Acute infections are associated with a transient increase

Transcription

1 Preceding and Poststroke s in Young Adults With First-Ever Ischemic Stroke Effect on Short-Term and Long-Term Outcomes Terttu Heikinheimo, MD; Jenna Broman, MD; Elena Haapaniemi, PhD; Markku Kaste, PhD; Turgut Tatlisumak, PhD; Jukka Putaala, PhD Background and Purpose Poststroke infections (PSIs) worsen the outcome in acute ischemic stroke, but the effect of preceding infection (PI) is controversial. Data on young patients are scarce. We characterized PI and PSI in young adults with first-ever stroke and studied whether they are associated with 3-month and long-term outcomes, recurrent vascular events, and death. Methods From our database of 1008 consecutive patients aged 15 to 49 years, we included in the present study those who had brain imaging done within the first 2 days from stroke onset. Outcomes were unfavorable at 3 months and during long-term follow-up, vascular events, and all-cause death. Logistic regression and Cox proportional models were used to determine associations between infections and clinical outcomes. Results A total of 681 patients (62.3% men) fulfilled the inclusion criteria. Of these, 70 (10.3%) had PI, most commonly upper respiratory tract infection, and 103 (15.1%) had PSI, most commonly pneumonia. After adjusting for sex, age, and risk factors, both PI (odds ratio, 2.86; 95% confidence interval, ) and PSI (odds ratio, 2.26; 95% confidence interval, ) were independently associated with unfavorable 3-month outcome. PSI was also associated with long-term (follow-up, 7.8±4.0 years) higher risk of all-cause death. Conclusions In young patients with ischemic stroke, both PI and PSIs are associated with unfavorable short-term outcome. PSIs are also associated with higher long-term mortality. (Stroke. 2013;44: ) Key Words: assessment, outcomes infection Acute infections are associated with a transient increase in the risk of vascular events, including ischemic stroke, regardless of patients age. 1 5 In some studies, preceding infections (PIs) have been linked particularly to large-vessel and cardioembolic ischemic strokes. 5,6 Also chronic infections, such as bronchial or dental infections, may have a potential to cause inflammation, to promote atherosclerosis, and to increase risk for atherothrombotic stroke. 7,8 Reported frequency of PI varies from 5.7% to 38.8%, respiratory tract infections being the most common type. 1 5 Patients with PI may represent a specific patient population with stroke having different response to acute ischemic injury compared with patients without PI. This might influence clinical outcome. Two recent studies found no independent influence between PI and 3-month outcome. 9,10 However, the latter study showed unfavorable 30-day outcome in those with PI. 10 Another stroke-related infectious phenomena are poststroke infections (PSIs) that occur during the acute phase (0 7 days after the stroke). PSIs have frequent medical complications: urinary tract infections occur in 10% to 29% and chest infections in 11.2% to 18% of the cases Stroke severity, treatment in intensive care unit, vomiting at stroke onset, having dysphagia, or nasogastric feeding tube make patients with stroke prone to PSI PSI worsens stroke outcome and is associated with increasing mortality 11,16,17 and risk of future vascular event after stroke Possible effect of PI or PSI on outcome in young adults with ischemic stroke is currently unknown. Young patients differ from other patients with stroke with regards to risk factors, causes, and outcome of stroke Therefore, the objectives of our study were to characterize PI and PSI in young patients with first-ever stroke and to explore whether they are associated with worse 3-month outcome and longterm mortality or recurrent vascular events. Patients and Methods Center and Patients This retrospective study is based on the Helsinki Young Stroke Registry data. The registry includes all consecutive 15- to 49-year-old Continuing medical education (CME) credit is available for this article. Go to to take the quiz. Received May 9, 2013; accepted September 5, From the Department of Neurology, Helsinki University Central Hospital, Helsinki, Finland. Guest Editor for this article was Christoph Kleinschnitz, MD. The online-only Data Supplement is available with this article at /-/DC1. Correspondence to Terttu Heikinheimo, MD, Department of Neurology, Helsinki University Central Hospital, Haartmaninkatu 4, FIN Helsinki, Finland. terttu.heikinheimo-connell@hus.fi 2013 American Heart Association, Inc. Stroke is available at DOI: /STROKEAHA

