Orolingual Angioedema During or After Thrombolysis for Cerebral Ischemia
Background and Purpose—Orolingual angioedema (OLAE) is a life-threatening complication of intravenous thrombolysis. Our objective was to compare outcomes of patients with and without OLAE.
Methods—We prospectively included consecutive patients who received intravenous thrombolysis for cerebral ischemia at Lille University Hospital. We examined tongue and lips every 15 minutes during thrombolysis and ≤30 minutes after. We evaluated the 3-month outcome with the modified Rankin scale (mRS) and compared outcomes of patients with and without OLAE.
Results—Of 923 consecutive patients, 20 (2.2%) developed OLAE. None of them needed oro-tracheal intubation. They were more likely to be under angiotensin-converting enzyme inhibitors (adjusted odds ratio [adjOR], 3.9; 95% confidence interval [CI], 1.6–9.7; P=0.005) to have total insular infarcts (OR, 5.0; 95% CI, 1.5–16.5; P=0.004) and tended to develop more symptomatic intracerebral hemorrhages. Results concerning angiotensin-converting enzyme inhibitors were not modified after adjustment for propensity scores (OR, 4.4; 95% CI, 1.6–11.9; P=0.004) or matched analysis based on propensity scores (OR, 3.4; 95% CI, 1.3–8.1; P=0.010). Patients with OLAE did not significantly differ at 3 months for the proportion of patients with mRS score of 0 to 1 (adjOR, 0.9; 95% CI, 0.3–2.1), mRS score of 0 to 2 (adjOR, 0.8; 95% CI, 0.1–1.8), and death (adjOR, 1.1; 95% CI, 0.3–3.8).
Conclusions—OLAE occurs in 1 of 50 patients who receive intravenous thrombolysis, 1 of 10 in case of total insular infarct, and 1 of 6 if they are under angiotensin-converting enzyme inhibitors. Their long-term outcome does not differ from that of other patients.
In eligible patients with acute cerebral ischemia, intravenous recombinant tissue-type plasminogen activator (r-tPA) increases the proportion of patients who survive without dependency.1–3 Orolingual angioedema (OLAE) is an acute swelling of the lips and the tongue that spontaneously resolves, but may be life threatening because of the risk of upper airways obstruction. Most studies reported isolated cases or small series, and larger series aimed at identifying predictors of OLAE, but not its influence on the 3-month outcome. The outcome of patients with OLAE may be influenced by an increased mortality because of upper airways obstruction. In survivors, more subtle changes in outcome might occur because of the consequence of a lower quality of the general management of thrombolysis in this context and of the uncertainty on the effect on the penumbra of drugs used to treat OLAE.
The aim of our study was to compare outcomes in a large cohort of patients with and without OLAE during or after intravenous thrombolysis for cerebral ischemia.
Patients and Methods
Analysis of the Lille Database
We used data prospectively collected in a registry of consecutive patients treated with intravenous r-tPA for cerebral ischemia in the stroke center of the Lille University Hospital, from September 30, 2003 to February 4, 2015. The general organization,4,5 eligibility criteria for thrombolysis,2,6,7 clinical assessment,8–10 imaging procedures, and treatment administration7 have been previously described.4,5 Patients from this cohort were included in previous studies of our group including one in which the occurrence of OLAE was reported11 and in a case report.12
Diagnosis and Management of OLAE
We defined OLAE as an unprovoked localized swelling and edema of the tongue or the lips that is not explained by a local bleeding and disappears completely within a few hours. In the protocol of thrombolysis, nurses were asked to inspect the tongue and the lips of the patient every 15 minutes during thrombolysis and ≤30 minutes after the end of the infusion. In case of swelling or edema of the tongue or the lips during this period, the nurses were asked to call immediately the senior neurologist of the stroke unit and the intensive care team, and an infusion of corticosteroids was immediately started together with antihistaminic therapy and oxygen therapy. The decision to intubate and to stop or not, r-tPA was left to the decision of the neurologist and of the intensivist in charge of the patient, on a case-by-case basis.
We performed the statistical analysis with the SPSS 22.0 package for windows. We used median values, interquartile ranges, and percentages. We compared groups for categorical variables, with the χ2 test with Yates’ correction or Fisher exact test when appropriate, and for continuous variables with the Mann–Whitney U test. We calculated odds ratios (OR) and 95% confidence intervals (95% CIs). For the need of statistical analyses, some of the missing data were replaced as follows: when time at onset was unknown, for example, in wake-up strokes, or data missing for weight, dose of r-tPA, blood pressure, serum glucose level, and platelet count, we replaced the missing value by the median value in the whole cohort.
