Abstract
Background
Pancreatic surgery with vascular reconstruction is increasingly performed to offer the benefits of surgical resection to patients with locally advanced disease. The short- and long-term patency rates and the clinical significance of thrombosis of such reconstructions are unknown.
Methods
We reviewed pancreatectomies requiring venous reconstruction from 1994 to 2011. We sought to identify predictors of acute (within 30 days) and late thrombosis. We compared survival of patients with thrombosis to patients with patent reconstructions.
Results
Of 203 pancreatectomies requiring venous reconstruction, acute thrombosis occurred in nine (4.4 %) cases and was associated with increased perioperative mortality (22.2 versus 4.6 %, p = 0.023). Even when nonfatal, acute thrombosis was associated with decreased median survival (7.1 versus 15.9 months, p = 0.011) and increased hazard of death (hazard ratio 8.6, confidence interval 3.7–19.9, p < 0.001). A late loss of patency was seen in 31.2 % of cases at a median of 9.5 months. Later loss of patency was not associated with decreased median survival or increased hazard of death.
Conclusions
Acute thrombosis of the portal venous reconstructions after pancreatectomy is associated with increased perioperative mortality and, even when nonfatal, is associated with decreased survival. Late loss of patency occurs in one-third of patients but does not affect survival.
This is a preview of subscription content, log in via an institution to check access.
Similar content being viewed by others
References
Allendorf, J.D., et al., Neoadjuvant chemotherapy and radiation for patients with locally unresectable pancreatic adenocarcinoma: feasibility, efficacy, and survival. J Gastrointest Surg, 2008. 12(1): p. 91–100.
Tseng J.F. et al. Pancreaticoduodenectomy with vascular resection: margin status and survival duration. J Gastrointest Surg, 2004. 8(8): p. 935–49; discussion 949–50.
Toomey, P., et al., Resection of portovenous structures to obtain microscopically negative margins during pancreaticoduodenectomy for pancreatic adenocarcinoma is worthwhile. Am Surg, 2009. 75(9): p. 804–9; discussion 809–10.
Smoot, R.L., J.D. Christein, and M.B. Farnell, Durability of portal venous reconstruction following resection during pancreaticoduodenectomy. J Gastrointest Surg, 2006. 10(10): p. 1371–5.
Castleberry, A.W., et al., The Impact of Vascular Resection on Early Postoperative Outcomes after Pancreaticoduodenectomy: An Analysis of the American College of Surgeons National Surgical Quality Improvement Program Database. Ann Surg Oncol, 2012. 19(13): p. 4068–77
Clark, W., et al., Targeting early deaths following pancreaticoduodenectomy to improve survival. J Gastrointest Surg, 2012. 16(10): p. 1869–74.
Merkow, R.P., et al., Post-discharge venous thromboembolism after cancer surgery: extending the case for extended prophylaxis. Ann Surg, 2011. 254(1): p. 131–7.
Katz, M.H., et al., Retroperitoneal dissection in patients with borderline resectable pancreatic cancer: operative principles and techniques. J Am Coll Surg, 2012. 215(2): p. e11-8.
Conflict of Interest
The authors have no conflicts of interest to disclose.
Author information
Authors and Affiliations
Corresponding author
Additional information
Discussion
Dr. Matthew Katz (Houston, Texas): Dr. Gawlas, Dr. Allendorf, and colleagues have put together a very nice manuscript and should be congratulated for both this work and, more importantly, their care of over 200 patients who required complex surgical operations to treat their pancreatic disease. Their data add to an increasing body of literature that clearly demonstrates that, while technically demanding, venous resection and reconstruction can and probably should be performed at the time of pancreatic resection in well-selected patients with appropriate tumor anatomy and disease biology. The study is the first I am aware of that has attempted to define predictors of both short- and long-term occlusive events following pancreatectomy with vascular resection and, for this reason alone, is worthy of significant attention among those of us who do these procedures routinely.
