Increased level of FAM19A5 is associated with cerebral small vessel disease and leads to a better outcome

Subjects

Cerebral infarction patients and healthy individuals admitted to the Institute of Cerebrovascular Diseases Cooperative Medical Institutions from May 2018 to June 2020 were continuously enrolled in this study. RSSI is defined as a recent deep small infarction caused by a single perforating artery occlusion, leading to ischemic necrosis of the brain tissue in its blood supply area. The inclusion criteria were as follows: RSSI patients showing typical cerebral infarction symptoms and physical signs with ischemic focus <20 mm in brain MRI. The exclusion criteria were as follows: patients with carotid or intracranial artery stenosis >50%, cardiogenic stroke, previous cardiocerebrovascular disease, severe liver or kidney function injury, and inflammation-related diseases, such as rheumatic immune disease and tumor. This study was conducted following the Declaration of Helsinki, was approved by the Ethics Review Board of the Affiliated Hospital of Qingdao University (QYFY WZLL 26787), and informed consent was taken from all individual participants.

Clinical analysis of subjects

Data, including age, sex, vital signs (blood pressure and pulse), history (hypertension and diabetes), personal history (smoking and drinking history), and basic laboratory examinations (glucose, triglyceride, total cholesterol, low-density lipoprotein, high-density lipoprotein, white blood cell count, and uric acid), were collected from medical records of all the subjects at the time of registration. All the healthy controls underwent electrocardiography (ECG) and brain imaging examinations (head computed tomography (CT) scan or head MR scan), while the RSSI patients received the following examinations: cerebral MRI, carotid artery examination, and cerebrovascular examination (arterial CTA or MRA), cardiac ultrasound, and 24-h Holter electrocardiogram. A total of 146 RSSI patients completed the head susceptibility-weighted imaging (SWI) examination. RSSI patients underwent laboratory examinations with respect to C-reactive protein, high sensitivity C-reactive protein, and homocysteine. We also evaluated the patients’ NIHSS score at admission and 14 days and calculated ΔNIHSS (NIHSS at admission–NIHSS at 14 days) and percentage of NIHSS improvement ((NIHSS at admission–NIHSS at 14 days)/NIHSS at admission). Also, the total MRI burden of patients in the RSSI group was estimated.

Serum analysis for FAM19A5 levels

Fasting blood was withdrawn within 24 h of admission, and the serum was collected by centrifugation at 1,000× g for 15 min and stored at −80 °C. The serum FAM19A5 level of all patients was determined by a double sandwich enzyme-linked immunosorbent assay kit (Jonln Biology, Wuhan, China), with inter-assay CV% < 10%, and intra-assay CV% < 15%, using 50 µL of each sample with no sample dilution. We strictly follow the operating instructions: the standard, specimen, and HRP-conjugate detection antibody were added to the microwells pre-coated with the FAM19A5 capture antibody in sequence. After incubation and washing thoroughly, the color was developed by TMB, and the optical density was measured at 450 nm on a microplate reader (Molecular Devices, San Jose, CA, USA).

Radiological assessment

All RSSI patients were examined using 3.0T MRI, and the images were read and analyzed by two professional neurologists; any differences were resolved through discussion with a third analyst. The volume of cerebral infarction was calculated using the ABC/2 method (Sims et al., 2009). Each imaging characteristic of the cSVD (WMH, EPVS, lacune, CMB) was evaluated separately. The lacune of presumed vascular origin has defined as a round or ovoid, subcortical, fluid-filled cavity (signal similar to CSF) between 3 mm and 15 mm in diameter, consistent with a previous acute small subcortical infarct or hemorrhage in the territory of one perforating arteriole. EPVS has defined as fluid-filled spaces that follow the typical course of a vessel as it goes through grey or white matter. The spaces have a signal intensity similar to CSF on all sequences. According to Potter scale, we counted the number of enlarged perivascular spaces and divided them into 0, 1, 2, 3, and 4 grades (0, 1–10, 11–20, 21–40, ≥40, respectively) (Potter et al., 2015). WMH has been defined as a signal abnormality of variable size in the white matter that shows the following characteristics: hyperintensity on T2-weighted images (T2WI) such as fluid-attenuated inversion recovery without cavitation (signal different from CSF); WMH was divided into periventricular WMH (PVWMH) and deep WMH (DWMH) that were graded by Fazekas scale, respectively (Wardlaw et al., 2013b), Only RSSI patients with head SWI were evaluated for CMB, which was defined as a small focal round lesion (diameter <10 mm), showing a low signal on SWI. The CMB was classified into lobar CMB and deep CMB. The presence of lacune, the presence of CMB, the number of EPVS ≥11, Fazekas scale ≥2, and each parameter assessed as one score were included in the total MRI burden of cSVD.

