Effects of SSRI exposure on hemorrhagic complications and outcome following thrombolysis in ischemic stroke

Abstract

Background

Selective serotonin reuptake inhibitors (SSRI) may interfere with platelet function, and pre-stroke SSRI treatment has been associated with increased hematoma volumes and mortality in hemorrhagic stroke patients. The effects of SSRI on the risk of hemorrhagic complications after thrombolysis in ischemic stroke patients are unclear.

Aims

To examine the effects of pre-stroke SSRI exposure on bleeding complications, functional outcome, and mortality following thrombolysis in ischemic stroke.

Methods

Data including standard demographic and clinical variables as well as baseline and follow-up stroke severity (measured by National Institutes of Health Stroke Score), functional outcome (measured by modified Rankin Scale) at 3 months, and mortality at 7 and 90 days were extracted from the Virtual International Stroke Trials Archive. Multivariable binary logistic regression was used for statistical analyses.

Results

Out of 1114 ischemic stroke patients treated with recombinant tissue-type plasminogen activator, 135 (12.1%) had previous SSRI exposure. Symptomatic intracranial hemorrhage occurred in 30 (2.7%) patients. Of those, 2 (1.5%, n = 135) were in the SSRI pretreatment group and 28 (2.9%, n = 979) were SSRI naive patients. Pre-stroke SSRI exposure in thrombolysed patients showed association with neither bleeding complications (P = .58) nor functional outcome (P = .38) nor mortality (P = .65).

Conclusions

Results from this large retrospective ad hoc database cohort study indicate that pre-stroke SSRI exposure in ischemic stroke patients who receive thrombolytic treatment is not associated with bleeding complications, functional outcome, or mortality.

Keywords

Stroke SSRI thrombolysis outcome intracerebral hemorrhage mortality

Introduction

Selective serotonin reuptake inhibitors (SSRI) are among the most commonly prescribed classes of medications.^1–3 Their broad therapeutic effect for a multitude of psychiatric disorders and their safe side effect profile have made them a first choice agent of many clinicians³ and multiple studies have reported beneficial effects of SSRI treatment in ischemic stroke patients.^4–11 On the other hand, SSRI use has been associated with bleeding complications in the gastrointestinal system and intracranially due to the inhibition of platelet serotonin reuptake and consequently reduced platelet function.³ Previous investigations have shown that SSRI exposure increases the risk of intracerebral bleeding, in particular when combined with oral anticoagulant treatment.¹ Pre-stroke treatment with SSRI has been linked to increased hematoma volumes and mortality in patients with hemorrhagic stroke, but showed no association with mortality in the general ischemic stroke population.⁵ The effects of SSRI pretreatment on bleeding complications after thrombolysis for ischemic stroke are currently unclear.

Hypothesis

In this study, we hypothesized that pre-stroke SSRI exposure in patients receiving recombinant tissue-type plasminogen activator (rt-PA) is associated with an increased risk of thrombolysis-related bleeding complications, unfavorable functional outcome, and increased mortality.

