Dissemination of Free Cancer Cells from the Gastric Lumen and from Perigastric Lymphovascular Pedicles during Radical Gastric Cancer Surgery

Transcription

1 Ann Surg Oncol (2011) 18: DOI /s ORIGINAL ARTICLE GASTROINTESTINAL ONCOLOGY Dissemination of Free Cancer Cells from the Gastric Lumen and from Perigastric Lymphovascular Pedicles during Radical Gastric Cancer Surgery Tae-Su Han, BS 1, Seong-Ho Kong, MD 2, Hyuk-Joon Lee, MD, PhD 1,2, Hye-Seong Ahn, MD 2, Keun Hur, PhD 1, Jieun Yu, MS 1, Woo-Ho Kim, MD, PhD 2,3, and Han-Kwang Yang, MD, PhD, FACS 1,2 1 Cancer Research Institute, Seoul National University College of Medicine, Seoul, Korea; 2 Department of Surgery, Seoul National University College of Medicine, Seoul, Korea; 3 Department of Pathology, Seoul National University College of Medicine, Seoul, Korea ABSTRACT Background. Manipulation and improper handling of a tumor during surgery may increase the risk of cancer cell dissemination after a curative gastrectomy. This study investigated the effect of improper handling of lymphovascular pedicles of stomach on tumor spillage during surgical procedure. Methods. Thirty-eight gastric cancer patients were enrolled. Three pairs of wash samples were obtained from each patient: (1) intraperitoneal wash samples obtained before (P0) and after gastrectomy (P1), (2) intragastric wash samples obtained before any manipulation (G0) and just before resection of the stomach (G1), and (3) ex vivo wash samples obtained by rinsing resected stomach with the lymphovascular pedicles closed by clips (S0) or with the pedicles open (S1). Cytologic examination was performed from all washes, and real-time reverse transcriptase polymerase chain reaction analysis for carcinoembryonic antigen was performed from washes P0, P1, S0, and S1. Results. Cytologic examination detected cancer cells in 34.2% (13 of 38) of G0 samples and in 39.5% (15 of 38) of G1 samples. The rate of conversion from G0-negative to G1-positive increased as T stage increased. Cytologic examination detected cancer cells in 2.6% (1 of 38) of S0 samples and in 13.2% (5 of 38) of S1 samples. The carcinoembryonic antigen mrna level of the S1 sample was 2-fold greater than that of the S0 sample in 50.0% (7 of 14). Ó Society of Surgical Oncology 2011 First Received: 11 September 2010; Published Online: 1 April 2011 H.-K. Yang, MD, PhD, FACS hkyang@snu.ac.kr Conclusions. Free cancer cells can be released from gastric lumen or lymphovascular pedicles opened during gastric cancer surgery, especially in advanced-stage disease. Care should be taken to minimize spillage from the gastric lumen and lymphovascular pedicles. Radical gastrectomy with extended lymph node dissection is considered to be the principle treatment for gastric cancer. However, many patients experience recurrence after radical gastrectomy, even if no residual tumor remains after surgical resection (R0 resection). There are three routes by which gastric cancer may spread: direct seeding via infiltration through the gastric wall, via blood vessels, and via perigastric lymphatic channels. 1 5 Peritoneal recurrence constitutes 32% to 54% of recurrences of gastric cancer. Peritoneal recurrence is thought to be caused by penetration of the gastric wall by the tumor and dissemination of cancer cells from lymph nodes. However, peritoneal recurrence has been reported even in T1 and N0 cases In these previous reports where peritoneal recurrence even in T1N0, peritoneal recurrence cannot be explained by cancer infiltration through gastric wall. Also in these reports, recurrence in T1N2 cases may be due to cancer cells spilled from lymphovascular vessels during radical surgery. In addition, it has been reported that free cancer cells were detected in the peritoneal cavity after radical gastric cancer surgery. 11 Cytologic evidence of free cancer cells in the peritoneum is regarded as an indicator of a poor prognosis in patients with gastric cancer Previous studies reported that colon and esophageal cancer patients who had negative preoperative cytology was converted to positive cytology after surgical procedures. 16,17 However, the sensitivity and

