Second and third hepatectomies for recurrent hepatocellular carcinoma are justified.

BACKGROUND: Liver resection is the main curative treatment for hepatocellular carcinoma (HCC), but recurrence rates are high. The remnant liver is the most common site of recurrence, but the role of repeat hepatectomy in the treatment of recurrent HCC is controversial. METHODS: Patients who underwent curative hepatectomy for HCC and subsequent repeat hepatectomy for recurrent HCC between 1990 and 2007 were reviewed retrospectively. Clinicopathological characteristics, and early- and long-term outcomes of patients who had a first, second, third and fourth hepatectomy were compared. RESULTS: Some 1177 patients underwent a first hepatectomy for HCC, and 149, 35 and eight patients respectively had a second, third and fourth hepatectomies for recurrence. There were no significant differences in early postoperative outcomes after first and repeat hepatectomies. Five-year disease-free and overall survival rates after first, second and third hepatectomies were 43.6, 31.8 and 33.8 per cent (P = 0.772), and 52.4, 56.4 and 59.4 per cent (P = 0.879), respectively. Patients undergoing second and third hepatectomies for recurrence had better survival rates than those who did not have a repeat hepatectomy, but not those after fourth hepatectomy. CONCLUSION: Second and third hepatectomies seem justified for hepatic recurrence of HCC. The role of fourth hepatectomy needs further investigation.

Induction of plasma protein secretion in a newly established human hepatoma cell line.

To study the expression and the regulation of hepatocyte markers, we have undertaken to establish human hepatoma cell lines of various phenotypes. We now report the establishment of a new human hepatoma cell line, HA22T/VGH. This cell line has many of the properties of human hepatocellular carcinoma. Only 5 of 15 plasma proteins investigated were detected in the medium of a 10-day-old HA22T/VGH culture. However, when the HA22T/VGH cells and a clonal derivative, C5, were cultured in an aggregated form, all 15 plasma proteins were found in the culture medium. These results indicate that hepatoma cell lines with different phenotypes can be established, and they provide a good experimental framework to investigate differentiation of human hepatocytes.

Transcripts of the insulin-like growth factors I and II in human hepatoma.

Using complementary DNAs of human insulin-like growth factors as probes, expressions of the insulin-like growth factors I and II mRNA were examined in seven human hepatoma tissues and their adjacent nontumorous livers. The level of insulin-like growth factor I mRNA in hepatoma was lower than that in the nontumorous liver control. This phenomenon was probably caused by the low expression of human growth hormone receptor in hepatoma tissues. The levels of insulin-like growth factor II mRNA vary among hepatomas. Some show elevated expression; some have diminished expression compared to their nontumorous liver counterparts. In four of the seven hepatomas, expression of fetal forms of insulin-like growth factor II transcripts was observed and may represent dedifferentiation of insulin-like growth factor II expression during hepatocarcinogenesis.

Consequences of opiate agonist and antagonist in myocardial ischaemia suggest a role of endogenous opioid peptides in ischaemic heart disease.

OBJECTIVE: The aim was to evaluate the effects of an opiate agonist (U50,488H) and an opiate antagonist (naloxone) in myocardial ischaemia. METHODS: A left thoracotomy was performed and the left coronary artery was ligated in adult Sprague-Dawley rats of either sex (350-400 g). Blood pressure, heart rate and electrocardiogram were measured before and after injections of U50,488H or naloxone and throughout the 30 min postligation period. RESULTS: Following coronary artery occlusion, all rats in the control group developed arrhythmias, bradycardia, and hypotension. U50,488H potentiated and naloxone attenuated the ischaemia induced arrhythmias, bradycardia, and hypotension. CONCLUSIONS: The potentiating and blocking effects of U50,488H and naloxone, respectively, suggest that endogenous opioid peptides are involved in the pathophysiology of myocardial ischaemia and play an important role in ischaemic heart disease.

