10.1245/ASO.2006.04.028
Annals of Surgical Oncology 13:221-228 (2006)
© 2006 Society of Surgical Oncology
Outcomes of Mass Screening for Gastric Carcinoma
Chikara Kunisaki, MD1,
Junko Ishino, MD2,
Susumu Nakajima, MD2,
Hisahiko Motohashi, MD2,
Hirotoshi Akiyama, MD1,
Masato Nomura, MD1,
Goro Matsuda, MD1,
Yuichi Otsuka, MD1,
Hidetaka Andrew Ono, MD1 and
Hiroshi Shimada, MD1
1 Department of Gastroenterological Surgery, Yokohama City University, Graduate School of Medicine, 3-9 Fukuura Kanazawaku, Yokohama 236-0004, Japan
2 Kanagawa Health Service Association, 58 Nihon Oodori, Nakaku, Yokohama 231-0021, Japan
Correspondence: Address correspondence and reprint requests to: Chikara Kunisaki, MD; E-mail: s0714{at}med.yokohama-cu.ac.jp.
 |
ABSTRACT
|
|---|
Background: Therapeutic results of gastric cancer have been improved by early detection of gastric cancer with the mass screening system in Japan. The objective of our study was to assess the efficacy of mass screening for gastric cancer by using a barium meal.
Methods: A series of 1050 patients (364 in the screened group and 686 in the nonscreened group) were included in this study from April 1992 to March 2000. Patient characteristics, therapeutic results, and prognostic factors were compared in the two groups.
Results: The screened patients tended to be younger and male, with tumors in the middle third of the stomach that were of a macroscopically superficial type, with a smaller diameter, and at an earlier stage. They had fewer metastatic lymph nodes and underwent more frequent curative resection. Among the screened patients with curatively resected disease, tumors tended to be of a smaller diameter, and there were fewer metastatic lymph nodes in both early and advanced cases. Disease-specific survival was significantly better in the screened cases among all registered and curatively resected patients. Mass screening achieved significantly better surgical results in early or advanced gastric cancer patients who received curative resection. Multivariate analysis revealed that mass screening was an independent prognostic factor (hazard ratio, .3949; P < .0001), together with depth of invasion, lymph node metastasis, age, and tumor diameter.
Conclusions: Mass screening by using barium meal examination for gastric cancer detects cancer at an early stage and produces good therapeutic results.
Key Words: Gastric cancer • Mass screening • Asymptomatic • Early gastric cancer • Barium meal examination
|
INTRODUCTION
|
|---|
Gastric cancer has continued to contribute substantially to cancer-related deaths throughout the world. However, its incidence has decreased in Japan.1 A major cause of this improvement is increasingly frequent detection of early gastric cancer by screening, as well as improved treatment.2 In Japan, mass screening is widely accepted, and the incidence of potentially curable early gastric cancer has increased.3 There are two screening programs in Japan: one involves a barium meal, followed by endoscopic examination in patients with abnormal findings,4 and the other involves endoscopic examination without a preceding barium meal.5 Early detection of gastric cancer permits curative management not only by surgical resection but also by endoscopic mucosal resection6,7 or by a minimally invasive technique such as laparoscopic gastrectomy.8,9 Early asymptomatic gastric cancer detected by screening can be treated less invasively and yields improved overall outcomes.
Few studies have compared therapeutic outcomes in cases of asymptomatic gastric cancer detected by screening and in cases of symptomatic gastric cancer.10,11 In this study, we compared the clinicopathologic characteristics and results of surgery in asymptomatic gastric cancers detected by means of an indirect barium meal conducted in the Kanagawa Health Service Association with those of symptomatic gastric cancers treated in the same period. Our aim was to assess the effectiveness of screening for gastric cancer.
|
PATIENTS AND METHODS
|
|---|
Patients
Between April 1992 and March 2000, 770,710 individuals were screened by barium meal examination for gastric cancer in the Kanagawa Health Service Association. Three hundred sixty-four cases (.047%) of asymptomatic gastric cancer were detected. In the same period, 686 individuals who were symptomatic and were diagnosed as having gastric cancer by nonscreening modalities underwent gastrectomy at the Department of Gastroenterological Surgery, Yokohama City University Graduate School of Medicine.