2 Dissemination of Tumor Cells during Gastrectomy 2819 the specificity of the cytologic examination are relatively low Therefore, reverse transcriptase polymerase chain reaction (RT-PCR) methods for detecting markers of gastric cancer have been used to improve the efficacy of detection of free cancer cells in the peritoneal cavity It has been reported that manipulation of a tumor during surgery increases carcinoembryonic antigen (CEA) mrna level in blood. 31 This report implies that surgical maneuvers could spread tumor cells into lymphovascular channels and the intragastric cavity. It is probable that free cancer cells are introduced into the peritoneal cavity when the surgeon opens the gastric wall or the lymphovascular vessels. Therefore, we hypothesized that during radical gastric cancer surgery, cancer cells from the gastric lumen and from the lymphovascular vessels can be sources of peritoneal seeding. In this study, we analyze the influence of manipulation of stomach on tumor spillage in gastric lumen and to determine whether improper closure of lymphovascular vessels results in spillage of cancer cells during radical gastric cancer surgery. MATERIALS AND METHODS Patients Thirty-eight patients in whom gastric cancer was histologically confirmed by endoscopic biopsy at the Seoul National University Hospital between 2007 and 2008 were enrolled. Informed consent was obtained before surgery. This study was approved by the institutional review board of the Seoul National University Hospital on June 29, 2006 (IRB No. H ). The clinicopathological features of the patients are summarized in Table 1. Surgical Procedure and Sampling Methods A nasogastric tube was inserted into the stomach before surgery. If the tumor is resectable, 200 ml of normal saline was delivered into the stomach through the nasogastric tube while the duodenum was clamped off by an intestinal clamp. As much of the fluid as possible was aspirated and designated as the first intragastric sample (G0). The peritoneal cavity was washed with 500 ml of saline, which was then aspirated with a suction tube to a clean bottle and designated as the first peritoneal sample (P0). Radical subtotal distal gastrectomies were performed for cancers located in the lower third and the distal part of the middle third of the stomach. Total gastrectomies with or without splenectomies were performed for cancers in the upper third and the proximal middle third of the stomach. D2 lymphadenectomies were performed according to the Japanese Classification of Gastric Carcinoma (second English TABLE 1 Clinicopathological features of 38 patients Characteristic Value Sex, M:F, n 25:13 (1.9:1) Age, y, mean (SD) 60.9 (11.7) Operation, n (%) Subtotal gastrectomy 30 (78.9) Total gastrectomy 8 (21.1) Histology, n (%) Differentiated (Pap, WD, MD) a 16 (42.1) Undifferentiated (PD, MUC, SRC) b 22 (57.9) Lauren classification, n (%) Intestinal 17 (44.7) Diffuse 13 (34.2) Mixed 8 (21.1) T classification, n (%) T1 7 (18.4) T2 15 (39.5) T3 16 (42.1) N classification, n (%) N0 12 (31.6) N1 14 (36.8) N2 8 (21.1) N3 4 (10.5) TNM stage, n (%) I 14 (36.8) II 4 (10.5) III 16 (42.1) IV 4 (10.5) a Pap papillary adenocarcinoma, WD well differentiated, MD moderately differentiated b PD poorly differentiated, MUC mucinous adenocarcinoma, SRC signet ring cell carcinoma edition). 32 The staging of gastric cancer was based on the 6th American Joint Committee on Cancer/International Union Against Cancer tumor, node, metastasis system (TNM) classification. 33 During surgery, all the named vessels such as right and left gastroepiploic arteries, right and left gastric arteries and lymphatic vessels were sealed securely with removable surgical clips (Hem-O-Lok, Weck, Raleigh, NC) and the electrothermal bipolar vessel sealer (LigaSure, Valleylab, Boulder, CO) or ultrasonic shears (Harmonic Scalpel, Ethicon Endo-Surgery Inc, Cincinnati, OH). During surgery, the proximal and distal resection lines of the stomach were closed with surgical clamps or linear staples to avoid spillage of intragastric contents. Just before the stomach was resected, intragastric irrigation using the nasogastric tube was repeated and collected (G1). Before closure of abdominal incision, peritoneal irrigation was repeated and collected (P1). The resected stomach was transferred to a basin containing 500 ml of saline and rinsed for 15 s. The saline was

