| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Divisions of Endocrinology (J.K.D., D.K.V.) and Cardiovascular Medicine (J.R.B., J.C., D.E.V.) and Department of Biostatistics (Q.C.), Vanderbilt University School of Medicine, Nashville, Tennessee 37232; and Department of Neurosurgery (L.S.B.), California Center for Pituitary Disorders at the University of California, San Francisco, San Francisco, California 94143
Address all correspondence and requests for reprints to: Jessica K. Devin, M.D., 715 Preston Research Building, 2220 Pierce Avenue, Nashville, Tennessee 37232. E-mail: JKS98{at}Alum.Dartmouth.org.
 |
Abstract |
|---|
 Top Abstract IntroductionPatients and MethodsResultsDiscussionReferences |
|---|
Objective: The objective of the study was to determine the fibrinolytic profile of adults with GHD in comparison with controls.
Study Design and Participants: This was a prospective, observational study including 12 adults with GHD. Twelve gender-, age-, and body mass index-matched adults served as controls.
Main Outcome Measures: The primary outcome measures were circadian plasma PAI-1 antigen with corresponding tissue-plasminogen activator (tPA) activity values. Endothelial function was assessed by flow-mediated vasodilation and fibrinolytic potential by venous occlusion test.
Results: Adults with GHD exhibited an unfavorable 24-h fibrinolytic profile characterized by a mean 62% elevation in PAI-1 antigen (2.77 ng/ml after adjustment for baseline PAI-1; P = 0.049) in the setting of a mean 24% reduction in tPA activity (–0.17 IU/ml after adjustment for baseline tPA; P = 0.003). Fibrinolytic response was defective in GHD, as demonstrated by a sustained elevation in PAI-1 activity greater than 4 IU/ml after venous occlusion [7.2 IU/ml (interquartile range 0.8–17.4); P = 0.018]. Endothelial function was impaired in GHD, as quantified by percent flow-mediated vasodilation over 120 sec [area under the curve 3.8 (interquartile range –2.4 to 7.9) vs. 12.8 (interquartile range 2.1–19.4); P = 0.043].
Conclusions: Adults with GHD demonstrate alterations in plasma fibrinolytic balance, including elevated levels of PAI-1 antigen with decreased tPA activity. These changes may contribute to the increased cardiovascular morbidity within this population.
|
Introduction |
|---|
|
TopAbstractIntroductionPatients and MethodsResultsDiscussionReferences |
|---|
Impaired fibrinolytic function has been recognized as an important determinant of cardiovascular risk in numerous populations. The impact of GHD on plasma fibrinolytic balance, however, has not been comprehensively investigated. Plasminogen activator inhibitor-1 (PAI-1) is the main physiologic inhibitor of tissue plasminogen activator (tPA) and is recognized as a primary regulator of the fibrinolytic system. Low fibrinolytic activity, as defined by PAI-1 excess, appears to be a risk factor for recurrent deep vein thrombosis and ischemic heart disease as well as stroke and insulin resistance (13). As expected, adults with GHD have previously demonstrated a significant elevation in PAI-1; furthermore, levels of PAI-1 decrease with several months of GH replacement (14, 15, 16). PAI-1 antigen, however, has a well-recognized circadian variation that appears to clinically impact our diurnal fibrinolytic activity; thus, a single measurement provides an inadequate view of an individual’s fibrinolytic capacity (17).
We hypothesized that the metabolic, hormonal, and inflammatory abnormalities associated with the GHD state would unfavorably impact plasma levels of PAI-1 antigen and tPA activity and that these alterations in turn may contribute to the increased vascular morbidity observed within this population. This study was therefore designed to characterize the 24-h fibrinolytic profile and potential of adults with GHD in comparison with gender-, age-, and body mass index (BMI)-matched controls.
|
Patients and Methods |
|---|
|
TopAbstractIntroductionPatients and MethodsResultsDiscussionReferences |
|---|
Adults with GHD were recruited from Vanderbilt University’s Pituitary Center between August 2005 and November 2006 at the time of diagnosis of GHD. Inclusion criteria included a peak response to GHRH-arginine stimulation test less than 9.5 ng/ml and age 18–65 yr. Exclusion criteria were: 1) cardiovascular disease (previous myocardial infarction or known coronary artery disease); 2) diabetes mellitus; 3) pregnant or nursing; or 4) current daily use of any drug known to affect the fibrinolytic system. Adults with GHD received additional hormone replacement therapy if indicated based on clinical signs and symptoms as well as the corresponding biochemical evaluation: low-dose ACTH stimulated cortisol less than 18 µg/dl; a decline in free T4; below-normal serum estradiol or free testosterone in the setting of inappropriately low gonadotropins; plasma hyperosmolarity with an inappropriately dilute urine and a clinical response to vasopressin. All subjects demonstrated adequate and stable replacement therapy at the time of study entry.