Immunohistochemical study of arginase in cancer of the stomach.

High levels of arginase have been detected in gastric adenocarcinoma. To examine the hypothesis that this is due to macrophage infiltration into the tumour, we localized the cellular distribution of arginase by immunohistochemical staining. We examined gastric adenocarcinomas and their corresponding normal tissues (n = 45), leiomyomas (n = 2), leiomyosarcomas (n = 3), human gastric adenocarcinoma cell lines (n = 3), and benign gastric ulcers (n = 4) by the avidin-biotin-peroxidase complex technique. Macrophages with strong arginase immunoreactivity were observed infiltrating both gastric normal and cancer tissues. No arginase immunoreactivity was observed in normal mucosal gland, muscular and serosal tissues or benign gastric ulcers. The immunoreactivity of arginase was positive but heterogeneous in most specimens of gastric adenocarcinoma (62.2%) and was absent from gastric intestinal metaplasia, leiomyomas and leiomyosarcomas. Among the 28 neoplasms with arginase immunoreactivity, scattered immunoreactivity was also noted in adjacent dysplastic glands in 12 (42.8%) specimens. Arginase immunoreactivity was observed in all three gastric cancer cell lines. Arginase is present in the cytoplasm but not in the nucleus. These data suggest that the high arginase levels in adenocarcinoma cancer tissues originate largely from cancer cells.

Carcinoid tumor of the pancreas causing the diarrheogenic syndrome: report of a case combined with multiple endocrine neoplasia, type I.

A case of carcinoid tumor of the pancreas with the watery diarrhea, hypokalemia, and hypochlorhydria syndrome in association with hyperparathyroidism and the amenorrhea-galactorrhea syndrome is presented. Resection of the three grossly enlarged, hyperplastic parathyroid glands restored eucalcemia in this patient. A subsequent excision of the 370 gm pancreatic carcinoid tumor resulted in a cure of the watery diarrhea and a return of the gastric acid secretion to normal. Immunocytochemical studies of the pancreatic tumor demonstrated a positive stain only for serotonin, and negative results for vasoactive intestinal polypeptide, pancreatic polypeptide, glucagon, insulin, cholecystokinin, gastrin, and calcitonin were obtained. These studies suggest that in this patient, serotonin was a causative agent of the watery diarrhea syndrome.

A clinical evaluation of ultrasonography in the diagnosis of acute appendicitis.

Recently ultrasonography (US) has proven effective in the diagnosis of acute appendicitis. However, the impact of US on surgical decision-making in clinical practice remains unclear. From August 1986 to July 1987, 62 patients with clinical signs of acute appendicitis received US examinations after initial clinical evaluations. According to clinical pictures, they were classified into two groups: group I, clinically typical (Alvarado's score greater than or equal to 7) and group II, clinically dubious (Alvarado's score less than or equal to 6) cases. All group I patients (n = 34) had surgery. Group II patients (n = 28) received close in-hospital observation on an every 4-hour basis for less than 24 hours. However, surgery was performed if Alvarado's score increased up to or above 7 or surgical indications became evident during observation. The pathologic reports and operative findings were retrospectively correlated with the US diagnosis. For group I patients, clinical examination without knowledge of US findings was correct in 30 (88.9%) cases, 26 (86.7%) of which were detected by US. Four (50%) of eight patients with negative US findings had acute appendicitis. In group II, 12 of 15 patients who had surgery had acute appendicitis, and 10 (83.3%) were detected by US. Two (11%) of 18 patients with negative US findings were finally proved to have acute appendicitis. The average duration from initial clinical examination to surgical decision fpr patients with acute appendicitis was 6.8 hours. There were no false-positives in either group. For patients with typical clinical presentation, US is not superior to clinical examination, and surgery is recommended even if US findings are negative.(ABSTRACT TRUNCATED AT 250 WORDS)

Hepatic resection for bilobar multicentric hepatocellular carcinoma: is it justified?