3 2820 T.-S. Han et al. collected and designated as S0. The resected stomach was then placed in another basin and sealed lymphovascular channels were opened by removing clips and opening up the coagulated vessels by scissors to simulate the inadequate sealing of the lymphovascular channels during surgery (ex vivo simulation model, Fig. 1). The stomach was rinsed again in the same way. This saline sample was designated as S1. A half of each sample (P0, P1, S0 and S1) was used for cytologic examination and the other half was used for CEA real-time RT-PCR analysis immediately after sampling. Pieces of primary tumor tissue from 32 patients whose tumors were large enough to be sampled without affecting the pathologic examination were collected for CEA mrna RT-PCR analysis. A strip of normal mucosa was also collected. As the tumors of the other six patients were too small to be sampled without affecting the pathologic examination, samples were only taken for cytologic examination in these cases. Cytologic Examination Each saline wash sample (intragastric G0 and G1, peritoneal P0 and P1, resected specimen S0 and S1) were centrifuged and the cell pellets were examined by Papanicolaou staining. CEA RT-PCR for Primary Tumors and Normal Mucosa Total RNA was extracted from 32 primary tumors and corresponding normal mucosa with Trizol (Molecular FIG. 1 Method used for obtaining the S0 and S1 samples from specimens after the gastrectomy. S0 implies the peritoneal washing fluid after our actual surgery. S1 implies peritoneal washing fluid after theoretical surgery with inadequate sealing of lymphovascular vessels (a). The resected stomach was transferred to a basin containing 500 ml of saline and rinsed for 15 s. The saline was collected and designated as the surgical sample 0 (S0) (b). The resected stomach was placed in the other basin and visible sealed lymphovascular channels were opened by removing clips and opening up the coagulate vessels by scissors to simulate the inadequate sealing of the lymphovascular channels during surgery (c) a b c

4 Dissemination of Tumor Cells during Gastrectomy 2821 Research Center, Inc., Cincinnati, OH), and cdna was produced with the Superscript III kit (Invitrogen, Carlsbad, CA) in accordance with the procedures suggested by the manufacturer. RT-PCR was performed for CEA with primers were 5 0 -AATAATAACGGGACCTATGCCTGT-3 0 (forward) and 5 0 -GGAGAAGTTCCAGATGCAGAGAC-3 0 (reverse). The primers for amplification of GAPDH mrna were 5 0 -ACCAGGGCTGCTTTTAACTCT-3 0 (forward) and 5 0 -GAGTCCTTCCACGATACCAAA-3 0 (reverse). The conditions for denaturing, annealing and extension of the template cdna were as follows: 94 C for 30 s, 57 C for 30 s and 72 C for 1 min for 28 cycles. PCR products were separated by electrophoresis on 2% agarose gels and visualized with ethidium bromide. Patients with primary tumors expressing strong CEA band were subjected for CEA realtime RT-PCR analysis for wash samples. CEA Real-Time RT-PCR for Wash Samples (P0, P1, S0, and S1) Wash samples were centrifuged at 5,000 rpm for 20 min. Total RNA was isolated from the cell pellet with Trizol (Molecular Research Center, Cincinnati, OH). cdna was produced from total RNA with a Superscript III kit (Invitrogen, Carlsbad, CA) in accordance with the procedures suggested by the manufacturer. RNA quantifications were conducted by real-time RT- PCR by the comparative Ct method with Applied Biosystems 7300 Real-Time PCR system (Applied Biosystems, Foster City, CA). The real-time RT-PCR reaction was conducted with a SyBR green 29 PCR Master Mix (Applied Biosystems, Foster City, CA) using the same CEA primers used in RT-PCR. The primers for