| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
(Stroke. 2005;36:9.)
© 2005 American Heart Association, Inc.
Original Contributions |
From the Department of Neurology (C.K., G.A.M., R.D., F.S., D.G.N., E.B.R., G.K.); the Department of Psychiatry (C.K.); the Institute of Clinical Chemistry and Laboratory Medicine (C.L., R.J.); the Institute of Arteriosclerosis Research (C.L., R.J., G.K.); and the Department of Epidemiology (K.B.), University of Münster, Germany.
Correspondence to Dr Carsten Konrad, Department of Psychiatry, University of Münster, Albert-Schweitzer-Str. 11, 48149 Münster, Germany. E-mail konradc{at}uni-muenster.de
| Abstract |
|---|
|
|
|---|
Methods Eighty patients with sCAD were compared with 80 age- and sex-matched healthy individuals.
1-antitrypsin (
1-AT) and
2-macroglobulin (
2-MG) levels, and
1-AT genotypes were assessed and compared between groups.
Results
1-AT and
2-MG levels as well as
1-AT genotypes did not differ significantly between patients and controls. The frequency of Z alleles in the patient group was higher than in the control group and than in other cohorts from Europe; however, the difference remained nonsignificant. All patients with Z alleles had internal carotid artery dissections.
Conclusions Overall, this data does not support the hypothesis that protease inhibitor levels or
1-AT genotypes play an important role in the etiology of sCAD. The present data does not exclude that the Pi-Z allele might have an influence on subgroups of sCAD, such as internal carotid artery dissections.
Key Words: alpha 1-antitryptsin alpha-macroglobulins dissection protease inhibitors risk factors
| Introduction |
|---|
|
|
|---|
2.6 per 100 000, spontaneous cervical artery dissection (sCAD) is a rare disease in neurological practice.1 Nevertheless, it is an important cause of stroke: 13% to 15.5% of strokes in adults <45 years2,3 and 30% to 40% of brain stem and cerebellar infarctions in this population occur on account of sCAD.4,5 Pathophysiologically, rupture of either the arterial intimal layer or the medial/adventitial layer, including vasa vasorum, causes an intramural arterial hematoma often leading to stenosis with a high risk of embolic brain infarction.6,7 Association of sCAD with heritable connective tissue diseases, such as Ehlers-Danlos syndrome type IV and Marfan syndrome,8,9 and ultrastructural connective tissue abnormalities in skin biopsies of sCAD patients6,10 suggest an important etiologic role of connective tissue aberrations in sCAD.
Recently, a number of case reports on patients with protease inhibitor deficiency and sCAD have been published, suggesting a possible etiologic role of protease inhibitor deficiency for sCAD.1113 In a small series of 22 patients with sCAD, 27.3% showed low levels of
1-antitrypsin (
1-AT),14 whereas no effect of
1-AT was observed in 35 patients with sCAD.15 Just recently, an investigation of
1-AT deficiency alleles in 74 patients did not suggest a causal relationship between
1-AT alleles and sCAD.16 For other arterial wall pathologies, an association of reduced antiproteolytic activity has repeatedly been observed, including patients with arterial aneurysms1719 or fibromuscular dysplasia.2022 On the basis of these observations, it has been postulated that an imbalance of proteolytic and antiproteolytic enzymatic activity might be a possible risk factor for sCAD.1113 However, larger studies on proteinases inhibitor levels and genotypes in patients with sCAD are missing so far.
