- Email: [email protected]
Plasma Immunoproteasome Predicts Early Hemorrhagic Transformation in Acute Ischemic Stroke Patients
ARTICLE IN PRESS
Plasma Immunoproteasome Predicts Early Hemorrhagic Transformation in Acute Ischemic Stroke Patients Xingyong Chen,
MD, PhD,
Yinzhou Wang, MD, PhD, Ming Fu, Qiong Cheng, MM, and Xu Zhang, BM
BM,
Huixin Lei,
BM,
Background and Purpose: Currently, blood biomarkers associated with an increased hemorrhagic transformation (HT) risk remain uncertain. We aimed to determine the significance of immunoproteasome as predictors of early HT in acute ischemic stroke patients. Methods: This study enrolled 316 patients with ischemic stroke. HT was assessed by computed tomography examination performed on day 5 ± 2 after stroke onset or immediately in case of clinical deterioration (CD). Plasma immunoproteasome subunits low molecular mass peptide 2 (LMP2), multicatalytic endopeptidase complex-like 1 (MECL-1), LMP7, interleukin-1β (IL1β), and high-sensitivity C-reactive protein (Hs-CRP) were measured with quantitative sandwich enzyme-linked immunosorbent assay kits. Factors associated with HT were analyzed using a multivariate logistic regression analysis. Results: There were 42 (13.3%, 42 of 316) patients who experienced HT. Compared with those patients without HT, plasma LMP2, MECL-1, LMP7, IL1β, and Hs-CRP concentrations on admission were significantly increased in patients with subsequent HT (P < .001). These protein concentrations increased with hemorrhage severity. Patients with CD caused by HT had the highest levels of LMP2 (1679.5 [1394.6136.6] pg/mL), MECL-1 (992.5 [849.7-1075.8] pg/mL), LMP7 (822.6 [748.6-1009.5] pg/mL), IL1β (113.2 [90.6-194.5] pg/mL), and Hs-CRP (30.0 [12.8-75.6] mg/L) (P < .05). Logistic regression analysis showed cardioembolism, LMP2, MECL-1, and LMP7 as independent predictors of HT (P < .05). Receiver operating characteristic curve analysis demonstrated LMP2 ≥ 988.3 pg/mL, MECL-1 ≥ 584.7 pg/mL, and LMP7 ≥ 509.0 pg/mL as independent factors associated with HT (P < .001). Conclusion: Evaluation of plasma levels of immunoproteasome could be helpful in the early prediction of HT in acute ischemic stroke patients. Key Words: Immunoproteasome—hemorrhagic transformation—cerebral infarction—risk factors. © 2016 National Stroke Association. Published by Elsevier Inc. All rights reserved.
From the Department of Neurology, Fujian Provincial Hospital, Fujian Medical University Provincial Clinical College, Fuzhou, China. Received July 21, 2016; revision received August 9, 2016; accepted August 17, 2016. Xingyong Chen and Yinzhou Wang contributed equally to this work. Author contributions: Xingyong Chen and Yinzhou Wang contributed in the study concept, acquisition of data, analysis and interpretation of the data, statistical analysis, and drafting and revision of the manuscript content. Ming Fu contributed in the acquisition and analysis of data. Qiong Cheng and Huixin Lei participated in the design of the study. Xu Zhang contributed in the revision of the manuscript content. Grant support: This study is supported by grants from Fujian Province Natural Science Fund (2013J01275, 2016J01432) and Young and Middle-aged Talents Training Project of Health and Family Planning Committee of Fujian Province (2015-ZQN-JC-5). Address correspondence to Yinzhou Wang, MD, PhD, and Xu Zhang, BM, Department of Neurology, Fujian Provincial Hospital, Fujian Medical University Provincial Clinical College, Fuzhou 350001, China. E-mail: [email protected]; [email protected]. 1052-3057/$ - see front matter © 2016 National Stroke Association. Published by Elsevier Inc. All rights reserved. http://dx.doi.org/10.1016/j.jstrokecerebrovasdis.2016.08.027
Journal of Stroke and Cerebrovascular Diseases, Vol. ■■, No. ■■ (■■), 2016: pp ■■–■■
1
ARTICLE IN PRESS X. CHEN ET AL.
2
Background Hemorrhagic transformation (HT) is bleeding into an area of ischemic brain after stroke. It occurs in as many as 10%-40% of patients with ischemic stroke.1,2 HT increases morbidity and mortality of acute ischemic stroke. It is difficult to predict a given patient’s probability of experiencing HT. Currently, several clinical parameters and blood biomarkers are known to be associated with an increased HT risk. Factors such as age, hypertension, atrial fibrillation (AF), baseline National Institutes of Health Stroke Scale (NIHSS) score, treatment with antiplatelet or thrombolytic agents, the presence of early ischemic changes on cranial computed tomography (CT) on admission, high levels of plasma matrix metalloproteinase (MMPs), activated protein C, estimated glomerular filtration rate, tight-junction proteins, and S100B have been related to HT after ischemic event.1-12 However, all studies in this field to date did not definitively elucidate the clinical relevance of any single marker or a panel of different markers. Immunoproteasome is a subtype of proteasome that contains three major catalytic subunits: β1i (also known as low molecular mass peptide 2 [LMP2]; proteasome subunit beta 9 [PSMB9]), β2i (also known as multicatalytic endopeptidase complex-like 1 [MECL-1]; PSMB10), and β5i (also known as LMP7; PSMB8).13 Immunoproteasome is present in several kinds of cells, including neurons, astrocytes, microglia, and endothelial cells in the brain areas.14,15 Our previous study had confirmed that augmentation of both immunoproteasome LMP2 and LMP7 is potentially involved in the inflammatory pathophysiological mechanism of ischemia stroke, and selective inhibition of the immunoproteasome subunit LMP7 offers a strong neuroprotection in middle cerebral artery occlusion (MCAO) rats.14 However, detailed knowledge on the relationship of immunoproteasome with human ischemia stroke is still not available. The aim of this prospective study in consecutive patients was therefore to assess (1) the rate of early HT in patients admitted for ischemic stroke, (2) the correlation between plasma immunoproteasome with early HT, and (3) the risk factors for early HT.
Subjects and Methods Patients We prospectively enrolled 316 patients with a first episode of ischemic stroke admitted within the first 72 hours of symptom onset to the neurology department of our hospital between March 2014 and September 2014. The study was approved by the local ethics committee and informed consent was obtained from each patient or a family member or relative. A diagnosis of acute ischemic stroke was made by stroke neurologists and was confirmed with CT or magnetic resonance imaging. A control group of 10 healthy subjects (men, 60%; mean age, 62.5 ± 11.9 years) was also studied. The exclusion criteria were patients who
received oral anticoagulants or heparin, with previous stroke, transient ischemic attack, or intracerebral or subarachnoid hemorrhage, inflammatory or infectious diseases, epilepsies, hematologic diseases, cancer, and severe liver and renal failure. Those who were previously dependent were excluded from the present study. In each case, past and present clinical histories, including medication use, were obtained by interviewing the patient or a family member in cases where the patient was aphasic or unconscious. Each patient’s hospital records were also reviewed for medical history. All patients underwent comprehensive clinical evaluation, blood pressure measurement, standard electrocardiogram (ECG), chest X-ray, blood routine, and fibrinogen and biochemical analyses. Stroke etiology was classified according to the Trial of Org 10172 in Acute Stroke Treatment (TOAST) criteria16: large artery atherosclerosis, cardioembolic, small vessel disease, and other causes and uncertain causes. Stroke subtypes were also divided into groups according to the size of infarction observed on CT (the formula Π/6 × a × b × c, where a and b are the largest perpendicular diameters measured on CT and c is the slice thickness): large infarction group (>10 cm3), medium infarction group (4.1-10 cm3), and small infarction group (<4 cm3).17 Clinical evaluations were performed on admission and every day during patients’ hospital stay. Stroke severity was assessed with the NIHSS. Symptomatic HT or clinical deterioration (CD) caused by HT was defined as an increase of ≥4 points in the NIHSS in combination with a visible HT on the head CT.5,18 The infarct volume and the occurrence of early HT were investigated on repeated CT examination performed after 5 ± 2 days from stroke onset or immediately in case of clinical neurologic deterioration. The type of HT was classified according to the European Cooperative Acute Stroke Study-II criteria.19 This classifies hemorrhagic infarction 1 (HI1) as small petechiae along the periphery of the infarct region; hemorrhagic infarction 2 (HI2) as more confluent petechiae within the infarct, without space-occupying effect; parenchymal hemorrhage 1 (PH1) as bleeding ≤30% of the infarcted area but with mild space-occupying effect; and parenchymal hemorrhage 2 (PH2) as bleeding >30% of the infarcted area with substantial space-occupying effect or as any hemorrhagic lesion outside the infarcted area. In cases of more than one hemorrhagic lesion on CT examination, the worst possible HT category was assumed. All CT evaluations were made by the same neuroradiologist who was blinded to the clinical and analytical results.