I have the following questions:
1. In my experience, veins go down for two primary reasons. The first is pancreatic fistula. What was the incidence of leak overall in your series of patients and was the incidence in the acute thrombosis group different from the no thrombosis group? Do you think it might be important to include this in your multivariate analysis? The second reason veins go down is cancer recurrence. You talk a little bit about the association between cancer and late thrombosis in the manuscript. Although malignancy was not a factor associated with late thrombosis in your model, the number of patients with benign disease was small and the confidence interval relating to that association was wide. Do you have any data relating the timing of late thrombosis with that of cancer recurrence?
2. I am somewhat surprised at the minimal use of antithrombotic medication used at your institution. Our group administers 5,000 U sq heparin upon induction to all patients who undergo pancreatectomy and start Lovenox immediately postoperatively. For patients who undergo vein resection, we administer IV heparin prior to clamping the vein and give rectal aspirin in addition to Lovenox following surgery. Clearly, there is no standard approach for the use of these drugs, even among high-volume institutions. In light of the data you have presented, does your group have any recommendations with regard to the administration of antithrombotic medications to patients who undergo pancreatectomy?
3. How are patients selected for these complicated operations at your institution? Does preoperative therapy play a role?
This really was a nice manuscript, and I am thankful to the SSAT, Dr. Gawlas, and Dr. Allendorf for the opportunity to review it.
Closing Discussant
Ms. Irmina Gawlas: Dr. Katz, thank you very much for your comments and your excellent questions. We appreciate your insights, especially in light of your extensive experience with these complex procedures. I will answer your questions to the best of my ability:
1. In our series, the overall pancreatic fistula rate is around 8 %. We, too, have observed that leak and vein graft failure sometimes go hand in hand, but because we sought to examine pre- and intraoperative contributors to thrombosis, we did not include this in our multivariable model. With regard to cancer recurrence, a majority (around 60 %) of the late thromboses were identified on imaging that concurrently noted a local tumor recurrence. It is certainly true that the uneven distribution of benign versus malignant disease may have precluded us from identifying malignancy as a predictor of late thrombosis.
2. It is certainly true that the use of antithrombotic medication during and after such procedures varies widely. We, too, have been increasingly giving IV heparin prior to clamping of the vein, but this is not standardized across all surgeons. Even so, we remain concerned about bleeding (and the highly associated mortality) in the immediate postoperative period. Thus, we are reticent to make recommendations about the routine use of antithrombotics, and we believe that more data are needed to identify an ideal protocol. However, there seems to be mounting evidence that the incidence and morbidity of thromboembolic events in surgical oncology in general and in hepatopancreaticobiliary oncologic surgery in particular are quite high. Thus, it is certainly possible that as more data emerge, it will justify more routine use of antithrombotics.
3. Our patient selection algorithm has evolved over the observation period. Over the past 10 years, we have been increasingly routinely referring patients with radiologic evidence of locally advanced disease and biopsy-proven pancreatic ductal adenocarcinoma for neoadjuvant chemotherapy. These patients are restaged after treatment, and resection is attempted if there is no progression of disease. Alternatively, some patients with borderline resectable tumors with venous involvement underwent resection without preoperative therapy. In some cases, this was because a pathologically confirmed diagnosis of adenocarcinoma could not be obtained or because of medical, psychosocial, or insurance reasons. In the study group, around 60 % of the patients did undergo neoadjuvant chemotherapy.
This work was presented as a plenary session on May 19, 2013 at the 54th Annual Meeting of the SSAT and on May 17, 2013 as an oral presentation at the 38th Annual SSAT Residents and Fellows Research Competition in Orlando, FL.
Rights and permissions
About this article
Cite this article
Gawlas, I., Epelboym, I., Winner, M. et al. Short-Term but Not Long-Term Loss of Patency of Venous Reconstruction During Pancreatic Resection Is Associated with Decreased Survival. J Gastrointest Surg 18, 75–82 (2014). https://doi.org/10.1007/s11605-013-2375-2
Received:
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s11605-013-2375-2