| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||
ABSTRACT |
Monday, May 7, 10:00 AM–12:00 PM
Abstracts 023-034
Moderator(s): Elmar Merkle and Aliya Qayyum
10:00 AM
023. Contrast-enhanced Ultrasound in Liver Transplant: First Results and Potential of Complications in the Postoperative Period
Clevert D. A.1,2*; Stickel M.3; Strautz T.1,2; Flach P.1,2; Becker C.1,2; Schoenberg S.1,2; Reiser M.1,2 1. Department of Clinical Radiology, Munich, Germany; 2. Department of Clinical Radiology, Passau, Germany; 3. Department of Surgery, Munich, Germany
Address correspondence to D. Clevert (Dirk.Clevert{at}med.uni-muenchen.de)
Objective: This study compared the efficacy of contrast-enhanced ultrasound (CE-US) using second generation contrast media versus CT or MRI in the assessment of vascular and biliary complications in postoperative follow-up of liver transplantation.
Materials and Methods: The study group consisted of 36 consecutive liver transplant recipients who underwent post-transplantation CE-US examination after developing ascites and/or unclear liver function tests. Real time CE-US was performed after a bolus injection of SonoVue (1.6 to 2.4 ml, Bracco, Italy) followed by 10-ml saline solution. Using contrast harmonic imaging (CHI) technique (Logiq 9, GE) and CE-US with CPS (Siemens, Sequoia, Acuson) with a 4-MHz transducer, a low mechanical index was chosen to avoid early destruction of the microbubbles (MI 0.1–0.2). In order to confirm the results, the patients underwent contrast-enhanced MR angiography (MRA) or CT angiography (CTA).
Results: Complications were identified in 16 of 36 patients (44.4 %). Five transplants (14 %) had hepatic artery thrombosis (n = 2) or significant stenosis (n = 3). Six transplants (16 %) developed portal vein stenosis (n = 4) or portal vein thrombosis (n = 2). MRA or CTA confirmed the findings of the CE-US in all 11 cases. Biliary stricture at the anastomotic site was detected in 5 patients. MR-cholangiography confirmed the findings of all strictures.
Conclusion: Due to advances in contrast-enhanced US, vascular and biliary complications in the postoperative period following liver transplantation can be reliably diagnosed noninvasively on the intensive care unit. CE-US shows vascular as well as biliary complications in the postoperative patient at high accuracy.
024. Solid Hypervascular Liver Lesions: Accurate Identification of True Benign Lesions on Delayed Hepatobiliary Phase MR Imaging After Gadobenate Dimeglumine
Morana G.2; Grazioli L.4; Schneider G.3*; Bondioni M. P.4; Guarise A.2; Kirchin M. A.1 1. Bracco Imaging SpA, Milano, Italy; 2. Ospedale Cà Foncello, Treviso, Italy; 3. Saarland University Hospital, Homburg, Germany; 4. University of Brescia, Brescia, Italy; 5. University of Verona, Verona, Italy
Address correspondence to G. Schneider (ragsne{at}uniklinik-saarland.de)
Objective: To determine the value of hepatobiliary phase MR imaging after gadobenate dimeglumine (MultiHance, Bracco) for accurate identification of true benign hypervascular liver lesions.
Materials and Methods: Retrospective assessment was performed of 468 patients with 785 lesions (267 focal nodular hyperplasia [FNH], 30 nodular regenerative hyperplasia [NRH], 124 hepatic adenoma/liver adenomatosis [HA/LA], 211 hepatocellular carcinomas [HCC], 11 fibrolamellar-HCC [FL-HCC], 16 peripheral cholangiocarcinomas [PCC], 17 capillary hemangiomas [CH] and 109 hyper- or hypovascular metastases) who underwent enhanced MR imaging with 0.05 mmol/kg gadobenate dimeglumine. Histological confirmation was available for all lesions except most FNH whose diagnosis was based on characteristic enhancement and two-year follow-up. Pre- and postcontrast T1-weighted images acquired during the dynamic and delayed hepatobiliary (1–3 h) phases were evaluated for lesion enhancement (hypo-, iso-, hyperintensity). Lesion enhancement was compared with histology findings and values for sensitivity, specificity, accuracy, PPV and NPV were determined.
