SS24 Cryopreserved Vein vs Autologous Vein in Portomesenteric Reconstruction During Oncologic Surgery

Abstract

In 2016, pancreatic cancer became the third leading cause of cancer-related death in the United States. While 5-year survival rates remain in the single digits, surgical resection offers the only potential for cure and can increase survival tenfold. As such, patients with portomesenteric vein involvement benefit from surgical resection with portomesenteric vein reconstruction (PVR). Studies demonstrate the feasibility of PVR with femoral and/or saphenous vein conduits; however, femoral vein (FV) harvest is associated with increased risk of deep vein thrombosis (DVT) and wound complications. Cryopreserved FV may be a reasonable alternative avoiding such morbidity. We sought to compare outcomes of autogenous vs cryopreserved FV conduit in patients undergoing PVR during pancreaticoduodenectomy. This was a retrospective review of our National Surgical Quality Improvement Program database for all patients undergoing pancreaticoduodenectomy with PVR from January 2010 to July 2016. Patient demographics (age, sex, and comorbidities) and tumor stage and pathology records were assessed. Postoperative complications, conduit patency, and perioperative mortality were compared between autogenous and cryopreserved vein groups. A total of 48 patients, 26 autogenous and 22 cryopreserved vein, underwent PVR. Demographics were similar between groups, although the autogenous vein group consisted of significantly more patients with diabetes (48% vs 14%; P = .015). Median follow-up was 6.7 months. Median survival was 24 months between both groups. Tumor stage did not differ significantly between groups. Conduit patency (Fig 1) and overall survival (Fig 2) did not differ between groups. While surgical site infection rates did not differ between the groups (7% v 0%; P = .493), there was a significantly higher rate of perioperative DVT in the autogenous vein group (57% v 14%; P = .002). These DVTs were notably more likely to be proximal to the site of FV harvest. There were no conduit infections in either group. Univariate analysis demonstrated autogenous vein conduit and diabetes mellitus were independent predictors of perioperative DVT. Patients undergoing oncologic pancreatic resection with PVR do not have significantly different patency or survival rates regardless of the conduit used. Since autogenous vein harvest is associated with significantly higher rates of perioperative DVT, cryopreserved vein should be considered for all PVR in pancreatic resection as it avoids the morbidity associated with autogenous vein harvest.Table IComparison of characteristics and outcomes in two types of conduitVariableAutologousa (n = 26)CryoVeina (n = 22)P valuebAge (years)64.9 ± 8.263.0 ± 14.8.583Gender.398 Male15 (57.7)10 (45.5) Female11 (42.3)12 (54.5)Diabetes12 (48.2)3 (13.6).015Hypertension11 (42.3)13 (59.1).247Obesity7 (26.9)4 (18.2).473Tobacco use.849 Never smoked11 (42.3)10 (45.5) Former smoker11 (42.3)10 (45.5) Current smoker4 (15.4)2 (9.0)Kidney failure2 (7.7)1 (4.5)1.000Pulmonary disease8 (30.8)5 (22.7).532Tumor stage.763 IB01 (4.8) IIA6 (23.1)5 (23.8) IIB19 (73.1)13 (61.9) III1 (3.8)1 (4.8) IV01 (4.8)Preoperative use Antiplatelet10 (38.5)2 (9.1).019 AnticoagulationTherapeutic02 (9.1).166Prophylactic24 (92.3)20 (90.9)Postoperative use Antiplatelet17 (65.4)20 (90.9).036 AnticoagulationTherapeutic8 (30.8)2 (9.1).084Deep vein thrombosis <30 days15 (57.7)3 (13.6).002 >30 days5 (19.2)1 (4.5).199Site of deep vein thrombosis<.0001 Proximal8 (30.8)0 Distal10 (38.5)0 No deep vein thrombosis8 (30.7)22 (100)Absence of wound infection Midline20 (76.9)18 (81.8).735 Leg wound <30 days24 (92.3)22 (100).493Hospital length of hospital (days)22.5 ± 16.816.6 ± 13.7.200Survival at last follow-up7 (26.9)12 (54.6).051aContinuous variables are summarized as mean ± standard deviation, and categoric variables are summarized as frequency (%).bContinuous variables are compared using Student t-test, and frequencies are compared using χ2 or Fisher exact test as appropriate. Bold values are statistically significant (P < .05). Open table in a new tab Table IIEvaluation of risk factors for prediction of conduit occlusionFactorUnivariate modelHR (95% CI)P valueaAge (years)1.04 (0.94-1.14).438Gender (female)0.78 (0.19-3.12).722Diabetes0.87 (0.21-3.65).846Hypertension0.33 (0.08-1.40).132Body mass index (kg/m2)0.97 (0.84-1.12).679Smoking Former smoker2.15 (0.41-11.18).364 Current smoker2.30 (0.21-25.52).498Kidney injuryNA1.000Pulmonary disease2.35 (0.58-9.57).233Preoperative Antiplatelet0.30 (0.04-2.44).261 AnticoagulationNA Statin0.80 (0.16-4.00).783Postoperative Antiplatelet0.88 (0.11-7.22).905 AnticoagulationNACryoVein conduit2.21 (0.52-9.33).280CI, Confidence interval; HR, hazard ratio.aCalculations are done using Cox proportional hazard model. Open table in a new tab Table IIIEvaluation of risk factors for prediction of deep vein thrombosis after vein graftingVariableUnivariate modelMultivariate modelOR (95% CI)P valueaOR (95% CI)P valueAge (years)1.00 (0.95-1.05).945Gender (male)0.86 (0.27-2.67).790Diabetes5.50 (1.42-21.30).01414.83 (1.58-139.52).018Hypertension1.00 (0.32-3.11)1.000BMI (kg/m2)1.00 (0.89-1.11).940Smoking Former smoker0.55 (0.16-1.91).3480.45 (0.08-2.65).380 Current smoker5.50 (0.54-55.49).14813.70 (0.61-309.76).100Kidney injury0.57 (0.05-6.76).657Pulmonary disease1.56 (0.43-5.59).499Preoperative Antiplatelet3.14 (0.79-12.42).1030.64 (0.06-6.50).709 Statin1.23 (0.31-4.98).767Postoperative Antiplatelet0.23 (0.05-1.01).0510.22 (0.02-1.98).176 Anticoagulationb3.58 (0.80-16.05).0963.43 (0.31-38.14).315CryoVein conduit0.10 (0.02-0.39).0010.18 (0.03-1.25).083CI, Confidence interval; HR, hazard ratio.aCalculations are done using logistic regression model. Bold values are statistically significant (P < . 05).bTherapeutic vs prophylactic. Open table in a new tab Fig 2Patient survival.View Large Image Figure ViewerDownload Hi-res image Download (PPT)