The first step of the analyses consisted in the determination of the proportion of patients of OLAE, with 95% CI. The second step consisted in a comparison of the main baseline characteristics of patients with and without OLAE. Because an association between insular infarcts and OLAE was reported in small series without controls,13 we analyzed this association in a subgroup of patients who underwent magnetic resonance imaging before thrombolysis to identify with certainty the acute event on diffusion-weighted images. For each patient with OLAE, we selected the 4 patients without, who had the closest NIHSS score and a lesion located in the same arterial territory divided into (1) middle cerebral artery territory, (2) anterior cerebral artery territory, (3) posterior cerebral artery territory, (4) lacunar, and (5) posterior fossa. When >4 patients met selection criteria, we selected in priority those of the same sex, then of the closest weight, and then of the closest age. The first magnetic resonance imaging scans performed at admission were analyzed by 2 investigators blinded to the occurrence of OLAE. We classified patients as having or not any insular lesion, and if yes, whether it was total (ie, involving at least 2/3 of the insula) or partial. Adjusted ORs were calculated from a logistic regression analysis with presence of OLAE as dependent variable and as explanatory variables, those of the variables that were present at stroke onset and associated with a P value of <0.20 at the bivariate analysis, age, diabetes mellitus, and arterial hypertension being forced into the model. Propensity scores have been generated by a logistic regression model. Variables for the propensity score were variables present before stroke, included step-by-step using an α value of 0.20: arterial hypertension, diabetes mellitus, hypercholesterolemia, previous myocardial infarction, previous atrial fibrillation, previous ischemic stroke, and previous modified Rankin scale (mRS) score of 0 or 1. We compared systolic blood pressure before thrombolysis and ≤6 hours after the bolus, between the patients with OLAE and the patients without selected for the imaging comparison.
The last step consisted in a comparison of outcomes at 3 months between patients with and without OLAE. The primary end point was the proportion of survivors without handicap (ie, mRS score of 0 or 1 or similar to the prestroke mRS) at 3 months. Secondary end points were the proportion of survivors without dependency (mRS score of 0–2 or similar to the prestroke mRS) at 3 months, occurrence of symptomatic intracerebral hemorrhage (s-ICH) according to the National Institute of Neurological Disorders and Stroke (NINDS)1 and the European Cooperative Acute Stroke Study (ECASS) 2 definitions,6 and death within 3 months.
The study was considered as observational by the internal review board (Comité de protection des personnes Nord Ouest IV, March 1, 2010; Comité consultatif sur le traitement de l’information en matière de recherche dans le domaine de santé, no. 10.677, December 14, 2010). The study has been performed in accordance with the ethical standards laid down in the 1964 declaration of Helsinki and its later amendments. Patients or their relatives gave informed consent for the inclusion of their data in a registry, in accordance with the French law concerning research in humans.
Review of the Literature
We performed a literature search until December 20, 2015, with Medline since 1966 and Embase since 1980, and we also selected articles cited in the references of the retrieved articles. We first selected abstracts according to the following algorithm: (((thrombolysis [All Fields] OR (“tissue plasminogen activator”[MeSH Terms] OR (“tissue”[All Fields] AND “plasminogen”[All Fields] AND “activator”[All Fields]) OR “tissue plasminogen activator”[All Fields] OR (“rt”[All Fields] AND “pa”[All Fields]) OR “rt pa”[All Fields])) OR (“fibrinolysis”[MeSH Terms] OR “fibrinolysis”[All Fields])) AND (((“stroke”[MeSH Terms] OR “stroke”[All Fields]) OR (“cerebral ischaemia”[All Fields] OR “cerebral infarction”[MeSH Terms] OR (“cerebral”[All Fields] AND “infarction” [All Fields]) OR “cerebral infarction” [All Fields] OR (“cerebral” [All Fields] AND “ischemia” [All Fields]) OR “cerebral ischemia” [All Fields] OR “brain ischemia” [MeSH Terms] OR (“brain” [All Fields] AND “ischemia” [All Fields]) OR “brain ischemia” [All Fields] OR (“cerebral” [All Fields] AND “ischemia” [All Fields]))) OR (“brain ischaemia” [All Fields] OR “brain ischemia” [MeSH Terms] OR (“brain” [All Fields] AND “ischemia” [All Fields]) OR “brain ischemia” [All Fields]))) AND ((“angioedema” [MeSH Terms] OR “angioedema” [All Fields]) OR angioneurotic [All Fields]). We did not make any restriction on the language. In case of redundancy, we considered the most recent publication. We excluded from the analysis articles reporting single cases or series of cases without any control group with patients treated with intravenous r-tPA for cerebral ischemia who did not develop OLAE.