Results: On hepatobiliary phase imaging 96.6% of FNH, 100% of NRH, 5.6% of HA/LA, 21.8% of HCC and 2.7% of metastases appeared hyper- or isointense, whereas 3.4% of FNH, 94.4% of HA/LA, 78.2% of HCC, 100% of FL-HCC, 100% of PCC, 100% of CH and 97.2% of metastases appeared hypointense. Assuming delayed phase lesion iso/hyperintensity as indicating true lesion benignity (FNH/NRH) values for sensitivity, specificity, accuracy, PPV and NPV of 97.0%, 88.5%, 91.7%, 83.7% and 98.0%, respectively, were obtained for identification of true benign lesions.
Conclusion: Hepatobiliary phase imaging after 0.05 mmol/kg gadobenate dimeglumine is highly accurate for distinguishing true benign focal liver lesions from other lesions. Lesions that are hypointense on hepatobiliary phase imaging should be considered for biopsy.
025. Automated Volumetric, RECIST, and WHO Measurements of Abdominal and Pelvic Lymph Nodes on MDCT: Preliminary Results
Rezvani M.1*; Yaghmai V.1; Soud H.2 1. Northwestern University-Feinberg School of Medicine, Chicago, IL; 2. Siemens Medical Solutions, Malvern, PA
Address correspondence to M. Rezvani (m-rezvani{at}northwestern.edu)
Objective: To assess the feasibility of automatically segmenting and measuring abdominal and pelvic lymph nodes volumetrically and by RECIST as well as WHO criteria.
Materials and Methods: Thirty lymph nodes on contrast-enhanced abdominal and pelvic CT scans of eleven patients with history of lymphoma were selected (para-aortic/para-caval n = 10; mesenteric n = 10; iliac chain n = 6; inguinal n = 4). For the purposes of this preliminary feasibility study, confluent lymph nodes and those indenting adjacent iso-dense anatomic structures were excluded. All scans had been acquired using multislice CT (4-, 16-, or 64-slice) with 5 mm slice thickness. Prototype software program (Siemens Medical Solutions, Forchheim, Germany) was utilized to measure the maximal diameter (RECIST criteria) and cross-product of diameters (WHO criteria) both manually and automatically. The manual measurements were performed independent of the automatic measurements by two board-certified radiologists. Pearson product moment correlation coefficient (r) was calculated. The volume of each lymph node was also automatically calculated by automatic segmentation. The quality of automated segmentation for each lymph node was evaluated in consensus by the radiologists on a five-point scale (1: poor–5: excellent).
Results: The prototype software successfully segmented all the lymph nodes. Segmentation scores were as follows: Score 5: 18 lymph nodes; Score 4: 10 lymph nodes; Score 3: 2 lymph nodes. The average quality of segmentation was 4.53 (SD = 0.63). The correlation coefficient between manual and automated measurements using RECIST and WHO criteria was 0.98 (p < 0.0001) and 0.99 (p < 0.0001), respectively.
Conclusion: Our preliminary data suggests that automated segmentation as well as volumetric, RECIST and WHO measurements of the abdominal and pelvic lymph nodes may be performed accurately. Its role in the treatment follow-up of oncology patients remains to be validated.
026. Functional Perfusion and Diffusion MR Imaging After Radiofrequency Ablation of Colorectal Liver Metastases: Multiparametric Assessment of Tumor Response
Assumpcao L. R.*; Gleisner A. L.; Pawlik T. M.; Choti M.; Bluemke D. A.; Kamel I. R. Johns Hopkins Hospital, Baltimore, MD
Address correspondence to L. Assumpcao (lassump1{at}jhmi.edu)
Objective: To evaluate the spectrum of changes on functional MR imaging after RFA of colorectal liver metastases.
Materials and Methods: The study included 13 patients (mean age, 59 years) who underwent surgical hepatic ablation of metastatic colorectal adenocarcinoma over a period of 18 months. MR imaging studies before and immediately (within 10 days) after treatment were evaluated. Imaging protocol included T2-weighted FSE images (matrix, 256 x 256; thickness, 8 mm; gap, 2 mm; TR, 5000; TE 100), BH diffusion-weighted echoplanar images (matrix, 128 x 128; thickness, 8 mm; gap, 2 mm; B value, 500; TR, 5000-6500; TE, 110), and BH unenhanced and contrast-enhanced T1-weighted 3D fat-suppressed GRE (matrix, 192 x 160; thickness, 4–6 mm; TE 1.2; flip angle, 15) in the arterial (20 sec) and portal venous (60 sec) phases. Images were evaluated by consensus of 2 MR radiologists. Tumor size, enhancement, and ADC values were recorded pre- and postablation and compared using the paired t-test.
