C. Pera et al.: Retroperitoneal Approach for Splenorenal Shunt
Invited Commentary W. Dean Warren, M.D. Emory University School of Medicine, Atlanta, Georgia, U.S.A.
We congratulate Drs. Pera, Visa, Rodes and Teres on the early results of the retroperitoneal approach for modified distal splenorenal shunt. Our experience with the retroperitoneal approach has been limited; however, we strongly recommend that this approach be utilized only under those conditions which make transabdominal operation extremely difficult. The 3-year follow-up of the prospective randomized study comparing selective and total decompression procedures has clearly shown that unless aggressive devascularization of the coronary venous system is accomplished, rapid recollateralization to the splenic venous system will occur, causing significant decrease in portal perfusion. Maintenance of portal perfusion to the liver is the sine qua non for prevention of encephalopathy. It is apparent from the authors' experience that the pitfalls of surgical technique are similar to those in the transabdominal approach. Specifically, the experience documents that the most difficult part of the dissection is the freeing of the splenic vein from the pancreas. Although we agree that the occlusion of the splenic vein can increase the pressure in the
retroperitoneal collaterals, it is usually easier to divide the splenic vein and dissect right-to-left to free the splenic vein from the pancreatic branches. The exposure of the pancreatic branches is greatly facilitated by this maneuver. In the retroperitoneal approach, it would seem reasonable to dissect the renal vein first and then transect the splenic vein, if desired. The authors state that they routinely divide the adrenal and gonadal vessel to improve mobility and exposure. If the left renal vein has a functional obstruction by either the aorta or retroperitoneal edema, the gonadal vein becomes a major outflow tract for renal venous flow. If the gonadal vein is divided, acute renal venous hypertension can occur. For this reason, if the preoperative left renal venogram demonstrates caudal flow in the gonadal vein, the gonadal vein is not ligated. We do, however, routinely divide the adrenal vein. We agree that marked care must be taken in approximating the veins for the anastomosis. Our experience has been that when the splenic vein is divided at its juncture with the superior mesenteric vein and sufficient dissection carried out toward the hilum of the spleen for proper positioning, it is often necessary to resect a portion of the splenic vein to prevent redundancy and kinking. We also utilize an oblique anastomosis with a continuous running suture on the posterior row and interrupted sutures on the anterior row.