Hee Chul Yu

Treatment Guidelines for Branch Duct Type Intraductal Papillary Mucinous Neoplasms of the Pancreas: When Can We Operate or Observe?

BackgroundThe objectives of this study were to investigate the clinicopathological features of branch intraductal papillary mucinous neoplasm (IPMN) and to determine safe criteria for its observation. Most clinicians agree that surgical resection is required to treat main duct-type IPMN because of its high malignancy rate. However, no definite treatment guideline (with respect to surgery or observation) has been issued on the management of branch duct type IPMN.MethodsWe retrospectively reviewed the clinicopathological data of 138 patients who underwent operations for IPMN between 1993 and 2006 at five institutes in Korea.ResultsOf 138 patients (mean age, 60.6 years; 87 men, 51 women), 76 underwent pancreatoduodenectomy, 39 distal pancreatectomy, 4 total pancreatectomy, and 20 limited pancreatic resection. There were 112 benign cases: 47 adenoma, 63 borderline cases, and 26 malignant cases, with 9 of these being noninvasive and 17 invasive. By univariate analysis, tumor size and the presence of a mural nodule were identified as meaningful predictors of malignancy. By receiver operating characteristic curve analysis, a tumor size of >2 cm was found to be the most valuable predictor of malignancy. When cases were classified according to tumor size and the presence of a mural nodule, the malignancy rate for a tumor ≤2 cm without a mural nodule was 9.2%, for a tumor of ≤2 cm plus a mural nodule was 25%, and for other conditions such as tumor >2 cm, >25%.ConclusionsMany branch duct IPMNs are malignant. Surgical treatment is recommended, except in cases that are strongly suspected to be benign or cases that present a high operative risk. Observation is only recommended in patients with a tumor size of ≤2 cm without a mural nodule.

Overexpression of VEGF and Angiopoietin 2: A Key to High Vascularity of Hepatocellular Carcinoma?

Hepatocellular carcinoma (HCC) is becoming one of the most common malignant tumors worldwide and is characterized by a high vascularity. Angiogenesis, formation of new microvessels, is critical for the growth and progression of various human solid tumors. Vascular endothelial growth factor (VEGF) and angiopoietins (Ang1 and Ang2) are endothelial cell–specific vasculogenic and angiogenic growth factors, but their expression and roles in HCC have not been extensively explored. The aim of this study was to determine the expression and cellular localization of VEGF, Ang1, and Ang2 in specimens of resected human HCC using in situ hybridization and immunohistochemical staining and to examine their relationship to microvessel density (MVD) and tumor size. We also investigated whether mutation of p53 protein might affect the expression of the above angiogenic growth factors. VEGF and Ang2 were strongly expressed and localized predominantly to cancer cells, whereas Ang1 was detected in supportive cells of large blood vessels, stromal cells, endothelial cells, and tumor cells. Expression of the VEGF protein and the Ang2 (but not Ang1) mRNA were strongly correlated with MVD (P < .05, P = .001) and tumor size (P < .05). There was also a strong correlation between VEGF protein and Ang2 mRNA expression (P < .001). However, no significant correlation was found between overexpression of p53 and the expression of VEGF, angiopoietins, or MVD. These findings suggest that overproduction of the angiogenic growth factors VEGF and Ang2 by HCC cells may increase vascularity and tumor growth in a paracrine manner. Our findings also suggest that interaction between VEGF and Ang2 may play a critical role in tumor angiogenesis in HCC.

Multicenter Analysis of Clinicopathologic Features of Intraductal Papillary Mucinous Tumor of the Pancreas: Is It Possible to Predict the Malignancy Before Surgery?

