Prognostic Factors for Survival in Node-Negative Gastric Cancer Patients Who Underwent Curative Resection

Introduction

Gastric cancer remains one of the most frequent malignancies and is the third leading cause of cancer-related death worldwide. Although the survival of patients with gastric cancer has been significantly improved due to advances, such as more reasonable lymph node dissections, chemotherapy regimens, and targeted therapy, the prognosis of gastric cancer patients is still not optimistic because of postoperative recurrence and metastasis (1). As an important cause of tumor relapse, the presence of lymph node metastases is a well-known predictor of poor outcome (2). Despite the fact that patients with node-negative gastric cancer have better overall survival (OS) rates compared to those with nodal metastases, a proportion of patients remain at risk of recurrence. Therefore, there is an urgent need to identify adverse prognostic factors. While many previous studies have clarified the relationship between the depth of tumor invasion and OS (3, 4), the role of other factors, including tumor size, tumor location, and the presence of lymphovascular invasion, remains uncertain (5, 6). This information could be significant for decisions about the administration of adjuvant therapy. Importantly, many node-negative patients with an increased risk of recurrence do not undergo chemotherapy. Therefore, if we identify useful factors that more accurately predict prognosis, more specialized strategies could be prepared for follow-up examination after surgery. These strategies will optimize existing treatment options and enable patients with node-negative gastric cancer to obtain better curative outcomes.

Methods

Results

Clinicopathologic characteristics

Of the 302 patients, there were 217 males and 85 females. Patient age ranged from 27 to 85 years old with a median age of 60 years. In total, 35 patients (11.6%) had a tumor located in the upper third, 63 (20.9%) in the middle third, and 204 (67.5%) in the lower third of the stomach. The median tumor size was 3.7 cm (0.5–13 cm). The patients were divided into two groups according to tumor size, a <4 cm group and a ⩾4 cm group. From stage T1 to T4, there were 131 (43.4%), 60 (19.9%), 32 (10.6%), and 79 (26.1%) patients, respectively. Highly differentiated tumors were observed in 177 (58.6%) patients, while poorly differentiated tumors were observed in the remaining 125 (41.4%) patients. Only 15 (5.0%) patients had lymphovascular invasion. In node-negative patients with AGC, 98 (57.3%) patients underwent postoperative chemotherapy, while 73 (42.7%) did not. Patients’ data are provided in Table 1.

OPEN IN VIEWER

Table 1

Clinicopathologic characteristics of node-negative gastric cancer patients.

Variable	N (%)	5-year survival rate (%)	p-value a
Gender			0.627
Male	217 (71.9)	62.0
Female	85 (28.1)	70.6
Age (years)			0.057
⩽60	128 (42.4)	71.9
>60	174 (57.6)	58.9
Tumor location			0.254
Upper 1/3	35 (11.6)	26.9
Middle 1/3	63 (20.9)	60.3
Lower 1/3	204 (67.5)	69.0
Tumor size (cm)			0.000
<4	197 (65.2)	75.4
⩾4	105 (34.8)	45.7
Type of gastrectomy			0.573
Proximal subtotal gastrectomy	33 (10.9)	43.6
Distal gastrectomy	220 (72.9)	66.4
Total gastrectomy	49 (16.2)	62.4
Depth of invasion			0.000
T1	131 (43.4)	81.7
T2	60 (19.9)	63.0
T3	32 (10.6)	59.4
T4	79 (26.1)	34.1
Histologic classification			0.002
Highly differentiated	177 (58.6)	71.8
Poorly differentiated	125 (41.4)	53.2
Lymphovascular invasion			0.134
Yes	15 (5.0)	46.7
No	287 (95.0)	65.9
Adjuvant chemotherapy b			0.138
Yes	98 (57.3)	54.1
No	73 (42.7)	45.2

a

p-values are generated through log-rank tests.

b

Advanced gastric cancer without lymph node metastasis.

Bold values: p<0.05.

Survival

The median OS was 108 months. The overall 1-, 3-, and 5-year survival rates were 80.0%, 69.0%, and 63.0%, respectively. Univariate analysis showed that gender, age, operation procedure, tumor location, presence of lymphovascular invasion, and postoperative chemotherapy had no relation to 5-year OS rates (Table 1). However, the following variables were associated with improved OS: smaller tumor size (<4 cm, p < 0.001), highly differentiated pathological classification (p = 0.002), and low tumor T stage (p < 0.001). Multivariate analysis of these variables revealed that they were the only independent prognostic factors for better OS. Data are provided in Table 2. The survival curves of patients according to these three factors are presented in Figs 1 to 3.

OPEN IN VIEWER

Table 2

Hazard ratio by age, tumor size, depth of invasion, histologic type, lymphovascular invasion, and adjuvant chemotherapy calculated by the Cox proportional hazard model.

Variable	HR	95.0% CI	p-value
Age (years)	1.114	0.686–1.810	0.663
Tumor size	1.706	1.064–2.738	0.027
Depth of invasion	1.251	1.036–1.510	0.020
Histologic classification	1.616	1.031–2.534	0.037
Lymphovascular invasion	0.953	0.399–2.277	0.914
Adjuvant chemotherapy	0.834	0.529–1.316	0.435

HR: hazard ratio; CI: confidence interval.

OPEN IN VIEWER

Fig. 1.