BACKGROUND: Hepatic resection for multiple hepatocellular carcinomas (HCCs) involving both lobes of the liver is rarely recommended because of high operative risks and low radicality. Thus the justification of hepatic resection for bilobar multicentric HCC remains undefined. METHODS: Two hundred eleven patients with HCC, who underwent curative hepatic resection, were studied retrospectively. The patients were divided into two groups. Group A consisted of 39 patients with bilobar (both sides of Cantlie's line) multicentric HCCs. Group B consisted of 172 patients with HCC with solitary or unilobar lesions. The backgrounds and resectional results of patients in groups A and B were compared. RESULTS: Patients in group A usually required multiple separate liver resections and a longer operative time. However, the operative blood loss, amount of blood transfused, and operative morbidity and mortality rates were not significantly different. Patients in group A showed higher incidences of associated satellite nodules, microscopic vascular invasion, and a lack of capsules. The 6-year disease-free and actuarial survival rates of patients in groups A and B were 30.5% and 41.8% (p = 0.17) and 42.9% and 51.4% (p = 0.12), respectively. For patients in group A the presence of satellite nodules in any resected tumor was the only independent unfavorable feature that influenced the actuarial survival rate after multivariate analysis. CONCLUSIONS: Liver resection is justified for bilobar multicentric HCCs in selected patients, if the tumors can be totally resected. Postoperative adjuvant therapies should be considered when satellite nodules are present in any resected tumor.

Hepatic resection of hepatocellular carcinoma in cirrhotic livers: is it unjustified in impaired liver function?

BACKGROUND: Resection for hepatocellular carcinoma in patients with cirrhosis and impaired liver function is usually unjustified because of higher surgical risks and poorer long-term prognosis. METHODS: A retrospective comparison of the background and resectional results of patients with cirrhosis and hepatocellular carcinoma was carried out between those with preoperative indocyanine green 15-minute retention rate of 20% or greater (group A, impaired function group, n = 36) and those with indocyanine green 15-minute retention rate of 10% or less (group B, normal function group, n = 34). RESULTS: The group A patients had significantly lower serum albumin level and higher serum bilirubin level, longer prothrombin time, higher incidence of associated esophageal varices, and poorer Child's classifications for cirrhosis. Although the tumor diameter in both groups was similar (A, 6.9 versus B, 7.1 cm; p = 0.94), the resected liver amount in group B was greater (227.4 versus 473.2 gm; p = 0.038) because of a greater extent of liver resection (p < 0.001) and a wider surgical margin (0.34 versus 1.85 cm; p < 0.0001). The amount of operative blood loss and blood transfusion, operative morbidity, and operative mortality were not different between the two groups. The pathologic characteristics and staging were also comparable. The 5-year disease-free and actuarial survival rates of groups A and B were 30.9% and 29.6% (p = 0.16) and 45.2% and 33.4% (p = 0.11), respectively. CONCLUSIONS: If the amount of resected nontumorous liver parenchyma could be reduced, resection of hepatocellular carcinoma in selected patients with cirrhosis and impaired liver function is still justified in spite of a narrow surgical margin.

Liver resection for hepatocellular carcinoma in octogenarians.

BACKGROUND: Liver resection is risky in patients aged > or = 80 years. Because of short life expectancies and improved nonoperative modalities, the role of liver resection in octogenarians with hepatocellular carcinoma (HCC) is unclear. METHODS: A retrospective review of the operative results of 260 patients with HCC between 1991 and 1997 was performed. According to the age at the time of operation, these patients were divided into 2 groups. Group 1 comprised 21 patients aged > or = 80 years, and group 2 comprised the other 239 younger patients. The backgrounds, pathologic features of the tumor, and operative results of the patients were compared. RESULTS: Octogenarians had a higher incidence of associated medical diseases, a higher incidence of negative serum hepatitis B surface antigen, a lower alpha-fetoprotein level, and a higher indocyanine green retention rate. Although octogenarians had a longer postoperative hospital stay, there were no significant differences between the 2 groups regarding operative morbidity and mortality. The 5-year disease-free and actuarial survival rates for octogenarians and younger patients were 50.6% and 35.3% (P = .15) and 40.9% and 59.3% (P = .46), respectively. CONCLUSION: Under meticulous preoperative assessments and postoperative care, liver resection for HCC is justified in selected octogenarians, with short- and long-term results comparable to those of younger patients.