Laboratory Tests Venous blood samples were collected on admission within 3 days of onset of acute ischemic stroke (blood samples from 292 patients were collected less than 24 hours after stroke onset, 17 less than 48 hours, and 7 less than 3 days after stroke onset). Ethylene Diamine Tetraacetic
ARTICLE IN PRESS IMMUNOPROTEASOME AND ISCHEMIC STROKE
Acid anticoagulant tubes were used to collect the blood. Plasma was immediately separated by centrifugation at 3000 rpm for 15 minutes at 4°C and immediately stored at −80°C until analysis. Plasma LMP2, MECL-1, LMP7, interleukin-1β (IL1β), and high-sensitivity C-reactive protein (Hs-CRP) were measured with commercially available quantitative sandwich enzyme-linked immunosorbent assay kits obtained from Shanghai Meilian Biological Technology Co., LTD (Shanghai, China). All tests were performed according to the manufacturer’s instructions. Detection was performed by researchers blinded to clinical and radiological data. All samples were tested in duplicate wells in the same assay and the average values were used.
Statistical Analysis Continuous variables are expressed as mean ± standard deviation and categorical variables are reported in percentages. Statistical intergroup differences were analyzed with Pearson χ2 test, two independent t tests, oneway analysis of variance, and least significant difference t test. Given that LMP2, MECL-1, LMP7, IL1β, and HsCRP levels were not normally distributed, they were expressed as median (interquartile range) and compared
3
between the two groups by the Mann–Whitney U test. A multivariate logistic regression analysis was performed to identify any independent factors of HT. Receiver operating characteristic curve analysis was performed to estimate the sensitivity, specificity, and predictive values of a specific concentration of plasma LMP2, MECL-1, and LMP7 for HT. The statistical analyses were performed using a commercially available software package (SPSS Inc., Chicago, IL, USA, version 13). Values of P < .05 were considered to be significant.
Results A total of 316 consecutive ischemic stroke patients (198 men, 62.7%) were enrolled in the present study. Table 1 shows the baseline clinical characteristics of the HT and non-HT groups. The mean age was 66.3 ± 12.4 years, 68.8 ± 10.6 years, and 65.9 ± 12.6 years in the total group, and in the HT and non-HT groups, respectively. There were 42 (13.3%, 42 of 316) subjects who experienced HT, including 13 (4.1%, 13 of 316) patients with HI1, 5 (1.6%, 5 of 316) patients with HI2, 7 (2.2%, 7 of 316) patients with PH1, and 17(5.4%, 17 of 316) patients with PH2.
Table 1. Baseline clinical characteristics of HT and non-HT Variable
Total (n = 316)
HT (n = 42)
Non-HT (n = 274)
P value
Age (years) Gender (male/female) Hypertension Diabetes mellitus Hyperlipidemia Atrial fibrillation NIHSS HT classification HI-1 HI-2 PH1 PH2 Clinical deterioration Infarct volume classification Large infarction Medium infarction Small infarction Stroke etiology Large artery atherosclerosis Cardioembolism Small vessel disease Other/unknown Plasma fibrinogen (g/L) Platelet count (×109/L) Blood uric acid (μmol/L) Glycosylated hemoglobin (%) Antithrombotic drugs use
66.3 ± 12.4 (198/118) 244 69 113 69 7.2 ± 4.6 42
68.8 ± 10.6 (23/19) 34 7 16 27 12.1 ± 4.8 42 (13.3%) 13 (4.1%) 5 (1.6%) 7 (2.2%) 17 (5.4%) 14 (33.3%)
65.9 ± 12.6 (175/99) 210 62 97 42 6.4 ± 4.1 0
.142 .257 .535 .384 .734 <.001 <.001
24 (8.8%)
<.001 <.001 <.001 .104
42 49 (15.5%) 145 (45.9%) 122 (38.6%)
27 (64.3%) 15 (35.6%) 0
22 (8.1%) 130 (47.4%) 122 (44.5%)
127 (40.2%) 70 (22.2%) 102 (32.3%) 17 (5.4%) 3.9 ± 2.2 170.1 ± 66.6 288.1 ± 109.4 6.6 ± 3.4 276
13 (30.9%) 27 (64.3%) 0 2 (4.8) 4.3 ± 1.2 168.8 ± 84.8 271.5 ± 111.7 6.6 ± 1.8 33
114 (41.6%) 43 (15.7%) 102 (37.2%) 15 (5.5%) 3.8 ± 2.3 170.3 ± 63.5 290.7 ± 109.0 6.6 ± 3.2 243
<.001 .126 <.001 .601 .182 .891 .292 .926 .066
Abbreviations: HI1, hemorrhagic infarction 1; HI2, hemorrhagic infarction 1; HT, hemorrhagic transformation; non-HT, non-hemorrhagic transformation; PH1, parenchymal hemorrhage 1; PH2, parenchymal hemorrhage 2.
ARTICLE IN PRESS <.001 <.001 <.001 <.001 <.001 200.2 (161.6-244.5) 208.8 (164.1-266.2) 179.3 (145.4-216.6) 31.3 (26.2-41.1) 4.9 (3.1-5.9) 613.9 (316.2-784.5) 386.9 (301.8-540.1) 272.1 (203.8-383.3) 44.1 (37.2-52.8) 5.7 (3.0-17.9) 661.9 (394.8-915.4) 411.9 (318.0-634.2) 299.1 (214.4-477.6) 46.4 (39.1-65.9) 6.2 (3.2-18.9) LMP2, median (pg/mL) (interquartile range) LMP5, median (pg/mL) (interquartile range) LMP7, median (pg/mL) (interquartile range) IL1β, median (pg/mL) (interquartile range) Hs-CRP, median (mg/L) (interquartile range)
1378.8 (1147.9-1737.7) 829.3 (656.8-987.5) 765.3 (640.6-934.9) 99.5 (83.6-121.7) 14.1 (7.2-41.8)
P value Control group (n = 10) Non-HT (n = 274) HT (n = 42) Total (n = 316) Variable
Table 2. Comparison of plasma LMP2, LPM5, LMP7, IL1β, and Hs-CRP levels between the HT and non-HT groups
Among this group, 14 patients (33.3%, 14 of 42) showed signs of CD (score increased ≥4 points in the NIHSS). In contrast, there were only 24 subjects (8.8%, 24 of 274) who displayed neurologic deterioration in the non-HT group. We assigned stroke etiology according to the TOAST criteria: large artery atherosclerosis, 40.2% (127 patients); cardioembolism, 22.2% (70 patients); small vessel disease, 32.2% (102 patients); and uncertain and other causes 5.4% (17 patients). On the other hand, stroke subtypes were also divided into large infarction group (49 patients), medium infarction group (145 patients), and small infarction group (122 patients) according to cranial CT performance. Interestingly, there was a tendency toward a higher incidence of cardioembolic, large volume infarction in the HT group than in the non-HT group (P < .001). There were significant differences in the NIHSS score (HT group [12.1 ± 4.8] versus non-HT group [6.4 ± 4.1]), frequencies of AF, large infarction, and cardioembolic stroke between the HT and non-HT groups (P < .001), whereas age, gender, peripheral blood plasma fibrinogen, blood platelet count, blood uric acid, glycosylated hemoglobin, and history of hypertension, diabetes mellitus, hyperlipidemia, and antithrombotic drug use were not significantly different between the HT and non-HT groups (P > .05) (Table 1). Plasma LMP2, MECL-1, LMP7, IL1β, and Hs-CRP concentrations on admission were significantly higher in patients with subsequent HT than in those without HT and in the control group, respectively (P < .001) (Table 2). Furthermore, we found that there were significant differences of plasma LMP2, MECL-1, LMP7, IL1β, and HsCRP protein levels in ischemic stroke patients with HI1, HI2, PH1, and PH2, respectively (P < .001) (Table 3). These protein concentrations increased with hemorrhage severity. Among all HT patients, the lowest concentrations of LMP2 (1086.5 [963.6-1193.3] pg/mL), MECL-1 (633.7 [528.8671.6] pg/mL), LMP7 (641.9 [590.1-734.5] pg/mL), IL1β (82.8 [69.2-96.4] pg/mL), and Hs-CRP (4.0 [2.8-10.4] mg/ L) were in the HI1 group (P < .001). However, the concentrations of LMP2 (1768.4 [1612.4-1930.8] pg/mL), MECL-1 (987.9 [886.7-1078.5] pg/mL), LMP7 (1009.3 [795.81032.2] pg/mL), IL1β (146.4 [104.1-199.2] pg/mL), and Hs-CRP (41.7 [18.8-81.6] mg/L) were higher in the PH2 group than in the HT group, respectively (P < .001) (Table 3). The similar effect was found for each stroke subtype according to stroke etiology. Table 4 indicated that the cardioembolic group shared the highest concentrations of LMP2 (767.5 [301.3-1260.8] pg/mL), MECL-1 (585.6 [409.6-814.9] pg/mL), LMP7 (514.9 [240.6-737.8] pg/ mL), IL1β (60.9 [40.6-94.3] pg/mL), and Hs-CRP (12.9 [4.331.9] mg/L) among these stroke subtypes (P < .001) (Table 4). Neurologic deterioration was observed in 33.3% (14 of 42) of patients with HT and in 8.8% (24 of 274) of those without HT (P < .001). Thirty-eight patients (14 patients
Abbreviations: Hs-CRP, high-sensitivity C-reactive protein; HT, hemorrhagic transformation; IL1β, interleukin-1β; LMP, low molecular mass peptide; non-HT, non-hemorrhagic transformation.
