Prognosis of gastric cancer patients with node-negative metastasis following curative resection : Outcomes of the survival and recurrence

1Gastrointestinal Cancer Surgery Division, Tianjin Medical University Cancer Hospital and City Key Laboratory of Tianjin Cancer Center; 2The Institute of Lung Cancer, Tianjin Medical University General Hospital, Tianjin, China Correspondence: Dr Jingyu Deng, Gastric Oncology Division, Tianjin Medical University Cancer Hospital and City Key Laboratory of Tianjin Cancer Center, Tianjin, China. Telephone 86-222-334-0123, fax 86-222-335-9984, e-mail dengery@126.com Received for publication May 26, 2008. Accepted June 5, 2008 J Deng, H Liang, D Sun, R Zhang, H Zhan, X Wang. Prognosis of gastric cancer patients with node-negative metastasis following curative resection: Outcomes of the survival and recurrence. Can J Gastroenterol 2008;22(10):835-839.

G astric cancer, the most common malignancy in the world, frequently reveals lymph node, peritoneum and liver metastases.More than one-half of gastric cancer patients have lymph node metastasis when they are initially diagnosed or operated on, which results in a poor prognosis (1)(2)(3)(4).Lymph node involvement is the most important indicator of overall survival (OS) of gastric cancer patients following curative resection, and the survival rates markedly decrease in relation to the increase in the number of metastatic-positive nodes (5)(6)(7)(8)(9)(10).Furthermore, many investigators have demonstrated that lymph node metastasis is an independent risk factor of gastric cancer recurrence following curative resection (9,11,12).The overall postoperative survival of lymph node-negative gastric cancer patients is significantly longer than that of lymph nodepositive gastric cancer patients.The overall postoperative recurrence rate of lymph node-negative gastric cancer patients is significantly lower than that of lymph node-positive gastric cancer patients (13)(14)(15)(16).
In light of these considerations, the aims of the present study were to analyze the data for elucidation of the overall prognosis of lymph node-negative gastric cancer patients following curative resection, and to use multivariate analyses to identify the independent risk factors of survival and recurrence of lymph node-negative gastric cancer patients following curative resection.

METHODS
Between January 1997 and December 2000, 308 gastric cancer patients underwent curative resection for histologically confirmed gastric carcinoma in the department of gastrointestinal cancer surgery at the Tianjin Medical University Cancer Hospital (Tianjin, China).Of these, 112 patients (36.4%) with lymph node-negative metastasis were identified by pathological examination and evaluation.Seventy patients were men, and 42 were women.Ages ranged from 33 to 72 years, with a mean age of 54.2 years.The tumour was located in the lower one-third of the stomach in 68 cases (60.7%), the middle one-third in 27 cases (24.1%) and the upper one-third in 17 cases (15.2%).According to the Japanese Classification of Gastric Carcinoma, 37 patients underwent limited lymph node dissection (D1 lymphadenectomy), and the remaining 75 patients underwent extended lymph node dissection (D2 lymphadenectomy).No patient died during the initial hospital stay or one month after surgery.Patients with fewer than 15 dissected lymph nodes after curative resection for gastric cancer were not included.
Medical records were reviewed to obtain patient data, including age at the time of surgery (70 years or younger, or older than 70 years), sex (male or female), location of the primary tumour (lower, middle or upper stomach), size of the primary tumour (4 cm or smaller, or larger than 4 cm), depth of the primary tumour invasion (T1, T2, T3 or T4), Lauren classification (intestinal, diffuse or mixed), presence of serosal involvement (no or yes), surgery (subtotal or total), lymphadenectomy (D1 or D2) and number of dissected lymph nodes (20 or fewer, or more than 20).After curative surgery, all patients were followed every six months for two years, then every year or until death.The median follow-up period was 84 months (range, six to 144 months).B-ultrasonography, computed tomography scans, chest x-rays and endoscopy results were obtained with every visit.
The median OS and median recurrence-free survival were determined using the Kaplan-Meier method.The log-rank test was used to determine univariate significance.Factors that were deemed to be of potential importance by univariate analyses (P<0.05) were included in the multivariate analyses.Multivariate analyses were performed by means of the Cox proportional hazards model, using the forward stepwise procedure for variable selection.Hazard ratios (HRs) and 95% CIs were generated.Significance was defined as P<0.05.All statistical analyses were performed using a statistical analysis program package (SPSS version 13.0; SPSS Inc, USA).