2 3332 Stroke December 2013 patients with first-ever ischemic stroke diagnosed and treated in the Department of Neurology, Helsinki University Central Hospital between January 1994 and May The local Ethics Committee has approved the study. We excluded patients that were treated as outpatients only or had the index stroke when hospitalized for another reason, and those whose time of symptom onset was unknown. We further excluded patients who did not have 3-month outcome data and those having >2 days from the onset of stroke symptoms to admission (brain imaging and confirmation of diagnosis) and first laboratory testing. The 2-day limit was set to minimize patient-related recall bias with respect to PIs and decrease the probability of misclassification of infections that emerged after stroke onset as PIs. Of the 1008 patients in the registry, we excluded 189 patients who were admitted later than 2 days from the onset of stroke symptoms, 48 who had unknown stroke onset time, 43 who were treated as outpatients only, and 23 who had stroke when already inpatient for another reason. After excluding the abovementioned patients, we excluded 67 patients with no aural temperature measured and 16 with missing inflammatory laboratory data. Furthermore, 2 patients had missing 3-month outcome data, 7 patients were lost to long-term follow-up, 5 refused to participate in the follow-up, and 13 had active malignancy, all of whom were excluded. Thus, 681 patients were included, of them 20 died during the first 30 days. Median duration of hospital stay until discharge to home or to a rehabilitation unit was 10 days (interquartile range, 7 17). Mean follow-up in surviving patients was 7.8 (±4.0) years. Definitions of s and Stroke Risk Factors Preceding acute, subacute, or chronic PI was defined as any signs of clinical infection or laboratory findings suggesting infection at stroke onset or reported symptoms of any infectious disease within the 4-week period before ischemic stroke. Our institutional guidelines instruct to inquire about PIs, especially from younger patients. PSI was defined as any acute infection occurring within 7 days after the onset of stroke. Fever was defined as aural temperature 38.0 C. Patients with fever only but no clinical signs of infection were not regarded as PI or PSI. The same applied to treatments with antibiotics without signs of infections. s were divided into 7 groups: (1) severe infections, including sepsis and endocarditis; (2) chest infections (defined as respiratory crackles combined with 1 of the following: fever, new purulent sputum, or positive chest radiography), including pneumonia and pleuritis caused by infection; (3) upper respiratory infections, including flu-like symptoms, sinusitis, bronchitis, and otitis media; (4) genitourinary infections, including urinary tract infections (defined as clinical symptoms combined with a positive urine dipstick examination for nitrite or pyuria), pyelonephritis, and endometritis; (5) gastrointestinal infections, including gastroenteritis (defined as signs and symptoms of infection in the digestion system); (6) skin and mucous infections, including infections caused by foreign objects such as intravenousline, dermatitis, and otitis externa; and (7) chronic PI, such as parodontitis or HIV infection. Patients with infections on multiple focuses (using the above-defined groups) or patients with both PI and PSI were also counted. Viral hepatitis was not considered as PI. Stroke risk factors were obtained from the medical records and clinical investigations during hospitalization. Hypertension (treated, or a history of hypertension as systolic blood pressure 140 mm Hg or diastolic blood pressure 90 mm Hg, or both), hypercholesterolemia (treated or total cholesterol level 5.0 mmol/l, low-density lipoprotein level 3.0 mmol/l, or high-density lipoprotein level <1.0 mmol/l), obesity (body mass index 30 kg/m 2 or patient clearly described as heavily obese), cardiovascular disease (prior diagnosis of coronary heart disease or myocardial infarction, heart failure [ejection fraction <55%], or peripheral arterial disease), atrial fibrillation, current smoking, and heavy alcohol use (estimated >200 g of alcohol per week consistently as reported by patient or next-of-kin) were recorded. Diabetes mellitus was defined as treated or presently diagnosed according to the 2004 World Health Organization criteria. Type 1 diabetes mellitus was distinguished from type 2 diabetes mellitus by the observance of insulin dependence within 1 year of diagnosis. 19 Patients with malignancy (diagnosed within 1 year or prior malignancy not in remission) were excluded from the present analysis. Data Collection Study researchers (J.B., T.H., and J.P.) reviewed all medical, laboratory, and imaging records from the patients hospital charts, reaching consensus on classification of infections when necessary. Aural temperature, inflammatory markers, such as leukocytes (reference range, cells/l) and C-reactive protein (CRP; reference range, <5 mg/l), on admission and the maximum values during the hospital stay were collected. Patients who did not have these tests performed were excluded from the analysis. Because high-sensitive CRP measurement was not used, CRP values of <5 mg/l were transformed to 2.5 mg/l to analyze CRP as a continuous variable. Urinary cultures and blood culture results were also obtained. Stroke Characteristics Clinical stroke subtype was classified according to the Trial of Org in Acute Stroke Treatment criteria. 