Results: A total of 41 tumors were evaluated (median pre-ablation size = 2.1 cm). Following RFA, lesion size increased to 5.0 cm (p-value < 0.001). Both arterial and venous enhancement significantly decreased after RFA (median decrease was 10% and 5%, respectively; p-value < 0.001 and < 0.001). Tumors had a median ADC value of 1.89 E-3 mm2/sec before treatment, and the median ADC value decreased significantly after treatment to 1.17 E-3 mm2/sec (p = 0.008). Prior to RFA, all lesions were hypointense on unenhanced T1-weighted images and slightly hyperintense on the T2-weighted images compared to the surrounding liver parenchyma. In contrast, following RFA, 80% of the lesions were hyperintense on unenhanced T1-weigted images and the remaining 20% were hypointense relative to the surrounding liver parenchyma. On T2-weighted images 83% of the lesions were hypointense and the remaining 17% were hyperintense compared with the surrounding liver parenchyma.
Conclusion: Following RFA of hepatic metastases the treated site had an apparent increase in size, most likely due to a safety zone of ablation created around the tumor at the time of ablation. A significant reduction in enhancement and ADC value was also noted, suggesting complete coagulative necrosis and tissue dehydration of the lesion. Perfusion and diffusion MR imaging may be useful in confirming the presence of tissue necrosis and cellular dehydration, and may serve as a surrogate marker of tumor response, especially in the absence of unenhanced T1- and T2-weighted changes.
027. Low-Kilovoltage/High mAs 64-Slice MDCT for Detection of Hypervascular Liver Tumors: a Phantom Study
Schindera S.*; Nelson R.; Paulson E.; Jaffe T.; Miller C.; Toncheva G.; Nguyen G.; DeLong D.; Kawaji K.; Yoshizumi T.; Mukundan S. Duke University Medical Center, Durham, NC
Address correspondence to S. Schindera (sebastian.schindera{at}duke.edu)
Objective: To investigate the effect of low kVp/high mAs 64-slice multi detector computed tomography (MDCT) for detection of hypervascular liver tumors on the contrast-to-noise ratio (CNR), image noise, lesion conspicuity, and radiation dose.
Materials and Methods: A custom liver phantom with a background attenuation value consistent with liver parenchyma during the late arterial phase (86 HU at 140 kVp and 120 HU at 80 kVp) contained 16 cylindrical cavities of various sizes (3, 5, 8 and 15 mm). The cavities were filled with various iodinated solutions to simulate hypervascular liver lesions. The phantom was scanned with a 64-slice MDCT scanner (LightSpeed VCT, GE Healthcare) at 140, 120, 100 and 80 kVp with corresponding tube current-time product settings at 225, 275, 420, and 675 mAs, respectively. Scanner configuration (64 x 0.625 mm) and beam pitch (1.375) were identical for all scans. For quantitative assessment, the CNRs for eight lesions filled with different iodinated solutions were calculated. For qualitative assessment, three independent radiologists graded on a three-point scale (1 = low, 2 = medium, 3 = high) the conspicuity of 16 lesions. An anthropomorphic phantom was scanned by using the four protocols to evaluate the effective dose. Statistical analysis consisted of the one-way and two-way analysis of variance (p < 0.05).
Results: Image noise increased by 45% as the tube voltage dropped from 140 kVp to 80 kVp (p < 0.001). At 80 kVp, the highest CNR for each of the eight lesions was measured. As the tube voltage decreased from 140 kVp to 80 kVp, the CNR increase of the eight lesions ranged from 147% to 5,100% (p < 0.001). The lower the CNR of a lesion at 140 kVp, the higher the percentage increased at 80 kVp. At 80 kVp, the highest lesion conspicuity for all lesions was seen (3.0 ± 0). For the 3-mm lesions, lesion conspicuity increased as the tube voltage dropped to 80 kVp. The effective radiation dose decreased by 57% from the 140-kVp protocol (11.1 ± 0.6 mSv) to the 80-kVp protocol (4.8 ± 0.3 mSv) (p < 0.001).
