| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Original Article |
1 Department of Surgery, Yonsei University College of Medicine, 134 Shinchon-dong Seodaemun-ku, Seoul 120-752, Korea
2 Cancer Metastasis Research Center, Yonsei University College of Medicine, 134 Shinchon-dong Seodaemun-ku, Seoul 120-752, Korea
3 Department of Surgical Oncology, Sir Run Run Shaw Hospital, College of Medicine, Zhejiang University, Hangzhou, 310016, China
4 Brain Korea 21 Project for Medical Science, Yonsei University College of Medicine, 134 Shinchon-dong Seodaemun-ku, Seoul 120-752, Korea
Correspondence: Address correspondence and reprint requests to: Sung Hoon Noh, MD, PhD; E-mail: sunghoonn{at}yumc.yonsei.ac.kr.
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONPATIENTS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Methods: From a prospective database of 1264 EGC patients between 1987 and 1997, 156 (12.4%) were found to have histologically confirmed node metastasis. A number of prognostic factors, including the N ratio, were evaluated by univariate and multivariate analysis.
Results: The recurrence rate of node-positive EGC was 16.7% (n = 26). The overall 5-year survival rate of all patients was 84.0%. It was 26.9% and 95.4% in patients with and without recurrence, respectively (P < .0001; log-rank test). The cutoff value of the N ratio was set at .07. The 5-year survival rate of patients with an N ratio <.07 was 94.0%; this was significantly higher than the rate (72.6%) for those with a ratio >.07 (P < .0001; log-rank test). Both univariate and multivariate analysis identified the N ratio as the most significant predictive factor for recurrence and overall survival. Regarding stage migration, it shows superiority in comparison to the number-based N classification.
Conclusions: The N ratio is a more effective and rational indicator for prognostic stratification of patients with lymph node–positive EGC than the current N classification of the tumor-node-metastasis system.
Key Words: Early gastric cancer • Lymph node metastasis • Metastatic lymph node ratio • Prognosis • Recurrence
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONPATIENTS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
In general, the presence of lymph node metastasis is recognized as the most significant prognostic factor for survival in patients with EGC.7–10 Unlike in advanced disease, most patients with lymph node–positive EGC fall into the N1 category; however, diverse prognoses are evident even among the same N classifications. Hence, it is practically impossible to discriminate which patients are at high risk of recurrence solely on the basis of the current N staging system. Furthermore, because the prevalence of EGC has been increasing,7,8 the proportion of those with recurrence is also likely to increase. Therefore, the precise identification of the patient subgroup with poor prognosis has substantial clinical importance. However, thus far, there have been few studies to address the prognostic stratification of lymph node–positive EGC within the current staging system.
Recently, a new prognostic factor has been proposed for patients with gastric cancer who undergo curative resection (R0 resection). This factor takes into account the ratio between metastatic and examined lymph nodes and seems to obviate the problems of overstaging or understaging incurred in the current tumor-node-metastasis system.11–17 The primary aim of this study was to evaluate objective and reliable prognostic factors for patients with lymph node–positive EGC and especially to assess the prognostic value of the ratio of metastatic to examined lymph nodes (N ratio) in identifying high-risk patient subgroups.
|
PATIENTS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONPATIENTS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
All patients in the study had the following standard operations: (1) total or distal subtotal gastrectomy, depending on the location and macroscopic appearance of the primary tumor, and (2) a D2 or more extended (D3) lymphadenectomy. The definitions for a D2 and D3 lymphadenectomy were based on the Japanese Research Society for Gastric Cancer rules; these classify the regional lymph nodes of the stomach into four compartments.18 D2 lymphadenectomy includes a complete dissection of compartments I and II, whereas a D3 lymphadenectomy includes that of compartments I, II, and III. Compartment I consists of the perigastric lymph nodes. Compartment II consists of the lymph nodes along the left gastric artery, the common hepatic artery, and the splenic artery and around the celiac axis. However, when the cancer is located in the lower third of the stomach, the lymph nodes along the splenic artery are classified as being in compartment III. Compartment III also consists of lymph nodes in the hepatoduodenal ligament, at the posterior aspect of the head of the pancreas, and at the root of the mesentery. Compartment IV consists of lymph nodes along the middle colic vessels and para-aortic lymph nodes.