X. CHEN ET AL.
4
HI1 (n = 13)
HI2 (n = 5)
PH1 (n = 7)
PH2 (n = 17)
P value
LMP2, median (pg/mL) (interquartile range) LMP5, median (pg/mL) (interquartile range) LMP7, median (pg/mL) (interquartile range) IL1β, median (pg/mL) (interquartile range) Hs-CRP, median (mg/L) (interquartile range)
1086.5 (963.6-1193.3) 633.7 (528.8-671.6) 641.9 (590.1-734.5) 82.8 (69.2-96.4) 4.0 (2.8-10.4)
1183.9 (1170.6-1376.4) 747.4 (703.7-849.4) 729.6 (687.2-829.3) 92.6 (86.4-107.3) 12.0 (8.9-28.3)
1473.0 (1254.7-1675.1) 845.4 (816.3-886.5) 555.1 (518.9-895.6) 110.9 (77.6-114.6) 21.4 (12.2-135.0)
1768.4 (1612.4-1930.8) 987.9 (886.7-1078.5) 1009.3 (795.8-1032.2) 146.4 (104.1-199.2) 41.7 (18.8-81.6)
<.001 <.001 <.001 <.001 <.001
Abbreviations: HI1, hemorrhagic infarction 1; HI2, hemorrhagic infarction 1; HT, hemorrhagic transformation; non-HT, non-hemorrhagic transformation; PH1, parenchymal hemorrhage 1; PH2, parenchymal hemorrhage 2.
Table 4. Comparison of plasma LMP2, LPM5, LMP7, IL1β, and Hs-CRP levels in different stroke subtypes
Variable
Large artery atherosclerosis (n = 127)
Cardioembolic (n = 70)
Small vessel disease (n = 102)
Other/unknown (n = 17)
P value
LMP2, median (pg/mL) (interquartile range) LMP5, median (pg/mL) (interquartile range) LMP7, median (pg/mL) (interquartile range) IL1β, median (pg/mL) (interquartile range) Hs-CRP, median (mg/L) (interquartile range)
651.9 (446.6-890.5) 458.1 (332.5-671.7) 310.2 (215.8-478.2) 47.6 (39.4-78.4) 7.8 (3.4-18.6)
767.5 (301.3-1260.8) 585.6 (409.6-814.9) 514.9 (240.6-737.8) 60.9 (40.6-94.3) 12.9 (4.3-31.9)
621.8 (308.2-784.4) 355.7 (285.3-442.9) 258.1 (178.1-341.6) 42.7 (33.8-50.5) 4.2 (2.6-12.8)
638.9 (368.0-921.7) 420.3 (301.7-619.7) 286.3 (202.8-458.9) 46.1 (43.1-53.2) 5.8 (2.4-15.6)
.029 <.001 <.001 <.001 <.001
Abbreviations: Hs-CRP, high-sensitivity C-reactive protein; IL1β, interleukin-1β; LMP, low molecular mass peptide.
ARTICLE IN PRESS
Variable
IMMUNOPROTEASOME AND ISCHEMIC STROKE
Table 3. Comparison of plasma LMP2, LPM5, LMP7, IL1β, and Hs-CRP levels in different types of HT
5
ARTICLE IN PRESS X. CHEN ET AL.
6
Table 5. Multivariate logistic regression analysis to identify independent factors of hemorrhagic transformation Hemorrhagic transformation
Variable
Odds ratio
95% confidence interval
P value
NIHSS Infarct volume (cm3) Cardioembolism LMP2 (pg/mL) LMP5 (pg/mL) LMP7 (pg/mL) IL1β (pg/mL) Hs-CRP (mg/L)
.807 3.928 5.621 1.002 1.007 1.009 1.015 .986
.592~1.102 .532~28.985 1.034~30.563 1.000~1.003 1.003~1.011 1.004~1.014 .995~1.034 .996~1.005
.177 .180 .046 .024 .001 <.001 .139 .147
Abbreviations: Hs-CRP, high-sensitivity C-reactive protein; IL1β, interleukin-1β; LMP, low molecular mass peptide.
with HT and 24 patients without HT) in the CD group had higher LMP2, MECL-1, LMP7, IL1β, and Hs-CRP protein levels than in the 278 patients without CD (28 patients with HT and 250 patients without HT). The median (interquartile range) plasma biomarker concentrations were higher in the CD group (LMP2: 1350.8 [568.8-1676.1] versus 643.8 [321.8-860.3] pg/mL, P < .001; MECL-1: 509.8 [382.7-908.9] versus 412.0 [315.3628.8] pg/mL, P = .007; LMP7: 765.3 [640.6-934.9] versus 277.6 [206.3-401.8] pg/mL, P < .001; IL1β: 87.2 [76.8107.9] versus 643.8 [37.3-53.9] pg/mL, P < .001; and Hs-CRP: 20.6 [11.7-41.8] versus 5.7 [3.0-16.4] mg/L, P < .001). In particular, patients with CD caused by HT had the highest levels of LMP2 (1679.5 [1394.6-136.6] pg/mL), MECL-1 (992.5 [849.7-1075.8] pg/mL), LMP7 (822.6 [748.61009.5] pg/mL), IL1β (113.2 [90.6-194.5] pg/mL), and Hs-CRP (30.0 [12.8-75.6] mg/L) (P < .05). Logistic regression analysis in the model including NIHSS, cardioembolism, infarct volume, and the studied biomarkers LMP2, MECL-1, LMP7, IL1β, and Hs-CRP showed cardioembolism (P = .046), LMP2 (P = .024), MECL-1 (P = .001), and LMP7 (P < .001) as independent predictors of HT (Table 5). The optimal cutoff point as determined by the receiver operating characteristic analysis to predictive HT was 988.3 pg/mL for LMP2 (sensitivity: 92.9%, specificity: 90.1%, area under the curve [AUC]: .965, 95% confidence interval [CI]: .945~.985), 584.7 pg/mL for MECL-1 (sensitivity: 90.5%, specificity: 81.0%, AUC: .899, 95% CI: .862~.936), and 509.0 pg/mL for LMP7 (sensitivity: 97.6%, specificity: 90.9%, AUC: .966, 95% CI: .948~.984) (P < .001).
