Journal of Gastrointestinal Surgery

, 15:1783

Predictable Factors for Lymph Node Metastasis in Early Gastric Cancer—Analysis of Single Institutional Experience

Authors

  • Man Sup Lim
    • Department of SurgeryHallym University Sacred Heart Hospital
    • Department of SurgeryHallym University Hangang Sacred Heart Hospital
  • Hyoungjune Im
    • Department of Occupation and Environmental MedicineHallym University Sacred Heart Hospital
  • Byung Seup Kim
    • Department of SurgeryHallym University Sacred Heart Hospital
  • Mi Yeol Lee
    • Department of SurgeryHallym University Sacred Heart Hospital
  • Jang Yong Jeon
    • Department of SurgeryHallym University Sacred Heart Hospital
  • Dae Hyun Yang
    • Department of SurgeryHallym University Kangnam Sacred Heart Hospital
  • Bong Hwa Lee
    • Department of SurgeryHallym University Sacred Heart Hospital
Original Article

DOI: 10.1007/s11605-011-1624-5

Cite this article as:
Lim, M.S., Lee, H., Im, H. et al. J Gastrointest Surg (2011) 15: 1783. doi:10.1007/s11605-011-1624-5

Abstract

Introduction

Prediction of lymph node metastasis in early gastric cancer (EGC) is very important to decide treatment strategies preoperatively. The aim of this study was to evaluate factors that predict the presence of lymph node metastasis and to indentify the differences between mucosal and submucosal gastric cancers.

Methods

A total of 376 patients with EGC who underwent gastrectomy from March 1999 through December 2007 were retrospectively identified. The clinopathological factors and biological markers (p53, Ki67) were analyzed.

Results

The rate of lymph node metastasis was 9.6% (mucosal cancer 2.8%, submucosal cancer 18.4%). Tumor size, depth of invasion, macroscopic type, and lymphovascular invasion were related to lymph node metastasis in EGC. When the carcinomas were confined to the mucosal layer, tumor size and lymphovascular invasion showed significant correlation with lymph node metastasis. On the other side, macroscopic type and lymphovascular invasion were association with lymph node metastasis in submucosal carcinoma.

Conclusion

The risk factors for lymph node metastasis in EGC are quite different depending on depth of tumor invasion. To predict lymph node metastasis in EGC, it is recommended that distinct assessment according to individual situation should be clearly established.

Keywords

Early gastric cancerLymph node metastasisp53KI67

Introduction

Early gastric cancer (EGC) is defined as when the tumor invasion is confined to the mucosa and submucosa, irrespective of the presence of a lymph node metastasis.1 The rate of EGC has been increasing worldwide with advanced diagnostic techniques and mass screening programs. The presence of lymph node metastasis is the most important prognostic factor for patients with EGC. Lymph node metastases are present in only 3–5% of patients with mucosal cancer versus 10–25% of those with submucosal cancer.26 Several clinocopathological factors, such as the tumor size, depth of invasion, histological type, and lymphatic and vascular invasion, are known to correlate with lymph node metastasis.68 Recently, several biological markers have been reported as useful predictors for lymph node metastases in gastric cancer. The p53 tumor-suppressor gene acts by modulating cell proliferation via G1 arrest checkpoint cell cycle.9 Mutation of p53 gene is one of the most frequent genetic lesions associated with gastric cancer and correlated with lymph node metastases in early gastric cancer.10 Ki67 is a nuclear non-histone protein and universally expressed among proliferating cells and absent in quiescent cells. Although little is known about the exact function of protein in cell division, Ki67 expression is evaluated as a marker of proliferation.11 Ki67 expression may be useful predictor for lymph node metastases in gastric cancer with submucosa invasion.12

Because the prognosis of patients with EGC has improved with radical gastrectomy, the treatment strategies for EGC have been focused on the improvement of quality of life.13 Although radical gastrectomy including lymph node dissection has been recognized as the standard surgical operation for EGC, unnecessary extended surgery could be avoided in patients with EGC without lymph node metastasis. Many retrospective studies on EGC have established an indication for minimal invasive surgery without lymph node dissection, such as endoscopic mucosal resection (EMR) or endoscopic submucosal dissection (ESD).