26 Stroke severity on admission was assessed by the National Institutes of Health Stroke Scale (NIHSS) and categorized as mild (0 6), moderate, 7 14 or severe ( 15). All patients underwent routine brain imaging, computed tomography, or MRI on admission. All images were evaluated by neuroradiologist, and image reports were evaluated by the clinicians. Measured from the images, the size of the largest ischemic lesion was classified on the basis of documented criteria 27 with the following modifications: small, <1.5-cm lesion in the anterior or posterior circulation; medium, lesion in a cortical superficial branch of the anterior cerebral artery, middle cerebral artery, posterior cerebral artery, or in a deep branch of middle cerebral artery or posterior cerebral artery, or lesion in internal border-zone territories; large posterior, >1.5-cm lesion involving brain stem or cerebellum or involving complete territory of posterior cerebral artery together with border-zone territories; and large anterior, lesion involving complete territory of anterior cerebral artery or middle cerebral artery, or lesion involving >1 artery territory. Lesion size in patients with clinical deficit lasting >24 hours but no visible ischemic lesion in any of the brain imaging was considered small. Outcomes Modified Rankin Scale was used to evaluate the functional outcome after 3 months from the index stroke; unfavorable outcome was defined as modified Rankin Scale 2 to 6. The surviving patients were contacted between November 2009 and January 2010 with the means of a structured telephone interview of patients, their next-of-kin, or nursing staff. Patients who could not be reached by telephone received a follow-up letter. Complete mortality data were obtained from Statistics Finland. Long-term outcome events included in the analysis were (1) nonfatal or fatal recurrent ischemic stroke; (2) composite vascular event of any stroke (ischemic or hemorrhagic), myocardial infarction, revascularization procedure, for example, coronary artery bypass, other arterial occlusive event, or vascular death; and (3) death from any cause. Evaluation of all patient records confirmed outcome events. Statistical Analyses Normality of continuous variables was assessed. Chi-square and Fisher exact tests allowed comparison of categorical variables. Student t test and Mann Whitney U test were used to compare normally and non-normally distributed continuous variables, respectively. Because some patients are likely to be hospitalized because of infections when having a stroke, we did a univariate comparison of the central characteristics between those 23 having inhospital stroke and the remaining cohort. Binary backward stepwise logistic regression analysis using likelihood ratio test was used to identify factors associated with infections and unfavorable 3-month outcome. The model assessing 3-month outcome was adjusted for age and sex, and variables known to correlate with unfavorable outcome in this patient population in

3 Heikinheimo et al s in the Young With Ischemic Stroke 3333 a univariate analysis were forced in the model (dyslipidemia, heart failure, peripheral arterial disease, smoking, heavy drinking, admission NIHSS score, causes, lesion size, lesion in multiple territories). 28 Furthermore, the recorded inflammatory parameters, including highest values of body temperature, leukocyte count, and CRP, were included in the model to control for potential stroke-induced inflammatory or thermoregulatory responses. PSI was included in the model investigating the effect of PI and vice versa. Because medical treatment and the use of antibiotic prophylaxis may have become more frequent during the study period, time trends of the effect of PI or PSI on functional outcome were further assessed with χ 2 test. Kaplan Meier log-rank test served for univariate analysis of the effect of PI and PSI on the risk for long-term end point events. Because mortality was the only long-term outcome, in which there was significant difference between the groups with or without PI and with or without PSI, only that end point was analyzed in a multivariate model. Thus, after exclusion of patients who died within 30 days from the index stroke, a Cox proportional hazard model with backward stepwise algorithm using likelihood ratio test was constructed. Age and sex and covariates known