Discussion In the present study, we observed an incidence of early HT of about 13.3%. There were higher NIHSS score and
more frequencies of AF, large infarction, and cardioembolic stroke in patients with HT. These proteins of immunoproteasome subunits LMP2, MECL-1, and LMP7 concentrations increased with hemorrhage severity. We suggested that high levels of LMP2, MECL-1, and LMP7 on admission may be useful in the prediction of early HT in acute ischemic stroke patients. HT of ischemic stroke occurs relatively frequently. Therefore, it is extremely important to find early specific markers heralding HT in cerebral infarction. Currently, most of the studies concerning the risk of secondary hemorrhage of ischemic stroke focus on neuroimaging. The present study showed that plasma concentrations of LMP2, MECL-1, and LMP7 proteins may be useful in the prediction of early HT in acute ischemic stroke patients. To the best of our knowledge, this is the first study that reported that plasma LMP2, MECL-1, and LMP7 protein levels significantly increased in acute cerebral infarction patients, especially in the HT with CD patients. Therefore, this study suggested that high level of immunoproteasome might be a predictor of early HT in acute ischemic stroke patients. The mechanism of immunoproteasome in HT may be linked to its actions on neurovascular unit integrity. First, immunoproteasome is a subtype of proteasome. The catalytic β1i, β2i, and β5i subunits display caspase-like, trypsin-like, and chymotrypsin-like proteolytic activities, and exhibit preferential substrate cleavage after acidic, basic, and hydrophobic amino acid residues, respectively.13 Several findings suggest the hypothesis that the integrity of the neurovascular unit, in particular the microvessel basal lamina and extracellular matrix, can be affected by proteases of all four families (select MMPs, serine proteases, select cysteine proteases, and heparanase).3,20 Of note, the activation of MMPs appears to play a primary role in basal lamina degradation and secondary HT of the ischemic area.3,12 Accumulated evidence showed that MMP2, MMP-3, MMP-9, and other MMPs had an association with HT in stroke context.3,12,20 It seemed that there is a correlation between MMPs and immunoproteasome. A recent study reported that inhibition of LMP2 suppressed the expression and activities of MMP-2 and MMP-9.21 Therefore, we suggested that immunoproteasome regulates MMPs involved in destroying the integrity of the neurovascular unit. Second, immunoproteasome potentially regulates the production of inflammation mediators (such as IL-1 and tumor necrosis factor-α)22,23 and reactive oxygen species,24,25 and leads to the dysfunction of blood-brain barrier (BBB). Third, hypoxia induces vascular endothelial growth factor (VEGF) expression. VEGF has an important role in vascular remodeling and angiogenesis. VEGF has a biphasic role in stroke, with early VEGF promoting BBB disruption and HT, and later VEGF promoting BBB integrity and vascular function.3 The ubiquitin-proteasome system plays a central role in finetuning the functions of core proangiogenic proteins,
ARTICLE IN PRESS IMMUNOPROTEASOME AND ISCHEMIC STROKE
including VEGF, VEGFR-2, angiogenic signaling proteins, and other non-VEGF angiogenic pathways. 26 Therefore, we postulated that immunoproteasome may regulate VEGF promoting BBB disruption and HT. In summary, it is difficult to elucidate the precise mechanisms of immunoproteasome implicated in the neurovascular unit integrity and secondary HT of the ischemic area. Several factors have been related to HT after an ischemic event. Patients with acute cardioembolic stroke frequently show HT.6 AF is the most common etiology of cardioembolic stroke and is associated with HT.8 In the present study, compared with the non-HT group, the percentage of AF was significantly increased in the HT group. Both this study and other studies suggested that cardioembolism is useful in the early prediction of HT.8 Besides, infarct volume and NIHSS score are often used to evaluate the prognosis of stroke,17,27 but these indicators for predicting HT remain controversial. It was reported that large infarct was an independent predictor of symptomatic HT, but the NIHSS score was not.6 However, another study did not find that larger infarct volumes or higher NIHSS scores were associated with increased likelihood of symptomatic or severe hemorrhage.1 Although we found that there was a tendency toward a higher incidence of large volume infarction and higher NIHSS score in the ischemia stroke with HT than without HT, the multivariable analysis showed that large infarct and NIHSS score were not independent predictors of HT. The present study has some limitations. First, the different period from onset to blood sample collection may influence the value of plasma protein concentration. It is better to assess data in the group where blood samples can be collected within 3-6 hours from onset, as Kazmierski reported.5 In our study, blood samples were acquired from 24 hours to 72 hours after onset. There is mounting evidence that dynamic inflammation response plays an important role in cerebral ischemia. This may be one of the explanations why the present results of Hs-CRP are different from other studies.6,10,28 Second, all biomarker studies in this field to date, including this one, should be regarded as pilot studies that did not definitively elucidate the clinical relevance of any single marker or a panel of different markers. Third, in the present study, the total percentage of medium and small infarction was 84.5% (267 of 316), whereas the large infarction was 15.5%. However, there were only 42 (13.3%) patients who experienced HT. Generally speaking, there was a tendency toward a higher incidence of HT in patients with large volume infarction than that without. Therefore, we could not rule out the possible presence of selection bias. Finally, the sample size was small. The number of outcomes was not sufficiently large for adequate multivariable analysis. Larger studies are needed to assess the initial NIHSS score, infarct volume, IL1β, and Hs-CRP, all of which were associated with HT in univariate analysis but not in
7
multivariable analysis. A larger study is therefore required to confirm our hypotheses. In conclusion, our study demonstrates that high plasma levels of immunoproteasome subunits LMP2, MECL-1, and LMP7 proteins are independently associated with early HT in acute ischemic stroke. We suggest that evaluation of plasma levels of immunoproteasome may be helpful in the early prediction of HT and in guiding the clinical therapy of patients with stroke. Further studies will be required to confirm our hypotheses and its underlying mechanism. Acknowledgment: We are grateful to Professor Junyang Chen (Department of Pediatric Research, Fujian Provincial Hospital, Fujian Medical University Provincial Clinical College, Fuzhou 350001, PR China) for conducting ELISA.
References 1. Marsh EB, Llinas RH, Schneider ALC, et al. Predicting hemorrhagic transformation of acute ischemic stroke. Medicine (Baltimore) 2016;95:e2430. 2. Terruso V, D’Amelio M, Di Benedetto N, et al. Frequency and determinants for hemorrhagic transformation of cerebral infarction. Neuroepidemiology 2009;33:261-265. 3. Jickling GC, Liu D, Stamova B, et al. Hemorrhagic transformation after ischemic stroke in animals and humans. J Cereb Blood Flow Metab 2014;34:185-199. 4. Sussman ES, Connolly EJ. Hemorrhagic transformation: a review of the rate of hemorrhage in the major clinical trials of acute ischemic stroke. Front Neurol 2013;4:69. 5. Kazmierski R, Michalak S, Wencel-Warot A, et al. Serum tight-junction proteins predict hemorrhagic transformation in ischemic stroke patients. Neurology 2012;79:1677-1685. 6. Lee JH, Park KY, Shin JH, et al. Symptomatic hemorrhagic transformation and its predictors in acute ischemic stroke with atrial fibrillation. Eur Neurol 2010;64:193-200. 7. Mendioroz M, Aacute F, Ndez-Cadenas I, et al. Endogenous activated protein c predicts hemorrhagic transformation and mortality after tissue plasminogen activator treatment in stroke patients. Cerebrovasc Dis 2009;28:143-150. 8. Paciaroni M, Agnelli G, Corea F, et al. Early hemorrhagic transformation of brain infarction: rate, predictive factors, and influence on clinical outcome: results of a prospective multicenter study. Stroke 2008;39:2249-2256. 9. Foerch C, Wunderlich MT, Dvorak F, et al. Elevated serum S100B levels indicate a higher risk of hemorrhagic transformation after thrombolytic therapy in acute stroke. Stroke 2007;38:2491-2495. 10. Di Napoli M, Schwaninger M, Cappelli R, et al. Evaluation of C-Reactive protein measurement for assessing the risk and prognosis in ischemic stroke: a statement for health care professionals from the CRP pooling project members. Stroke 2005;36:1316-1329. 11. Lindley RI, Wardlaw JM, Sandercock PA, et al. Frequency and risk factors for spontaneous hemorrhagic transformation of cerebral infarction. J Stroke Cerebrovasc Dis 2004;13:235-246. 12. Castellanos M, Leira R, Serena J, et al. Plasma metalloproteinase-9 concentration predicts hemorrhagic transformation in acute ischemic stroke * editorial comment. Stroke 2003;34:40-46.
ARTICLE IN PRESS X. CHEN ET AL.
8 13. Angeles A, Fung G, Luo H. Immune and non-immune functions of the immunoproteasome. Front Biosci 2012;17:1904-1916. 14. Chen X, Zhang X, Wang Y, et al. Inhibition of immunoproteasome reduces infarction volume and attenuates inflammatory reaction in a rat model of ischemic stroke. Cell Death Dis 2015;6:e1626. 15. Kremer M, Henn A, Kolb C, et al. Reduced immunoproteasome formation and accumulation of immunoproteasomal precursors in the brains of lymphocytic choriomeningitis virus-infected mice. J Immunol 2010;185:5549-5560. 16. Adams HJ, Bendixen BH, Kappelle LJ, et al. Classification of subtype of acute ischemic stroke. Definitions for use in a multicenter clinical trial. TOAST. Trial of Org 10172 in Acute Stroke Treatment. Stroke 1993;24:3541. 17. Chen X, Zhan X, Chen M, et al. The prognostic value of combined NT-pro-BNP levels and NIHSS scores in patients with acute ischemic stroke. Intern Med 2012;51:2887-2892. 18. Shi ZS, Loh Y, Walker G, et al. Endovascular thrombectomy for acute ischemic stroke in failed intravenous tissue plasminogen activator versus nonintravenous tissue plasminogen activator patients: revascularization and outcomes stratified by the site of arterial occlusions. Stroke 2010;41:1185-1192. 19. Hacke W, Kaste M, Fieschi C, et al. Randomised doubleblind placebo-controlled trial of thrombolytic therapy with intravenous alteplase in acute ischaemic stroke (ECASS
20. 21.
22.
23.
24.
25.
26. 27.
28.
II). Second European-Australasian Acute Stroke Study Investigators. Lancet 1998;352:1245-1251. Del ZG. The neurovascular unit, matrix proteases, and innate inflammation. Ann N Y Acad Sci 2010;1207:46-49. Wang HX, Wang HM, Lin HY, et al. Proteasome subunit LMP2 is required for matrix metalloproteinase-2 and -9 expression and activities in human invasive extravillous trophoblast cell line. J Cell Physiol 2006;206:616-623. Muchamuel T, Basler M, Aujay MA, et al. A selective inhibitor of the immunoproteasome subunit LMP7 blocks cytokine production and attenuates progression of experimental arthritis. Nat Med 2009;15:781-787. Basler M, Dajee M, Moll C, et al. Prevention of experimental colitis by a selective inhibitor of the immunoproteasome. J Immunol 2010;185:634-641. Shang F, Taylor A. Ubiquitin-proteasome pathway and cellular responses to oxidative stress. Free Radic Biol Med 2011;51:5-16. Seifert U, Bialy LP, Ebstein F, et al. Immunoproteasomes preserve protein homeostasis upon interferon-induced oxidative stress. Cell 2010;142:613-624. Rahimi N. The ubiquitin-proteasome system meets angiogenesis. Mol Cancer Ther 2012;11:538-548. Garcia-Berrocoso T, Giralt D, Bustamante A, et al. B-type natriuretic peptides and mortality after stroke: a systematic review and meta-analysis. Neurology 2013;81:1976-1985. O’Donnell MJ, Berge E, Sandset PM. Are there patients with acute ischemic stroke and atrial fibrillation that benefit from low molecular weight heparin? Stroke 2006;37:452-455.