The purpose of this study was to evaluate the factors that can be predicted the presence of lymph node metastasis and to identify the difference between mucosal and submucosal gastric cancers in point of lymph node metastasis. Although this process is retrospective study, authors focused on the factors that could be evaluated preoperatively.

Material and Methods

Patients

This study enrolled 376 early gastric cancer cases; those were pathologically proven after gastrectomy with lymph node dissection at Hallym University Sacred Heart Hospital between March 1999 and December 2007. The standard operative procedures were performed for EGC: distal subtotal or total gastrectomy depending on the location of gastric cancer with D2 or more extended lymphadenectomy. Patient characteristics including age and sex were collected, and information on tumor size, depth of invasion, gross appearance, histological type, and lymphovascular invasion was also retrieved from medical records.

The depth of tumor invasion was classified as mucosa and submucosa carcinoma. The maximum diameter of tumor was recorded as tumor size. The carcinomas were classified into four macroscopic types: elevated (type I or IIa), flat (IIb), depressed (IIc or III), and mixed type which include both elevated and depressed. Tumor histology was classified into two groups: the differentiated group, which included papillary adenocarcinoma and well- or moderately differentiated adenocarcinoma, and the undifferentiated group, which included poorly or undifferentiated adenocarcinoma, mucinous, and signet ring cell carcinoma. This study was approved by the Institutional Review Board of the Hallym University Sacred Heart Hospital, and informed consent was obtained from all individuals.

Immunohistochemistry

For immunohistochemistry, the avidin–biotin complex method was performed on 4-μm-thick tissue sections. The sections were deparaffinized with xylene for 15 min and treated in a microwave oven using 0.01 mol/l citrate buffer (pH 6.0) for 30 min and incubated with monoclonal antibody against recombinant human p53 (Zymed, San Francisco, CA, USA) and Ki67 (Zymed, San Francisco, CA, USA). In each experiment, both positive and negative controls were used. For the staining of p53 and Ki67, only nuclear expression was recognized as specific immunostaining. The immunohistochemical score was estimated by the percentage of the immunoreactive cells: A score of 0 indicates no positive cells or essentially none (0 or lesser than 1%); a score of 1 indicates some positive cells (1–10%); a score of 2 indicates well-defined area of positive cells (11–25%); a score of 3 indicates extensive area of positive cells (26–50%); a score of 4 indicates almost every cell stained (>50%). Slides were examined by one investigator who was blinded to clinicopathological data.

Statistical Analysis

Associations between lymph node metastasis and clinicopathological parameters were analyzed using the chi-square test (or Fisher’s exact test when appropriate). Multivariate logistic regression analyses were used to assess the associated factors for lymph node metastasis. All statistical analyses were conducted using R (version 2.12.1; R foundation for Statistical Computing, Vienna, Austria). Statistical significance was established at p < 0.05.

Results

Relationship Between Clinicopathological Findings and Lymph Node Metastasis in All Patients with EGC

The study includes 238 men and 138 women and involved 213 mucosal carcinomas and 163 submucosal carcinomas. The mean age was 59.2 years (range, 29–81 years). The mean number of retrieved lymph nodes was 26.0. Among the 376 patients with EGC, 36 (9.6%) were histologically shown to have lymph node metastasis and 340 (90.4%) had no lymph node metastasis. Lymph node metastasis was associated with tumor size, depth of invasion, macroscopic type, and lymphovascular invasion (Table 1). A tumor larger than 2 cm and submucosal invasion were association with lymph node metastasis (p = 0.0002, p < 0.0001). In elevated type of carcinoma, the rate of lymph node metastasis was higher (p = 0.0018). In the presence of lymphatic or vascular invasion, the rate of lymph node metastasis was also higher (p < 0.0001). There was no significant difference in gender, age, tumor location, histology, and immunoreactivity for p53 or Ki67.
Table 1

Relationship between clinicopathological findings and lymph node metastasis in patients with early gastric cancer (376 patients)

Variables

Lymph node metastasis

p value

Negative (n = 340)

Positive (n = 36)

Gender

  

0.169

 Male

219 (92.0)

19 (8.0)

 

 Female

121 (87.7)

17 (12.3)

 

Age (years)

  

0.727

 <60

154 (91.1)

15 (8.9)

 

 ≥60

186 (89.9)

21 (10.1)

 

Tumor location

  

0.518

 Upper 1/3

26 (92.9)

