Role of Low Albumin Levels in Pediatric Patients with Central Venous Catheters at Risk for Thrombosis

Introduction: The incidence of venous thromboembolism (VTE) in children has been increasing by about 10% per year for the past decade. This rise in VTE in children is attributed to improved survival of pediatric subspecialty patients as well as increased utilization of life-saving measures such as central venous catheters (CVCs) that can lead to catheter related thromboses (CRTs). More than 50% of venous thromboses in children and 90% of thromboses in neonates are CRTs, making catheter placement the single most significant risk factor for thrombosis in children. Association studies in adults have suggested an inverse relationship between serum albumin and risk of VTE. Given the paucity of data about albumin levels in pediatric patients in relation to VTE, we first performed a retrospective study to determine the prevalence of low albumin levels in pediatric patients with CRTs. We then initiated a prospective study to evaluate the incidence of hypoalbuminemia in pediatric patients at the time of CVC placement.

Methods: We initially performed a retrospective study of 78 pediatric patients with a documented thrombotic event referred to the pediatric hematology service at University of Arizona between Aug 2014 and Jul 2017. Data collected included patient demographics, clinical diagnoses, type and location of thrombosis, presence or absence of a CVC, and albumin level within four days of the thrombotic event. We then designed a prospective study of pediatric patients undergoing CVC placement. Baseline data collected included patient demographics, clinical diagnoses, type of CVC placed, and laboratory results for albumin, factor VIII activity, D-dimer, and antithrombin activity within 4 days of CVC placement.

Results: In the retrospective cohort of 78 consecutive pediatric patients with thrombosis, 84 thrombotic events were documented, of which 41 (49%) were CRTs. CRTs occurred in 22/35 (63%) patients less than 12 years of age and 14/42 (33%) patients aged 12 to 23 (P = 0.02). Serum albumin levels were available for 53 thrombotic events. Albumin levels were below the normal reference range in 27/30 (90%) of CRTs compared with 12/23 (52%) non-CRTs (P = 0.003).

In the prospective study, data have been analyzed for the first 58 patients enrolled ages 2 months-17 years. At baseline, the prevalence of serum albumin below the lower limit of normal was 71% (41/58). Factor VIII activity was above 200% in 28/58 patients (48%), D-dimer was elevated in 48/58 patients (83%), and antithrombin was below 85% in 21/58 patients (36%). The mean duration of follow-up is 12 months. Four of the first 58 (6.9%) patients enrolled have had a CRT, all of whom had low albumin and elevated D-dimer levels at baseline.

Conclusion: Hypoalbuminemia is highly prevalent in pediatric patients with thrombosis and also in patients with CVCs, who are at increased risk of thrombosis. In our retrospective analysis of pediatric patients with thrombosis, a low serum albumin level was documented in 74% of patients with thrombosis, including 90% of patients with a CRT. Preliminary analysis of the first 58 patients enrolled in a prospective study of patients with CVCs demonstrates a high prevalence (71%) of low albumin levels at the time of CVC placement. Further analysis is needed to define the value of serum albumin in thrombotic risk prediction and further explore potential prothrombotic mechanisms in CRT.