Conclusion: Our phantom study data support the use of 80 kVp with a very high mAs setting for 64-slice MDCT during the late arterial phase to detect hypervascular liver tumors. By applying this technique, the CNR and the conspicuity of hypervascular tumors can be significantly increased while decreasing the radiation dose.
028. Does Contrast Concentration Matter? A Double-blind Comparison of Contrast Enhancement after Iopamidol or Iodixanol in Patients Undergoing Contrast Enhanced Multidetector CT of the Liver
Sahani D. V.3*; Lepanto L.1; Nelson R.2; Heiken J. P.4 1. CHUM Hospital St. Luc, Montreal, Canada; 2. Duke University, Durham, NC; 3. Massachusetts General Hospital, Boston, MA; 4. Washington University, St. Louis, MO
Address correspondence to D. Sahani (dsahani{at}partners.org)
Objective: We conducted a multicenter, double-blind, randomized, parallel-group study to compare the efficacy of contrast enhancement obtained with iopamidol-370 (370 mgI/mL) and iodixanol-320 (320 mgI/mL) in patients undergoing contrast-enhanced multidetector CT (MDCT) examinations of the liver.
Materials and Methods: A total of 121 patients were randomized to receive equi-iodine doses (40 gI) of iopamidol-370 or iodixanol-320 intravenously via power injector at a rate of 4 mL/sec. CT of the liver was performed using 16-slice MDCT scanners using predefined imaging protocols. Images were randomized and assessed by two independent readers unaffiliated with the study centers and blinded to subject identity, site information, dose, and contrast agent administered. Contrast density in Hounsfield Units (HU) was measured at pre-defined regions of interest (ROIs) during the arterial and portal venous phase of enhancement. Maximum enhancement was evaluated using a mixed effect model. ROIs clustered within a patient were treated as repeated measures. The difference in HU between the two contrast agents and its 95% confidence interval were estimated.
Results: A total of 60 patients received iopamidol-370 while 61 received iodixanol-320. Mean age, gender distribution, weight, total iodine dose, and dose/body weight were comparable in the two groups. In the arterial phase iopamidol-370 provided significantly higher HU values in abdominal aorta (Reader 1: 301.3 ± 80.2 vs. 273.6 ± 65.9 HU, 95% CI [6.1, 56.8], p = 0.02; Reader 2: 302.0 ± 73.6 vs. 275.1 ± 62.9 HU, 95% CI [2.3, 51.3], p = 0.03). No significant differences in HU were observed between the two contrast agents at the level of inferior vena cava, main portal vein, and normal liver parenchyma during the portal phase of enhancement.
Conclusion: When the same injection rate and iodine dose are used for MDCT of the liver, iopamidol-370 provides significantly greater enhancement during the arterial phase and similar enhancement during the portal venous phase compared to iodixanol-320. Use of the higher concentration agent resulted in better enhancement with a lower administered volume.
029. Comparison of Moderate Versus High Concentration Contrast Material Followed by Rapid Saline Flush at 64-slice CT (MSCT) for the Detection of Hepatocellular Carcinoma (HCC) in Cirrhotic Patients
Marin D.*; Catalano C.; Michele D.; Guerrisi A.; De Filippis G.; Passariello R. Università degli Studi di Roma "La Sapienza", Rome, Italy
Address correspondence to D. Marin (danielemarin{at}hotmail.it)
Objective: To prospectively assess aortic and hepatic enhancement and the detectability of hypervascular HCC between two contrast materials with moderate and high iodine concentrations.