All control participants were recruited from Vanderbilt University’s General Clinical Research Center (GCRC) Recruitment Registry or through advertisement on the Vanderbilt Medical Center campus. Controls were recruited on a case-by-case basis after the recruitment of each adult with GHD and were closely matched for gender, age within 5 yr, and BMI. Exclusion criteria were identical with those previous listed with the addition of a history of pituitary disease or any type of radiation to the region involving the pituitary gland.
Informed consent was obtained from all patients before initiation of study procedures. The study was approved by the Institutional Review Board at Vanderbilt University.
Study protocol
All study visits were identical and took place in the GCRC at Vanderbilt University. All participants arrived at 0730 h after an 8-h overnight fast and full night of sleep. Baseline assessments included vital signs, waist and hip circumference, height, and weight; a clinical exam was performed by study physician (J.K.D.). None of the participants reported any recent illness at the time of study visit. Additional cardiovascular risk factors, such as tobacco use and family history of premature cardiovascular disease in a first-degree relative, were assessed; a complete list of medications was obtained.
A peripheral iv line was placed and baseline fasting laboratories were drawn at 0800 h. Circadian laboratories (GH, PAI-1 antigen, tPA activity) were additionally obtained at 0800 h and every 2 h for 24 h thereafter. Assessment of PAI-1 antigen and tPA activity was not performed at 1600 h due to interference from the venous occlusion test. All laboratories were drawn with the participant in a supine position and after a 10-min period of rest. Flow-mediated dilation was performed while fasting at 1000 h, as described below. The venous occlusion test was performed at 1400 h. Meal times and content were standardized. Participants were not allowed to leave the GCRC during study visits and were not permitted to nap, exercise, eat between meals, or smoke during study visits.
Biochemical assays
All plasma for analysis of fibrinolytic parameters was collected in a Stabilyte tube (Trinity Biotech USA; St. Louis., MO) and immediately stored at –80 C. Plasma was aliquoted into separate microfuge tubes at the time of collection to avoid repeated freeze-thaw cycles. Plasma tPA activity was obtained by Chromolize tPA assay (Trinity Biotech USA). The intraassay coefficient of variation (CV) is 11.0% and the interassay CV, 5.3%. Plasma tPA antigen was determined by TintElize tPA assay (Trinity Biotech USA). The intraassay CV is 10%. PAI-1 antigen was determined by TintElize PAI-1 assay (Trinity Biotech USA). This kit is specific for endothelial-type human PAI-1 in the active and inactive (latent) forms as well as PAI-1 complexed with tPA. The interassay CV is 7.3% and the intraassay CV is 7.6%. PAI-1 activity was determined by Chromolize PAI-1 assay (Trinity Biotech USA). The intraassay CV is 3.0%. Plasma fibrinogen was analyzed by the Clauss method using the STA-R Evolution coagulation analyzer (DIAGNOSTICA STAGO Inc.; Parsippany, NJ).
Plasma GH levels, IGF-I, aldosterone, and insulin were measured in Vanderbilt University’s Hormone Assay Core. All samples were run in duplicate with internal controls. Human GH double antibody was obtained from MP Biomedicals (Solon, OH) for RIA; the intraassay CV is 8.3%. Similarly, plasma aldosterone double antibody was obtained from MP Biomedicals for RIA; the intraassay CV is 8.5%. Plasma insulin was measured using human insulin-specific RIA (Linco Research, St. Charles, MO); intraassay CV is 6.4%. Plasma IGF-I was measured using the quantitative determination of IGF-I (with IGF binding protein blocked) RIA (ALPCO Diagnostics, Windham, NH). The 50th percentiles for IGF-I (nanograms per milliliter) reported within the various age groups are: 198 for 20–30 yr, 188 for 30–40 yr, 178 for 40–50 yr, and 169 for 50–60 yr. The intraassay CV is 5.0%.