The screening procedure for asymptomatic participants is shown in Fig. 1
. When abnormal findings were detected in the initial barium meal examination (indirect examination; 8 series of x-ray films), a more detailed barium meal examination (direct examination) was undertaken (11 series of x-ray films). If, as a result, gastric cancer was strongly suspected, endoscopic examination with a biopsy was performed. Abnormal findings leading to an endoscopic examination after a barium meal examination were as follows: decreased caliber of the gastric lumen, stenosis, deformity, rigidity, indentation, the presence of a niche or a filling defect in the wall, flattening of the randwall, pooling of the barium, irregularity in the gastric area, change in the gastric fold (e.g., thickening, enlargement, or tortuosity), or presence of a polypoid lesion.
The target population in the screening program consisted of participants who underwent a self-financed yearly health checkup or an annual institutional health checkup. The cost of the barium meal examination was $43 per person. Participants who underwent the more detailed barium meal examination paid $136 for it and $118 for the endoscopic examination.
Clinicopathologic variables (patient age and sex, location of the tumor, macroscopic appearance, tumor diameter, depth of invasion, lymph node metastasis, stage, and curability) and survival were retrieved retrospectively and compared between the symptomatic patients (nonscreened group) and asymptomatic patients (screened group). These factors were recorded prospectively according to the International Union Against Cancer (UICC) TNM: Classification of Malignant Tumours.12
Preoperative evaluation was performed by oral barium meal examination, gastrofiberscopy with biopsy, and computed tomography (CT). Clinical lymph node metastasis was diagnosed by CT; irregularly shaped lymph nodes and nodes with diameters >10 mm were strongly positive indications.
Of the 1050 registered patients, 918 (87.4%) underwent curative resection. Among the 1050 registered patients, the macroscopically superficial type of carcinoma was observed in 522 patients, the well-defined type was observed in 131 patients, and the ill-defined type was observed in 397. The mean tumor diameter was 49.0 ± 35.5 mm. Histologically, the diKerentiated type of tumor was observed in 514 patients, and the undiKerentiated type, in 536. Five hundred twenty-two patients (49.7%) had T1 tumors, 264 (25.1%) had T2 tumors, and 264 (25.1%) had T3 or T4 tumors. Therefore, there were 522 (49.7%) cases of early gastric cancer (T1) and 528 (50.3%) cases of advanced gastric cancer (T2, T3, and T4). At least 15 lymph nodes were dissected pathologically in each patient. Lymph node involvement was detected in 415 patients: pN1 in 141, pN2 in 110, pN3 in 86, and unknown in 78, according to the UICC tumor-node-metastasis classification. Lymph node metastasis was diagnosed as unknown in many of the patients who received palliative resection because the extent of lymph node dissection was insuMcient. Lymph node dissection was performed in the Japanese manner.13 All patients were followed up according to our standard protocol (at least once every 12 weeks for 5 years), which includes tumor marker studies, gastrofiberscopy, abdominal ultrasonography, CT, and chest radiography. A preliminary diagnosis of peritoneal dissemination based on physical examination or imaging modalities was confirmed by aspiration biopsy. The mean follow-up duration was 69.0 ± 42.0 months.
Statistical Analysis
Statistical analyses were performed with SPSS statistical software version 10.0 for Windows (SPSS Inc., Chicago, IL). Statistical significance was defined as P < .05. Disease-specific survival was calculated by using the Kaplan-Meier method, and differences between groups were assessed by log-rank tests. To evaluate the effect of clinicopathologic factors on long-term survival, potential prognostic factors were analyzed with a forward-condition Cox proportional hazards regression model. The following variables were inserted: patient age and sex, location of the tumor, macroscopic appearance, tumor diameter, histological type, depth of invasion, lymph node metastasis, and screening. Patient characteristics were compared by the two-tailed Fisher’s exact test or by the 
2 test with the Yates correction, as appropriate. Quantitative variables were compared by using the Student’s t-test and are expressed as mean ± SD.
