Association of antinuclear antibodies with the risk of intracranial arterial stenosis

Patient characteristics

The baseline demographic and clinical characteristics, laboratory findings are shown in Table 1. A total of 2,492 patients were included in our statistical analysis. There were 1056 (42.4%) males and 1436 (57.6%) females with a median age of 70 (63-78) years. The positive rates of ANAs were significantly different among the multiple IAS group, single IAS group and no IAS group (p<0.001).

Association between ANAs and IAS burden

Association between ANAs and IAS burden is shown in Table 2. Multivariate logistic regression analysis of the overall subjects showed that although ANAs was associated with the single IAS (Adjusted OR1=1.451,95% CI=1.142-1.843, p=0.002), the association did not reach statistical significance after adjustment for all potential confounders (adjusted OR2=1.252, 95%CI=0.920-1.705, P=0.153). After adjusting for all confounding factors, compared with the ANAs-negative patients, the adjusted OR for two or more IAS in ANAs-positive patients was 3.737 (95% CI=2.676-5.220, p<0.001).

Subgroup analysis in different gender groups

Subgroup analysis in different gender groups showed that the risks for two or more IAS in ANAs-positive male and female subgroups were 3.142 (95%CI=1.886-5.234, P<0.001) and 4.525 (95%CI=2.841-7.207, P<0.001) times higher, respectively, compared with the ANAs-negative patients. However, among the male and female subgroups, there was no significant association between ANAs positivity and a single IAS (Table 3).

Subgroup analysis in different age groups

Subgroup analysis according to age showed that the associations between ANAs and IAS burden were not the same in different age groups. ANAs positivity was significantly associated with single and multiple IAS in individuals ≤ 60 years (adjusted OR2=3.855, 95% CI=1.625-9.144, p=0.002 for single IAS; adjusted OR2=5.655, 95%CI=1.574-20.324, p=0.008 for two or more IAS). However, ANAs positivity was only associated with two or more IAS in individuals aged between 61 to 75 years (adjusted OR2=4.393, 95% CI=2.674-7.216, p<0.001) and those older than 75 years old (adjusted OR2=4.083, 95% CI=2.245-7.423, p<0.001) (Table 4).

Dose-effect relationship between ANAs and IAS burden

To further validate the association between ANAs and IAS, we analyzed the association between the number and grade of ANAs and IAS burden in the 572 ANAs-positive patients. The distribution and grades of positive ANAs are shown in Supplementary Figure 1. Defined ANAs-Count as the sum of ANAs positivity items in ANAs-positive individuals; defined ANAs-Max as the highest grade of ANAs positivity items in ANAs-positive individuals (+=1, ++=2, +++=3).

The IAS burden was taken as a dependent variable and the ANAs-Count as well as ANAs-Max were taken as independent variables. After adjusting for all confounding factors, multivariate logistic regression analysis showed that there was no significant association of ANAs-Count and ANAs-Max with single or multiple IAS in the 572 ANAs-positive patients (Table 5).

Study participants

Study subjects with acute ischemic stroke (AIS) (<7 days of onset) or transient ischemic attack (TIA) confirmed by neuroimaging were prospectively recruited from September 2014 to February 2018 in the department of neurology at Qingdao Municipal Hospital. The diagnosis of AIS and TIA conforms to the standards of the American Heart Association/American Stroke Association [22, 23]. We excluded individuals who:(1) less than 40 years old; (2) had undergone incomplete vascular imaging and laboratory tests; (3) had severe stenosis (≥70%) or occlusion confirmed by doppler ultrasonography of extracranial carotid artery, brachiocephalic trunk, subclavian artery, and extracranial vertebral artery; (4) had undergone intracranial or extracranial cerebral artery stenting or balloon angioplasty; (5) had undergone carotid endarterectomy; (6) were diagnosed with moyamoya disease; (7) had been clinically diagnosed with cerebral embolism; (8) did not provide written informed consent.

On admission, the information was collected and recorded in detail, such as baseline data, age, gender, blood pressure, and risk factors for ischemic stroke. In addition to routine blood tests and medical history collection, ANA test and MRA were performed in patients who agreed and signed an informed consent form as well. This study was approved by the Institutional Ethics Committee of Qingdao Municipal Hospital. Finally, 2492 patients were included and divided into three groups according to the vascular lesion burden showed by MRA.

Ethics statement

This study was conducted in accordance with the Helsinki declaration and the protocol of this study was approved by the Institutional Ethics Committees of Qingdao Municipal Hospital. The written informed consent was obtained from all studied participants directly or from the authorized person.

Definition of diseases and risk factors

Hypertension: defined as a history of hypertension or required regular treatment with antihypertensive agents before admission or diagnosed at discharge. Diabetes mellitus: defined as a history of diabetes mellitus or being treated for diabetes mellitus or glycosylated hemoglobin≥7% or diagnosed at discharge. Dyslipidemia: history of dyslipidemia, or currently taking lipid-lowering drugs for hyperlipidemia, or fit the following criterion: the serum total cholesterol (TC) >5.2mmol/L, or low-density lipoprotein (LDL) > 3.37mmol/L, or high-density lipoprotein (HDL) < 1.04 mmol/L, or triglyceride > 1.7mmol/L. Hyperuricemia was defined as serum uric level >357ummol/L or currently taking uric-acid-lowering drugs. History of previous stroke: once suffering from cerebral infarction or cerebral hemorrhage. Coronary heart disease: previous history of coronary heart disease or diagnosed at discharge. Atrial fibrillation: patients with atrial fibrillation diagnosed by electrocardiogram or previously diagnosed as paroxysmal atrial fibrillation. In addition, smoking habit was defined as a patient who had smoked continuously for 6 months ≥1 cigarette a day. Drinking habit was defined as the average alcohol consumption on average for more than half a year from past or present ≥2 units/d (16g/d).