to associate with higher mortality in this patient population (dyslipidemia, smoking, hypertension, cardiovascular disease, type 1 and type 2 diabetes mellitus, heavy drinking, stroke severity, causes, and lesion size) were forced in the model. This model also included the same above-mentioned inflammatory parameters. Statistical analyses were performed using SPSS version 20 for Macintosh (IBM Corp, Armonk, NY). Two-sided significance was set at P<0.05. Results Preceding s A total of 70 (10.3%) patients had PI during the 4-week period before the index stroke. The only risk factor more common in patients with PI was heart failure. We found no difference in stroke causes according to Trial of Org in Acute Stroke Treatment, symptom severity (NIHSS), lesion topography, or its size between those with and without PI. Patients with PI had significantly higher admission CRP values but no difference in admission leukocyte count or body temperature. The highest measured levels of CRP, leukocyte count, and body temperature during hospitalization were observed in patients with PI (Table 1). The most common PI subtype was upper respiratory tract infection (Table 2). Median duration of hospital stay was similar in patients with and without PIs (12 [7 19] days versus 10 [7 17] days; P=0.413). The modified Rankin Scale distribution of patients with and without PI is shown in Figure 1. After adjustment for demographics, risk factors, stroke severity and subtype, lesion size, presence of multiterritorial lesions, PSI, and highest levels of inflammatory parameters, PI remained independently associated with unfavorable 3-month outcome (odds ratio, 2.86; 95% confidence interval [CI], ; P=0.002; Table I in the online-only Data Supplement). We found no time trend in the effect of PI and functional outcome (data not shown). PIs had no effect on recurrent ischemic stroke (log-rank, P=0.323) or composite of vascular end points (P=0.157). After exclusion of patients who died within 30 days from the index stroke, we did not find increased risk of all-cause death in those with PI compared with those without (P=0.500). Poststroke s One or more PSIs occurred in 103 (15.1%) patients. Except for more frequent (P=0.045) migraine among subjects without PSI, clinical risk factors did not differ between the 2 groups. Patients with PSI had more severe stroke symptoms on admission (NIHSS), larger lesion size, and more often bilateral or multiterritorial lesions when compared with the patients without PSI. Large-artery atherosclerosis, high-risk cardioembolism, and carotid artery dissection were main causes of first-ever stroke in patients with PSI (Table 3). On admission, leukocyte count and CRP, but not body temperature, were higher in patients who developed PSI. Regarding highest recorded values of leukocyte count, CRP, and body temperature, all were significantly higher in patients with PSI (Table 3). The most common PSI subtype was lower respiratory tract infections (Table 2). Median duration of hospital stay was longer in patients with PSI than in those without (20 [9 37] days versus 10 [7 14] days; P<0.001). The modified Rankin Scale distribution of patients with and without PSI is shown in Figure 1. After adjustment for demographics, risk factors, stroke severity and subtype, lesion size, presence of multiterritorial lesions, and PI, patients with PSI had exhibited higher risk for unfavorable outcome when compared with those without PSI (odds ratio, 2.26; 95% CI, ; P=0.031; Table I in the online-only Data Supplement). There was no time trend in the effect of PSI and functional outcome (data not shown). As with PI, PSI had no statistically significant effect on the long-term risk of recurrent ischemic stroke (log-rank, P=0.124) or composite of vascular events (P=0.170) in our cohort. Nevertheless, patients who had PSIs continued to exhibit higher mortality during the long-term follow-up with a profound early mortality (Figure 2). After exclusion of those who died within 30 days and after adjusting for demographics, comorbidities known to associate with higher mortality in this patient population, stroke severity and causes, lesion size, and inflammatory parameters; hazard ratio for PSI was 2.37 (95% CI, ; P=0.002; Table II in the online-only Data Supplement). s in Multiple Foci and Patients With Both PI and PSI Of those with PI, 4 patients had an infection in more than one focus without univariate effect on 3-month unfavorable outcome (odds ratio, 2.9; 95% CI, ). A total of 13 patients had PSI in multiple foci that increased the probability of unfavorable outcome 5-fold (univariate odds ratio, 5.2; 95% CI, ). All 16 patients with both PI and PSI (Table 4) had unfavorable 3-month outcome. Differences Between Patients With Inhospital Stroke and the Remaining Cohort Compared with the included cohort, the excluded patients with inhospital stroke (23 patients) had more frequently any PIs (42.1% versus 10.3%; P<0.001) or PSIs (42.1% versus 15.1%; P=0.005). Specifically, they more often had pre-existing severe infections (10.5% versus 0.6%; P=0.010) or chest infections (21.1% versus 1.2%; P<0.001) and, correspondingly, their PSIs were likewise more frequently had severe infections (10.5% versus 0.7%; P=0.014) or chest

