Low cut-off values increase diagnostic performance of protein S assays

Abstract

Conflicting data have been reported on the accuracy of protein S (PS) assays for detection of hereditary PS deficiency. In this study we assessed the diagnostic performance of two total PS antigen assays, four free PS assays and three PS activity assays in a group of 28 heterozygous carriers of mutations in PROS1 and 165 control subjects. Several control groups were formed, one of healthy volunteers and - because PS levels are influenced by oral contraception and pregnancy, and assays measuring PS activity may be influenced by the presence of the factor V Leiden mutation -, we also investigated the influences of these factors. All nine PS assays detected significantly reduced PS levels in subjects with a PROS1 mutation. Eight out of nine PS assays showed a 100% sensitivity and 100% specificity to detect heterozygous carriers of mutations in PROS1 with values far below the lower limit of the reference values obtained from healthy volunteers. Low specificities were found in subjects with a factor V Leiden mutation and in pregnant women. At lower cut-off levels, equal to the highest PS value found in heterozygous carriers of mutations in PROS1, the specificity considerably increased in these subjects. When using low cut-off levels equal to the highest PS value found in heterozygous carriers of mutations in PROS1, ensuring 100% sensitivity, the specificity in all study groups increases considerably, by which misclassification can be maximally avoided.