Home parenteral nutrition‐associated thromboembolic and bleeding events: results of a cohort study of 236 individuals

Abstract

Sparse or outdated studies focus on thrombotic and bleeding risk in home parenteral nutrition (HPN). 236 HPN patients followed at a single center for a total of 684 patient‐years were evaluated. Rates of venous thrombosis and major bleeding, and prevalence of vena cava syndrome are provided. Anticoagulants might reduce thrombosis risk, but population‐specific safety concerns remain. Summary BackgroundHome parenteral nutrition (HPN) is necessary for patients with intestinal failure. Recurrent catheter‐related thrombosis (CRT) is common, leading to infectious complications, pulmonary embolism, vascular access loss and intestinal transplantation. The efficacy and safety of anticoagulants are unknown in this setting and based on sparse and low‐quality observational data. ObjectivesOur aim was to estimate the incidence of thromboembolic, bleeding and anticoagulant‐related complications in HPN patients, and evaluate risk factors for first venous thrombosis (VT). MethodsThis retrospective cohort study included all adult patients followed for long‐term HPN at our center between 1986 and 2014. Primary outcomes were symptomatic objectively diagnosed VT, encompassing CRT and venous thromboembolism, and major bleeding. Secondary outcomes were vena cava syndrome and heparin‐induced thrombocytopenia or hypersensitivity. ResultsA total of 236 patients were included (median HPN duration, 17 months) and 136 received anticoagulants at HPN onset (57.6%). Overall, the annual incidence of first VT was 11.4% (95% confidence interval [95% CI], 8.6–14.7%); VT was associated with a personal history of thrombosis (adjusted hazard ratio, 2.22; 95% CI, 1.06–4.64), whereas anticoagulation seemed to account only for a mild protection (adjusted hazard ratio, 0.72; 95% CI, 0.36–1.44). The annual incidence of major bleeding was 4.3% for patients on anticoagulants vs. 1.8% for those off anticoagulants. Vena cava syndrome developed in 20.7% of patients with VT. One patient had isolated heparin‐induced thrombocytopenia (0.6%) and four had heparin hypersensitivity (2.5%). ConclusionsPatients on HPN have a significant risk of venous thrombosis, major bleeding and vena cava syndrome. Anticoagulants might reduce thrombosis risk, but population‐specific safety concerns remain.

Home Parenteral Nutrition in Adult Patients With Chronic Intestinal Failure: Catheter‐Related Complications Over 4 Decades at the Main Danish Tertiary Referral Center

“…Over the 4 decades, the clinically symptomatic CRVT incidence was 0.11 per 1000 CVC days. We acknowledge that by our definition, mainly clinically visible and vessel occlusive thrombi would be detected, and therefore the actual CRVT incidence may be higher . The hazard for CRVT was almost 3 times greater in women compared with men, while no significant increase in the hazard for CRVT was seen with increased age, history of mesenteric vascular events, or malignancy.…”

Parenteral Nutrition in Interventional Radiology: Consideration and Management of the Patient

“…20 The risk of thrombosis is about 10% per year of PN. 35 The risk of thrombosis is higher in patients with a history of venous thrombosis; the risk of first venous thrombosis is 13.3/100 patient years, and the rate of recurrent venous thrombosis is 24.5/100 patient years. 35 Additional risk factors for venous thrombosis include malignancy, thrombophilia, associated catheter infection, catheter diameter, and location of the catheter tip.…”

“…35 The risk of thrombosis is higher in patients with a history of venous thrombosis; the risk of first venous thrombosis is 13.3/100 patient years, and the rate of recurrent venous thrombosis is 24.5/100 patient years. 35 Additional risk factors for venous thrombosis include malignancy, thrombophilia, associated catheter infection, catheter diameter, and location of the catheter tip. 22 Placement of the catheter tip at the cavoatrial junction (outside of the right atrium due to concern for arrhythmia) is recommended to prevent thrombosis related to PN.…”

“…2,[36][37][38] Systemic anticoagulation for the prevention of catheterrelated thrombosis is not recommended, as research has demonstrated little to no benefit with this therapy. 35,39 In addition, Barco et al 35 cited multiple complications of anticoagulation, including heparin-induced thrombosis (0.6%), heparin sensitivity (2.5%), and death (1.8%), with an estimated 4% risk per year of major bleeding events in patients treated with anticoagulation.…”

Cost-of-illness model for venous thromboembolism

“…As previously stated, both the US and EU-28 costs [3,4,6] have many costs that have not been included -most specifically indirect costs which often are significantly more than direct medical costs, but also direct costs of complications that can be triggered by index VTE, such as arterial cardiovascular events which may occur at least 1.5 fold more frequently in patients that acquire an idiopathic VTE as compared to those who do not [13] or the development of vena cava syndrome as a long-term complication of catheter-related thrombosis [14]. We are hopeful that more comprehensive cost-of-illness models will be built, both incidence-based and prevalence-based, including all costs associated with the disease to better define overall costs, hospital-associated costs and, importantly, preventable costs to aid in determining where resources may best be utilized for preventative care.…”