Methods

Data source and processing

Data from ischemic stroke patients comprising standard demographic and clinical variables were extracted from the Virtual International Stroke Trials Archive (VISTA) database (http://vistacollaboration.org),¹² including age, sex, onset-to-treatment time (OTT), administration of rt-PA, pre-stroke SSRI exposure, stroke severity by National Institutes of Health Stroke Score (NIHSS) at admission (baseline NIHSS) and in the course of 24 to 48 h (NIHSS_24–48h), modified Rankin Scale (mRS) at 3 months from stroke onset, 90 days mortality, time to death, admission platelet count, baseline glycemia, previous stroke, pre-admission statin and antiplatelet therapy, hypertension, diabetes mellitus, history of myocardial infarction (MCI), atrial fibrillation (aFib), ischemic heart disease (IHD), congestive heart failure (CHF), and transient ischemic attacks (TIA). Adverse events (AE) records were screened for bleeding complications, which were then categorized into intra- and extracranial bleeding events. If available, grading information was retrieved for intracerebral hemorrhages according to the European Cooperative Acute Stroke Study (ECASS) radiological classification of post-thrombolysis brain hemorrhage.^13–15 Based on previous reports, we considered any new occurrence of intracranial hemorrhage within 36 h from rt-PA administration as symptomatic intracranial hemorrhage (SICH) and adverse side effect of rt-PA thrombolysis, if the bleeding was associated with an NIHSS increase of ≥4 points or leading to death within 48 h from thrombolysis treatment.^15–21 In addition, we regarded any grade 1 (PH1) and grade 2 (PH2) parenchymal hematoma (according to ECASS classification) which newly emerged within 36 h from thrombolysis as a relevant complication of rt-PA administration, regardless of the availability of NIHSS data.^13–15 Because the available data provided time points in the format of days from admission, a limit of ≤2 days rather than 36 h was chosen for defining early and rt-PA-related bleeding complications. Functional outcome at three months was dichotomized into favorable outcome (mRS 0–2), reflecting independence in the daily activities of life, versus unfavorable outcome (mRS 3–6), indicating death or dependency. Mortality at seven days was inferred from time to death data.

Statistical analysis

Patients were categorized into two groups based on pre-stroke SSRI exposure. To test variables for confounding effects, pre-stroke SSRI exposure as the independent treatment variable and one by one the possible confounders (age, sex, OTT, baseline NIHSS, admission platelet count and glycemia, previous stroke, pre-admission statin and antiplatelet therapy, MCI, aFib, IHD, CHF, TIA) entered a bivariable binary logistic regression model and the adjusted odds ratio (OR) was calculated. When the adjusted OR deviated from the crude OR by >5%, a variable was recorded as possible confounder. A multivariable model with the treatment variable and all variables that shifted the crude OR by >5% was then built. Moreover, due to their well-documented relation to outcome in patients with acute ischemic stroke, the variables age, baseline NIHSS, admission platelet count, and baseline glycemia entered multivariable analysis regardless of bivariable testing results.^6,16 Two-sided values of P<.05 were considered as statistically significant in all tests. Cohen's ƒ² effect sizes of .02, .15, and .35 were regarded as small, medium, and large. All statistics were performed using statistical software IBM SPSS Statistics 20.0 for Windows.

Results

Altogether 1114 patient records were included in the analyses, mean age 72 years (range 36–98; SD 11), 57.6% of patients were male. A total of 135 (36.8%) patients had previous SSRI treatment. In SSRI naive patients and pre-stroke SSRI users, OTT (P = .64), baseline NIHSS (P = .12), and admission glycemia (P = .48) were comparable. Mean age was lower (P = .01) in the SSRI group and female sex prevailed (P<.001). Pre-admission statin (P<.001) and antiplatelet therapy (P<.001) were less common and admission platelet count was higher (P = .001) in patients with pre-stroke SSRI treatment (Table 1).

Table 1.

Characteristics of SSRI naive patients vs. those treated with SSRI before stroke.