4 3334 Stroke December 2013 Table 1. Baseline Characteristics and Their Univariate Association With Preceding s All (n=681) No Preceding (n=611) Preceding (n=70) P Value Demographics Age, y 44 (38 47) 44 (38 47) 42 (36 47) Male 424 (62.3) 379 (62.0) 45 (64.3) Risk factors Dyslipidemia 398 (58.4) 357 (58.4) 41 (58.6) Cigarette smoking 299 (43.9) 270 (44.2) 29 (41.4) Hypertension 255 (37.4) 230 (37.6) 25 (35.7) Obesity 75 (11.0) 68 (11.1) 7 (10.0) Cardiovascular disease Coronary heart disease 32 (4.7) 29 (4.7) 3 (4.3) Heart failure 33 (4.8) 25 (4.1) 8 (11.4) Myocardial infarction 25 (3.7) 22 (3.6) 3 (4.3) Peripheral arterial disease 10 (1.5) 8 (1.3) 2 (2.9) History of transient ischemic attack 57 (8.4) 54 (8.8) 3 (4.3) Diabetes mellitus, type 1 27 (4.0) 22 (3.6) 5 (7.1) Diabetes mellitus, type 2 43 (6.3) 42 (6.9) 1 (1.4) Atrial fibrillation 32 (4.7) 29 (4.7) 3 (4.3) Migraine 120 (17.6) 106 (17.3) 14 (20.0) Heavy drinking 101 (14.8) 96 (15.7) 5 (7.1) Obstructive sleep apnea 27 (4.0) 24 (3.9) 3 (4.3) Gravidity or postpartum period Stroke characteristics NIHSS score, median (range) 3 (2 8) 3 (2 8) 4 (1 10) Infarct size Small 288 (42.3) 261 (42.7) 27 (38.6) Medium 180 (26.4) 160 (26.2) 20 (28.6) Large posterior 84 (12.3) 77 (12.6) 7 (10.0) Large anterior 129 (18.9) 113 (18.5) 16 (22.9) Territorial distribution Single territory 648 (95.2) 583 (95.4) 65 (92.9) Multiple territory 33 (4.8) 28 (4.6) 5 (7.1) Lesion laterality Left 316 (48.1) 288 (48.9) 28 (41.2) Right 276 (42.0) 244 (41.4) 32 (47.1) Bilateral 65 (9.9) 57 (9.7) 8 (11.8) Stroke causes Large-artery atherosclerosis 55 (8.1) 70 (11.5) 11 (15.7) Low-risk source of CE 81 (11.9) 70 (11.5) 11 (15.7) High-risk source of CE 74 (10.9) 61 (10.0) 13 (18.6) Small-vessel disease 80 (11.7) 74 (12.1) 6 (8.6) Internal carotid artery dissection 56 (8.2) 47 (7.7) 9 (12.9) Vertebral artery dissection 58 (8.5) 53 (8.7) 5 (7.1) Other, nondissection 66 (9.7) 61 (10.0) 5 (7.1) Undetermined causes 211 (31.0) 193 (31.6) 18 (25.7) Inflammatory parameters Admission leukocyte count, 10 9 cells/l 8.3 ( ) 8.2 ( ) 8.5 ( ) Highest leukocyte count, 10 9 cells/l 9.1 ( ) 9.1 ( ) 10.1 ( ) Admission C-reactive protein, mg/l 3 (3 7) 3 (3 6) 4 (3 23) <0.001 Highest C-reactive protein, mg/l 5 (3 15) 3 (3 14) 12 (3 51) <0.001 Admission body temperature, C 36.6 ( ) 36.6 ( ) 36.6 ( ) Highest body temperature, C 37.2 ( ) 37.1 ( ) 37.4 ( ) Data are median (interquartile range) or n (%). CE indicates cardioembolism; and NIHSS, National Institutes of Health Stroke Scale.

5 Heikinheimo et al s in the Young With Ischemic Stroke 3335 Table 2. Frequencies of Preceding and Poststroke s Preceding s (n=70) Poststroke s (n=103) Severe infections 4 (5.7) 5 (4.9) Chest infections 8 (11.4) 52 (50.5) Upper respiratory tract infections 38 (54.3) 11 (10.7) Genitourinary infections 3 (4.3) 36 (35.0) Gastrointestinal infections 9 (12.9) 2 (1.9) Skin and mucous infections 8 (11.4) 13 (12.6) Chronic infections 4 (5.7) NA One patient may have had 1 infections. NA indicates not applicable. infections (26.3% versus 7.6%; P=0.014). Patients with inhospital stroke had more commonly unfavorable 3-month outcome (78.9% versus 52.7%; P=0.024). Discussion The main findings in our study were that among young patients with ischemic stroke, 10% had infection preceding the event, whereas 15% had a PSI. A more important fact is that both of these manifestations were independent of each other and independent of well-known robust prognostic factors associated with more than doubled risk of unfavorable 3-month outcome. Apart from heart failure as a risk factor, PI did not affect the stroke characteristics and length of hospital stay. However, our results suggest that PI increases the disability and risk of death during the first 3 months after ischemic stroke in young patients. This is contradicting the previous studies that included general stroke patients 9,10 and also hemorrhagic strokes. 9 Frequency of PI was lower in these studies than in our study, 5.7% and 9.7%, respectively, possibly because young onset stroke carries a higher relative risk of stroke associated with infection as a triggering mechanism. 4,29 Our findings are based on retrospective data, so they must be replicated in a prospective sample with standardized collection of history of recent infections, possibly accompanied by serology and plasma biomarkers. 