Survival and recurrence analyses of lymph node-negative and lymph node-positive gastric cancer patients
In the present retrospective study, data from 308 consecutive patients who underwent curative resection for primary gastric cancer between January 1997 and December 2000 were examined.Of these patients, 112 (36.4%) had lymph node-negative metastasis and 196 (63.6%) had lymph node-positive metastasis.The five-year survival rate of lymph node-negative gastric cancer patients was significantly higher than that of lymph node-positive gastric cancer patients (85.7% versus 33.2%; log-rank P<0.001) following curative resection.Furthermore, the median OS of lymph node-negative gastric cancer patients was significantly longer than that of lymph node-positive gastric cancer patients following curative resection (114 versus 29 months; log-rank P<0.001).The survival curves for lymph node-negative gastric cancer patients and lymph node-positive gastric cancer patients are illustrated in Figure 1.Twentyfive of 112 lymph node-negative and 144 of 196 lymph nodepositive gastric cancer patients presented with recurrence.The median recurrence-free survival of lymph node-negative gastric cancer patients was significantly longer than that of lymph node-positive gastric cancer patients following curative resection (40 versus 18 months; log-rank P<0.001).The diseasefree survival curves for lymph node-negative and lymph node-positive gastric cancer patients are illustrated in Figure 2.After all recurrences were identified, the median postrecurrence survival of lymph node-negative gastric cancer patients who presented with recurrence was six months, whereas that of lymph node-positive gastric cancer patients who presented with recurrence was four months (log-rank P=0.021).

Univariate and multivariate survival analyses of lymph node-negative gastric cancer patients
Using univariate analyses, five factors were found to have statistically significant associations with OS of lymph nodenegative gastric cancer patients after curative surgery.These factors were sex, age, operative type, depth of primary tumour

Postoperative months Cumulative survival probability
Node-positive metastatic patients P<0.001 Figure 1) Overall survival of patients by status of lymph node metastasis.The Kaplan-Meier curve compares overall survival (defined as months after surgery) between gastric cancer patients with node-negative metastasis and those with node-positive metastasis invasion and presence of serosal involvement.The postoperative survival results of all lymph node-negative gastric cancer patients who underwent curative resection were as follows: the median OS of male patients was longer than that of female patients (119.4 versus 102.7 months; log-rank P=0.025); patients who were older than 70 years of age had a significantly shorter median OS than those who were 70 years of age or younger (60.0 versus 116.2 months; log-rank P=0.007); patients who underwent subtotal gastrectomy had a longer median OS than those who underwent total gastrectomy (116.6 versus 91.3 months; log-rank P=0.031); patients with serosal involvement had a shorter median OS than those without serosal involvement (101.4 versus 126.2 months; log-rank P<0.001); and the more deeply the primary tumour invaded, the shorter the median OS of lymph node-negative gastric cancer patients (T4, 72.0 months; T3, 102.5 months; T2, 116.6 months; and T1, 127.9 months; logrank P=0.002).
All five variables listed above were included in a multivariate Cox proportional hazards model (forward stepwise procedure) to adjust for the effects of covariates (Table 1).In that model, sex (HR=0.203;P=0.001), operative type (HR=0.343;P=0.035) and the presence of serosal involvement (HR=0.098;P<0.001) showed significant correlations with OS of lymph node-negative gastric cancer patients after curative surgery.

Univariate and multivariate recurrence analyses of lymph node-negative gastric cancer
Recurrence of lymph node-negative gastric cancer during followup was detected in 25 of 112 patients (22.3%) included in the study group; 13 patients (52.0%) showed local recurrence, six (24.0%) showed peritoneal dissemination, two (8.0%) showed hepatic metastases, and four (16.0%) showed local recurrence and peritoneal dissemination simultaneously.The median disease-free survival was 40 months (95% CI 29.8 to 50.2 months).In most cases (21 of 25 cases, 84.0%), recurrence occurred within five years after curative surgery; in only four cases (16.0%), recurrence occurred after five years.With univariate analysis, four factors were found to have statistically significant associations with recurrence of lymph node-negative gastric cancer patients after curative surgery.The factors were as follows: lymphadenectomy, depth of primary tumour, presence of serosal involvement and number of dissected nodes.The postoperative recurrent results in all lymph node-negative gastric cancer patients who underwent curative resection were as follows: patients who underwent D2 lymphadenectomy had a longer median disease-free survival than those who underwent D1 lymphadenectomy (49.8 versus 23.5 months; log-rank P=0.006); patients with serosal involvement had a shorter median disease-free survival than those without serosal involvement (36.7 versus 75.3 months; log-rank P=0.012); patients with more than 20 dissected nodes had a longer median disease-free survival than those with 20 or fewer dissected nodes (47.5 versus 21.4 months; log-rank P=0.012); and the more deeply the primary tumour invaded, the shorter the median disease-free survival of lymph node-negative gastric cancer patients was (T4, 15.0 months; T3, 39.3 months; T2, 71.7 months; and T1, 85.0 months; log-rank P=0.015) after curative surgery.
All four variables listed above were included in a multivariate Cox proportional hazards model (forward stepwise procedure) to adjust for the effects of covariates (Table 2).In that model, lymphadenectomy (HR=23.578;P=0.001), the presence of serosal involvement (HR=0.052;P=0.011) and the number of dissected nodes (HR=0.108;P=0.006) showed significant correlations with overall recurrence of lymph nodenegative gastric cancer patients after curative surgery.

