Comparison of results between pylorus-preserving pancreaticoduodenectomy and subtotal stomach-preserving pancreaticoduodenectomy: report at a single cancer institute.

BACKGROUND/AIMS: Pylorus-preserving pancreaticoduodenectomy (PPPD) has the advantage of achieving good nutritional status postoperatively, but delayed gastric empty (DGE) is a frequent complication leading to a longer fasting period. Subtotal stomach-preserving pancreaticoduodenectomy (SSPPD) is an alternative option to preserve nutritional status and shorten the fasting period. We retrospectively compared clinical results between PPPD and SSPPD. METHODOLOGY: PPPD was performed in 28 patients and SSPPD in 27, between 2000 and 2009. RESULTS: Pancreatic carcinoma was more frequent in the SSPPD group (p = 0.041). Operating time was longer in the SSPPD group (610 min) than in the PPPD group (540 min; p = 0.031). Blood loss was greater in the SSPPD group (1810 mL) than in the PPPD group (1306 mL; p = 0.048). Period of NG intubation and fasting period were shorter in the SSPPD group (6 days and 9 days, respectively) compared to the PPPD group (15 days and 19 days, respectively; p <0.01 each). Severe DGE was 7% in the SSPPD group and 46% in the PPPD group (p <0.01). Postoperative complications and nutritional status in the early period did not differ between groups, although incidence of fatty liver was higher in the SSPPD group (78%) than in the PPPD group (25%; p <0.01). CONCLUSIONS: SSPPD is a useful alternative for pancreaticoduodenectomy. Further prospective studies with longer follow-up are warranted to clarify the superiority and problems associated with this procedure.

Does fibrin glue prevent biliary and pancreatic fistula after surgical resection?

BACKGROUND/AIMS: Efficacy of fibrin glue to prevent biliary or pancreas fistula at the resected edge of the liver or pancreas is controversial. We examined surgical results of fibrin glue use in patients who underwent hepatectomy or pancreatectomy to assess the efficacy of its use. METHODOLOGY: Subjects were divided into two groups; the fibrin glue group in hepatectomy (n=228) and in pancreatectomy (n=113), and the non-fibrin glue group in hepatectomy (n=94) and in pancreatectomy (n=24). In case of hepatectomy, the fibrin glue was sprayed on the cut-surface or anastomotic site of hepatico-jejunostomy. In case of pancreatectomy, the fibrin glue was sprayed on the anastomotic site of pancreato-jejunostomy or closed pancreatic stump. RESULTS: In the hepatectomy group, uncontrolled ascites were more frequent in the fibrin glue group (p<0.05). The use of fibrin glue for both groups has been less frequent in recent years. Prevalence of biliary fistula was not significantly different between groups. Hospital stay in the fibrin glue group was significantly longer than that in the non-fibrin glue group, and was not significantly different between hepatectomy or pancreatectomy groups. There was no significant difference of any complications including pancreatic fistula between groups. Prevalence of pancreatic fistula was not significantly different between the fibrin glue group and the non-fibrin glue group. CONCLUSIONS: Use of fibrin glue did not prevent biliary or pancreatic fistula in patients who underwent hepatectomy and pancreatectomy with or without enteric anastomosis.

Extended right hepatectomy for hilar bile duct carcinoma using the modified liver hanging maneuver.

To achieve complete extended right hepatectomy or trisectionectomy for a bismuth type IV hilar bile duct carcinoma, we propose the application of Belghiti's liver hanging maneuver (LHM) using a small nasogastric tube. This small nasogastric tube was placed in the cut plane: the top of the tube was placed between the hepatic veins. The tube was placed along the border between the left lateral sector and Spiegel's caudate lobe and the bottom of the tube was placed at the left side of the umbilical Glissonian pedicle. Hepatic parenchyma was transected using a vascular sealing device. Hepatic transection was always targeted to the tube and, eventually, a cut line of left hepatic ducts remained. We report the case of a 76-year-old female and an 83-year-old female with widely extended hilar bile duct carcinomas showing Bismuth type IV. Applying the modified LHM for extended right hepatectomy, the cut planes were easily and adequately obtained in patients with hilar bile duct carcinoma.

