Outcomes of surgical, percutaneous endoscopic, and percutaneous radiologic gastrostomies.

OBJECTIVES: To evaluate and compare outcomes and complications in patients having undergone gastrostomy by surgical (SG), percutaneous endoscopic (PEG), or percutaneous radiological (PRG) procedure. DESIGN: Retrospective analysis. SETTING: University-based tertiary care center. PATIENTS: Of 82 patients who met inclusion criteria, 14 patients (median age, 40 years) received a surgical tube placement (SG), in 24 patients (median age, 55 years) a PEG procedure was performed, and in 44 patients (median age, 57 years) the tube was placed under fluoroscopic guidance (PRG). Indications for gastrostomy were similar in all groups, representing mainly cancer of the oropharyngeal, head and neck region (51 [61%]) as well as the upper gastrointestinal tract (6 [8%]), neurological disorders (15 [18%]), and others (10 [13%]). MAIN OUTCOME MEASURES: Catheter function rates, major and minor procedure-related complications, and survival. RESULTS: Median follow-up was 17.2 months. Ten patients (71%) died in the SG group 7 to 855 days (median, 67 days) after the procedure, 7 patients (29%) died 5 to 263 days (median, 103 days) after PEG placement, and 30 patients (68%) died within 3 to 621 days (median, 112 days) after PRG, of their underlying disease or disease-related complications; 1 procedure-related death occurred 6 days after radiological tube placement. We observed a rate of minor complications of 43% (6 patients), 33% (8), and 36% (16) and a major complication rate of 14% (2 patients), 17% (4), and 11% (5) in the SG, PEG, and PRG groups, respectively. Tube function rates at 1 year were 67% (9 patients) and 68% (20) in the SG and PEG groups, respectively, and 10% lower (39) in the PRG group, although the difference was not statistically significant. CONCLUSIONS: There is no major difference between SG, PEG, and PRG concerning procedure-related complications. Tube function tends to be inferior after radiological tube placement.

Arterial hepatic embolization of unresectable hepatocellular carcinoma using a cyanoacrylate/lipiodol mixture.

PURPOSE: A survival analysis in 16 patients with unresectable hepatocellular carcinoma (HCC) undergoing transcatheter arterial embolization (TAE) using a combination of lipiodol and N-butyl-2-cyanoacrylate (5:1) was performed in a retrospective study. METHODS: A combination of lipiodol and N-butyl-2-cyanoacrylate (5:1) was used for TAE. All patients had disease compatible with Okuda stages I and II. RESULTS: Twenty-four embolizations were done; five patients had more than one embolization. Median alpha-fetoprotein levels declined from 116 to 48.6 ng/ml. A median of 0.3 ml cyanoacrylate was administered per patient. Median survival was 8.5 months (range 2-49 months). After a median follow-up of 4 years, 12 patients have died (75%). Okuda stage I and II patients had a median survival time of 34.4 and 5.5 months respectively. Few side effects (19%) were seen. CONCLUSION: We conclude that the TAE procedure used [lipiodol and N-butyl-2-cyanoacrylate (5:1)] is safe and produced only few side effects, thus constituting a valuable therapeutic option for patients with Okuda stage I and II HCC.

[Treatment of arterial aneurysms of the pelvic leg vessels using dacron covered nitinol stents]. / Behandlung arterieller Aneurysmen der Becken-Bein-Gefässe mittels Dacron-ummantelter Nitinol-Stents.

PURPOSE: To evaluate the success rates of the implantation of stent grafts in the treatment of peripheral aneurysms. MATERIALS AND METHODS: In 13 patients with 15 aneurysms at the common iliac artery (n = 6), external iliac artery (n = 1), hypogastric artery (n = 2), femoral artery (n = 2) or popliteal artery (n = 4), implantation of dacron-covered nitinol stents was performed. The patients were followed up for three to 20 months (mean, 8.8 months) with intravenous digital subtraction angiography, CT or colour-coded Doppler sonography. RESULTS: In all cases, the aneurysm was successfully occluded after stent implantation. In one case with a popliteal aneurysm, kinking of the vessel caused thrombosis of the stent. The stent was successfully reopened. The aneurysm however, had to be surgically treated 9 months later. The primary and secondary patency rates at 6 months were 93% and 100%, respectively. CONCLUSION: The method described might be an alternative therapy to surgery. There are some limitations in respect of treatment of popliteal aneurysms.

