Fibrinogen Decrease after Intravenous Thrombolysis in Ischemic Stroke Patients Is a Risk Factor for Intracerebral Hemorrhage Laura Vandelli, MD,* Marco Marietta, MD,† Mariaelena Gambini, PhD,‡ Milena Cavazzuti, MD,* Tommaso Trenti, MD,x Maria A. Cenci, MD,x Federica Casoni, MD,* Guido Bigliardi, MD,* Roberta Pentore, MD,* Paolo Nichelli, MD, PHD,* and Andrea Zini, MD*
Background: Intravenous thrombolysis is an effective treatment in acute stroke patients, but it increases the risk of intracerebral hemorrhages. Our aim is to establish if fibrinogen depletion increases the risk of intracerebral hemorrhage after intravenous thrombolysis for acute ischemic stroke. Methods: In 104 ischemic stroke patients, treated with intravenous thrombolysis, we assessed the rate of intracerebral hemorrhages documented by computed tomographic scan at 24 hours and within 7 days post-treatment. Fibrinogen levels were determined at 2 hours after therapy: patients were classified as belonging to ‘‘low fibrinogen group’’ if levels decreased to less than 2 g/L and/or by 25% or more. Fibrinogen levels and other known hemorrhagic risk factors were studied using univariate and multivariate analyses. Results: During the first 7 days, an intracerebral hemorrhage was detected in 24 patients (23.1%), and only 6 of these (5.8%) experienced symptomatic bleeding; 41 patients were included in the low fibrinogen group. Among the 24 hemorrhages, 18 occurred in the low fibrinogen group and 6 in the ‘‘normal fibrinogen group’’: the bleeding rate in the low fibrinogen group was significantly higher (43.9%) than that in the normal fibrinogen group (9.5%; odds ratio [OR] 7.43, P ,.001). Univariate and multivariate analyses revealed that only clinical severity (OR 1.15, P , .001) and hypofibrinogenemia (OR 7.47, P , .001) were significantly associated with brain bleeding at 7 days and at 24 hours (P 5.008). Conclusions: An early fibrinogen reduction seems to increase the risk of intracerebral hemorrhage after rtPA treatment in ischemic stroke. Fibrinogen assessment could be a rapid, inexpensive, and widely available tool to help the identification of patients at higher risk of bleeding. Key Words: Fibrinogen—acute stroke—intracerebral hemorrhage—risk factors—thrombolysis—rtPA. Ó 2015 by National Stroke Association
From the *Stroke Unit, Department of Neuroscience, University of Modena and Reggio Emilia, Nuovo Ospedale Civile S. AgostinoEstense, AUSL Modena, Modena; †Section of Haematology, Department of Oncology and Haematology, Policlinico of Modena, University of Modena and Reggio Emilia, Modena; ‡Environmental Protection Agency, ARPA Emilia Romagna District of Reggio Emilia, Reggio Emilia; and xClinical Pathology-Toxicology, Nuovo Ospedale Civile S. Agostino-Estense, AUSL Modena, Modena, Italy. Received March 21, 2014; revision received September 2, 2014; accepted September 6, 2014.
394
Disclosure: The authors report no conflicts of interest. Address correspondence to Andrea Zini, MD, Neurology Clinic, Stroke Unit, Department of Neuroscience, University of Modena and Reggio Emilia, Nuovo Ospedale Civile S.Agostino-Estense, AUSL Modena, via Giardini, 1355 Baggiovara, 41100 Modena, Italy. E-mail: [email protected]. 1052-3057/$ - see front matter Ó 2015 by National Stroke Association http://dx.doi.org/10.1016/j.jstrokecerebrovasdis.2014.09.005
Journal of Stroke and Cerebrovascular Diseases, Vol. 24, No. 2 (February), 2015: pp 394-400
FIBRINOGEN DECREASE AFTER INTRAVENOUS THROMBOLYSIS IN STROKE PATIENTS
Introduction Thrombolytic therapy with administration of intravenous recombinant tissue plasminogen activator (iv rtPA) has been shown to improve long-term functional outcome, and it is recommended for the treatment of eligible acute ischemic stroke patients.1 This therapy, however, is associated with an increased risk of symptomatic intracerebral hemorrhages (sICH). Symptomatic intracranial bleeding in stroke patients treated with iv rtPA is approximately 6%-8%2-4 and is associated with a worse clinical outcome. It is, however, not known what proportion of a worse outcome is attributable to sICH as there is an overlap between the risk factors for thrombolysis-associated sICH and those for poor outcome after thrombolytic therapy with no sICH.2 Many studies have evaluated sICH risk factors in patients receiving thrombolytic therapy, and 2 recent systematic reviews identified the most relevant ones: hyperglycemia, early ischemic changes on computed tomography (CT)/magnetic resonance imaging scan, clinical stroke severity assessed by the National Institutes of Health Stroke Scale (NIHSS) score on admission, advanced age, and high blood pressure.2,3,5-10 The pathophysiological mechanisms of hemorrhagic cerebral transformations could result from the reperfusion of intracranial arteries whose integrity has been disrupted by cerebral ischemia, with an alteration in the blood–brain barrier and an increase in capillary permeability.3 Within this complex pathophysiological process, a central role is played by fibrinogen. rtPA binds to plasminogen within the clot, converting it to plasmin that is a proteolytic enzyme capable of breaking cross-links between fibrin molecules and so dissolving clots, releasing fibrin(ogen) degradation products (FDPs). It is important to note, however, that plasmin is fairly nonspecific in its activity and, besides fibrin, will also break down other circulating proteins, including fibrinogen.11,12 In short, although rtPA is relatively selective for clot-associated fibrin, it can produce a systemic fibrinolytic state (with a secondary hypofibrinogenemia and a D-dimer increase12) and bleeding complications, which were first reported in the setting of the treatment of acute myocardial infarction (AMI). The Thrombolysis in Myocardial Infarction trial13 showed that both streptokinase and rtPA caused a decrease in fibrinogen levels and an increase in FDP and that the rate of hemorrhagic events was higher in patients with increased FDP in both treatment groups and in patients with reduced fibrinogen levels in the rtPA group. Moreover, there were more hemorrhages in patients with greater plasma changes, underlining the importance of coagulation parameters.13,14 Collen et al15 analyzed coagulation and fibrinolysis parameters during intravenous rtPA infusions in patients with AMI and showed that the extent of fibrinogen breakdown is occasionally very
395
important, with a decrease of fibrinogen level less than 1.0 g/L observed in 27% of rtPA-treated patients. Experience and information on the subject of ischemic stroke are poor and sparse. Only 1 recent work by Matosevic et al16 evaluated the extent of fibrinogen depletion in rtPA thrombolysis in stroke patients and its association with intracranial and extracranial bleeding: they showed that a fibrinogen decrease is a significant predictor for bleeding risk and that a temporal relationship exists with the manifestation of hemorrhage. For these reasons, we decided to study early modifications in fibrinogen levels pre- and post-thrombolysis, to assess if they are significantly involved in iv rtPArelated ICH in ischemic stroke patients.
Materials and Methods Study Population We studied consecutive patients with acute ischemic stroke treated with iv infusion of Alteplase (Actilyse, a tPA produced by recombinant DNA technology) from January 1, 2010, to January 1, 2011, in our stroke unit. Patient’s selection was made according to the Safe Implementation of Thrombolysis in Stroke-Monitoring Study criteria and the European Cooperative Acute Stroke Study-3 time window (,4.5 hours). A few of treated patients were included in randomized control trials, such as International Stroke Trial-3, Synthesis Expansion, and Thrombolysis in Elderly Stroke Patients in Italy trial, an Italian randomised controlled trial in patients older than 80 years. The following data were collected on each patient: age, gender, clinical stroke severity assessed by NIHSS preand post-thrombolysis, fibrinogen level at admission and 2 hours after the end of rtPA infusion, baseline glycemia, baseline platelet count, and blood pressure levels.