Three-dimensional fusion images of hepatic vasculature and bile duct used for preoperative simulation before hepatic surgery.

BACKGROUND/AIMS: Recent developments in radiological technology allowed acquisition of images with high spatial resolution that facilitate effective 3-dimensional (3D) reconstruction of fusion images. Present study utilized 3D cholangiography and angiography with multi-detector-row computed tomography (MDCT) to acquire information regarding operative simulations. METHODOLOGY: 3D-fusion images were evaluated in 39 patients with hepatobiliary malignancies who underwent surgical resections. RESULTS: An aberrant branch of segment 3 over the umbilical portal vein, a large hepatoma compressed the hilar vessels, an aberrant branch of the caudate lobe vasculature in case of metastatic liver tumor with a right-sided umbilical portal vein and transected biliary leakage were clearly observed by 3D imaging system. Four patients with intrahepatic cholangiocarcinoma underwent multiple biliary stent placements and adequate placement of biliary stents was possible. In 22 patients with extrahepatic biliary carcinomas, visualization of the extent of tumor invasion by 3D-fusion images was equivalent to conventional cholangiography. In 2 patients, adequate placement of multiple stents could be visualized with this system. In 2 patients who underwent hepatectomy, more extended cancer invasion was observed than was visualized by 3D-fusion images. CONCLUSIONS: 3D fusion images were very useful for preoperative simulations in order to understand relationships between tumors and adjacent vasculatures.

Usefulness of omental wrapping to prevent biliary leakage and delayed gastric emptying in left hepatectomy.

BACKGROUND/AIMS: To identify the clinical significance of the omental wrapping (OW) technique after left hepatectomy to reduce bile leakage and delayed gastric emptying. We examined clinical and surgical parameters after left hepatectomy with or without biliary reconstruction in 79 patients. METHODOLOGY: This was a retrospective study of data from 14 patients undergoing OW compared to 65 patients in the control group. RESULTS: Bile leakage and delayed emptying after hepatectomy were observed in 15 and 11 patients, respectively. Gender, background liver function, liver diseases and preoperative liver function tests were not significantly different between both groups. Prevalence of extent of hepatectomy, existence of segment 1 resection, biliary-enteric anastomosis, operating time and blood loss were also not significantly different between groups. Prevalence of bile leakage was similar between the OW and the control group (14 vs. 20%) (p=0.91). Prevalence of delayed gastric emptying was not significantly different between groups, but this complication was not observed in the OW group in comparison with the control group (0% vs. 20%) (p=0.31). Prevalence of other complications and hospital stay after hepatectomy were similar between groups. CONCLUSIONS: Significant differences for preventing left hepatectomy related complications were not found; however, it is possible that OW could reduce delayed gastric emptying.

Experience of surgical resection for hilar cholangiocarcinomas at a Japanese single cancer institute.

BACKGROUND/AIMS: Surgical resection is a radical treatment option for hilar bile duct carcinoma (HBDC); however, it is still difficult to cure and postoperative morbidity is high at this stage. METHODOLOGY: We examined the demographics, surgical records and outcome in 38 patients with hilar cholangiocarcinoma undergoing operation. RESULTS: Five patients (13%) underwent probe laparotomy because of peritoneal dissemination or liver metastasis. Of 33 patients, extended hemi-hepatectomy was performed in 32 patients. Postoperative complications were observed in 46% including hepatic failure in 3 and hospital death was observed in 4 patients. Advanced tumor stage more than stage III was observed in 23 patients. Curability of operation was A in 5 patients, B in 17 and C in 11 and postoperative adjuvant chemotherapy was administered in 24% including photodynamic therapy in 3. Tumor recurrence was observed in 41% of HBDC patients. The 3- and 5-year tumor-free survival was 38% and 10%, respectively and 3- and 5-year overall survival was 48% and 32%, respectively. By comparison with tumor stage or final curability, survival rates were not significantly different between groups. CONCLUSIONS: Surgical resection is still the only curative treatment option to improve patient survival even in advanced stage HBDC.