|
RESULTS
|
|---|
Characteristics of Screened and Nonscreened Patients
Among the registered patients, there were significant diKerences between screened and nonscreened patients in age, sex, location of tumor, macroscopic appearance, tumor diameter, lymph node metastasis, depth of invasion, stage, and curability, but not in histological type. The screened patients tended to be younger and male, with tumors in the middle third of the stomach that were of a macroscopically superficial type, had a smaller diameter, and were at an earlier stage. These patients had fewer metastatic lymph nodes, more frequently underwent curative resection, and had more superficial invasion (Table 1). Quite similar results were obtained in the patients with curatively resected disease (Table 2). Moreover, the screened patients among the curatively resected patients with early gastric cancer tended to be younger, with tumors located in the middle third of the stomach and of a smaller tumor diameter; they also had fewer metastatic lymph nodes (Table 3). The screened patients among the curatively resected patients with advanced gastric cancer tended to have tumors of a histologically undifferentiated type, in addition to the characteristics of the screened early gastric cancer patients already mentioned (Table 3).
Survival
In the total population of registered patients, there was a diKerence in overall survival between the screened and nonscreened patients (P < .0001). The cumulative 5-year survival rate in the former was significantly better than in the latter: 89.4% vs. 66.5% (Fig. 2). In the curatively resected patients, there was also a difference in disease-specific survival between the screened and nonscreened patients (P < .0001). The cumulative 5-year survival rate in the former was significantly better than in the latter: 92.6% vs. 77.6% (Fig. 3). Moreover, screening significantly affected therapeutic outcome in patients with early gastric cancer (T1). The 5-year survival rate was 99.6% in the screened patients and 95.9% in the nonscreened patients (P = .0003). Screening also affected therapeutic outcome in patients with advanced gastric cancer (T2, T3, and T4). Five-year survival was 79.1% in screened patients and 57.9% in nonscreened patients (P < .0001; Fig. 4). When patients were classified by tumor stage, there was also a significant difference between screened and nonscreened patients in the survival of those with stage IA, IB, IIIB, and IV tumors (Table 4).
View larger version (13K):
[in this window]
[in a new window]
|
FIG. 2. Survival of all registered patients. There was a significant difference (P < .0001) in disease-specific survival between screened (n = 364) and nonscreened (n = 686) patients.
|
|
View larger version (12K):
[in this window]
[in a new window]
|
FIG. 3. Survival of screened versus nonscreened patients with curative resection. There was a significant difference (P < .0001) in disease-specific survival between screened (n = 337) and nonscreened (n = 581) patients.
|
|
View larger version (14K):
[in this window]
[in a new window]
|
FIG. 4. Survival of early gastric cancer patients with curative resection. There was a significant difference (P = .0003) in disease-specific survival between early gastric cancer patients who were screened (n = 222) and those who were not (n = 300). There was also a significant difference (P < .0001) in disease-specific survival between advanced gastric cancer patients who were screened (n = 115) and those who were not (n = 281).
|
|
Prognostic Factors
The Cox proportional hazards regression model revealed that depth of invasion, lymph node metastasis, patient age, tumor diameter, and screening independently aKected prognosis in the patients with curatively resected disease (Table 5).
|
DISCUSSION
|
|---|
Our study indicates that mass screening for gastric cancer by using barium meal examination has a significant survival benefit at an acceptable financial cost. Since 1963, screening has been conducted by the Kanagawa Health Service Association with the aim of detecting gastric cancer at an early stage. Before this initiative, many gastric cancers were diagnosed only at an advanced stage, and, consequently, the results of surgery were poor. Therefore, nationwide screening for gastric cancer was initiated in Japan as a public health policy and has continued successfully since then. Recently, the incidence of early gastric cancer among all gastric cancers treated was reported to be 
50% as a result of mass screening.2 Moreover, in our study, 60% of the patients identified by screening had early gastric cancer. Our results showed that screening increased the frequency of curative resection even in advanced gastric cancer and resulted in an improvement in overall outcome. Therefore, promoting the value of screening to the general public is important. However, there have been conflicting reports concerning the efficacy of screening for gastric cancer in other countries.14,15 In Japan, gastric cancers constitute a high proportion of all cancers, whereas in the countries from which conflicting reports have emerged, gastric cancer represents a lower proportion of all cancers. Furthermore, the effectiveness of screening via a barium meal for diagnosing gastric cancer may be lower in other countries. Therefore, efforts should be made to increase both the practice and the efficacy of screening in such countries.
