Markus Düx

Prospective Comparison of Endoscopy, Endosonography and Computed Tomography for Staging of Tumours of the Oesophagus and Gastric Cardia

Background: Local and multimodal therapeutic strategies for tumours of the oesophagus and gastric cardia, require precise preoperative staging. Endosonography is considered the most accurate staging method, while computed tomography (CT) has limitations especially in the evaluation of local infiltration. Macroscopic endoscopic evaluation was reported to be accurate in selected series, but no study has yet compared all three staging modalities. Methods: One hundred and seventeen unselected patients with tumours of the oesophagus and gastric cardia were prospectively staged first by the endoscopic macroscopic appearance and then by endosonography. All patients had preoperative CT scans, however, only the 36 patients receiving the scans at our institution were included in the study. The preoperative staging results were then compared to postoperative histology which was available as the gold standard in all included patients. Kappa statistics were used to exclude chance agreement of the clinical staging results with the pathohistological findings. Differences between the resulting ĸ values for the different staging modalities were analysed with a jack-knife test. Results: Endoscopic macroscopic staging and endosonography (accuracy 67 and 69%, weighted ĸ 0.78 and 0.84) were significantly more accurate than CT (accuracy 33%, weighted ĸ 0.44) for determination of the T category (p = 0.006 and p = 0.001). After exclusion of tumours of the cardia (n = 33), the accuracy of macroscopic and endosonographic staging (accuracy 72 and 75%, weighted ĸ 0.86 and 0.88) increased and remained more accurate than CT (accuracy 50%, weighted ĸ 0.62). The main pitfall in our series in staging the T category was the overestimation of T2 tumours in the cardia as T3 or even as T4 tumours due to the inability to visualise the serosa. The accuracy of predicting lymph node metastasis was 68% for macroscopic endoscopic, 79% for endosonographic, and 67% for CT staging. Only endosonographic staging was significantly different from chance agreement with histology (weighted ĸ = 0.56). Endosonographic staging was significantly more accurate than endoscopic macroscopic and CT staging (p = 0.03). Conclusions: Endosonography is the most accurate staging modality for overall preoperative staging of oesophageal and cardial tumours. Endoscopic macroscopic staging allows a reasonably accurate assessment of the T category.

Helical hydro-CT for diagnosis and staging of gastric carcinoma.

PURPOSE The purpose of this work was to define the accuracy of helical hydro-CT (HHCT) in the diagnosis and staging of gastric carcinoma. METHOD One hundred twelve patients with gastric carcinoma were preoperatively imaged by HHCT. Gastric distension was achieved by ingestion of up to 1,500 ml of water. Bolus tracking was performed, and peristalsis was minimized by intravenously administered spasmolytics. Contrast material was then injected, and helical scanning was performed at the time of peak enhancement of the liver. CT images were analyzed for tumor infiltration of the gastric wall, and TNM staging criteria were applied according to the International Union Against Cancer (UICC) classification. The results were correlated with histopathologic findings. RESULTS One hundred two of 115 (89%) gastric carcinomas were correctly diagnosed by HHCT. Small malignant ulcers (< or =2 cm) that corresponded to early gastric carcinoma were not visible on CT scans. T and N staging accuracies were 51% each; abdominal M staging was correct in 79% of all cases. The positive and negative predictive values of HHCT to foresee curative resection of gastric carcinoma were 75 and 84%, respectively. CONCLUSION Mural thickening as well as marked contrast enhancement of the gastric wall are firmly related to gastric carcinoma. The accuracy of HHCT is acceptable for M staging but inadequate for local staging of gastric carcinoma. Nonetheless, HHCT is a useful guide for choosing between tumor resection and nonoperative treatment of patients. We therefore recommend HHCT as the method of choice for preoperative imaging of gastric carcinoma.

Retained surgical sponge with migration into the duodenum and persistent duodenal fistula.

Abstract. Migration of a retained surgical sponge into the bowel is a rare cause of bowel obstruction. Thus far, there have not been any reports that the site of initial migration of the sponge was identified by imaging studies or surgical exploration because the onset of symptoms is usually delayed. Unique about the case presented herein is that a barium meal follow-through study revealed a duodenual fistula that had developed after uneventful cholecystectomy due to a retained surgical sponge that had migrated into the duodenum and obstructed the distal jejunum. Imaging findings are presented and discussed.