Detection of ANA and blood biochemical analysis

Three milliliters of fasting venous blood samples were drawn with a non-anticoagulant vacuum tube in the morning, and were preserved at -80°C until antinuclear antibody (ANA) and ANAs tests were done. As recommended by the manufacturer, all the blood samples were screened for ANA with indirect immunofluorescence technique (IFAT) on HEp-2 cells by using the commercially available kit ANA HEp-2 (HEp-2 cells, Euroimmun Hangzhou medical laboratory diagnostics GmbH, Hangzhou, China). ANA titers of at least 1:100 were considered positive. Simultaneously, blood samples were identified by immunoblotting for ANAs spectrum (anti-nRNP, Sm, SS-A, Ro-52, SS-B, Scl-70, PM-Scl, Jo-1, CENP-B, PCNA, dsDNA, Nukleosome, Histone, Ribosomal P protein, AMA-M2) using individual ENA kit (Euroimmun Hangzhou medical laboratory diagnostics GmbH, Hangzhou, China). According to the instructions, the procedures were carried out and the results were interpreted based on the degree of staining of the antigen colored ribbons. As recommended by the manufacturer, the results are assessed to be (-), (+), (++) (+++). Patients who are positive for the ANA screening test and/or are positive for one or more of the specific ANAs are considered to be ANAs-positive individuals. Blood biochemical analysis was conducted by BECKMAN COULTER (AU5800) automatic biochemistry analyzer (Beckman Coulter, Inc. Brea, USA).

Imaging assessment

All the patients underwent conventional MRI on a 3.0-T MR scanner (Signa Excite HD 3.0T system, GE Medical Systems, USA) within two weeks following admission. MRI sequences included three-dimensional time of flight magnetic resonance angiography (3D TOF MRA), T2/ T1-weighted imaging, fluid-attenuated inversion recovery sequences (FLAIR), and diffusion-weighted imaging (DWI). MR angiograms were obtained using the following imaging parameters: repetition time/echo time (TR/TE), 21/3ms, flip angle 15°, the field of view (FOV) 240mm×240mm, matrix 320×160, and slice thickness 0.7mm without interslice gaps. All MRI/MRA images were stored in digital format and were read independently by two readers (Yongan Sun and Lin Tong) who were blinded to subjects’ clinical information. A third appointed senior reader would decide the final score if there were disagreements of larger than 10% on the degree of stenosis.

The following arterial segments were assessed: bilateral intracranial internal carotid arteries (ICA), middle cerebral arteries (MCA) M1/M2, anterior cerebral arteries (ACA) A1/A2, posterior cerebral arteries (PCA) P1/P2, intracranial vertebral arteries (V4) and basilar artery (BA). Based on the original and reconstructed images, the severity of intracranial stenosis on MRA was classified into four grades (no or <50% stenosis, 50% to 69% stenosis, 70% to 99% stenosis, and occlusion) using the published method Warfarin-Aspirin Symptomatic Intracranial Disease (WASID) study [24]. Focal flow void found on MRA with distal filling was regarded as severe stenosis (70%–99%). Absence of distal filling on MRA would be regarded as 100% occlusion. For the internal carotid artery, the intracranial location was defined as the distal end of the ophthalmic artery [25]. A significant artery stenosis lesion was defined as ≥50% stenosis or occlusion of the major intracranial cerebral arteries. IAS burden was defined as the total number of intracranial arteries with significant stenosis (≥50% stenosis) [26]. According to their IAS burden, the patients were divided into the group with no significant stenosis and the group with significant stenosis. The latter was further divided into single stenosis group and multiple stenosis group.

Statistical analyses

Continuous variables that conform to a normal distribution, were expressed as mean ± standard deviation; continuous variables that do not conform to a normal distribution, were expressed as median and interquartile range (IQR). Continuous variables were compared by Wilcoxon test and categorical variables were compared by Chi-square test. In the regression analysis, if the parallel test hypothesis was rejected, a multinomial logistic regression model was used. In logistic regression model, OR1 was adjusted for stroke risk factors (age, gender, mean arterial pressure, stroke history, hypertension, diabetes mellitus, coronary heart disease, atrial fibrillation, hyperuricemia, smoking habits and drinking habits), and OR2 was adjusted for the items in OR1 plus blood laboratory results (blood glucose, serum uric acid, serum homocysteine, serum triglyceride, serum cholesterol, serum high density lipoprotein, serum low density lipoprotein, neutrophil and lymphocyte counts). All P values were 2-sided and evaluated at the 0.05 α level. All analyses were performed in Stata 12.0 (Stata Corp, College station, TX).