The clinical aspect of retained gastric antrum.

Thirty symptomatic patients with retained gastric antrum proved pathologically and/or by isotopic visualization were studied and treated from 1968 to 1983. The latent periods from the antral exclusion to the occurrence of anastomotic ulcers after a subtotal gastrectomy with Billroth's type II reconstruction varied from a few days to 19 years, with an average of 2.8 years. Fasting serum gastrin levels were normal in 14 of 21 patients and were intermittently high in some patients. The basal to maximal acid-output ratio was greater than 0.6 in 69% of the patients; primary cimetidine treatment was effective in three of five patients. Thirty-five operations on 27 patients were divided into six groups; all of these 27 patients eventually underwent resection of retained gastric antrum. We concluded that resection remains the best treatment for anastomotic ulcer related to retained gastric antrum. Additional truncal vagotomy did not provide additional benefit to these patients. Furthermore, cimetidine can be useful to control the symptoms for preoperative preparation or definitive treatment in high-risk patients.

Sodium pertechnetate Tc 99m antral scan in the diagnosis of retained gastric antrum.

Retained gastric antrum (RGA) is a major factor in recurrent peptic ulcer. We studied 121 patients with proven anastomotic ulcers following subtotal gastrectomy and Billroth II reconstruction with sodium pertechnetate Tc 99m to determine the presence of RGA. Of the patients, 59 required surgery, 22 had RGA, and 16 had a positive scan for RGA. This noninvasive examination has 100% specificity. If RGA is identified before operation, minimal time is wasted in exploration of the abdomen, which is especially important in dealing with emergency cases. When an antral scan is negative for RGA, the surgeon is still advised to search for this condition.

An appraisal of liver and portal vein resection for hepatocellular carcinoma with tumor thrombi extending to portal bifurcation.

BACKGROUND: The role of surgical resection for hepatocellular carcinoma with tumor thrombi involving the major portal vein is controversial because of a high operative risk and poor prognosis. Previously, a resection was performed only when the tumor thrombi were limited to the first branch of the portal vein without extension to the portal bifurcation. HYPOTHESIS: Concomitant liver and portal vein resection may be beneficial in patients with hepatocellular carcinoma with tumor thrombi extending to portal bifurcation. DESIGN: Retrospective review. SETTING: University hospital, tertiary referral center. PATIENTS: Among 368 patients with hepatocellular carcinoma who underwent a curative resection, portal vein involvement occurred in 112 patients. Fifteen of the 112 patients underwent a concomitant liver and portal vein resection owing to extension of tumor thrombi to the portal bifurcation (group 1). The remaining 97 patients did not need portal vein resection (group 2). INTERVENTION: Surgical indications, procedures, and results of pathological examination of resected specimens were assessed in patients in group 1. The clinicopathological characteristics, operative morbidity and mortality, and operative results were compared between the 2 groups. MAIN OUTCOME MEASURES: Disease-free and actuarial survival rates. RESULTS: Intramural tumor infiltration was found at the site of thrombi adhesion to the portal vein cuff in 11 of 15 patients in group 1. Owing to patient selection bias, patients in group 1 were significantly younger and had better liver function and greater resected liver weight. The operative time, postoperative hospitalization, operative blood loss, amount of blood transfusion, and operative morbidity and mortality did not differ significantly between the 2 groups. The 5-year disease-free survival rates of groups 1 and 2 were 21.6% and 20.4% (P =.19), respectively, while the actuarial survival rates were 26. 4% and 28.5% (P =.33), respectively. CONCLUSION: Liver resection with partial resection of the portal vein is justified in selected patients with hepatocellular carcinoma with tumor thrombi extending to portal bifurcation.