Effects of Vascular Perfusion on Coagulation Size in Radiofrequency Ablation of Ex Vivo Perfused Bovine Livers

Objectives:A standardized perfused ex vivo bovine liver model was used to evaluate the effect of organ perfusion on coagulation size and energy deposition during radiofrequency ablation (RFA) procedures. Materials and Methods:Bovine livers were perfused in a tank after rinsing the prepared liver vessels with anticoagulants. Tyrodes solution, oxygenated and heated to 36.5°C, was used as perfusion medium. A flow and pressure controlled pump regulated Portal vein circulation; a dialysis machine provided pulsatile arterial circulation. Impedance-guided radiofrequency ablations were performed with 4-cm LeVeen electrodes with and without underlying liver perfusion. Two-dimensional diameters (Dv, Dh) of each ablation area were measured after dissecting the livers. Results:In 4 bovine livers weighing 8.85 ± 0.83 kg per organ (min, 7.7 kg; max, 9.7 kg) altogether 40 RF ablations were performed. A total of 20 ablations were generated with underlying liver perfusion (group 1) and 20 ablations with no liver perfusion (group 2). In group 1, Dv was 28.4 ± 5.3 mm, Dh 38.6 ± 7.8 mm, and energy deposition 36.9 ± 18.0 kJ. The 20 ablation areas generated without liver perfusion displayed statistically significant differences, with Dv being 35.7 ± 6.5 mm (P = 0.001), Dh 49.5 ± 9.4 mm (P = 0.001), and energy deposition 25.5 ± 13.0 kJ (P = 0.018). Conclusion:The model reproduced the cooling effect of perfused tissue during RFA. The ablation areas produced under perfusion conditions had smaller diameters despite longer exposure times and higher energy deposition.

MRI for staging of gastric carcinoma : First results of an experimental prospective study

PURPOSE Our goal was to define the accuracy of MRI in the staging of gastric carcinomas. METHOD Twenty consecutive surgical specimens were imaged immediately after gastrectomy for gastric carcinoma. Imaging was performed with a 1.0 T imaging system. T1-weighted, T2-weighted, and opposed phase images were acquired and analyzed for tumor infiltration of the gastric wall and the presence of perigastric lymph nodes. T and N stages were classified according to the International Union Against Cancer classification. Finally histopathologic staging of the specimens was compared with staging by MRI. RESULTS In gastric specimens, three to five layers of the gastric walls were visible. There were typical signal intensity patterns on T1-weighted, T2-weighted, and opposed phase images. Tumor diagnosis and lymph node detection were best achieved by opposed phase imaging. Nineteen of 20 (95%) carcinomas were localized by MRI; T staging accuracy was 65%. The sensitivity to detect metastatic lymph nodes was 87%, specificity 60%. N staging accuracy (nodes positive versus negative) was 80%. CONCLUSION High resolution MRI of gastric tumors is possible ex vivo. MRI enabled differentiation of up to five layers of the gastric wall, and therefore staging of gastric carcinomas is technically possible. However, to evaluate the exact role of MRI as a staging tool of gastric carcinomas, a correlation between MR morphology and the histologic structure of the gastric wall has to be achieved first.

TNM staging of gastrointestinal tumors by hydrosonography: results of a histopathologically controlled study in 60 patients

Abstract.Background: The study is a prospective evaluation of preoperative TNM staging of gastrointestinal tumors by hydrosonography (HUS). Methods: Sixty patients with suspected gastric or colorectal cancer underwent HUS for TNM staging. All patients were operated on and the tumors completely removed when possible. HUS findings were correlated with histopathologic staging. Results: HUS correctly localized tumors in 75% of patients. T stage accuracy was low for gastric cancers (41%). N staging of gastric cancers was accurate in 68% of all cases and was highly specific (100%). Staging was more accurate for colorectal tumors (70%), especially with respect to infiltration of other structures (sensitivity 100%, specificity 95%). N staging, however, was not reliable, mostly owing to impaired examination conditions. Conclusion: HUS easily misses tumors of the gastric cardia and distal part of the rectum. T staging of colorectal tumors with HUS is highly accurate, reaching 92% if the tumor is localized. T1 cancers of the stomach tend to be overstaged, and serosal infiltration by gastric cancers is often misjudged. With the exception of cardial gastric and distal rectal cancers, HUS comes close to endosonography for staging gastrointestinal tumors. HUS does not require intraluminal access.

