Martin-Schild S, Hallevi H, Albright KC, Khaja AM, Barreto AD, Gonzales NR, Grotta JC, Savitz SI. Aggressive Blood Pressure–Lowering Treatment Before Intravenous Tissue Plasminogen Activator Therapy in Acute Ischemic Stroke. Arch Neurol. 2008;65(9):1174-1178. doi:10.1001/archneur.65.9.1174
Copyright 2008 American Medical Association. All Rights Reserved. Applicable FARS/DFARS Restrictions Apply to Government Use.2008
Patients with acute ischemic stroke (AIS) commonly have elevated blood pressure (BP). Guidelines have recommended against treatment with intravenous tissue plasminogen activator (tPA) when aggressive measures such as continuous infusion with nicardipine hydrochloride are required to maintain BP lower than 185/110 mm Hg. We evaluated the effect of elevated BP and its management on clinical outcomes after tPA therapy in AIS.
To evaluate safety and outcome in patients with AIS who require treatment to lower BP before tPA therapy and to compare safety and outcome in patients who received aggressive treatment with nicardipine with those who received labetalol hydrochloride before tPA.
Retrospective review of medical records for all patients who received intravenous tPA within 3 hours of AIS onset.
One hundred seventy-eight patients with AIS treated with tPA.
Main Outcome Measures
Occurrence of symptomatic intracerebral hemorrhage and neurologic deterioration.
Fifty patients required BP lowering before tPA therapy. Twenty-four of these patients (48%) received nicardipine either after labetalol or as first-line therapy. Patients requiring antihypertensive agents had higher baseline blood glucose concentrations, incidence of hypertension, and National Institutes of Health Stroke Scale scores. The rate of adverse events and of modified Rankin score at discharge were not significantly different in patients without BP-lowering treatment compared with patients given either labetalol or nicardipine before intravenous tPA therapy.
Blood pressure lowering before intravenous tPA therapy, even using aggressive measures, may not be associated with a higher rate of hemorrhage or poor outcome. Data suggest that patients with AIS requiring aggressive treatment to lower BP should not be excluded from receiving tPA therapy. A prospective study is needed to support these conclusions.
As many as 10% of otherwise eligible patients do not receive tissue plasminogen activator (tPA), the only proved therapy for acute ischemic stroke (AIS), because of severely elevated blood pressure (BP).1 In the past, guidelines recommended against giving tPA to treat AIS when aggressive measures (such as continuous infusion or more than 2 infusions of antihypertensive agents) are required to maintain BP lower than 185/110 mm Hg.2 According to an independent tPA review committee in charge of reviewing and reanalyzing the pivotal National Institute of Neurological Disorders and Stroke Trial, “It is biologically plausible that hypertension management could affect clinical outcome”3 in patients with AIS treated with tPA. However, the committee was unable to address this question because of insufficient data. To our knowledge, no study has specifically addressed the safety of aggressive treatment of BP as a potential factor affecting outcome in patients with AIS treated with tPA.
One reason to withhold tPA therapy in patients with persistently elevated BP exceeding the recommended limit despite antihypertensive therapy is increased risk of hemorrhagic transformation (HT).2,4 The evidence that elevated BP increases the risk of bleeding from tPA is largely derived from the cardiac literature,5,6 and there was no evidence from the National Institute of Neurologic Disorders and Stroke tPA study that elevated BP increases the risk of bleeding from tPA.4 In that tPA stroke trial, 19% of enrolled patients had BP higher than 185/110mm Hg, and about 5% had BP exceeding this limit at randomization.3,4 Though treatment with labetalol hydrochloride, nicardipine hydrochloride, or transdermal nitroglycerin ointment (nitropaste) was permitted, there was no protocol for the use of these medications before tPA treatment that included the dose or time of medication administration; thus, the effect of elevated BP and its management on clinical outcomes could not be definitively assessed. A retrospective analysis of the European Cooperative Acute Stroke Study II showed a weak association between severe hypertension and hemorrhagic transformation after tPA therapy (odds ratio, 1.02;95% confidence interval, 1.00-1.03; P = .02).7 Other studies of patients with AIS treated with tPA failed to identify baseline BP as an independent predictor for intracranial hemorrhage (ICH)8 and poor outcome.9- 12
An additional problem is that lowering BP rapidly from very high values may be detrimental by reducing penumbral blood flow, which may increase infarct volume and result in worse outcome.13,14 It is not known whether BP reduction to levels just less than 185/110 mm Hg is harmful in AIS and whether any harmful effect is outweighed by the beneficial effect of thrombolysis.