Prediction and limitation of hepatic tumor resection without blood transfusion in cirrhotic patients.

BACKGROUND: The need for blood transfusion in cirrhotic liver resection is difficult to determine because of inaccurate estimation of operative blood loss. Moreover, blood transfusion is detrimental to cirrhotic patients. OBJECTIVE: To investigate the predictors and limitations of hepatectomy without blood transfusion for cirrhotic patients. DESIGN: Retrospective study. SETTING: University hospital, a tertiary referral center. PATIENTS: A consecutive 163 cirrhotic patients underwent resection for liver tumor(s) under a policy of restrictive blood transfusion. INTERVENTIONS: Estimated blood losses and clinicopathological features of patients who received and those who did not receive a blood transfusion were compared. MAIN OUTCOME MEASURES: Estimated operative blood losses, preoperative assessments, and operative procedures. RESULTS: There were 48 patients in the group who received a blood transfusion, with 1275 +/- 650 mL (mean +/- SE) of blood transfused, and 115 patients in the group who did not receive a blood transfusion. From discriminant analysis, the cutoff value of estimated blood loss for blood transfusion was 1685 mL. Tumor size and site of hepatectomy were found to be independent variables influencing blood transfusion under logistic regression analysis. CONCLUSIONS: Most cirrhotic patients tolerate hepatectomy without blood transfusion when the estimated operative blood loss is less than 1600 mL. Hepatectomy can be performed in cirrhotic patients without blood transfusion if the tumor is small (<5 cm), and/or the resection area is confined to Couinaud segments II, III, and VI. In this study, the largest amount of estimated blood loss in cirrhotic liver resection without blood transfusion was 2350 mL, but the uppermost limit remains to be determined.

An alternative approach of choledocholithotomy via laparoscopic choledochotomy.

OBJECTIVE: To evaluate the safety and feasibility of laparoscopic choledocholithotomy via choledochotomy for the treatment of choledocholithiasis. DESIGN: A prospective series of 1332 consecutive patients who underwent laparoscopic cholecystectomies, with a mean follow-up of 21.2 months. SETTING: University-affiliated referral center. PATIENTS: Forty-three patients (3%) with documented common bile duct stones from January 1991 to February 1995. INTERVENTIONS: Laparoscopic choledocholithotomy with choledochotomy and T tube drainage were performed in 40 patients. Postoperative endoscopic sphincterotomy after laparoscopic cholecystectomy was performed in three patients. MAIN OUTCOME MEASURES: Documented removal of common bile duct stones and procedure-related complications. RESULTS: Laparoscopic choledocholithotomy via choledochotomy was successful in 35 (88%) of 40 patients in whom this procedure was attempted. The mean (+/- SD) operation time was 191.3 +/- 75.4 minutes, and the mean (+/- SD) length of postoperative stay was 10.4 +/- 2.7 days. Seven complications (18%) were recorded, including three major complications (8%) and two retained stones (5%). CONCLUSIONS: Laparoscopic choledocholithotomy via choledochotomy can be performed safely, without increasing the morbidity rate as compared with that of open choledocholithotomy. Thus, some of the advantages of minimally invasive surgery are preserved.

Hepatic segmentectomy for curative resection of primary hepatocellular carcinoma.