Analysis of the Lille Database
During the study period, we treated 923 consecutive patients with intravenous r-tPA for cerebral ischemia. Their demographic and baseline characteristics, their outcomes, and the number of missing data for each variable are detailed in Table I in the online-only Data Supplement. Missing data were replaced by the median value of the whole cohort, in 21 patients (2.3%) for onset-to-needle time, in 6 (0.7%) for weight and dose of r-tPA, in 4 (0.4%) for platelet count, in 3 (0.3%) for blood pressure, and in 1 (0.1%) for serum glucose level.
Characteristics of OLAE
Twenty patients (2.2%) developed an OLAE. In 18 patients, the OLAE started in the side opposite to the brain lesion, and in 2, the edema was not clearly lateralized at time of diagnosis. The OLAE started from 15 to 105 minutes after the bolus of r-tPA, with mean of 70 minutes. None of the 20 patients needed intubation or epinephrine, and the infusion of r-tPA was not stopped in any patient. The delay between onset of OLAE and complete recovery was not reliable in 17 patients, because of difficulty to identify exactly the time of recovery. In 3 patients in whom it was considered reliable, this delay was of 90 minutes in 2 patients and 225 minutes in 1 patient.
Comparison of Baseline Characteristics Between Patients With and Without OLAE
The comparison of baseline characteristics between patients with and without OLAE are detailed in Table. Patients with OLAE were more likely to be under angiotensin-converting enzyme inhibitors (ACE-I; OR, 3.8; 95% CI, 1.6–9.3 and adjOR, 3.9; 95% CI, 1.6–9.7) and to have atherothrombotic strokes (OR, 2.7; 95% CI, 1.0–7.7 and adjOR, 2.9; 95% CI, 1.03–8.4). They were also more likely to develop s-ICH according to the NINDS (OR, 3.2; 95% CI, 1.2–8.5) but not to the ECASS 2 (OR, 2.3; 95% CI, 0.7–8.0) criteria. None of the patients with OLAE were under angiotensin receptor blockers. The overall results concerning the influence of ACE-I were not modified after adjustment for propensity scores (OR, 4.4; 95% CI, 1.6–11.9; P=0.004) or matched analysis based on propensity scores (OR, 3.4; 95% CI, 1.3–8.1; P=0.010). We could not perform a stratification per quintiles of propensity scores because of the small number of events per quintile (mean=4 events per quintile).
Patients with and without OLAE did not differ for other baseline characteristics. There was a tendency for patients who were smokers, who received higher doses of r-tPA, and who developed a malignant infarct, to develop more often OLAE (0.10>P values>0.05).
The comparison of the frequency of insular infarcts was conducted in a subgroup of 19 of the 20 patients with OLAE who underwent magnetic resonance imaging at baseline (95% of all patients with OLAE) and 77 controls. Two patients selected as controls could not be separated on the basis of sex, age, weight, and baseline NIHSS score. Therefore, both were maintained in the control group. In this sub analysis, the 19 patients with OLAE did not differ from those without for age (P=0.590), sex (P=0.760), weight (P=0.963), dose of r-tPA (P=0.893), baseline NIHSS score (P=0.905), and arterial territory (P=1.000). The presence of insular infarcts of any side and any size was not associated with the occurrence of OLAE (11/19 versus 37/77, P=0.442). Patients with OLAE were more likely to have a total insular infarct than those without (7/19 versus 8/77; OR, 5.0; 95% CI, 1.5–16.5; P=0.004), but they did not differ for partial infarcts (4/19 versus 29/77; OR, 0.4; 95% CI, 0.1–1.5; P=0.172), and for the side of infarct (right 4/19 versus 20/57; OR for right infarcts, 1.9; 95% CI, 0.7–5.6; P=0.231 and left 4/19 versus 20/57; OR for left infarcts, 0.8; 95% CI, 0.2–2.6; P=0.657).