Characteristics	SSRI naive (n = 979)	SSRI before stroke (n = 135)	P
Age (y), mean (range, SD)^a	72 (41–98, 10)	69 (36–92, 13)	.001
Sex (female), n (%)^b	386 (39.4)	86 (63.7)	<.001
OTT (h), median (range, IQR)^b	3.5 (1–6, 1) [N = 962]	3.7 (1–6, 1) [N = 134]	.64
Baseline NIHSS, median (range, IQR)^b	14 (3–30, 9)	13 (5–27, 9)	.12
Platelet count (G/l), median (range, IQR)^b	215 (90–688, 79) [N = 785]	232 (117–501, 114) [N = 98]	.001
Admission glycemia (mmol/l), median (range, IQR)^b	6.7 (2.7–28.0, 2.4) [N = 765]	6.7 (3.6–21.7, 2.5) [N = 114]	.48
Pre-stroke statin, n (%)^c	506 (51.7)	48 (35.6)	<.001
Pre-stroke antiplatelets, n (%)^c	801 (81.8)	56 (41.5)	<.001
Previous stroke, n (%)^c	226 (23.3) [N = 971]	37 (27.8) [N = 133]	.28
Hypertension, n (%)^c	795 (81.9) [N = 971]	100 (75.8) [N = 132]	.10
Diabetes mellitus, n (%)^c	252 (26.0) [N = 971]	33 (25.0) [N = 132]	.92
History of myocardial infarction, n (%)^c	255 (26.3) [N = 971]	27 (20.5) [N = 132]	.17
Atrial fibrillation, n (%)^c	307 (31.6) [N = 971]	33 (25.0) [N = 132]	.13
Ischemic heart disease, n (%)^c	368 (47.5) [N = 775]	43 (37.7) [N = 114]	.06
Congestive heart failure, n (%)^c	109 (13.1) [N = 830]	17 (13.8) [N = 123]	.78
Transient ischemic attacks, n (%)^c	86 (11.1) [N = 777]	11 (9.6) [N = 115]	.75

IQR: interquartile range; N, number of samples, if data were not available from all patients; NIHSS: National Institutes of Health Stroke Score; OTT: onset-to-treatment time; SSRI: selective serotonin reuptake inhibitors. P-values, ^aStudent t-test, ^bMann–Whitney U-test, ^cFisher's exact test.

Intracranial bleeding complications

Out of 1114 thrombolysed patients with ischemic stroke, 163 (14.6%) patients developed any type of adverse hemorrhagic event within two days from thrombolysis. Any type of hemorrhagic transformation and/or intracranial hemorrhage was reported in 86 (52.8%) of these patients (7.7% of the study population). SICH occurred in 30 (2.7%) out of 1114 thrombolysed patients and in 28 (2.9%) SSRI naive versus 2 (1.5%) pre-exposed patients. Early PH1 or PH2 were observed in a total of 34 (3.1%) patients and in 30 (3.1%) SSRI naive versus 4 (3.0%) pre-exposed patients (Table 2). Pre-stroke SSRI exposure prior to rt-PA treatment was neither associated with early ICH in general (Adj. OR, .60; 95% CI, .17–2.10; P = .42) nor linked to symptomatic ICH (Adj. OR, .55; 95% CI, .06–4.71; P = .58) or early PH1 and PH2 bleeding complications (Adj. OR, .31; 95% CI, .04–2.67; P = .29) (Table 3).

Table 2.

Bleeding complications, outcome and mortality in SSRI naive patients vs. those treated with SSRI before stroke

Outcome	SSRI naive (n = 979)	SSRI before stroke (n = 135)	P
Early intracranial hemorrhage, n (%)^a	79 (8.1)	7 (5.2)	.30
Symptomatic (SICH), n (%)^a	28 (2.9)	2 (1.5)	.57
HI1 or HI2, n (%)^a	4 (.4)	0 (0)	>.99
PH1 or PH2, n (%)^a	30 (3.1)	4 (3.0)	>.99
Early ECH ( ≤ 2d), n (%)^a	76 (7.8)	11 (8.1)	.86
NIHSS (24–48 h), median (range, IQR)^b	11 (0–38, 12) [N = 778]	10 (0–38, 12) [N = 120]	.29
mRS, median (range, IQR)^b	3 (0–6, 4) [N = 941]	3 (0–6, 2) [N = 133]	.56
Mortality (90 days), n (%)^a	203 (20.7)	16 (11.9)	.02
Mortality (7 days), n (%)^a	93 (9.5)	6 (4.4)	.46
Time to death (days), median (range, IQR)^b	10 (0–104, 22) [N = 205]	22 (3–92, 35) [N = 17]	.07

