RESUMO
UNLABELLED: Aim of the study was to elucidate value of HAS-BLED and HEMORR2HAGES scales for prediction bleedings in patients receiving long-term warfarin (W) therapy. MATERIAL AND METHODS: The study involved 119 patients (72 men) aged 60.9 ± 9.6 years with atrial fibrillation or venous thromboembolic complications. Follow up period was 5.6 ± 3.4 years. All bleedings were categorized as 1) single bleeding with INR > 4.0 during the 1st month of W therapy; 2) any single bleeding after 1st month of W therapy; 3) recurrent bleedings. CYP2C9 and VKORC1 (G3673A) genotypic variants were determined by PCR. Patients were divided into low (< 3 points of HAS-BLED scale, n = 58; < 4 points of HEMORR2HAGES scale, n = 109) and high (3 points of HAS-BLED scale, n = 61, ≥ 4 points of HEMORR2HAGES scale, n = 10) bleeding risk groups. RESULTS: There was no relationship between total HAS-BLED, HEMORR2HAGES scores and numbers of all as well as category 1 and 2 bleedings. The difference in bleeding frequency between high and low risk groups was significant only for recurrent bleedings. There were 22 (36.1%) and 5 (8.6%) recurrent bleedings among 61 and 58 patients with high and low-risk HAS-BLED score, respectively (p = 0.0048). Recurrent bleedings also occurred more frequently among patients with high risk (7/10, 70%) compared with low risk (20/109, 18.35%) HEMORR2HAGES score (p = 0.018). Subgroups of high and low bleeding risk according to HAS-BLED and HEMORR2HAGES scores differed only by proportion of patients with recurrent bleedings. High W sensitivity represented by 2*/2*, 2*/3*, 3*/3* CYP2C9 and/or AA VKORC1 homozygosis was detected in 25 of 119 patients. Six of 8 patients (75%) with category 1 bleedings were carriers of any polymorpism. CONCLUSION: HAS-BLED and HEMORR2HAGES scales performed best in predicting recurrent bleedings in patients on long term W therapy. Single bleedings with INR > 4.0 during 1st month of W therapy were associated with reduced W metabolism (AA VKORC1 or/and CYP2C9 allelic variants 2*/2*, 2*/3*, 3*/3*).