The resected specimen was carefully examined for accurate pathologic staging according to the Japanese Research Society for Gastric Cancer rules.18 Depth of invasion was precisely determined by examining the deepest portion of the gastric wall invasion. The classification of the dissected lymph nodes was verified by surgeons reviewing the excised specimens after surgery. All lymph nodes retrieved were stained with hematoxylin and eosin and examined for metastasis by specialized pathologists who used light microscopy. For the purpose of a comparative prognostic factor study, nodal status was classified according to both the fourth and the fifth edition of the tumor-node-metastasis classification system of the International Union Against Cancer (UICC) and the American Joint Committee on Cancer (AJCC).19,20
Aside from the number and location of metastatic lymph nodes, the N ratio was also evaluated. The N ratio was defined as the number of metastatic lymph nodes over the number of retrieved lymph nodes.
Determination of the N Ratio Cutoff Value
The log-rank test was used to find the appropriate cutoff point for the N ratio; it identified two subgroups of patients with remarkably different survival rates. When the N ratio was <.07, the 5-YSR was 94.0%, well above the overall 5-YSR of 84.0% (see Results). When the N ratio was >.07, the 5-YSR was 72.6% (see Results). Other cutoff points were also evaluated; however, significant differences in survival rates and the most even distribution of patients across the two groups were observed only when the abovementioned cutoff point was used. Consideration was also given to the average number of retrieved lymph nodes across the two groups, because this number is the denominator of the N ratio and could, therefore, substantially affect the ratio. In conclusion, the cutoff value of the N ratio was set at .07.
Follow-Up
All patients were initially evaluated at 1 month after surgery; follow-up was then at 2- to 4-month intervals and then twice a year thereafter. Patient outcome information for recurrence and survival was obtained through routine outpatient clinic follow-ups and from admission records. For our study, follow-up ended on June 30, 2002; the median follow-up period was 92.3 months (range, 1.8–182.3 months). Thirty-one patients (19.9%) had died, and four patients (2.6%) were lost to follow-up. Information on survival for these four patients was obtained from the Korean National Statistics Registry Database. Unfortunately, other detailed information, such as recurrence patterns, if present, could not be obtained.
Determination of recurrence was made by clinical and radiological examination or by operation. The main patterns of recurrence were recorded as the first site of detectable failure at the time of diagnosis. Survival duration was calculated from the date of operation to the date of death.
Statistical Analysis
All statistical analyses were conducted by using the statistical program SPSS 11.5 (SPSS Inc., Chicago, IL). Pearson’s correlation coefficient was used to assess the correlation between the number of retrieved nodes and the number of metastatic nodes, the number of retrieved nodes and the N ratio, and the number of metastatic nodes and the N ratio. The risk factors associated with tumor recurrence were determined by univariate and multivariate analysis. In univariate analysis, two-tailed 
2 tests for categorical variables and two-tailed Student’s t-tests for continuous variables were used for statistical comparisons. For multivariate analysis, logistic regression analysis was performed in a forward stepwise selection of variables with a likelihood ratio test. The hazard ratio, in the logistic regression analysis, was defined as the ratio of the probability that an event (tumor recurrence) would occur to the probability that it would not occur. The prognostic power of covariates was expressed by calculation of a hazard ratio with a 95% confidence interval (CI).
Cumulative survival rates were estimated by using the Kaplan-Meier product-limit method. Univariate influences of certain parameters on survival were assessed with the log-rank test. All statistically significant variables found by univariate analysis were used in the multivariate analysis by using the Cox proportional hazards model in forward stepwise regression. Statistically significant factors in the Cox analysis were expressed as a hazard ratio and a 95% CI. In all statistical analyses, a P value of <.05 was considered significant.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONPATIENTS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
The mean metastatic to retrieved lymph node ratio was .1 ± .1 (range, .01–.91). The mean number of lymph nodes that had been removed for the group with an N ratio <.07 was 37.2 ± 13.3 (n = 83) and was 35.1 ± 14.7 (n = 73) for the group with an N ratio 
.07. There was no significant difference between these numbers (P = .51). The mean number of metastatic nodes, however, was significantly different between the two groups: 1.36 ± .69 (n = 83) and 6.22 ± 7.53 (n = 73) for the <.07 and .07 groups, respectively (P < .0001).
