Main

DL defines abnormalities in the lipid profile caused by both genetic and acquired metabolic disorders. Most of the genetic lipid disorders are dominantly inherited and include familial hypercholesterolemia, familial combined hyperlipidemia, familial hypertriglyceridemia, and familial hypoalphalipoproteinemia (1). Acquired abnormalities in lipid metabolism are most commonly related to obesity (2). Persistent abnormalities of the lipid profile during childhood are associated with accelerated atherosclerosis and increased cardiovascular risk in early adult life (3).

Atherosclerosis is a chronic inflammatory response of the endothelium to metabolic and physical injuries leading to gradual accumulation of plaque in the vessel wall (4). The pathogenesis of atherosclerosis begins during childhood with the appearance of intimal lesions in the aorta of young children and fatty streaks in coronary arteries of adolescents. These lesions occur at the same vascular sites of atherosclerotic plaques observed in older ages (5–7). The process of atherogenesis leads to abnormal endothelial cell function that can be detected in the early stages, before anatomic evidence of plaque formation occurs (4, 8).

Studies in asymptomatic adults at risk for future myocardial infarction suggest that increased baseline plasma concentrations of tPA and PAI-1, reflecting reduced fibrinolytic activity, and increased plasma concentrations of fibrinogen or vWF are sensitive (nonspecific) markers of endothelial cell dysfunction, and useful predictors of future cardiovascular morbidity and mortality (9–17).

The present study investigated whether abnormalities in plasma concentrations of components of the fibrinolytic or coagulation systems are present in children with DL.

METHODS

Patient population.

All consecutive children with DL attending a follow-up clinic at the Hospital for Sick Children, Toronto, Ontario, Canada, and who were at least 8 y of age were considered for inclusion in this study. Age-matched healthy children on no medication and attending a school in Oakville, Ontario, Canada, were also asked to participate as a control group. Before entry into the study, informed consent was obtained from both the children and their parents. The study was approved by the research ethics boards of the Hospital for Sick Children and the Children's Hospital at McMaster University, Hamilton, Ontario, Canada.

Blood sampling.

All blood samples were collected before and after VO by clean venepuncture from the antecubital fossa between 0800 and 1000 h after overnight fasting. Preocclusion blood samples were obtained from one arm without the use of a tourniquet and after 15 min of resting supine. VO was performed on the other arm using a sphygmomanometer inflated to mid-diastolic–systolic blood pressure for 10 min. Postocclusion samples were obtained from the occluded arm before deflation of the blood pressure cuff. Blood samples were collected into 0.105 M buffered trisodium citrate (9 parts blood:1 part citrate) for tPA antigenic assay, PAI-1 antigenic and activity assays, plasminogen antigen and activity assays, α2-AP, D-dimer, α2-M, fibrinogen, and vWF. Another sample was collected into acidified sodium citrate tubes (pH 4.3; Stabilyte tubes; American Diagnostica, Greenwich, CT, USA) for tPA activity assays. All blood samples were centrifuged at 1700 ×g for 15 min at 4°C within 20 min of collection, and platelet-poor plasma was aliquoted and frozen at −70°C until time of assay. All postocclusion results were corrected for hemoconcentration using the correction factor F, calculated as follows:EQUATION

where HCT1 and HCT2 are hematocrit values before and after VO, respectively (18).

Laboratory assays.

The following assays were performed following the manufacturer's guidelines: tPA antigen (Imubind, tPA; American Diagnostica), tPA activity (Spectrolyse tPA/PAI; American Diagnostica), PAI-1 antigen (Imubind, PAI-1; American Diagnostica), PAI-1 activity (Spectrolyse PAI-1; American Diagnostica), plasminogen antigen (Imuclone; American Diagnostica), plasminogen activity (Actichrome PLG; American Diagnostica), and α2-AP activity (Actichrome; American Diagnostica) using an automated coagulation laboratory (ACL 300; Milan, Italy); D-dimer (Dimertest Gold EIA; American Diagnostica); fibrinogen by semiautomated Clauss method (ST4; Diagnostica Stago, Asnières, France); vWF antigen by ELISA (Affinity Biologicals, Hamilton, Ontario, Canada); and α2-M antigen immunologically by radial immunodiffusion using commercially available antibody (Cedarlane, Laboratories Limited, Hornby, Ontario, Canada).