OBJECTIVES: To evaluate the feasibility and results of segmentectomy for curative resection of hepatocellular carcinoma and to compare the clinicopathological findings of the patients according to the tumor location in the liver. DESIGN: Case series. SETTING: A tertiary care center. PATIENTS: Seventy-five patients with Child's grade A or B liver function who had hepatocellular carcinoma that was confined to one segment and who underwent segmentectomy for curative resection of the tumor. The patients were divided into four groups: group P (posterior segmentectomy, n = 23); group A (anterior segmentectomy, n = 10); group M (medial segmentectomy, n = 16); and group L (lateral segmentectomy, n = 26). MAIN OUTCOME MEASURE: Disease-free survival rate. RESULTS: Seventy-three percent of the patients had cirrhosis of the liver. The surgical mortality and morbidity rates were 5.3% and 36.0%, respectively. The 1-, 3-, and 5-year disease-free survival rates were 61.9%, 39.1%, and 26.3%, respectively, and were not significantly different among the four groups (P = .86). Group L had the least operative blood loss and shortest operative time when compared with the other three groups (P < .05). The postoperative liver function changes were mild and transient in the four groups of patients. With regard to pathological factors, only tumor size differed among the groups (tumors in group L were significantly larger than those in the other three groups, P < .05). Forty-three percent of the recurrent tumors were solitary in the early stage, with 81% involving the segment(s) adjacent to the resected one and 57% being confined solely to the segment adjacent to the resected segment. Patients having recurrent hepatocellular carcinomas had significantly larger tumors at the time of resection than did those without recurrence (P = .03). CONCLUSIONS: Hepatic segmentectomy is an effective therapeutic approach for small hepatocellular carcinomas and can be done safely even in patients with chronic liver disease and impaired liver function.

A comparison of regional versus systemic drug injection. Adriamycin concentration in peripheral blood and gastric stump (post-Billroth II gastrectomy) in the dog.

Gastric stump cancer has a low resection rate and poor response to systemic chemotherapy. This study attempted to achieve higher drug concentrations in target tissues by way of regional arterial injection instead of systemic venous administration. A total of 12 male mongrel dogs that had undergone post-Billroth II gastrectomies were randomly divided into two groups: in group A, 1.27 mg/kg adriamycin (ADM) was injected through the left gastric artery; in group B, the same dose of ADM was injected into a vein in the left front leg. Blood samples were taken at various time intervals, and the dogs were sacrificed 2 h after drug administration. Tissues were removed from various parts of the gastric stump for measurement of the ADM concentration. The ADM content in the jejunum, heart, liver, spleen, and pancreas was also determined. The results were as follows: (1) the ADM concentration in the gastric stump near its lesser curvature and stomal mucosa was significantly higher in group A than in group B (P less than 0.05). (2) The ADM concentrations in the adjacent organs (heart, liver, and pancreas) were also significantly higher in group A than in group B (P less than 0.05). (3) The ADM levels in the venous blood were significantly higher in group B than in group A (P less than 0.05). These results indicate that a chemotherapeutic drug given through the left gastric artery provides a higher drug concentration in the area where gastric stump cancer frequently occurs and that a lower systemic blood level may cause fewer adverse drug effects. The high concentration of ADM in the heart may not be a good indication, but it may serve as an important signal either to select a less cardiotoxic drug or to monitor heart function cautiously during drug therapy.

Results of curative gastrectomy for carcinoma of the distal third of the stomach.