Plasma Immunoproteasome Predicts Early Hemorrhagic Transformation in Acute Ischemic Stroke Patients Xingyong Chen,
MD, PhD,
Yinzhou Wang, MD, PhD, Ming Fu, Qiong Cheng, MM, and Xu Zhang, BM
BM,
Huixin Lei,
BM,
Background and Purpose: Currently, blood biomarkers associated with an increased hemorrhagic transformation (HT) risk remain uncertain. We aimed to determine the significance of immunoproteasome as predictors of early HT in acute ischemic stroke patients. Methods: This study enrolled 316 patients with ischemic stroke. HT was assessed by computed tomography examination performed on day 5 ± 2 after stroke onset or immediately in case of clinical deterioration (CD). Plasma immunoproteasome subunits low molecular mass peptide 2 (LMP2), multicatalytic endopeptidase complex-like 1 (MECL-1), LMP7, interleukin-1β (IL1β), and high-sensitivity C-reactive protein (Hs-CRP) were measured with quantitative sandwich enzyme-linked immunosorbent assay kits. Factors associated with HT were analyzed using a multivariate logistic regression analysis. Results: There were 42 (13.3%, 42 of 316) patients who experienced HT. Compared with those patients without HT, plasma LMP2, MECL-1, LMP7, IL1β, and Hs-CRP concentrations on admission were significantly increased in patients with subsequent HT (P < .001). These protein concentrations increased with hemorrhage severity. Patients with CD caused by HT had the highest levels of LMP2 (1679.5 [1394.6136.6] pg/mL), MECL-1 (992.5 [849.7-1075.8] pg/mL), LMP7 (822.6 [748.6-1009.5] pg/mL), IL1β (113.2 [90.6-194.5] pg/mL), and Hs-CRP (30.0 [12.8-75.6] mg/L) (P < .05). Logistic regression analysis showed cardioembolism, LMP2, MECL-1, and LMP7 as independent predictors of HT (P < .05). Receiver operating characteristic curve analysis demonstrated LMP2 ≥ 988.3 pg/mL, MECL-1 ≥ 584.7 pg/mL, and LMP7 ≥ 509.0 pg/mL as independent factors associated with HT (P < .001). Conclusion: Evaluation of plasma levels of immunoproteasome could be helpful in the early prediction of HT in acute ischemic stroke patients. Key Words: Immunoproteasome—hemorrhagic transformation—cerebral infarction—risk factors. © 2016 National Stroke Association. Published by Elsevier Inc. All rights reserved.
From the Department of Neurology, Fujian Provincial Hospital, Fujian Medical University Provincial Clinical College, Fuzhou, China. Received July 21, 2016; revision received August 9, 2016; accepted August 17, 2016. Xingyong Chen and Yinzhou Wang contributed equally to this work. Author contributions: Xingyong Chen and Yinzhou Wang contributed in the study concept, acquisition of data, analysis and interpretation of the data, statistical analysis, and drafting and revision of the manuscript content. Ming Fu contributed in the acquisition and analysis of data. Qiong Cheng and Huixin Lei participated in the design of the study. Xu Zhang contributed in the revision of the manuscript content. Grant support: This study is supported by grants from Fujian Province Natural Science Fund (2013J01275, 2016J01432) and Young and Middle-aged Talents Training Project of Health and Family Planning Committee of Fujian Province (2015-ZQN-JC-5). Address correspondence to Yinzhou Wang, MD, PhD, and Xu Zhang, BM, Department of Neurology, Fujian Provincial Hospital, Fujian Medical University Provincial Clinical College, Fuzhou 350001, China. E-mail: [email protected]; [email protected]. 1052-3057/$ - see front matter © 2016 National Stroke Association. Published by Elsevier Inc. All rights reserved. http://dx.doi.org/10.1016/j.jstrokecerebrovasdis.2016.08.027
Journal of Stroke and Cerebrovascular Diseases, Vol. ■■, No. ■■ (■■), 2016: pp ■■–■■
1
ARTICLE IN PRESS X. CHEN ET AL.
2
Background Hemorrhagic transformation (HT) is bleeding into an area of ischemic brain after stroke. It occurs in as many as 10%-40% of patients with ischemic stroke.1,2 HT increases morbidity and mortality of acute ischemic stroke. It is difficult to predict a given patient’s probability of experiencing HT. Currently, several clinical parameters and blood biomarkers are known to be associated with an increased HT risk. Factors such as age, hypertension, atrial fibrillation (AF), baseline National Institutes of Health Stroke Scale (NIHSS) score, treatment with antiplatelet or thrombolytic agents, the presence of early ischemic changes on cranial computed tomography (CT) on admission, high levels of plasma matrix metalloproteinase (MMPs), activated protein C, estimated glomerular filtration rate, tight-junction proteins, and S100B have been related to HT after ischemic event.1-12 However, all studies in this field to date did not definitively elucidate the clinical relevance of any single marker or a panel of different markers. Immunoproteasome is a subtype of proteasome that contains three major catalytic subunits: β1i (also known as low molecular mass peptide 2 [LMP2]; proteasome subunit beta 9 [PSMB9]), β2i (also known as multicatalytic endopeptidase complex-like 1 [MECL-1]; PSMB10), and β5i (also known as LMP7; PSMB8).13 Immunoproteasome is present in several kinds of cells, including neurons, astrocytes, microglia, and endothelial cells in the brain areas.14,15 Our previous study had confirmed that augmentation of both immunoproteasome LMP2 and LMP7 is potentially involved in the inflammatory pathophysiological mechanism of ischemia stroke, and selective inhibition of the immunoproteasome subunit LMP7 offers a strong neuroprotection in middle cerebral artery occlusion (MCAO) rats.14 However, detailed knowledge on the relationship of immunoproteasome with human ischemia stroke is still not available. The aim of this prospective study in consecutive patients was therefore to assess (1) the rate of early HT in patients admitted for ischemic stroke, (2) the correlation between plasma immunoproteasome with early HT, and (3) the risk factors for early HT.
Subjects and Methods Patients We prospectively enrolled 316 patients with a first episode of ischemic stroke admitted within the first 72 hours of symptom onset to the neurology department of our hospital between March 2014 and September 2014. The study was approved by the local ethics committee and informed consent was obtained from each patient or a family member or relative. A diagnosis of acute ischemic stroke was made by stroke neurologists and was confirmed with CT or magnetic resonance imaging. A control group of 10 healthy subjects (men, 60%; mean age, 62.5 ± 11.9 years) was also studied. The exclusion criteria were patients who
received oral anticoagulants or heparin, with previous stroke, transient ischemic attack, or intracerebral or subarachnoid hemorrhage, inflammatory or infectious diseases, epilepsies, hematologic diseases, cancer, and severe liver and renal failure. Those who were previously dependent were excluded from the present study. In each case, past and present clinical histories, including medication use, were obtained by interviewing the patient or a family member in cases where the patient was aphasic or unconscious. Each patient’s hospital records were also reviewed for medical history. All patients underwent comprehensive clinical evaluation, blood pressure measurement, standard electrocardiogram (ECG), chest X-ray, blood routine, and fibrinogen and biochemical analyses. Stroke etiology was classified according to the Trial of Org 10172 in Acute Stroke Treatment (TOAST) criteria16: large artery atherosclerosis, cardioembolic, small vessel disease, and other causes and uncertain causes. Stroke subtypes were also divided into groups according to the size of infarction observed on CT (the formula Π/6 × a × b × c, where a and b are the largest perpendicular diameters measured on CT and c is the slice thickness): large infarction group (>10 cm3), medium infarction group (4.1-10 cm3), and small infarction group (<4 cm3).17 Clinical evaluations were performed on admission and every day during patients’ hospital stay. Stroke severity was assessed with the NIHSS. Symptomatic HT or clinical deterioration (CD) caused by HT was defined as an increase of ≥4 points in the NIHSS in combination with a visible HT on the head CT.5,18 The infarct volume and the occurrence of early HT were investigated on repeated CT examination performed after 5 ± 2 days from stroke onset or immediately in case of clinical neurologic deterioration. The type of HT was classified according to the European Cooperative Acute Stroke Study-II criteria.19 This classifies hemorrhagic infarction 1 (HI1) as small petechiae along the periphery of the infarct region; hemorrhagic infarction 2 (HI2) as more confluent petechiae within the infarct, without space-occupying effect; parenchymal hemorrhage 1 (PH1) as bleeding ≤30% of the infarcted area but with mild space-occupying effect; and parenchymal hemorrhage 2 (PH2) as bleeding >30% of the infarcted area with substantial space-occupying effect or as any hemorrhagic lesion outside the infarcted area. In cases of more than one hemorrhagic lesion on CT examination, the worst possible HT category was assumed. All CT evaluations were made by the same neuroradiologist who was blinded to the clinical and analytical results.