More aggressive management of BP may increase the number of patients with AIS who are eligible for tPA therapy. In the Memorial Hermann Hospital, we have been treating patients with severely elevated BP and AIS with intravenous agents to lower BP to the recommended limits so that tPA may be administered. The most common antihypertensive medications used are labetalol as intravenous (IV) boluses and nicardipine as continuous infusion. In many cases, nicardipine infusion is begun after administration of labetalol to control BP. In some cases, we initiate treatment of elevated BP with nicardipine as a first-line agent. Although labetalol (≤40 mg) before tPA therapy had been formally recommended,2 nicardipine was only recently included in the new American Heart Association guidelines15 as an acceptable BP-lowering agent before tPA therapy, yet no data support the safety of nicardipine in patients with AIS treated with IV tPA. The purpose of this study was to evaluate safety and outcome in patients with AIS who require BP-lowering treatment before tPA administration and to compare safety and outcome in patients who received aggressive treatment with nicardipine vs those who received labetalol before tPA therapy.
An open, retrospective, nonrandomized, observational study was performed. Medical records were reviewed for all patients identified from our prospectively collected stroke registry who received IV tPA therapy because of AIS from January 1, 2004, through December 13,2006. Patients treated with BP-lowering medications other than labetalol or nicardipine, treated after the 3-hour window for therapy, treated with adjuvant intra-arterial therapy, or enrolled in clinical trials were excluded. The study was approved by the institutional review boards for the University of Texas Health Science Center at Houston and the Memorial Hermann Healthcare system.
Data collected included time of symptom onset and patient arrival, all recorded vital signs, clinical findings, baseline computed tomographic findings, comorbidities, medication history, dosage and time of administration of antihypertensive agents in the emergency department, and time of tPA administration. We defined cortical signs as the presence of aphasia, visual field cut, conjugate gaze deviation, or inattention on the National Institutes of Health Stroke Scale (NIHSS). Early ischemic changes included sulcal effacement, loss of gray-white junction, insular ribbon sign, subtle hypodensity, and basal ganglia obscuration. All patients were treated at the same emergency department by the same stroke team (all of the authors). Both labetalol and nicardipine have been the primary agents used to control BP in this setting by our stroke team, and our use of these drugs was standardized and published in our stroke program handbook.16 Labetalol hydrochloride, usually 10 to 20 mg, was given as an IV bolus and was repeated as necessary. Nicardipine hydrochloride, 5 mg/h, was given as a continuous intravenous infusion, and the dose was increased every 10 to 15 minutes as needed, but no more than 15 mg/h.
The primary goal of this study was to compare the safety of BP-lowering treatment before tPA therapy. We defined standard BP lowering (SL) as monotherapy with labetalol hydrochloride in a dose not exceeding 40 mg, and aggressive BP lowering (AL) as the need for additional nicardipine treatment or nicardipine treatment alone.
Blood pressure was monitored before, during, and after tPA administration. All patients underwent baseline computed tomography and follow-up neuroimaging. All patients including those receiving nicardipine were admitted to the same stroke unit or the neurologic intensive care unit if intubation was needed. No arterial line was required for nicardipine infusion. Patients receiving labetalol or nicardipine were all treated according to the same BP-management guidelines for post-tPA treatment.2 Both drugs were replaced with oral agents as soon as possible.