Materials and Methods: One hundred twenty consecutive patients with cirrhosis underwent multiphasic CT performed with a 64-slice scanner (Sensation 64, Siemens). Fifty-four patients with hypervascular HCC were identified. Sixty-two patients were randomly assigned to receive iomeprol 400 mg I/kg (Iomeron 400; Bracco Imaging) with protocol A; 58, iodixanol 320 mg I/kg (Visipaque 320, Amersham Health) with protocol B. In both protocols, each patient received the same iodine load per body weight (521 mg/kg) with the same injection duration (25 seconds), followed by rapid saline flush using a double power injector. CT imaging parameters were: collimation, 0.6 x 64 mm; section thickness, 3 mm; mAs, 250; and kVp, 120. A triple-phase protocol was started 18 (arterial-), 60 (hepatic-venous-), and 180 (equilibriumphase) seconds after the trigger (threshold level set at an increase of 150 HU over the baseline CT number of the aorta). Enhancement of aorta and liver were measured in all patients. Tumor-to-liver contrast was measured in 54 patients with hypervascular HCC. Statistical analysis was performed with Mann-Whitney U test.
Results: Medians of aortic and hepatic enhancement during three phases were 310, 190, 75 HU and 35, 50, 33 HU with protocol A. Corresponding values were 270, 185, 73 HU and 23, 47, 33 HU with protocol B. Medians of tumor-to-liver contrast were 48, 1, and –0.5 with protocol A and 40, 0, and –2.4 with protocol B. During the first phase, a significantly higher aortic and hepatic enhancement and lesion conspicuity were registered with protocol A.
Conclusion: High iodine concentration of contrast material showed a higher conspicuity than medium concentration of contrast material for the detection of hypervascular HCC. No significant difference was observed during the hepatic-venous- and equilibrium-phase in terms of aortic and hepatic enhancement between both contrast agents.
030. Abdominal Lymph Nodes Seen on Coronal Reformations From Isotropic Voxels Using 16-slice MDCT: Do We Really See Them Better Than on the Axial Scan?
Chlebek S. J.*; Jaffe T. A.; Ho L. Duke University Medical Center, Durham, NC
Address correspondence to S. Chlebek (schlebek{at}gmail.com)
Objective: To test the hypothesis that abdominal lymph node enlargement identified on coronal reformations obtained with isotropic MDCT data sets is equal to the axial data sets of the abdomen and pelvis.
Materials and Methods: From June 2003 to October 2003, 105 consecutive patients with unexplained abdominal pain underwent 16-slice MDCT (Lightspeed 16, GE Healthcare) with coronal reformations. Protocol: oral contrast; 150 mL iopamidol at 3 mL/second; 16 x 0.625 mm, pitch 1.75, 17.5 mm/rotation, 0.5 seconds. Axial images were reconstructed first at 5 mm at 5-mm intervals and second at 06.25 mm at 0.5 mm intervals, and further reformatted coronally at 3 mm at 5-mm intervals. On a workstation (AW, GE Healthcare), two independent blinded readers reviewed a randomized set of scans (axials and coronals) and identified lymph nodes in four intraabdominal locations: root of the mesentery, right lower quadrant, upper retroperitoneum (diaphragm to iliac bifurcation), and lower retroperitoneum (below iliac bifurcation). One month later readers reviewed the scan obtained in the other imaging plane for the same findings. p values from the signed rank test of differences in the rates of positive calls between the two scans were obtained.
Results: The study included 74 women and 31 men, average age 48 (range 19–85 years). 53 cases were read first in the axial plane. p values for the signed rank test of differences are as follows: root of mesentery, 0.11; right lower quadrant mesentery, 0.28; upper retroperitoneum, 0.006; and lower retroperitoneum, 0.20.
Conclusion: With the exception of the upper retroperitoneum, there is no statistically significant difference in identification of abdominal lymph nodes between the axial and coronal planes. Readers are more likely to identify upper retroperitoneal lymph nodes in the axial plane.
031. Can Body Water Content Measured with Bio-electric Impedance Method Allow Contrast Material Dose Optimization for MDCT of Abdomen?
Namasivayam S.2*; Kalra M. K.2; Udayasankar U. K.1; Sahani D. V.2; Small W. C.1 1. Emory University Hospital, Atlanta, GA; 2. Massachusetts General Hospital, Boston, MA
Address correspondence to S. Namasivayam (namasivayam2005{at}yahoo.com)
Objective: Intravenous iodinated contrast material gets primarily distributed in the fluid compartment of the body. We hypothesize that total body water (TBW) content if available could predict the enhancement on CT and hence, the contrast material dose. However, no published reports are available on the correlation of enhancement on CT with TBW and still most centers used fixed dose of contrast material. Thus, the purpose of our study was to determine the correlation of enhancement on abdominal MDCT with TBW content and compare the correlation with that of body weight.