Younger and male patients preponderated in the screened group. This is partly because this group consisted of young and middle-aged men who were employed and able to take advantage of the free annual health screenings oKered by their employers. In addition, the tumors detected by screening were frequently in the middle third of the stomach. This may be because macroscopically depressed early gastric cancer in Japan tends to be located around the side of the lesser curvature of the middle third of the stomach. Comparison of all the other pathologic variables confirmed that screening detects cancer at an earlier stage. As a result, surgical outcomes tend to be favorable. Among early gastric cancers (T1), tumors of a smaller diameter and with fewer metastatic lymph nodes were more common in the screened group, and the higher survival rate among the cases of screened early cancer is attributable to these factors, especially because lymph node metastasis is reported to be an independent prognostic factor in early gastric cancer.16 Because advanced cancers (T2, T3, and T4) were frequently identified at an earlier stage in the screened group, it is not surprising that their outcome was also significantly better. The histologically undifferentiated type of tumor was common in the screened patients with advanced gastric cancer. However, because histological type is not an independent prognostic factor in curatively resected gastric cancer,17 this may not have affected survival.
Comparison of survival by stage in the two groups showed significant diKerences for stages IA, IB, IIIB, and IV. For the same stages, lymph node metastasis and depth of invasion were quite similar. However, other clinicopathologic factors may diKer; even at the same stage, the survival of symptomatic patients (nonscreened group) may be aKected by adverse factors such as anemia due to bleeding, poor nutrition, body weight loss due to poor oral intake, and poor performance status. Moreover, lymph node metastasis was defined in this study in accordance with the UICC/tumor-node-metastasis classification (number of metastatic lymph nodes). Therefore, there may have been diKerent anatomical distributions of metastatic lymph nodes in the two groups, even at equivalent stages. In addition, differences in patient numbers at each stage may have aKected the outcome. To clarify these matters, a more detailed or larger-scale study should be conducted.
The Cox proportional hazards regression model revealed that depth of invasion, lymph node metastasis, age, tumor diameter, and screening independently aKected prognosis. This result shows that screening reduces mortality by 60% in curatively resected gastric cancers. Other clinicians have reported similar results.18 Clearly, screening plays an important role in detecting early gastric cancer and in improving outcomes.
Recently, screening with an initial endoscopy has become more popular in Japan. It is generally considered that an endoscopic examination and biopsy can detect smaller early gastric cancers compared with barium meal examination. However, good technique in the barium meal examination and interpretation of the films by experienced doctors and technicians can provide satisfactory results. In another study, there was no diKerence in the survival of patients with asymptomatic gastric cancers between cases detected by endoscopy and those detected by barium meal examination.19 Appropriate therapy for early tumors regardless of the method of detection should produce good results. We consider that, in view of its benefits and cost-effectiveness, barium meal examination is entirely adequate for detecting early gastric cancers.
In conclusion, screening for gastric cancer is eKective in detecting early-stage gastric cancer and results in satisfactory therapeutic outcomes. Continued promotion of screening for gastric cancer should ensure good therapeutic results in the future. In countries where the incidence of gastric cancer is fairly high, the establishment of Japanese-style screening for gastric cancer and its popularization, not only among medical workers, but also among the general public, would be beneficial.
Received for publication April 26, 2005.
Accepted for publication August 26, 2005.
|
REFERENCES
|
|---|
- Tominaga S, Aoki K, Fujimoto I, Kurihara M. Cancer mortality and morbidity statistics. Japan and the world—1994. Gann Monogr Cancer Res 1994; 41:175–7.