Comparison of Outcomes Between Patients With and Without OLAE
Patients with OLAE did not significantly differ at 3 months for the proportion of patients with mRS score of 0 to 1 (8/20 versus 396/903; adjOR, 0.9; 95% CI, 0.3–2.1), mRS score of 0 to 2 (10/20 versus 514/903; adjOR, 0.8; 95% CI, 0.1–1.8), and death (3/20 versus 134/903; adjOR, 1.1; 95% CI, 0.3–3.8).
Evolution of Systolic Blood Pressure in Patients With OLAE
The evolution over time of systolic blood pressure in patients with and without OLAE did not show any significant difference at any of the 16 measures between baseline and 360 minutes after bolus.
Review of the Literature
The search strategy led to the selection of 50 abstracts. We added 3 abstracts from the reference lists of these articles. After reading of these 53 articles, 18 were selected for the combined analysis.11,13–29 Seventeen articles were published in English11,13–17,19–29 and 1 in German.18 The flow chart of the selection of articles is presented in Figure.
None of these 18 articles11,13–29 reported information on the prestroke mRS, alcohol consumption, weight, NIHSS at 2 and 24 hours and at 7 days, presence of infarct on imaging at 22 to 36 hours, s-ICH, infection, malignant infarct, and outcome at 3 months. The incidence of OLAE in these 18 previous studies and in our study, and their combined analysis, are detailed in Table II in the online-only Data Supplement. The overall incidence reported in previous articles was 1.8% (95% CI, 1.6%–2.1%) before inclusion of our study and increased ≤1.9% (95% CI, 1.6%–2.1%) after.
Thirteen studies11,14,17–19,21–28 did not compare any baseline characteristics between patients with and without OLAE and could not be included in a combined analysis of potential predisposing factors. Three studies could be used only for a combined analysis on ACE-I.15,16,20 Only 2 studies13,29 could be included with ours in a combined analysis for several items. Patients with OLAE were more likely to be under ACE-I before thrombolysis (OR, 8.6; 95% CI, 5.5–13.5) and to have arterial hypertension (OR, 2.5; 95% CI, 1.4–4.4), diabetes mellitus (OR, 1.8; 95% CI, 1.01–3.3), and hypercholesterolemia (OR, 2.0; 95% CI, 1.2–3.2; Table III in the online-only Data Supplement). The presence of insular infarcts was described in several studies, but without any comparison between patients with and without OLAE. None of these studies have evaluated the outcome at 3 months.
Our study has shown that (1) ≈2% of patients treated with intravenous r-tPA for cerebral ischemia developed OLAE; (2) OLAE occurs during thrombolysis in half of the patients, and soon after the end of infusion in the other half; (3) most OLAE start on the side opposite to the lesion; (4) OLAE are more likely to occur in patients with ACE-I and total insular infarcts; (5) depending on the definition used, patients with OLAE have—or tend to have—more s-ICH, that is not explained by significant variations in systolic blood pressure; and (6) although OLAE may be life threatening, it does not influence the 3-month outcome.
The strengths of our study are the large sample size, the absence of patients lost to follow-up, and a prospective collection of data with few missing data. This is the first study that compared the prevalence of insular lesions and the influence of their size with that of controls, evaluated the time course of blood pressure, and evaluated the influence of OLAE on the 3-month outcome.
Our study has also limitations. Our protocol of screening by nurses stops after 90 minutes, and we cannot exclude that mild cases may have occurred after this delay. The number of cases reported in the literature is still of limited size, with a total of <200 cases reported in series that included controls. Another limitation is that we did not determine C-1 esterase levels, because of the observational design of the study.
The incidence of OLAE in our study was slightly higher than that reported in most previous studies,11,13–28 except that of Hurford et al,29 probably because of the systematic evaluation by nurses during the first 90 minutes, allowing the detection of mild cases that would have remained undetected otherwise.
R-tPA used in vitro at a therapeutic concentration generates significant quantities of kinins from human plasma.30 OLAE associated with r-tPA treatment results directly from plasmin-mediated release of bradykinin.30 Our study and the results of the combined analysis strongly support the hypothesis that patients under ACE-I are at risk for OLAE during thrombolysis. Bradykinin is converted to inactive metabolites by ACE, and the inhibition of this enzyme leads to increased levels of bradykinin.30 We did not use icatibant, a selective bradykinin B2 receptor antagonist indicated for treatment of acute attacks of hereditary angioedema, but this could be an option.
