Reconstruction of an Occluded Portal Vein During Pancreatic Resection

Abstract

Background: Pancreatic Ductal Adenocarcinoma (PDAC) is one of the most common malignancies associated with thrombotic events. While there is research present on various techniques of portal vein reconstruction, there is limited published data on the techniques and/or considerations of reconstruction in the setting of complete portal vein occlusion. We therefore sought to analyze and present our experience of this clinical scenario. Methods: This was a retrospective analysis of a prospectively collected database. All patients who underwent portal vein resection and/or reconstruction during a pancreatic resection were included. Post-operatively, all patients underwent a contrast-enhanced CT scan on post-operative day 1 to assess for any portal vein thrombus. Results: Pancreatic resection with portal vein reconstruction was performed in 183 patients. Complete PV occlusion was seen in 12 patients at the time of surgery. In those patients with an occluded PV, reconstruction options included primary repair with end-end anastomosis (n = 2) or use of an interposition graft (n = 9). Interposition graft conduits included the left renal vein (n = 6), tubularized bovine pericardium (n = 3), and femoral vein (n = 1). Post-operative portal vein thrombus was seen in 4/12 patients. The majority of patients (n = 7) were discharged on therapeutic anticoagulation, 4 were discharged on an antiplatelet, and 1 patient received neither. Conclusions: Based on our series, we would recommend attempting PV reconstruction in these patients with an interposition graft (with autologous left renal vein or bovine pericardium). We believe that with this technique, the post-operative thrombosis risk is similar to PV reconstructions in non-occluded patients.

1. Introduction

Pancreatic Ductal Adenocarcinoma (PDAC) is one of the most common malignancies associated with thrombotic events [1]. The rate of visceral thrombosis has been reported as greater than 20%, with portal vein thrombosis (PVT) reported as the most common subtype [2,3]. PVT has been classified into four categories: Grade 1: <50% PVT; Grade 2: >50% PVT; Grade 3: complete portal vein (PV) and proximal Superior Mesenteric Vein (SMV) thrombosis; and Grade 4: complete PV and entire SMV thrombosis [4].

The degree of PVT impacts resectability during pancreatic surgery. The international consensus defines the presence of PVT as borderline resectable when there is a “suitable vessel proximal and distal to the site of involvement allowing for safe and complete resection and vein reconstruction”. It is defined as locally advanced when the PVT is “unreconstructible due to tumor involvement/occlusion” [5]. The application of this definition is highly dependent on surgeon experience with portal vein (PV) reconstruction, especially in the setting of a completely occluded portal vein.

While there is research present on various techniques of portal vein reconstruction, there is limited published data on the techniques and/or considerations of reconstruction in the setting of complete portal vein occlusion. This can lead to discordant decision making within the specialty when faced with this situation.

Our institutional preference is to reconstruct the resected portal vein when encountered with this situation. We therefore sought to analyze and present our experience of this clinical scenario. We hope to determine whether resection and reconstruction of an occluded portal vein is a safe and effective technique.

2. Methods

This was a retrospective analysis of a prospectively collected database at a single institution from 1 January 2017 to 15 December 2023. All patients who underwent portal vein resection and/or reconstruction during a pancreatic resection were included in the database. Patients with borderline resectable or locally advanced disease would undergo neoadjuvant therapy (generally three months but at the discretion of the medical oncologist). Patients with arterial involvement would be considered for additional neoadjuvant chemoradiotherapy prior to any surgical intervention.

Complications were collected prospectively, both through chart review and bi-weekly ward rounds by an independent data collector. The data collector is a non-physician researcher who would round with the inpatient surgical team to collect the complication data, which was inputted into a secure database. Clarifications were provided in real-time if there was any uncertainty as to the complication. Post-operative thrombosis was classified as thrombus in either the SMV and/or the portal vein. If there was doubt as to whether a PV thrombus was present and the patient was started on treatment, that was classified that as PVT. Severe complications were classified as Clavien Dindo Grade III+.

All operative notes were reviewed to confirm portal vein resection was performed and the method of reconstruction. This could be performed as a venous resection alone or in combination with arterial resection. Patients were excluded if portal vein resection was part of an ex vivo resection, or if vascular resection of the IVC or aorta was performed.

Portal vein occlusion was determined at the most recent CT scan prior to operative intervention. This was determined by the radiologist report and independent verification by a study author. Post-operatively, all patients underwent a contrast-enhanced CT scan on post-operative day 1 to assess for any portal vein thrombus.

