Venous Thromboembolism Prophylaxis in High-Risk Pediatric Oncology Patients

Introduction

Pediatric oncology patients are at increased risk for hospital-acquired venous thromboembolism (HA-VTE), which is associated with increased morbidity in this population. Up to 20% of pediatric oncology patients with central venous catheters (CVC) experience HA-VTE. However, much debate exists over the use of pharmacologic thromboprophylaxis to prevent HA-VTE in the pediatric oncology population, primarily due to lack of evidence regarding safety and efficacy. Pediatric oncology patients have many well-established risk factors for VTE including adolescence, obesity, CVC presence, and receipt of asparaginase chemotherapy. Given the uncertainty regarding prevention strategies in this population, research is needed regarding the use of prophylactic anticoagulation in pediatric oncology patients.

Objective: To evaluate the use of prophylactic anticoagulation in pediatric oncology patients at risk for HA-VTE.

Methods

A retrospective, single center chart review of pediatric oncology patients who received clinical care at a tertiary care children's hospital was performed. Over 1600 patients were captured from a departmental database of all new oncologic diagnoses identified between January 1, 2014 and December 31, 2014. Patients were excluded if they received care at another institution or did not have an additional risk factor for HA-VTE outside of cancer diagnosis. Defined risk factors were age >12 years, body mass index (BMI) >30, CVC placement, receipt of asparaginase, and history of thrombosis. 514 patients met inclusion criteria. Patient age, BMI, history of thrombophilia, type of CVC, cancer type, steroid administration and asparaginase administration were analyzed as potential risk factors for VTE. Logistic regression was used to model odds of thrombus as a function of risk factors. A classification and regression tree (CART) model was also fitted to identify predictors of thrombus.

Results

Of 514 patients identified as high-risk for VTE, only 19 (3.7%) received prophylaxis. Most patients received enoxaparin for prophylaxis. Almost 15% (75/514) developed VTE during the study period. Of those who developed a VTE, 9% (7/75) were on prophylaxis at the time of VTE. 100% of the patients that developed VTE had a CVC in place. Obese patients were most likely to develop VTE with an odds ratio of 2.2 (95% CI 1-4.3, p 0.032). 28% (18/65) of obese patients developed VTE, yet only 6% (4/65) of obese patients received prophylaxis. Of these obese patients, 50% (2/4) developed VTE despite prophylaxis. Patients age >12 years were more likely to have VTE than those age <12 years (OR 1.1; 95% CI 1-1.1, p 0.023). Patients with Hickman catheters also had twice the odds of developing VTE (OR 2.1; 95% CI 1-4.1, p 0.046). In evaluating VTE occurrence by malignancy type, patients with hepatic tumors had the highest occurrence of thrombosis (8/12; 87%) followed by patients with myeloid leukemia (10/36; 28%), and lymphoblastic leukemia (28/170; 16%). The CART model was able to accurately predict occurrence or non-occurrence of thrombus in 87% of patients in the dataset based on the risk factors of obesity, age, and cancer type.

Discussion

It is well-established that malignancy is a significant risk factor for the development of VTE in pediatric oncology patients. Yet due to concern for bleeding events in these patients who frequently have thrombocytopenia, pharmacologic prophylaxis may not be used despite additional high-risk factors for VTE. This study demonstrates that pharmacologic prophylaxis is rarely used, even among high-risk pediatric oncology patients. For each risk group, most patients prescribed prophylaxis did not develop thrombus, however VTE still occurred in some. Risk stratification for pediatric oncology patients is essential given potential bleeding risk. We were able to identify additional high-risk features such as obesity, type of CVC, and cancer type. The CART model for high-risk pediatric populations had high specificity (88%) but poor sensitivity (23%). Application of this model could inform appropriate use of thromboprophylaxis in pediatric oncology patients to reduce the incidence of HA-VTE in this population. Further research is needed regarding safety and efficacy of prophylaxis, but this study provides important data to guide future steps.

Amos:Genetech: Consultancy, Honoraria; Sanofi: Consultancy, Honoraria.