The major inhibitors of human proteinases are
1-AT and
2-macroglobulin (
2-MG). The 2 most important genetic variants leading
1-AT deficiency are named S and Z alleles. The normal variant is called M allele. S and Z alleles are caused by point mutations leading to amino acid exchanges (S allele: glutamic acid 264 to valine; Z allele: glutamic acid 342 to lysine). Severe
1-AT deficiency is in the majority of cases associated with the Pi-ZZ genotype and causes severe damage of connective tissues in lungs, liver, and skin.23 In this study, we tested the hypothesis that reduced antiproteolytic enzyme activity may increase the risk of cervical arterial dissections by assessing
1-AT serum levels beyond the acute phase,
1-AT genotype, and
2-MG serum levels in a large group of patients with sCAD and age- and sex-matched healthy controls population.2430
| Methods |
|---|
|
|
|---|
Biochemical and Genetic Analysis
For biochemical and genetic analysis, venous blood samples were taken in the early morning after overnight fasting (12 hours). The parameters
1-AT and
2-MG were determined immunonephelometically with the Dade Behring BNII system using polyclonal antisera from rabbit against these 2 proteins. C-reactive protein (CRP) was measured with the Roche Hitachi 747 system using Roche high-sensitive CRP reagents. For our laboratory, normal values are 90 to 200 µmol/L for
1-AT, 130 to 300 µmol/L for
2-MG, and <0.5 mg/dL for CRP. DNA was extracted from EDTA-anticoagulated blood samples with magnetic beads by Tecan DNA sample preparation system and frozen until analysis at 20°C. The
1-AT genotype was determined with the Light Cycler (Roche Diagnostics) according to a method from Aslanidis et al31 using the following primers and probes from TIB MOLBIOL (Berlin, Germany): Pi*S allele: sense GGTGCCTATGATGAAGCGTTTAGGC; antisense AGGTGTGGGCAGCTTCTTGGTCA; probe TTCTTCCTGCCTGATGAGGGGAAACTA-fluorescein; reporter LC-red640-GCACCTGGAA
-ATGAAC-p and for the Pi*Z allele: sense TCCACGTGAGCCTTGCTCGAGGCCTG, antisense TTGGGTGGG
-ATCACCACTTTTC, probe CTCCAGGCCG-TGCATAAGGCTGT-fluorescein; reporter LC-red640-GACCAT- CGACGAGAAAGGG-p.
Statistical Analysis
Data were analyzed using SPSS for Windows, release 11.5.1. Patients and control subjects were matched according to age and sex. Baseline demographic characteristics of the study population were compared using
2 test for categorical data (with Fisher exact correction for cell sums <5) or Wilcoxon signed rank test for continuous data. CRP was classified in steps of 1 mg/dL.
1-AT and
2-MG levels were compared between groups using Wilcoxon signed rank test. In addition, we analyzed whether significant level differences between both groups existed after taking age and gender into account. This was done in 2 ANOVA models with
1-AT and log-10transformed
2-MG as dependent variables and case status, age (in years), and gender as independent variables. Assumptions of normality were checked before the use of the ANOVA models and not found to be violated. Colinearities between factors were examined using Spearman correlations and ANOVA models.
1-AT genotypes were compared between groups using
2 test with Fisher exact correction. Post hoc dissections were stratified into internal carotid artery (ICA) and vertebral artery (VA) dissections. The risk to experience ICA or VA dissection associated with the Pi-Z allele was determined using a binary logistic regression model.
| Results |
|---|
|
|
|---|
2 test with Fisher exact correction showed no significant differences between patients and controls concerning smoking, diabetes, and hypertension (P=0.24, P=1.00, and P=0.09, respectively). More than 70% of CRP levels were <0.5 mg/dL in both groups and >90% below 1 mg/dL. CRP did not significantly differ between patients and controls (Wilcoxon P=0.129). Sixteen patients could remember a minor trauma before the symptoms of dissection, including working with hyperextension of the neck, rapid neck movements, etc. Fourteen patients had cervical chiropractic manipulation to alleviate neck pain before the diagnosis of sCAD. SCAD occurred in 4 vessels in 2 patients (both vertebral and both ICAs), 3 vessels in 1 patient (both vertebral and left ICA), 2 vessels in 14 patients (6 both vertebral arteries, 1 left vertebral and basilary artery, and 7 both ICAs), in 63 patients in 1 artery (22 vertebral and 41 ICAs). Fifty-seven patients experienced completed ischemic infarction, 23 had transient ischemic or local compressive symptoms. Median time between sCAD and blood withdrawal was 35.3 months (mean 40.7±29.0).
|
1-AT and
2-MG
Median
1-AT levels were 130.5 µmol/L in the patient group and 124.5 µmol/L in the control group (Wilcoxon P=0.384). Median
2-MG levels were 159.5 µmol/L for patients and 146.5 µmol/L for controls (Wilcoxon P=0.359). All
1-AT levels <90 µmol/L were associated with either Pi-S or Pi-Z alleles.
