risk factors for disease related death like tumor grade, LVSI, tumor stage, and nodal stage were analyzed for significance. Similar to recurrences, only LVSI (p ≤ 0.0001) and nodal stage (p = 0.0441) were found to be significant risk factors associated with deaths. Tumor grade and T stage did not show any significance for deaths (Table 3).

Table 3.

Survival and risk factors

Variables	Recurrences		Deaths
	N = 46 (45%)	p value (Fisher’s exact test)	N = 58 (56.8%)	p value (Fisher’s exact test)
Grade
Well/mod differentiated	21 (45.7%)	0.1633	27 (46.6%)	0.1096
Poorly differentiated	25 (54.3%)		31 (53.4%)
LVSI	29 (63%)	< 0.00001	32 (55.1%)	< 0.00001
T stage
T1/T2	11 (23.9%)	0.6353	13 (22.4%)	1
T3/T4	35 (76.1%)		45 (77.6%)
N stage
N0/N1	20 (43.5%)	0.0051	29 (50%)	0.0441
N2/N3	26 (56.5%)		29 (50%)

Median overall survival for all stages was 23 months (95% CI of 18–28) and median disease-free survival was 26 months (95% CI 20–30). Stage-specific survival rates for stages I, II, III, and IV were 60, 27, 18, and 10 months, respectively (Fig. 1). Comparison of survival curves by log rank test showed significance between stage groups (p = 0.0083). Median follow-up time calculated by reverse Kaplan–Meier curve method was 32 months (95% CI of 24–71) (Fig. 2). Overall survival rates for 2nd, 3rd, and 4th year were 31.72%, 22.52%, and 7.43%, respectively.

Fig. 1.

Kaplan–Meier analysis of stagewise survival

Fig. 2.

Kaplan–Meier curve for disease-free interval

Discussion

Our study included 102 gastric cancer patients who underwent gastrectomy over a period of 6 years. Most of the tumors were locally advanced type affecting the antropyloric region, and majority were pT2 and above or node positive tumors. Adjuvant chemotherapy completion rate for all planned 6 cycles of chemotherapy was 80%. Median overall survival was 23 months. Two-year and 3-year overall survival rates were 31% and 22%, respectively. Comparison of stagewise survival as compared by log rank test showed significance for stage groups. An analysis of risk factors like tumor grade, LVSI, tumor stage, and nodal stage showed only LVSI and nodal stage to be significant for both recurrences and deaths.

In the present study, median age of presentation was 60 years which is about 6–8 years later than other Indian studies conducted at Tata Memorial Hospital (TMH) [13], Christian Medical College (CMC) Vellore [6], and Cancer Institute Adyar (CIA) Chennai [14]. Compared to American cancer registries, this is about 8 years earlier [15]. Males were most commonly affected about 70% which is comparable to other Indian studies [6] [13] and is in line with data from the American cancer surveillance database, Surveillance, Epidemiology, and End Results (SEER) [16]. Eighty-one percent of tumors were antropyloric growths compared to 61% in CMC study and 75% in the TMH study. However, in western population, the trend is more towards proximal tumors [17].

A total of 91% (n = 93) patients presented at locally advanced stage (II, III, IV) against 90% in CMC study [6] keeping in lines with trends from regions where screening is not done such as America, Europe, and China [18] [19] [20]. More than 90% of tumors in our series were moderately or poorly differentiated tumors showing similarity with other Indian studies [6] [13]. Mean number of nodes harvested was 19.3 versus 13 nodes compared to CMC series. However the nodal harvest was only comparable to 18 from the TMH series.

Overall postoperative morbidity observed in our series was 16.7% comparable to 16.7% in CIA series. Major postoperative complication was anastomotic leak observed in 6.1% versus 6.8% in CMC series, 0.3% in CIA series and 30-day mortality rate of 2.9% as against 4.2% in CMC series, 1.5% in TMH series, and 1.8% in CIA series. Average hospital stay was 9.5 days versus 15.5 days in CMC, 13.5 days in TMH series, and 15.6 days in CIA series which is much lower in our series [6] [14] [13]. The present study showed an acceptable morbidity and mortality rates comparable to larger Indian studies conducted at CIA, CMC, and TMH.

Median disease-free survival was 28 months with a median follow-up period of 32 months (range 24–71 months). Stage-specific survival rates for stages I, II, III, and IV were 60, 27, 18, and 10 months, respectively. Two-year, three-year, and four-year survival rates were 31.7%, 22.5%, and 7.4% respectively in our series whereas 3-year and 5-year overall survival was 67.9% and 59.3% in CMC series and 58% 5-year survival in TMH series which showed much better survival rates than our study. The reported 5-year survival in India is between 5 and 15% [7]. Our data was not mature to calculate the 5-year survival rates. Higher 5-year survival rates in the TMH series may be due to usage of neoadjuvant chemotherapy in 68.6% of patients in TMH and the study population represent mostly North India population [13]. CMC series subset analysis showed slightly lower 5-year survival of 52% in Southern India compared to North-east 66% and North India 59% [6]. Lower survival rates in our study may be due to older age of presentations compared to other Indian studies, inability to complete all planned adjuvant chemotherapy in 20% of patients, and that all our patients underwent primary surgery. The lack of information regarding the impact of perioperative chemotherapy on survival in our population needs to be explored.

In our series, we noticed 45% recurrences against 44% in CMC series and multiple factors analyzed by univariate analysis showed only lymphovascular invasion (LVSI) (p = 0.0001) and higher nodal burden (p = 0.0051) significant which was consistent with CMC series. Tumor grade (p = 0.1633) and tumor stage (p = 0.6353) did not show significance with recurrence. Our cohort had 56% deaths against 26% in CMC series among which higher nodal burden (p = 0.0441) and LVSI (p = 0.0001) were found to be significant by univariate analysis and this was comparable to Tata Memorial [13] and Christian Medical Colleges studies [6]. We have experienced higher rates of disease-related deaths which indicates the need to change from primary surgery followed by chemotherapy from approach towards perioperative or neoadjuvant strategies.

Major limitations of our study are retrospective nature and moderate sample size. Adjuvant chemotherapy completion rate of only 80% and non-implementation of perioperative chemotherapy could have impacted survival. Available data in this series is not mature to calculate 5-year survival rate.

Conclusion

Gastric cancer in south India continues to have dismal prognosis with 2-year survival rate of 31%. Results from our study indicate that most of the patients present in locally advanced stages with several high risk histological features. More disease-related deaths in our cohort direct us towards the need to change and adopt to emerging standard of care approaches like perioperative and neoadjuvant chemotherapy strategies to improve survival outcomes in the future.

Declarations

Conflict of Interest

The authors declare no competing interests.