There was a significant correlation between the number of retrieved nodes and that of metastatic nodes by Pearson’s correlation test (r = .331; P < .0001; Table 1
). There was no significant association between the N ratio and the number of retrieved nodes (r = –.099; P = .218), but there was a strong relationship between the N ratio and the number of metastatic nodes (r = .735; P < .0001; Fig. 1). Thus, with respect to stage migration, the N ratio shows superiority in comparison to the number-based N classification of the current tumor-node-metastasis system.
|
|
. The most common pattern of recurrence was hematogenous metastasis (34.6%), followed by locoregional (19.2%) and peritoneal (11.5%) failure. Hematogenous recurrence occurred in the liver, bone, lung, and brain in seven, four, two, and two patients, respectively. The mean interval between the initial operation and the development of recurrence was 23.4 months and ranged from 4 to 61 months. The clinicopathologic features of patients with and without recurrence are listed in Table 3. Tumor size, location, gross appearance, histological classification, and depth of invasion (mucosa vs. submucosa) had no significant relationship to the incidence of recurrence. Of note, the mean N ratio was significantly higher for patients with recurrence (.238; 95% CI, .158–.318) than for those without recurrence (.078; 95% CI, .066–.09; P = .001). The N ratio, number-based N classification (fifth edition of the tumor-node-metastasis staging system), and location-based N classification (4th edition of the tumor-node-metastasis staging system) were found to be significantly associated with tumor recurrence by the univariate 2 test. On multivariate analysis using a logistic regression model, however, the metastatic lymph node ratio emerged as the only significant prognostic determinant for recurrence (hazard ratio, 7.521; 95% CI, 1.984–28.517; P = .003).
|
|
The overall 5-YSR was 84.0% (95% CI, 83.9%–84.1%). The rates for patients with and without recurrence were 26.9% (95% CI, 26.8%–27.0%) and 95.4% (95% CI, 95.3%–95.5%), respectively (P < .0001; log-rank test; Fig. 2). The 5-YSR of patients with an N ratio less than the cutoff value of .07 was 94.0% (95% CI, 93.9%–94.1%), which was significantly longer that that of patients with a ratio >.07, which was 72.6% (95% CI, 72.5%–72.7%; P < .0001; log-rank test). Tumor recurrence, depth of invasion (mucosa vs. submucosa), location-based N classification (fourth edition of the tumor-node-metastasis staging system), number-based N classification (fifth edition of the tumor-node-metastasis staging system), and N ratio were significant predictors of survival on univariate log-rank test.
|
). Survival curves according to these prognostic factors are shown in Fig. 3.
|
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONPATIENTS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Accurate disease staging is of fundamental importance in clinical practice, especially when adjuvant treatment is to be considered, as well as for planning a follow-up schedule. In EGC, metastasis to the lymph nodes is considered the main prognostic factor after curative resection.7–10 Lymph node status assessment, therefore, is an essential part of stage classification.
In this study, we reviewed the data from a prospectively maintained database to investigate objective and reliable prognostic factors for lymph node–positive EGC. Of the various prognostic indicators evaluated, which included submucosal invasion, tumor size and location, histological differentiation, macroscopic appearance, N classification (both number and location based), and the N ratio, only the N ratio was associated with both recurrence and poor survival by multivariate analysis. In the current AJCC/UICC tumor-node-metastasis guidelines, the N classification is based on the number of metastatic lymph nodes, and this has been recognized as a simple, reliable, and objective method for predicting the prognosis of gastric cancer patients. Even though this classification provides relatively reliable information on nodal status, its secondary purpose is to allow prognostic stratification of patients into homogenous groups, with respect to prognosis, after surgery. However, the N classification of the current tumor-node-metastasis staging system fails to meet this goal, at least for EGC with lymph node metastasis.