3. Results

Pancreatic resection with portal vein reconstruction was performed in 183 patients in the database. Reconstruction options were as follows: end-end anastomosis (n = 103), interposition graft (n = 55), venorrhaphy/primary closure (n = 19), and patch repair (n = 7). Interposition graft conduits included the left renal vein (n = 42), femoral vein (n = 2), splenic vein (n = 1), bovine pericardium (n = 7), homograft (n = 2), and PTFE (n = 1). Post-operative PV thrombus was seen in 14 (7.7%) patients. Severe complications were seen in 31 (16.9%) patients. There were five deaths within 90 days of the index operation.

Complete PV occlusion was seen in 12 patients at the time of surgery (

Table 1. Characteristics of included patients.

Age (Sex)	Operation	Pathology	Type of PV Reconstruction	Additional Vascular Reconstruction	Antiplatelet or Anticoagulation on Discharge	Post-Operative Portal Vein Thrombus	Other Complications (Overall Clavien Dindo Grade)
54 (M)	Whipple	PDAC	Interposition Graft (Left Renal Vein)	SMA	Antiplatelet	No	No
57 (M)	Whipple	PDAC	Interposition Graft (Bovine Pericardium)	No	Antiplatelet	No	No
36 (M)	Whipple	PNET	Interposition Graft (Femoral Vein)	SMA	Therapeutic Anticoagulation	No	Left Vocal Cord Paralysis (I)
59 (F)	Whipple	PDAC	Interposition Graft (Left Renal Vein)	SMA	Therapeutic Anticoagulation	Yes—POD 6 occlusive thrombus, thrombectomy performed with non-occlusive re-thrombosis	Bleeding requiring takeback to the OR with temporary abdominal closure and admission to the ICU (IV)
63 (F)	Whipple	PDAC	Interposition Graft (Left Renal Vein)	No	Antiplatelet	No	No
72 (F)	Whipple	PDAC	Interposition Graft (Left Renal Vein)	CHA	Therapeutic Anticoagulation (prescribed prior to surgery)	No	Pneumonia (II)
60 (F)	Total Pancreatectomy	PDAC	Interposition Graft (Bovine Pericardium)	CHA	Therapeutic Anticoagulation (for other medical co-morbidities)	No	Bleeding requiring takeback to the OR (III)
24 (F)	Whipple	PNET	Interposition Graft (Left Renal Vein)	No	Therapeutic Anticoagulation	No	Intra-abdominal fluid collection requiring IR drain (III)
60 (F)	Whipple	PNET	End-End Anastomosis	No	Therapeutic Anticoagulation	Yes—POD 1 Thrombus	No (II)
60 (M)	Whipple	PDAC	End-End Anastomosis	No	Therapeutic Anticoagulation c/b bleeding so switched to Antiplatelet	Yes—POD 1 Thrombus, thrombectomy performed with re-thrombosis	No (III)
60 (F)	Total Pancreatectomy	PDAC	Interposition Graft (Left Renal Vein)	No	Therapeutic Anticoagulation	Yes—POD 11 thrombus	Ascites (II)
78 (M)	Total Pancreatectomy	PNET	Interposition Graft (Bovine Pericardium)	No	No	No	No

PDAC: Pancreatic Ductal Adenocarcinoma, PNET: Pancreatic Neuroendocrine Tumor, SMA: Superior Mesenteric Artery, CHA: Common Hepatic Artery, POD: post-operative day.

). The age range at the time of surgery was 24–78 years and seven of the patients were female. A pancreatoduodenectomy was performed in 9 patients and 3 patients underwent a total pancreatectomy. The disease pathologies were PDAC (n = 8) and PNET (n = 4). Portal vein reconstruction alone was performed in 7 patients, with 5 patients undergoing combined arterial and venous reconstruction (SMA, n = 3; CHA, n = 2). Severe complications were seen in three patients and there was no 90-day mortality in this subgroup.

In those patients with an occluded PV, reconstruction options included primary repair with end-end anastomosis (n = 2) or use of an interposition graft (n = 9). Interposition graft conduits included the left renal vein (n = 6), tubularized bovine pericardium (n = 3), and femoral vein (n = 1). Post-operative portal vein thrombus was seen in 4/12 patients, with two patients returning to the operating room for portal vein thrombectomy. Both patients who underwent a portal vein thrombectomy had evidence of re-thrombosis on re-imaging. Both patients who underwent an end-end anastomosis had evidence of a PVT on POD 1. Of the patients who underwent an interposition graft, one had a PVT detected on POD 6 and one had a PVT detected on POD 11. The majority of patients (n = 7) were discharged on therapeutic anticoagulation, 4 were discharged on an antiplatelet, and 1 patient received neither.