Hematocrits were measured on a Coulter STK-S (Coulter Canada, Burlington, ON, Canada). The lipid profile including TC, HDL cholesterol, LDL cholesterol, and triglycerides was measured on a Vitros 950 Chemistry System (Ortho Clinical Diagnostics Inc, Rochester, NY, USA).

Statistical analysis.

Plasma concentrations of components of the fibrinolytic and coagulation system of preocclusion and postocclusion samples are presented as mean ± 95% CI. Ninety-five percent CI were calculated from SEM, assuming symmetric distribution of all the fibrinolytic and coagulation variables measured. To evaluate the interrelationship between the amount of the free active fraction of tPA (tPA activity) and the total tPA (tPA antigen), consisting of both the free active and the complexed inactive fraction of tPA, tPA activity/tPA antigen ratios in patients and control subjects were calculated and results presented as mean (± 95% CI). From previous published studies in adults, results of tPA and PAI-1 were considered the primary outcomes. Results of the other fibrinolytic and coagulation components were considered secondary outcomes. Statistically significant differences of mean values of fibrinolytic and coagulation components between patients and control subjects were tested by two-tailed t test. To assess whether fibrinolytic and coagulation components varied between children with familial DL and children with obesity-related DL as well as between children with hypercholesterolemia and children with hypertriglyceridemia, significant differences of mean values were assessed by one-way ANOVA with post hoc comparisons (Tukey-Kramer multiple comparisons test). Probability values less than 0.05 were considered statistically significant.

RESULTS

Patient population.

Thirty-six children with DL (14 girls and 22 boys) aged 8 to 16 y (mean, 12.5 y; median, 12.4 y) were included in this study. Twenty-three children had familial hypercholesterolemia, four children had familial combined hyperlipidemia, and nine children had obesity-related hyperlipidemia with a mean body mass index of 29.6. Results of the lipid profile and the calculated TC/HDL and LDL/HDL ratios are presented in Table 1 (19, 20). At the time of the study three children were receiving oral antihypercholesterolemic therapy with colestipol. All children had normal blood pressures. Twenty-six healthy children (14 girls, and 12 boys), aged 13 to 18 y (mean and median, 16 y), were included as control subjects.

Table 1 Lipid profile of children with DL
Full size table

Preocclusion results of components of the fibrinolytic and coagulation system.

Overall mean plasma concentrations (± 95% CI) of components of the fibrinolytic and the coagulation systems in preocclusion samples in patients and control subjects are summarized in Table 2. Mean plasma concentrations of tPA activity were significantly decreased in patients compared with control subjects, whereas mean tPA antigen plasma concentrations were similar in both populations (Table 2). Patients also showed significantly increased plasma concentrations of plasminogen (antigen and activity) and α2-M relative to control subjects, as well as highly significantly increased plasma concentrations of fibrinogen compared with control subjects (Table 2).

Table 2 Preocclusion results of fibrinolytic and coagulation components
Full size table

Compared with patients with familial DL, patients with obesity-related DL showed increased plasma concentrations of fibrinogen (mean, 3.73 g/L; 95% CI, 3.14 to 4.32 g/L versus mean, 3.11 g/L; 95% CI, 2.85 to 3.373 g/L; p < 0.05). Compared with patients with hypercholesterolemia, patients with hypertriglyceridemia showed increased plasma concentrations of both tPA antigen (mean, 7.29 ng/mL; 95% CI, 5.50 to 9.07 ng/mL versus mean, 4.78 ng/mL; 95% CI, 4.07 to 5.49 ng/mL) and plasminogen activity (mean, 1.45 U/mL; 95% CI, 1.19 to 1.71 U/mL versus mean, 1.24 U/mL; 95% CI, 1.2 to 1.28 U/mL; p < 0.05).

In all patients, plasma concentrations of components of the fibrinolytic and the coagulation systems in preocclusion samples showed no linear correlation with the results of the different lipid classes.