7 (7.1)

 

 Middle 1/3

121 (92.4)

10 (7.6)

 

 Lower 1/3

193 (88.9)

24 (11.1)

 

Tumor size (cm)

  

0.0002

 <2

115 (98.2)

2 (1.7)

 

 ≥2

225 (86.9)

34 (13.1)

 

Depth of invasion

  

<0.0001

 Mucosa

207 (97.2)

6 (2.8)

 

 Submucosa

133 (81.6)

30 (18.4)

 

Macroscopic type

  

0.0018

 Elevated

19 (70.4)

8 (29.6)

 

 Flat

100 (94.3)

6 (5.7)

 

 Depressed

201 (90.5)

21 (9.5)

 

 Mixed

20 (95.2)

1 (4.8)

 

Histology

  

0.500

 Differentiated

190 (91.3)

18 (8.7)

 

 Undifferentiateda

150 (89.3)

18 (10.7)

 

Lymphovascular invasion

  

<0.0001

 Absent

320 (93.8)

21 (6.2)

 

 Present

20 (57.1)

15 (42.9)

 

P53

  

0.275

 0–2 weak

148 (91.4)

14 (8.6)

 

 3–4 strong

99 (83.2)

20 (16.8)

 

Ki67

  

0.397

 0–2

96 (90.6)

10 (9.4)

 

 3–4

165 (86.8)

25 (13.2)

 

Relationship Between Clinicopathological Findings and Lymph Node Metastasis in Mucosal Gastric Carcinoma

Among the 213 patients with mucosal gastric cancer, only six (2.8%) were histologically shown to have lymph node metastasis. When the carcinomas were confined to the mucosal layer, tumor size and lymphovascular invasion showed significant correlation with lymph node metastasis (p = 0.039, p = 0.002) (Table 2). However, there was no significant difference in gender, age, tumor location, macroscopic type, histology, and immunoreactivity for p53 or Ki67.
Table 2

Relationship between clinicopathological findings and lymph node metastasis in patients with mucosal gastric carcinomas (213 patients)

Variables

Lymph node metastasis

p value

Negative (n = 207)

Positive (n = 6)

Gender

  

0.560

 Male

128 (97.7)

3 (2.3)

 

 Female

79 (96.3)

3 (3.7)

 

Age (years)

  

0.106

 <60

97 (95.1)

5 (4.9)

 

 ≥60

110 (99.1)

1 (0.9)

 

Tumor location

  

0.577

 Upper 1/3

12 (92.3)

1 (7.7)

 

 Middle 1/3

71 (97.3)

2 (2.7)

 

 Lower 1/3

124 (97.6)

3 (2.4)

 

Tumor size (cm)

  

0.039

 <2

91 (100)

0 (0)

 

 ≥2

116 (95.1)

6 (4.9)

 

Macroscopic type

  

0.982

 Elevated

10 (100)

0 (0)

 

 Flat

79 (96.3)

3 (3.7)

 

 Depressed

111 (97.4)

3 (2.6)

 

 Mixed

7 (100)

0 (0)

 

Histology

  

0.204

 Differentiated

125 (98.4)

2 (1.6)

 

 Undifferentiateda

82 (95.4)

4 (4.6)

 

Lymphovascular invasion

  

0.002

 Absent

206

4

 

 Present

1

2

 

 P53

  

0.159

 0–2 weak

87 (94.6)

5 (5.4)

 

 3–4 strong

52 (100)

0 (0)

 

 Ki67

  

0.605

 0–1

38 (95.0)

2 (5.0)

 

 2–4

111 (97.4)

3 (2.6)

 

Relationship Between Clinicopathological Findings and Lymph Node Metastasis in Submucosal Gastric Carcinoma

Among the 163 patients with submucosal gastric cancer, 30 (18.4%) were histologically shown to have lymph node metastasis. When the carcinomas were confined to the submucosal layer, macroscopic type and lymphovascular invasion were association with lymph node metastasis (Table 3). In the elevated type of submucosal carcinoma, the rate of lymph node metastasis was higher than that of others groups (p = 0.022). In the presence of lymphatic or vascular invasion, the rate of lymph node metastasis was higher (p = 0.0008). However, no significant association was found between tumor size and lymph node metastasis in submucosal carcinoma.
Table 3

