Abstract
BACKGROUND: Antithrombin, a key regulator of the coagulation cascade, is often decreased in patients with sepsis-associated disseminated intravascular coagulation (DIC). Antithrombin is commonly supplemented when activity levels fall to 70% or below in Japan. While there is considerable interindividual variability in antithrombin activity following treatment, the factors contributing to this variability remain unclear. This study aims to identify the determinants of post-treatment antithrombin activity levels and to investigate the potential association between antithrombin activity and bleeding risk. METHODS: We conducted a retrospective analysis using data from the post-marketing surveillance of antithrombin concentrate in patients with sepsis-associated DIC. Changes in antithrombin activity were calculated as: (Day 1 activity - baseline activity [%]) divided by the daily dose (international units [IU] per kilogram of body weight). Logistic regression analysis was employed to identify factors associated with changes in antithrombin activity following supplementation and factors related to bleeding risk. Additionally, Kaplan-Meier survival curves were used to examine the relationship between antithrombin activity and 28-day survival outcomes. RESULTS: A total of 1,524 patients were included in the analysis. The median baseline antithrombin activity was 49%, which increased to 74% on day 1 post-treatment. The mean change in antithrombin activity was 0.99% /IU/kg and followed a normal distribution. The SOFA score ≥ 13 (p = 0.035) and FDP score ≥ 3 (≥ 25μg/mL), part of the JAAM DIC score, (p = 0.016) were significantly associated with lower antithrombin activity increase. Patients achieving ≥ 1% /IU/kg increase showed a higher 28-day survival rate (relative risk: 0.72, p = 0.004). No significant association was found between antithrombin doses or activity changes and bleeding risk. CONCLUSION: A higher SOFA score and FDP level were associated with a smaller increase in post-treatment antithrombin activity. There was no clear association between antithrombin doses and bleeding risk. The present study suggests the necessity of individualized dosing beyond weight-based regimens.