Background and study aims: Open or laparoscopic gastrojejunostomy is an established treatment for malignant
duodenal obstruction but may be associated with significant morbidity and mortality.
The purpose of this study was to develop a model for an entirely endoscopic gastrojejunostomy
to treat duodenal obstruction, and to compare this with the laparoscopic technique.
Methods: During the first part of the study the endoscopic technique was developed and tested
in porcine nonsurvival and survival experiments (n = 12). During the second part of
the study (n = 10), endoscopic gastrojejunostomy for duodenal occlusion was compared
with laparoscopic gastrojejunostomy in a survival randomized controlled trial (RCT).
For both groups duodenal occlusion was achieved by the laparoscopic approach.
Results: In the RCT, the median times for laparoscopic vs. endoscopic gastrojejunostomy were
70 minutes (interquartile range [IQR] 65 – 75) vs. 210 minutes (IQR 197 – 220; P = 0.01). There was a trend toward increased anastomotic diameter at necropsy in the
laparoscopic group (2 cm, IQR 2 – 3) compared to the endoscopic group (1.8 cm, IQR
1.6 – 1.8; P = 0.06). One animal in the endoscopic group died secondarily to bowel ischemia from
volvulus of the jejunal loop. One animal in the laparoscopic group was prematurely
sacrificed due to extensive pulmonary congestion and edema. All anastomoses were intact
and patent.
Conclusions: Purely endoscopic gastrojejunostomy using the developed technique and devices is
feasible and can result in adequate relief of duodenal obstruction. Endoscopic anastomoses
tend to be smaller than laparoscopic anastomoses, with the procedures being more time-consuming
and associated with higher complication rates.