Plasmatic Fibrinogen Dosage Peripheral blood samples were drawn from each patient, on entry to the emergency room and 2 hours after the end of fibrinolytic infusion. Fibrinogen levels were determined on fresh plasma obtained from blood collected in 5-mL vacuum tubes containing .5 mL sodium citrate (.129 mol/L). The assay was performed using the HemosIL Fibrinogen-C kit assay based on the Clauss method on ACL TOP Coagulation Systems (Instrument Laboratory, Milan, Italy). The Fibrinogen-C kit uses an excess of thrombin to convert fibrinogen to fibrin in diluted plasma. At high thrombin and low fibrinogen concentrations, the rate of reaction is a linear function of the fibrinogen concentration.
Neuroimaging On admission, all patients underwent a noncontrast brain CT scan, which was repeated about 24 hours and
L. VANDELLI ET AL.
396
7 days post-treatment; an additional control scan was conducted whenever neurologic worsening was observed in the first 7 days. We considered every intracranial bleeding that occurred within the first 7 days post-thrombolysis, including both symptomatic and asymptomatic hemorrhages, classifying them according to the Safe Implementation of Thrombolysis in Stroke radiological criteria. Early hemorrhages (early ICH) were those occurring within 24 hours post-thrombolytic treatment. Hemorrhagic transformations were categorized as sICH if patients had neurologic deterioration with an NIHSS increase of more than 1 point and if the CT scan revealed blood at any site in the brain irrespective of its extension (National Institute of Neurological Disorders and Stroke classification). All neuroimaging studies were reviewed by 2 independent, certified, expert neurologists blinded to fibrinogen-level assessment.
Statistical Analysis We assessed fibrinogen-level changes after thrombolytic therapy. A threshold of fibrinogen plasma value critical for the development of bleeding is not reported in the literature. Some studies13,15 have reported a correlation between bleeding and fibrinogen values less than 1.5 or 1.0 g/L or less than 1.0 and .5 g/L. A more recent study, which evaluated the role of fibrinogen concentrate substitution therapy in patients with massive hemorrhage and low plasma fibrinogen concentrations, considered 2 g/L to be the threshold value.17 On these grounds, we performed a logistic regression analysis that analyzed different levels of both absolute fibrinogen plasma values and percentage decrease. We analyzed the following absolute fibrinogen plasma levels as independent variables: lower than 3, 2.5, 2, and 1. We then tested the percentage decrease of fibrinogen values calculated as the percentage of the proportion between the difference of fibrinogen values after and before treatment, divided by the fibrinogen value before treatment: % decrease5½ðfibrinogen value after treatment2 fibrinogen value before treatmentÞ= fibrinogen value before treatment%: For the analysis, we considered only negative values that indicate a decrease in the fibrinogen values after the treatment, and we chose to analyze decreases exceeding 45%, 40%, 35%, 30%, 25%, 20%, 15%, and 10%. After these analyses, we set the cutoff values as fibrinogen absolute levels after treatment below 2 g/L and/or 25% decrease or more, between pre- and post-treatment. This choice derives from the comparison of the logistic regression results between the infarction (dependent variable) and the different absolute values of fibrinogen and between the infarction (dependent variable) and the different values of percentage decrease in fibrinogen: in
both cases, we chose the values associated with the result showing a greater statistical significance (lower P-value). A Pubmed search was conducted looking for sICH risk factors after thrombolytic therapy, focusing primarily on systematic reviews.2,3,5-10 From these data, the following variables were indicated as being most closely related to hemorrhagic transformation: hyperglycemia, clinical stroke severity assessed by baseline NIHSS, age, sex, and high blood pressure. A logistic regression analysis was then conducted on each single parameter. The study population was splitted into 2 groups: a ‘‘low fibrinogen group’’ and a ‘‘normal fibrinogen group’’, to compare important statistical data (eg, mean age, mean NIHSS, mean fibrinogen level pre- and postthrombolysis) and to evaluate any significant differences between the groups. The low fibrinogen group included all patients with post-thrombolysis fibrinogen levels less than 2 g/L and/or a 25% or more decrease of fibrinogen levels respect to baseline, whereas all other patients were included into the normal fibrinogen group. Univariate and multivariate logistic regression analyses were performed to evaluate if fibrinogen levels and other parameters (age, sex, NIHSS, baseline glycemia, baseline blood pressure) were independent risk factors for ICH. We also analyzed the occurrence of ICH and sICH (within 24 hours and 7 days), comparing low and normal fibrinogen groups using contingency tables with Pearson chi-square, to verify a possible association between fibrinogen levels and early onset of ICH or sICH.
Results A total of 104 patients were included in this study, 66 (63.5%) men and 38 (36.5%) women. The mean age was 66.6 years (standard deviation [SD] 13.5; range 21-91) and the mean baseline NIHSS was 10.9 (SD 7.2; range 2-34). On the seventh day follow-up CT scan, 24 patients (23.1%) presented ICH (3 HI1, 8 HI2, 2 PH1, 7 PH2, and 4 PHr1), 14 of which developed in the first 24 hours (3 HI1, 3 HI2, 1 PH1, 4 PH2, and 3 PHr1). Only 6 patients (5.8%) experienced sICH according to the NINDS classification. Mean fibrinogen pretreatment level was 3.34 g/L (SD 1.1) with a significant decrease after thrombolysis (2.47 g/L, 226%, SD .8; P , .001 respect to baseline). Logistic analysis found a significant association between baseline fibrinogen value and ICH (P 5 .048) that became more significant when comparing ICH and fibrinogen post-thrombolysis level (P 5 .029). A logistic regression analysis of the most well-known ICH risk factors (age, sex, baseline NIHSS, 24-hour NIHSS, pre-rtPA arterial blood pressure, pre-rtPA glycemia; baseline platelet count) showed that only baseline NIHSS (P , .001) and 24-hour NIHSS (P , .001) were significantly associated with ICH (Table 1).
FIBRINOGEN DECREASE AFTER INTRAVENOUS THROMBOLYSIS IN STROKE PATIENTS
397
Table 1. Univariate logistic regression analysis: risk factors for ICH post-thrombolysis Variable Fibrinogen level ,2.0 g/L and/or .25% decrease Fibrinogen level ,2.0 g/L Fibrinogen level decreased .25% Platelet count Sex Age Glycemia Blood systolic pressure Baseline NIHSS 24-h NIHSS
OR
P
95% CI
7.43 ,.001 2.620-21.100 2.83 .051 .994-8.075 8.36 ,.001 3.002-23.309 1.00 .485 .996-1.008 .83 .710 .318-2.180 1.19 .138 .945-1.499 1.01 .054 1.000-1.019 1.01 .569 .983-1.031 1.16 ,.001 1.082-1.253 1.11 ,.001 1.051-1.168
Abbreviations: ICH, intracerebral hemorrhage; NIHSS, National Institutes of Health Stroke Scale. Bold values indicate P , .05.
A subsequent logistic regression analysis comparing different post-rtPA fibrinogen cutoff values (,1.0, ,2.0, ,2.5, and ,3.0 g/L) showed that a cutoff value less than 2 g/L offered the best predictive risk factor for ICH, although it did not reach a statistical significance (P 5.051, odds ratio [OR] 5 2.83). The regression analysis, after a class-based analysis with different rates of fibrinogen reduction (45%, 40%, 35%, 30%, 25%, 20%, 15%, and 10%), showed that a higher ICH risk lies in a 25% or more decrease in fibrinogen levels. Based on the results of regression analysis, the patient cohort was divided in the 2 main groups: low fibrinogen group and normal fibrinogen group as described earlier. The low fibrinogen group was composed of 41 patients (39.4%); 6 of them (5.8% of the whole population) had post-rtPA fibrinogen levels less than 2 g/L, 21 (20.2% of the whole population) had a decrease of 25% or more respect to baseline and 14 patients had a reduction in both values (13.5%). Among the 24 patients who presented ICH, 18 were in the low fibrinogen group (Figs 1 and 2) and 6 were in the normal fibrinogen group. The ICH rate during the first 7 days in the low fibrinogen group was significantly higher (43.9%) than that (9.5%) in the normal fibrinogen group (OR 7.43, P , .001; Table 1). The difference remained statistically significant when the analysis was restricted to the patients showing only a fibrinogen decrease greater than 25% (P , .001; OR 8.36; Table 1). Patient characteristics included in the low fibrinogen and normal fibrinogen groups are summarized in Table 2, showing the absence of significant differences except for baseline NIHSS scores and baseline fibrinogen levels. A further subanalysis divided patients into mild (NIHSS score 0-7), moderate (NIHSS score 8-15), and severe (NIHSS score .15) clinical severity groups: similar
Figure 1. A noncontrast brain computed tomographic scan of a patient belonging to low fibrinogen group that experienced post-thrombolysis parenchymal hematoma in the contest of a left frontotemporal ischemic lesion (PH2 type according to the Safe Implementation of Thrombolysis in Stroke radiological criteria), with associated intraventricular hemorrhage.