DISCUSSION
In gastric cancer, the presence or absence of lymph node metastasis is one of the most important prognostic indicators of relative clinical and pathological factors that could influence the prognosis of patients following curative resection (1)(2)(3)(4)(17)(18)(19)(20).Although several investigators reported that the postoperative prognosis of lymph node-negative gastric cancer patients was significantly better than that of lymph node-positive gastric cancer patients, some researchers found that a minority of lymph node-negative gastric cancer patients had recurrence and poor survival (21)(22)(23)(24)(25). Identification of the factors associated with survival and recurrence of lymph node-negative gastric cancer patients following curative resection is important for choosing the appropriate recipients of sequentially essential adjuvant therapy.Figure 2) Disease-free survival of patients by status of lymph node metastasis.The Kaplan-Meier curve compares overall disease-free survival (defined as the time after surgery) between gastric cancer patients with node-negative metastasis and thoses with node-positive metastasis  Several investigators have demonstrated that older gastric cancer patients had a shorter OS than younger patients (26)(27)(28).In the present study, we found a similar relationship between age and OS in all lymph node-negative gastric cancer patients by using the Kaplan-Meier method and the logrank test.However, the results of multivariate analyses showed that age was not a statistically significant independent factor of OS in lymph node-negative gastric cancer patients.Some researchers showed no differences in OS regarding sex in gastric cancer patients, regardless of tumour stages (29,30).In a study performed on gastric cancer patients (31), women had a better prognosis than men.Conversely, in the present study, we found that male lymph node-negative gastric cancer patients had a longer median OS than female lymph node-negative gastric cancer patients, and identified sex as an independent factor of OS of lymph node-negative gastric cancer patients after curative surgery.Whether operative types of gastrectomy correlate with the postoperative survival of gastric cancer patients is still controversial.Many studies found that total gastrectomy offered no more survival benefits for gastric cancer patients than subtotal gastrectomy, regardless of the location of resectable lesions (32)(33)(34).We demonstrated that there is a close relationship between subtotal gastrectomy and long OS of lymph node-negative gastric cancer patients after curative surgery, which was not previously reported.To our knowledge, in addition to the number of positive lymph nodes, the depth of primary tumour invasion is another independent prognostic predictor of lymph node-negative gastric cancer (35).Our results showed that the depth of primary tumour invasion could not significantly affect the recurrence of lymph node-negative gastric cancer patients until the serosa of the stomach was infiltrated by a primary tumour.Similar results were reported by other authors (25,36,37).Although Japanese surgeons routinely perform extended lymphadenectomy for gastric cancer patients without distant metastasis, many western investigators have reported that D2 or more extended lymphadenectomy could not improve the prognosis of gastric cancer significantly compared with D1 lymphadenectomy, which is generally accepted by surgeons in western countries (38)(39)(40)(41)(42)(43).Nevertheless, other investigators have stated that D2 lymph node dissection might improve the survival of patients with lymph node-negative gastric carcinoma (25,44,45).In the present study, we found that D2 lymphadenectomy did not contribute to extending OS, but did contribute to extending disease-free survival of lymph node-negative gastric cancer patients.Many studies demonstrated that the number of dissected lymph nodes was not relative to the prognosis of gastric cancer patients after curative surgery (46)(47)(48).Our results revealed that the number of dissected lymph nodes did not affect OS but did affect disease-free survival.

CONCLUSIONS
Male lymph node-negative gastric cancer patients who underwent subtotal gastrectomy and presented without serosal invasion had long-term survival after curative surgery for gastric cancer.Meanwhile, D1 lymphadenectomy, few dissected lymph nodes and serosal invasion were the independent risk factors of recurrence for lymph node-negative gastric cancer after curative surgery for gastric cancer.

TABLE 1
Multivariate analysis of factors affecting the overall survival of 112 patients with lymph node-negative metastasis

TABLE 2
Multivariate analysis of factors affecting overall recurrence of lymph node-negative metastasis in 112 patients