Persistent neutrophilia is a marker for an increased risk of venous thrombosis

Abstract

In patients with cancer and myeloproliferative disorders, leukocytosis has been associated with an increased venous thromboembolic (VTE) risk. Our goal was to determine whether persistent neutrophilia (PN), not associated with known causes such as malignancies, infections or steroids, is independently associated with VTE. All adult patients with >3 outpatient complete blood counts (CBCs) within 3 years were included. PN was defined as having an absolute neutrophil count >95 % (>2SD) of the population (≥7.8 × 10(9)/L) on at least three CBCs, at least 2 months apart. Separate analyses for neutrophil counts ≥9 × 10(9)/L and ≥10 × 10(9)/L were also performed. Blood counts from inpatients were excluded. Primary outcome was diagnosis of VTE, as determined by ICD-9 codes. Odds ratios were adjusted for diabetes, smoking, obesity, gender, and age. Charlson score was utilized as a morbidity measure. Data on 43,538 outpatients were collected. Although there was no association of VTE with neutrophil counts ≥7.8 × 10(9)/L, patients with ≥9.0 × 10(9)/L neutrophils were twice as likely to be diagnosed with VTE compared to those with normal neutrophil counts (OR 2.0, 95 % CI 1.3, 3.1; p = 0.003). Patients with neutrophil counts ≥10.0 × 10(9)/L were at an even higher risk (OR 2.3, 95 % CI 1.2, 4.8; p = 0.019). Charlson scores significantly modified this risk when incorporated into analysis. Elevated neutrophil counts are associated with an increased risk of venous thrombosis even when they are not due to cancer, infection or steroids. In patients with significant comorbidities, neutrophilia may be a marker of VTE risk.

Keywords: Diabetes; Leukocytosis; Neutrophilia; Obesity; Venous thrombosis.