Fasting lipid profile and lipoprotein (a) [Lp(a)] were analyzed at the time of collection in Vanderbilt University’s Lipid Core Laboratory. Plasma Lp(a) was determined using the Macro Lp(a) ELISA kit (Trinity Biotech USA; Jamestown, NY). The interassay CV is 8.8% and the intraassay CV is 5.0%. Total cholesterol, high-density lipoprotein (HDL), and triglycerides were determined using ACE cholesterol, HDL-C, and triglyceride reagent, respectively, and performed on the Alfa Wasserman ACE clinical chemistry system (ALFA Wasserman, West Caldwell, NJ). The intraassay CV of the ACE cholesterol assay is 1.8% and the interassay CV is 2.0%. The intraassay CV of the triglycerides assay is 1.4% and the interassay CV is 1.9%. The intraassay CV of the HDL-cholesterol assay is 2.3% and the interassay CV is 3.9%. Low-density lipoprotein (LDL) was calculated using the Friedwald LDL equation. All of the participants were found to have a triglyceride level less than 400 mg/dl.
C-reactive protein and plasma renin activity were determined at the time of collection in Vanderbilt University Hospital’s clinical laboratory using the Cobas Integra 800 clinical chemistry system (Roche Diagnostics, Indianapolis, IN) and RIA, respectively. Serum glucose was determined in the GCRC Core Laboratory using the VITROS 250 chemistry analyzer system (Ortho-Clinical Diagnostics, Inc., Rochester, NY). The within laboratory CV is 1.7%.
Plasma inflammatory cytokines (IL-6, IL-8, and IL-10) were analyzed in triplicate in Vanderbilt University’s Immunology Core Laboratory using the BD cytometric bead array human inflammation kit (BD Biosciences, San Diego, CA). The intraassay CVs of IL-8, IL-6, and IL-10 are 5.0, 9.0, and 7.9%, respectively.
Flow-mediated dilation
Brachial artery flow-mediated dilation was measured in the dominant arm with ultrasound. All subjects fasted overnight and had refrained from exercise and the intake of caffeine for 10 h. Brachial artery ultrasound images were obtained with a high-resolution cardiovascular ultrasound system (Logiq 700 Expert series; GE, Waukesha, WI) using a 10-MHz GE vascular linear array probe.
Baseline images of the brachial artery were obtained by scanning the artery in longitudinal section 5–10 cm above the elbow of the dominant arm. Gain and depth settings were set to optimize images of the interface between the lumen and arterial wall; settings were not changed during the study. Brachial artery diameter was defined as the distance from the anterior intima-vessel lumen interface to the posterior intima-vessel lumen interface. Boundaries for diameter measurements were identified manually with electronic calipers. Ischemia was then induced by inflating a sphygmomanometric cuff around the proximal forearm to a pressure 40 mm Hg greater than the patient’s systolic blood pressure for 4 min. Brachial artery diameter measurements were taken at time points 30, 60, 90, and 120 sec after cuff deflation.
All measurements were obtained at the time of the procedure in a blinded fashion by a single operator (J.K.D.). The average of three successive measurements was used to obtain the resting brachial artery diameter. Intraobserver variability was calculated from these measurements and found to be 3.1% over the course of the study. Flow-mediated dilation was expressed as a percentage at 30, 60, 90, and 120 sec after cuff deflation: (brachial diameter at that time point minus the baseline diameter)/(baseline diameter (millimeters).
Venous occlusion test
Fibrinolytic potential was assessed by a venous occlusion test as previously described (18, 19). A defective fibrinolytic response to venous occlusion was defined as residual PAI-1 activity greater than 4 IU/ml, an increase in tPA activity less than 3.5-fold resting value, or a tPA antigen increase less than 3-fold resting value. Evidence of a sustained elevation of PAI-1 antigen after venous occlusion was also assessed.
Statistical methods
Results are expressed as median with interquartile range (IQR) unless otherwise specified. Hemolyzed samples taken for determination of fibrinolytic parameters were not included in the analyses.