30 However, the fact that PI was independently associated with adverse 3-month outcome may even now have clinical implications because detection of a recent infection in a patient with stroke may serve as one Figure 1. Three-month modified Rankin Scale distribution according to preceding infection (PI) or poststroke infection (PSI). PI+ indicates patients with preceding infection; PI, patients without preceding infection; PSI+, patients with poststroke infection; and PSI, patients without poststroke infection. Table 3. Baseline Characteristics and Their Univariate Association With Poststroke s No Poststroke (n=578) Poststroke (n=103) P Value Demographics Age, y 44 (37 47) 45 (38 47) Male 363 (62.8) 61 (59.2) Risk factors Dyslipidemia 345 (59.7) 53 (51.5) Cigarette smoking 256 (44.3) 43 (41.7) Hypertension 220 (38.1) 35 (34.0) Obesity 67 (11.6) 8 (7.8) Cardiovascular disease Coronary heart disease 27 (4.7) 5 (4.9) Heart failure 24 (4.2) 9 (8.7) Myocardial infarction 18 (3.1) 7 (6.8) Peripheral arterial disease 9 (1.6) 1 (1.0) History of transient ischemic 51 (8.8) 6 (5.8) attack Diabetes mellitus, type 1 22 (3.8) 5 (4.9) Diabetes mellitus, type 2 37 (6.4) 6 (5.8) Atrial fibrillation 28 (4.8) 4 (3.9) Migraine 109 (18.9) 11 (10.7) Heavy drinking 81 (14.0) 20 (19.4) Obstructive sleep apnea 24 (4.2) 3 (2.9) Gravidity or postpartum period Stroke characteristics NIHSS score 3 (1 6) 11 (3 18) <0.001 Infarct size <0.001 Small 272 (47.1) 16 (15.5) Medium 159 (27.5) 21 (20.4) Large posterior 60 (10.4) 24 (23.3) Large anterior 87 (15.1) 42 (40.8) Territorial distribution Single territory 557 (96.4) 91 (88.3) Multiple territories 21 (3.6) 12 (11.7) Lesion laterality <0.001 Left 272 (49.1) 44 (42.7) Right 240 (43.3) 36 (35.0) Bilateral 42 (7.6) 23 (22.3) Stroke causes <0.001 Large-artery atherosclerosis 42 (7.3) 13 (12.6) Low-risk source of CE 72 (12.5) 9 (8.7) High-risk source of CE 58 (10.0) 16 (15.5) Small-vessel disease 77 (13.3) 3 (2.9) Internal carotid artery 39 (6.7) 17 (16.5) dissection Vertebral artery dissection 49 (8.5) 9 (8.7) Other, nondissection 59 (10.2) 7 (6.8) Undetermined causes 182 (31.5) 29 (28.2) (Continued )

6 3336 Stroke December 2013 Table 3. Continued No Poststroke (n=578) Poststroke (n=103) P Value Inflammatory parameters Admission leukocyte count, 8.1 ( ) 9.9 ( ) < cells/l Highest leukocyte count, 8.7 ( ) 12.4 ( ) < cells/l Admission C-reactive 3 (3 6) 3 (3 13) <0.001 protein, mg/l Highest C-reactive 2.5 (2.5 10) 58 (15 138) <0.001 protein, mg/l Admission body 36.6 ( ) 36.6 ( ) temperature, C Highest body temperature, C 37.1 ( ) 37.9 ( ) <0.001 Data are median (interquartile range) or n (%). CE indicates cardioembolism; and NIHSS, National Institutes of Health Stroke Scale. relevant selection criterion for immediate antimicrobial therapies, which are currently being tested in randomized trials, although to date results have been inconsistent. 31 The reasons why PI would lower odds for good recovery after ischemic stroke are unknown but some could be proposed. First, inflammatory processes are important factors in the cause of ischemic cerebrovascular disease, as well as in the pathophysiology of cerebral ischemia. 4,5,7,8,32 Second, the pathogenesis of atherosclerosis starts already when young, and the prevalence of atherosclerotic findings in intra- and extracranial arteries of young patients with stroke increases already at early middle age. 33 Existing intracranial atherosclerosis may hamper penumbral salvage in brain ischemia. Third, as observed in healthy children, even minor infections can cause endothelial dysfunction. 34 This effect could explain the triggering effect of infection to ischemic stroke not only on the endothelium, but also on clinical outcome. Endothelium of the arteries is also part of a neurovascular unit with neurons and supporting nerve cells that plays part in the cellular Survival rate (%) No post-stroke infection Post-stroke infection Years to death Figure 2. Kaplan Meier estimates showing cumulative long-term mortality risk according to the presence of poststroke infection (log-rank, P<0.001). Mean survival time for patients with and without poststroke infection was 10.3 (95% confidence interval [CI], ) years and 14.3 (95% CI, ) years, respectively. Table 4. Cross-Table of Frequencies of Preceding and Poststroke s Without Preceding s (n=611) With Preceding (n=70) Without poststroke infection (n=578) 524 (85.5) 54 (77.1) With poststroke infection (n=103) 87 (14.2) 16 (22.9) Data are n (%). inflammation when ischemic brain injury happens. 35 It has been suggested that arterial dissection, which is a common cause of stroke in the young patients, may be caused by vessel wall inflammation triggered by PI. 36 Inflammatory agents, such as cytokines and neutrophils, cause secondary inflammation and thus induce secondary damage in the penumbral area of the stroke. 