Transjugular intrahepatic portosystemic shunt and cardiac arrhythmias.

The approach to the liver for a transjugular intrahepatic porto-systemic shunt (TIPS) is through the venous system. Because catheter and guidewire system traverses the heart, cardiac arrhythmias may be expected during the procedure. We have prospectively investigated the incidence of such dysrhythmias during TIPS implantation. Twelve consecutive patients, 4 women and 8 men aged 26 to 75 years (mean, 58 +/- 13 years), were studied. Before and on the day of TIPS implantation, a 24-hour Holter recording was performed. Transjugular intrahepatic portosystemic shunt implantation was performed under local anesthesia (lidocaine) and sedoanalgesia (midazolam and fentanyl). None of the patients had concomitant cardiac disease or electrolyte disturbances. In all patients except one, TIPS implantation was successful without any technical complications. A mean of 43 +/- 5.3 hours of Holter recording was performed before and after TIPS implantation. All recordings obtained during this control period were considered inconspicuous. The mean heart rate was significantly higher during the implantation procedure of 136 +/- 37 minutes' duration (83 +/- 20 beats per minute vs 70 +/- 19 beats per minute; p < 0.01). Nine of the 12 patients experienced episodes of nonsustained supraventricular tachycardias, and one patient had two sustained supraventricular tachycardias. Frequent episodes of nonsustained ventricular tachycardias developed in 75% of the patients. It seems clear that TIPS implantation is frequently associated with supraventricular and ventricular tachyarrhythmias even in patients with apparently good cardiac condition at the beginning of the procedure. Thus close cardiac monitoring with resuscitation equipment immediately available throughout the procedure is mandatory.

Endovascular AAA treatment: expensive prestige or economic alternative?

OBJECTIVES: To compare the costs of endovascular aneurysm treatment versus open surgery during the perioperative period. METHODS: Retrospective analysis of a consecutive series of 44 patients undergoing infrarenal abdominal aneurysm repair from February 1995 to March 1996 at a university teaching hospital. RESULTS: No endovascular procedure was converted to open repair. Operative time was shorter for endovascular treatment (207.6 min vs. 229.1 min, n.s.), as well as postoperative intensive care unit stay (ICU, 22.7 h vs. 55.0 h, p = 0.017) and the postoperative recovery period (5.6 days vs. 13.3 days, p < 0.001). Open surgery generated significantly more costs (25,374.07 ECU vs. 22,268.78 ECU, p < 0.001), despite evaluation and a more expensive endovascular procedure (10,699.48 ECU vs. 4032.01 ECU, p < 0.001). During the study, costs for open surgery exceeded the cost for endovascular treatment by 13.95%. CONCLUSIONS: Endovascular aneurysm treatment is cost effective and less expensive than open surgery. The main reason for cost saving is faster patient recovery after surgery, associated with a shorter LOS in the patients treated with endovascular procedure.

Evaluation of abdominal aortic aneurysm for stent-graft placement: comparison of gadolinium-enhanced MR angiography versus helical CT angiography and digital subtraction angiography.