4. Discussion

The role of portal vein resection and reconstruction is not a new concept in the management of pancreatic malignancy. Some would argue that invasion of the portal system could be seen as a consequence of tumor location, rather than an absolute indicator of aggressive tumor biology [6]. Therefore, portal vein resection should be in the armamentarium of all complex hepatobiliary centers. However, some may classify complete portal vein occlusion as a contraindication to proceed with resection. In our series, we demonstrate that an occluded portal vein can be reconstructed with acceptable outcomes.

In the setting of PV occlusion, it can be debated whether there is any need for reconstruction at all. It can be argued that patients who have chronic, complete portal vein thrombosis usually have developed collateralization of their mesenteric blood flow prior to resection. Therefore, if these collaterals are sufficient for drainage pre-operatively, then they should be suitable post-operatively (based on the assumption that they will not be sacrificed and/or ligated during the operative exploration). Consequently, the flow through a reconstructed PV may be low due to the collateral circulation, therefore increasing the risk of post-operative thrombosis. In our cohort, PV thrombosis was seen in 5.8% of patients (10/173) who underwent reconstruction for tumor involvement and 33.3% (4/12) in those patients who underwent reconstruction with occluded PV. However, at our institution, the preference is to reconstruct these patients. Our concern is that some of these collaterals may need to be sacrificed during the resection. It is very difficult to assess whether the remaining collaterals would be sufficient and we think the safest approach is to re-establish normal anatomical portal flow.

Even though there is a marked difference, it can be argued that the method of reconstruction in this cohort may play a role. In the literature, an end-end anastomosis or venorrhaphy is usually preferred due to the lower reported rates of post-operative thrombosis, with one study reporting no post-operative PVT for patients undergoing this type of reconstruction [7,8,9]. In our patient group who underwent PV reconstruction in this manner for tumor involvement, a thrombosis rate of 5.0% (6 out of 121; End-end 5.0%, 5/101; venorrhaphy 5.2%, 1/19) was reported. However, both patients who underwent end-end anastomosis in the setting of occlusion had evidence of PVT on POD 1. All anastomoses were performed in the same fashion, which raises the question of whether this type of reconstruction has a higher thrombosis rate in the setting of portal vein occlusion.

In this series, when using an interposition graft, our thrombosis rate was 2/10 (20%). Whilst this is higher than the PVT rate for interposition grafts for tumor involvement in our cohort (7.1%, 3/42), it is comparable to the thrombosis rate with interposition grafts for portal vein reconstruction in the literature [7,8,9]. Therefore, we raise the argument that an attempt at PV reconstruction in the setting of occlusion may still be warranted, as the majority of reconstructions will remain patent.

There are also no definitive guidelines on the management of thrombosis of the reconstructed portal vein. In the management of index PVT, the use of anticoagulation therapy has been shown to have a favorable risk–benefit ratio [10]. Surgical thrombectomy can be attempted but has been associated with a high recurrence rate [11]. In both patients in this series that underwent surgical thrombectomy, re-thrombosis of the PV occurred. A recent systematic review demonstrated the heterogeneity in the management of post-operative portal vein thrombosis. It showed that whilst anticoagulation is safe, there is limited evidence that it improves clinical outcomes or primary patency [12]. In our cohort, patients that were discharged on post-operative anticoagulation, the duration of the therapy was at the discretion of the individual provider (surgeon, medical oncologist, or hematologist).

This series must be assessed within the context of its limitations. There are not enough patients to draw any statistical conclusions from this series. However, portal vein reconstruction is rare, and even rarer in the setting of complete occlusion. Our institutional preference is to reconstruct these patients and, therefore, there may be a bias as a consequence. Therefore, a suitably powered study would likely need to be an international effort. In addition, it is not possible to analyze whether there is a survival difference with such few patients when compared to non-occlusive portal vein resections. This is a retrospective analysis of a database focusing on short-term outcomes and therefore comes with its associated limitations. Aside from the POD 1 imaging, there was no consistent imaging to assess for PVT in the post-operative period. It is possible that small, asymptomatic PVT may have gone undiagnosed in this patient population.

5. Conclusions

We present our series of PV reconstructions during pancreatic surgery in the setting of a completely occluded PV. Based on our series, we would recommend attempting PV reconstruction in these patients with an interposition graft (with autologous left renal vein or bovine pericardium). We believe that with this technique, the post-operative thrombosis risk is similar to PV reconstructions in non-occluded patients. In the setting of post-operative thrombosis, we feel there is limited utility of surgical thrombectomy and systemic anticoagulation should likely be considered.