amplification of GAPDH mrna were 5 0 -TGGAGTCCACTGGCGTCTTC-3 0 (forward) and 5 0 -TTCACACCCATGACGAACATG-3 0 (reverse). The cycling parameters were 2 min at 50 C and 10 min at 95 C, followed by 40 cycles of 15 s at 95 C and 1 min at 60 C. Immunohistochemical Staining of CEA in Gastric Cancer Tissues Four-micron sections from paraffin blocks of the primary tumor were cut for immunohistochemical staining with the anti-cea antibody (M7072, Dako, Glostrup, Denmark). Paraffin sections were deparaffinized by incubation with xylene and rehydrated in alcohol. Endogenous peroxidase was blocked with 3% hydrogen peroxide. Sections were incubated for 1 h at room temperature with anti- CEA antibody (M7072, Dako, Glostrup, Denmark). After rinsing with phosphated-buffered saline for 10 min, the sections were exposed to the secondary antibody, a biotinylated anti-mouse IgG antibody for 30 min. Antibody binding was visualized with amino ethyl carbazole, and the sections were counterstained with Mayer s hematoxylin for 10 s. Statistical Analysis Statistical analysis was performed by the v 2 test, Fisher s exact test, t-test, and Wilcoxon signed rank test. A P value of less than 0.05 was regarded as statistically significant. All statistical tests were performed by SPSS version 17.0 software (SPSS, Chicago, IL). RESULTS Cytologic Examination of Intragastric Wash Samples (G0, G1) The proportion of positive cytologic results for the G0 and G1 samples increased as T stage increased. The percentage of positive results was increased by surgical manipulation (34.2% in G0 to 39.5% in G1), but the difference was not statistically significant (Table 2). Five patients showed positive conversion from G0 to G1 after surgical manipulation of the stomach (T1; 1 case, T2; 1 case, T3; 3 cases). CEA mrna and Protein Expression in the Primary Gastric Cancer Tissues Sixteen primary tumors out of 32 expressed a strong CEA band. Expression of CEA protein in these 16 tumors was validated by immunohistochemical staining of the primary tumor tissues with the monoclonal mouse anti- CEA antibody (data not shown). Further analysis CEA real-time RT-PCR of P0, P1, S0 and S1 was done only in these 16 cases with primary tumor expressing a strong CEA band. Cytologic Examination and CEA Real-Time RT-PCR Analysis of Intraperitoneal Wash Samples (P0, P1) Cytologic examination did not reveal any free cancer cells in the P0 samples (0 of 38). However, one patient had TABLE 2 Results of cytologic examination of intragastric wash samples (G0, G1) according to T stage T stage Intragastric washing cytology G0 G1 P value 1 1/7 (14.3%) 2/7 (28.6%) /15 (33.3%) 5/15 (33.3%) /16 (43.8%) 8/16 (50.0%) Total 13/38 (34.2%) 15/38 (39.5%) 0.480

5 2822 T.-S. Han et al. free cancer cells in the P1 sample (2.6%, 1 of 38). The stage of this patient was T3N2M0. CEA real-time RT-PCR analysis was performed on the P0 and P1 wash samples from the 16 patients who expressed high levels of CEA mrna in their primary tumor tissue. Four of these patients were excluded from the analysis because poor RNA sample quality of cell pellets (data not shown). Four of 12 patients (33.3%, 4 of 12) showed an increase in CEA mrna expression of more than 2-fold from the P0 sample to the P1 sample (Fig. 2a). The mean CEA mrna level did not differ significantly between the P0 and P1 samples (Fig. 2b). Cytologic Examination and CEA Real-Time RT-PCR Analysis of Wash Samples from the Ex Vivo Simulation Model (S0, S1) In S0 samples, positive cancer cytology was observed in only one case (T3N2M0) (2.6%, 1 of 38), whose peritoneal wash sample after operation (P1) was also positive. In S1 samples, cytology was positive in 5 cases (13.2%, 5 of 38), especially those with advanced stage cancers. In four patients (10.5%, 4 of 38), conversion from a negative result for the S0 sample to a positive result for the S1 sample was identified. The difference in cytology positivity between S0 and S1 was statistically significant (P = 0.046) (Table 3). In CEA real-time RT-PCR analysis of the S0 and S1 wash samples of the 16 patients