CT and endoscopic ultrasound in comparison to endoluminal MRI: preliminary results in staging gastric carcinoma.

PURPOSE To prospectively compare diagnostic parameters of a newly developed endoluminal MRI (endo-MRI) concept with endoscopic ultrasound (EUS) and hydro-computer tomography (Hydro-CT) in T-staging of gastric carcinoma on one patient collective. MATERIAL AND METHODS 28 consecutive patients (11 females, 17 males, age range 46-87 years, median 67 years) referred for surgery due to a gastric malignancy were included. Preoperative staging by EUS was performed in 14 cases and by Hydro-CT in 14 cases within a time frame of 2 weeks. Ex vivo endo-MRI examination of gastric specimens was performed directly after gastrectomy within a time interval of 2-3h. EUS data were acquired from the clinical setting whereas Hydro-CT and endo-MRI data were evaluated in blinded fashion by two experienced radiologists and one surgeon well experienced in EUS on gastric carcinomas. RESULTS Histopathology resulted in 4 pT1, 17 pT2, 3 pT3 and 2 pT4 carcinomas with 2 gastric lymphomas which were excluded. Overall accuracy for endo-MRI was 75% for T-Staging of the 26 carcinomas. EUS achieved 42.9% accuracy; endo-MRI in this subgroup was accurate in 71.4%. Hydro-CT was correct in 28.6%, accuracy for endo-MRI in this subgroup was 71.4%. CONCLUSION The direct comparison of all three modalities on one patient collective shows that endo-MRI is able to achieve adequate staging results in comparison with clinically accepted methods like EUS and Hydro-CT in classifying the extent of tumor invasion into the gastric wall. However the comparison is limited as we compared in vivo routine clinical data with experimental ex vivo data. Future investigations need to show if the potential of endo-MRI can be transferred into a clinical in vivo setting.

Der Stellenwert der CT in der Akutdiagnostik der Divertikulitis

ZusammenfassungZielsetzung. Bei der akuten Divertikulitis entscheidet eine exakte Stadieneinteilung der Erkrankung über das therapeutische Vorgehen.Die Diagnostik mittels Computertomographie (CT) wird der klinischen Einteilung der akuten Divertikulitis und den sich daraus ableitenden therapeutischen Strategien gegenübergestellt. Material und Methode. Die Durchführung der CT bei Verdacht auf akute Divertikulitis erfolgt nach transanaler, intraluminaler Kontrastierung des Darms mit positivem wasserlöslichem Kontrastmittel, um eine Penetration oder Darmperforation nachzuweisen. Anschließend erfolgt eine intravenöse Kontrastmittelapplikation, um das Kontrastverhalten der abdominellen Strukturen/Organe, insbesondere der Darmwand, zu beurteilen und Abszesse zu diagnostizieren. Die CT-Morphologie der verschiedenen Stadien der akuten Divertikulitis wird beschrieben und anhand von Beispielen vorgestellt. Ergebnisse. Die CT kombiniert bei der notfallmäßigen Abklärung einer akuten Divertikulitis als einziges Verfahren Sicherheit mit Genauigkeit.Auf der einen Seite ist die CT schnell und damit sicher im Hinblick auf die Überwachung des Patienten.Auf der anderen Seite erlaubt die CT eine exakte Stadieneinteilung der Entzündung mit einer Sensitivität und Spezifität von bis zu 100%. Durch den sicheren Nachweis von Komplikationen wie die Ausbildung eines Abszesses, einer Penetration oder Darmperforation,hat die CT unmittelbaren Einfluss auf die Therapie des Patienten. Diskussion. Die CT ist bei Verdacht auf eine akute Divertikulitis aufgrund der hohen Relevanz für die Therapie und das Management von Komplikationen die diagnostische Methode der Wahl.AbstractPurpose. In acute diverticulitis accurate diagnosis and staging are mandatory to decide on the treatment of the patient.The impact of computed tomography (CT) on the treatment of acute diverticulitis will be discussed. Material and Method. CT is performed after distension of the distal colon by means of positive,water-soluble contrast media to depict intestinal perforation or penetration. Then intravenous contrast material is administered and spiral scanning is repeated to judge enhancement patterns of the abdominal structures/organs especially of the intestinal wall and to diagnose abscess formation. CT-morphologies of different stages of acute diverticulitis will be described. Results. CT imaging is the only diagnostic method that in case of an acute diverticulitis combines safety with accuracy. On the one hand, it is fast and therefore safe with respect to patient control and on the other hand, it allows accurate staging of the inflammatory process reaching a sensitivity and specificity of up to 100%, each. CT is an appropriate tool to diagnose acute diverticulitis complicated by abscess formation, intestinal penetration or perforation and therefore has direct impact on the treatment of the patient. Discussion. If acute diverticulitis is suspected CT is the method of choice for imaging because of its high impact on the choice of therapy and on the management of complications.

