Role of Routine Laboratory Tests in Assessing Risk of Recurrent Venous Thrombosis: Results from the MEGA Follow-Up Study.

Abstract

Kidney function, measured as estimated glomerular filtration rate (eGFR), glucose levels and haematologic variables (blood cell count) are easily obtainable tests, and have been associated with increased risk of first venous thrombosis (VT). Whether these routine tests can identify patients at risk of recurrence is unclear. Our aim was to investigate the predictive value of serum glucose levels, eGFR and haematologic variables in patients at risk of recurrent VT. Patients with a first VT were followed from discontinuation of anticoagulant treatment. Percentile categories of eGFR, glucose levels and haematologic variables were established. Crude incidence rates with 95% confidence intervals (CIs) of recurrence were estimated in each percentile category. Cox regression models were used to compare groups, adjusted for age and sex. Of 2,106 patients followed for a median of 6.9 years, 326 developed recurrence (incidence rate, 2.7/100 patient-years; 95% CI, 2.5-3.1). The adjusted hazard ratio for recurrence was 1.5 (95% CI, 0.9-2.4) in the lowest eGFR percentile category (< 59 mL/min/1.73 m2) versus the reference (≥86 mL/min/1.73 m2). Stratification by unprovoked or provoked first events yielded similar results. The combination of a first unprovoked VT with renal dysfunction was associated with a threefold increased risk of recurrence compared with those with first provoked VT and normal kidney function (hazard ratio, 3.1, 95% CI, 1.6-5.9). Glucose levels and haematologic variables were not associated with increased recurrence risk. Testing glucose levels and haematologic variables did not identify patients at increased risk of recurrent VT. Renal dysfunction tests may have some predictive value, particularly in combination with other variables.