1-AT correlated with
2-MG and with CRP (Spearman correlation coefficient rS=0.247, P=0.004 and rS=0.235, P=0.006, respectively). The time between sCAD and blood withdrawal had no effect on
1-AT or log-transformed
2-MG levels (ANOVA P=0.939 and P=0.926, respectively). Gender had a significant influence on
1-AT and log-transformed
2-MG levels (ANOVA P=0.007 and P=0.001, respectively), with male subjects having lower median
1-AT and
2-MG levels than female subjects (124.0: 131.0 µmol/L and 142.0: 179.0 µmol/L, respectively). Age, smoking habits, diabetes, or hypertension had no influence on
1-AT and log-transformed
2-MG concentrations.
1-AT serum levels were significantly influenced by the
1-AT genotype: subjects with MM-genotype had a median
1-AT level of 127.5 µmol/L, MS had 118.0 µmol/L, MZ had 80.6 µmol/L, and ZZ had 20.9 µmol/L (ANOVA P=0.001).
One of the main results is that
1-AT and
2-MG levels did not differ significantly between patients and controls (Table 2). After adjustment for age and gender, there was still no significant difference between cases and controls. Introduction of CRP as a covariate does not change this result. Pathological
1-AT values occurred in 8.6% of patients and 6.3% of controls, which were not significant in the
2 test. Pathological
2-MG values were measured in 17.9% and 27.5%, which were not significant in the
2 test either.
|
1-AT Genotypes
An equal number of Pi-MM genotypes was identified in patients and controls. The allele frequencies found in the control group are 0.9521 for Pi-M, 0.0342 for Pi-S, and 0.0137 for Pi-Z. The allele frequencies in the patients with sCAD were 0.9333 for Pi-M, 0.0267 for Pi-S, and 0.0400 for Pi-Z. There was an almost equal distribution of the Pi-MS genotype (Table 3). The overall distribution of
1-AT genotypes did not differ statistically between patients and controls (
2 test with Fisher exact correction P=0.839). However, the Z allele tended to occur slightly more often in patients: four patients had Pi-MZ and 1 patient ZZ genotype, compared with 2 controls with MZ-genotype. All patients with Z allele had sCAD of the ICA, 1 of them of both ICAs, none of them had VA dissection (Table 4). Post hoc stratification according to ICA- and VA-dissection did not show a significant effect of the occurrence of a Z allele on the occurrence of ICA- or VA-dissections (binary logistic regression).
|
|
| Discussion |
|---|
|
|
|---|
1-AT genotypes have so far been described in patients with arterial aneurysms and fibromuscular dysplasia. A number of case reports demonstrated an association
1-AT deficiency with arterial aneurysms.12,21,32,33 Protease-antiprotease imbalance was revealed by measurements of the elastase-
1-AT balance in tissue of aortic aneurysms and in the serum of patients with ruptured cerebral aneurysms.17,34 Genetic studies provided divergent results, some suspecting an association of
1-AT variants with arterial aneurysms,19 others questioning such an association.3537 For fibromuscular dysplasia, an association with
1-AT deficiency is also suspected.2022
For sCAD, 6 case reports and 1 case series suggest an etiologic role of
1-AT for spontaneous arterial dissections: dissection of the internal and external iliac arteries occurred in a 34-year-old man with Pi-SZ genotype,38 a dissecting hematoma of the left coronary trunk in an
1-AT-deficient 46-year-old woman,39 ICA dissections occurred in a 38-year-old woman and a 50-year-old male with Pi-MZ genotype,12,40 and sCAD with multiple aneurismal dilatations in a man with M1S genotype.33 Our group reported a 45-year-old male patient with Pi-ZZ genotype who had spontaneous ICA dissection with embolic middle cerebral artery occlusion and was successfully treated by systemic thrombolysis.11 Recently, an increased rate of lowered
1-AT levels was observed in a series of 22 sCAD patients,14 whereas no association of
1-AT with sCAD was found in 35 patients (16 in the acute phase, 19 in the convalescent phase).15 The reports cited above raise the possibility that reduced levels of protease inhibitors, such as
