Impaired cerebrovascular reactivity may predict delayed cerebral ischemia after aneurysmal subarachnoid hemorrhage

Open AccessPublished:October 17, 2019DOI:https://doi.org/10.1016/j.jns.2019.116539

      Highlights

      • Impaired cerebrovascular reactivity (CVR) may predict delayed cerebral ischemia (DCI).
      • CVR is reduced on the ipsilateral side after aneurysm treatment, regardless of DCI development.
      • Ipsi- and contralateral CVR need to be considered separately.
      • Contralateral CVR seems best suited as predictor of delayed cerebral ischemia.

      Abstract

      Introduction

      Delayed cerebral ischemia (DCI) is a major cause of disability and death after aneurysmal subarachnoid hemorrhage. The literature suggests that impaired cerebrovascular reactivity (CVR) may be a predictor for DCI; still no CVR based prediction model has been developed. Increased knowledge about possible predictors of DCI can improve patient management in high-risk patients and allow for shorter hospital stay in low-risk patients.

      Method

      CVR was examined in 42 patients with aneurysmal subarachnoid hemorrhage and 37 patients treated for unruptured intracranial aneurysm, using acetazolamide test with transcranial Doppler monitoring of blood flow velocities. Patients were followed for development of DCI, separated into clinical deterioration and radiographic infarction.

      Results

      For all patients, regardless of aneurysm rupture status, CVR was on average 5.5 percentage points lower on the ipsilateral side of aneurysm treatment. Patients with clinical deterioration due to DCI had lower CVR than patients without DCI, and the difference was larger on the contralateral side (33.9% vs. 49.2%). Two prediction models were constructed for clinical deterioration due to DCI. The area under the receiver operating characteristic curve was 0.82 in the model using established predictors, and 0.86 in the model that also included CVR.

      Conclusion

      Our findings support the hypothesis that impaired CVR may be an independent predictor of clinical deterioration due to DCI, and may assist in identifying patients at risk after aneurysmal subarachnoid hemorrhage. Ipsilateral CVR reduction occurs in all patients after aneurysm treatment, regardless of DCI development, thus highlighting the need to evaluate ipsi- and contralateral CVR separately.

      Keywords

      1. Introduction

      Delayed cerebral ischemia (DCI) is a major cause of disability and death after aneurysmal subarachnoid hemorrhage (aSAH) [
      • Roos Y.B.
      • de Haan R.J.
      • Beenen L.F.
      • et al.
      Complications and outcome in patients with aneurysmal subarachnoid haemorrhage: a prospective hospital based cohort study in the Netherlands.
      ,
      • Kassell N.F.
      • Torner J.C.
      • Haley Jr., E.C.
      • et al.
      The international cooperative study on the timing of aneurysm surgery. Part 1: overall management results.
      ]. Identification of patients at high risk of developing DCI can improve patient management, and valid predictors of DCI could allow for shorter hospital stay in patients at low risk. Large amount of subarachnoid blood and poor clinical admission status are known predictors for DCI [
      • Adams Jr., H.P.
      • Kassell N.F.
      • Torner J.C.
      • et al.
      Predicting cerebral ischemia after aneurysmal subarachnoid hemorrhage: influences of clinical condition, CT results, and antifibrinolytic therapy. A report of the Cooperative Aneurysm Study.
      ,
      • Fisher C.M.
      • Kistler J.P.
      • Davis J.M.
      Relation of cerebral vasospasm to subarachnoid hemorrhage visualized by computerized tomographic scanning.
      ,
      • van Gijn J.
      • Kerr R.S.
      • Rinkel G.J.
      Subarachnoid haemorrhage.
      ,
      • de Oliveira Manoel A.L.
      • Jaja B.N.
      • Germans M.R.
      • et al.
      The VASOGRADE: a simple grading scale for prediction of delayed cerebral ischemia after subarachnoid hemorrhage.
      ,
      • Crobeddu E.
      • Mittal M.K.
      • Dupont S.
      • et al.
      Predicting the lack of development of delayed cerebral ischemia after aneurysmal subarachnoid hemorrhage.
      ]. Other predictors, like smoking and hydrocephalus have been suggested [
      • de Rooij N.K.
      • Rinkel G.J.
      • Dankbaar J.W.
      • et al.
      Delayed cerebral ischemia after subarachnoid hemorrhage: a systematic review of clinical, laboratory, and radiological predictors.
      ]. Still, additional information is needed to make prediction more accurate.
      Cerebrovascular reactivity (CVR) is often impaired in the early phase after aSAH [
      • Dernbach P.D.
      • Little J.R.
      • Jones S.C.
      • et al.
      Altered cerebral autoregulation and CO2 reactivity after aneurysmal subarachnoid hemorrhage.
      ,
      • Seiler R.W.
      • Nirkko A.C.
      Effect of nimodipine on cerebrovascular response to CO2 in asymptomatic individuals and patients with subarachnoid hemorrhage: a transcranial Doppler ultrasound study.
      ,
      • Abe K.
      • Nishimura M.
      • Yoshiya I.
      Local cerebral blood flow and CO2 reactivity during prostaglandin E1-induced hypotension in patients undergoing cerebral aneurysm surgery.
      ,
      • Abe K.
      • Demizu A.
      • Mima T.
      • et al.
      Carbon dioxide reactivity during prostaglandin E1 induced hypotension for cerebral aneurysm surgery.
      ,
      • Shinoda J.
      • Kimura T.
      • Funakoshi T.
      • et al.
      Acetazolamide reactivity on cerebral blood flow in patients with subarachnoid haemorrhage.
      ,
      • da Costa L.
      • Houlden D.
      • Rubenfeld G.
      • et al.
      Impaired cerebrovascular reactivity in the early phase of subarachnoid hemorrhage in good clinical grade patients does not predict vasospasm.
      ,
      • Tanaka A.
      • Yoshinaga S.
      • Nakayama Y.
      • et al.
      Cerebral blood flow and the response to acetazolamide during the acute, subacute, and chronic stages of aneurysmal subarachnoid hemorrhage.
      ], especially in patients with poor clinical grade [
      • Abe K.
      • Demizu A.
      • Mima T.
      • et al.
      Carbon dioxide reactivity during prostaglandin E1 induced hypotension for cerebral aneurysm surgery.
      ,
      • Carrera E.
      • Kurtz P.
      • Badjatia N.
      • et al.
      Cerebrovascular carbon dioxide reactivity and delayed cerebral ischemia after subarachnoid hemorrhage.
      ,
      • Ishii R.
      Regional cerebral blood flow in patients with ruptured intracranial aneurysms.
      ,
      • Abe K.
      • Demizu A.
      • Kamada K.
      • et al.
      Prostaglandin E1 and carbon dioxide reactivity during cerebral aneurysm surgery.
      ]. Several test methods have shown reduced CVR in patients with aSAH compared with unruptured intracranial aneurysms (UIAs) [
      • Dernbach P.D.
      • Little J.R.
      • Jones S.C.
      • et al.