Evaluation of surgical resection for pancreatic carcinoma at a Japanese single cancer institute.

BACKGROUND/AIMS: Surgical resection is a radical treatment option for pancreatic carcinoma (PC); however, it is still difficult to cure and patient prognosis is poor at this stage. METHODOLOGY: We examined the demographics, surgical records and outcome in 64 patients with hilar PC undergoing surgical resection. RESULTS: Pancreatoduodenectomy (PD) was carried out in 48 patients, distal pancreatectomy (DP) in 14 and total pancreatectomy in two. Postoperative complications were observed in 18 patients (28%) but no hospital deaths. All stage I patients showed carcinoma in situ of intraductal papillary mucinous carcinoma (IPMC). Postoperative adjuvant chemotherapy was performed in 15 patients (23%) using gemcitabine or S-1. Cancer recurrence was observed in 36 patients (56%) and 31 died of carcinoma. The 5-year cancer-free and overall survival rate was 12% and 14%, respectively. CA19-9 level, morphological type, T category, lymph node metastasis, extrapancreatic nerve plexus invasion, retropancreatic tissue invasion, distal bile duct invasion, duodenal invasion and arterial system invasion were significant poor prognostic factors; however, portal vein system invasion was not significantly associated with prognosis. Cancer infiltration at bile duct cut-end and dissected peripancreatic tissue margin and presence of residual tumor showed a poor prognosis. Surgical prognosis in only non-invasive IPMC was satisfactory. CONCLUSIONS: Radically extended surgical resection is necessary and newly effective adjuvant chemotherapy is a promising modality to improve patient survival in PC patients.

Evaluation of surgical resection for gallbladder carcinoma at a Japanese cancer institute.

BACKGROUND/AIMS: Surgical resection is a radical treatment option for gallbladder carcinoma (GBC); however, it is still difficult to cure and patient prognosis is poor. An assessment of the surgical results and chemotherapy options may elucidate effective treatments. METHODOLOGY: We retrospectively examined the demographics, surgical records and outcome in 33 patients with GBC undergoing surgical resection. RESULTS: Postoperative cancer recurrence was observed in 36% of patients. Mean cancer-free survival time was 84 months and 3-year cancer-free survival rate was 70% Mean overall survival time was 96 months and 5-year overall survival rate was 52%. The 3-year cancer-free survival and the 5-year overall survival were significantly different between the final tumor stages (p<0.001). Higher CEA and CA19- 9 level were significantly related to poor overall survival (p<0.05). Macroscopically, papillary type tumor showed significantly better overall survival compared to nodular or flat types (p<0.05). Degree of invasion, node metastasis, moderate or poor differentiation, vascular or perineural invasion and invasion of the liver or hepatoduodenal ligament were significantly associated with poor overall survival (p<0.05). A cancerfree margin at the hepatic cut end and dissected periductal structures showed a significantly poor prognosis (p<0.05). The overall survival in final curability A was significantly associated with better curability than B or C (p<0.05). CONCLUSIONS: Radically extended surgical resection for GBC is necessary to obtain improved patient survival and new adjuvant chemotherapy would be expected to improve results after surgery.

Prediction of indocyanine green retention rate at 15 minutes by correlated liver function parameters before hepatectomy.