- Kong SH, Park do J, Lee HJ, et al. Clinicopathologic features of asymptomatic gastric adenocarcinoma patients in Korea. Jpn J Clin Oncol 2004; 34:1–7.[Abstract/Free Full Text]
- Fukao A, Tsubono Y, Tsuji I, Hisamichi S, Sugahara N, Takano A. The evaluation of screening for gastric cancer in Miyagi prefecture, Japan. A population-based case-control study. Int J Cancer 1995; 60:45–8.[Medline]
- Matsukuma A, Furusawa M, Tomoda H, Seo Y. A clinicopathological study of asymptomatic gastric cancer. Br J Cancer 1996; 74:1647–50.[Medline]
- Mori Y, Arita T, Shimoda K, Yasuda K, Yoshida T, Kitano S. Effect of periodic endoscopy for gastric cancer on early detection and improvement of survival. Gastric Cancer 2001; 4:132–6.[Medline]
- Arai J, Yamada H, Maruyama M. Initial radiographic findings of early gastric cancer detected in health check programs and human "dry dock" (multiphasic screening) collective health checks and treated by endoscopic mucosal resection. Gastric Cancer 2002; 5:35–42.[Medline]
- Kondo H, Gotoda T, Ono H, et al. Percutaneous traction-assisted EMR by using an insulation-tipped electrosurgical knife for early-stage gastric cancer. Gastrointest Endosc 2004; 59:284–8.[Medline]
- Kitano S, Shiraishi N. Current status of laparoscopic gastrectomy for cancer in Japan. Surg Endosc 2004; 18:182–5.[Medline]
- Kitano S, Shiraishi N, Kakisako K, Yasuda K, Inomata M, Adachi Y. Laparoscopy-assisted Billroth-I gastrectomy (LADG) for cancer: our 10 years’ experience. Surg Laparosc Endosc Percutan Tech 2002; 12:204–7.[Medline]
- Ohshima A, Hirata N, Ubukata T, Umeda K, Fujimoto I. Evaluation of a mass screening program for stomach cancer with a case-control study design. Int J Cancer 1986; 38:829–33.[Medline]
- McCulloch P, Brown P, Martin B, Williams E. The effects of an awareness-raising program for patients and primary care physicians on the early detection of gastro-oesophageal cancer. Surgery 2003; 133:154–61.[Medline]
- Sobin LH, Wittekind CH. TNM: Classification of Malignant Tumours. 6th ed. New York: Wiley-Liss, 2002.
- Hagiwara A, Sawai K, Sakakura C, et al. Complete omentectomy and extensive lymphadenectomy with gastrectomy improves the survival of gastric cancer patients with metastases in the adjacent peritoneum. Hepatogastroenterology 1998; 45:1922–9.[Medline]
- Liorens P. Gastric cancer mass survey in Chile. Semin Surg Oncol 1991; 7:339–43.[Medline]
- Pisani P, Oliver WE, Parkin DM, Alvarez N, Vivas J. Case-control study of gastric cancer screening in Venezuela. Br J Cancer 1994; 69:1102–5.[Medline]
- Kunisaki C, Shimada H, Nomura M, et al. Surgical outcome of serosa-negative advanced gastric cancer. Anticancer Res 2004; 24:3169–76.[Medline]
- Kunisaki C, Shimada H, Takahashi M, et al. Implication of extended lymph node dissection stratified for advanced gastric cancer. Anticancer Res 2003; 23:4181–6.[Medline]
- Mizoue T, Yoshimura T, Tokui N, et al. Japan Collaborative Cohort Study Group. Prospective study of screening for stomach cancer in Japan. Int J Cancer 2003; 106:103–7.[Medline]
- Kubota H, Kotoh T, Masunaga R, et al. Impact of screening survey of gastric cancer on clinicopathological features and survival: retrospective study at a single institution. Surgery 2000; 128:41–7.[Medline]
TOP
ABSTRACT
INTRODUCTION
