Association of Phosphodiesterase 4D With Ischemic Stroke
A Population-Based Case-Control Study
Background and Purpose— The Phosphodiesterase 4D (PDE4D) gene was reported recently to be associated with ischemic stroke in an Icelandic population. The association was found predominately with large vessel and cardioembolic stroke. However, 2 recent reports were unable to confirm this association, although a trend toward association with cardioembolic stroke was reported. None of the reports included significant proportions of blacks. We tested for genotype and haplotype association of polymorphisms of the PDE4D gene with ischemic stroke in a population-based, biracial, case-control study.
Methods— A total of 357 cases of ischemic stroke and 482 stroke-free controls from the same community were examined. Single nucleotide polymorphisms (SNPs) were chosen based on significant associations reported previously. Linkage disequilibrium (LD), SNP, and haplotype association analysis was performed using PHASE 2.0 and Haploview 3.2.
Results— Although several univariate associations were identified, only 1 SNP (rs2910829) was found to be significantly associated with cardioembolic stroke among both whites and blacks. The rs152312 SNP was associated with cardioembolic stroke among whites after multiple comparison corrections. The same SNP was not associated with cardioembolic stroke among blacks. However, significant haplotype association was identified for both whites and blacks for all ischemic stroke, cardioembolic stroke, and stroke of unknown origin. Haplotype association was identified for small vessel stroke among whites.
Conclusions— PDE4D is a risk factor for ischemic stroke and, in particular, for cardioembolic stroke, among whites and blacks. Further study of this gene is warranted.
Stroke is the third leading cause of death and the leading cause of disability in the United States. Ischemic stroke represents the most common subtype of stroke and accounts for 80% of all strokes. In 2002, investigators of the Iceland DeCODE study reported linkage of the chromosome 5q13 region with ischemic stroke.1 In 2003, the same group of investigators reported that the phosphodiesterase 4D (PDE4D) gene within their region of interest was associated with ischemic stroke.2 The findings in both studies were particularly strong for carotid and cardiogenic stroke.
However, 2 recent reports were unable to confirm this association. Lohmussaar et al reported on 639 cases of stroke compared with 736 population-based controls and found no association between single markers or haplotypes of PDE4D with ischemic stroke.3 Nominal levels of significance were reported for a few SNPs and cardiogenic stroke, which were not significant after correction for multiple comparisons. Bevan et al reported that when comparing 737 white patients with ischemic stroke and 933 community-based controls, no association was identified for ischemic stroke overall.4 They reported that 6 of the 19 examined SNPs were associated with cardioembolic stroke and 2 with large vessel disease. They did not identify any association with carotid artery IMT or carotid plaque.
To date, no study has reported on the importance of the gene among blacks or in a US population. Blacks have a higher risk of stroke and, in particular, a higher rate of cardioembolic stroke than whites. We genotyped 6 single nucleotide polymorphisms (SNPs) of the PDE4D gene found to be significantly associated with ischemic stroke in the DeCODE study among a cohort of ischemic stroke patients and controls from the greater Cincinnati population to test the hypothesis that genotype or haplotype association of the region with ischemic stroke or subtypes of ischemic stroke exist.
The greater Cincinnati/northern Kentucky region includes 2 southern Ohio counties and 3 contiguous northern Kentucky counties that abut the Ohio River. Included in this area are 17 hospitals. This study was approved by the institutional review board at all participating hospitals.
We ascertained all strokes within the study population between January 1999 to December 1999. Study nurses screened the medical records of all inpatients with primary or secondary stroke-related International Classification of Diseases 9th revision (codes 430 to 436) discharge diagnoses from the 17 acute-care hospitals. Study nurses gathered information regarding symptoms, physical examination findings, past medical/surgical history, medication use before stroke, social history/habits, and emergency room evaluation, neurological evaluation, diagnostic test results, treatments, outcome, and type of insurance. Classification of race/ethnicity was as self-reported in the medical record.