Statistical analysis

The normality of quantitative data distribution was assessed using the Kolmogorov–Smirnov test. Continuous variables with normal distribution were described as mean ± standard deviation (SD), continuous variables with non-normal distribution were described as median (interquartile range (IQR)), and qualitative data were described as quantity (percentage). Multivariable logistic regression analysis compared the difference in FAM19A5 expression between the RSSI group and healthy controls, while multivariable linear regression analysis compared the correlation between FAM19A5 and the infarct volume. In order to verify the correlation between the imaging characteristics of total MRI burden and FAM19A5, we first determined the confounding factors by univariate logistic analysis, and those with P-value < 0.1 were included in multivariable logistic analysis for adjustment. Due to the high incidence of WMH and EPVS, the variables were classified into different grades that were analyzed by multivariate ordinal logistic regression. Next, lacune and CMB were analyzed by multivariate binary logistic regression with the classification into the presence and absence. Deep and lobar CMB and, as well as PVWMH and DWMH were independently analyzed by multivariate logistics. P < 0.05 was considered statistically significant. In the RSSI group, the correlation between FAM19A5 levels and clinical scales (NIHSS scale at NIHSS scale at admission, NIHSS scale at 14 days, ΔNIHSS, percentage of NIHSS improvement, and total MRI burden of cSVD) were assessed, in which Spearman’s test performed the comparisons between continuous variables. All statistical analyses were performed using IBM SPSS Statistics for Windows, version 26.0 (IBM Corporation, Armonk, NY, USA).

Recruitment and baseline characteristics of subjects

A total of 609 subjects were selected in our study from 1784 RSSI patients and healthy individuals from May 2018 to June 2020 (Fig. 1). There were 1,342 acute infarction patients were included in the filter process; among them, 573 patients with severe large artery stenosis and occlusion were excluded, and 188 patients were excluded because of cardiac diseases or other diseases that may lead to cerebral infarctions. Finally, 344 RSSI patients were included in the current study, the mean age was 61.874 ± 11.359 years, and 211 (63.2%) patients were males. Also, 265 healthy controls were strictly filtered from 541 patients undergoing health checkups, the mean age was 61.411 ± 11.978 years, and 115 (43.4%) patients were males (Table 1).

FAM19A5 level and infarct volume of RSSI

The current study showed a positive correlation between FAM19A5 and recent small subcortical infarct. Mean ± standard deviation of serum FAM19A5 in the RSSI and healthy control groups was 681.894 ± 1225.014 and 432.582 ± 777.871 pg/mL, respectively (Fig. 2). After adjustment for age, sex, hypertension, diabetes, smoking, pulse, triglyceride, cholesterol, low-density lipoprotein, and glucose, our results showed that the expression of FAM19A5 was higher in the RSSI group than the control group (odds ratio (OR) (95% confidence interval (CI)) = 1.265 (1.032–1.549), P = 0.023). The linear regression analysis of infarct persons showed the correlation between infarct volume and FAM19A5 (OR (95% CI) = 1.587 (1.324–1.901), P < 0.001) (Table 2).

FAM19A5 and imaging characteristics of total MRI burden

Serum FAM19A5 levels in imaging characteristics of cSVD total MRI burden are depicted in Fig. 3. A positive correlation between the total MRI burden of cSVD and FAM19A5 levels was established (P = 0.041) and assessed by univariate and multivariable regression analysis. (Table 3) Age, sex, hypertension, diabetes, smoking, drinking, pulse, blood pressure, volume of cerebral infarction, glucose, C-reactive protein, triglyceride, total cholesterol, low-density lipoprotein, high-density lipoprotein, white blood cell count, uric acid, and homocysteine were included in the adjustment process.

After adjustment for age, hypertension, pulse, systolic blood pressure, total cholesterol, low-density lipoprotein, C-reactive protein, and high sensitivity C-reactive protein, multivariate ordinal logistic regression analysis showed a positive correlation between FAM9A5 and white matter hypertension on Fazekas scale (OR (95% CI) =1.842 (1.459–2.326), P < 0.001). (Table 3) Our further analysis showed that FAM19A5 was correlated with both periventricular white matter hypertensity (OR (95% CI) = 1.560 (1.311–1.855), P < 0.001) and deep white matter hypertensity (OR (95% CI) = 1.557 (1.268–1.910), P < 0.001).

After adjustment for age, sex, hypertension, smoking, total cholesterol, and low-density lipoprotein, multivariate ordinal logistic regression analysis showed a positive correlation between FAM9A5 and EPVS Potter scale (OR (95% CI) = 1.434 (1.211–1.698), P < 0.001).

The current study did not identify any significant correlation between FAM9A5 and the presence or absence of lacunae (OR (95% CI) = 0.980 (0.801–1.199), P = 0.844) and CMB (OR (95% CI) = 0.895 (0.659–1.216), P = 0.479). We also found that deep microhemorrhage and cortical microhemorrhage were not related to the FAM19A5 level.

Correlation between FAM19A5 and outcomes

The mean ± SD of NIHSS at admission, NIHSS at 14 days, and ΔNIHSS was 3.22 ± 2.88, 2.52 ± 2.66, and 0.69 ± 1.27, respectively, and the mean ± SD of the percentage of NIHSS improvement was 25.4 ± 48.5. Spearman’s correlation analysis demonstrated that the FAM19A5 level was not significantly correlated with NIHSS at admission and 14 days after admission. However, it was positively correlated with ΔNIHSS (P = 0.031) and the percentage of NIHSS improvement (P = 0.011) (Table 4).