Portal Vein Reconstruction With Interposition Cryopreserved Descending Thoracic Aortic Homograft

Abstract

Improvements in chemoradiation have rendered more complex pancreatic cancers involving the portomesenteric venous axis amenable to surgical resection. Portal vein reconstruction (PVR) has thus become an essential component of the overall operative approach. For significant defects requiring segmental replacement, various conduits have been proposed, but the optimal repair remains unknown. Here, we present short-term results of the largest series to date using interposition cryopreserved cadaveric descending thoracic aortic homograft (CCDTAH) for PVR. Patient health information was harvested retrospectively from a prospectively maintained database. Data were summarized using standard statistical techniques. Fourteen patients, six men and eight women, underwent PVR with CCDTAH from 2014 to 2020. Average age was 63 years. All patients had pancreatic cancer (93% adenocarcinoma, 7% neuroendocrine tumor) and underwent either a Whipple procedure (71%) or pancreatectomy and splenectomy (29%). Mean follow-up was 6.6 months (range, 1.0-27.1 months). Seven patients (50%) developed early recurrent oncologic disease (EROD). Five patients (36%) died during the follow-up period, four from EROD, with a mean survival of 4.8 months (range, 1.4-10 months). PVR technical success was 100%, and there were no intraoperative complications. Major perioperative complications included a pancreatic leak, which was not associated with graft pseudoaneurysm, and a chyle leak requiring percutaneous drainage. The graft remained patent in the former patient for 2 months before occluding in the setting of EROD, whereas the graft in the latter patient has remained patent through follow-up out to 8 months. All grafts were patent at discharge. Three grafts thrombosed during postdischarge follow-up, all of which occurred in the setting of EROD (two of these patients died at 3 and 7 months, respectively). Five other patients developed EROD, two of whom died during follow-up, without evidence of graft thrombosis. The 30-day, 3-month, and 9-month primary patency rates were 100%, 86%, and 71%, respectively. Pancreatic cancer remains a highly morbid disease despite surgical resection with curative intent, and early recurrence is the main driver of PVR thrombosis. In the absence of EROD, interposition bypass with CCDTAH for PVR is associated with acceptable patency rates, at least in the short term. When taken together with its optimal size, rapid availability, excellent technical success, favorable risk profile, and avoidance of harvest site complications, it is a reasonable choice for conduit in this clinical setting. Longer term follow-up is required to ensure continued durability.