In addition to the retrospective screening, we implemented a prospective case ascertainment technique. Admission logs of all local emergency departments were screened for stroke-like symptoms. Eligible patients were approached for consent for the genetic substudy. The chart was then abstracted. A study physician reviewed every abstract and all available neuroimaging to determine whether a stroke had occurred. The physician assigned stroke category and mechanism to each event based on all available information. See the online appendix (available at http://stroke.ahajournals.org) for details of subtype classification.
The Genetic and Environmental Risk Factors of Hemorrhagic Stroke (GERFHS) study (NS-36695) is a population-based case-control study of hemorrhagic stroke (1998 to 2002; 2003 to present). Cases of hemorrhagic stroke are invited to enroll into an interview and genetic sampling study. If they agree, cases are then matched to 2 population-based controls by age, race, and gender, which are identified by random digit dialing methods (response rate 49%).
Cases of ischemic stroke from the genetic substudy of the ischemic stroke study were matched to controls obtained for the hemorrhagic stroke study. Hemorrhagic stroke cases were excluded from analysis. A total of 482 controls had genotyping data available for analysis. This group of controls was significantly younger than the cases (66 versus 73 years). Thus, cases were matched to controls by age (±5 years), race, and gender. For the overall ischemic stroke analysis, 303 controls with matching demographics to cases were identified, and 54 controls were matched to 2 cases. In the subtype analyses, cases were compared with up to 3 controls randomly selected by age, race, and gender criteria from the overall pool of controls. Knowledge of risk factors including genotype or haplotype data were not used in selecting controls.
Ischemic stroke cases had genetic sampling through buccal swabs (n=174), blood samples (n=83), or both (n=193). Controls from the GERFHS study were exclusively sampled using buccal cell swabs. To ensure sufficient quantities of DNA for genotyping, all samples underwent whole genome amplification using the improved primer extension preamplification methodology.5 This procedure has been standardized in our laboratory.
The following polymorphisms were found to be significantly associated with ischemic stroke in the Iceland DeCODE study: SNP45, which was significantly associated, was found to be monomorphic in our population, rs1396476 (SNP89), rs2910829 (SNP87), rs966221 (SNP83), rs702553 (SNP56), and rs152312 (SNP41). Each polymorphism was genotyped using the TaqMan assay.
All polymorphisms were tested for conformation with Hardy-Weinberg expectations in both blacks and whites. Allele and genotype association was calculated using χ2 and trend tests. Case-control comparisons were performed using contingency tables that include a permutation test of 10 000 iterations, and the permuted P values are provided. Permutation testing provides a more conservative estimate than univariate P values. Multiple testing correction was performed using the Bonferroni method and given in the text where positive.
Linkage disequilibrium was calculated using HAPLOVIEW. The 95% CI of the D′ statistic did not cross 0 for any pair examined, and the entire region was combined as a 5-SNP haplotype. Haplotypes were constructed using HELIXTREE and the frequencies double-checked using HAPLOVIEW 3.2. HELIXTREE and HAPLOVIEW use an expectation-maximization method of constructing haplotypes. The parameters included exclusion of subjects with >25% missing genotypes.
Multivariate logistic regression was performed using backward elimination of risk factors found to be significant in univariate comparison (P<0.20 initially; P<0.10 thereafter) for ischemic stroke and the subtype of cardioembolic stroke. Analyses were race stratified.
Among the final sample size of 357 cases of ischemic stroke and 303 population-based controls for the total ischemic stroke comparisons, there were no significant differences in the percent black (29.4% versus 23.4%; P=0.08), percent female (56.3% versus 55.8%; P=0.89), or average age (69 versus 68 years; P=0.24). All SNPs were found to be in Hardy-Weinberg equilibrium. Genotyping results are presented in Tables 1 and 2⇓ for all ischemic stroke as well as the subtypes of ischemic stroke among whites and blacks, respectively. Although several univariate associations were identified, only 1 SNP (rs2910829) was found to be associated with cardioembolic stroke among both whites and blacks. The rs152312 SNP was associated with cardioembolic stroke among whites after multiple comparison corrections.