BACKGROUND: Indocyanine green retention rate at 15 min (ICGR15) is a useful marker of liver function in deciding on the extent of hepatectomy. To determine ICGR15 regardless of liver condition, we sought to establish a formula for converted ICGR15 based on conventional blood tests and technetium-99 m galactosyl human serum albumin ((99m)Tc-GSA) scintigraphy. MATERIALS AND METHODS: We measured liver function parameters, including ICGR15, in 307 patients, including 265 liver cancer patients without biliary obstruction (no obstruction group) and 42 with biliary obstruction (obstruction group). RESULTS: In the no obstruction group, multiple regression analysis identified blood pool clearance ratio (HH15), liver uptake ratio (LHL15) calculated by heart and liver activity between 3 and 15 min after injection of (99m)Tc-GSA, and serum hyaluronic acid as significant correlates (P < 0.05). The calculated converted ICGR15 was then equal to 0.02∗HA + 0.276∗(HH15∗100)-0.501∗(LHL15∗100) + 41.41. The mean difference between actual and converted ICGR15 was significantly lower in the obstruction than in the no obstruction group (P = 0.031). A significantly larger proportion of patients of the obstruction group had lower converted ICGR15 than those of the no obstruction group (P = 0.045). CONCLUSION: The converted ICGR15 is useful for evaluating hepatic function in patients with biliary obstruction who plan to undergo major hepatectomy.

Usefulness of sonazoid-ultrasonography during hepatectomy in patients with liver tumors: A preliminary study.

BACKGROUND AND OBJECTIVES: To improve diagnostic accuracy of intraoperative ultrasonography (IOUS), we investigated the usefulness of new contrast medium of microbubble agent, Sonazoid as a preliminary study. METHOD: We examined IOUS in 50 patients with liver tumors who underwent hepatectomy. Sonazoid was administrated intravenously and Kupffer-phase images of the tumor were observed before hepatectomy. Sonazoid was reinjected to observe the tumor vasculature. RESULTS: The tumors included hepatocellular carcinoma (HCC) in 25 patients, intrahepatic cholangiocarcinoma in 3, colorectal liver metastasis in 14, gastrointestinal stromal tumor in 1, and benign hematoma in 1. Liver tumors were clearly detected as perfusion defect in most cases. Small lesions (<1 cm), extra-capsular tumor growth, and portal vein tumor thrombus were also clearly detected on the Sonazoid-IOUS. Small occult tumors were detected in five cases. Differential diagnosis with suspicious non-tumorous lesions and benign mass was possible based on vascular findings at the early phase. In comparison with hepatectomy for HCC under conventional IOUS, the proportion of patients with positive surgical margin (0%) tended to be lower than that of the control group (P = 0.073). CONCLUSION: Sonazoid-IOUS is a promising useful tool to detect the precise tumor margin and small tumors, hence allowing curative hepatectomy or intraoperative ablation.

Left hepatectomy accompanied by a resection of the whole caudate lobe using the dorsally fixed liver-hanging maneuver.

A resection of the caudate lobe often needs to be combined with a hemi-hepatectomy for hilar cholangiocarcinoma or a liver tumor in segment 1. To achieve complete resection of the whole caudate lobe, the cut line between the right edge of the paracaval portion and the right lateral sector should be precisely controlled. The liver-hanging maneuver (LHM) is a useful anterior approach that does not require mobilization of the remnant liver. However, the precise set-up of the cut line of the right edge has not been optimized in previous reports. We herein introduce a new modification of LHM that we named the "dorsally fixed liver-hanging maneuver" (DF-LHM) based on the results in five patients who underwent left hepatectomy combined with a total resection of segment 1. This technique provided adequate cut planes along the right edge of the caudate lobe, shortening the transection time and reducing intraoperative blood loss. The DF-LHM may represent a new key technique for this type of hepatectomy, and further applications for other anatomical resections can be modeled on the strategy.

Intraductal papillary growth of liver metastasis originating from colon carcinoma in the bile duct: report of a case.