Materials and Methods: The institutional review board approved our study. Fifty-one consecutive patients (M:F, 24:27; mean, 53 years) referred for contrast-enhanced MDCT of abdomen were included. All CT studies were performed using a 16-slice MDCT scanner. Body weight and height of all patients were recorded. TBW content of all patients was measured using bio-electric impedance method (BIA-450, Biodynamics Corporation) before CT examination. Contrast injection protocol was 130 mL of 350 mg I/mL iohexol at 3mL/second with 30-mL saline chaser at 3 mL/second. Scanning was started at a threshold enhancement of 50 HU in liver using a bolus-tracking program. CT attenuation values of liver, portal vein (PV), and aorta were measured for all exams. Correlation of TBW and weight with enhancement was determined using Pearson's correlation coefficient.
Results: Average TBW and body weight of patients were 42 ± 10.5 L (mean ± standard deviation), and 86.7 ± 21.4 kg, with a range of 27.5–71.3 L and 50–152.9 kg, respectively. Average attenuation values of liver, PV, and aorta were 138.1 ± 17.9 HU, 209.0 ± 37.1 HU, and 199.1 ± 57.2 HU, respectively. TBW showed moderate inverse correlation with liver (r = –0.51; p < 0.01), PV (r = –0.54; p < 0.001), and aortic (r = –0.5; p = 0.012) enhancement. Likewise, liver (r = –0.54; p < 0.001), PV (r = –0.6; p < 0.001), and aortic (r = –0.56; p< 0.001) enhancement had moderate inverse correlation with body weight. The correlation of TBW with liver, PV, and aortic enhancement was comparable to the correlation of weight.
Conclusion: Liver, PV, and aortic enhancement on abdominal MDCT showed a moderate inverse correlation with TBW. Thus, patients with lower TBW content would need less contrast material dose for CT and vice versa. The correlation of TBW with enhancement on abdominal MDCT is comparable to that of body weight.
032. Recurrent Intrahepatic Cholangiocarcinoma after Surgical Resection: Recurrence Patterns and Prognostic Factors Based on Clinico-radiologic Features
Park M.*; Lee W.; Park M.; Park K.; Jeon T.; Lim H. K. Samsung Medical Center, Seoul, South Korea
Address correspondence to M. Park (minjung06.park{at}samsung.com)
Objective: To elucidate the recurrence patterns and prognostic factors of intrahepatic cholangiocarcinoma after curative resection based on clinicoradiologic features.
Materials and Methods: During a 9-year period, we collected 75 patients with intrahepatic cholangiocarcinoma (M:F = 48:27, mean age; 58 years) who survived over 3 months after curative resection and had pre-operative CT. We evaluated the recurrence rate, recurrence site, cumulative recurrence-free survival rate, and median time to disease recurrence. In addition, we evaluated various clinical (age, sex, symptom, operation type and preoperative treatment), laboratory (tumor markers), pathologic (cell type and growth pattern), and CT (tumor size, tumor location, invasion of the liver capsule, and the presence of satellite nodule, vascular invasion and lymphadenopathy) features with particular attention to CT features, and analyzed them with the use of univariate and multivariate analyses. All CT images were analyzed by consensus of three radiologists.
Results: The follow-up period was 3–123 months. Forty-one patients were revealed to have recurrent tumor with the recurrence rate of 48%. Recurrence sites were the liver (n = 25), lymph node (n = 9), resection margin (n = 8), peritoneal seeding (n = 6), lung (n = 3), skin (n = 1), bone (n = 1), and brain n = 1). The 1-, 3-, and 5-year cumulative recurrence-free survival rates were 55%, 36%, and 34%, respectively. The median survival was 15 months. A univariate analysis showed the tumor size, satellite nodule, and vascular invasion were significantly related with recurrence (p < 0.05). A multivariate stepwise Cox's hazard model revealed that the satellite nodule was a statistically independent factor.
Conclusion: The recurrence rate of intrahepatic cholangiocarcinoma after curative resection is relatively high, and the most common recurrence site is the liver. The prognosis of patients with recurrent intrahepatic cholangiocarcinoma is substantially poor, and some CT findings such as the tumor size, satellite nodule and vascular invasion at CT can be used as significant prognostic factors for the recurrence of intrahepatic cholangiocarcinoma.