Haplotype construction results are presented in Tables 3 and 4⇓ for cases and controls of ischemic stroke divided by race. Significant association of the PDE4D region with ischemic stroke was identified for both whites (P<0.0001) and blacks (P=0.0004), which remained significant after adjustment for multiple testing (P=0.0002 and P=0.0008, respectively). In addition, the region appeared to be associated with cardioembolic, small vessel, and stroke of unknown or undetermined type among whites. Among blacks, haplotype association for cardioembolic stroke and stroke of unknown origin was identified. All of these findings were significant after multiple comparison correction. No association was identified for large vessel ischemic stroke. The haplotype GCTTC was found to be signif-icantly associated with ischemic stroke, cardioembolic, small vessel, and stroke of unknown type among whites. The same haplotype was associated with cardioembolic stroke among blacks.
Using multivariate analysis and controlling for the presence of hypertension, diabetes, hypercholesterolemia, previous heart disease, smoking (current and previous), first degree relative with stroke, body mass index, and frequent alcohol use, none of the polymorphisms were significantly associated with the total group of ischemic stroke among blacks and whites (data not shown). The same risk factors were considered for the subtype of cardioembolic stroke. Among whites, both rs152312T (P=0.011) and rs2910829 (P=0.0075) were significantly associated with cardioembolic stroke after controlling for all other significant risk factors. No significant association was found among blacks for any of the polymorphisms tested. In addition, we examined the association of each SNP with atrial fibrillation and found no significant association (data not shown).
We report association of the PDE4D region with ischemic stroke based on haplotype and genotype frequencies for both whites and blacks. Although each of the SNPs chosen were found to be associated with ischemic stroke in the Iceland DeCODE study, we were unable to replicate the majority of those findings. Indeed, only the rs152312 SNP was found to be associated with the total group of ischemic stroke among whites, and this difference was not significant after correction for multiple comparisons or in multivariate analysis. However, individual SNPs were found to be associated with cardioembolic stroke.
Haplotypes consist of the sequence of each polymorphism on the same chromosome. An overall haplotype association compares the total distribution of haplotypes among cases and controls. Using this methodology, significant haplotype association of the PDE4D region was identified for both whites and blacks, which remained significant after multiple testing corrections. The haplotype GCTTC was found to be significantly elevated among white cases compared with controls for total ischemic stroke as well as several subtypes of ischemic stroke and was also found to be significantly associated with cardioembolic stroke among blacks.
Table 5 presents the replication efforts regarding PDE4D and ischemic stroke to date. Lohmussaar et al identified “haplotype-tagging” SNPs across the PDE4D region and examined them among cases and controls from a central European region.3 They were unable to identify a positive association with ischemic stroke. However, only 1 of the 11 SNPs that they examined was positive in the original publication by Gretasdottir et al.2 Bevan et al examined SNPs that were originally identified as being part of “at-risk” haplotypes and were also unable to identify association with stroke overall.4 Additional SNPs were also examined, which were negative in result. Both of these studies largely examined SNPs that were negative previously, yet both reported trends toward association with cardioembolic and large vessel stroke for these previously negative SNPs.
Recently, Meschia et al examined linkage and association of the PDE4D gene with ischemic stroke using 249 subjects from 104 families and an additional 329 cases and 215 controls from a case-control study along with 48 cases and 48 controls from a stroke repository.6 Although they were unable to confirm linkage among their families, they did report significant SNP association. Finally, Nilsson-Ardnor, in a separate study of ischemic stroke within a Swedish population, identified linkage (LOD=2.06) at marker D5S424, which resides within the PDE4D gene.7 Although the level of linkage was only moderate, the report is an important replication of the importance of the gene to ischemic stroke.
It should be noted that few SNPs have been examined across >2 replication studies, and the results have varied by study regarding which SNPs are positively associated. This lack of consistency requires further study. The SNPs may be in linkage disequilibrium (LD) with a disease-causing SNP, and this level of LD may vary by population, leading to conflicting results. Alternatively, an interaction with another risk factor (genetic or environmental) may be required to lead to an increased risk of stroke.