whose primary tumor expressed CEA mrna, two pairs of samples were excluded because of poor RNA sample quality (data not shown). Seven of 14 patients (50.0%) showed an increase in CEA mrna expression of more than 2-fold from the S0 sample to the S1 sample (Table 3, Fig. 2c). The mean CEA mrna level for the S1 samples was significantly higher than that for the S0 samples (P = 0.035) (Fig. 2d). DISCUSSION The aim of this study was to analyze the effect of the surgical manipulation of gastric cancer on cancer cell FIG. 2 Results of CEA realtime RT-PCR analysis. The CEA mrna levels of the P0 samples were similar to those of the P1 samples (a). The mean CEA mrna levels of the 2 intraperitoneal wash samples (P0 and P1) did not differ significantly (P = 0.533) (b). The CEA mrna levels of the S1 samples tended to be greater than those of the S0 samples (c). The mean CEA mrna levels for the 2 specimen wash samples (S0 and S1) were significantly different (P = 0.035) (d). CEA mrna expression level of S0 samples that were simulated as ex vivo was similar to expression level of P1 samples a c b d

6 Dissemination of Tumor Cells during Gastrectomy 2823 TABLE 3 Results of cytologic examination and CEA real-time RT-PCR analysis of specimen wash samples (S0 and S1) according to TNM stage a TNM stage Cytologic examination (n = 38) CEA real-time RT-PCR (n = 14) S0 S1 P value S1/S0 [ 2 b I 0/14 (0.0%) 1/14 (7.1%) /3 (66.7%) II 0/4 (0.0%) 0/4 (0.0%) /1 (0.0%) III 1/16 (6.3%) 3/16 (18.8%) /8 (37.5%) IV 0/4 (0.0%) 1/4 (25.0%) /2 (100.0%) Total 1/38 (2.6%) 5/38 (13.2%) /14 (50.0%) a Ex vivo wash samples obtained from resected stomach and perigastric tissues with the lymphovascular pedicles clamped off (S0) or open (S1) b In CEA real-time PCR, S1/S0 [ 2 showed a more than 2-fold increase from S0 sample to S1 sample spillage and to provide scientific evidence for the oncologic principle of less manipulation to be able to prevent cancer cell spillage during radical gastric cancer surgery. However, this oncologic principle has rarely been studied in gastric cancer because of ethical issues. Furthermore, because now we use clips or energy-based devices to close lymphovascular channels as much as possible, we cannot reproduce the situation of inadequately controlled lymphovascular channels during an actual operation. Therefore, we designed a simple ex vivo model of inadequate lymphovascular control, which was designated as the S1 sample. Because the cancer cell can also spread via the gastric lumen or the peritoneal cavity, we examined the presence of cancer cells with irrigation of the gastric lumen before and after manipulation (G0, G1) as well as peritoneal washes (P0, P1). Intragastric cancer cell positivity increased by stage as well as by surgical maneuver. Five out of thirty eight patients (13.2%) converted from negative intragastric cytology (G0) to positive cytology (G1) during the operation. Additionally, about half of the T3 patients showed positivity in G1. Interestingly, free cancer cells were found from intragastric fluid even in the patients with T1 gastric cancer (G0; 1 of 7, G1; 2 of 7). These results suggest that manipulation of the stomach with gastric cancer can increase the detachment of tumor cells into the gastric lumen and that these tumor cells could spread into the peritoneal cavity when the lumen of the stomach is opened during the operation. This result suggests that the free cancer cells in the gastric lumen can be the source of peritoneal recurrence in serosa-noninfiltrating node negative gastric cancer patients. In most cases, the resection lines are well controlled in oncologic surgery, and the purpose of resection line control is mainly to prevent potential bacterial contamination. However, this study showed that the surgeon should control the resection line not only to prevent the bacterial contamination but also to prevent spillage of the free cancer cell. Gastric irrigation with saline solution through a nasogastric tube may be a helpful preventive method to minimize spillage of gastric cancer cells. In the