Mechanisms and effects of low plasminogen activator inhibitor type 1 activity

Sammanfattning: Bleeding diathesis is a common reason for investigation of the haemostatic system and in about 50% of the cases an abnormality is found. Plasminogen activator inhibitor type 1 (PAI-1) is an inhibitor of tissue- and urokinase plasminogen activator (tPA/uPA). High PAI-1 levels are associated with arterial and venous thrombotic conditions but are also seen in pregnancy, inflammatory and infectious diseases, diabetes and overweight. Low levels of PAI-1 have been reported in case studies as a reason for bleeding tendency. Studies of low PAI-1 activity as a cause for bleeding in large populations have so far been lacking. The aim of this thesis was to systematically study the prevalence of low PAI-1 activity in patients with bleeding diathesis, the clinical significance, evidence for hyperfibrinolysis and regulation. The method for determination of PAI-1 activity had been improved at our laboratory to more precisely measure lower levels of PAI-1 activity. We studied prospectively 586 patients with bleeding tendency with analysis of PAI-1 activity in addition to the routine investigation and compared with two control groups. The bleeding problems were clinically evaluated and compared between those with low PAI-1 and normal/high PAI-1. The prevalence of low PAI-1 among patients was 23% compared to 13% in blood donors and 10% in healthy controls. No specific bleeding symptom predicted for low PAI-1 activity. The 4G/5G polymorphism in the PAI-1 gene did not differ between patients and controls (paper I). From the original 586 patients 424 were further investigated for laboratory evidence of hyperfibrinolysis with plasmin-antiplasmin complex (PAP) and D-dimer. PAP was higher among patients with low PAI-1 activity compared to normal/high PAI-1, but the coagulation activity measured by D-dimer did not differ (paper III). In 181 samples from the same cohort we analysed C-peptide, proinsulin, high sensitivity C-reactive protein and interleukin 6 (IL-6). Body mass index (BMI), exogenous hormones, sex and age were registered. Low BMI, oral estrogens, young age and low C-peptide were significantly associated with low PAI-1 activity. After adjustments the effect of C-peptide, IL-6 and young age disappeared and low BMI remained the strongest predictor of low PAI-1 activity (paper IV). In 62 patients referred for transurethral resection of prostate (TURP) due to prostatic hyperplasia samples for PAI-1 were taken before surgery and analysed after hospitalisation. Haemoglobin was measured before surgery and the day of discharge, per-and postoperative bleeding was registered as well as bleeding complications, amount of resected prostate and days of hospitalisation. Bleeding complications occurred in 75% of the patients with low PAI-1 activity compared to 28% of the patients with normal/high PAI-1 activity. This became borderline significant after adjustments (paper II). In conclusion, low PAI-1 activity seems to be more prevalent among patients with bleeding tendency than controls, but the bleeding risk in patients with low PAI-1 activity seems to be of minor clinical significance. Low BMI is the strongest predictor of low PAI-1, which is associated with higher fibrinolytic activity than in patients with normal/high PAI-1.