Postocclusion results of components of the fibrinolytic and coagulation system.

Overall mean plasma concentrations (± 95% CI) of components of the fibrinolytic and the coagulation systems in postocclusion samples in patients and control subjects are summarized in Table 3. Mean plasma concentrations of tPA antigen were significantly decreased in patients relative to control subjects (Table 3). Patients also showed significantly increased plasma concentrations of plasminogen (antigen and activity), and α2-M relative to control subjects, as well as highly significant increased plasma concentrations of fibrinogen (Table 3).

Table 3 Postocclusion results of fibrinolytic and coagulation components
Full size table

Compared with patients with familial DL, patients with obesity-related DL showed decreased plasma concentrations of tPA activity (mean, 0.92 U/mL; 95% CI, 0.48 to 1.37 U/mL versus mean, 1.91 U/mL; 95% CI, 1.49 to 2.33 U/mL; p < 0.05) and increased plasma concentrations of both plasminogen antigen (mean, 0.34 μg/mL; 95% CI, 0.27 to 0.41 μg/mL versus mean, 0.29 μg/mL; 95% CI, 0.27 to 0.31 μg/mL; p < 0.05) and PAI-1 activity (mean, 18.51 U/mL; 95% CI, 11.30 to 25.72 U/mL versus mean, 9.44 U/mL; 95% CI, 7.39 to 11.48 U/mL; p < 0.01). No statistically significant differences in the plasma concentrations of all fibrinolytic and coagulation components measured were observed between patients with hypercholesterolemia and those with hypertriglyceridemia.

In all patients, plasma concentrations of components of the fibrinolytic and the coagulation systems in postocclusion samples showed no linear correlation with the results of the different lipid classes.

tPA activity/tPA antigen ratios.

The mean tPA activity/tPA antigen ratio was significantly decreased in patients (mean, 0.28; 95% CI, 0.21 to 0.33) relative to control subjects (mean, 0.43; 95% CI, 0.30 to 0.56; p = 0.02) in preocclusion samples, reflecting increased levels of inactive tPA relative to free tPA in children with DL compared with control subjects (Fig. 1). In postocclusion samples, mean tPA activity/tPA antigen ratios were no longer statistically different between patients (mean, 0.30; 95% CI, 0.15 to 0.43) and control subjects (mean, 0.24; 95% CI, 0.19 to 0.29; p = 0.55;Fig. 1). When compared with preocclusion ratios, tPA activity/tPA antigen ratios in postocclusion samples were unchanged in patients, but significantly decreased in control subjects (p = 0.01), suggesting a decreased fibrinolytic response to VO in children with DL compared with control subjects (Fig. 1).

Figure 1
figure 1

tPA activity/tPA antigen ratios of preocclusion (Pre VO) and postocclusion (Post VO) samples in patients with DL and control subjects. Data are shown as mean (bar) ± 95% confidence intervals (vertical lines). *p value indicating a significant difference in tPA activity/tPA antigen ratios of preocclusion samples between patients and control subjects. **p value indicating a significant difference in tPA activity/tPA antigen ratios between preocclusion and postocclusion samples in control subjects.

Full size image

No statistically significant differences in tPA activity/tPA antigen ratios were observed between patients with familial DL and patients with obesity-related DL as well as between patients with hypercholesterolemia and those with hypertriglyceridemia.

DISCUSSION

Dyslipidemias are major risk factors for the development of atherosclerotic heart disease in young adults. Epidemiologic data in adults suggest that impaired fibrinolysis is a sensitive (nonspecific) marker of endothelial dysfunction and a strong predictor of future cardiovascular disease (9–13). Although increased serum lipid levels are readily identified in children with DL, no previous studies have determined whether impaired fibrinolysis commences during childhood. Results of this study show a reduced fibrinolytic activity in asymptomatic children with DL, suggesting that progressive endothelial dysfunction may occur early, which may justify early pharmacologic intervention in these children.