Relationship between clinicopathological findings and lymph node metastasis in patients with submucosal gastric carcinomas (163 patients)

Variables

Lymph node metastasis

p value

Negative (n = 133)

Positive (n = 30)

Gender

  

0.119

 Male

91 (85.1)

16 (14.9)

 

 Female

42 (75.0)

14 (25.0)

 

Age (years)

  

0.413

 <60

57 (85.1)

10 (14.9)

 

 ≥60

76 (79.2)

20 (20.8)

 

Tumor location

  

0.182

 Upper 1/3

14 (93.3)

1 (6.7)

 

 Middle 1/3

50 (86.2)

8 (13.8)

 

 Lower 1/3

69 (76.7)

21 (23.3)

 

Tumor size (cm)

  

0.170

 <2

24 (92.3)

2 (7.7)

 

 ≥2

109 (79.6)

28 (20.4)

 

Macroscopic type

  

0.022

 Elevated

9 (52.9)

8 (47.1)

 

 Flat

21 (87.5)

3 (12.5)

 

 Depressed

90 (83.3)

18 (16.7)

 

 Mixed

13 (92.9)

1 (7.1)

 

Histology

  

0.659

 Differentiated

65 (80.3)

16 (19.7)

 

 Undifferentiateda

68 (82.9)

14 (17.1)

 

Lymphovascular invasion

  

0.0008

 Absent

114 (87.0)

17 (13.0)

 

 Present

19 (59.4)

13 (40.6)

 

P53

  

0.208

 0–2 weak

61 (83.6)

12 (16.4)

 

 3–4 strong

47 (73.4)

17 (26.6)

 

Ki67

  

0.381

 0–2

34 (73.9)

12 (26.1)

 

 3–4

78 (81.2)

18 (18.8)

 

Multivariate Analysis of Risk Factors

Multivariate logistic regression analyses were used to assess the associated factors for lymph node metastasis. In multivariate logistic regression for lymph node metastasis, the independent factors were age (dichotomized as <60, ≥60), tumor size (<2 cm, ≥2 cm), lymphovascular invasion (absent, present), and macroscopic type (categorized as elevated, flat, depressed, mixed). In mucosal gastric carcinoma group, the multivariate logistic regression could not be performed appropriately because of the small number of patients with lymph node metastasis (n = 6). On the other hand, lymphovascular invasion (relative risk 4.57, 95% confidence interval (CI) 1.74–12.24, p = 0.0021) and macroscopic type were independently associated with lymph node metastasis in submucosal gastric carcinoma (Table 4).
Table 4

Multivariate analysis of risk factors for lymph node metastasis

Variables

Mucosal gastric carcinoma

Submucosal gastric carcinoma

Adjusted odds ratio (95% CI)

p value

Adjusted odds ratio (95% CI)

p value

Age (<60, ≥60)

0.15 (0.01–1.25)

0.1282

1.07 (0.42–2.75)

0.8923

Tumor size (<2 cm, ≥2 cm)

8.63 × 107 (0–∞)

0.9950

3.31 (0.78–23.68)

0.1510

LVI (absent, present)

72.2 (4.08–3493.97)

0.0077

4.57 (1.74–12.24)

0.0021

Macroscopic type (elevated)

 Flat

7.44 × 107 (0–∞)

0.9983

0.11 (0.02–0.57)

0.0122

 Depressed

2.61 × 107 (0–∞)

0.9984

0.17 (0.05–0.59)

0.0054

 Mixed

0.58

0.9999

0.05 (0.00–0.37)

0.0119

This table demonstrates risk factors for lymph node metastasis in early gastric cancer. As shown, multivariate analysis in patients with mucosal gastric carcinoma is not fully appropriate because the cases with lymph node metastasis were too small