Early hemorrhage, occurrence at ≤2 days from thrombolysis; ECASS: European Cooperative Acute Stroke Study; HI: hemorrhagic infarction according to the ECASS radiological classification of post-thrombolysis brain hemorrhage; HI1: HI grade 1; HI2: HI grade 2; mRS: modified Rankin Scale at three months; IQR: interquartile range; NIHSS: National Institutes of Health Stroke Score; PH: parenchymal hematoma according to ECASS classification; PH1: PH grade 1; PH2: PH grade 2; SICH: symptomatic intracranial hemorrhage (early occurrence, NIHSS increase of ≥4 points or leading to death within 2 days from thrombolysis). NIHSS (24–48 h), follow-up NIHSS at 24–48 h from admission. P-values, ^aFischer's exact test, ^bMann–Whitney U-test.

Table 3.

Binary logistic regression including pre-stroke SSRI exposure to predict bleeding complications, functional outcome at three months and mortality

Outcome	Crude OR	95% CI	P	Adj. OR	95% CI	P
Early intracranial hemorrhage^a,b,c,d	.62	.28–1.38	.24	.60	.17–2.10	.42
Symptomatic ICH (SICH)^a,b,c,d,e	.50	.12–2.13	.35	.55	.06–4.71	.58
PH1 or PH2^{a,b,c,d,e,f,g}	.97	.34–2.79	.95	.31	.04–2.67	.29
Early extracranial hemorrhage^b,c,d,e,h,i	1.05	.55–2.04	.88	.65	.20–2.04	.46
Unfavorable outcome^a,b,c,g,i	1.03	.71–1.50	.86	1.33	.72–2.47	.38
Mortality (7 days)^b,c,d,e,h,j	.66	.23–1.84	.43	1.15	.27–4.87	.85
Mortality (90 days)^b,c,d,e,i	.51	.30–.89	.02	.84	.39–1.80	.65

Adj. OR: adjusted odds ratio; CI: confidence interval; early hemorrhage, occurrence at ≤2 days from thrombolysis; ECASS: European Cooperative Acute Stroke Study; NIHSS: National Institutes of Health Stroke Score; OR: odds ratio; PH: parenchymal hematoma according to the ECASS radiological classification of post-thrombolysis brain hemorrhage; PH1: PH grade 1; PH2: PH grade 2; SICH: symptomatic intracranial hemorrhage (early occurrence, NIHSS increase of ≥4 points or leading to death within two days from thrombolysis). Adjustment for confounders including age, baseline NIHSS, admission platelet count, baseline glycemia, and if noted ^asex, ^bpre-stroke antiplatelets, ^cischemic heart disease, ^dtransient ischemic attack, ^econgestive heart failure, ^fpre-stroke statin, ^ghypertension, ^honset-to-treatment time, ⁱatrial fibrillation, ^jdiabetes mellitus.

Extracranial bleeding complications

Out of 163 (14.6%) patients with early hemorrhagic events, 87 (53.4%) patients in total and 76 (7.8%) SSRI naive versus 11 (8.1%) pre-exposed patients developed extracranial bleeding complications (Table 2). Pre-stroke SSRI exposure was not associated with early extracranial hemorrhages (Adj. OR, .65; 95% CI, .20–2.04; P = .46) (Table 3).

Functional outcome at three months

Functional outcome data were available for 1074 patients (96.4%). Unfavorable outcome occurred in a total of 663 patients (61.7%) and in 580 (59.2%) SSRI naive versus 83 (61.5%) pre-exposed patients. An association of pre-stroke SSRI exposure with unfavorable mRS was not observed. Adjustment for confounders only had a moderate effect and did not reveal significant associations (Adj. OR, 1.33; 95% CI, .72–2.47; P = .36) (Table 3).