The primary end points were adverse events, discharge disposition, and modified Rankin score at discharge. Adverse events included symptomatic hemorrhage (sICH), all grades of hemorrhagic transformation at neuroimaging performed after tPA therapy, and neurologic deterioration (ie, increase in NIHSS score of >2). All primary outcomes were collected by a data abstractor blinded to the purpose of the study and to the treatment group. All computed tomographic scans and magnetic resonance images and reports were reviewed by either one of us (S.M.-S. or H.H.) who was blinded to the BP-lowering treatment. Grading of hemorrhagic changes was performed using previously established criteria.8 The daily progress notes were reviewed for all patients with a type 1 or type 2 parenchymal hematoma to determine whether the hemorrhage was symptomatic. Symptomatic ICH was defined as any parenchymal hematoma likely to be the cause of neurologic deterioration. Good outcomes were defined as a discharge Modified Rankin Scale score of 0 to 2 or discharge to home or rehabilitation.
Statistical analyses were performed using commercially available software (SPSS version 15.0; SPSS, Inc, Chicago, Illinois). Continuous variables were reported as mean (SD) when the distribution was normal and as median (interquartile range) when the distribution was nonnormal. Categorical variables were analyzed using χ2 and Fisher exact tests as appropriate. Bivariate logistic regression was used for analysis of dichotomous outcome variables.
We identified 178 patients who were treated with IV tPA within 3 hours of symptom onset of AIS and who met all inclusion and exclusion criteria. Blood pressure medications were given to 50 patients (28.1%) to meet the recommended BP level (<185/110 mm Hg) before administration of tPA. Of these, 26 patients (52%) required only labetalol therapy to reach the target BP (SL group) and 42 patients (48%) required nicardipine therapy in addition to labetalol or alone (AL group). Patient baseline characteristics are given in Table 1.
Patients who required BP-lowering medications had higher baseline NIHSS scores, baseline blood glucose concentrations, and history of hypertension rates. As expected, initial systolic and diastolic BP were significantly higher in patients who required antihypertensive agents before tPA therapy. There was no difference in time from arrival to tPA administration and onset of AIS to tPA administration between the 2 groups (Table 1).
At univariate analysis (Table 2), we found no significant difference in sICH, neurologic deterioration, or any grade of hemorrhagic transformation between patients who received antihypertensive therapy and those who did not. After controlling for age, baseline NIHSS score, and blood glucose concentration at admission, administration of BP-lowering medications before tPA therapy was not associated with a poor outcome.
There were no significant differences in age, NIHSS scores, blood glucose concentrations, or rates of early ischemic changes on baseline computed tomographic scans (Table 3) between the SL and AL groups. There was a trend toward higher initial systolic BP in the AL group, and the times from arrival to tPA administration and from AIS onset to tPA administration were substantially longer in the AL group.
There was no difference in the rates of any grade of hemorrhagic transformation or symptomatic hemorrhagic transformation (Table 4). Although nonsignificant, mortality and the rate of neurologic deterioration were higher in the AL group. None of the instances of neurologic deterioration were related to BP lowering. The rate of good outcome was higher in the AL group.
Approximately 30% of our patients with AIS seen within the time window for IV tPA therapy had BP higher than the suggested safe limit for thrombolytic therapy. The rate of admission BP higher than 185/110mm Hg in our population was higher than that in the National Institute of Neurological Disorders and Stroke4 tPA trial (19% of enrolled patients). We observed several important differences between patients who required BP-lowering treatment and those who did not. They had more severe strokes and their blood glucose concentration was higher, predicting they would have a worse outcome if all other factors were equal. As expected, they more frequently had a history of hypertension. Though rates of adverse events (sICH, neurologic deterioration, and in-hospital death) were higher in the group who received BP-lowering agents, we were underpowered to detect the significance of the magnitude of differences observed in our patients.