MRI of pouch-related fistulas in ulcerative colitis after restorative proctocolectomy.

PURPOSE Our purpose was to determine the value of MRI in diagnosing pouch-related fistulas in patients with ulcerative colitis and to compare pulse sequences with and without contrast enhancement in their performance of visualization. METHOD Forty-four patients with pelvic symptoms after restorative proctocolectomy underwent MRI. All 26 patients with pouch-related fistulas were treated surgically; 18 patients with pouchitis were treated conservatively. MRI was performed at 1.0 T with T1-weighted FLASH sequences before and after administration of Gd-DTPA, T2-weighted and proton density-weighted turbo SE sequences, and a T2-weighted fat saturation sequence. Images were analyzed for the presence of fistula; pulse sequences were additionally compared for best visualization on a four point scale of diagnostic confidence. RESULTS MRI detected 23 of 26 cases of fistulas; there were no false-positive diagnoses. Surgery revealed fistulas in three cases in which no pathology was found on MRI. Two patients had a short sinus tract at the pouch-anal anastomosis, and a third patient had a pouch-vaginal fistula. The Gd-enhanced FLASH sequence obtained the highest score, and second best was the T2-weighted fat saturation technique. CONCLUSION MRI is a valuable technique for diagnosing pouch-related fistulas, However, there are limitations in detection of short sinus tracts and pouch-vaginal fistulas. Highest diagnostic confidence is obtained with a Gd-enhanced FLASH sequence, which might be helpful after pelvic surgery or if the fact saturation technique is equivocal.

Magnetic resonance imaging for local staging of gastric carcinoma: results of an in vitro study.

Objective: Preoperative staging of gastric carcinoma is limited by the fact that available imaging modalities do not enable accurate evaluation of the depth of infiltration of the gastric wall. The aim of this study was to evaluate the efficiency of conventional magnetic resonance imaging (MRI) in local staging of gastric carcinoma. Methods: Sixty-five specimens of patients with proven gastric carcinoma were examined immediately after gastrectomy. Examination was performed with a 1-T MRI and included T1-weighted, T2-weighted, and opposed phase images. Images were analyzed for the number of visible wall layers and their signal intensity characteristics, for tumor localization and depth of infiltration. T-stage was classified according to the TNM system. Finally, the staging by MRI was compared with the histopathological staging of the specimens. Results: The mucosal, submucosal, and proper muscle layers could be differentiated by the typical signal intensities. Depiction of the subserosa or serosa was not possible. In 65 specimens, 67 carcinomas were found by the pathologist. Sixty-four of 67 (96%) histologically proven carcinomas were correctly localized by MRI; T-staging accuracy was 50% only, mainly because of overstaging pT2 tumors as T3. Conclusions: MRI enables differentiation of gastric wall layers and, therefore, technically allows the evaluation of the local tumor stage of gastric carcinomas. However, infiltration of the subserosal and serosal layer cannot be proved accurately. Overstaging pT2 tumors is one of the most predominant problems. Yet further technical developments in high-resolution imaging of the gastric wall may improve T-staging in the near future and overcome todays staging limitations.