Mortality

Vital status was available for 1114 (100%) patients. By day 90, 219 patients (19.7%) had died. In patients with pre-stroke SSRI exposure, mortality was 11.9% (n = 16) compared to 20.7% (n = 203) in patients without prior SSRI therapy (Table 1). In univariable analysis, pre-stroke SSRI exposure was associated with lower 90-day mortality (crude OR, .51; 95% CI, .30–.89; P = .02, Cohen's ƒ², .01). However, after adjustment for confounders, the association was no longer significant. Mortality at seven days was not associated with pre-stroke SSRI use (Table 3).

Combined pre-stroke antiplatelet and SSRI treatment

In view of the significantly lower prevalence of antiplatelet therapy in patients with pre-stroke SSRI treatment, further exploratory analyses were performed in those patients with pre-stroke antiplatelet therapy (n = 857). In compliance with results from the general study population, pre-stroke SSRI treatment in this subset of patients showed no association with bleeding complications (e.g. SICH, Adj. OR 1.24, 95% CI .13–11.59, P = 0.85), functional outcome (Adj. OR .81, 95% CI .31–2.12, P = .67), or mortality (e.g. 90-day mortality, Adj. OR .90, 95% CI .09–8.85, P = .93).

Discussion

Our results suggest that pre-stroke SSRI exposure prior to rt-PA thrombolysis for ischemic stroke does not increase the risk of intra- or extracranial bleeding complications and is not associated with adverse outcome or increased mortality.

Positive associations of SSRI treatment with structural and functional recovery from brain damage following ischemic stroke have been reported by a growing number of studies with post-stroke SSRI administration and linked to a variety of physiological mechanisms.^4–11 In contrast to potential beneficial associations in ischemic stroke patients, SSRI treatment has been linked to bleeding complications and a higher incidence of cerebrovascular events in the general population.^1,4,22 Such adverse effects of SSRI exposure have mainly been attributed to the inhibition of platelet serotonin reuptake and consequently reduced platelet aggregation and activity.³ In ischemic stroke patients, SICH is a devastating complication of thrombolysis treatment and associated with high morbidity and mortality.^15,23 It has previously been reported that pre-stroke SSRI use is linked to increased mortality in hemorrhagic stroke patients, but not in patients with ischemic stroke.⁵ Other authors have associated pre-stroke SSRI use with higher mortality in the general stroke population,²⁴ unfortunately without reporting the prevalence of hemorrhagic versus ischemic stroke. We investigated ischemic stroke patients in more detail and found that mortality was not associated with pre-stroke SSRI exposure following rt-PA thrombolysis. Lower mean age in the SSRI group may explain lower mortality in univariable comparison.

Reports on functional outcome following rt-PA thrombolysis in pre-stroke SSRI users are inconclusive. One study investigating 476 acute ischemic stroke patients who received thrombolysis treatment reported a trend towards association with unfavorable outcome at three months from rt-PA administration,⁶ however, non-significant and based on a relatively small sample of pre-stroke SSRI users (n = 22). A significant association with unfavorable outcome was only found in patients with cortical stroke, which we were unable to separately assess in our study. In a more recent analysis of 239 ischemic stroke patients, SSRI pretreatment (n = 51) was linked to favorable outcome at discharge, in comparison to post-stroke SSRI administration (n = 188).¹¹ Functional outcome was assessed noticeably earlier than in previous reports^6,10 and in our study, and hospital stay of patients with pre-stroke SSRI exposure (median seven days; interquartile range [IQR], five days) was significantly shorter (P<.001) compared to patients who were newly treated with SSRI (median 11d; IQR, 6d). The authors did not investigate patients who received neither pre- nor post-stroke SSRI and excluded patients who suffered from severe stroke with infaust prognosis. In comparison, our data showed no relation between pre-stroke SSRI exposure and functional outcome at three months. The conflicting results could be related to different inclusion criteria and outcome measures (discharge vs. three-month follow-up) or might be attributed to unknown factors introduced in the later course of the disease, which were not accounted for in this study and in previous comparisons. Data on post-stroke SSRI exposure were not available in our study. While it is likely that patients with pre-stroke SSRI continued SSRI treatment after the index stroke, variations in post-stroke treatment may explain deviating results.