      Altered cerebral autoregulation and CO2 reactivity after aneurysmal subarachnoid hemorrhage.
      ,
      • Seiler R.W.
      • Nirkko A.C.
      Effect of nimodipine on cerebrovascular response to CO2 in asymptomatic individuals and patients with subarachnoid hemorrhage: a transcranial Doppler ultrasound study.
      ,
      • Abe K.
      • Nishimura M.
      • Yoshiya I.
      Local cerebral blood flow and CO2 reactivity during prostaglandin E1-induced hypotension in patients undergoing cerebral aneurysm surgery.
      ,
      • Abe K.
      • Demizu A.
      • Mima T.
      • et al.
      Carbon dioxide reactivity during prostaglandin E1 induced hypotension for cerebral aneurysm surgery.
      ,
      • Schmieder K.
      • Jarus-Dziedzic K.
      • Wronski J.
      • et al.
      CO2 reactivity in patients after subarachnoid haemorrhage.
      ], but this has not been confirmed for the transcranial Doppler (TCD) and acetazolamide (AZ) test (TCD-AZ test) [
      • Bothun M.L.
      • Haaland O.A.
      • Logallo N.
      • et al.
      Cerebrovascular reactivity after treatment of unruptured intracranial aneurysms - a transcranial Doppler sonography and acetazolamide study.
      ]. With one exception [
      • da Costa L.
      • Houlden D.
      • Rubenfeld G.
      • et al.
      Impaired cerebrovascular reactivity in the early phase of subarachnoid hemorrhage in good clinical grade patients does not predict vasospasm.
      ], the literature suggests that impaired CVR may be associated with vasospasm and can be a potential predictor for DCI after aSAH [
      • Tanaka A.
      • Yoshinaga S.
      • Nakayama Y.
      • et al.
      Cerebral blood flow and the response to acetazolamide during the acute, subacute, and chronic stages of aneurysmal subarachnoid hemorrhage.
      ,
      • Carrera E.
      • Kurtz P.
      • Badjatia N.
      • et al.
      Cerebrovascular carbon dioxide reactivity and delayed cerebral ischemia after subarachnoid hemorrhage.
      ,
      • Ishii R.
      Regional cerebral blood flow in patients with ruptured intracranial aneurysms.
      ,
      • Kimura T.
      • Shinoda J.
      • Funakoshi T.
      Prediction of cerebral infarction due to vasospasm following aneurysmal subarachnoid haemorrhage using acetazolamide-activated 123I-IMP SPECT.
      ,
      • Frontera J.A.
      • Rundek T.
      • Schmidt J.M.
      • et al.
      Cerebrovascular reactivity and vasospasm after subarachnoid hemorrhage: a pilot study.
      ,
      • da Costa L.
      • Fisher J.
      • Mikulis D.J.
      • et al.
      Early identification of brain tissue at risk for delayed cerebral ischemia after aneurysmal subarachnoid hemorrhage.
      ,
      • Voldby B.
      • Enevoldsen E.M.
      • Jensen F.T.
      Cerebrovascular reactivity in patients with ruptured intracranial aneurysms.
      ,
      • Nakagawara J.
      • Wada K.
      • Takeda R.
      • et al.
      Prediction of cerebral ischemia due to cerebral vasospasm in SAH using SPECT and 123I-IMP with acetazolamide test.
      ,
      • Nakagawara J.
      • Wada K.
      • Takeda R.
      • et al.
      Incidence of clinically significant cerebral vasospasm in SAH estimated by local cerebral vasodilatory capacity using acetazolamide-activated 123I-IMP SPECT.
      ,
      • Messeter K.
      • Brandt L.
      • Ljunggren B.
      • et al.
      Prediction and prevention of delayed ischemic dysfunction after aneurysmal subarachnoid hemorrhage and early operation.
      ,
      • Mourier K.L.
      • George B.
      • Raggueneau J.L.
      • et al.
      Value of the measurement of cerebral blood flow before and after diamox injection in predicting clinical vasospasm and final outcome in aneurysmal subarachnoid hemorrhage.
      ,
      • Tran Dinh Y.R.
      • Lot G.
      • Benrabah R.
      • et al.
      Abnormal cerebral vasodilation in aneurysmal subarachnoid hemorrhage: use of serial 133Xe cerebral blood flow measurement plus acetazolamide to assess cerebral vasospasm.
      ,
      • Hassler W.
      • Chioffi F.
      CO2 reactivity of cerebral vasospasm after aneurysmal subarachnoid haemorrhage.
      ]. However, sample sizes have been limited, methodology has varied and inconsistent and outdated definitions of DCI have been used. To our knowledge, no CVR based prediction model has yet been developed.
      The main objective of this study was to assess if impaired CVR can be a predictor for DCI after aSAH. In adjunct, we wanted to assess the relationship between aneurysm rupture status and CVR assessed by TCD-AZ test, in order to improve understanding of CVR in this patient group.

