Study Setting

The study was conducted at the Neurology Department of Calcutta National Medical College, following approval by the Institutional Ethics Committee.

• Study design—Hospital-based prospective interventional study.

• Study duration—February 2017 to September 2019.

• Inclusion criteria—Consecutive acute ischemic stroke (AIS) patients presenting within 4.5 hours and receiving thrombolysis.

We obtained the written informed consent from the relatives of the patients included in this study. After admission, we recorded a quick history and systemic examination was performed. From the medical history prior to the stroke, we calculated the CHADS2 scores for all patients (CHADS2, congestive heart failure, hypertension, age ≥75 years, diabetes mellitus, stroke [double weight],transient ischemic attack [TIA], or thromboembolism). We analyzed the following variables: age, gender, domestic arrangements (living alone or with family members), ongoing medications, histories of vascular risk factors including hypertension, diabetes mellitus, heart diseases (ischemic heart disease, low ejection fraction [EF], and atrial fibrillation [AF], either as a history of AF or AF diagnosed during the index admission by electrocardiography), TIA, current or former smoking, and hypercholesterolemia. At admission, the stroke severity was ascertained by a neurologist experienced in NIHSS evaluation and was classified as mild (0–4), moderate (5–15), or severe (16–42). According to the Bamford criteria,8 strokes were classified in total anterior circulation stroke (TACS), partial anterior circulation stroke, posterior circulation stroke, and lacunar stroke. Complete hemograms, coagulation profiles, liver, and renal function tests were recorded in the registry at admission and again 24 to 48 hours later. The cardiological profile (electrocardiogram and transthoracic echocardiography) and investigations for vasculitis (antinuclear and antiphospholipid antibodies) were done. The SEDAN and ASPECTS (Alberta Stroke Program Early CT Score) scores were calculated for all patients. The SEDAN score, (0–6 points), which is based on blood sugar, early infarct sign, dense artery sign, age (>75 years), and NIHSS score (>10), had been suggested in European Cooperative Acute Stroke Study to predict the risk of sICH after intravenous thrombolysis.9 The patient’s vital parameters were monitored using multiparameter monitors. The entire drug amount (0.9 mg/kg of alteplase [Actilyse by Boehringer Ingelheim]) was dissolved in the diluent provided by the manufacturer to attain a concentration of 1mg/mL. Ten percent of the total dose was administered as a bolus and rest as an infusion over 60 minutes, using a syringe pump. Acute stroke management and secondary prevention in these patients followed current European Stroke Organization guidelines.10 Repeat brain CT was done 36 hours after admission. Postthrombolysis ICH was defined as any hemorrhage in the brain within 36 hours after thrombolysis which was detected on imaging. ICH was classified as sICH and asymptomatic ICH. sICH was defined as a CT documented hemorrhage associated with clinical deterioration in the form of a decline in NIHSS score of ≥4 points or death. Following discharge, the patients were followed up monthly for 6 months through neurological examination and review of records. The outcome of patients with sICH and patients without sICH was evaluated by measuring intrahospital mortality and mRS at 6 months and classified as favorable (score 0–1) or unfavorable (score 2–6).

Statistical Analysis

We entered the data on SPSS 20.0 software. The baseline characteristics were compared in two groups, using student t-test for continuous variables and Chi-square test for categorical variables. The risk factors for intracranial hemorrhage and poor outcomes were evaluated using univariate analysis followed by multiple logistic regressions with variables having two-tailed p-value < 0.05, to establish the independent risk factors. The total fibrinogen concentration was examined by tertiles of increasing levels to evaluate for possible threshold effects. The first tertile served as the reference value.

Exclusion criteria—patients who were not thrombolysed and those with inadequate follow-up data were excluded.

Discussion on Predictors of an Unfavorable Outcome

In our study, multivariate analysis identified previous TIA, low EF, CHADS2 score≥2, occurrence of TACS, and upper tertiles of age, MABP at admission, CBG at admission, NIHSS score at admission, and fibrinogen level to be independent predictors of mRS score≥2. TACS, low EF, followed by CHADS2 score >2 had the highest odds for unfavorable outcomes. In Cetiner’s study, pretreatment NIHSS scores, presence of proximal hyperintense middle cerebral artery sign in noncontrast CT brain, cardioembolic stroke, and age ≥80 years were significantly higher in the group with unfavorable functional outcomes.12 In the Zhao’s study on the factors associated with thrombolysis outcome in AIS patients with AF, heart failure, higher systolic BP, blood glucose, and NIHSS score independently predicted unfavorable outcomes. Thrombolysis independently predisposed to a favorable outcome but did not significantly reduce mortality.13 Huang et al. described that patients with no history of diabetes, leukoaraiosis, low blood glucose level, low systolic BP level, and low baseline NIHSS score before thrombolysis had better outcome.14 Thus, the findings of the majority of the studies performed worldwide were in unison with our results.

