Methionine-Loading and Random Homocysteine Tests Have No Added Value in Risk Assessment for Venous and Arterial Thrombosis.

Hyperhomocysteinemia is a risk factor for venous and arterial thrombosis. Different diagnostic strategies are used to identify subjects at risk of thrombosis, related to hyperhomocysteinemia. Measurements of fasting and methionine-loading levels are usually recommended. Alternatively, random homocysteine measurements may simplify the procedure. Random levels < 10 and > 20 μmol/l are considered to indicate normohomocysteinemia and hyperhomocysteinemia, respectively, while consecutive fasting and methionine-loading tests are required at levels 10–20 μmol/l. We performed a study to assess the most suitable strategy in a large cohort of families with hereditary (index) deficiencies of protein S, protein C or antithrombin. Random, fasting and methionine-loading homocysteine samples were measured in 713 relatives. According to predefined cut-off levels hyperhomocysteinemic and normohomocysteinemic relatives were identified and their absolute risks of thrombosis were compared. Relatives with random homocysteine levels > 20 μmol/l were not at risk of venous or arterial thrombosis compared to relatives with levels < 10 μmol/l (relative risks 0.9 [95% CI, 0.4–2.3] and 1.7 [0.5–5.7], respectively). Fasting hyperhomocysteinemia (homocysteine levels > 18.5 μmol/l) was associated with an increased risk of venous and arterial thrombosis (relative risks 2.6 [1.3–4.8] and 3.7 [1.5–8.4)], respectively) (Table). Relatives with normal fasting homocysteine levels, but methionine-loading hyperhomocysteinemia (homocysteine levels > 58.8 μmol/l) were not at risk; relative risk 0.8 (0.2–1.9) for venous thrombosis and 1.1 (0.2–3.9) for arterial thrombosis. Exclusion of relatives with an index deficiency did not alter the risk estimates, while annual incidences of normohomocysteinemic relatives decreased to 0.19% per year (0.12–0.29), which is comparable with the annual incidence in the normal population. As only fasting homocysteine identified subjects at risk of thrombosis, random homocysteine and methionine-loading tests can be omitted in clinical practice.