NIHSS values were found in each group, in particular in the ‘‘severe’’ one (Table 3). In conclusion, the univariate analysis showed that the only variables significantly associated with ICH were baseline NIHSS, 24-hour NIHSS, and belonging to the
Figure 2. A noncontrast brain computed tomographic scan of another low fibrinogen group patient that presented multiple parenchymal hematomas after intravenous recombinant tissue plasminogen activator infusion: the largest one in the contest of a left parieto-occipital ischemic lesion (PH2 type) and some other minor contralateral parenchymal hematomas, remote from the ischemia (PHr1 type).
L. VANDELLI ET AL.
398
Table 2. Comparison of patient characteristics between normal and low fibrinogen groups Low fibrinogen Normal group, N group, N
Variable Age (mean) Sex (male %) Baseline NIHSS (mean) 24-h NIHSS (mean) Glycemia (mean) Blood pressure (systolic) Blood pressure (diastolic) Platelet count Fibrinogen-level prethrombolysis (mean)
66.71 65.9 12.8 9.34 135.78 148.63 78.54 207.73 362.21
66.46 61.9 9.71 7.27 133.86 146.78 78.71 192.11 311.72
P NS NS .031* NS NS NS NS NS .028*
Abbreviation: NIHSS, National Institutes of Health Stroke Scale. Bold values indicate P , .05. *P , .05.
low fibrinogen group (fibrinogen levels post-thrombolysis ,2 g/L and/or a $25% decrease). This finding was confirmed by the multivariate analysis (Table 4). Considering early ICH, 14 events were found, 10 in the low fibrinogen group (71.4% of early ICH and 9.62% of total ICH) and 4 in the normal fibrinogen group (28.6% of early ICH and 3.85% of total ICH), showing that hemorrhagic risk remains significantly higher in the low fibrinogen group when analyzing only ICH less than 24-hour subgroup (P 5 .008; Table 5). Of the 6 sICH reported, 5 (83.3%) occurred during the first 24 hours post-thrombolysis. The logistic regression analysis for sICH and early sICH did not reveal statistical significance between the normal and low fibrinogen groups (P . .05; Table 5).
Discussion The development of ICH is the most serious adverse event related to iv rtPA as it carries a relevant morbidity
Table 3. Subgroup analysis of baseline clinical severity between low and normal fibrinogen groups
Clinical severity Mild (NIHSS score 0-7) Moderate (NIHSS score 8-15) Severe (NIHSS sore .15)
Fibrinogen group Mean Normal Low Normal Low Normal Low
4.83 4.31 10.48 11.36 20.58 22.14
P*
n
SD
30 13 21 14 12 14
1.46 1.75 .319 2.58 2.50 .324 3.34 4.82 .355
Abbreviation: NIHSS, National Institutes of Health Stroke Scale. *Significance P , .05.
Table 4. Multivariate logistic regression analysis of risk factors for ICH post-thrombolysis Variable
OR
P
95% CI
Fibrinogen level ,2.0 g/L and/or .25% decrease Baseline NIHSS Glycemia Blood systolic pressure
7.47
,.001
2.26-24.74
1.15 1.01 1.00
,.001 .131 .802
1.06-1.25 1.00-1.02 .97-1.03
Abbreviations: ICH, intracerebral hemorrhage; NIHSS, National Institutes of Health Stroke Scale. Bold values indicate P , .05.
and mortality.4 In our study, we found an overall ICH rate within the first 7 days of 23.1%, which is lower than that reported from previous studies.18 Similarly, we found 6 patients with sICH (5.8%), a rate comparable with those reported in the literature (6%-8%).2-4 Many studies in the literature have reported various risk factors for ICH, the most relevant of which are age, clinical severity, pre-rtPA arterial blood pressure, and pre-rtPA glycemia.2,3 In our study, we found that the only already known variables significantly associated with the development of ICH are baseline NIHSS (OR 1.16, P , .001) and 24-hour NIHSS (OR 1.11, P , .001) (Table 1) because neither baseline blood glucose nor systolic blood pressure pretreatment attained the statistical significance. On the other hand, we demonstrated by logistic regression analysis that a decrease in post-thrombolysis fibrinogen levels of less than 2 g/L and/or of 25% or more (low fibrinogen group) was statistically significantly related with post-rtPA ICH. Indeed, in this patient group, the ICH rate within 7 days was 43.9%, significantly higher than the 9.5% of the normal fibrinogen group (OR 7.4, P , .001) (Table 1). When the 2 subgroups of patients included in the low fibrinogen group were separately analyzed, however, the results were
Table 5. Logistic regression analysis of fibrinogen decrease as risk factor for ICH, early ICH, sICH, and early sICH Low Normal Total Subgroups of fibrinogen fibrinogen hemorrhage (n 5 63) (n 5 41) (n 5 104) ICH Early ICH sICH Early sICH
6 (9.5%) 4 (28.6%) 2 (33.3%) 2 (40%)
18 (43.9%) 10 (71.4%) 4 (66.7%) 3 (60%)
24 14 6 5
P ,.001* .008* .586 .480
Abbreviations: ICH, intracerebral hemorrhage; sICH, symptomatic intracerebral hemorrhage. Bold values indicate P , .05. *P , .05.