Limitation of our study is the heterogeneous indication for SSRI treatment and its uncontrolled and non-randomized administration with insufficient study data on dosage and/or duration of pre-stroke treatment. Presumably the majority of SSRI users in this study received treatment because of depression, which is by itself associated with increased ischemic stroke risk,^7,25 unfavorable functional outcome,²⁶ higher mortality,²⁵ and confounding comorbidity such as cardiac disease.²⁷ Despite having performed a detailed search for possible confounders in the analysis, confounding by indication may be a source of bias in this study. Furthermore, sufficient data were not available to further discriminate between different types of SSRI and to assess potential effects of other antidepressants. Finally, the results of our study have to be interpreted with caution because of the retrospective, non-randomized nature of the analysis. Notwithstanding the aforementioned limitations, the strength of this study is a large prospective patient cohort with rigorously collected detailed baseline demographics, precise AE reporting, and standardized manner and timing of outcome assessments.

Conclusion

Results from this large retrospective comparison indicate that pre-stroke SSRI exposure in ischemic stroke patients receiving rt-PA is not associated with bleeding complications, functional outcome or mortality. However, due to the retrospective and non-randomized nature of this study, unbiased proof of the SSRI-thrombolysis safety cannot be claimed. As far as currently available data indicate, there is no signal to consider pre-stroke SSRI exposure to be a risk with regard to rt-PA thrombolysis.

Footnotes

Acknowledgments

VISTA-Acute Steering Committee members: KR Lees (Chair), A Alexandrov, PM Bath, E Bluhmki, N Bornstein, C Chen, L Claesson, SM Davis, G Donnan, HC Diener, M Fisher, M Ginsberg, B Gregson, J Grotta, W Hacke, MG Hennerici, M Hommel, M Kaste, P Lyden, J Marler, K Muir, N Venketasubramanian, R Sacco, A Shuaib, P Teal, NG Wahlgren, S Warach, and C Weimar.

Authors' contributions

CS searched the literature, analyzed the data, and wrote the paper. JF and WL contributed to and reviewed the paper. MS extracted the data, reviewed, and made critical revisions of the paper.

Declaration of conflicting interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

References

Hackam

Mrkobrada

. Selective serotonin reuptake inhibitors and brain hemorrhage: a meta-analysis. Neurology 2012; 79: 1862–1865.

McCrea

Sammon

Nazareth

Petersen

. Initiation and duration of selective serotonin reuptake inhibitor prescribing over time: UK cohort study. Br J Psychiatry 2016; 209: 421–426.

Gaduputi

Patel

Sakam

Chime

Balar

Kumar

. Serotonin reuptake inhibitors and post-gastrostomy bleeding: reevaluating the link. Ther Clin Risk Manag 2015; 11: 1283.

Mortensen

Larsson

Johnsen

Andersen

. Post stroke use of selective serotonin reuptake inhibitors and clinical outcome among patients with ischemic stroke: a nationwide propensity score-matched follow-up study. Stroke 2013; 44: 420–426.

Mortensen

Larsson

Johnsen

Andersen

. Impact of prestroke selective serotonin reuptake inhibitor treatment on stroke severity and mortality. Stroke 2014; 45: 2121–2123.

Miedema

Horvath

Uyttenboogaart

et al.

Effect of selective serotonin re-uptake inhibitors (SSRIs) on functional outcome in patients with acute ischemic stroke treated with tPA. J Neurol Sci 2010; 293: 65–67.

Siepmann

Penzlin

Kepplinger

et al.

Selective serotonin reuptake inhibitors to improve outcome in acute ischemic stroke: possible mechanisms and clinical evidence. Brain Behav 2015; 5: e00373.

Mead

Hsieh

C-F

Lee

et al.

Selective serotonin reuptake inhibitors (SSRIs) for stroke recovery. In: Mead

(ed). Cochrane Database of Systematic Reviews, Chichester, UK: John Wiley & Sons Ltd, 2012, pp. CD009286.