37 Taken together, existing atherosclerosis, endothelial dysfunction, and its adverse effect on penumbra and neuroregeneration might explain the negative effect on clinical prognosis after stroke in the setting of PI. In contrast to short-term effect, PI did not affect long-term outcomes (recurrent ischemic stroke, other vascular events, or all-cause mortality) in our study. A likely explanation is that the transient nature of acute infection, most of them being banal upper respiratory tract infections, acts as a temporary risk factor for stroke and extends to have effect on short-term outcome only. 3 Compared with PI, the association between PSI and adverse outcome after stroke is based on much more solid evidence We confirmed that young age does not protect from the deleterious effects of PSI. As in previous studies, also in our patients, PSI was more frequent in those with more severe stroke: they had higher NIHSS scores, larger lesion sizes, and more often multiterritorial or bilateral distribution of lesions. Unlike with PI, in patients having a PSI most of the inflammatory parameters were increased: leukocyte count and CRP on admission and the highest values of leukocyte count, CRP, and body temperature. PSI also increased the long-term all-cause mortality. There were trends toward increased risk of recurrent ischemic stroke and composite of vascular events (P 0.17). Leukocytosis during acute phase has been used as a predictor of developing PSI: ischemic stroke causes inflammation at its site and potentially increases the size of stroke lesion. Inflammation also causes peripheral immunosuppression that makes patient more vulnerable to PSI thus worsening the outcome. 38 The alteration in immunoresponse and risk of poststroke pneumonia can be predicted by recently developed, validated PANTHERIS score that takes into account the high leukocyte count (> cells/l) within 24 hours from admission and usual risk factors (such as age and size of stroke) of stroke-associated pneumonia. 39 The main limitation of our study is its retrospective nature and the extensive time period it is covering. Another limitation is that we did not use more detailed laboratory testing during the acute phase of stroke; for instance, high-sensitivity CRP and differential count of leukocytes on admission would have given more detailed information. As mentioned above, there is a need of prospective study where history of PI is carefully taken from all the patients and their next-of-kin and where plasma biomarkers are included in the protocol. The definition of PI largely depends on the history taken if symptoms of infection have resolved before stroke that carries a risk of

7 Heikinheimo et al s in the Young With Ischemic Stroke 3337 reporting bias especially with aphasic patients and those with very severe stroke. The strengths include a fairly large wellinvestigated patient population and the long follow-up period. According to our results, it seems that the diagnosis of PI or PSI independently reflects to 3-month outcome of stroke. Disclosures This work was supported by the Helsinki University Central Hospital (Dr Putaala, TKK ; Dr Haapaniemi, TKK ; Dr Tatlisumak, TKK ), the Finnish Medical Foundation (Drs Putaala and Haapaniemi), the Paavo Nurmi Foundation (Dr Haapaniemi), and the Emil Aaltonen Foundation (Dr Haapaniemi). References 1. Grau AJ, Buggle F, Heindl S, Steichen-Wiehn C, Banerjee T, Maiwald M, et al. Recent infection as a risk factor for cerebrovascular ischemia. Stroke. 1995;26: Zurrú MC, Alonzo C, Brescacín L, Romano M, Cámera LA, Waisman G, et al. Recent respiratory infection predicts atherothrombotic stroke: case- control study in a Buenos Aires healthcare system. Stroke. 2009;40: Guiraud V, Amor MB, Mas JL, Touzé E. Triggers of ischemic stroke: a systematic review. Stroke. 2010;41: Syrjänen J, Valtonen V, Iivanainen M, Kaste M, Huttunen JK. Preceding infection as an important risk factor for ischaemic brain infarction in young and middle aged patients. BMJ. 1988;296: Nencini P, Sarti C, Innocenti R, Pracucci G, Inzitari D. Acute inflammatory events and ischemic stroke subtypes. Cerebrovasc Dis. 2003;15: Paganini-Hill A, Lozano E, Fischberg G, Perez Barreto M, Rajamani K, Ameriso SF, et al. and risk of ischemic stroke: differences among stroke subtypes. Stroke. 2003;34: Syrjänen J, Peltola J, Valtonen V, Iivanainen M, Kaste M, Huttunen JK. Dental infections in association with cerebral infarction in young and middle-aged men. J Intern Med. 1989;225: Grau AJ, Buggle F, Ziegler C, Schwarz W, Meuser J, Tasman AJ, et al. Association between acute cerebrovascular ischemia and chronic and recurrent infection. Stroke. 