BackgroundDespite recently increasing numbers of reports on intraductal papillary mucinous tumors (IPMTs), difficulties still remain in terms of diagnosis, treatment, and prognosis. The purpose of this multicenter study was to evaluate the clinicopathologic features of IPMT in Korea and to suggest predictive criteria for malignancy in IPMT.MethodsWe retrospectively reviewed the clinicopathologic data of 208 patients who underwent operations for IPMT between 1993 and 2002 at 28 institutes in Korea.ResultsOf the 208 patients (mean age, 61 years), 147 were men and 61 were women. A total of 124 patients underwent pancreatoduodenectomy, 42 underwent distal pancreatectomy, 17 underwent total pancreatectomy, and 25 underwent limited pancreatic resection. There were 128 benign cases (adenoma, n = 62; borderline, n = 66) and 80 malignant cases (noninvasive, n = 29; invasive, n = 51). A significant difference in 5-year survival was observed between the benign and malignant groups (92.6% vs. 65.3%; P = .006). Of the six factors (age, location, duct dilatation, mural nodule, main duct type, and tumor size) that showed statistical differences by univariate analysis between the benign and malignant groups, three were significant by multivariate analysis—namely, mural nodule (P = .009), tumor size (P = .023), and a dilated duct size (P = .010).ConclusionsA significant proportion of IPMTs are malignant, although the overall prognosis of IPMT is superior to that of ordinary pancreatic cancer. Radical surgery is recommended for IPMT with the predictors of malignancy: mural nodule, tumor size (⩾30 mm), and dilated duct size (⩾12 mm).

A Prospective Randomized Controlled Study Comparing Outcomes of Standard Resection and Extended Resection, Including Dissection of the Nerve Plexus and Various Lymph Nodes, in Patients With Pancreatic Head Cancer

Objective:To prospectively evaluate the survival benefit of dissection of the nerve plexus and lymphadenectomy in patients with pancreatic head cancer. Background:Despite randomized controlled trials on the extent of surgery in pancreatic cancer, attempts have been made to perform more extended resections. Methods:A total of 244 patients were enrolled; of these, 200 were randomized to undergo standard resection or extended resection, with the latter including the dissection of additional lymph nodes and the right half of the nerve plexus around the superior mesenteric artery and celiac axis. We evaluated 167 patients from 7 centers who fulfilled all of the required criteria. Result:Operation time was longer and estimated blood loss was higher in the extended resection group than in the standard resection group, but the R0 resection rate was comparable. The mean number of lymph nodes retrieved per patient was higher in the extended resection group than in the standard resection group (33.7 vs 17.3; P < 0.001). The morbidity rate was slightly higher in the extended resection group than in the standard resection group. Two patients in the extended resection group died in hospital. Median survival after R0 resection was similar in the extended resection and standard resection groups (18.0 vs 19.0 months; P = 0.239) regardless of lymph node metastasis. Adjuvant chemoradiation had a positive impact on overall survival. Conclusions:This study suggests that extended lymphadenectomy with dissection of the nerve plexus does not provide a significant survival benefit compared with standard resection in pancreatic head cancer. Standard resection can be performed safely and efficiently, without negatively affecting oncologic efficacy or long-term survival, when compared with extended pancreaticoduodenal resection. (NCT00679913)?

Detection of hepatocellular carcinoma: gadoxetic acid-enhanced 3-dimensional magnetic resonance imaging versus multi-detector row computed tomography.

Purpose: The aim of this study was to compare the diagnostic accuracy and sensitivity of gadoxetic acid-enhanced magnetic resonance imaging (MRI) with multi-detector row computed tomography (MDCT) for the detection of hepatocellular carcinomas (HCCs). Materials: Sixty-two patients (81 HCCs) who underwent MDCT and gadoxetic acid-enhanced MRI using a 3-dimensional volumetric interpolated technique with a mean interval of 7 days (range, 3-11 days) were included in this study. Two observers reached a consensus on 2 sets of images: the gadoxetic acid set (unenhanced, early dynamic, 10-minute, and 20-minute hepatocyte phase images) and the 3-phase MDCT. Diagnostic accuracy and sensitivity were evaluated using the alternative-free response receiver operating characteristic method. Results: There was a trend toward increased area under the receiver operating characteristic curve (Az value) for the gadoxetic acid set (0.963) as compared with the MDCT (0.930), but no significant difference was found (P = 0.41). Sensitivity of the gadoxetic acid set (91.4%) was better than that of the MDCT (71.6%; P = 0.0001). There were 12 lesions that showed only arterial hypervascularization on MDCT but showed arterial hypervascularization and delayed hypointensity on the gadoxetic acid set. Conclusions: Gadoxetic acid-enhanced MRI, including hepatocyte phase imaging, is more sensitive than MDCT for the detection of HCCs.

Estimation of standard liver volume for liver transplantation in the Korean population

The standard liver volume (LV) of a recipient is estimated in liver transplantation to determine the minimum LV necessary for the recipient. Simple linear formulas of LV estimation were developed for the Japanese and Caucasian populations. The present study examined the applicability of the reported formulas to the Korean population. Liver density (LD) was determined by analyzing 24 healthy livers. Data of liver weight (LW), body weight (BW), body height (BH), body surface area (BSA), and age were obtained from 652 postmortem examination reports (age, 42.4 ± 16.5 years) showing normal livers. The LV of each subject was estimated by LW / LD and the relationships between LV, BW, BSA, and age were analyzed. LD was 1.04 ± 0.07 kg/L. LV / BW decreased as age increased in the children but leveled off in the adults; the rate of increase in LV along with BSA in individuals with BSA <1.2 m2 appeared less than the corresponding rate in individuals with BSA ≥1.2 m2. The Japanese formula produced underestimates for the Korean population (226.9 ± 289.4 mL), while the Caucasian formula produced random errors (−30.64 ± 281.5 mL). A better LV estimation formula was established: LV (mL) = 21.585 × BW (kg)0.732 × BH (cm)0.225 (adjusted R2 = 0.59; SE = 275.8 mL). In conclusion, this study indicates that a nonlinear or piecewise linear model is more desirable than a simple linear model for LV estimation in children and adults, because LV / BW and LV / BSA are not constant with age and BSA. (Liver Transpl 2004;10:779–783.)

Comparison of gadoxetic acid-enhanced MRI and superparamagnetic iron oxide-enhanced MRI for the detection of hepatocellular carcinoma

AIM To compare the diagnostic accuracy and sensitivity of gadoxetic acid-enhanced MRI and superparamagnetic iron oxide (SPIO)-enhanced magnetic resonance imaging (MRI) for the detection of hepatocellular carcinoma (HCC). MATERIALS AND METHODS Eighty-nine patients (118 HCCs) who underwent three-dimensional gadoxetic acid-enhanced MRI and SPIO-enhanced MRI with a mean interval of 4.7 days (range 3-7 days), were included in this study. Two observers reviewed the gadoxetic acid set (unenhanced, early dynamic, 10 and 20 min hepatocyte-phase images) and SPIO set [unenhanced and ferucarbotran-enhanced T1-, T2-turbo spin-echo (TSE), and T2* weighted imaging (WI)] in consensus. Diagnostic accuracy and sensitivity were evaluated using the alternative-free response receiver operator characteristic (ROC) method. RESULTS The area under ROC curve (Az value) and sensitivity of the gadoxetic acid set (Az 0.964; sensitivity 90.7%) were significantly higher than those of the SPIO set (Az 0.830; sensitivity 84.7%; p<0.05). There were 14 and seven lesions that were verified only on the gadoxetic acid set and only on the SPIO set, respectively. Four HCCs were clearly revealed as hypointense only on gadoxetic acid-enhanced hepatocyte phase imaging, but were occult on other sequences, including the SPIO set. CONCLUSION Gadoxetic acid-enhanced MRI is better than SPIO-enhanced MRI for the detection of HCCs.

A hybrid semi-automatic method for liver segmentation based on level-set methods using multiple seed points

The present study developed a hybrid semi-automatic method to extract the liver from abdominal computerized tomography (CT) images. The proposed hybrid method consists of a customized fast-marching level-set method for detection of an optimal initial liver region from multiple seed points selected by the user and a threshold-based level-set method for extraction of the actual liver region based on the initial liver region. The performance of the hybrid method was compared with those of the 2D region growing method implemented in OsiriX using abdominal CT datasets of 15 patients. The hybrid method showed a significantly higher accuracy in liver extraction (similarity index, SI=97.6 ± 0.5%; false positive error, FPE = 2.2 ± 0.7%; false negative error, FNE=2.5 ± 0.8%; average symmetric surface distance, ASD=1.4 ± 0.5mm) than the 2D (SI=94.0 ± 1.9%; FPE = 5.3 ± 1.1%; FNE=6.5 ± 3.7%; ASD=6.7 ± 3.8mm) region growing method. The total liver extraction time per CT dataset of the hybrid method (77 ± 10 s) is significantly less than the 2D region growing method (575 ± 136 s). The interaction time per CT dataset between the user and a computer of the hybrid method (28 ± 4 s) is significantly shorter than the 2D region growing method (484 ± 126 s). The proposed hybrid method was found preferred for liver segmentation in preoperative virtual liver surgery planning.

Detection of small hepatocellular carcinoma: intraindividual comparison of gadoxetic acid-enhanced MRI at 3.0 and 1.5 T.

Purpose:To investigate the diagnostic efficacy of gadoxetic acid-enhanced magnetic resonance imaging (MRI) at 3.0 T for the detection of hepatocellular carcinomas (HCC) and compare with that at 1.5 T. Materials:Forty patients with 54 HCCs (size range: 0.6–2.0 cm) underwent gadoxetic acid-enhanced MRIs at both 1.5 and 3.0 T with 3 to 8 days interval. The MRIs were compared quantitatively by measuring tumor-liver contrast-to-noise ratio, and qualitatively by evaluating tumor-liver contrast using matched pairs analysis. Diagnostic accuracy and sensitivity were also evaluated by the consensus readings of 2 reviewers using the alternative-free response receiver operator characteristic (ROC) method. Results:Although the tumor-liver contrast-to-noise ratio for the arterial phase was significantly higher at 3.0 T than at 1.5 T (30.2 ± 21.4 vs. 35.2 ± 22.9; P = 0.04), we found similar values for the hepatocyte phase (38.2 ± 24.6 vs. 38.4 ± 25.3; P = 0.762). Matched pairs analysis indicated that the relative tumor-liver contrast was better in 7 and 9 lesions in the arterial phase and hepatocyte phase at 3.0 T, respectively, than those at 1.5 T. The diagnostic accuracy and sensitivity of 3.0-T imaging [Az, 0.988; 92.6% (n = 50)] were slightly higher than those of 1.5-T imaging [Az, 0.981; 88.9% (n = 48)], but the difference was not statistically significant (P = 0.487). Conclusions:Gadoxetic acid-enhanced MRIs at 1.5 and 3.0 T showed similar diagnostic performances for detecting small HCCs. However, there was a tendency toward increased reader confidence for the arterial phase and hepatocyte phase with 3.0 T compared with 1.5 T.

Anatomical Variations of the Origin of the Segment 4 Hepatic Artery and Their Clinical Implications

The threat of ischemic complications following massive resection, especially in living donor hepatectomy or split liver transplantation, has been haunting surgeons for many years. Postmortem dissections of 62 livers were performed to investigate anatomical variations of the principal artery for segment 4 (A4). The origin of A4 was examined separately in the liver with (n = 46) or without (n = 16) an aberrant left hepatic artery (abLHA). A4s were found to be extrahepatic or intrahepatic branches of the right hepatic artery (RHA), left hepatic artery, or proper hepatic artery and were categorized into 4 different types according to their origins. The RHA type, originating from the RHA or right anterior hepatic artery (RAHA), was the most common pattern in our series. The A4 roots had a strong tendency of stemming from the RHA (n = 12) even in the livers with abLHA (n = 16). Among the RHA‐type A4s, the A4 arising from RAHA (n = 2) is supposed to be the most dangerous variant because it can cause an ischemic change in the remaining part of the liver after right hepatectomy. In conclusion, in the era of living donor liver transplantation, paying particular attention to the point of origin of A4 is a prerequisite, especially when the lateral section is relatively small. Arterial injuries to A4 during split liver transplantation may also increase the risk of hepatic artery thrombosis and ischemic cholangiopathy. Liver Transpl 14:1180–1184, 2008.