Laboratory Tests Venous blood samples were collected on admission within 3 days of onset of acute ischemic stroke (blood samples from 292 patients were collected less than 24 hours after stroke onset, 17 less than 48 hours, and 7 less than 3 days after stroke onset). Ethylene Diamine Tetraacetic
ARTICLE IN PRESS IMMUNOPROTEASOME AND ISCHEMIC STROKE
Acid anticoagulant tubes were used to collect the blood. Plasma was immediately separated by centrifugation at 3000 rpm for 15 minutes at 4°C and immediately stored at −80°C until analysis. Plasma LMP2, MECL-1, LMP7, interleukin-1β (IL1β), and high-sensitivity C-reactive protein (Hs-CRP) were measured with commercially available quantitative sandwich enzyme-linked immunosorbent assay kits obtained from Shanghai Meilian Biological Technology Co., LTD (Shanghai, China). All tests were performed according to the manufacturer’s instructions. Detection was performed by researchers blinded to clinical and radiological data. All samples were tested in duplicate wells in the same assay and the average values were used.
Statistical Analysis Continuous variables are expressed as mean ± standard deviation and categorical variables are reported in percentages. Statistical intergroup differences were analyzed with Pearson χ2 test, two independent t tests, oneway analysis of variance, and least significant difference t test. Given that LMP2, MECL-1, LMP7, IL1β, and HsCRP levels were not normally distributed, they were expressed as median (interquartile range) and compared
3
between the two groups by the Mann–Whitney U test. A multivariate logistic regression analysis was performed to identify any independent factors of HT. Receiver operating characteristic curve analysis was performed to estimate the sensitivity, specificity, and predictive values of a specific concentration of plasma LMP2, MECL-1, and LMP7 for HT. The statistical analyses were performed using a commercially available software package (SPSS Inc., Chicago, IL, USA, version 13). Values of P < .05 were considered to be significant.
Results A total of 316 consecutive ischemic stroke patients (198 men, 62.7%) were enrolled in the present study. Table 1 shows the baseline clinical characteristics of the HT and non-HT groups. The mean age was 66.3 ± 12.4 years, 68.8 ± 10.6 years, and 65.9 ± 12.6 years in the total group, and in the HT and non-HT groups, respectively. There were 42 (13.3%, 42 of 316) subjects who experienced HT, including 13 (4.1%, 13 of 316) patients with HI1, 5 (1.6%, 5 of 316) patients with HI2, 7 (2.2%, 7 of 316) patients with PH1, and 17(5.4%, 17 of 316) patients with PH2.
Table 1. Baseline clinical characteristics of HT and non-HT Variable
Total (n = 316)
HT (n = 42)
Non-HT (n = 274)
P value
Age (years) Gender (male/female) Hypertension Diabetes mellitus Hyperlipidemia Atrial fibrillation NIHSS HT classification HI-1 HI-2 PH1 PH2 Clinical deterioration Infarct volume classification Large infarction Medium infarction Small infarction Stroke etiology Large artery atherosclerosis Cardioembolism Small vessel disease Other/unknown Plasma fibrinogen (g/L) Platelet count (×109/L) Blood uric acid (μmol/L) Glycosylated hemoglobin (%) Antithrombotic drugs use
66.3 ± 12.4 (198/118) 244 69 113 69 7.2 ± 4.6 42
68.8 ± 10.6 (23/19) 34 7 16 27 12.1 ± 4.8 42 (13.3%) 13 (4.1%) 5 (1.6%) 7 (2.2%) 17 (5.4%) 14 (33.3%)
65.9 ± 12.6 (175/99) 210 62 97 42 6.4 ± 4.1 0
.142 .257 .535 .384 .734 <.001 <.001
24 (8.8%)
<.001 <.001 <.001 .104
42 49 (15.5%) 145 (45.9%) 122 (38.6%)
27 (64.3%) 15 (35.6%) 0
22 (8.1%) 130 (47.4%) 122 (44.5%)
127 (40.2%) 70 (22.2%) 102 (32.3%) 17 (5.4%) 3.9 ± 2.2 170.1 ± 66.6 288.1 ± 109.4 6.6 ± 3.4 276
13 (30.9%) 27 (64.3%) 0 2 (4.8) 4.3 ± 1.2 168.8 ± 84.8 271.5 ± 111.7 6.6 ± 1.8 33
114 (41.6%) 43 (15.7%) 102 (37.2%) 15 (5.5%) 3.8 ± 2.3 170.3 ± 63.5 290.7 ± 109.0 6.6 ± 3.2 243
<.001 .126 <.001 .601 .182 .891 .292 .926 .066
Abbreviations: HI1, hemorrhagic infarction 1; HI2, hemorrhagic infarction 1; HT, hemorrhagic transformation; non-HT, non-hemorrhagic transformation; PH1, parenchymal hemorrhage 1; PH2, parenchymal hemorrhage 2.
ARTICLE IN PRESS <.001 <.001 <.001 <.001 <.001 200.2 (161.6-244.5) 208.8 (164.1-266.2) 179.3 (145.4-216.6) 31.3 (26.2-41.1) 4.9 (3.1-5.9) 613.9 (316.2-784.5) 386.9 (301.8-540.1) 272.1 (203.8-383.3) 44.1 (37.2-52.8) 5.7 (3.0-17.9) 661.9 (394.8-915.4) 411.9 (318.0-634.2) 299.1 (214.4-477.6) 46.4 (39.1-65.9) 6.2 (3.2-18.9) LMP2, median (pg/mL) (interquartile range) LMP5, median (pg/mL) (interquartile range) LMP7, median (pg/mL) (interquartile range) IL1β, median (pg/mL) (interquartile range) Hs-CRP, median (mg/L) (interquartile range)
1378.8 (1147.9-1737.7) 829.3 (656.8-987.5) 765.3 (640.6-934.9) 99.5 (83.6-121.7) 14.1 (7.2-41.8)
P value Control group (n = 10) Non-HT (n = 274) HT (n = 42) Total (n = 316) Variable
Table 2. Comparison of plasma LMP2, LPM5, LMP7, IL1β, and Hs-CRP levels between the HT and non-HT groups
Among this group, 14 patients (33.3%, 14 of 42) showed signs of CD (score increased ≥4 points in the NIHSS). In contrast, there were only 24 subjects (8.8%, 24 of 274) who displayed neurologic deterioration in the non-HT group. We assigned stroke etiology according to the TOAST criteria: large artery atherosclerosis, 40.2% (127 patients); cardioembolism, 22.2% (70 patients); small vessel disease, 32.2% (102 patients); and uncertain and other causes 5.4% (17 patients). On the other hand, stroke subtypes were also divided into large infarction group (49 patients), medium infarction group (145 patients), and small infarction group (122 patients) according to cranial CT performance. Interestingly, there was a tendency toward a higher incidence of cardioembolic, large volume infarction in the HT group than in the non-HT group (P < .001). There were significant differences in the NIHSS score (HT group [12.1 ± 4.8] versus non-HT group [6.4 ± 4.1]), frequencies of AF, large infarction, and cardioembolic stroke between the HT and non-HT groups (P < .001), whereas age, gender, peripheral blood plasma fibrinogen, blood platelet count, blood uric acid, glycosylated hemoglobin, and history of hypertension, diabetes mellitus, hyperlipidemia, and antithrombotic drug use were not significantly different between the HT and non-HT groups (P > .05) (Table 1). Plasma LMP2, MECL-1, LMP7, IL1β, and Hs-CRP concentrations on admission were significantly higher in patients with subsequent HT than in those without HT and in the control group, respectively (P < .001) (Table 2). Furthermore, we found that there were significant differences of plasma LMP2, MECL-1, LMP7, IL1β, and HsCRP protein levels in ischemic stroke patients with HI1, HI2, PH1, and PH2, respectively (P < .001) (Table 3). These protein concentrations increased with hemorrhage severity. Among all HT patients, the lowest concentrations of LMP2 (1086.5 [963.6-1193.3] pg/mL), MECL-1 (633.7 [528.8671.6] pg/mL), LMP7 (641.9 [590.1-734.5] pg/mL), IL1β (82.8 [69.2-96.4] pg/mL), and Hs-CRP (4.0 [2.8-10.4] mg/ L) were in the HI1 group (P < .001). However, the concentrations of LMP2 (1768.4 [1612.4-1930.8] pg/mL), MECL-1 (987.9 [886.7-1078.5] pg/mL), LMP7 (1009.3 [795.81032.2] pg/mL), IL1β (146.4 [104.1-199.2] pg/mL), and Hs-CRP (41.7 [18.8-81.6] mg/L) were higher in the PH2 group than in the HT group, respectively (P < .001) (Table 3). The similar effect was found for each stroke subtype according to stroke etiology. Table 4 indicated that the cardioembolic group shared the highest concentrations of LMP2 (767.5 [301.3-1260.8] pg/mL), MECL-1 (585.6 [409.6-814.9] pg/mL), LMP7 (514.9 [240.6-737.8] pg/ mL), IL1β (60.9 [40.6-94.3] pg/mL), and Hs-CRP (12.9 [4.331.9] mg/L) among these stroke subtypes (P < .001) (Table 4). Neurologic deterioration was observed in 33.3% (14 of 42) of patients with HT and in 8.8% (24 of 274) of those without HT (P < .001). Thirty-eight patients (14 patients
Abbreviations: Hs-CRP, high-sensitivity C-reactive protein; HT, hemorrhagic transformation; IL1β, interleukin-1β; LMP, low molecular mass peptide; non-HT, non-hemorrhagic transformation.