Normal serum lipid and lipoprotein levels differ markedly in children to adults, and age-specific ranges are well-documented (21). There is considerable evidence that hypercholesterolemic adults have abnormal lipid levels in childhood (22, 23). Family screening is frequently performed to detect affected children in the belief that early intervention will be of benefit, although the age at which intervention should be optimally initiated remains unknown.

Impaired flow-mediated dilation of femoral and brachial arteries reflecting endothelial dysfunction can be demonstrated in children with familial DL as young as 8 y of age (8, 24, 25). In addition, 66% of asymptomatic teenagers with heterozygous familial hypercholesterolemia and positive family history have abnormal cardiac stress thallium scans (26). Stress thallium scans have also been shown to highly correlate with angiographic abnormalities in homozygous familial hypercholesterolemia (27).

Although decreased fibrinolytic activity is considered a useful predictor of endothelial dysfunction in adults with atherosclerosis (10–13), there is still some controversy about whether decreased fibrinolytic activity is a consequence or a primary process in the pathophysiology of endothelial dysfunction. Abnormalities of the fibrinolytic system have not previously been investigated in children with DL. Major components of the fibrinolytic system, including tPA and, in part, PAI-1, are produced and released by endothelial cells (28). VO stress testing is a sensitive measure of the fibrinolytic system, measuring both the baseline (preocclusion) and endothelial-stimulated (postocclusion) fibrinolytic potential (29).

Our results show decreased fibrinolytic activity in asymptomatic children with DL before and after VO. Two different mechanisms responsible for the impaired fibrinolytic activity could be identified. Before VO, decreased fibrinolytic activity was related to decreased tPA activity in conjunction with a relative increase of complexed inactive tPA relative to free active tPA and to increased plasma concentration of α2-M. Although plasma concentrations of tPA antigen have been shown to reflect an inhibitory effect of PAI-1 on tPA activity, tPA complexes are not only formed with PAI-1 but also with other minor inhibitors including α2-M (11, 30, 31). Because plasma concentrations of PAI-1 antigen and activity were similar in both populations, the finding of increased plasma concentrations of α2-M in our patients suggests the presence of an increased inhibitory effect of this proteinase inhibitor on tPA. After VO, decreased fibrinolytic activity was related to decreased plasma concentration of tPA antigen. Impaired fibrinolytic response to VO has been related to either a deficient release of tPA during VO or increased PAI-1 activity levels before VO (32, 33). Because plasma concentrations of PAI-1 antigen and activity were similar in both populations, and no increased residual PAI-1 activity was present in patients relative to control subjects, impaired fibrinolytic response to VO in our patients likely reflects impaired release of tPA from the vascular endothelial cells, suggesting the presence of endothelial cell damage.

In our study, patients with obesity-related DL showed significantly decreased plasma concentrations of tPA activity and increased plasma concentrations of PAI-1 activity after VO compared with patients with familial DL. Although these findings suggest that the different types of DL may be associated with different degrees of severity of endothelial dysfunction and hypofibrinolysis, the number of patients in this study does not allow definite conclusions to be made.

Our results also show increased plasma concentrations of plasminogen, fibrinogen, and α2-M in children with DL. α2-M is a major plasma proteinase inhibitor that acts as a molecular trap for proteinase molecules including factor Xa, thrombin, tPA, kallikrein, and plasmin, hindering their ability to react with other proteins (31, 34–36). Increased concentrations of plasminogen and fibrinogen may reflect increased inhibitory effect of α2-M not only on tPA, as described previously, but also on kallikrein as well as on factor Xa and thrombin, respectively. Thus, α2-M may play a role in decreasing the fibrinolytic activity in children with DL and may be a useful marker of endothelial cell dysfunction in patients at increased risk for arterial disease. Further study is required to clarify the extent to which α2-M inhibits the fibrinolytic system relative to the coagulation system in these patients.

CONCLUSIONS

In summary, a decreased fibrinolytic activity is present in children with DL, suggesting the presence of endothelial cell dysfunction and possible early presymptomatic atherosclerosis. Further studies are required to determine possible cutoff values of fibrinolytic and coagulation components as useful indicators of both disease progression and response to therapy in children with DL, and whether the age of onset of fibrinolytic abnormalities predicts severity of future cardiovascular disease.