      2. Methods

      2.1 Participants

      Patients treated at the Department of Neurosurgery, Haukeland University Hospital between February 2011 and May 2013 were prospectively included. Inclusion criteria were: age ≥ 18 years and treatment for saccular intracranial aneurysms with endovascular coiling or surgical clipping. Exclusion criteria were: previous treatment of intracranial aneurysms, giant aneurysms treated with proximal artery occlusion, carotid stenosis (>50%) or occlusion, lack of transtemporal bone window in TCD examination, DCI present at admission, moribund patients, and contraindications to acetazolamide.

      2.1.1 Clinical, sonographic and radiographic assessment

      In all patients demographics, body weight, smoking status, hypertension, aneurysm location, and treatment modality was recorded. In patients with ruptured aneurysms, the clinical status upon admission was evaluated with Glasgow Coma Scale (GCS) [
      • Teasdale G.
      • Jennett B.
      Assessment of coma and impaired consciousness. A practical scale.
      ] and World Federation of Neurological Surgeons (WFNS) scale [
      • Report of World Federation of Neurological Surgeons Committee on a Universal Subarachnoid Hemorrhage Grading Scale
      ]. Admission computed tomographic (CT) scan was classified according to the modified CT Fisher scale [
      • Frontera J.A.
      • Claassen J.
      • Schmidt J.M.
      • et al.
      Prediction of symptomatic vasospasm after subarachnoid hemorrhage: the modified fisher scale.
      ].
      Patients were monitored clinically, sonographically and radiographically for development of DCI, vasospasm and cerebral infarction. Neurological status was scored routinely three times a day, and more frequent as regarded necessary. As a pragmatic choice, sonographic examinations were performed whenever possible, in order to avoid conflicts with patient investigations or treatment. Blood flow velocities in major intracranial arteries were assessed with transcranial color-coded sonography (TCCS) using a portable Philips CX50 system (5–1 MHz sector array probe) and Lindegaard Index was calculated [
      • Lindegaard K.F.
      • Nornes H.
      • Bakke S.J.
      • et al.
      Cerebral vasospasm diagnosis by means of angiography and blood velocity measurements.
      ]. TCCS was performed daily and more frequent if neurological deterioration occurred. Neuroimaging, i.e. CT or magnetic resonance imaging (MRI) with or without angiography and digital subtraction angiography (DSA) was performed tailored to the clinical situation. An experienced neuroradiologist (GM) assessed all angiograms, CT and MRI scans retrospectively for vasospasm and cerebral infarctions.

      2.1.2 Cerebrovascular reactivity

      CVR testing was performed with acetazolamide (AZ). The AZ dose was 1000 mg for patients weighing <80 kg, and 15 mg/kg for patients weighing ≥80 kg. The maximum dose was 1500 mg. CVR testing was performed after aneurysm treatment to avoid risk of (re-)rupture. Patients treated for unruptured intracranial aneurysms (UIA) were examined once, whereas patients treated for aSAH underwent serial testing during the hospital stay. The time interval between examinations was minimum 24 h. To avoid steal phenomenon and neurological deterioration, CVR testing was not performed after a patient was diagnosed with DCI. TCD was used to monitor blood flow velocities in the middle cerebral arteries (MCA) after intravenous injection of acetazolamide (AZ) [
      • Bothun M.L.
      • Haaland O.A.
      • Logallo N.
      • et al.
      Cerebrovascular reactivity after treatment of unruptured intracranial aneurysms - a transcranial Doppler sonography and acetazolamide study.
      ]. Cerebrovascular reactivity was calculated as the maximum percentage change in MFV in MCA after administration of acetazolamide: CVR (%) = [(MFVAZ –MFVBASELINE) / MFVBASELINE] x 100, where MFVBASELINE is mean blood flow velocity before acetazolamide and MFVAZ is mean blood flow velocity (maximum change) after acetazolamide. In patients with a paradoxical velocity reduction after AZ due to a steal phenomenon, CVR will be a negative value.