peritoneal cavity, only one patient (2.6%, 1 of 38) showed positive conversion of their cytology from P0 to P1 after the operation. The patient was T3N2M0 (lymph node metastasis in 11 out of 38 resected lymph nodes). Considering the high number of metastatic nodes in this patient, the cytology conversion is not surprising. A low incidence (2.6%, 1 of 38) of positive results from the peritoneal cavity after surgery and low levels of CEA real-time RT-PCR amplification in P1 as well as P0 suggests that our operation adequately controlled spillage of cancer cells from the gastric lumen or the lymphovascular channels. In addition, previous studies reported that extensive intraoperative peritoneal lavage (EIPL) can be a useful prophylactic strategy for peritoneal recurrence in patients with gastric cancer. 11,34,35 These reports accorded with our results that peritoneal recurrence was caused by releasing of free cancer cells during gastric cancer surgery. Other investigators reported that the positive cytologic examination ranged from 16 to 43%. 12,18,36 However, most of these studies included 50 to 90% of gastric cancer patients with peritoneal carcinomatosis. Our study only included gastric cancer without evident distant metastasis, which may explain the low positive rate of cytologic examination. Therefore, additional real-time RT-PCR for CEA was used to improve the efficacy of detection of free cancer cells in the peritoneal cavity in our study. Although the method of CEA real-time RT-PCR has a high sensitivity of cancer cell detection, it also can be contaminated by CEA expressed in normal mucosa or blood cell components if the cycle number is too much increased, even though it has been reported that CEA mrna expression levels that had been detected in the peripheral blood samples from health volunteers were relatively low. 11,37 In previous reports, CEA real-time RT-PCR was performed 45 or more cycles, regardless of CEA expression level of primary tumor. 11,23,37 In this study, however, CEA amplification was limited to 40

7 2824 T.-S. Han et al. cycles only in samples with high CEA expression level in the primary tumor to minimize contamination. Regarding cancer cells from the lymphovascular channels, only one patient (2.6%, 1 of 38) who had 11 positive metastatic lymph nodes was cytology positive in the S0 sample. This patient s peritoneal wash sample (P1) was also the one that converted from negative to positive during the operation. Considering the P1 sample is theoretically similar to S0 sample, this result is quite reasonable. Four patients (10.5%, 4 of 38) showed positive conversion of cytologic examination from S0 to S1 after the lymphovascular channels were opened, and one of them was stage I. This means that lymphovascular cancer cell spillage can be accompanied even in stage I, if lymphovascular channels are not properly secured. The CEA mrna levels of seven of 14 patients (50.0%) also increased more than twice after the lymphovascular vessels on specimen side were opened (S1) compared to the sealed channels (S0). In the previous report, free cancer cells were detected in the peritoneal cavity, which suggests that cancer cells can be spilled out from lymphovascular channels. 11 In this report, using ex vivo specimen, we clearly demonstrated that cancer cells are spilled out from lymphovascular channels. And we emphasize that during radical gastric cancer surgery, lymphovascular vessels should be carefully secured either with clips or energy-based devises to avoid cancer cell spillage, which can result in peritoneal seeding. These results also do not recommend performing a sharp dissection with scissors or scalpel, which will open lymphovascular vessels during dissection. Our results of intragastric cytologic study (G0, G1) as well as ex vivo surgical specimen study (simulating improper closure of lymphovascular vessels; S0, S1) can provide possible explanation for the peritoneal recurrence in early gastric cancer either without lymph node metastasis or with lymph node metastasis. In conclusion, this