Survival curves were generated by the Kaplan–Meier method for gastric cancer patients who were divided into four groups, pT1 (n = 131), pT2 (n = 60), pT3 (n = 30), or pT4 (n = 79), according to the depth of tumor invasion. The log-rank test was used to analyze statistical significance.

OPEN IN VIEWER

Fig. 2.

Survival curves were generated by the Kaplan–Meier method for gastric cancer patients who were divided into two groups, differentiation group (n = 175) and undifferentiation group (n = 125), according to histological types of tumor.

OPEN IN VIEWER

Fig. 3.

Survival curves were generated by the Kaplan–Meier method for gastric cancer patients who were divided into two groups, <4 cm group (n = 197) and ⩾4 cm group (n = 103), according to the tumor size.

Discussion

Identifying high-risk factors for the recurrence of gastric cancer helps to accurately distinguish between high-risk and low-risk patients. In this context, individual interventions could be performed to maximize individualized therapy effectiveness. Lymph node metastases are generally recognized as one of the most significant predictors of poor outcome, which is closely related to tumor recurrence. However, this fact does not suggest that node-negative patients are safe. It was reported that the 5-year OS varied from 72% to 97.1% in node-negative patients. Although this survival was significantly better than node-positive patients (8), some patients still experienced recurrence and metastasis.

A growing number of researchers have reported that the depth of tumor invasion and lymph nodes metastases were crucial prognostic factors in gastric cancer. In particular, the depth of tumor invasion has been commonly regarded as an independent predictor related to the survival rate, which is independent of lymph node metastasis (9, 10). The consensus was that the depth of tumor invasion was related to local recurrence. The probability of tumor cells shedding into the peritoneal cavity is increased when the depth of invasion is beyond the serosa and can simultaneously promote tumor cell metastasis. Kim et al. (11) argued that the prognosis of patients would be aggravated by lymph node micrometastasis, which was closely related to advanced T stage. We also found that advanced T stage was a negative predictor; patients with early disease had a longer OS than those patients with an advanced T stage.

Tumor size is another predictor of poor outcome in patients with lung cancer, thyroid cancer, and breast cancer, and tumor size is included in the tumor-node-metastasis (TNM) staging system (12). However, the relationship between tumor size and gastric cancer patient prognosis is still controversial. Using univariate survival analyses, Kim et al. (11) identified tumor size as a significant prognostic factor affecting 5-year OS. However, Baiocchi et al. (3) reported that some parameters associated with a better survival excluded tumor size. In our study, the 5-year OS of patients with tumor <4 cm (75.4%) was significantly better than those patients with tumor ⩾4 cm (45.7%). Multivariate analysis also confirmed that tumor size was also useful for predicting prognosis. Yokota et al. (13) found that tumor size was a predictor of survival according to univariate, but not multivariate, analysis. Still, other studies have reported that tumor size is an independent prognostic factor (14). The effects of invasion depth and lymph node metastasis on tumor size make it difficult to identify the most important prognostic factors. Thus, to eliminate lymph node effects, Xu et al. (15) analyzed the effects of tumor size on prognosis in patients with node-negative gastric cancer. The results showed that tumor size and depth of tumor invasion were independent prognostic factors for patient survival. With increasing tumor size, the probability of micrometastases migrating from the tumor through the lymph vessels is also increased (4). At the same time, tumors that penetrated over the submucosa would be more likely to invade the blood vessels, thus enhancing the likelihood of metastatic disease.

In addition, histological classification was also considered a prognostic factor, but with quite different conclusions. Seshadri et al. (16) reported that the tumor histology did not correlate significantly with survival. Nevertheless, Kim et al. (11) published a report in which histologic differentiation was found to be a statistically significant prognostic factor affecting 5-year OS. Our analysis showed that the 5-year OS of patients with poorly differentiated gastric cancer (53.2%) was significantly lower than that of those patients with highly differentiated gastric cancer (71.8%). We believe that more malignant poorly differentiated gastric cancer tumor cells were more aggressive and more likely to develop peritoneal metastasis.

At the same time, this study indicated that the patient’s gender, age, surgical method, tumor location, lymphovascular invasion, and postoperative chemotherapy were not independent predictors of prognosis. For patients with AGC, surgery combined with chemotherapy or radiotherapy was considered to be an effective treatment option (17, 18), especially in countries where D2 resection is the standard surgical procedure. Numerous meta-analyses have also suggested a survival benefit associated with adjuvant chemotherapy in patients with AGC (19, 20). However, for patients with node-negative gastric cancer, the role of chemotherapy was not recognized in some previous studies (21, 22). The results of our study showed that the 5-year OS of the chemotherapy group and the non-chemotherapy group were 54.1% and 45.2%, respectively, and Log-rank univariate survival analysis showed these differences were not statistically significant (p = 0.138).

This study has some limitations. First, our study was performed on 302 gastric cancer patients, who were negative for lymph node metastasis. Hence, the 5-year OS (63.0%) was slightly less than that reported in the literature. Second, the objective detection rates of early gastric cancer in China were low. Consequently, some patients with early gastric cancer who underwent endoscopic submucosal dissection (ESD) treatment instead of gastrectomy were excluded. As such, further studies with multiple centers that enroll more patients are needed to confirm our results.

Conclusion

This study demonstrated that large tumor size, poorly differentiated histologic classification, and deep tumor invasion depth were associated with an unfavorable 5-year OS rate for patients with node-negative gastric cancer. Such cases have an increased risk of tumor recurrence and should be considered for individual follow-up and radical treatment.