Continuous and categorical data were compared between groups using the Wilcoxon rank sum and Fisher’s exact test, respectively. A ratio of tPA activity was defined as tPA activity after venous occlusion to tPA activity before venous occlusion. A one-sample t test was used to compare the ratios from each group to the test value of 3.5, whereas a defective fibrinolytic response represents an increase in tPA activity less than 3.5-fold resting value. The mean difference of the ratio from the test value of 3.5 is reported with the 95% confidence interval. The mean percent flow-mediated dilation was plotted vs. time. Area under the curve (AUC) was calculated for each subject and groups compared using the Wilcoxon rank sum test. Finally, circadian PAI-1 antigen and tPA activity levels were compared between groups using linear mixed model analysis with spatial power structure. The repeated measure in each analysis was hours and the factor was study group; baseline values of PAI-1 antigen and tPA activity were used as covariates. Circadian GH, PAI-1 antigen, and tPA activity levels were also summarized using AUC; Spearman correlation was used to determine the association between GH AUC vs. PAI-1 antigen AUC as well as GH AUC vs. tPA activity AUC.
Sample size was calculated with P.S. software (version 2.1.31; http://biostat.mc.vanderbilt.edu/twiki/bind/view/Main/PowerSampleSize) using the design for an independent two-sample t test (20). Based on previous studies, we anticipated that a sample size of six patients in each group would provide sufficient power (0.9) to detect a clinically significant change in PAI-1 antigen (
= 10 ng/ml) with an 
= 0.05 and 
= 5 ng/ml (15, 16, 21). We therefore designed this study to include 12 patients in each arm to account for early dropout and missing data during study visits.
All data were maintained within a secure, Internet-based data collection system. All statistical analyses were performed using SPSS Inc. (version 13; Chicago, IL) and SAS Institute Inc. (version 9.1; Cary, NC).
|
Results |
|---|
|
TopAbstractIntroductionPatients and MethodsResultsDiscussionReferences |
|---|
The study included 12 adults with GHD and 12 control participants (Table 1
). Two adults with GHD were prescribed antihypertensive therapy. None of the study participants were receiving statin therapy, and groups were matched in regard to oral estrogen use (n = 4 females/group). Both males with GHD received transdermal testosterone replacement for central hypogonadism.
|
Table 2 summarizes analyses of fasting laboratories. Adults with GHD exhibited significantly higher levels of C-reactive protein (P = 0.043), total cholesterol (P = 0.004), LDL (P = 0.030), triglycerides (P = 0.001), and IL-8 (P = 0.043), compared with controls. Additionally, glucose homeostasis was impaired in adults with GHD, as characterized by an elevated fasting insulin (P = 0.007) and homeostasis model assessment index (P = 0.008), in the setting of normal fasting blood glucose. Alterations in the renin-angiotensin-aldosterone system were demonstrated by a significant elevation in aldosterone in adults with GHD (P = 0.033) as well as a trend toward a higher plasma renin activity (P = 0.104), compared with controls.
|
Adults with GHD exhibited an unfavorable 24-h fibrinolytic profile characterized by a mean 62% elevation in PAI-1 antigen (2.77 ng/ml after adjustment for baseline PAI-1; P = 0.049) in the setting of a mean 24% reduction in tPA activity (–0.17 IU/ml after adjustment for baseline tPA; P = 0.003) (Fig. 1). Notably, the characteristic diurnal rhythm of PAI-1 antigen appeared lost in adults with GHD. Twenty-four-hour GH levels, as summarized by AUC, were inversely correlated with PAI-1 antigen AUC [Spearman correlation coefficient (rs) = –0.463; P = 0.023]. Alternatively, GH AUC was directly correlated with tPA activity AUC (rs = 0.440; P = 0.031) (Fig. 2).
|
|
Adults with GHD demonstrated an impaired fibrinolytic potential as defined by a sustained elevation in PAI-1 activity greater than 4 IU/ml after venous occlusion [7.2 IU/ml (0.8–17.4) vs. 0 IU/ml (0–2.8); P = 0.018]. PAI-1 antigen is consistently elevated in adults with GHD both before [14.1 ng/ml (7.1–23.3) vs. 5.0 (3.0–9.4); P = 0.035] and after [14.1 ng/ml (10.2–25.1) vs. 5.8 (2.4–7.9); P = 0.012] venous occlusion, compared with matched controls (Fig. 3). Adults with GHD have increased tPA antigen both before [6.5 ng/ml (4.2–9.7) vs. 3.6 (3.1–4.8); P = 0.025] and after [13.2 ng/ml (7.6–21.9) vs. 7.2 (4.3–10.7); P = 0.025] venous occlusion secondary to an increase in PAI-1/tPA complexes. (Data are available in 10 controls and 12 adults with GHD.) Neither adults with GHD (n = 12; mean difference 2.84, 95% CI –0.42–6.09; P = 0.082) nor matched controls (n = 9; mean difference 2.08, 95% CI –1.62–5.79; P = 0.231) exhibited a defective increase in tPA activity after venous occlusion.