study demonstrates using ex vivo simulation that free cancer cell release is increased into the gastric lumen and from lymphovascular channel during radical gastric cancer surgery. This result indicates the experimental and theoretical background of the oncologic surgical principle that manipulation during gastric cancer surgery should be minimized. Also, this study indirectly showed that the spillage of free cancer cells from gastric lumen or from lymphovascular vessels could be prevented by using an energy-based device and a carefully controlled resection line. Therefore, on the basis of this study, surgeons should take care to prevent any spillage of gastric contents into peritoneal cavity during surgery and make secure the closure of lymphovascular vessels, especially on the specimen side during dissection. ACKNOWLEDGMENT This study was supported by a grant from the Seoul National University College of Medicine Research Fund ( ), which was a donation from Heung Soon Park. The authors would like to thank Dr. Cristina Ferrone for her review. REFERENCES 1. Moriguchi S, Maehara Y, Korenaga D, et al. The relationship between prognostic-significance of pathological type and the degree of gastric wall invasion in gastric-cancer. Cancer J. 1992;5: Iriyama K, Miki C, Kitagawa T. Influence of tumour penetration on period of death after gastric cancer surgery. Br J Surg. 1997;84: Noguchi Y. Blood vessel invasion in gastric carcinoma. Surgery. 1990;107: Ikeguchi M, Katano K, Oka A, et al. Relationship between hematogenic metastasis of gastric cancer and the maximum extent of venous invasion by cancer cells in the gastric wall. Hepatogastroenterology. 1995;42: Maehara Y, Oshiro T, Baba H, et al. Lymphatic invasion and potential for tumor growth and metastasis in patients with gastric cancer. Surgery. 1995;117: Maehara Y, Hasuda S, Koga T, et al. Postoperative outcome and sites of recurrence in patients following curative resection of gastric cancer. Br J Surg. 2000;87: Schwarz RE, Zagala-Nevarez K. Recurrence patterns after radical gastrectomy for gastric cancer: prognostic factors and implications for postoperative adjuvant therapy. Ann Surg Oncol. 2002; 9: Roviello F, Marrelli D, de Manzoni G, et al. Prospective study of peritoneal recurrence after curative surgery for gastric cancer. Br J Surg. 2003;90: Lee HJ, Kim YH, Kim WH, et al. Clinicopathological analysis for recurrence of early gastric cancer. Jpn J Clin Oncol. 2003; 33: Ichiyoshi Y, Toda T, Minamisono Y, et al. Recurrence in early gastric cancer. Surgery. 1990;107: Marutsuka T, Shimada S, Shiomori K, et al. Mechanisms of peritoneal metastasis after operation for non-serosa-invasive gastric carcinoma: an ultrarapid detection system for intraperitoneal free cancer cells and a prophylactic strategy for peritoneal metastasis. Clin Cancer Res. 2003;9: Boku T, Nakane Y, Minoura T, et al. Prognostic significance of serosal invasion and free intraperitoneal cancer cells in gastric cancer. Br J Surg. 1990;77: Bonenkamp JJ, Songun I, Hermans J, et al. Prognostic value of positive cytology findings from abdominal washings in patients with gastric cancer. Br J Surg. 1996;83: Kodera Y, Yamamura Y, Shimizu Y, et al. Peritoneal washing cytology: prognostic value of positive findings in patients with gastric carcinoma undergoing a potentially curative resection. J Surg Oncol. 1999;72: Homma Y, Ushida S, Yamada M, et al. Positive peritoneal washing cytology in multiple cavities can predict poor prognosis of advanced gastric cancer patients. Ann Surg Oncol. 2010;17: Hase K, Ueno H, Kuranaga N, et al. Intraperitoneal exfoliated cancer cells in patients with colorectal cancer. Dis Colon Rectum. 1998;41: Doki Y, Kabuto T, Ishikawa O, et al. Does pleural lavage cytology before thoracic closure predict both patient s prognosis and site of cancer recurrence after resection of esophageal cancer. Surgery. 2001;130: Abe S, Yoshimura H, Tabara H, et al. Curative resection of gastric cancer: limitation of peritoneal lavage cytology in predicting the outcome. J Surg Oncol. 1995;59:226 9.