PURPOSE: To determine whether magnetic resonance (MR) angiography can be used alone to evaluate abdominal aortic aneurysms (AAAs) for endovascular placement of stent grafts. MATERIALS AND METHODS: Sixty-one patients with AAAs underwent gadolinium-enhanced MR angiography of the abdominal aorta and pelvic arteries. Measurements of the size and extent of the AAAs were compared with helical computed tomographic (CT) and digital subtraction angiographic measurements; 95% confidence intervals for the differences in the means were determined. RESULTS: Because of the larger field of view, MR angiography was superior to CT angiography in assessing visceral iliac artery disease. Both modalities were equal in evaluating the proximal extent of the AAA (mean difference, -0.16 mm; 95% CI, -0.31, 0.64) and in measuring all aortic dimensions (e.g., mean difference in the proximal neck diameter, -0.74 mm; 95% CI, -0.98, -0.49). MR angiography was inferior to CT angiography in depicting accessory renal arteries (seven of 12) and in grading renal artery stenoses (sensitivity, 100% [95% CI, 0.90, 1.00]; specificity, 84% [95% CI, 0.74, 0.91]). CONCLUSION: Gadolinium-enhanced MR angiography is a fast, reliable means of providing all the information relevant to the preoperative assessment of endovascular aortic stent-graft placement.

Spiral CT angiography in the assessment of abdominal aortic aneurysms after stent grafting: value of maximum intensity projections.

PURPOSE: Our goal was to evaluate the value of maximum intensity projections (MIPs) rendered from spiral CT in the assessment of abdominal aortic aneurysms after stent graft implantation. METHOD: Spiral CT of the aorta was performed in 28 patients 1 week after implantation of covered aortic stent grafts. MIPs were evaluated in a blinded fashion. Transaxial scans and intravenous angiographs were considered the reference standard. RESULTS: Stent deformity was evident on the MIPs in 5 cases and stent angulation was seen in 24 cases. The excluded aneurysmal sac was visualized in 14 (50%) cases. MIPs depicted the patent inferior mesenteric artery in 20 of 23 cases (87%). Renal artery occlusion (n = 3), leaks (n = 8), and parietal thrombi, (n = 3) were detected on the MIPs in all cases. CONCLUSION: MIPs are a valuable tool in the assessment of aortic stent grafts.

Radiographic detection of intrabronchial malpositions of nasogastric tubes and subsequent complications in intensive care unit patients.

OBJECTIVE: The aim of our study was to illustrate the radiographic spectrum of the intrabronchial malposition of nasogastric tubes and subsequent complications, and to discuss the role of radiography in the detection of such malpositions. DESIGN: Retrospective clinical investigation. SETTING: Tertiary care university teaching hospital. PATIENTS AND METHODS: We reviewed chest radiographs of 14 intensive care patients with nasogastric tubes malpositioned in the tracheobronchial tree. The site and anatomic location of the malposition were recorded. Complications due to tube malpositioning were monitored on follow-up radiographs and on computed tomographic examinations, which were available in 4 patients. RESULTS: Nine of 14 nasogastric tubes were inserted in the right and 5 in the left tracheobronchial tree. Tube tips were malpositioned in the lower lobe bronchi (50%), the intermediate bronchus (36%), and the main bronchi (14%). There was perforation of the bronchial system with subsequent pneumothorax in 4 patients. In 4 other patients, pneumonia developed at the former site of the malpositioned tube tip. Radiographic detection of nasogastric tube malpositioning was prompt in 9 patients and delayed in 5 patients. CONCLUSIONS: Whereas clinical signs of nasogastric tube malpositioning in intensive care patients may be absent or misleading, chest radiography can accurately detect nasogastric tube malpositions in the tracheobronchial tree, may prevent complications, and avoid the use of further costly or invasive diagnostic techniques.

Azygos arch cannulation by central venous catheters: radiographic detection of malposition and subsequent complications.

The aim of this study was to determine the frequency of radiographically evident central venous catheter misplacement in the azygos arch and to analyze whether the frequency of azygos arch cannulation is dependent on the anatomical site of catheter insertion. We reviewed 1,287 postprocedural examinations and 3,441 follow-up examinations. Catheters had been inserted through the left (6%) or right (15%) internal jugular veins and through the left (32%) or right (46%) subclavin veins. Radiographs were analyzed for possible catheter malposition in the azygos arch and for complications related to this malposition. Catheter malposition in the azygos arch was seen on 16/1,287 (1.2%) postprocedural radiographic examinations. Of the 16 malpositioned catheters, 11 (69%) had been inserted in the left subclavian vein, three (19%) in the left jugular vein, two (12%) in the right subclavian vein, and none (0%) in the right jugular vein. There was a statistically significant difference in the frequency of azygos arch cannulation between left- and right-sided catheters (p = 0.001). All complications consisted of venous perforations and were seen in three of 16 cases (19%). Azygos arch cannulation is a rare but hazardous central venous catheter malposition that occurs early after catheter insertion and carries a substantial risk for complication. The risk for azygos arch cannulation is substantially increased if catheters are inserted in left-sided veins. Because of the severity of subsequent complications, radiologists should be vigilant in the detection of this rare malposition.

Common bile duct obstruction due to malignancy: treatment with plastic versus metal stents.

PURPOSE: To compare the clinical efficacy and treatment costs of plastic versus metal biliary stents. MATERIALS AND METHODS: In a randomized trial, 101 patients with malignant common bile duct obstruction underwent transhepatic stent implantation and were followed up until death. Patients were stratified into risk and nonrisk groups. Forty-nine patients received 12-F plastic stents, and 52 received expandable metal stents. Plastic endoprostheses were placed in a two-step procedure; metal stent, in a single procedure. Kaplan-Meier analyses were used to compare patient survival and stent patency rates. RESULTS: The 30-day mortality rate was significantly lower for metal stents (five of 52 [10%]) than plastic stents (12 of 49 [24%]; P = .05). The obstruction rate was 19% (10 of 52; median patency, 272 days) for metal stents and 27% for plastic stents (13 of 49; median patency, 96 days; P < .01). Median time until death or obstruction was longer for metal stents (122 vs 81 days; P < .01). Placement of metal stents was associated with shorter hospital stay (10 vs 21 days; P < .01) and lower cost ($7,542 vs $12,129; P < .01). CONCLUSION: Use of self-expanding metal stents appears to show substantial benefits for patients and to be cost-effective.

Two cases of fractured esophageal nitinol stents.

We report here on two cases of fractured nitinol stents in the esophagus. In case 1, the correctly inserted stent broke spontaneously shortly after insertion. In a second case, a nitinol stent broke after laser application due to tumor ingrowth with massive bleeding. In both cases, a second stent was implanted in order to reestablish food passage. The fracture of the stent in case 1 seemed to be caused by defective material, whilst in case 2 the stent broke because of thermal overstrain during laser application. The use of electrocoagulation or laser in the stent area should therefore be avoided; argon plasma coagulation may offer an effective alternative in treating tumor ingrowth. As there was a risk from piercing broken filaments with the second stent, covering a fracture using stents with tight walls or plastic tubes seems to be a more effective approach than the inserting an uncovered stent type.

[Spiral-CT [corrected] in chronic lung thromboembolism. / Spiral-CT bei chronisch-thrombembolischer Lungenerkrankung.

Chronic thromboembolic pulmonary hypertension (CTEPH) is a rare result of recurrent pulmonary embolism and is treated by pulmonary thromboendarterectomy. Knowledge of the exact location of the thrombi is necessary in planning this operation. To date, pulmonary, angiography is the diagnostic imaging gold standard. Since the introduction of spiral CT excellent vascular opacification of the pulmonary arteries has become feasible, and thrombi in the pulmonary arteries can be visualized directly. Spiral CT is superior to angiography in demonstrating thrombi in the central pulmonary arteries, whereas angiography proves superior to CT in the evaluation of abnormalities within segmental arteries. The sensitivity of spiral CT in confirming the diagnosis of CTEPH is reported to be more than 90%. According to the literature and based on our own results, the decision concerning operability is possible on the basis of spiral CT images in more than 80% of patients with CTEPH. Spiral CT as a non-invasive tool may be used for CTEPH screening, for postoperative follow-up after pulmonary thrombendarterectomy and, combined with pulmonary angiography, to optimize operation planning.

Bronchial wall thickness: appropriate window settings for thin-section CT and radiologic-anatomic correlation.

PURPOSE: To analyze the influence of computed tomographic (CT) window settings on bronchial wall thickness and to define appropriate window settings for its evaluation. MATERIALS AND METHODS: Three inflation-fixed lungs were scanned with a section thickness of 1.5 mm by using a high-spatial-frequency algorithm. Wall thickness in 10 bronchial specimens was measured with planimetry. Window centers were altered in a range of -200 to -900 HU and window widths in a range of 400-1,500 HU. Relative and absolute differences between CT and planimetric values were calculated. CT and planimetric measures were correlated. Inter- and intraobserver variabilities were determined. RESULTS: Window widths less than 1,000 HU resulted in a substantial overestimation of bronchial wall thickness, whereas widths greater than 1,400 HU resulted in an underestimation of bronchial wall thickness. There was no interaction between "width" and "center" regarding their influence on bronchial walls (F = 0.23; P = .99). Correlation between CT and planimetry was statistically significant (r = .85; P = .0001). Differences between the two observers were not statistically significant; results of the measurements of the two observers correlated well (r = .97; P = .001). CONCLUSION: Bronchial wall thickness on thin-section CT scans should be evaluated with window centers between -250 and -700 HU and with window widths greater than 1,000 HU. Other than window settings, notably window widths less than 1,000 HU, can lead to substantial artificial thickening of bronchial walls.

[Contrast media in peripheral angiography: does cost reduction mean loss of quality and safety? A reevaluation based on a randomized double-blind comparative study of ioxaglate versus iopromide]. / Kontrastmittel in der peripheren Angiographie: Bedeutet Kostenreduktion Verlust an Qualität und Sicherheit? Eine Reevaluierung anhand einer randomisierten Doppelblind-Vergleichsstudie von Ioxaglat versus Iopromid.

PURPOSE: To evaluate whether meglumine-sodium-ioxaglate (Hexabrix) and iopromide (Ultravist) are identically appropriate for peripheral angiography. Outcome variables were pain, image quality and adverse events. METHODS: Sixty patients were included in a randomised double-blind study. In all patients an intraarterial digital subtraction angiography (i.a. DSA) of iliac and peripheral arteries was performed. RESULTS: Analysis of the study revealed no significant difference between both contrast media in terms of the main and additional outcome variables. In comparison to iopromide, ioxaglate caused milder pain sensations (VAS 4.70 vs. 7.76, p = 0.25). Mild adverse events were observed more frequently in ioxaglate angiography (11% vs. 0%, p = 0.1). CONCLUSION: Both contrast media seem to be appropriate for peripheral angiography using DSA technique. Ioxaglate causes a cost reduction of about 20-35%. However, an increase of mild adverse reactions up to 11% to 15% has to be accepted.

Spiral computed tomographic angiography of the renal arteries: a prospective comparison with intravenous and intraarterial digital subtraction angiography.

PURPOSE: To assess the accuracy of computed tomographic angiography (CTA) in the evaluation of the renal arteries in comparison with intravenous (IVDSA) and intraarterial digital subtraction angiography (IADSA). METHODS: In 18 patients, 35 CTAs and DSAs (27 IADSA, 8 IVDSA) of the renal arteries were performed. CTA was done with 2-3 mm collimation,2-4 mm/sec table speed, after intravenous injection of 80 ml of contrast medium at 4 ml/sec with a scanning delay time of 14-21 sec. No previous circulation time curve was performed. CTA data were reconstructed with maximum intensity projection (MIP) and shaded surface display (SSD). The presence of stenosis was assessed on a three-point rating scale (grade 1-3). The quality of the examinations; visualization of the ostium, the main artery, and its branches; vessel sharpness, linearity, and intraluminal contrast filling were evaluated. We compared CTA with DSA. RESULTS: CTA had 96% sensitivity, 77% specificity, and 89% accuracy in the detection of stenoses > 50%. Due to technical errors two stenoses were erroneously diagnosed as positive but there were no false negative diagnoses. The quality of CTA was good in 56% and moderate in 34% of cases. Visualization of the ostium and main artery was graded as 1.74 (out of 2) points and of the renal branches as 1.02 (out of 2) points and of the renal branches as 1.02 (out of 2) points. The quality of CTA images was worse than that of IADSA in 52%, equal in 41%, and better in 7% of cases. CTA was equal to IVDSA in 25% and better in 75% of the cases. CONCLUSION: CTA is an accurate noninvasive method for the evaluation of renal arteries. Examination quality is essential for the diagnosis. CTA is limited in its ability to visualize the branches of the renal artery and accessory arteries. CTA seems to be superior to IVDSA.

Successful exclusion of a large femoropopliteal aneurysm with a covered nitinol stent.

A 70-year-old woman presented with a large femoro-popliteal aneurysm. A covered nitinol stent was implanted successfully and complete exclusion of the aneurysm was achieved. At follow-up 5 months later the stent was still patent and the patient was free of symptoms. However, moderate stenosis was seen at the proximal end of the stent.

Modifications of the paraspinous muscle flap: anatomy and clinical application.

Soft-tissue defects of the back, particularly involving the paravertebral tissues, are generally covered with myocutaneous, muscle, or fasciocutaneous flaps. The case of a 64-year-old man with a paravertebral malignant fibrous histiocytoma is reported. To ensure adequately radical margins, the ipsilateral trapezius and latissimus dorsi muscles as well as the costal periosteum and the spinous processes were resected between T9 and T12. The resulting defect was covered with a pedicled latissimus dorsi flap and an island flap of the paravertebral muscles. Prompted by this case, we studied the blood supply of the paravertebral muscles in 10 cadavers. The vasculature was visualized after flushing with colored latex and microsurgical dissection. Another 4 specimens were subjected to angiography and tomography. In the majority of cases (8 of 10), three perforators emerging from the intercostal arteries were identified. These were found to communicate in a longitudinal and vertical direction. Before piercing the fascia, they ramified in three layers matching the layers of the paravertebral muscles. Since the intercostal arteries were shown to communicate through anastomoses of adequate caliber, the paravertebral muscles appear to be useful candidates for proximally or distally pedicled transposition or island flaps.

Covered self-expanding transhepatic biliary stents: clinical pilot study.

PURPOSE: We report our preliminary results with a new type of self-expanding covered stent for treatment of malignant biliary obstruction. METHODS: Wallstents, fully covered with high elasticity polyurethane, with an unconstrained diameter of 10 mm and a total length of 69 mm, were placed transhepatically under fluoroscopic guidance in five patients. The length of the biliary obstruction varied between 30-50 mm. At 1 and 3 months (82-98 days) clinical assessment, serum bilirubin measurement, and ultrasound examination of the biliary tree were performed. RESULTS: Initial uncomplicated deployment of the stents and internal drainage was possible in all patients. Distal stent migration resulted in early biliary reobstruction in one patient. At 3-month follow-up, partial reobstruction, most probably due to sludge formation, was found in another patient. CONCLUSION: Our initial results indicate that the covered, self-expanding Wallstent endoprosthesis can be reliably and safely deployed transhepatically for malignant biliary obstruction.

Palliative treatment of obstructing esophageal cancer with nitinol stents: value, safety, and long-term results.

OBJECTIVE: The purpose of our study was to evaluate the effectiveness of nitinol stents for palliation of dysphagia due obstructing esophageal cancer, safety of stent placement, and long-term results. SUBJECTS AND METHODS: Self-expanding uncoated nitinol stents were inserted either radiologically (14 patients) or endoscopically (12 patients) on an outpatient basis in 26 consecutive patients with dysphagia grade 3 or 4 caused by incurable malignant obstructions in the middle or distal third of the esophagus (n = 22) or at esophagojejunal anastomoses (n = 4). No esophagotracheal fistulas were seen in any patient. In 22 patients prior treatments had failed. Following insertion, the stent lumen was dilated to the maximum diameter. Finally, esophagography or esophagoscopy was done to confirm the position of the stent and patency of the esophageal lumen. Twenty-four hours after the procedure, esophageal function was investigated by a barium swallow. Patients were encouraged to ingest solid food thereafter. Improvement in dysphagia was evaluated 1 week after stent placement and during monthly interviews. Complications were defined as major (aspiration, bleeding, stent misplacement or dislocation, perforation) or minor (reflux esophagitis, chest pain, pharyngeal discomfort). Tumor ingrowth or overgrowth was considered a treatment failure. Twenty-three patients (88%) were followed until death: three patients (12%) were followed for a mean of 14 months. RESULTS: Exact positioning of the stent and dilation to its maximum diameter were technically feasible in all patients. No stents were placed in the stomach. Patency of the esophageal lumen was successfully restored in 25 patients. In one patient a broken strut of the stent after dilation caused a partial obstruction, which was detected endoscopically. Two patients had recurrent dysphagia due to tumor ingrowth or overgrowth, one after 1 month and the other after 3 months. In these patients an additional overlapping stent was successfully placed. No procedure-related mortalities or major complications occurred. The mean dysphagia grade of 3.5 was improved to a mean grade of 0.6 after stent placement. All patients could take liquids within the first 24 hr. Fifteen patients improved to dysphagia grade 0, seven patients to grade 1, and four patients to grade 2 within 1 week after the procedure. Twenty-three patients (88%) died during the follow-up period (mean survival, 5 months) as a result of their disease. Latest evaluation of the mean dysphagia grade was 0.7. Three patients (12%) are still alive (mean survival, 14 months) with a dysphagia grade 1 in one patient and grade 0 in two. CONCLUSION: Implantation of nitinol stents proved to be an effective and safe method of palliating severe dysphagia in patients with obstructing esophageal cancer. The improvement in dysphagia was impressive and long lasting. Placement of the stents was feasible without major procedure-related complications.

Malignant insulinoma: permanent hepatic artery embolization of liver metastases--preliminary results.

PURPOSE: To evaluate permanent hepatic artery embolization of liver metastases of malignant insulinoma as a therapeutic procedure. METHODS: Three female patients had persistent severe hypoglycemia after distal pancreatectomy because of a malignant insulinoma. Computed tomography (CT) and CT-portography (CTAP) were used for tumor assessment and follow-up and demonstrated multiple hypervascular metastases 0.5-3 cm in diameter in both lobes of the liver. Unilobar sequential transcatheter embolization of the hepatic artery was performed with an interval of 1-2 months between the procedures. Permanent occlusion was achieved by using a mixture of n-butyl-2-cyanoacrylate and ethiodized oil as an embolizing agent. RESULTS: In all patients, embolization of the hepatic artery was technically feasible and complete occlusion could be obtained. In two patients, collaterals originating from the right inferior phrenic artery were embolized superselectively 3 months after bilobar embolization. CTAP at that time revealed marked decrease in tumor size of more than 50%. All patients responded to the treatment as confirmed by normalization of measurable hormone levels, glucose levels, and disappearance of symptoms. Two patients are still alive after 24 and 31 months from the time of the first embolization. Current investigations revealed normal laboratory data and no further tumor progression in the liver. The third patient died 15 months after the first embolization; she also had developed ileus due to local recurrence of the primary tumor and lymph node metastases. CONCLUSION: Hepatic arterial embolization appears to be an effective means of palliation for liver metastases of malignant insulinoma. Long-term improvement seems most likely to be the result of extensive ischemia from permanent occlusion.