Cramer

. Drugs to enhance motor recovery after stroke. Stroke 2015; 46: 2998–3005.

10.

Chollet

Tardy

Albucher

J-F

et al.

Fluoxetine for motor recovery after acute ischaemic stroke (FLAME): a randomised placebo-controlled trial. Lancet Neurol 2011; 10: 123–130.

11.

Siepmann

Kepplinger

Zerna

et al.

The effects of pretreatment versus de novo treatment with selective serotonin reuptake inhibitors on short-term outcome after acute ischemic stroke. J Stroke Cerebrovasc Dis 2015; 24: 1886–1892.

12.

Ali

Bath

PMW

Curram

et al.

The virtual international stroke trials archive. Stroke 2007; 38: 1905–1910.

13.

Hacke

. Intravenous thrombolysis with recombinant tissue plasminogen activator for acute hemispheric stroke. The European Cooperative Acute Stroke Study (ECASS). JAMA 1995; 274: 1017–1025.

14.

Fiorelli

Bastianello

von Kummer

et al.

Hemorrhagic transformation within 36 hours of a cerebral infarct : relationships with early clinical deterioration and 3-month outcome in the European Cooperative Acute Stroke Study I (ECASS I) cohort. Stroke 1999; 30: 2280–2284.

15.

Trouillas

von Kummer

. Classification and pathogenesis of cerebral hemorrhages after thrombolysis in ischemic stroke. Stroke 2006; 37: 556–561.

16.

Miller

Simpson

Silver

. Safety of thrombolysis in acute ischemic stroke: a review of complications, risk factors, and newer technologies. Neurohospitalist 2011; 1: 138–147.

17.

The National Institute of Neurological Disorders and Stroke rt-PA Stroke Study Group. Tissue plasminogen activator for acute ischemic stroke. N Engl J Med 1995; 333: 1581–1588.

18.

The NINDS t-PA Stroke Study Group. Intracerebral hemorrhage after intravenous t-PA therapy for ischemic stroke. Stroke 1997; 28: 2109–2118.

19.

del Zoppo

Higashida

Furlan

Pessin

Rowley

Gent

. PROACT: a phase II randomized trial of recombinant pro-urokinase by direct arterial delivery in acute middle cerebral artery stroke. Stroke 1998; 29: 4–11.

20.

von Kummer

Broderick

Campbell

BCV

et al.

The Heidelberg bleeding classification. Stroke 2015; 46: 2981–2986.

21.

Hacke

Donnan

Fieschi

et al.

Association of outcome with early stroke treatment: pooled analysis of ATLANTIS, ECASS, and NINDS rt-PA stroke trials. Lancet 2004; 363: 768–774.

22.

C-S

Wang

S-C

Cheng

Y-C

Gau

SS-F

. Association of cerebrovascular events with antidepressant use: a case-crossover study. Am J Psychiatry 2011; 168: 511–521.

23.

Seet

RCS

Rabinstein

. Symptomatic intracranial hemorrhage following intravenous thrombolysis for acute ischemic stroke: a critical review of case definitions. Cerebrovasc Dis 2012; 34: 106–114.

24.

Ried

Jia

Feng

et al.

Selective serotonin reuptake inhibitor treatment and depression are associated with poststroke mortality. Ann Pharmacother 2011; 45: 888–897.

25.

Pan

Sun

Okereke

Rexrode

. Depression and risk of stroke morbidity and mortality. JAMA 2011; 306: 1241.

26.

Aron

Staff

Fortunato

McCullough

. Prestroke living situation and depression contribute to initial stroke severity and stroke recovery. J Stroke Cerebrovasc Dis 2015; 24: 492–499.

27.

Lichtman

Bigger

Blumenthal

et al.

Depression and coronary heart disease: recommendations for screening, referral, and treatment. Circulation 2008; 118: 1768–1775.