Morphologically, liver metastases from colorectal carcinoma usually form as nodular tumor masses, whereas intraductal papillary growth in the bile duct is rare. A 65-year-old man underwent right hemicolectomy for advanced colon carcinoma, and histology of the primary carcinoma confirmed moderately differentiated adenocarcinoma with subserosal invasion, no vascular infiltration, and no lymph node metastasis. A liver tumor was found in the right paramedian Glisson pedicle and intraductal growth of cholangiocarcinoma was seen on imaging. We performed right hepatectomy and macroscopically, the resected specimen contained a growth in the bile duct lumen similar to cholangiocarcinoma. Histological examination revealed intraductal papillary proliferation of well-differentiated adenocarcinoma without vascular infiltration or lymph node metastasis in the hepatic hilum. Immunohistochemical staining revealed that the tumor cells were negative for cytokeratin-7 and positive for cytokeratin-20. Based on these findings, liver metastasis from colon carcinoma was diagnosed. Liver metastasis from colorectal carcinoma rarely arises as intraductal papillary growth in the bile duct, but the possibility of liver metastases with unusual morphology must be borne in mind for patients with a history of carcinoma in the digestive tract.

Relationship of hepatic functional parameters with changes of functional liver volume using technetium-99m galactosyl serum albumin scintigraphy in patients undergoing preoperative portal vein embolization: a follow-up report.

BACKGROUND: To identify predictors of changes in functional hepatic volumes after portal vein embolization (PVE) before hepatectomy, we examined the relationship between hepatic functional parameters and changes in functional volume of the embolized and non-embolized liver based on a previous volumetric analysis. MATERIAL AND METHODS: Subjects were 24 patients who underwent PVE, which was performed through the trans-ileocolic vein (n = 4) or by percutaneous transhepatic puncture (n = 20). The RI liver volume parameter was measured by liver scintigraphy with technetium-(99m) galactosyl human serum albumin ((99m)Tc-GSA). Computed tomography (CT) volume parameter was also measured. RESULTS: Significant atrophy of the embolized liver and hypertrophy of the non-embolized liver (change of 72 ± 108 cm(3) and 111 ± 91 cm(3), respectively) (change of 7.8%) was observed after PVE. The change in these RI volume parameters (change of 173 ± 175 cm(3) and 145 ± 137 cm(3) , respectively) (16.5%) was significantly greater than CT volume parameters (P < 0.01). CT vol and RI vol in the embolized and non-embolized liver were well correlated (r = 0.75 and 0.69, respectively). However, the correlation between CT and RI volume parameters in the embolized and non-embolized liver after PVE was very weak (r = 0.17 and 0.03, respectively). Only alkaline phosphatase level correlated negatively with atrophic CT volume parameter of the embolized liver (r = -0.455, P < 0.05). When compared with CT volume parameter, more parameters were significantly correlated with changes of RI volume parameter in the embolized liver: pre-PVE pressure; ICGR15; and serum levels of hyaluronate, total bilirubin, albumin, and alkaline phosphatase. Only platelet count was significantly correlated with hypertrophy of the non-embolized liver. CONCLUSION: RI volume parameter might more accurately reflect functional changes in the embolized liver and non-embolized liver than CT volume parameter. Correlated parameters might allow us to predict the functional effect of PVE.

Usefulness of the combination procedure of crash clamping and vessel sealing for hepatic resection.

BACKGROUND: Minimization of blood loss during resection of the hepatic parenchyma in hepatectomy remains a major problem. The usefulness of the LigaSure sealing system has been reported. OBJECTIVES: To evaluate the efficacy of combination procedure of LigaSure and forceps clamping for control blood loss and transection time in hepatectomy. METHODS: Here, we report our experience with the combination technique of LigaSure Precise, a clamp forceps type, and crush clamping method for hepatic transection in 33 patients who underwent hepatectomy. RESULTS: The combination technique allows fast and bloodless transection even along the major intrahepatic vessels. Blood loss and transection time were significantly reduced in the group of LigaSure use (P < 0.05). Efficient hemostasis could be achieved also in patients with extensive liver injury such as cirrhosis. The rates of postoperative intraabdominal abscess formation in the combination technique of LigaSure and crush clamping were lower compared with the conventional crush clamping method (P < 0.05). CONCLUSIONS: The combined use of LigaSure Precise and crush clamping technique is safe and allows rapid completion of hepatic resection.