033. CT Distinction of Adenomyomatosis and Gallbladder Cancer
Ching B. H.; Coakley F. V.; Westphalen A. C.; Joe B. N.; Qayyum A.; Yeh B. M.* University of California, San Francisco, San Francisco, CA
Address correspondence to B. Yeh (ben.yeh{at}radiology.ucsf.edu)
Objective: To investigate if CT findings can help in the distinction of adenomyomatosis and gallbladder cancer.
Materials and Methods: We retrospectively identified 39 patients with pathologically proven adenomyomatosis (n = 22) or gallbladder cancer (n = 17) who had undergone preoperative abdominal CT studies. The patients consisted of 25 women and 14 men with a mean age of 60 years (range, 20 to 86). Two readers reviewed all CT images and independently evaluated a variety of morphological findings, including the presence, location, and cystic appearance of gallbladder wall thickening, the appearance of direct liver invasion by a gallbladder mass; and the presence of liver metastases, gallstones, or lymphadenopathy.
Results: A cystic appearance of gallbladder wall thickening was significantly more common in patients with adenomyomatosis (8 and 8 patients) than with carcinoma (0 and 2 patients [p < 0.005 and p = 0.08] for readers 1 and 2, respectively). Direct invasion of the liver by a gallbladder mass was not seen by either reader in patients with adenomyomatosis, but was seen in 4 and 3 patients with carcinoma (p < 0.05 and p = 0.07, for readers 1 and 2, respectively). Focal thickening at the fundal region, presence of mural calcifications, an hourglass shape of the gallbladder, regional lymphadenopathy, liver metastases, and gallstones did not differ significantly in frequency between patients with adenomyomatosis and gallbladder carcinoma. Liver metastases were seen in 5 and 4 patients with adenomyomatosis by readers 1 and 2, respectively, due to metastatic tumor from a non-gallbladder source.
Conclusion: A cystic appearance of gallbladder wall thickening suggests a diagnosis of adenomyomatosis while direct hepatic invasion by a gallbladder mass suggests gallbladder carcinoma. The most characteristic features of gallbladder adenomyomatosis is a cystic appearance of focal wall thickening while the presence direct hepatic invasion by a gallbladder mass suggests carcinoma.
034. MRCP: Does Parallel Imaging (ASSET) Improve Image Quality and Duct Visualization Using 2D Thick Slab Breath Hold Single-shot Fast Spin Echo (SSFSE) Imaging?
Furlan A.*; Almusa O.; Hosseinzadeh K. UPMC, Department of Radiology, Division of Abdominal Imaging, Pittsburgh, PA
Address correspondence to A. Furlan (furlana{at}upmc.edu)
Objective: To compare two thick-slab SSFSE MRCP sequences with and without ASSET, in terms of image quality and visibility of the biliary tree and pancreatic duct.
Materials and Methods: 39 patients (18 M, 21 F, mean age 51.8 years) underwent MRCP between October 2005 and March 2006 at 1.5-T field strength using an 8-channel torso array coil. Multi-angled coronal thick-slab breath hold SSFSE imaging was performed without (TR/TE: 2507-3528/900ms, ETL: 222) and with integrated parallel imaging technique (TR/TE: 1240-1738/900ms, ETL: 103, ASSET factor 2). Images were retrospectively reviewed by two abdominal radiologists by consensus. A 4-point rating scale was used to grade the overall image quality and the visibility of 10 segments of the pancreatico-biliary ductal system. A comparison of the two techniques was conducted using a Wilcoxon signed rank test and a p value < 0.05 was considered to be statistically significant.
Results: Overall image quality was improved without the use of parallel imaging, however, the results were not statistically significant (p = 0.06). With respect to duct visualization, parallel imaging improved duct conspicuity of the common bile duct, medial and lateral segments of the left hepatic lobe, and posterior segment of the right hepatic lobe. However, a significant improvement in visibility (p = 0.03) was only achieved for the medial segment of the left hepatic lobe.
Conclusion: Improvement in overall image quality was not observed with parallel imaging. However, application of parallel imaging to thick-slab MRCP sequence can improve visibility of the smallest caliber second order branches.
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
| ||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