BACKGROUND: Carcinoma of the distal stomach metastasizes to lymph nodes along the bile duct, pancreatic head, and duodenum. We reviewed the results of patients who underwent operation for carcinoma of the stomach, and placed special emphasis on survival related to lymph node metastasis. STUDY DESIGN: We conducted a prospective study of 258 consecutive patients with adenocarcinoma of the distal third of the stomach who underwent curative gastrectomies. RESULTS: Most of the patients (193 [75 percent]) had advanced cancer (extension beyond the submucosa). A subtotal gastrectomy was the procedure most commonly performed (89 percent). Combined organ resection was performed in 100 patients (39 percent). The operative morbidity was 17 percent; the most frequent complications were chylous leakage, anastomotic insufficiency, and intra-abdominal infection. Ninety-four patients (36 percent) had tumor recurrence, with local recurrence occurring in 45 patients. In 40 of the 45 cases, the local recurrence resulted from remnant lymph nodes (LNs), or soft tissues in the gastric bed. Lymph node metastases were observed in 152 patients (59 percent). Excluding five patients (2 percent) who died, the overall five-year cumulative survival rate was 53 percent. The five year survival rate was 98 percent for patients with TNM stage I disease; 68 percent for patients with stage II disease; 40 percent for patients with stage III disease; and 10 percent for patients with stage IV disease. The survival rate for patients with n0, n1, n2, and n3 disease was respectively 92 percent, 45 percent, 30 percent, and 20 percent. For patients with metastatic LNs in the hepatoduodenal ligament, the five-year survival rate was 20 percent. CONCLUSIONS: These data suggest that survival rate relates to the extent of LN metastasis. It appears that systematic lymph node dissection may have a beneficial effect. However, the efficacy of radical lymph node dissection can only be determined by prospective, randomized clinical trials with a proper study design.

Morbidity and mortality after radical gastrectomy for patients with carcinoma of the stomach.

BACKGROUND: This study sought to analyze the morbidity and mortality rates after radical gastrectomy for carcinoma, since the operation has been criticized as too morbid for the benefits it may provide. STUDY DESIGN: A prospective study of 474 patients who underwent radical gastrectomy was conducted. RESULTS: The overall morbidity and mortality rates were 20.1 and 3.0 percent, respectively. The morbidity and mortality rates fell significantly from 27.0 to 15.7 percent (p = 0.003) and 5.5 to 1.1 percent (p < 0.001), respectively, after the first 200 cases. By logistic regression analysis, it was found that male gender, combined organ(s) resection, extended lymphadenectomy, respiratory system disease, and tumor location were significantly related to postoperative morbidity. In regard to the extent of lymphadenectomy, relative to R2 resection (n = 102), the odds ratio for morbidity after R3 resection (n = 217) was 2.13, and for R4 resection (n = 155) it was 3.12. Age older than 65 years, total gastrectomy, combined organ(s) resection, and respiratory system disease were factors that negatively affected operative mortality. CONCLUSIONS: These observations suggested that radical gastrectomy can be performed with an acceptable risk of morbidity and mortality in a general hospital.

Optimal surgical strategy for potentially curable serosa-involved gastric carcinoma with intraperitoneal free cancer cells.

BACKGROUND: Radical gastrectomy with systematic lymphadenectomy (RG) remains controversial in the treatment of gastric carcinoma. On the other hand, the prognosis of gastric carcinoma, in the presence of intraperitoneal free cancer cells, is poor. The optimal surgical strategy for serosa-involved gastric carcinoma with intraperitoneal free cancer cells remains undefined. STUDY DESIGN: A prospective study of intraperitoneal cytologic washing was conducted on 134 patients with potentially curable serosa-involved gastric carcinoma who underwent RG. During the same period, 28 patients with resectable tumors who received palliative simple gastrectomies because of the presence of gross incurable conditions (eg, multiple hepatic metastases, peritoneal carcinomatosis, or extra-abdominal metastasis) were used as controls. RESULTS: Intraperitoneal free cancer cells were found in 26 patients (19.4 percent). Compared with the remaining 108 patients without free cancer cells, there were no significant differences in the clinicopathologic characteristics and pathologic stages. No patients died after palliative resection, but five patients died after RG. The patients with free intraperitoneal free cancer cells had a poorer long-term prognosis after RG than those without free cancer cells (p < 0.0001). The prognosis for such patients was similar to the prognosis of those who underwent palliative resection. CONCLUSIONS: A peritoneal washing cytologic examination is mandatory before resection for potentially curable serosa-involved gastric carcinoma. When free cancer cells appear in the washing fluid, the cancer is incurable. Simply gastrectomy without additional lymphadenectomy is the optimal strategy for treatment.