Strategy of treatment for hepatocellular carcinomas with vascular infiltration in patients undergoing hepatectomy.

BACKGROUND AND OBJECTIVES: Vascular infiltration (VI) is an important prognostic factor for hepatocellular carcinoma (HCC) and predictive parameters are necessary to preoperatively decide treatment strategies in patients with HCC. METHODS: Relationships between presence and degree of VI in the portal and hepatic veins and bile duct, and post-hepatectomy survival were examined in 271 HCC patients who underwent hepatectomy. RESULTS: VI was observed in 81 patients (30%). Disease-free and overall survival rates was significantly lower in patients with VI than in patients without VI, and became poorer according to the degree of infiltration (P < 0.01). Multiple, increased size, non-meeting of Milan criteria, irregular macroscopic findings and increased PIVKA-II levels were associated with degree of VI in portal vein (P < 0.01). Increased size and increased PIVKA-II level were associated with degree of VI in hepatic vein (P < 0.05). Non-meeting of Milan criteria was associated with degree of infiltration in bile duct (P = 0.034). Survival was significantly better following anatomical resection than with non-anatomical resection and, furthermore, survival was better with surgical margins >5 mm than with shorter margins in patients who underwent non-anatomical resection. CONCLUSION: Adequate extent of operative procedures, but not limited resection with short margins, is useful when predictive parameters associated with VI are observed.

Selecting treatment for hepatocellular carcinoma based on the results of hepatic resection and local ablation therapy.

BACKGROUND: First-line treatment for

Preoperative diagnosis of lymph node metastasis in biliary and pancreatic carcinomas: evaluation of the combination of multi-detector CT and serum CA19-9 level.

BACKGROUND: It is difficult to diagnose lymph node metastasis in biliary and pancreas carcinomas before surgery. AIM: The aim of this study was to assess the utility of the combination of multi-detector computed tomographic (MDCT) findings and serum carbohydrate antigen (CA)19-9 level in the diagnosis of lymph node metastasis in biliary and pancreas carcinomas. METHODS: The subjects were 139 patients with biliary and pancreas carcinomas who underwent surgical resection. We calculated the positive predictive values (PPV), sensitivities, specificities, positive likelihood ratios (PLR) and accuracies of diagnosis by MDCT alone, serum CA19-9 level alone, and their combination. RESULTS: The PPV and sensitivity were higher for node metastasis in hepatoduodenal ligament than in common hepatic artery (CHA) or para-aortic region (PAR). Specificity, accuracy and PLR were highest for CHA in biliary carcinoma. With pancreatic carcinoma, PLR was slightly higher in PAR compared to other regions. The sensitivity of CA19-9 for node metastasis was higher than that of MDCT, while the PPV, specificity, accuracy and PLR were low for both biliary and pancreas carcinoma. The combination of positive CT findings and high CA19-9 level had the highest positive rate for node metastasis for both types of carcinomas. Nodes around the supra-mesenteric vein could not be fully observed on CT. CONCLUSION: The combination of high-resolution MDCT and CA19-9 is useful for the diagnosis of lymph node metastasis in biliary and pancreas carcinomas.

Comparison of outcome of hepatectomy with thoraco-abdominal or abdominal approach.

BACKGROUND/AIMS: Thoraco-abdominal approach is a suitable choice for hepatectomy to secure good view for mobilization. The aim of this study was to assess efficacy of thoraco-abdominal approach (TAA) for hepatectomy. METHODOLOGY: There were compared clinicopathological data, surgical results and postoperative complications of 425 consecutive patients who underwent hepatectomy via abdominal (AA) (n = 147) or TAA (n = 278). RESULTS: Blood loss and operating time were significantly higher in TAA than AA group (970 vs. 830 ml and 408 vs. 372 min.) (p < 0.05). Prevalence of pleural effusion was significantly higher in TAA than AA group (24 vs. 9%) (p < 0.01). However, proportions of patients who developed hepatic complications such as biloma (14 vs. 23%), and wound infection (8 vs.25%) were significantly less in TAA than AA group (p < 0.05). Hospital stay after hepatectomy and mortality were similar between both groups. Presence of chronic viral hepatitis, lower platelet count, higher level of serum hyaluronic acid, larger blood loss and TAA correlated significantly with thoracic complications (p < 0.05). Multivariate analysis showed that increased blood loss (p = 0.011), but not TAA, was a significant determinant of thoracic complications (p = 0.08). CONCLUSIONS: TAA can be considered a relatively safe approach for hepatectomy with minimal abdominal complications nevertheless of frequent pleural effusion.

Stepwise education for pancreaticoduodenectomy for young surgeons at a single Japanese institute.

BACKGROUND/AIMS: Expert technique and special anatomical or physiological knowledge are needed in the field of pancreatic surgery. The establishment of basic policies and operative techniques for pancreaticoduodenectomy (PD) and stepwise training for young pancreatic surgeons are necessary. METHODS: We scheduled PD for ampullar, biliary and pancreas carcinoma, and evaluated types of pancreatic anastomosis or results by each operator such as a chief, fellowship and resident doctors (> 5 years after graduation). RESULTS: Based on a questionnaire distributed to young residents (n = 30), only half of them have experienced PD or PPPD, which was related to operating volume at the hospital. Post-operative complications at the teaching hospital were observed in 50 of 88 patients (56%). Post-operative complications were not significantly correlated with the type of anastomosis; however, duct-to-mucosa anastomosis of the pancreas might decrease pancreatic fistula (0% vs. 26% in pancreaticogastrostomy and 13% in pancreaticojejunostomy without duct-to-mucosa anastomosis). Based on the stepwise education protocol of technique, patient demographics, the surgical records and the post-operative complications were not significantly different between experienced teaching surgeons, fellowship surgeons and senior residents, although the time of operation and anastomosis tended to be longer in resident surgeons (p = 0.22). CONCLUSION: Competent operative techniques for inexperienced surgeons and the achievement of safe resection at each stage are our educational goals for PD.

Relationship between period of survival and clinicopathological characteristics in patients with hepatocellular carcinoma who underwent hepatectomy.

BACKGROUND/AIMS: Cancer death in the early period after hepatectomy remains problematic in patients with hepatocellular carcinoma (HCC). We examined the relationship between clinicopathological parameters' and survival periods in 234 HCC patients who underwent hepatectomy. METHODOLOGY: Patients were divided into four groups: Group 1, survival > 5 years; Group 2, survival for 2-5 years; Group 3, cancer death at 2-5 years; and Group 4, cancer death in < 2 years. RESULTS: Numbers of patients in each subgroup were: Group 1, n = 87 (37%); Group 2, n = 44 (19%); Group 3, n = 46 (20%); and Group 4, n = 57 (24%). Child-Pugh B status, blood loss > 1500 ml, multiple tumors, tumor size > 5 cm, not meeting Milan criteria, irregular macroscopic findings, invasion of Glissonian pedicle, invasion of hepatic vein, higher modified Japan Integrated Staging score (3-5), long-term ascites after hepatectomy and postoperative tumor recurrence within 12 months were frequent in Group 4 ( p < 0.05). Multivariate analysis revealed AFP level > or = 1000 ng/ml (hazard ratio (HR), 2.6) and early tumor relapse (HR, 8.1) as independently related parameters ( p < 0.05). CONCLUSIONS: Careful follow-up for early tumor relapse may be important for improving postoperative outcomes in HCC patients with high preoperative AFP levels.