According to the fifth edition AJCC/UICC tumor-node-metastasis guidelines, metastasis in one to six regional lymph nodes is classified as N1 disease. 20 Because most cases with lymph node–positive EGC fall into the N1 category, the current N classification has a serious drawback for discriminating between subgroups of patients with diverse prognoses. In our series, only 3 (1.9%) of 156 patients had N3 disease, and 20 (12.8%) had N2 disease. Considering the low rate of lymph node involvement, as well as the small number of positive nodes actually involved, a novel classification system would be required to provide the precise prognostic stratification of lymph node–positive EGC.
Because of these current limitations, we evaluated the prognostic relevance of the N ratio, along with the number- and location-based N classification systems. The N ratio has been proposed as a new prognostic factor that is independent of the number of lymph nodes removed and examined and that avoids the stage-migration phenomenon related to the current tumor-node-metastasis staging system. A multitude of studies have demonstrated that the N ratio is a simple way to measure the efficacy of lymphadenectomy.11–17 Its application has proven to be a good alternative in helping to avoid stage migration.12,14 This also enables physicians to plan a more appropriate follow-up strategy that is based on more reliable and predictable prognostic information.
Our results are in line with the accumulating body of evidence indicating that the N ratio is a powerful prognostic parameter and further demonstrate its usefulness in identifying a high-risk subset of patients within the same N1 category (n = 133). This has substantial practical importance because most lymph node–positive cases have between one and six positive nodes and fall into the N1 category; however, this current N classification cannot provide prognostic stratification to identify high-risk patients within this category. If the cutoff value of the N ratio is set at .07, then 83 patients in our study were <.07 and 50 were >.07. The 5-YSR was 95.2% and 88.0%, respectively. Furthermore, the recurrence rate was also significantly different between the two groups: 3.6% vs. 22%, respectively (P = .001). This signifies that even in EGC patients within the same N1 classification, a substantial proportion of them will experience recurrence after curative surgery and that the N ratio can be a better predictor of recurrence than the currently available tumor-node-metastasis classification.
In addition, correlation tests revealed that the N ratio is not related to the stage-migration phenomenon, whereas the number-based current N classification is associated with the stage-migration phenomenon (any increase of metastatic lymph nodes according to the increasing number of retrieved/examined nodes).21,22 The N ratio correlated significantly (r = .735; P < .0001) with the number of metastatic nodes, a current parameter for N classification in the tumor-node-metastasis system. At the same time, however, it was found not to be associated with the number of retrieved nodes (r = –.099; P = .218).
To our knowledge, this study is the first to confirm that the N ratio is of greater prognostic value than the N classification of the current tumor-node-metastasis staging system in prognostic stratification of node-positive EGC. Aside from the current study, we performed an additional analysis of 16 patients from the same database. These patients were considered to undergo curative resection and had <15 lymph nodes retrieved (data not shown). Although it is possible that these cases could have been inadequately staged or not staged on the basis of the current N classification, our analysis revealed that the N ratio may identify high-risk patients even with <15 examined nodes. This implies that the N ratio has practical significance in the era of minimally invasive and limited surgery for EGC. In the setting of limited lymphadenectomy for EGC, the number of lymph nodes resected may be strictly limited and even <15. Therefore, an appropriate assessment of nodal status according to the current tumor-node-metastasis system would be hindered, and prognostic information may not be obtained. However, application of the N ratio would enable surgeons not only to classify such cases, but also to predict a more accurate prognosis and risk of recurrence. Further clinical evaluation of this finding, however, should be performed in future studies with appropriate designs and a valid number of cases.
In conclusion, the results of this study indicate that the N ratio is a more effective and rational indicator for prognostic stratification of patients with lymph node–positive EGC than the current N classification of the tumor-node-metastasis system. In patients with a ratio >.07, a closer follow-up strategy for early detection of recurrence may be indicated. Furthermore, identification of patients at high risk of recurrence and poor prognosis on the basis of the N ratio could help in the consideration of adjuvant chemotherapy.
|
ACKNOWLEDGMENTS |
|---|
Received for publication April 13, 2005. Accepted for publication September 5, 2005.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONPATIENTS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
This article has been cited by other articles:
 |
|
 |
|
  X. Wang, J.-J. Wang, and F. Wan A Common Misuse of Stepwise Regression in Studies of Ratio of Metastatic Lymph Nodes for Gastric Cancer Ann. Surg. Oncol., June 1, 2008; 15(6): 1805 - 1806. [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