When comparing the SL and AL groups, no differences were found in the baseline characteristics except for initial systolic BP. Patients in the AL group had (nonsignificant) higher rates of death and neurologic deterioration compared with those in the SL group. None of these adverse events were ascribed to either low BP or uncontrolled BP. Given the small sample, these results are in line with the odds, but may also indicate that patients requiring AL are sicker to begin with. It is also possible that higher initial BP may affect recanalization in response to tPA, as has been recently reported.12 A larger prospective study may be required to address this question. That patients in the AL group had a higher rate of good outcome and shorter length of stay provides reassurance that AL may not be harmful.
Our analysis shows an important clinical finding: there was significant delay in administration of tPA therapy when AL was required. There are several reasons why this may have occurred. First, these patients may have had more resistant BP requiring a longer time to reach the target BP. Second, although labetalol is readily available in our emergency department and can be administered within 1 minute of an order, nicardipine requires preparation that includes mixing and setting up of an infusion system, which introduces delay. The difference in time to drug delivery was not monitored but may be as long as 5 to 10 minutes. Third, there is a tendency to start nicardipine infusion cautiously at low infusion rates (5 mg/h), and peak activity may be seen several minutes later. Our study was not powered to examine differences in outcome resulting from such delay but, in theory, it can be important.17
This study has several limitations. First, the results reflect a single center's experience. Because our study was retrospective and included a small sample of patients who received AL or SL therapy, our results are vulnerable to type II error. We tried to minimize this by addressing nonsignificant trends in the results. The retrospective study design precluded standardization of BP measurements, resulting in a variable number of recordings and variable dosages of labetalol before the addition of nicardipine. We also do not have detailed information on BP treatment and response after tPA infusion. Inasmuch as all patients were routinely admitted to the same stroke unit and received standardized BP treatment, the chances of a significant bias are minimal. The presence of small amounts of petechial hemorrhage may have been missed in some patients but likely would not have clinical significance. Our institution's overall low rate of sICH may have prevented detection of a difference if one existed. Second, we do not have accurate data for patients who were not treated because of uncontrolled hypertension. This may cause bias toward patients with less resistant hypertension. Our experience suggests that once using AL these patients are rarely encountered.
Overall, the results of the present study provide the first experimental support for the revised American Heart Association guidelines allowing tPA therapy in patients requiring aggressive BP management and also provide support for the use of nicardipine in patients with AIS who are eligible for thrombolytic therapy. Aggressive control of severely elevated BP is feasible and should not automatically exclude otherwise eligible patients with AIS from receiving thrombolytic therapy. This study supports the need for a prospective investigation to further determine the adverse events and outcomes in patients with AIS who require BP lowering before thrombolytic therapy.
Correspondence: Sean I. Savitz, MD, Vascular Neurology Program, Department of Neurology, University of Texas Health Science Center at Houston, 6431 Fannin St, Medical School Bldg Room 7.128, Houston, TX 77030 (Sherylmartinschild@yahoo.com).
Accepted for Publication: April 8, 2008.
Author Contributions: Dr Martin-Schild had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis. Study concept and design: Martin-Schild, Albright, Khaja, Barreto, Gonzales, and Grotta. Acquisition of data: Martin-Schild, Hallevi, Khaja, and Barreto. Analysis and interpretation of data: Martin-Schild, Hallevi, Albright, Grotta, and Savitz. Drafting of the manuscript: Martin-Schild, Albright, Khaja, Barreto, Gonzales, Grotta, and Savitz. Critical revision of the manuscript for important intellectual content: Martin-Schild, Hallevi, Barreto, Grotta, and Savitz. Statistical analysis: Hallevi and Albright. Obtained funding: Martin-Schild, Khaja, and Barreto. Administrative, technical, and material support: Hallevi. Study supervision: Gonzales, Grotta, and Savitz.
Financial Disclosure: Dr Grotta has received lecturing honoraria from PDL BioPharma, Inc.
Funding/Support: This study was supported by training grant 5-T32-NS007412-09 from the National Institutes of Health to the University of Texas Medical School at Houston Stroke Program (Drs Martin-Schild, Barreto, and Khaja).
Additional Information: PDL BioPharma, Inc, was not a sponsor of this study.