1-AT and
2-MG, might be a risk factor for sCAD. However, patient numbers studied so far are much too small to generalize for the results, and underlying genetics were only analyzed in single cases. One very recent investigation of
1-AT deficiency alleles in 74 patients did not find a relationship between
1-AT alleles and sCAD; however,
1-AT levels and other protease inhibitors were not investigated.16
This study investigates the protease inhibitor hypothesis, including a systematic genetic analysis in a relatively large number of patients, considering the low annual incidence of this disease.1 Because of this fact, a retrospective study design was chosen, accepting the limitation that only patients who were able and willing to give written informed consent and blood samples up to 10 years after sCAD could be included. A strength of this study is the well-matched control population originating from a large population-based study.2430 The allele frequencies found in the present control cohort are within the range estimations of 0.9272 to 0.9708 for Pi-M, 0.0176 to 0.0564 for Pi-S, and 0.0074 to 0.0153 for Pi-Z for different European regions.41 Because the regional provenance plays an important role, it is advantageous that the present control cohort was recruited within the same region of Germany (Westfalia) than the patient group. Rare
1-AT deficiency alleles and null alleles genotypes occurring with frequencies like 1.1x10(4) for Pi*Mmalton, 2.5x10(5) for Pi*Mcobalt, or 1.410(4) for all null alleles combined were not included in the genetic analysis. However, these variants lead to a dramatic reduction of
1-AT serum levels to 3% to 15% of their normal values or <1% for null alleles.42 That all
1-AT serum levels <90 µmol were associated with either Pi-S or Pi-Z alleles makes it unlikely that we overlooked one of the rare variants. Acute phase reactions leading to artificially increased
1-AT levels were avoided by testing patients >1 month after dissection. Nevertheless, the statistic power of the present sample is insufficient to test whether a combination of genetically determined protease inhibitor deficiency with acquired risk factors, such as smoking,43,44 recent infection,15 or reduced vitamin levels18 may increase the risk of sCAD.
In summary, this data does not support the hypothesis that protease inhibitor levels or
1-AT genotypes play an important role in the etiology of sCAD. The frequency of Z alleles in the patient group was higher than in the control group and than in other cohorts from Europe; however, the difference remained nonsignificant. Surprisingly, all patients with Z alleles had ICA and not VA dissections. Differences in the pathobiology of VA and ICA dissections have so far not been described. However, our group has reported that VA dissections are more often preceded by minor trauma-like chiropractic manipulations than ICA dissections, pointing toward different causative mechanisms.45 Future research might therefore consider VA and ICA dissections separately.
| Acknowledgments |
|---|
Received April 22, 2004; revision received August 10, 2004; accepted October 6, 2004.
| References |
|---|
|
|
|---|
This article has been cited by other articles:
![]() |
G Kuhlenbaumer, C Konrad, S Kramer, B Kis, D Nabavi, R Dittrich, and E B Ringelstein The collagen 1A2 polymorphism rs42524, which is associated with intracranial aneurysms, shows no association with spontaneous cervical artery dissection (sCAD) J. Neurol. Neurosurg. Psychiatry, January 1, 2006; 77(1): 124 - 125. [Full Text] [PDF] |
||||
![]() |
S. M. Rubinstein, S. M. Peerdeman, M. W. van Tulder, I. Riphagen, and S. Haldeman A Systematic Review of the Risk Factors for Cervical Artery Dissection Stroke, July 1, 2005; 36(7): 1575 - 1580. [Abstract] [Full Text] [PDF] |
||||
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
Stroke Home | Subscriptions | Archives | Feedback | Authors | Help | AHA Journals Home | Search Copyright © 2005 American Heart Association, Inc. All rights reserved. Unauthorized use prohibited. |