Ischemic stroke patients within 12 hours from symptoms onset were prospectively included in the study between April 2013-July 2014. A cohort of control subjects matched by sex and age was included.
Inclusion criteria were: hospitalized patients with first-episode of IS within 12 hours from symptoms onset; age >18 years; previously independent for their daily living activities (modified Rankin Scale (mRS) ≤1).
Exclusion criteria were: presence of intracerebral hemorrhage confirmed by neuroimaging; previous IS; cancer or severe systemic disease that determine a life expectancy lower than 6 months; infections during the last 30 days before admission; chronic inflammatory disease; pregnancy; renal replacement therapy; treatment with steroids, immunosuppressive and immunomodulatory drugs or antibiotics during the last 30 days before admission; periodontal disease; and fever in the previous 72h (axillary temperature over 38ºC). Patients with active infection (axillary temperature > 37.5ºC and leukocyte levels >15000/μL or <4000/μL), cough and spitting, voiding dysfunction, diarrhea and clinical signs of endocarditis or meningitis.
On the other hand, a cohort of subjects without any neurological, inflammatory or infectious disease was included as control group. The selection of these control subjects was made by inviting the patient´s relatives to participate in the study. Control subjects were matched to patients by gender and age.
Clinical variables and Neuroimaging studies
All patients were admitted in the Stroke Unit of University Clinical Hospital of Santiago de Compostela and treated according to the guidelines of the Cerebrovascular Diseases Study Group of the Spanish Society of Neurology . Medical history recording demographic data, potential vascular risk factors, blood counts, biochemistry and coagulation tests, 12-lead ECG, chest radiography, carotid and transcranial ultrasonography and Computed Tomography (CT) or Magnetic Resonance Imaging (MRI) were performed at admission.
To evaluate neurologic deficit, the National Institute of Health Stroke Scale (NIHSS) was performed at admission, 24, 48 and 72 hours, at discharge, and at 3 months. END was defined as an increase of 4 points or more in NIHSS assessment between baseline and any other NIHSS evaluation during the first 72 hours. Functional outcome was evaluated at discharge and at 3 months by mRS. NIHSS and mRS were evaluated by internationally certified neurologists. Stroke etiology was classified according to TOAST criteria .
We evaluated the incidence of any infection during the hospitalization period. A protocol has been implemented in order to evaluate the presence of infections during the acute phase of stroke. The following tests were performed in those patients who showed an axillar temperature >37.5 ºC in two different determinations separated by 1 hour, or in patients with one axillar temperature determination >38ºC: blood counts, biochemistry analysis and blood culture; physicians made a clinical suspicion regarding the infection origin. During the etiological examination of the infection origin, empiric antibiotherapy was started according to clinical suspicion. Once the antibiogram was obtained, specific antibiotic treatment was started in case of positive cultures.
To evaluate infarct volume, a control CT was performed between 4th-7th days after IS. Infarct volume was quantified in cubic centimeters (cm3) and was assessed according to the formula 0.5xAxBxC, where A and B correspond to higher diameters in perpendicular direction and C to the number of 10 mm slices where infarct volume is present . All neuroimaging evaluations were made by the same neuroradiologist blinded to clinical and laboratory data.
Quantification of Treg
Circulating levels of Treg were measured by flow cytometry according to methods and using the markers described elsewhere [21-23]. Prior to patient´s inclusion, we selected 20 IS patients who matched inclusion/exclusion criteria to evaluate the optimal temporal profile for the quantification of Treg during the acute phase of IS. Blood samples were collected with an evacuated tube system (Vacutainer) in EDTA tubes at baseline, 24, 48, 72 hours and at days 4, 5 and 7. Based on to this temporal profile, we obtained blood samples in the more relevant time-points for Treg evaluation (i.e. at admission, 48 and 72 hours and 7th day).
Blood samples were processed within 3 hours after collection by a single researcher who was blinded to patients’ clinical, biochemical or radiological results. Circulating Treg were analyzed for the expression of specific surface antigens with direct flow cytometry (BD FACSAria IIu, BD, Franklin Lakes, NJ, USA). In brief, 50 μL of peripheral blood were labelled with 10 μL of FITC-conjugated anti-CD4 (BD, Franklin Lakes, NJ, USA), 10 μL of PE-conjugated anti-CD25 (BD, Franklin Lakes, NJ, USA), and 10 μL of Alexa Fluor® 647-conjugated anti-CD127 (BD, Franklin Lakes, NJ, USA) monoclonal antibodies. We considered Treg as CD4+/CD25+/CD127- staining cells in the lymphocyte gate. In all analyses, 2.5×105 events were acquired, using a FACSAria IIu analyzer (BD, Franklin Lakes, NJ, USA), and processed using the PC FACSDiva software program (BD, Franklin Lakes, NJ, USA). Treg count was expressed as percentage of Treg over total analyzed lymphocytes.
Blood samples, drawn from all patients at admission, and at 24±6, 48±12, and 72±12 hours, were collected in glass chemistry test tubes, centrifuged at 3000 rpm during 10 minutes, and immediately frozen and stored at -80 ºC. Serum levels of IL-10 were measured using an immunodiagnostic IMMULITE 1000 System (Siemens Healthcare España, Madrid, Spain). Determinations were performed in an independent laboratory blinded to clinical and neuroimaging data.
The primary endpoint was good functional outcome (mRS ≤2) at 3 months. Infarct volume and presence of END were evaluated as secondary outcome variables. The development of infections during hospitalization was recorded as safety variable. Finally, we analyzed the correlation between circulating Treg and serum levels of IL-10 in order to investigate the possible mechanism of action of Treg.
Sample size was calculated using the statistical EPIDAT 3.1 software, considering that those patients within the highest quartile regarding Treg levels during the first week after stroke achieve a 25% more frequency of good outcome at 3 months compared with those with Treg levels in the lowest quartile. The minimum calculated sample size was 172 patients in order to obtain a statistical power of 80% with a significant difference level of 0.05.
Results were expressed as percentages for categorical variables and as mean (SD) or median and range (25th and 75th percentiles) for the continuous variables depending on whether their distribution was normal or not. The Kolmogorov-Smirnov test was used for testing the normality of the distribution. Proportions were compared using the chi-square or Fisher test, while the continuous variables between groups were compared with the Student’s t or the Mann-Whitney tests depending on whether their distribution was normal or not, respectively. In case of more than 3 groups, variables were compared using ANOVA test. Bivariate correlations were performed using Pearson’s (normally distributed variables) or Spearman (variables without normal distribution) coefficients.
ROC curves were used to establish the best cut-off point for Treg levels that optimally predicted good functional outcome.
The independent association of circulating Treg levels with good functional outcome at 3 months and the risk of infections was assessed by logistic regression analysis; while their independent influence on infarct volume was assessed by multiple linear regression models. Each logistic regression analysis or multivariable linear regression model was adjusted for those significant variables in the bivariate analysis. Residual plots were examined to detect potential non-linear relationships between the outcome variable and continuous independent variables. Results were expressed as adjusted odds ratios (ORs) or Beta estimate with the corresponding 95% confidence intervals (95% CI). A p-value <0.05 was considered to be statistically significant in all tests. The statistical analysis was conducted in SPSS 20.0 (IBM, Chicago, IL, USA) for Mac.