The value of free enterprise.

from peripheral blood leukocytes as described previously (5). Amplification of genomic DNA was performed by PCR as described (1). The G3 A substitution at nucleotide position 20210 was detected by allele-specific restriction digestion by using HindIII (1). Among the 27 patients, 4 (14.8%) were heterozygous for the G 20210 3 A mutation in the prothrombin gene and 1 patient (3.7%) was homozy-gous, whereas 22 patients (81.5%) did not carry the mutation (Table 1). Among the 245 healthy control persons, 239 subjects (97.6%) did not carry the G 20210 3 A mutation, whereas 6 persons (2.4%) were het-erozygous, and none was homozy-gous for the mutation (Table 1). The difference of frequencies between the group of patients (18.5%) and the controls (2.4%) was statistically significant (P Ͻ0.005, Fisher's Exact test; integral part of GraphPad Prism TM , Ver. 2.0, Graph-Pad Software, Inc.). The mutation G 20210 3 A in the prothrombin gene increased the risk for pulmonary em-bolism approximately ninefold (odds ratio, 9.1; 95% confidence interval, 2.6 –32.1) compared with noncarriers of the mutation (Table 1). Among the 27 patients, 3 (11.1%) were heterozygous or homozygous for factor V Leiden mutation, and 2 (7.4%) had a protein C deficiency. One subject (3.7%) exhibited protein S deficiency, whereas no patient had antithrombin III deficiency. The patient homozygous for the G 20210 3 A mutation in the prothrombin gene was also heterozygous for factor V Leiden mutation. However, no further coincidence existed between carriers of the G 20210 3 A mutation in the prothrombin gene and the above-described risk factors for thrombo-philia. It has been reported recently that the prevalence of factor V Leiden mutation is high in patients with pulmonary embolism associated with deep venous thrombosis, whereas it is similar to that of controls in subjects with isolated pulmonary embolism (6). In our study, 2 patients of 17 with isolated pulmonary embolism (12%) carry the factor V Leiden mutation compared with 1 patient of 10 with pulmonary embo-lism related to deep venous thrombosis (10%), which was not a significant difference. Furthermore, we found the G 20210 3 A mutation in the prothrombin gene in 3 of 10 patients (30%) with pulmonary embolism associated with deep venous thrombosis and in 2 of 17 patients (12%) with isolated pulmonary embolism. Because of the small number of patients in the subgroups, these differences were not significant (P ϭ 0.4). Although our study is based on a …