|
Adults with GHD exhibited impaired endothelial function as demonstrated by reduced flow-mediated vasodilation [AUC 3.8 (–2.4 to 7.9) vs. 12.8 (2.1–19.4); P = 0.043] (Fig. 4). One pair of study participants was not able to complete the entire protocol secondary to inadequate venous access. They were able to complete the ultrasonography, however, and their results are therefore included in this analysis for a total of n = 13/group.
|
|
Discussion |
|---|
|
TopAbstractIntroductionPatients and MethodsResultsDiscussionReferences |
|---|
Other investigators have previously noted elevated levels of PAI-1 antigen and activity in adults with GHD (14, 15, 16, 22). The circadian variation of PAI-1 antigen, however, may confound results obtained at a single time point. Additionally, polymorphisms exist within the promoter of the PAI-1 gene, which directly influence circadian oscillation and result in increased levels of PAI-1 in the morning only (23). The results from our study are therefore novel in that they are the first to systemically define an altered fibrinolytic profile and potential in adults with GHD.
An inappropriate elevation in PAI-1 may be directly mediated by reduced secretion of IGF-I or GH or secondary to physical features characteristic of the GHD state. PAI-1 is synthesized by vascular endothelial and smooth muscle cells, hepatocytes, and visceral adipose tissue (24). Major determinants of PAI-1 plasma levels include genetic factors, specifically a single guanosine insertion/deletion at –675 bp upstream of the transcription start site, as well as gender, tobacco use, age, and the extent of visceral obesity (25). PAI-1 expression is up-regulated in the settings of hypoxia, hyperinsulinemia, infection, and estrogen deficiency. Glucose, very low-density lipoprotein, glucocorticoids, TNF-, and TGF-ß directly influence gene expression at specific regulatory elements within the PAI-1 promoter (13, 26). Angiotensin II induces the expression of PAI-1 within the vasculature; a correlation between plasma renin activity and aldosterone levels with plasma PAI-1 antigen levels has been observed in adults post-myocardial infarction and in controls on a low-salt diet (21).
PAI-1 is transcriptionally regulated by numerous growth factors; it is therefore possible that GH may act directly through a specific regulatory element to affect PAI-1 expression. GH has been reported to increase PAI-1 secretion in adipocytes in vitro (27). There is, however, a substantial body of evidence that suggests that the ability of GH to modulate PAI-1 is through effects on the nitric oxide (NO) system. An infusion of GH into the brachial artery significantly increases forearm blood flow and comparably decreases forearm vascular resistance in healthy subjects; these changes are accompanied by an increase in NO (28). Similarly, GH replacement in adults with GHD improves forearm vasodilator response (28, 29, 30). Short-term GH administration induces endothelial NO synthase (eNOS) expression in the vasculature of hypophysectomized rats (31). In vitro, treatment of cultured human endothelial cells with GH leads to a significant increase in eNOS gene and protein expression, resulting in enhanced NO release (32). Interestingly, animal models demonstrate that inhibition of NO synthase enhances vascular expression of PAI-1 (33). GHD itself has been described as a relatively NO-deficient state, whereas replacement with GH normalizes urinary nitrate as well as cGMP excretion and concomitantly improves peripheral arterial resistance (34). PAI-1 levels may therefore be inappropriately elevated in the GHD state through release of NO-mediated inhibition.
GHD is characterized as well by a low level of IGF-I. Experimental evidence has demonstrated that IGF-I is also able to directly activate eNOS via the serine/threonine kinase Akt (35). Thum et al. (32) further demonstrated that treatment of aged mice with IGF-I increases eNOS expression. Therefore, it also appears that an increase in IGF-I may contribute to enhancement of the NO system and alternatively that a low level of IGF-I may enhance PAI-1 production.
