Procarboxypeptidase U (proCPU, EC 220.127.116.11), also known as thrombin-activatable fibrinolysis inhibitor (TAFI), is considered to be a link between coagulation and fibrinolysis . During coagulation and fibrinolysis the active enzyme, carboxypeptidase U (CPU), can be generated from proCPU. CPU exerts its antifibrinolytic effect by removing C-terminal lysine residues from partially degraded fibrin , thereby decreasing the enhanced plasminogen activation and preventing an acceleration of fibrinolysis. Elevated levels of proCPU have been found to be a mild risk factor for venous thrombosis (odds ratio 1.7; 95% confidence interval 1.1–25)  and were observed in patients with angina pectoris and ischemic stroke. From cross-sectional studies it is known that proCPU concentrations increase with age in women not receiving postmenopausal hormone therapy (HT) [2,3]. Significantly lower values have been observed in postmenopausal women using HT, compared with non-users , whereas other studies did not confirm these findings.
Raloxifene is a second-generation selective estrogen receptor modulator (SERM) that has estrogen-like effects on bone and cardiovascular risk factors, and estrogen receptor antagonistic effects on the breast and uterus. Several studies have reported beneficial effects of raloxifene on various cardiovascular risk markers [4–7]. Raloxifene has been reported to reduce the risk of cardiovascular events in osteoporotic postmenopausal women at high cardiovascular risk . However, like oral HT, raloxifene increases the risk of venous thromboembolic disease . Recently, we reported a randomized, controlled trial in which raloxifene reduced proCPU plasma concentrations in healthy postmenopausal women during a 2-year study period suggesting a possible preventive effect on atherothrombotic disease .
HMR 3339 (4-chloro-11β-(4-(2-(diethylamino)ethoxy)phenyl)-estra-1,3,5(10)-triene-3, 17β-diol) is a newly developed SERM for the prevention of postmenopausal osteoporosis, that binds to the human recombinant estrogen receptor and shows tissue-specific agonistic and antagonistic activity in in-vitro and in-vivo models. Preliminary data suggest a favorable effect on menopausal symptoms and the cardiovascular system, as well as beneficial activity on bone and a favorable safety profile on the breast and uterus. Recently, we reported that HMR 3339 therapy was associated with an anti-atherogenic lipid profile  and with reduced C-reactive protein levels . In the present study we investigated the short-term effects of three doses of HMR 3339 on plasma proCPU concentrations and compared the effects with raloxifene.
For this study, patient selection and inclusion procedures have been reported previously . In short, 118 healthy, non-hysterectomized postmenopausal women were included in this 12-week multicenter, randomized, placebo-controlled, double-blind, phase II study. The investigation conformed to the principles outlined in the Declaration of Helsinki. The Institutional Review Boards of all participating centers approved the protocol. Written informed consent was obtained from each participant before entry into the study. Women were assigned randomly to either placebo (n = 22), HMR 3339 2.5 mg day−1 (HMR 2.5 group; n = 25, Hoechst Marion Roussel R&D, Romainville, France), HMR 3339 10 mg day−1 (HMR 10 group; n = 24), HMR 3339 50 mg day−1 (HMR 50 group; n = 24), or raloxifene 60 mg day−1 (Rlx 60 group; n = 23, Eli Lilly and Company, Indianapolis, IN, USA).
At baseline and after 4 and 12 weeks of treatment, venous blood samples were collected and stored at −80 °C until analysis. Plasma proCPU concentrations were determined by converting the zymogen to its active form and subsequently measuring the carboxypeptidase activity with a colorimetric assay . The intra-assay coefficient of variation had a mean of 3.7%.
Statistical analysis was performed using the Statistical Package for the Social Sciences 9.0 (SPSS Inc., Chicago, IL, USA). We compared baseline measurements and means of the individual percentage changes from baseline between groups using standard parametric tests. Analysis of covariance (ancova) for repeated measurements with the baseline value of proCPU as a constant covariate was used for comparisons among and between the groups. A two-tailed P < 0.05 was considered statistically significant.
At baseline, no significant differences were found between the groups in either demographic characteristics or in proCPU concentrations. Compared with placebo, ancova with the baseline value of proCPU as a constant covariate, showed a decrease in proCPU concentrations after 12 weeks in the HMR 2.5 group (P = 0.102), HMR 10 group (P = 0.005), HMR 50 group (P < 0.001), and in the Rlx group (P = 0.005). The mean percentage reduction from baseline vs. placebo after 12 weeks of treatment was 5.8% (P = 0.099) in the HMR 2.5 group, 7.0% (P = 0.039) in the HMR 10 group, 10.9% (P = 0.003) in the HMR 50 group and 10.5% (P = 0.002) in the Rlx group (Fig. 1). Post-hoc analyses revealed no significant differences in the percentage change from baseline after 4 or 12 weeks between the four treatment groups.
Until now, three randomized controlled trials have been published that investigated the effect of HT or raloxifene on proCPU concentrations [5,11,12]. Post et al. described unchanged proCPU levels after 12 weeks in women treated with unopposed oral estradiol 2 mg or oral estradiol 2 mg combined with dydrogesterone 10 mg . However, these authors found a reduction compared to placebo of 5.9% in women treated with oral estradiol 2 mg combined with trimegestone 0.5 mg . Bladbjerg et al. reported that HT during 5–6 years had no overall effect on proCPU concentrations . Recently, we reported a randomized controlled trial in which proCPU vs. placebo was reduced with 4.3% by raloxifene 60 mg and with 4.0% by raloxifene 150 mg after 24 months, but no significant changes were found in the oral HT group (conjugated equine estrogens 0.625 mg combined continuously with medroxyprogesterone acetate 2.5 mg) . Our present findings with HMR 3339 and raloxifene are in line with the reductions found for raloxifene previously .
High proCPU concentrations are associated with an increased risk for a thrombotic event  and for recurrent venous thromboembolism. It has been suggested that this is caused by a reduced fibrinolytic capacity . Data about the role of proCPU in arterial thrombosis are conflicting. Some studies reported an association between arterial thrombotic events and elevated plasma proCPU levels, whereas others reported an association between low proCPU levels and coronary disease. These contradictory results may be explained by the different enzyme-linked immunosorbent assays (ELISAs) used, which do not all recognize the same polymorphisms .
When we compare HMR 3339 and raloxifene it can be concluded that HMR 3339 acts as raloxifene with regard to plasma proCPU concentrations, as well as for the lipid profile . Because raloxifene has been found to have cardioprotective effects in osteoporotic postmenopausal women at high cardiovascular risk , but also increases the risk of venous thromboembolic disease , the same cardiovascular profile may also be applicable to HMR 3339. However, randomized controlled trials with hard clinical endpoints are needed to confirm this hypothesis.