1997;28: Roquer J, Cuadrado-Godia E, Giralt-Steinthauer E, Jimena S, Jiménez- Conde J, Martínez-Rodríguez JE, et al. Previous infection and stroke: a prospective study. Cerebrovasc Dis. 2012;33: Grabska K, Gromadzka G, Członkowska A. s and ischemic stroke outcome. Neurol Res Int. 2011;2011: Westendorp WF, Nederkoorn PJ, Vermeij JD, Dijkgraaf MG, van de Beek D. Post-stroke infection: a systematic review and meta-analysis. BMC Neurol. 2011;11: Davenport RJ, Dennis MS, Wellwood I, Warlow CP. Complications after acute stroke. Stroke. 1996;27: Aslanyan S, Weir CJ, Diener H-C, Kaste M, Lees KR. Pneumonia and urinary tract infection after acute ischaemic stroke: a tertiary analysis of the GAIN international trial. Eur Neurol. 2004;11: Kumar S, Selim MH, Caplan LR. Medical complications after stroke. Lancet Neurol. 2010;9: Vargas M, Horcajada JP, Obach V, Revilla M, Cervera A, Torres F, et al. Clinical consequences of infection in patients with acute stroke: is it prime time for further antibiotic trials? Stroke. 2006;37: Hong KS, Kang DW, Koo JS, Yu KH, Han MK, Cho YJ, et al. Impact of neurological and medical complications on 3-month outcomes in acute ischaemic stroke. Eur J Neurol. 2008;15: Vermeij FH, Scholte op Reimer WJM, de Man P, van Oostenbrugge RJ, Franke CL, de Jong G, et al. Stroke-associated infection is an independent risk factor for poor outcome after acute ischemic stroke: data from the Netherlands stroke survey. Cerebrovasc Dis. 2009;27: Kwan J, Hand P. after acute stroke is associated with poor short-term outcome. Acta Neurol Scand. 2007;115: Wartenberg KE, Stoll A, Funk A, Meyer A, Schmidt JM, Berrouschot J. after acute ischemic stroke: risk factors, biomarkers, and outcome. Stroke Res Treat. 2011;2011: Putaala J, Metso AJ, Metso TM, Konkola N, Kraemer Y, Haapaniemi E, et al. Analysis of 1008 consecutive patients aged 15 to 49 with first-ever ischemic stroke: the Helsinki young stroke registry. Stroke. 2009;40: Haapaniemi H, Hillbom M, Juvela S. Lifestyle-associated risk factors for acute brain infarction among persons of working age. Stroke. 1997;28: You RX, McNeil JJ, O Malley HM, Davis SM, Thrift AG, Donnan GA. Risk factors for stroke due to cerebral infarction in young adults. Stroke. 1997;28: Naess H, Nyland HI, Thomassen L, Aarseth J, Myhr KM. Etiology of and risk factors for cerebral infarction in young adults in western Norway: a population-based case-control study. Eur J Neurol. 2004;11: Marini C, Totaro R, Carolei A. Long-term prognosis of cerebral ischemia in young adults. National Research Council Study Group on Stroke in the Young. Stroke. 1999;30: Leys D, Bandu L, Hénon H, Lucas C, Mounier-Vehier F, Rondepierre P, et al. Clinical outcome in 287 consecutive young adults (15 to 45 years) with ischemic stroke. Neurology. 2002;59: Adams HP Jr, Bendixen BH, Kappelle LJ, Biller J, Love BB, Gordon DL, et al. Classification of subtype of acute ischemic stroke. Definitions for use in a multicenter clinical trial. TOAST. Trial of Org in Acute Stroke Treatment. Stroke. 1993;24: Paciaroni M, Agnelli G, Corea F, Ageno W, Alberti A, Lanari A, 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: Putaala J, Strbian D, Mustanoja S, Haapaniemi E, Kaste M, Tatlisumak T. Functional outcome in young adult ischemic stroke: impact of lipoproteins. Acta Neurol Scand. 2013;127: Grau AJ, Buggle F, Becher H, Zimmermann E, Spiel M, Fent T, et al. Recent bacterial and viral infection is a risk factor for cerebrovascular ischemia: clinical and biochemical studies. Neurology. 1998;50: Fullerton HJ, Elkind MS, Barkovich AJ, Glaser C, Glidden D, Hills NK, et al. The vascular effects of infection in Pediatric Stroke (VIPS) Study. J Child Neurol. 2011;26: Westendorp WF, Vermeij JD, Vermeij F, Den Hertog HM, Dippel DW, van de Beek D, et al. Antibiotic therapy for preventing infections in patients with acute stroke. The Cochrane Database Syst Rev. 2012;1:CD Lindsberg PJ, Grau AJ. Inflammation and infections as risk factors for ischemic stroke. Stroke. 2003;34: von Sarnowski B, Schminke U, Tatlisumak T, Putaala J, Grittner U, Kaps M, et al. Prevalence of stenoses and occlusions of brain-supplying arteries in young stroke patients. Neurology. 2013;80: Charakida M, Donald AE, Terese M, Leary S, Halcox JP, Ness A, et al. Endothelial dysfunction in childhood infection. Circulation. 2005;111: del Zoppo GJ. Inflammation and the neurovascular unit in the setting of focal cerebral ischemia. Neuroscience. 2009;158: Pfefferkorn T, Saam T, Rominger A, Habs M, Gerdes LA, Schmidt C, et al. Vessel wall inflammation in spontaneous cervical artery dissection: a prospective, observational positron emission tomography, computed tomography, and magnetic resonance imaging study. Stroke. 2011;42: Heiss WD. The ischemic penumbra: how does tissue injury evolve? Ann N Y Acad Sci. 2012;1268: Vogelgesang A, Becker KJ, Dressel A. Immunological consequences of ischemic stroke. J Neuroimmunol. 2011;231: Harms H, Grittner U, Droge H, Meisel A. Predicting post-stroke pneumonia: the PANTHERIS score. Acta Neurol Scand. 2013;128: