Y-90 with Resin Microspheres without Routine Embolization of the Gastroduodenal Artery

Summary

Ward et al present a retrospective analysis of gastrointestinal complications in patients following radioembolization with resin microspheres without routine embolization of the gastroduodenal artery. In this study, 62 consecutive patients underwent Yttrium-90 (Y90) radioembolization with resin microspheres for cholangiocarcinoma or metastatic liver disease. Prior to treatment, preparatory angiography was performed with embolization of high-risk hepaticoenteric anastomoses and administration of technetium-99m macroaggregated albumin. On the day of treatment, repeat digital subtraction angiography and cone-beam computed tomography (CT) was utilized to identify and embolize additional hepaticoenteric anastomoses prior to resin microsphere administration. In addition, if the treatment vessel was less than ~2 cm distal to a hepaticoenteric anastomotic vessel that could not receive coil embolization, administration of RE was done via an antireflux microcatheter (ARMC) (Surefire Infusion System). Patients were discharged with a 1-month prescription for a proton pump inhibitor and a 10-day steroid taper. Follow-up was scheduled for 1 and 3 months post-treatment with an examination focused to detect symptoms of gastrointestinal complications. During planning angiography, 68% of patients had one or more hepaticoenteric vessels embolized, with the right gastric artery being the most commonly embolized. The gastroduodenal artery was embolized in only three patients, two during planning angiography and one immediately prior to radioembolization. ARMC was used in 14% (10 of 69) treatments. Stasis of flow was identified in three patients and administration of resin microspheres was terminated in these patients prior to complete administration of the prescribed dose. Clinical follow-up was available in 60 out of 62 patients and no patients experienced Grade 3 or 4 gastrointestinal symptoms (hematemesis, melena, severe nausea, vomiting, or abdominal pain). No patients required upper endoscopy or extension of prophylactic proton pump inhibitor.



Figure 2. (a) Celiac arteriography in a 48-year-old woman with metastatic neuroendocrine tumor who required embolization of the GDA demonstrated an accessory left hepatic artery (black arrow) from the left gastric artery, a right hepatic artery (white arrow), and a middle hepatic artery supplying segment 4 (dashed white arrow). The GDA (dashed black arrow) originated from the right hepatic artery distal to the origin of the middle hepatic artery. The middle hepatic artery demonstrated a tight kink at its origin, precluding use of an antireflux microcatheter, but supplied approximately 400 mL of diffuse tumor in segment 4. Redistribution of the middle hepatic artery territory by coil embolization was rejected because of irregular intrahepatic vessels occluded by tumors that may have jeopardized successful redistribution. (b) Plug embolization of the GDA (black arrow) and coil embolization of the right gastric artery (white arrow) were performed before microsphere administration in 3 separate doses, administered via the replaced left hepatic artery, the middle hepatic artery, and the right hepatic artery.

Commentary

Routine embolization of the gastroduodenal artery has previously been recommended in the context of Y90 radioembolization with resin microspheres to prevent gastrointestinal ulceration. This action has potential risks; however, including coil migration, recanalization, increased procedure costs, and development of hepaticoenteric collateral vessels prior to treatment. In addition, the overall risk of gastrointestinal ulceration following radioembolization is relatively low, with only 3.7% of patients demonstrating symptoms in the randomized multi-center SIRFLOX trial. Ward et al demonstrate that routine embolization of the gastroduodenal artery may not be necessary given developments that increase the safety of resin microsphere administration (eg. cone-beam CT and ARMC). In addition, 5% dextrose was utilized instead of sterile water for suspension and flushing of resin microspheres during treatment in order to decrease vasospasm and vascular stasis. Microspheres were also administered a day before official calibration so the activity per sphere was 65 Gy/sphere instead of 50 Gy/sphere. This allowed for administration of approximately 23% fewer microspheres and decreased the risk of vascular stasis. With these precautions, there were no observed gastrointestinal complications at follow-up for any of the patients in the study. This study was limited as a retrospective single center study with all procedures performed by two operators with significant experience in resin microsphere radioembolization, potentially limiting the generalizability of the results. In addition, patients did not undergo routine upper endoscopy and biopsy, so the authors were unable to rule out subclinical gastrointestinal ulceration. The results; however, suggest that leaving the gastroduodenal artery intact prior to radioembolization with resin microspheres is safe and has the potential to save procedural time and costs.

Click here for abstract

Ward TJ, Louie JD, Sze DY. Yttrium-90 Radioembolization with Resin Microspheres without Routine Embolization of the Gastroduodenal Artery. J Vasc Interv Radiol. 2017 Feb;28(2):246-253.

Post Authors:
Jeffrey Forris Beecham Chick, MD, MPH, DABR
Assistant Professor of Vascular and Interventional Radiology
Vice Quality Assurance and Safety Officer
Venous Health Program Faculty
University of Michigan Health System
Michigan Medicine

Thomas J. An, BA
Vanderbilt University School of Medicine