Our study also supports the hypothesis that patients with total insular infarcts are at risk for OLAE during thrombolysis. A possible mechanism is that insular infarcts might induce autonomic dysregulation. The brain tissue damage itself could be a trigger for OLAE via the generation of bradykinine.31 This hypothesis is in line with the evidence in a single-center study of 1 case of OLAE out of 24 patients who were treated by intravenous r-tPA for the second time and did not develop OLAE at the first thrombolysis.32 This hypothesis is also supported by the rare occurrence of OLAE in patients who received intravenous r-tPA for other indications than stroke, including stroke mimics.6 The absence of association between OLAE and partial insular infarcts may be explained by the necessity to reach a certain threshold of lesion volume in the insula to generate autonomic dysregulation.
We also found a statistical association in the meta-analysis between the occurrence of OLAE and preexisting arterial hypertension and at a lower degree with diabetes mellitus and hypercholesterolemia. In the absence of individual data, we could not adjust on the presence of ACE-I, and it is likely that these associations are found because these patients are more often under ACE-I than the general population.
We found that patients with OLAE were more likely to develop s-ICH according to the NINDS1 criteria and tended to with the ECASS 2 criteria.6 This association between OLAE and s-ICH cannot be explained by differences in the volume of brain lesions because the baseline stroke severity did not differ between patients with and without OLAE. Our study does not support the hypothesis that this increased bleeding risk is the consequence of increased systolic blood pressure when OLAE occurs, because of the absence of difference between patients with and without OLAE. We cannot exclude an interference with the treatments administered for the OLAE, slight blood pressure variations that cannot be identified with this sample size, or just a relationship with the location of the infarct in an area more prone to bleeding. Finally, a chance finding cannot be excluded, because of the small number of OLAE and of sICH.
OLAE is a rare complication that does not influence outcome if detected and properly treated early enough. They occur in only 1 of 50 patients treated by intravenous thrombolysis for stroke (95% CI, 1 of 47 to 1 of 62), but this proportion increases to 1 of 10 in case of total insular infarct, and 1 of 6 in case of on-going treatment with ACE-I. In survivors, the long-term outcome is similar to that of other patients. The frequency of OLAE being low, our study can only indicate a trend in terms of relation with the outcome.
Sources of Funding
This study was supported by University Lille 2, Institut national de la santé et de la recherche médicale U1171. Dr Myslimi was supported partly by a department-to-department grant of the European Academy of Neurology.
Drs Dequatre-Ponchelle, Moulin, Cordonnier, and Leys are investigators of ECASS 3 and ECASS 4 (no personal funding, fees given to ADRINORD). Dr Leys has been a member of scientific advisory boards and has participated to symposia organized by Boehringer- Ingelheim (no personal funding, fees given to ADRINORD). The other authors report no conflicts.
The online-only Data Supplement is available with this article at http://stroke.ahajournals.org/lookup/suppl/doi:10.1161/STROKEAHA.116.013334/-/DC1.
- Received March 2, 2016.
- Revision received April 25, 2016.
- Accepted April 27, 2016.
- © 2016 American Heart Association, Inc.
- Hacke W,
- Kaste M,
- Fieschi C,
- von Kummer R,
- Davalos A,
- Meier D,
- et al
- Lyden P,
- Brott T,
- Tilley B,
- Welch KM,
- Mascha EJ,
- Levine S,
- et al
- Adams HP Jr,
- Bendixen BH,
- Kappelle LJ,
- Biller J,
- Love BB,
- Gordon DL,
- et al
- van Swieten JC,
- Koudstaal PJ,
- Visser MC,
- Schouten HJ,
- van Gijn J
- Zinkstok SM,
- Engelter ST,
- Gensicke H,
- Lyrer PA,
- Ringleb PA,
- Artto V,
- et al
- Wallon D,
- Girardie P,
- Bombois S,
- Lucas C
- Hill MD,
- Buchan AM
- Winkler DT,
- Fluri F,
- Fuhr P,
- Wetzel SG,
- Lyrer PA,
- Ruegg S,
- et al
- Tsivgoulis G,
- Alexandrov AV,
- Chang J,
- Sharma VK,
- Hoover SL,
- Lao AY,
- et al
- Hurford R,
- Rezvani S,
- Kreimei M,
- Herbert A,
- Vail A,
- Parry-Jones AR,
- et al
- Molinaro G,
- Gervais N,
- Adam A
- Gröger M,
- Lebesgue D,
- Pruneau D,
- Relton J,
- Kim SW,
- Nussberger J,
- et al