X. CHEN ET AL.
4
HI1 (n = 13)
HI2 (n = 5)
PH1 (n = 7)
PH2 (n = 17)
P value
LMP2, median (pg/mL) (interquartile range) LMP5, median (pg/mL) (interquartile range) LMP7, median (pg/mL) (interquartile range) IL1β, median (pg/mL) (interquartile range) Hs-CRP, median (mg/L) (interquartile range)
1086.5 (963.6-1193.3) 633.7 (528.8-671.6) 641.9 (590.1-734.5) 82.8 (69.2-96.4) 4.0 (2.8-10.4)
1183.9 (1170.6-1376.4) 747.4 (703.7-849.4) 729.6 (687.2-829.3) 92.6 (86.4-107.3) 12.0 (8.9-28.3)
1473.0 (1254.7-1675.1) 845.4 (816.3-886.5) 555.1 (518.9-895.6) 110.9 (77.6-114.6) 21.4 (12.2-135.0)
1768.4 (1612.4-1930.8) 987.9 (886.7-1078.5) 1009.3 (795.8-1032.2) 146.4 (104.1-199.2) 41.7 (18.8-81.6)
<.001 <.001 <.001 <.001 <.001
Abbreviations: HI1, hemorrhagic infarction 1; HI2, hemorrhagic infarction 1; HT, hemorrhagic transformation; non-HT, non-hemorrhagic transformation; PH1, parenchymal hemorrhage 1; PH2, parenchymal hemorrhage 2.
Table 4. Comparison of plasma LMP2, LPM5, LMP7, IL1β, and Hs-CRP levels in different stroke subtypes
Variable
Large artery atherosclerosis (n = 127)
Cardioembolic (n = 70)
Small vessel disease (n = 102)
Other/unknown (n = 17)
P value
LMP2, median (pg/mL) (interquartile range) LMP5, median (pg/mL) (interquartile range) LMP7, median (pg/mL) (interquartile range) IL1β, median (pg/mL) (interquartile range) Hs-CRP, median (mg/L) (interquartile range)
651.9 (446.6-890.5) 458.1 (332.5-671.7) 310.2 (215.8-478.2) 47.6 (39.4-78.4) 7.8 (3.4-18.6)
767.5 (301.3-1260.8) 585.6 (409.6-814.9) 514.9 (240.6-737.8) 60.9 (40.6-94.3) 12.9 (4.3-31.9)
621.8 (308.2-784.4) 355.7 (285.3-442.9) 258.1 (178.1-341.6) 42.7 (33.8-50.5) 4.2 (2.6-12.8)
638.9 (368.0-921.7) 420.3 (301.7-619.7) 286.3 (202.8-458.9) 46.1 (43.1-53.2) 5.8 (2.4-15.6)
.029 <.001 <.001 <.001 <.001
Abbreviations: Hs-CRP, high-sensitivity C-reactive protein; IL1β, interleukin-1β; LMP, low molecular mass peptide.
ARTICLE IN PRESS
Variable
IMMUNOPROTEASOME AND ISCHEMIC STROKE
Table 3. Comparison of plasma LMP2, LPM5, LMP7, IL1β, and Hs-CRP levels in different types of HT
5
ARTICLE IN PRESS X. CHEN ET AL.
6
Table 5. Multivariate logistic regression analysis to identify independent factors of hemorrhagic transformation Hemorrhagic transformation
Variable
Odds ratio
95% confidence interval
P value
NIHSS Infarct volume (cm3) Cardioembolism LMP2 (pg/mL) LMP5 (pg/mL) LMP7 (pg/mL) IL1β (pg/mL) Hs-CRP (mg/L)
.807 3.928 5.621 1.002 1.007 1.009 1.015 .986
.592~1.102 .532~28.985 1.034~30.563 1.000~1.003 1.003~1.011 1.004~1.014 .995~1.034 .996~1.005
.177 .180 .046 .024 .001 <.001 .139 .147
Abbreviations: Hs-CRP, high-sensitivity C-reactive protein; IL1β, interleukin-1β; LMP, low molecular mass peptide.
with HT and 24 patients without HT) in the CD group had higher LMP2, MECL-1, LMP7, IL1β, and Hs-CRP protein levels than in the 278 patients without CD (28 patients with HT and 250 patients without HT). The median (interquartile range) plasma biomarker concentrations were higher in the CD group (LMP2: 1350.8 [568.8-1676.1] versus 643.8 [321.8-860.3] pg/mL, P < .001; MECL-1: 509.8 [382.7-908.9] versus 412.0 [315.3628.8] pg/mL, P = .007; LMP7: 765.3 [640.6-934.9] versus 277.6 [206.3-401.8] pg/mL, P < .001; IL1β: 87.2 [76.8107.9] versus 643.8 [37.3-53.9] pg/mL, P < .001; and Hs-CRP: 20.6 [11.7-41.8] versus 5.7 [3.0-16.4] mg/L, P < .001). In particular, patients with CD caused by HT had the highest levels of LMP2 (1679.5 [1394.6-136.6] pg/mL), MECL-1 (992.5 [849.7-1075.8] pg/mL), LMP7 (822.6 [748.61009.5] pg/mL), IL1β (113.2 [90.6-194.5] pg/mL), and Hs-CRP (30.0 [12.8-75.6] mg/L) (P < .05). Logistic regression analysis in the model including NIHSS, cardioembolism, infarct volume, and the studied biomarkers LMP2, MECL-1, LMP7, IL1β, and Hs-CRP showed cardioembolism (P = .046), LMP2 (P = .024), MECL-1 (P = .001), and LMP7 (P < .001) as independent predictors of HT (Table 5). The optimal cutoff point as determined by the receiver operating characteristic analysis to predictive HT was 988.3 pg/mL for LMP2 (sensitivity: 92.9%, specificity: 90.1%, area under the curve [AUC]: .965, 95% confidence interval [CI]: .945~.985), 584.7 pg/mL for MECL-1 (sensitivity: 90.5%, specificity: 81.0%, AUC: .899, 95% CI: .862~.936), and 509.0 pg/mL for LMP7 (sensitivity: 97.6%, specificity: 90.9%, AUC: .966, 95% CI: .948~.984) (P < .001).
Discussion In the present study, we observed an incidence of early HT of about 13.3%. There were higher NIHSS score and
more frequencies of AF, large infarction, and cardioembolic stroke in patients with HT. These proteins of immunoproteasome subunits LMP2, MECL-1, and LMP7 concentrations increased with hemorrhage severity. We suggested that high levels of LMP2, MECL-1, and LMP7 on admission may be useful in the prediction of early HT in acute ischemic stroke patients. HT of ischemic stroke occurs relatively frequently. Therefore, it is extremely important to find early specific markers heralding HT in cerebral infarction. Currently, most of the studies concerning the risk of secondary hemorrhage of ischemic stroke focus on neuroimaging. The present study showed that plasma concentrations of LMP2, MECL-1, and LMP7 proteins may be useful in the prediction of early HT in acute ischemic stroke patients. To the best of our knowledge, this is the first study that reported that plasma LMP2, MECL-1, and LMP7 protein levels significantly increased in acute cerebral infarction patients, especially in the HT with CD patients. Therefore, this study suggested that high level of immunoproteasome might be a predictor of early HT in acute ischemic stroke patients. The mechanism of immunoproteasome in HT may be linked to its actions on neurovascular unit integrity. First, immunoproteasome is a subtype of proteasome. The catalytic β1i, β2i, and β5i subunits display caspase-like, trypsin-like, and chymotrypsin-like proteolytic activities, and exhibit preferential substrate cleavage after acidic, basic, and hydrophobic amino acid residues, respectively.13 Several findings suggest the hypothesis that the integrity of the neurovascular unit, in particular the microvessel basal lamina and extracellular matrix, can be affected by proteases of all four families (select MMPs, serine proteases, select cysteine proteases, and heparanase).3,20 Of note, the activation of MMPs appears to play a primary role in basal lamina degradation and secondary HT of the ischemic area.3,12 Accumulated evidence showed that MMP2, MMP-3, MMP-9, and other MMPs had an association with HT in stroke context.3,12,20 It seemed that there is a correlation between MMPs and immunoproteasome. A recent study reported that inhibition of LMP2 suppressed the expression and activities of MMP-2 and MMP-9.21 Therefore, we suggested that immunoproteasome regulates MMPs involved in destroying the integrity of the neurovascular unit. Second, immunoproteasome potentially regulates the production of inflammation mediators (such as IL-1 and tumor necrosis factor-α)22,23 and reactive oxygen species,24,25 and leads to the dysfunction of blood-brain barrier (BBB). Third, hypoxia induces vascular endothelial growth factor (VEGF) expression. VEGF has an important role in vascular remodeling and angiogenesis. VEGF has a biphasic role in stroke, with early VEGF promoting BBB disruption and HT, and later VEGF promoting BBB integrity and vascular function.3 The ubiquitin-proteasome system plays a central role in finetuning the functions of core proangiogenic proteins,
ARTICLE IN PRESS IMMUNOPROTEASOME AND ISCHEMIC STROKE