8 Dissemination of Tumor Cells during Gastrectomy Cetin B, Atalay C, Aslan S, et al. Peritoneal carcinoembryonic antigen level for predicting locoregional and distant spread of gastric cancer. Surg Today. 2005;35: Li JK, Zheng M, Miao CW, et al. Peritoneal lavage cytology and carcinoembryonic antigen determination in predicting peritoneal metastasis and prognosis of gastric cancer. World J Gastroenterol. 2005;11: Suzuki T, Ochiai T, Hayashi H, et al. Peritoneal lavage cytology findings as prognostic factor for gastric cancer. Semin Surg Oncol. 1999;17: Bando E, Yonemura Y, Takeshita Y, et al. Intraoperative lavage for cytological examination in 1,297 patients with gastric carcinoma. Am J Surg. 1999;178: Kodera Y, Nakanishi H, Ito S, et al. Prognostic significance of intraperitoneal cancer cells in gastric carcinoma: analysis of real time reverse transcriptase polymerase chain reaction after 5 years of followup. J Am Coll Surg. 2006;202: Tamura N, Iinuma H, Takada T. Prospective study of the quantitative carcinoembryonic antigen and cytokeratin 20 mrna detection in peritoneal washes to predict peritoneal recurrence in gastric carcinoma patients. Oncol Rep. 2007;17: Yonemura Y, Fujimura T, Ninomiya I, et al. Prediction of peritoneal micrometastasis by peritoneal lavaged cytology and reverse transcriptase polymerase chain reaction for matrix metalloproteinase-7 mrna. Clin Cancer Res. 2001;7: Sakakura C, Takemura M, Hagiwara A, et al. Overexpression of dopa decarboxylase in peritoneal dissemination of gastric cancer and its potential as a novel marker for the detection of peritoneal micrometastases with real-time RT-PCR. Br J Cancer. 2004; 90: Shimomura K, Sakakura C, Takemura M, et al. Combination of L- 3-phosphoserine phosphatase and CEA using real-time RT-PCR improves accuracy in detection of peritoneal micrometastasis of gastric cancer. Anticancer Res. 2004;24: Nakanishi H, Kodera Y, Torii A, et al. Detection of carcinoembryonic antigen-expressing free tumor cells in peritoneal washes from patients with gastric carcinoma by polymerase chain reaction. Jpn J Cancer Res. 1997;88: Wang Z, Zhang X, Xu H, et al. Detection of peritoneal micrometastasis by reverse transcriptase polymerase chain reaction for heparanase mrna and cytology in peritoneal wash samples. J Surg Oncol. 2005;90: Fujiwara Y, Doki Y, Taniguchi H, et al. Genetic detection of free cancer cells in the peritoneal cavity of the patient with gastric cancer: present status and future perspectives. Gastric Cancer. 2007;10: Miyazono F, Natsugoe S, Takao S, et al. Surgical maneuvers enhance molecular detection of circulating tumor cells during gastric cancer surgery. Ann Surg. 2001;233: Japanese Gastric Cancer A. Japanese classification of gastric carcinoma. 2nd English ed. Gastric Cancer. 1998;1: Greene FL, American Joint Committee on Cancer., American Cancer Society. AJCC cancer staging manual. 6th ed. New York: Springer-Verlag; Shimada S, Tanaka E, Marutsuka T, et al. Extensive intraoperative peritoneal lavage and chemotherapy for gastric cancer patients with peritoneal free cancer cells. Gastric Cancer. 2002; 5: Kuramoto M, Shimada S, Ikeshima S, et al. Extensive intraoperative peritoneal lavage as a standard prophylactic strategy for peritoneal recurrence in patients with gastric carcinoma. Ann Surg. 2009;250: Kodera Y, Nakanishi H, Yamamura Y, et al. Prognostic value and clinical implications of disseminated cancer cells in the peritoneal cavity detected by reverse transcriptase polymerase chain reaction and cytology. Int J Cancer. 1998;79: Nakanishi H, Kodera Y, Yamamura Y, et al. Rapid quantitative detection of carcinoembryonic antigen-expressing free tumor cells in the peritoneal cavity of gastric-cancer patients with realtime RT-PCR on the lightcycler. Int J Cancer. 2000;89:411 7.