Considering in-hospital mortality alone, multivariate analysis identified CHADS2 score >2, upper tertiles of MABP, CBG, NIHSS, and fibrinogen level as the independent predictor of mortality. In his study, Bateman et al described that the thrombolysed patients had a higher in-hospital mortality rate compared with the nonthrombolysed patients (11.4 vs. 6.8%). His multivariate logistic regression showed that older age, Asian/Pacific Islander race, congestive cardiac failure, and AF/flutter to be independent predictors, whereas, overall ischemic stroke volume, was not a significant predictor of in-hospital mortality after thrombolysis.15 Heuschmann et al reported a 10% mortality following thrombolysis with advanced age and altered level of consciousness significantly influencing in-hospital deaths.16

Discussion on Postthrombolysis Hemorrhage

Eight (13.33%) patients had an sICH in our study, which was higher compared with other Indian studies— 0 (Padma et al),17 6.3% (Sharma and Sharma),18 1.7% (Salam et al),195.5% (Boddu et al),20 and other international studies—4.4% (Bateman et al),15 and 7.1% (Heuschmann et al).16 The rate of fatal postthrombolysis hemorrhage was 6.67% in our study. Whiteley et al meta-analyzed data of 6,756 participants from nine trials of thrombolysis with alteplase versus controls.21 Thrombolysis increased parenchymal hemorrhage, occurring in 6.8% patients allocated to alteplase group versus 1.3% of controls; and of fatal intracerebral hemorrhage 2.7% among thrombolysed subjects versus 0.4% among nonthrombolysed patients. However, the proportional increase in sICH was similar irrespective of age, treatment delay, or baseline NIHSS score, but the absolute excess risk of sICH was elevated with increasing stroke severity ranging from 1.5% (0.8–2.6%) for strokes with NIHSS 0 to 4 to 3.7% (2.1–6.3%) for NIHSS 22 or more (p = 0·0101).21 In our study, multivariate analysis identified CHADS2 score >2, occurrence of TACS, AF, low EF, antiplatelet therapy, anticoagulant therapy, and upper tertiles of NIHSS score, fibrinogen level, MABP, CBG, SEDAN score, and lower tertile of ASPECT score as independent predictors of postthrombolysis hemorrhage. Upper tertile of NIHSS score, low EF, followed by CHADS2 score >2, had the highest odds for hemorrhagic conversion. Chenna et al inferred that increased door to needle time, prior diabetes and hypertension, and higher NIHSS score and blood sugar at presentation increased the occurrence of sICH.6 The Canadian Alteplase for Stroke Effectiveness Study results showed that the door-to-needle time was an independent predictor of sICH and there was a nonsignificantly higher trend toward the occurrence of sICH in the 3 to 4.5 hours group compared with 0 to 3 hours group (7.8 vs. 3.8%, p = 0.06).22 However, in our study, the occurrence of postthrombolysis hemorrhage was unrelated to the exact time of thrombolysis within the 4.5-hour window period. Cardioembolism, NIHSS on admission > 20, serum glucose on admission>9.0 mmol/L, and age ≧ 70 years were independent risk factors for sICH in Chinese patients with AIS treated with alteplase, with cardioembolism carrying the highest odds.5 Modrego described earlier microbleeds on brain magnetic resonance imaging (MRI), leukoaraiosis, and previous antiplatelet drug use or statin use, uncontrolled hypertension, higher plasma glucose, lower platelet count, surgery within the preceding 3 months, NIHSS>25, early signs of infarction on brain CT, ASPECT ≤ 7, stroke symptoms duration≥4.5 hours, symptoms of unknown onset time, an apparent diffusion coefficient (ADC) ratio (ADC in the affected area/ADC in the normal contralateral area) <0.65, FLAIR (fluid-attenuated inversion recovery) hyperintensity, and focal hyperperfusion in the damaged area after reperfusionon arterial spin labeling in a ratio higher than 1.5 (ipsilateral/contralateral) as independent predictors of sICH.23 The increased rate of sICH in our study was possibly because of the fact that among patients who were thrombolysed on the basis of early ischemic changes in brain CT scans, 11 turned out to be TACS on subsequent postthrombolysis imaging and four (36.36%) patients among them developed hemorrhage. Also, we thrombolysed eight patients on warfarin, among whom four (50%) developed hemorrhage.