LVI lymphovascular invasion

Discussion

Gastric cancer is the most prevalent cancer in Korea, which accounted for 18.3% of the whole cancer cases.14 According to a report issued by the Korean Gastric Cancer Association, the proportion of patients with early stage gastric cancer has increased from 28.6% in 1995 to 32.8% in 1999 and 47.4% in 2004.15 Lymph node metastasis is known to be one of the major negative prognostic factors for gastric cancer. The average 5-year survival rate in patients with EGC is over 90%, and it is up to 94.2% in patients without lymph node metastasis.16 In this study, the rate of lymph node metastasis in EGC was 9.6%, and the risk factors of lymph node metastasis were tumor size, depth of invasion, macroscopic type, and lymphovascular invasion. When EGC was subdivided into mucosal and submucosal carcinomas, the risk factors of lymph node metastasis were quite different from each other. When the carcinomas were confined to the mucosal layer, the incidence of lymph node metastasis is 2.8%. It is similar to other reports.6,17 Tumor size and lymphovascular invasion showed significant correlation with lymph node metastasis in mucosal carcinoma. On the other side, the incidence of lymph node metastasis is higher in submucosal carcinoma (18.4% vs. 2.8%). It is also similar to other reports.18 In submucosal carcinoma, macroscopic type and lymphovascular invasion showed significant correlation with lymph node metastasis. Tumor size was not risk factor of lymph node metastasis in submucosal carcinoma.

Many studies have been carried out to evaluate the risk of lymph node metastasis in gastric cancer. Tumor size, depth of invasion, histological type, gross appearance, and presence of lymphatic or vascular invasion were related to lymph node metastasis in EGC.6,19,20 When early gastric cancer was subdivided into mucosal and submucosal carcinomas, the risk factors of lymph node metastasis were different each other. Li et al. reported that the tumor size and lymphovascular invasion were independent risk factors for lymph node metastasis in case of intramucosal undifferentiated EGC.21 Song et al. reported that histological differentiation, increased submucosal vascularity, and invasion of tumor cells into the muscularis mucosae were correlated with the lymph node status of intramucosal gastric carcinoma.22 Park et al. reported that tumor size and depth of invasion were significantly correlated with lymph node metastasis,7 and An et al. also reported that tumor size and lymphatic involvement were independent risk factors for lymph node metastasis in EGC with submucosal invasion.8 Shen et al. reported that histological classification, macroscopic type, tumor size, depth of gastric carcinoma infiltration, and the presence of vascular or lymphatic invasion were significantly and independently related to lymph node metastasis. For intramucosal cancer, tumor size was the unique risk factor for lymph node metastasis. For submucosal cancer, histological classification and tumor size were independent risk factors for lymph node metastasis.20

In addition to these clinicopathological factors, many studies have been carried out to evaluate the role of biological markers in lymph node metastasis. p53 is a tumor-suppressor gene that plays an important role in the cell cycle and mediates the induction of apoptosis or recovery from DNA. Because of such a crucial role in the controlling of genome integrity, p53 has earned the name “guardian of the genome.” When mutant p53 protein is expressed, these functions are often blocked. It has been reported that p53 acts as a tumor growth regulator, and a correlation between p53 positivity and the presence of regional lymph node metastasis has been noted.12,2325 Ki67 is a nuclear antigen that is detected in proliferating but not resting cells. A high Ki67 labeling index (LI) at the deepest site of tumor penetration indicates an environment that promotes invasion and metastasis.12,26,27 Overexpression of p53 has been demonstrated to be related to lymph node metastasis in EGC.10,28 Goishi et al. reported that low Ki67 LI meant the low possibility of lymph node metastasis in EGC.12 However, Ohashi et al. reported that both p53 and Ki67 did not correlate to lymph node metastases in EGC.23 In this study, we concluded that both p53 and Ki67 did not correlate to lymph node metastasis in EGC.

Prior to operation, many investigators want to find the factors that associate with lymph node metastasis in EGC because the treatment strategies are different depending on whether lymph node metastasis is present or not. These treatments affect preservation of body function and maintenance of quality of life. It is reasonable to expect that useful information could be obtained by comparing the clinicopathological features and immunohistochemical expression of biological markers.

This research was retrospective and with small cases in single institution; furthermore, only two biological markers among a lot of immunohistochemical markers were evaluated. The value of these factors for lymph node metastasis needs to be verified by large-scale prospective studies.

In summary, prediction of lymph node metastasis in EGC is very important to decide the treatment strategies preoperatively. It is also important to estimate lymph node metastasis after EMR or ESD. The risk factors for lymph node metastasis in EGC are quite different depending on the state of things, such as depth of tumor invasion. To predict lymph node metastasis in EGC, it is recommended that distinct assessment according to individual situation must be clearly established.

Copyright information

© The Society for Surgery of the Alimentary Tract 2011