including VEGF, VEGFR-2, angiogenic signaling proteins, and other non-VEGF angiogenic pathways. 26 Therefore, we postulated that immunoproteasome may regulate VEGF promoting BBB disruption and HT. In summary, it is difficult to elucidate the precise mechanisms of immunoproteasome implicated in the neurovascular unit integrity and secondary HT of the ischemic area. Several factors have been related to HT after an ischemic event. Patients with acute cardioembolic stroke frequently show HT.6 AF is the most common etiology of cardioembolic stroke and is associated with HT.8 In the present study, compared with the non-HT group, the percentage of AF was significantly increased in the HT group. Both this study and other studies suggested that cardioembolism is useful in the early prediction of HT.8 Besides, infarct volume and NIHSS score are often used to evaluate the prognosis of stroke,17,27 but these indicators for predicting HT remain controversial. It was reported that large infarct was an independent predictor of symptomatic HT, but the NIHSS score was not.6 However, another study did not find that larger infarct volumes or higher NIHSS scores were associated with increased likelihood of symptomatic or severe hemorrhage.1 Although we found that there was a tendency toward a higher incidence of large volume infarction and higher NIHSS score in the ischemia stroke with HT than without HT, the multivariable analysis showed that large infarct and NIHSS score were not independent predictors of HT. The present study has some limitations. First, the different period from onset to blood sample collection may influence the value of plasma protein concentration. It is better to assess data in the group where blood samples can be collected within 3-6 hours from onset, as Kazmierski reported.5 In our study, blood samples were acquired from 24 hours to 72 hours after onset. There is mounting evidence that dynamic inflammation response plays an important role in cerebral ischemia. This may be one of the explanations why the present results of Hs-CRP are different from other studies.6,10,28 Second, all biomarker studies in this field to date, including this one, should be regarded as pilot studies that did not definitively elucidate the clinical relevance of any single marker or a panel of different markers. Third, in the present study, the total percentage of medium and small infarction was 84.5% (267 of 316), whereas the large infarction was 15.5%. However, there were only 42 (13.3%) patients who experienced HT. Generally speaking, there was a tendency toward a higher incidence of HT in patients with large volume infarction than that without. Therefore, we could not rule out the possible presence of selection bias. Finally, the sample size was small. The number of outcomes was not sufficiently large for adequate multivariable analysis. Larger studies are needed to assess the initial NIHSS score, infarct volume, IL1β, and Hs-CRP, all of which were associated with HT in univariate analysis but not in
7
multivariable analysis. A larger study is therefore required to confirm our hypotheses. In conclusion, our study demonstrates that high plasma levels of immunoproteasome subunits LMP2, MECL-1, and LMP7 proteins are independently associated with early HT in acute ischemic stroke. We suggest that evaluation of plasma levels of immunoproteasome may be helpful in the early prediction of HT and in guiding the clinical therapy of patients with stroke. Further studies will be required to confirm our hypotheses and its underlying mechanism. Acknowledgment: We are grateful to Professor Junyang Chen (Department of Pediatric Research, Fujian Provincial Hospital, Fujian Medical University Provincial Clinical College, Fuzhou 350001, PR China) for conducting ELISA.
References 1. Marsh EB, Llinas RH, Schneider ALC, et al. Predicting hemorrhagic transformation of acute ischemic stroke. Medicine (Baltimore) 2016;95:e2430. 2. Terruso V, D’Amelio M, Di Benedetto N, et al. Frequency and determinants for hemorrhagic transformation of cerebral infarction. Neuroepidemiology 2009;33:261-265. 3. Jickling GC, Liu D, Stamova B, et al. Hemorrhagic transformation after ischemic stroke in animals and humans. J Cereb Blood Flow Metab 2014;34:185-199. 4. Sussman ES, Connolly EJ. Hemorrhagic transformation: a review of the rate of hemorrhage in the major clinical trials of acute ischemic stroke. Front Neurol 2013;4:69. 5. Kazmierski R, Michalak S, Wencel-Warot A, et al. Serum tight-junction proteins predict hemorrhagic transformation in ischemic stroke patients. Neurology 2012;79:1677-1685. 6. Lee JH, Park KY, Shin JH, et al. Symptomatic hemorrhagic transformation and its predictors in acute ischemic stroke with atrial fibrillation. Eur Neurol 2010;64:193-200. 7. Mendioroz M, Aacute F, Ndez-Cadenas I, et al. Endogenous activated protein c predicts hemorrhagic transformation and mortality after tissue plasminogen activator treatment in stroke patients. Cerebrovasc Dis 2009;28:143-150. 8. Paciaroni M, Agnelli G, Corea F, et al. Early hemorrhagic transformation of brain infarction: rate, predictive factors, and influence on clinical outcome: results of a prospective multicenter study. Stroke 2008;39:2249-2256. 9. Foerch C, Wunderlich MT, Dvorak F, et al. Elevated serum S100B levels indicate a higher risk of hemorrhagic transformation after thrombolytic therapy in acute stroke. Stroke 2007;38:2491-2495. 10. Di Napoli M, Schwaninger M, Cappelli R, et al. Evaluation of C-Reactive protein measurement for assessing the risk and prognosis in ischemic stroke: a statement for health care professionals from the CRP pooling project members. Stroke 2005;36:1316-1329. 11. Lindley RI, Wardlaw JM, Sandercock PA, et al. Frequency and risk factors for spontaneous hemorrhagic transformation of cerebral infarction. J Stroke Cerebrovasc Dis 2004;13:235-246. 12. Castellanos M, Leira R, Serena J, et al. Plasma metalloproteinase-9 concentration predicts hemorrhagic transformation in acute ischemic stroke * editorial comment. Stroke 2003;34:40-46.
ARTICLE IN PRESS X. CHEN ET AL.
8 13. Angeles A, Fung G, Luo H. Immune and non-immune functions of the immunoproteasome. Front Biosci 2012;17:1904-1916. 14. Chen X, Zhang X, Wang Y, et al. Inhibition of immunoproteasome reduces infarction volume and attenuates inflammatory reaction in a rat model of ischemic stroke. Cell Death Dis 2015;6:e1626. 15. Kremer M, Henn A, Kolb C, et al. Reduced immunoproteasome formation and accumulation of immunoproteasomal precursors in the brains of lymphocytic choriomeningitis virus-infected mice. J Immunol 2010;185:5549-5560. 16. Adams HJ, Bendixen BH, Kappelle LJ, et al. Classification of subtype of acute ischemic stroke. Definitions for use in a multicenter clinical trial. TOAST. Trial of Org 10172 in Acute Stroke Treatment. Stroke 1993;24:3541. 17. Chen X, Zhan X, Chen M, et al. The prognostic value of combined NT-pro-BNP levels and NIHSS scores in patients with acute ischemic stroke. Intern Med 2012;51:2887-2892. 18. Shi ZS, Loh Y, Walker G, et al. Endovascular thrombectomy for acute ischemic stroke in failed intravenous tissue plasminogen activator versus nonintravenous tissue plasminogen activator patients: revascularization and outcomes stratified by the site of arterial occlusions. Stroke 2010;41:1185-1192. 19. Hacke W, Kaste M, Fieschi C, et al. Randomised doubleblind placebo-controlled trial of thrombolytic therapy with intravenous alteplase in acute ischaemic stroke (ECASS
20. 21.
22.
23.
24.
25.
26. 27.
28.
II). Second European-Australasian Acute Stroke Study Investigators. Lancet 1998;352:1245-1251. Del ZG. The neurovascular unit, matrix proteases, and innate inflammation. Ann N Y Acad Sci 2010;1207:46-49. Wang HX, Wang HM, Lin HY, et al. Proteasome subunit LMP2 is required for matrix metalloproteinase-2 and -9 expression and activities in human invasive extravillous trophoblast cell line. J Cell Physiol 2006;206:616-623. Muchamuel T, Basler M, Aujay MA, et al. A selective inhibitor of the immunoproteasome subunit LMP7 blocks cytokine production and attenuates progression of experimental arthritis. Nat Med 2009;15:781-787. Basler M, Dajee M, Moll C, et al. Prevention of experimental colitis by a selective inhibitor of the immunoproteasome. J Immunol 2010;185:634-641. Shang F, Taylor A. Ubiquitin-proteasome pathway and cellular responses to oxidative stress. Free Radic Biol Med 2011;51:5-16. Seifert U, Bialy LP, Ebstein F, et al. Immunoproteasomes preserve protein homeostasis upon interferon-induced oxidative stress. Cell 2010;142:613-624. Rahimi N. The ubiquitin-proteasome system meets angiogenesis. Mol Cancer Ther 2012;11:538-548. Garcia-Berrocoso T, Giralt D, Bustamante A, et al. B-type natriuretic peptides and mortality after stroke: a systematic review and meta-analysis. Neurology 2013;81:1976-1985. O’Donnell MJ, Berge E, Sandset PM. Are there patients with acute ischemic stroke and atrial fibrillation that benefit from low molecular weight heparin? Stroke 2006;37:452-455.