The study by Meschia et al was the first to include a significant proportion of blacks. A limitation of their study design is that cases and controls were drawn from different populations, and, as described above, regional differences in allele frequency may affect the results. Our findings are consistent with theirs in respect to overall association. Our results were stratified by race given the different allele frequencies seen among blacks and whites. Recent studies suggest that race may not be an important factor in genetic effects. A review by Ioannidis et al found that of 43 gene-disease association studies reported in the literature that were stratified by race, 32 (74%) had odds ratios in the same direction for both groups,8 and large differences were only seen in 14% of the studies examined. Their conclusion was that genetic risk is usually consistent across racial groups. Our allele frequencies were very different by race, but separation by race may have led to a loss of power. Differences by race may therefore be attributable to type I or type II statistical error rather than true racial differences. Further studies with larger samples are needed to confirm these findings.
In conclusion, there are now several “positive” replication studies of the association of PDE4D with ischemic stroke. The negative studies that exist largely examined SNPs that were negative in the original PDE4D application. A consistent trend in the literature toward association with cardioembolic stroke appears to exist. Yet, no SNP has been consistently found to be positive across all studies suggesting that the “causative” mutation has not been identified. Further study of this gene is warranted.
Details of our subtype assignment and quality control results have been published previously. Our stroke subtype definitions were adapted from the Classification of Cerebrovascular Diseases III and epidemiologic studies performed in Rochester, Minnesota.
Definition of Subtypes
A. Cardioembolic: (Rochester Criteria) This category includes myocardial infarction within 6 weeks of stroke onset; acute congestive heart failure, mitral stenosis confirmed by clinical examination, echocardiography, or autopsy; artificial heart valve; atrial fibrillation or atrial flutter on electrocardiography; thrombus in the atrium, ventricle or on the aortic or mitral valve identified by echocardiography or coronary angiography; left ventricular aneurysm identified by echocardiography or coronary angiography; sick sinus syndrome identified by monitoring of cardiac rhythm. Patients with an akinetic or hypokinetic wall segment by echocardiogram also are included.
B. Large Vessel Stroke: This category required occlusion or 50% stenosis of the internal carotid artery by carotid ultrasound/duplex studies or 50% stenosis of the carotid, middle, anterior and/or posterior cerebral, vertebral or basilar arteries by angiography or magnetic resonance angiography in at least 1 plane that is in a vascular distribution consistent with stroke symptoms. Distinction between resolving embolism and primary disease of the intracranial vessel was made by the neuroradiologist.
C. Small Vessel Stroke: [Either condition a, b, or c is true] Condition A: Brain images show a deep infarct 1.5 cm in its maximal diameter that is appropriate to a clinical classical lacunar syndrome. Condition B: Brain images show no lesion to explain the clinical syndrome, and the clinical presentation is one (including the following) classically associated with a small deep infarct. Pure motor hemiplegia: hemiparesis or hemiplegia involving the face, arm and leg equally or arm and leg equally without other neurological findings. Although mild sensory symptoms can be present, there is no sensory loss on examination that is related to the infarct. Pure sensory stroke: isolated sensory loss or disturbance involving the entire hemiface and hemibody or the hemibody alone. There may be incidental motor weakness from another cause. Ataxia-hemiparesis: hemiparesis with ipsilateral ataxia. Paresis is more commonly crural. Dysarthria/clumsy hand syndrome: dysarthria with a clumsy hand. Facial weakness is possible. Hemiballismus, hemiathetosis or hemidystonia: must be acute onset. Sensorimotor stroke: weakness and sensory loss involving face, arm and leg equally without other neurological findings. Condition C: CT scan shows a deep infarct of 1.5 cm in its maximal diameter that is appropriate to the clinical syndrome, but the syndrome is not one of the classical syndromes for lacunar stroke.
D. Other Cause: Cerebral infarction that is caused by another clearly identified cause of stroke (e.g. traumatic arterial dissection, post coronary-bypass graft surgery, post-carotid endarterectomy, acquired immune deficiency syndrome (AIDS), cocaine use, etc.).
