Bariatric gastric artery embolization can modulate “hunger hormone” levels

Charles Kim, Clifford Weiss and Aravind Arepally
Charles Kim, Clifford Weiss and Aravind Arepally

An animal study using a porcine model has identified some of the sequelae of gastric artery embolization for systemic suppression of the hunger hormone, ghrelin. Ghrelin levels may be reduced by embolizing its production centres in the gastric fundus; however, incomplete fundal embolization and concerns for non-target embolization remain obstacles to the success of the procedure, say researchers.

“Bariatric gastric artery embolization can significantly suppress systemic levels of the appetite hormone, ghrelin, and impact weight gain; complications were seen in 33% of treated animals in the form of gastric ulceration, possibly due to non-target embolization. Gastroprotective agents may provide a role in further refining this technique,” said Charles Kim, assistant professor of Radiology, Division of Vascular and Interventional Radiology, Duke University Medical Center, Durham, USA.


Ghrelin is the most potent appetite stimulant and 90% is produced in the stomach fundus. The level of the hormone rises before a meal, falls after eating and increases with dieting, said the researchers.

“The procedure was pioneered by interventional radiologist, Aravind Arepally, based on the concept that interventional radiologists can destroy cells by blocking the blood supply. Most ghrelin secreting cells are in the fundus and starving this area of its blood supply, decreases ghrelin levels, which decreases hunger. The procedure is performed through a needle hole and there was no evidence of dangerous stomach necrosis. FDA-cleared devices are used in the procedures,” Kim said.

Kim and colleagues performed this study in healthy growing swine (weight range, 30.3–47kg; n=12). They performed targeted gastric artery embolization in six swine with the infusion of 40μ calibrated particles (4–6mLs) selectively into the gastric arteries that supply the fundus. Six control animals underwent a sham procedure with 5mL of saline.

“Weight and fasting plasma ghrelin levels were obtained in animals at baseline and at weeks one to eight. Endoscopy was performed on nine of 12 animals at three weeks postprocedure. Gross pathologic examination of the stomach mucosa was performed post-necropsy in all animals.

Contiguous tissue sections from 11 stomachs (fundus, body, and antrum) were sent for histopathologic analysis. Statistical testing was carried out using the Wilcoxon rank-sum test,” said Kim.


The investigators found that change in ghrelin levels over time was statistically significantly between different groups. “In treated animals, ghrelin levels were significantly suppressed at weeks one to six and week eight relative to baseline mean (p<0.05). Treated animals weighed significantly less at week one and weeks three to eight (p<0.05).

“Three week endoscopic evaluation of treatment animals demonstrated ulcers in 2/5 (40%), a distal oesophageal stricture in one animal (20%), and mild-moderate gastritis in five animals (100%). Control animals demonstrated mild gastritis in 2/4 animals (50%). Eight-week histopathologic evaluation of treatment animals demonstrated ulcers in 2/6 (33%) and mild to moderate gastritis in six animals (100%). Control animals demonstrated mild to moderate gastritis in five animals (10%). These results correlated well with prior endoscopic findings.

There is a need to refine the technique and adjunct medications before beginning trials in humans. The procedure is a potential adjunct or alternative to surgery,” said Kim.

Increasing the accuracy of the procedure

A separate study employed new X-ray visible embolic beads (XEBs) and C-arm cone beam computed tomography to increase the accuracy of gastric artery embolization. This included minimising incomplete fundal embolization and addressing the concerns of non-target embolization.

Investigators, Clifford R Weiss, The Johns Hopkins University School of Medicine, Baltimore, USA, and colleagues used custom-made, barium sulphate-containing, highly uniform, alginate, X-ray visible beads (~50μm) as the embolic agent in this study. Arteries supplying the gastric fundus in two swine were determined using cone-beam CT (dynaCT from Siemens Healthcare).

The researchers used a femoral approach, to select the coeliac axis with a 5F Mickelson catheter (from Cook). After angiogram, each of the 3–4 fundal arteries was selected using a Renegade Hi-Flow catheter (from Boston Scientific) and a 0.016 Fathom wire (from Stryker).

“At each location a pre-embolization angiogram and cone-beam CT (25–50% contrast, 1cc/sec for 10 seconds) were performed, the artery was embolized under direct fluoroscopic visualisation, and a non-contrast post embolization cone-beam CT was completed. Prior to sacrifice, a black tissue dye was injected from the coeliac artery to confirm post-embolic circulation,” Weiss said.

The investigators mapped the vasculature and selectively embolized the gastric tree of two swine. “The beads were visible during both fluoroscopic delivery and cone-beam CT in all studies, and the additional 3D information from cone-beam CT allowed efficient gastric arterial tree mapping and targeting. Reflux and non-target injection was visualised at time of delivery, and embolic coverage was confirmed intraoperatively via cone-beam CT. “The combination of X-ray visible beads and cone-beam CT facilitated accurate mapping of the gastric fundus and enhanced safety by reducing the risk of back flow, thus preventing non-target tissue embolization. It allows the interventional radiologists to better see where they are going, and see where they have been. It allows for complete fundal embolization and a better assessment of outcomes. It may allow interventional radiologists to determine if re-embolization is needed. Further study will be necessary to evaluate gastric artery embolization as a therapy for morbid obesity,” said Weiss.