FIBRINOGEN DECREASE AFTER INTRAVENOUS THROMBOLYSIS IN STROKE PATIENTS
somewhat different. Indeed, considering only patients with fibrinogen decreases of 25% or more, the difference remains statistically significant (OR 8.36, P , .001), whereas those with fibrinogen levels less than 2 g/L did not, although the results were at limit of significance (OR 2.83, P 5 .051). Our data are in line with results recently published by Matosevic et al,16 which showed that a 6-hour reduction in fibrinogen levels greater than 2 g/L after thrombolysis can—as with stroke severity, diabetes, and subtherapeutic Vitamin K Antagonists pretreatment—heighten the risk of sICH and systemic bleeding. Moreover, as also reported in the Matosevic study, we found that a higher baseline fibrinogen level seems to enhance the risk of a more significant decrease post-treatment. This suggests a correlation between extent of fibrinogen depletion (delta pre- and post-rtPA) and baseline fibrinogen values that could be explained by a ‘‘consumption effect,’’ that is, a more active early fibrinogen degradation coagulopathy. These findings support the hypothesis that postthrombolysis intracranial bleedings are, at least partially, linked to an early rtPA-related coagulopathy that causes a fibrinogen breakdown because of rtPA’s relative selectivity for clot-associated fibrin.11,12 This type of coagulopathy, with an early increase in FDP and fibrinogen consumption, was first described in iv thrombolytic therapy used for AMI.13-15 Some years later, Trouillas19 studied the possible correlation between acquired hypofibrinogenemia and sICH after thrombolysis in stroke patients and found that an increase in FDP, like D-dimer, 2 hours post-thrombolysis was statistically significant and appeared to be a predictive factor for early parenchymal hematomas (,24 hours). Nevertheless, we consider FDP a less specific parameter than fibrinogen of such coagulopathy: in fact, D-dimer, which is the most frequently measured FDP, always increases after a cerebral ischemic event, even without thrombolysis12 as it is a by-product of clot degradation and remodeling that constantly occurs during vessel occlusion. Other factors have been implicated in the genesis of rtPA-related sICH, such as the pretreatment levels of fibrinolysis inhibitors (ie, plasminogen activator inhibitor-17,8 or matrix metalloproteinases-920), but studies gave conflicting results. Moreover, the methods for determining plasma levels of these biomarkers are both very expensive and too slow for use in emergencies, and therefore, they cannot be used to stratify sICH risk in rtPA therapy decision making. On the contrary, determining fibrinogen levels is very quick, easy, and cheap and can be performed in any hospital laboratory. The decrease in fibrinogen was demonstrated to be a significant risk factor for bleeding, even when the analysis was restricted only to early (ie, within 24 hours
399
from treatment) ICH (P 5 .008; Table 5). This aspect is very important as some authors have suggested that only early ICH can regarded as linked to thrombolysis because of the rtPA short plasma half-life (around 5 minutes). Matosevic et al16 found that most of bleeding occurred between 2 and 24 hours post-thrombolysis, an interval that broadly overlaps with the period in which fibrinogen degradation coagulopathy develops. For this reason, they evaluated the fibrinogen decrease 6 hours post-rtPA therapy and its correlation with hemorrhages. On the other hand, other studies on post-fibrinolysis hemostatic changes in myocardial infarction patients have shown that fibrinogen reaches its lowest point some time between 90 minutes and 3 hours.21,22 On these grounds, and on the pharmacologic basis of the short rtPA halflife, we decided to restrict the analysis of fibrinogen levels to 2 hours post-thrombolysis, assuming a very early fibrinogen-related coagulopathy. In our study, the analysis of sICH, either early or total, failed to show a statistically significant correlation with fibrinogen levels (Table 5), but this finding could be explained by the very small number of patients with sICH, making it difficult to reach reliable conclusions. One limitation of our study was the significant difference in baseline NIHSS score that existed between the normal and low fibrinogen groups. This factor could diminish the relevance of our results as it makes it less clear whether the greater bleeding rate in the low fibrinogen group was attributable to the fibrinogen decrease itself or to clinical severity. As a final point, the limited number of included patients did not allow a reliable stratification of study population in several subgroups of patients, such as those with sICH and early sICH. In conclusion, an early decrease in fibrinogen levels seems to represent a risk factor for ICH within the first 7 days after rtPA therapy and even within the first 24 hours in ischemic stroke patients. This result confirms the hypothesis that iv rtPA causes a related coagulopathy because of relative selectivity of Alteplase for clotassociated fibrin, which results in fibrinogen breakdown. We, therefore, recommend assessing fibrinogen levels before and 2 hours postinfusion of rtPA in all thrombolyzed patients: this would allow the identification of patients at higher bleeding risk. The rapidity, cheapness, and wide availability of such a test further reinforce the utility value of this parameter. Further RCT studies could assess the preventive role of therapies like fibrinogen intravenous replacement or plasma transfusion in the low fibrinogen group in avoiding post-thrombolysis ICH. Acknowledgment: The authors thank all neurologists and nurses working in Neurology Clinic at Nuovo Ospedale Civile S.Agostino-Estense.
L. VANDELLI ET AL.
400
References 1. Adams HP, Del Zoppo G, Alberts MJ, et al. A guideline from the American Heart Association/American Stroke Association Stroke Council, Clinical Cardiology Council, Cardiovascular Radiology and Intervention Council, and the Atherosclerotic Peripheral Vascular Disease and Quality of Care Outcomes in Research Interdisciplinary Working Groups. Stroke 2007;38:1655-1711. 2. Lansberg MG, Albers GW, Wijman CA. Symptomatic intracerebral hemorrhage following thrombolytic therapy for acute ischemic stroke: a review of the risk factors. Cerebrovasc Dis 2007;24:1-10. 3. Derex L, Nighoghossian N. Intracerebral haemorrhage after thrombolysis for acute ischaemic stroke: an update. J Neurol Neurosurg Psychiatry 2008;79:1093-1099. 4. Seet RC, Rabinstein AA. Symptomatic intracranial hemorrhage following intravenous thrombolysis for acute ischemic stroke: a critical review of case definitions. Cerebrovasc Dis 2012;34:106-114. 5. Trouillas P, Von Kummer R. Classification and pathogenesis of cerebral hemorrhages after thrombolysis in ischemic stroke. Stroke 2006;37:556-561. 6. Derex L, Hermier M, Adeleine P, et al. Clinical and imaging predictors of intracerebral hemorrhage in stroke patients treated with intravenous tissue plasminogen activator. J Neurol Neurosurg Psychiatry 2005;76:70-75. 7. Ribo M, Montaner J, Molina CA, et al. Admission fibrinolytic profile is associated with symptomatic hemorrhagic transformation in stroke patients treated with tissue plasminogen activator. Stroke 2004;35:2123-2127. 8. Cocho D, Borrell M, Marti-Fabregas J, et al. Pretreatment hemostatic markers of symptomatic intracerebral hemorrhage in patients treated with tissue plasminogen activator. Stroke 2006;37:996-999. 9. Kellert L, Sykora M, Gumbinger C, et al. Blood pressure variability after intravenous thrombolysis in acute stroke does not predict intracerebral hemorrhage but poor outcome. Cerebrovasc Dis 2012;33:135-140. 10. K€ ohrmann M, Schellinger PD. Symptomatic intracranial hemorrhage after thrombolysis. Cerebrovasc Dis 2007; 23:83-84. 11. Klabunde RE. Cardiovascular Physiology Concepts. 2nd ed. Lippincott Williams & Wilkins/Baltimore 2011. 12. Francis CW, Kornberg A. Fibrinogen- and fibrindegradation products during fibrinolytic therapy. Ann N Y Acad Sci 1992;667:310-323. 13. Rao AK, Pratt C, Breke A, et al. Thrombolysis in Myocardial Infarction (TIMI) trial-phase I. Hemorrhagic mani-
14.
15.
16.
17.
18.
19.
20.
21.
22.
festations and changes in plasma fibrinogen and the fibrinolytic system in patients treated with recombinant tissue plasminogen activator and streptokinase. J Am Coll Cardiol 1988;11:1-11. Bovill EG, Tracy RP, Knatterud GL, et al. Hemorrhagic events during therapy with recombinant tissue plasminogen activator, heparin, and aspirin for unstable angina (Thrombolysis In Myocardial Ischemia, phase IIIB trial. Am J Cardiol 1997;79:391-396. Collen D, Bounameaux H, De Cock F, et al. Analysis of coagulation and fibrinolysis during intravenous infusion of recombinant human tissue-type plasminogen activator in patients with acute myocardial infarction. Circulation 1986;73:511-517. Matosevic B, Knoflach M, Werner P, et al. Fibrinogen degradation coagulopathy and bleeding complications after stroke thrombolysis. Neurology 2013;80: 1216-1224. Fenger-Eriksen C, Jensen TM, Kristensen BS, et al. Fibrinogen substitution improves whole blood clot firmness after dilution with hydroxyethyl starch in bleeding patients undergoing radical cystectomy: a randomized, placebo-controlled clinical trial. J Thromb Haemost 2009;7:795-802. Dzialowski I, Warwick Pexman JH, Barber PA, et al. Asymptomatic hemorrhage after thrombolysis may not be benign prognosis by hemorrhage type in the Canadian Alteplase for Stroke Effectiveness Study Registry. Stroke 2007;38:75-79. Trouillas P, Derex L, Philippeau F, et al. Early fibrinogen degradation coagulopathy is predictive of parenchymal hematomas in cerebral rtpa thrombolysis. A study in 157 cases. Stroke 2004;35:1323-1328. Montaner J, Molina CA, Monasterio J, et al. Matrix metalloproteinase-9 pretreatment level predicts intracranial hemorrhagic complications after thrombolysis in human stroke. Circulation 2003;107:598-603. Granger CB, Becker R, Tracy RP, et al. Thrombin generation, inhibition and clinical outcomes in patients with acute myocardial infarction treated with thrombolytic therapy and heparin: results from the GUSTO-I Trial. GUSTO-I Hemostasis Substudy Group. Global Utilization of Streptokinase and TPA for Occluded Coronary Arteries. J Am Coll Cardiol 1998;31:497-505. Stangl K, Laule M, Tenckhoff B, et al. Fibrinogen breakdown, long-lasting systemic fibrinolysis, and procoagulant activation during alteplase double-bolus regimen in acute myocardial infarction. Am J Cardiol 1998; 81:841-847.
Background: Intravenous thrombolysis is an effective treatment in acute stroke patients, but it increases the risk of intracerebral hemorrhages. Our aim is to establish if fibrinogen depletion increases the risk of intracerebral hemorrhage after intravenous thrombolysis for acute ischemic stroke. Methods: In 104 ischemic stroke patients, treated with intravenous thrombolysis, we assessed the rate of intracerebral hemorrhages documented by computed tomographic scan at 24 hours and within 7 days post-treatment. Fibrinogen levels were determined at 2 hours after therapy: patients were classified as belonging to ‘‘low fibrinogen group’’ if levels decreased to less than 2 g/L and/or by 25% or more. Fibrinogen levels and other known hemorrhagic risk factors were studied using univariate and multivariate analyses. Results: During the first 7 days, an intracerebral hemorrhage was detected in 24 patients (23.1%), and only 6 of these (5.8%) experienced symptomatic bleeding; 41 patients were included in the low fibrinogen group. Among the 24 hemorrhages, 18 occurred in the low fibrinogen group and 6 in the ‘‘normal fibrinogen group’’: the bleeding rate in the low fibrinogen group was significantly higher (43.9%) than that in the normal fibrinogen group (9.5%; odds ratio [OR] 7.43, P ,.001). Univariate and multivariate analyses revealed that only clinical severity (OR 1.15, P , .001) and hypofibrinogenemia (OR 7.47, P , .001) were significantly associated with brain bleeding at 7 days and at 24 hours (P 5.008). Conclusions: An early fibrinogen reduction seems to increase the risk of intracerebral hemorrhage after rtPA treatment in ischemic stroke. Fibrinogen assessment could be a rapid, inexpensive, and widely available tool to help the identification of patients at higher risk of bleeding. Key Words: Fibrinogen—acute stroke—intracerebral hemorrhage—risk factors—thrombolysis—rtPA. Ó 2015 by National Stroke Association
From the *Stroke Unit, Department of Neuroscience, University of Modena and Reggio Emilia, Nuovo Ospedale Civile S. AgostinoEstense, AUSL Modena, Modena; †Section of Haematology, Department of Oncology and Haematology, Policlinico of Modena, University of Modena and Reggio Emilia, Modena; ‡Environmental Protection Agency, ARPA Emilia Romagna District of Reggio Emilia, Reggio Emilia; and xClinical Pathology-Toxicology, Nuovo Ospedale Civile S. Agostino-Estense, AUSL Modena, Modena, Italy. Received March 21, 2014; revision received September 2, 2014; accepted September 6, 2014.
394
Disclosure: The authors report no conflicts of interest. Address correspondence to Andrea Zini, MD, Neurology Clinic, Stroke Unit, Department of Neuroscience, University of Modena and Reggio Emilia, Nuovo Ospedale Civile S.Agostino-Estense, AUSL Modena, via Giardini, 1355 Baggiovara, 41100 Modena, Italy. E-mail: [email protected]. 1052-3057/$ - see front matter Ó 2015 by National Stroke Association http://dx.doi.org/10.1016/j.jstrokecerebrovasdis.2014.09.005
Journal of Stroke and Cerebrovascular Diseases, Vol. 24, No. 2 (February), 2015: pp 394-400
FIBRINOGEN DECREASE AFTER INTRAVENOUS THROMBOLYSIS IN STROKE PATIENTS
Introduction Thrombolytic therapy with administration of intravenous recombinant tissue plasminogen activator (iv rtPA) has been shown to improve long-term functional outcome, and it is recommended for the treatment of eligible acute ischemic stroke patients.1 This therapy, however, is associated with an increased risk of symptomatic intracerebral hemorrhages (sICH). Symptomatic intracranial bleeding in stroke patients treated with iv rtPA is approximately 6%-8%2-4 and is associated with a worse clinical outcome. It is, however, not known what proportion of a worse outcome is attributable to sICH as there is an overlap between the risk factors for thrombolysis-associated sICH and those for poor outcome after thrombolytic therapy with no sICH.2 Many studies have evaluated sICH risk factors in patients receiving thrombolytic therapy, and 2 recent systematic reviews identified the most relevant ones: hyperglycemia, early ischemic changes on computed tomography (CT)/magnetic resonance imaging scan, clinical stroke severity assessed by the National Institutes of Health Stroke Scale (NIHSS) score on admission, advanced age, and high blood pressure.2,3,5-10 The pathophysiological mechanisms of hemorrhagic cerebral transformations could result from the reperfusion of intracranial arteries whose integrity has been disrupted by cerebral ischemia, with an alteration in the blood–brain barrier and an increase in capillary permeability.3 Within this complex pathophysiological process, a central role is played by fibrinogen. rtPA binds to plasminogen within the clot, converting it to plasmin that is a proteolytic enzyme capable of breaking cross-links between fibrin molecules and so dissolving clots, releasing fibrin(ogen) degradation products (FDPs). It is important to note, however, that plasmin is fairly nonspecific in its activity and, besides fibrin, will also break down other circulating proteins, including fibrinogen.11,12 In short, although rtPA is relatively selective for clot-associated fibrin, it can produce a systemic fibrinolytic state (with a secondary hypofibrinogenemia and a D-dimer increase12) and bleeding complications, which were first reported in the setting of the treatment of acute myocardial infarction (AMI). The Thrombolysis in Myocardial Infarction trial13 showed that both streptokinase and rtPA caused a decrease in fibrinogen levels and an increase in FDP and that the rate of hemorrhagic events was higher in patients with increased FDP in both treatment groups and in patients with reduced fibrinogen levels in the rtPA group. Moreover, there were more hemorrhages in patients with greater plasma changes, underlining the importance of coagulation parameters.13,14 Collen et al15 analyzed coagulation and fibrinolysis parameters during intravenous rtPA infusions in patients with AMI and showed that the extent of fibrinogen breakdown is occasionally very
395
important, with a decrease of fibrinogen level less than 1.0 g/L observed in 27% of rtPA-treated patients. Experience and information on the subject of ischemic stroke are poor and sparse. Only 1 recent work by Matosevic et al16 evaluated the extent of fibrinogen depletion in rtPA thrombolysis in stroke patients and its association with intracranial and extracranial bleeding: they showed that a fibrinogen decrease is a significant predictor for bleeding risk and that a temporal relationship exists with the manifestation of hemorrhage. For these reasons, we decided to study early modifications in fibrinogen levels pre- and post-thrombolysis, to assess if they are significantly involved in iv rtPArelated ICH in ischemic stroke patients.
Materials and Methods Study Population We studied consecutive patients with acute ischemic stroke treated with iv infusion of Alteplase (Actilyse, a tPA produced by recombinant DNA technology) from January 1, 2010, to January 1, 2011, in our stroke unit. Patient’s selection was made according to the Safe Implementation of Thrombolysis in Stroke-Monitoring Study criteria and the European Cooperative Acute Stroke Study-3 time window (,4.5 hours). A few of treated patients were included in randomized control trials, such as International Stroke Trial-3, Synthesis Expansion, and Thrombolysis in Elderly Stroke Patients in Italy trial, an Italian randomised controlled trial in patients older than 80 years. The following data were collected on each patient: age, gender, clinical stroke severity assessed by NIHSS preand post-thrombolysis, fibrinogen level at admission and 2 hours after the end of rtPA infusion, baseline glycemia, baseline platelet count, and blood pressure levels.
Plasmatic Fibrinogen Dosage Peripheral blood samples were drawn from each patient, on entry to the emergency room and 2 hours after the end of fibrinolytic infusion. Fibrinogen levels were determined on fresh plasma obtained from blood collected in 5-mL vacuum tubes containing .5 mL sodium citrate (.129 mol/L). The assay was performed using the HemosIL Fibrinogen-C kit assay based on the Clauss method on ACL TOP Coagulation Systems (Instrument Laboratory, Milan, Italy). The Fibrinogen-C kit uses an excess of thrombin to convert fibrinogen to fibrin in diluted plasma. At high thrombin and low fibrinogen concentrations, the rate of reaction is a linear function of the fibrinogen concentration.
Neuroimaging On admission, all patients underwent a noncontrast brain CT scan, which was repeated about 24 hours and
L. VANDELLI ET AL.
396
7 days post-treatment; an additional control scan was conducted whenever neurologic worsening was observed in the first 7 days. We considered every intracranial bleeding that occurred within the first 7 days post-thrombolysis, including both symptomatic and asymptomatic hemorrhages, classifying them according to the Safe Implementation of Thrombolysis in Stroke radiological criteria. Early hemorrhages (early ICH) were those occurring within 24 hours post-thrombolytic treatment. Hemorrhagic transformations were categorized as sICH if patients had neurologic deterioration with an NIHSS increase of more than 1 point and if the CT scan revealed blood at any site in the brain irrespective of its extension (National Institute of Neurological Disorders and Stroke classification). All neuroimaging studies were reviewed by 2 independent, certified, expert neurologists blinded to fibrinogen-level assessment.
Statistical Analysis We assessed fibrinogen-level changes after thrombolytic therapy. A threshold of fibrinogen plasma value critical for the development of bleeding is not reported in the literature. Some studies13,15 have reported a correlation between bleeding and fibrinogen values less than 1.5 or 1.0 g/L or less than 1.0 and .5 g/L. A more recent study, which evaluated the role of fibrinogen concentrate substitution therapy in patients with massive hemorrhage and low plasma fibrinogen concentrations, considered 2 g/L to be the threshold value.17 On these grounds, we performed a logistic regression analysis that analyzed different levels of both absolute fibrinogen plasma values and percentage decrease. We analyzed the following absolute fibrinogen plasma levels as independent variables: lower than 3, 2.5, 2, and 1. We then tested the percentage decrease of fibrinogen values calculated as the percentage of the proportion between the difference of fibrinogen values after and before treatment, divided by the fibrinogen value before treatment: % decrease5½ðfibrinogen value after treatment2 fibrinogen value before treatmentÞ= fibrinogen value before treatment%: For the analysis, we considered only negative values that indicate a decrease in the fibrinogen values after the treatment, and we chose to analyze decreases exceeding 45%, 40%, 35%, 30%, 25%, 20%, 15%, and 10%. After these analyses, we set the cutoff values as fibrinogen absolute levels after treatment below 2 g/L and/or 25% decrease or more, between pre- and post-treatment. This choice derives from the comparison of the logistic regression results between the infarction (dependent variable) and the different absolute values of fibrinogen and between the infarction (dependent variable) and the different values of percentage decrease in fibrinogen: in
both cases, we chose the values associated with the result showing a greater statistical significance (lower P-value). A Pubmed search was conducted looking for sICH risk factors after thrombolytic therapy, focusing primarily on systematic reviews.2,3,5-10 From these data, the following variables were indicated as being most closely related to hemorrhagic transformation: hyperglycemia, clinical stroke severity assessed by baseline NIHSS, age, sex, and high blood pressure. A logistic regression analysis was then conducted on each single parameter. The study population was splitted into 2 groups: a ‘‘low fibrinogen group’’ and a ‘‘normal fibrinogen group’’, to compare important statistical data (eg, mean age, mean NIHSS, mean fibrinogen level pre- and postthrombolysis) and to evaluate any significant differences between the groups. The low fibrinogen group included all patients with post-thrombolysis fibrinogen levels less than 2 g/L and/or a 25% or more decrease of fibrinogen levels respect to baseline, whereas all other patients were included into the normal fibrinogen group. Univariate and multivariate logistic regression analyses were performed to evaluate if fibrinogen levels and other parameters (age, sex, NIHSS, baseline glycemia, baseline blood pressure) were independent risk factors for ICH. We also analyzed the occurrence of ICH and sICH (within 24 hours and 7 days), comparing low and normal fibrinogen groups using contingency tables with Pearson chi-square, to verify a possible association between fibrinogen levels and early onset of ICH or sICH.
Results A total of 104 patients were included in this study, 66 (63.5%) men and 38 (36.5%) women. The mean age was 66.6 years (standard deviation [SD] 13.5; range 21-91) and the mean baseline NIHSS was 10.9 (SD 7.2; range 2-34). On the seventh day follow-up CT scan, 24 patients (23.1%) presented ICH (3 HI1, 8 HI2, 2 PH1, 7 PH2, and 4 PHr1), 14 of which developed in the first 24 hours (3 HI1, 3 HI2, 1 PH1, 4 PH2, and 3 PHr1). Only 6 patients (5.8%) experienced sICH according to the NINDS classification. Mean fibrinogen pretreatment level was 3.34 g/L (SD 1.1) with a significant decrease after thrombolysis (2.47 g/L, 226%, SD .8; P , .001 respect to baseline). Logistic analysis found a significant association between baseline fibrinogen value and ICH (P 5 .048) that became more significant when comparing ICH and fibrinogen post-thrombolysis level (P 5 .029). A logistic regression analysis of the most well-known ICH risk factors (age, sex, baseline NIHSS, 24-hour NIHSS, pre-rtPA arterial blood pressure, pre-rtPA glycemia; baseline platelet count) showed that only baseline NIHSS (P , .001) and 24-hour NIHSS (P , .001) were significantly associated with ICH (Table 1).
FIBRINOGEN DECREASE AFTER INTRAVENOUS THROMBOLYSIS IN STROKE PATIENTS
397
Table 1. Univariate logistic regression analysis: risk factors for ICH post-thrombolysis Variable Fibrinogen level ,2.0 g/L and/or .25% decrease Fibrinogen level ,2.0 g/L Fibrinogen level decreased .25% Platelet count Sex Age Glycemia Blood systolic pressure Baseline NIHSS 24-h NIHSS
OR
P
95% CI
7.43 ,.001 2.620-21.100 2.83 .051 .994-8.075 8.36 ,.001 3.002-23.309 1.00 .485 .996-1.008 .83 .710 .318-2.180 1.19 .138 .945-1.499 1.01 .054 1.000-1.019 1.01 .569 .983-1.031 1.16 ,.001 1.082-1.253 1.11 ,.001 1.051-1.168
Abbreviations: ICH, intracerebral hemorrhage; NIHSS, National Institutes of Health Stroke Scale. Bold values indicate P , .05.
A subsequent logistic regression analysis comparing different post-rtPA fibrinogen cutoff values (,1.0, ,2.0, ,2.5, and ,3.0 g/L) showed that a cutoff value less than 2 g/L offered the best predictive risk factor for ICH, although it did not reach a statistical significance (P 5.051, odds ratio [OR] 5 2.83). The regression analysis, after a class-based analysis with different rates of fibrinogen reduction (45%, 40%, 35%, 30%, 25%, 20%, 15%, and 10%), showed that a higher ICH risk lies in a 25% or more decrease in fibrinogen levels. Based on the results of regression analysis, the patient cohort was divided in the 2 main groups: low fibrinogen group and normal fibrinogen group as described earlier. The low fibrinogen group was composed of 41 patients (39.4%); 6 of them (5.8% of the whole population) had post-rtPA fibrinogen levels less than 2 g/L, 21 (20.2% of the whole population) had a decrease of 25% or more respect to baseline and 14 patients had a reduction in both values (13.5%). Among the 24 patients who presented ICH, 18 were in the low fibrinogen group (Figs 1 and 2) and 6 were in the normal fibrinogen group. The ICH rate during the first 7 days in the low fibrinogen group was significantly higher (43.9%) than that (9.5%) in the normal fibrinogen group (OR 7.43, P , .001; Table 1). The difference remained statistically significant when the analysis was restricted to the patients showing only a fibrinogen decrease greater than 25% (P , .001; OR 8.36; Table 1). Patient characteristics included in the low fibrinogen and normal fibrinogen groups are summarized in Table 2, showing the absence of significant differences except for baseline NIHSS scores and baseline fibrinogen levels. A further subanalysis divided patients into mild (NIHSS score 0-7), moderate (NIHSS score 8-15), and severe (NIHSS score .15) clinical severity groups: similar
Figure 1. A noncontrast brain computed tomographic scan of a patient belonging to low fibrinogen group that experienced post-thrombolysis parenchymal hematoma in the contest of a left frontotemporal ischemic lesion (PH2 type according to the Safe Implementation of Thrombolysis in Stroke radiological criteria), with associated intraventricular hemorrhage.
NIHSS values were found in each group, in particular in the ‘‘severe’’ one (Table 3). In conclusion, the univariate analysis showed that the only variables significantly associated with ICH were baseline NIHSS, 24-hour NIHSS, and belonging to the
Figure 2. A noncontrast brain computed tomographic scan of another low fibrinogen group patient that presented multiple parenchymal hematomas after intravenous recombinant tissue plasminogen activator infusion: the largest one in the contest of a left parieto-occipital ischemic lesion (PH2 type) and some other minor contralateral parenchymal hematomas, remote from the ischemia (PHr1 type).
L. VANDELLI ET AL.
398
Table 2. Comparison of patient characteristics between normal and low fibrinogen groups Low fibrinogen Normal group, N group, N
Variable Age (mean) Sex (male %) Baseline NIHSS (mean) 24-h NIHSS (mean) Glycemia (mean) Blood pressure (systolic) Blood pressure (diastolic) Platelet count Fibrinogen-level prethrombolysis (mean)
66.71 65.9 12.8 9.34 135.78 148.63 78.54 207.73 362.21
66.46 61.9 9.71 7.27 133.86 146.78 78.71 192.11 311.72
P NS NS .031* NS NS NS NS NS .028*
Abbreviation: NIHSS, National Institutes of Health Stroke Scale. Bold values indicate P , .05. *P , .05.
low fibrinogen group (fibrinogen levels post-thrombolysis ,2 g/L and/or a $25% decrease). This finding was confirmed by the multivariate analysis (Table 4). Considering early ICH, 14 events were found, 10 in the low fibrinogen group (71.4% of early ICH and 9.62% of total ICH) and 4 in the normal fibrinogen group (28.6% of early ICH and 3.85% of total ICH), showing that hemorrhagic risk remains significantly higher in the low fibrinogen group when analyzing only ICH less than 24-hour subgroup (P 5 .008; Table 5). Of the 6 sICH reported, 5 (83.3%) occurred during the first 24 hours post-thrombolysis. The logistic regression analysis for sICH and early sICH did not reveal statistical significance between the normal and low fibrinogen groups (P . .05; Table 5).
Discussion The development of ICH is the most serious adverse event related to iv rtPA as it carries a relevant morbidity
Table 3. Subgroup analysis of baseline clinical severity between low and normal fibrinogen groups
Clinical severity Mild (NIHSS score 0-7) Moderate (NIHSS score 8-15) Severe (NIHSS sore .15)
Fibrinogen group Mean Normal Low Normal Low Normal Low
4.83 4.31 10.48 11.36 20.58 22.14
P*
n
SD
30 13 21 14 12 14
1.46 1.75 .319 2.58 2.50 .324 3.34 4.82 .355
Abbreviation: NIHSS, National Institutes of Health Stroke Scale. *Significance P , .05.
Table 4. Multivariate logistic regression analysis of risk factors for ICH post-thrombolysis Variable
OR
P
95% CI
Fibrinogen level ,2.0 g/L and/or .25% decrease Baseline NIHSS Glycemia Blood systolic pressure
7.47
,.001
2.26-24.74
1.15 1.01 1.00
,.001 .131 .802
1.06-1.25 1.00-1.02 .97-1.03
Abbreviations: ICH, intracerebral hemorrhage; NIHSS, National Institutes of Health Stroke Scale. Bold values indicate P , .05.
and mortality.4 In our study, we found an overall ICH rate within the first 7 days of 23.1%, which is lower than that reported from previous studies.18 Similarly, we found 6 patients with sICH (5.8%), a rate comparable with those reported in the literature (6%-8%).2-4 Many studies in the literature have reported various risk factors for ICH, the most relevant of which are age, clinical severity, pre-rtPA arterial blood pressure, and pre-rtPA glycemia.2,3 In our study, we found that the only already known variables significantly associated with the development of ICH are baseline NIHSS (OR 1.16, P , .001) and 24-hour NIHSS (OR 1.11, P , .001) (Table 1) because neither baseline blood glucose nor systolic blood pressure pretreatment attained the statistical significance. On the other hand, we demonstrated by logistic regression analysis that a decrease in post-thrombolysis fibrinogen levels of less than 2 g/L and/or of 25% or more (low fibrinogen group) was statistically significantly related with post-rtPA ICH. Indeed, in this patient group, the ICH rate within 7 days was 43.9%, significantly higher than the 9.5% of the normal fibrinogen group (OR 7.4, P , .001) (Table 1). When the 2 subgroups of patients included in the low fibrinogen group were separately analyzed, however, the results were
Table 5. Logistic regression analysis of fibrinogen decrease as risk factor for ICH, early ICH, sICH, and early sICH Low Normal Total Subgroups of fibrinogen fibrinogen hemorrhage (n 5 63) (n 5 41) (n 5 104) ICH Early ICH sICH Early sICH
6 (9.5%) 4 (28.6%) 2 (33.3%) 2 (40%)
18 (43.9%) 10 (71.4%) 4 (66.7%) 3 (60%)
24 14 6 5
P ,.001* .008* .586 .480
Abbreviations: ICH, intracerebral hemorrhage; sICH, symptomatic intracerebral hemorrhage. Bold values indicate P , .05. *P , .05.
FIBRINOGEN DECREASE AFTER INTRAVENOUS THROMBOLYSIS IN STROKE PATIENTS
somewhat different. Indeed, considering only patients with fibrinogen decreases of 25% or more, the difference remains statistically significant (OR 8.36, P , .001), whereas those with fibrinogen levels less than 2 g/L did not, although the results were at limit of significance (OR 2.83, P 5 .051). Our data are in line with results recently published by Matosevic et al,16 which showed that a 6-hour reduction in fibrinogen levels greater than 2 g/L after thrombolysis can—as with stroke severity, diabetes, and subtherapeutic Vitamin K Antagonists pretreatment—heighten the risk of sICH and systemic bleeding. Moreover, as also reported in the Matosevic study, we found that a higher baseline fibrinogen level seems to enhance the risk of a more significant decrease post-treatment. This suggests a correlation between extent of fibrinogen depletion (delta pre- and post-rtPA) and baseline fibrinogen values that could be explained by a ‘‘consumption effect,’’ that is, a more active early fibrinogen degradation coagulopathy. These findings support the hypothesis that postthrombolysis intracranial bleedings are, at least partially, linked to an early rtPA-related coagulopathy that causes a fibrinogen breakdown because of rtPA’s relative selectivity for clot-associated fibrin.11,12 This type of coagulopathy, with an early increase in FDP and fibrinogen consumption, was first described in iv thrombolytic therapy used for AMI.13-15 Some years later, Trouillas19 studied the possible correlation between acquired hypofibrinogenemia and sICH after thrombolysis in stroke patients and found that an increase in FDP, like D-dimer, 2 hours post-thrombolysis was statistically significant and appeared to be a predictive factor for early parenchymal hematomas (,24 hours). Nevertheless, we consider FDP a less specific parameter than fibrinogen of such coagulopathy: in fact, D-dimer, which is the most frequently measured FDP, always increases after a cerebral ischemic event, even without thrombolysis12 as it is a by-product of clot degradation and remodeling that constantly occurs during vessel occlusion. Other factors have been implicated in the genesis of rtPA-related sICH, such as the pretreatment levels of fibrinolysis inhibitors (ie, plasminogen activator inhibitor-17,8 or matrix metalloproteinases-920), but studies gave conflicting results. Moreover, the methods for determining plasma levels of these biomarkers are both very expensive and too slow for use in emergencies, and therefore, they cannot be used to stratify sICH risk in rtPA therapy decision making. On the contrary, determining fibrinogen levels is very quick, easy, and cheap and can be performed in any hospital laboratory. The decrease in fibrinogen was demonstrated to be a significant risk factor for bleeding, even when the analysis was restricted only to early (ie, within 24 hours
399
from treatment) ICH (P 5 .008; Table 5). This aspect is very important as some authors have suggested that only early ICH can regarded as linked to thrombolysis because of the rtPA short plasma half-life (around 5 minutes). Matosevic et al16 found that most of bleeding occurred between 2 and 24 hours post-thrombolysis, an interval that broadly overlaps with the period in which fibrinogen degradation coagulopathy develops. For this reason, they evaluated the fibrinogen decrease 6 hours post-rtPA therapy and its correlation with hemorrhages. On the other hand, other studies on post-fibrinolysis hemostatic changes in myocardial infarction patients have shown that fibrinogen reaches its lowest point some time between 90 minutes and 3 hours.21,22 On these grounds, and on the pharmacologic basis of the short rtPA halflife, we decided to restrict the analysis of fibrinogen levels to 2 hours post-thrombolysis, assuming a very early fibrinogen-related coagulopathy. In our study, the analysis of sICH, either early or total, failed to show a statistically significant correlation with fibrinogen levels (Table 5), but this finding could be explained by the very small number of patients with sICH, making it difficult to reach reliable conclusions. One limitation of our study was the significant difference in baseline NIHSS score that existed between the normal and low fibrinogen groups. This factor could diminish the relevance of our results as it makes it less clear whether the greater bleeding rate in the low fibrinogen group was attributable to the fibrinogen decrease itself or to clinical severity. As a final point, the limited number of included patients did not allow a reliable stratification of study population in several subgroups of patients, such as those with sICH and early sICH. In conclusion, an early decrease in fibrinogen levels seems to represent a risk factor for ICH within the first 7 days after rtPA therapy and even within the first 24 hours in ischemic stroke patients. This result confirms the hypothesis that iv rtPA causes a related coagulopathy because of relative selectivity of Alteplase for clotassociated fibrin, which results in fibrinogen breakdown. We, therefore, recommend assessing fibrinogen levels before and 2 hours postinfusion of rtPA in all thrombolyzed patients: this would allow the identification of patients at higher bleeding risk. The rapidity, cheapness, and wide availability of such a test further reinforce the utility value of this parameter. Further RCT studies could assess the preventive role of therapies like fibrinogen intravenous replacement or plasma transfusion in the low fibrinogen group in avoiding post-thrombolysis ICH. Acknowledgment: The authors thank all neurologists and nurses working in Neurology Clinic at Nuovo Ospedale Civile S.Agostino-Estense.
L. VANDELLI ET AL.
400
References 1. Adams HP, Del Zoppo G, Alberts MJ, et al. A guideline from the American Heart Association/American Stroke Association Stroke Council, Clinical Cardiology Council, Cardiovascular Radiology and Intervention Council, and the Atherosclerotic Peripheral Vascular Disease and Quality of Care Outcomes in Research Interdisciplinary Working Groups. Stroke 2007;38:1655-1711. 2. Lansberg MG, Albers GW, Wijman CA. Symptomatic intracerebral hemorrhage following thrombolytic therapy for acute ischemic stroke: a review of the risk factors. Cerebrovasc Dis 2007;24:1-10. 3. Derex L, Nighoghossian N. Intracerebral haemorrhage after thrombolysis for acute ischaemic stroke: an update. J Neurol Neurosurg Psychiatry 2008;79:1093-1099. 4. Seet RC, Rabinstein AA. Symptomatic intracranial hemorrhage following intravenous thrombolysis for acute ischemic stroke: a critical review of case definitions. Cerebrovasc Dis 2012;34:106-114. 5. Trouillas P, Von Kummer R. Classification and pathogenesis of cerebral hemorrhages after thrombolysis in ischemic stroke. Stroke 2006;37:556-561. 6. Derex L, Hermier M, Adeleine P, et al. Clinical and imaging predictors of intracerebral hemorrhage in stroke patients treated with intravenous tissue plasminogen activator. J Neurol Neurosurg Psychiatry 2005;76:70-75. 7. Ribo M, Montaner J, Molina CA, et al. Admission fibrinolytic profile is associated with symptomatic hemorrhagic transformation in stroke patients treated with tissue plasminogen activator. Stroke 2004;35:2123-2127. 8. Cocho D, Borrell M, Marti-Fabregas J, et al. Pretreatment hemostatic markers of symptomatic intracerebral hemorrhage in patients treated with tissue plasminogen activator. Stroke 2006;37:996-999. 9. Kellert L, Sykora M, Gumbinger C, et al. Blood pressure variability after intravenous thrombolysis in acute stroke does not predict intracerebral hemorrhage but poor outcome. Cerebrovasc Dis 2012;33:135-140. 10. K€ ohrmann M, Schellinger PD. Symptomatic intracranial hemorrhage after thrombolysis. Cerebrovasc Dis 2007; 23:83-84. 11. Klabunde RE. Cardiovascular Physiology Concepts. 2nd ed. Lippincott Williams & Wilkins/Baltimore 2011. 12. Francis CW, Kornberg A. Fibrinogen- and fibrindegradation products during fibrinolytic therapy. Ann N Y Acad Sci 1992;667:310-323. 13. Rao AK, Pratt C, Breke A, et al. Thrombolysis in Myocardial Infarction (TIMI) trial-phase I. Hemorrhagic mani-
14.
15.
16.
17.
18.
19.
20.
21.
22.
festations and changes in plasma fibrinogen and the fibrinolytic system in patients treated with recombinant tissue plasminogen activator and streptokinase. J Am Coll Cardiol 1988;11:1-11. Bovill EG, Tracy RP, Knatterud GL, et al. Hemorrhagic events during therapy with recombinant tissue plasminogen activator, heparin, and aspirin for unstable angina (Thrombolysis In Myocardial Ischemia, phase IIIB trial. Am J Cardiol 1997;79:391-396. Collen D, Bounameaux H, De Cock F, et al. Analysis of coagulation and fibrinolysis during intravenous infusion of recombinant human tissue-type plasminogen activator in patients with acute myocardial infarction. Circulation 1986;73:511-517. Matosevic B, Knoflach M, Werner P, et al. Fibrinogen degradation coagulopathy and bleeding complications after stroke thrombolysis. Neurology 2013;80: 1216-1224. Fenger-Eriksen C, Jensen TM, Kristensen BS, et al. Fibrinogen substitution improves whole blood clot firmness after dilution with hydroxyethyl starch in bleeding patients undergoing radical cystectomy: a randomized, placebo-controlled clinical trial. J Thromb Haemost 2009;7:795-802. Dzialowski I, Warwick Pexman JH, Barber PA, et al. Asymptomatic hemorrhage after thrombolysis may not be benign prognosis by hemorrhage type in the Canadian Alteplase for Stroke Effectiveness Study Registry. Stroke 2007;38:75-79. Trouillas P, Derex L, Philippeau F, et al. Early fibrinogen degradation coagulopathy is predictive of parenchymal hematomas in cerebral rtpa thrombolysis. A study in 157 cases. Stroke 2004;35:1323-1328. Montaner J, Molina CA, Monasterio J, et al. Matrix metalloproteinase-9 pretreatment level predicts intracranial hemorrhagic complications after thrombolysis in human stroke. Circulation 2003;107:598-603. Granger CB, Becker R, Tracy RP, et al. Thrombin generation, inhibition and clinical outcomes in patients with acute myocardial infarction treated with thrombolytic therapy and heparin: results from the GUSTO-I Trial. GUSTO-I Hemostasis Substudy Group. Global Utilization of Streptokinase and TPA for Occluded Coronary Arteries. J Am Coll Cardiol 1998;31:497-505. Stangl K, Laule M, Tenckhoff B, et al. Fibrinogen breakdown, long-lasting systemic fibrinolysis, and procoagulant activation during alteplase double-bolus regimen in acute myocardial infarction. Am J Cardiol 1998; 81:841-847.