E. Ischemic stroke of uncertain cause: relatively rapid onset of a major focal neurological deficit that persists >24 hours or is fatal and cannot be attributed to another cause. This category is used when a patient does not meet any of the above criteria.
Quality Control Measures:
A study physician reviewed every possible case, and all available neuroimaging studies to determine whether a stroke or TIA had occurred. The physician assigned stroke/TIA category and mechanism to each event based on all available information, using definitions listed and previously reported. Agreement between nurse abstractors and physicians for TIA case/not a case was 93% (κ=0.84). A previous study has documented that residents of our 5-county study region exclusively seek care at the hospitals included in our study. [Broderick, 1996 #83].
Intraclass correlation for stroke subtype was 0.87 and κ ranged from 0.71 to 0.93 between individuals.
In addition the incidence of ischemic stroke study, subjects who were alive at the time of abstraction and deemed ‘a case’ by the nurse abstractor and a study physician were approached for enrollment. These subjects are the basis of the current manuscript. A total of 519 subjects were approached. Of these, 451 cases agreed to enrollment and 68 declined. Of the 451 cases who agreed, 23 patients died before discharge but were included into the study for analysis. In the overall incidence study, 2291 cases of ischemic stroke were identified.
This study was funded by the National Institute of Neurological Disorders and Stroke (R-O1-NS36695; R-O1-NS30678), the National Institute of Environmental Health Sciences (R-O1-ES-06096), and Elizabeth B. Lips Memorial Fund of the Greater Cincinnati Foundation.
- Received October 6, 2005.
- Revision received October 21, 2005.
- Accepted October 28, 2005.
Gretarsdottir S, Sveinbjornsdottir S, Jonsson HH, Jakobsson F, Einarsdottir E, Agnarsson U, Shkolny D, Einarsson G, Gudjonsdottir HM, Valdimarsson EM, Einarsson OB, Thorgeirsson G, Hadzic R, Jonsdottir S, Reynisdottir ST, Bjarnadottir SM, Gudmundsdottir T, Gudlaugsdottir GJ, Gill R, Lindpaintner K, Sainz J, Hannesson HH, Sigurdsson GT, Frigge ML, Kong A, Gudnason V, Stefansson K, Gulcher JR. Localization of a susceptibility gene for common forms of stroke to 5q12. Am J Hum Genet. 2002; 70: 593–603.
Gretarsdottir S, Thorleifsson G, Reynisdottir ST, Manolescu A, Jonsdottir S, Jonsdottir T, Gudmundsdottir T, Bjarnadottir SM, Einarsson OB, Gudjonsdottir HM, Hawkins M, Gudmundsson G, Gudmundsdottir H, Andrason H, Gudmundsdottir AS, Sigurdardottir M, Chou TT, Nahmias J, Goss S, Sveinbjornsdottir S, Valdimarsson EM, Jakobsson F, Agnarsson U, Gudnason V, Thorgeirsson G, Fingerle J, Gurney M, Gudbjartsson D, Frigge ML, Kong A, Stefansson K, Gulcher JR. The gene encoding phosphodiesterase 4D confers risk of ischemic stroke. Nat Genet. 2003; 35: 131–138.
Lohmussaar E, Gschwendtner A, Mueller JC, Org T, Wichmann E, Hamann G, Meitinger T, Dichgans M. ALOX5AP gene and the PDE4D gene in a central European population of stroke patients. Stroke. 2005; 36: 731–736.
Bevan S, Porteous L, Sitzer M, Markus HS. Phosphodiesterase 4D gene, ischemic stroke, and asymptomatic carotid atherosclerosis. Stroke. 2005; 36: 949–953.
Nilsson-Ardnor S, Wiklund PG, Lindgren P, Nilsson AK, Janunger T, Escher SA, Hallbeck B, Stegmayr B, Asplund K, Holmberg D. Linkage of ischemic stroke to the PDE4D region on 5q in a Swedish population. Stroke. 2005; 36: 1666–1671.