The GHD state itself is associated with numerous metabolic and physiognomic abnormalities that may contribute to elevated plasma PAI-1 levels. A deficiency in GH enhances the activity of the 11ß-hydroxysteroid dehydrogenase type II isoenzyme, which converts cortisone to cortisol. This leads to increased levels of cortisol in key organs, such as the liver and adipose tissue, which may then drive production of PAI-1 (36, 37). Adults with GHD demonstrated a propensity toward hyperinsulinemia as well as visceral obesity manifested in this study population by a higher waist to hip ratio. Additionally, inflammation as well as very low-density lipoprotein-mediated gene transcription may further enhance PAI-1 secretion; adults with GHD demonstrated a significant elevation in triglycerides and IL-10 (38).
Finally, study participants with GHD demonstrated a higher plasma renin activity and aldosterone than matched controls. This is compatible with the decrease in total body water and extracellular fluid volume previously described in adults with GHD. Plasma renin activity does not change after GH replacement, whereas aldosterone levels decrease (39, 40). An elevation in PAI-1 expression may therefore be partially mediated by an activated renin-angiotensin-aldosterone system secondary to the decrease in fluid volume observed in this population.
The current study was designed to control for many of the characteristics of the GHD state that may serve to drive expression of PAI-1 antigen. Specifically, controls were identical with respect to gender and closely matched for age and BMI. There were no significant differences between our control group and adults with GHD in the incidence of tobacco use, use of antihypertensive medication or estrogens, or family history of premature cardiovascular disease. None of our study participants reported a recent illness at the time of their study visit. All adults with hypopituitarism were taking stable and appropriate doses of glucocorticoid, thyroid, and sex steroid replacement at the time of their study visit. There may be, however, additional presently unrecognized variables that contribute to an impaired fibrinolytic balance for which we did not account. Additionally, the duration of GHD may influence the extent of fibrinolytic imbalance; this variable was difficult to accurately define in our participants.
The applicability of the current findings is limited because participants were largely young, otherwise healthy females. We would suggest, however, that this limitation actually lends strength to our findings. A young, otherwise healthy female population would be most expected to demonstrate normal or low levels of plasma PAI-1 antigen. Instead, the diagnosis of GHD alone imparts a significant alteration in the fibrinolytic profile. We hypothesize that increasing age, male gender, and other risk factors would further increase plasma PAI-1 levels (13). Further studies are needed to define the influence of gender and additional hormonal deficiencies on PAI-1 in adults with GHD.
In conclusion, this study demonstrates that the fibrinolytic potential is significantly altered in the setting of GHD. These changes may be secondary to various characteristics of the GHD itself or due directly to a decrease in GH and IGF-I. Pharmacological replacement with GH in this setting may not only restore fibrinolytic balance but also ultimately improve cardiovascular morbidity and mortality in this population. It is furthermore tempting to speculate that the age-mediated decline in GH secretion and concomitant increase in PAI-1 antigen may in part be responsible for the increased incidence of cardiovascular disease in physiological senescence.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
This study opened for enrollment in August 2005 and was therefore retroactively registered (www.clinicaltrials.gov) in November 2006 as NCT00397319.
Author Disclosure Summary: J.K.D. received grant support (August 2005) from Pfizer Pharmaceutical. L.S.B. consults for Auxilium Pharmaceutical and received lecture fees from Astellas Pharma Inc. and Unimed Pharmaceutical. D.K.V., Q.C., J.R.B., J.C., and D.E.V. have nothing to declare.
First Published Online June 19, 2007
Abbreviations: AUC, Area under the curve; BMI, body mass index; CV, coefficient of variation; eNOS, endothelial NO synthase; GCRC, General Clinical Research Center; GHD, GH deficiency; HDL, high-density lipoprotein; IQR, interquartile range; LDL, low-density lipoprotein; Lp(a), lipoprotein (a); NO, nitric oxide; PAI, plasminogen activator inhibitor; rs, Spearman correlation coefficient; tPA, tissue-plasminogen activator.
Received March 22, 2007.
Accepted June 13, 2007.
|
References |
|---|
|
TopAbstractIntroductionPatients and MethodsResultsDiscussionReferences |
|---|
activates the human plasminogen activator inhibitor-1 gene through a distal nuclear factor 
B site. J Biol Chem 279:18127–18136
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Endocrinology | Endocrine Reviews | J. Clin. End. & Metab. |
| Molecular Endocrinology | Recent Prog. Horm. Res. | All Endocrine Journals |
