Association of antenatal antithrombin activity with perinatal liver dysfunction: A prospective multicenter study.

Abstract

Liver dysfunction with decreased antithrombin (AT) activity and/or thrombocytopenia is life threatening in pregnant women. Whether AT is clinically useful for prediction of liver dysfunction remains unclear. A total of 541 women were registered prospectively at gestational week 34.7 (20.0-41.4) with available data on antenatal AT and platelet count (PLC). Liver dysfunction defined as serum aspartate aminotransferase>45IU/L concomitant with lactate dehydrogenase>400IU/L occurred in five women antenatally (≤2weeks before delivery) and in 17 women post-partum (within 1week post-partum). Median (5th-95th) antenatal value was 85 (62-110)% for AT and 202 (118-315)×109 /L for PLC in the 541 women and was significantly lower in women with than without perinatal liver dysfunction; 75 (51-108) versus 86 (62-110)% and 179 (56-244) versus 203 (121-316)×109 /L, respectively. Nineteen (86%) women with liver dysfunction showed AT≤62% or thrombocytopenia (PLC≤118×109 /L) perinatally, but five lacked thrombocytopenia throughout the perinatal period. The best cut-off (AT, 77%; PLC, 139×109 /L) suggested by receiver operating characteristic curve gave antenatal AT and PLC sensitivity of 59% and 41% with positive predictive value of 8.6% and 14%, respectively, and combined use of AT and PLC improved sensitivity to 73% (16/22) with positive predictive value of 9.2% for prediction of perinatal liver dysfunction. Reduced AT not accompanied by thrombocytopenia can precede liver dysfunction. Clinical introduction of AT may enhance the safety of pregnant women.