Discussion on Fibrinogen As a Biomarker

The “early fibrinogen degradation coagulopathy” hypothesis described a biological phenomenon predictive of postthrombolysis hemorrhage related to an early loss of fibrinogen and consequently, leading to the increase of fibrinogen degradation products (FDPs). Studies showed that a decrement in 2 hours postthrombolysis fibrinogen levels to less than 2 g/L increased the odds of an early sICH by 12.82 times.24 Our study also demonstrated that the serum fibrinogen levels fell significantly after thrombolysis, the mean fibrinogen decreasing by 7% (p < 0.0001) and we recorded an sICH rate of 13.33%. The Alteplase–Tenecteplase Trial Evaluation for Stroke Thrombolysis study also demonstrated that alteplase was associated with prolongation of prothrombin time, reduction of plasminogen and fibrinogen levels, and elevation of FDP and D-dimer levels.25 In their study on whether prethrombolytic plasmatic markers of coagulation and fibrinolysis are correlated to recanalization and functional outcome, Martí-Fàbregas found no relation between fibrinogen level and functional outcome.26 However, our study demonstrated that the odds of poor outcome and sICH increased progressively with increasing levels of fibrinogen. AIS predisposes to mitochondrial dysfunction and increased nitric oxide (NO) production which dilates blood vessels, thereby maintaining perfusion. NO stimulates free radical production and generates peroxynitrite, which irreversibly nitrates proteins. During AIS, fibrinogen is upregulated as an acutephase reactant. AIS also increases the permeability of the blood–brain barrier, thereby, allowing the extravasation of plasma proteins into the brain parenchyma. This potentiates fibrinogen nitrotyrosination. Initially, nitro-fibrinogen retards clot formation, but later it produces harmful fibrinolysis-resistant clots and induces brain parenchymal damage.27 Immunofluorescence analysis of the affected human brain areas also showed that both fibrinogen and nitrotyrosinated fibrinogen spread into the brain parenchyma after AIS. Measurement of caspase-3 enzyme activity in the human neuroblastoma cell line assayed the toxicity of fibrinogen and nitrotyrosinated fibrinogen. Nitrotyrosinated fibrinogen induced higher activation of caspase 3, which is integral in the cellular apoptosis process. Cell survival assays also demonstrated a greater neurotoxic effect of nitrotyrosinated fibrinogen at all tested concentrations. These studies suggest that fibrinogen, and especially nitrotyrosinated fibrinogen, plays a prominent role in the progression of brain tissue damage in the penumbral region, thereby, determining the final extent of the ischemic insult.28 29 Hence, higher elevation in the fibrinogen levels produces greater neuronal and local vascular damage, resulting in a higher chance of postthrombolysis sICH. Bagoly et al’s review article demonstrated that the assessment of pre- and postthrombolysis plasma fibrinogen levels was potentially useful to predict postthrombolysis sICH but more data are needed to establish its association with unfavorable outcome.24

The study was limited by small sample size, suffers from the probability of selection bias for not including a control group consisting of patients not receiving thrombolytic treatment, our inability to estimate FDP levels in all patients, lack of a long-term follow-up, and lack of MRI-based protocols prior to thrombolysis, thereby, hindering our ability to test leukoaraiosis, microbleeds, and low ADC coefficient as predictors of sICH. Therefore, the results should be confirmed in multicenter cohort studies.

Hence, this study made the following conclusions:

1. Independent predictors of unfavorable outcome after thrombolysis were past TIA, CHADS2 score >2, occurrence of TACS, low EF, higher age, MABP, CBG, and NIHSS score.

2. Independent predictors of sICH were CHADS2 score > 2, occurrence of TACS, AF, low EF, ongoing antiplatelet and anticoagulant therapy, higher MABP, CBG, SEDAN, and NIHSS scores and lower ASPECTS score.

Thus, CHADS2 score >2, low EF, occurrence of TACS and higher NIHSS score, MABP, and CBG at admission were the common risk factors significantly predicting postthrombolysis sICH and morbidity.

3. Higher fibrinogen levels were significantly related with higher rates of sICH, morbidity, and mortality. After thrombolysis, the level of fibrinogen significantly reduced. Thus, it is a potential biomarker.

4. The performance of thrombolysis within 3 hours or between 3 and 4.5 hours after the onset of symptoms did not affect morbidity, mortality, or the occurrence of sICH.