      2.1.3 Clinical deterioration and cerebral infarction due to delayed cerebral ischemia

      All patients with aSAH, regardless of presence of ischemic symptoms or not, were given 60 mg nimodipine orally every four hours to prevent DCI [
      • Steiner T.
      • Juvela S.
      • Unterberg A.
      • et al.
      European Stroke Organization guidelines for the management of intracranial aneurysms and subarachnoid haemorrhage.
      ,
      • Dorhout Mees S.M.
      • Rinkel G.J.
      • Feigin V.L.
      • et al.
      Calcium antagonists for aneurysmal subarachnoid haemorrhage.
      ]. Critically ill patients and patients with swallowing difficulties were given nimodipine as intravenous infusion 2 mg/h. Patients with severe and refractory radiographic vasospasm and clinical deterioration due to DCI were given intraarterial nimodipine [
      • Biondi A.
      • Ricciardi G.K.
      • Puybasset L.
      • et al.
      Intra-arterial nimodipine for the treatment of symptomatic cerebral vasospasm after aneurysmal subarachnoid hemorrhage: preliminary results.
      ]. Patient management was not influenced by CVR results.
      We distinguished between clinical deterioration and radiographic infarction due to DCI [
      • Vergouwen M.D.
      • Vermeulen M.
      • van Gijn J.
      • et al.
      Definition of delayed cerebral ischemia after aneurysmal subarachnoid hemorrhage as an outcome event in clinical trials and observational studies: proposal of a multidisciplinary research group.
      ,
      • Vergouwen MD and Participants in the International Multi-Disciplinary Consensus Conference on the Critical Care Management of Subarachnoid H
      Vasospasm versus delayed cerebral ischemia as an outcome event in clinical trials and observational studies.
      ].
      Clinical deterioration (DCIclinical) was defined as a new focal neurological impairment or ≥ 2 points reduction in Glasgow Coma Score, lasting for minimum 1 h, and not appearing immediately after aneurysm occlusion. Other causes of deterioration were excluded by clinical assessment, cerebral CT or MRI, and laboratory analyses.
      Radiographic infarction (DCIinfarction) was defined as a new infarction identified on CT or MR scans during the hospital stay, within six weeks after aSAH. Infarctions present on the admission or immediate postoperative CT, and hypodensities resulting from the clipping or coiling procedure, ventricular catheter placement or intraparenchymal hematoma were not regarded as cerebral infarctions from DCI.
      Angiographic vasospasm was defined as arterial narrowing present on CTA, MRA or DSA, not attributable to atherosclerosis, catheter-induced spasm, or vessel hypoplasia. Angiographic vasospasm was classified as none, mild (<33% reduction in arterial diameter), moderate (34–66%) or severe (≥67%), and was also categorized with the Nathal grading scale [
      • Nathal E.
      • Lopez-Gonzalez F.
      • Rios C.
      Angiographic scale for evaluation of cerebral vasospasm.
      ].
      Sonographic vasospasm was defined as time-averaged peak velocity (TAPV) ≥120 cm/s in any cerebral artery and Lindegaard index ≥3 [
      • Lindegaard K.F.
      • Nornes H.
      • Bakke S.J.
      • et al.
      Cerebral vasospasm diagnosis by means of angiography and blood velocity measurements.
      ]. Severe sonographic vasospasm was defined as TAPV ≥200 cm/s and Lindegaard index ≥6.

      2.1.4 Ethics

      The study was conducted in accordance with the Declaration of Helsinki (2013) of the World Medical Association, and was approved by the local ethics committee (the Regional Committees of Western Norway for Medical and Health Research Ethics, approval number 2011/144). Written informed consent was obtained from all patients or a legal representative.

      2.2 Statistical analysis

      We reported CVR separately as ipsilateral and contralateral values to avoid that mean values could obscure side differences [
      • Bothun M.L.
      • Haaland O.A.
      • Logallo N.
      • et al.
      Time course of cerebrovascular reactivity in patients treated for unruptured intracranial aneurysms: a one-year transcranial doppler and acetazolamide follow-up study.
      ]. Due to the wide range of reported CVR measured by TCD and AZ in healthy subjects [
      • Sorteberg W.
      • Lindegaard K.F.
      • Rootwelt K.
      • et al.
      Effect of acetazolamide on cerebral artery blood velocity and regional cerebral blood flow in normal subjects.
      ,
      • Piepgras A.
      • Schmiedek P.
      • Leinsinger G.
      • et al.
      A simple test to assess cerebrovascular reserve capacity using transcranial Doppler sonography and acetazolamide.
      ,
      • Mancini M.
      • De Chiara S.
      • Postiglione A.
      • et al.
      Transcranial Doppler evaluation of cerebrovascular reactivity to acetazolamide in normal subjects.
      ,
      • Dahl A.
      • Russell D.
      • Nyberg-Hansen R.
      • et al.
      Simultaneous assessment of vasoreactivity using transcranial Doppler ultrasound and cerebral blood flow in healthy subjects.
      ,
      • Dahl A.
      • Russell D.
      • Rootwelt K.
      • et al.
      Cerebral vasoreactivity assessed with transcranial Doppler and regional cerebral blood flow measurements. Dose, serum concentration, and time course of the response to acetazolamide.
      ,
      • Ulrich P.T.
      • Becker T.
      • Kempski O.S.
      Correlation of cerebral blood flow and MCA flow velocity measured in healthy volunteers during acetazolamide and CO2 stimulation.
      ,
      • Hamann G.F.
      • Stoll M.
      • Jost V.
      • et al.
      Time course of acetazolamide effect in normal persons.
      ,
      • Karnik R.
      • Valentin A.
      • Winkler W.B.
      • et al.
      Sex-related differences in acetazolamide-induced cerebral vasomotor reactivity.
      ,
      • Patrick J.T.
      • Fritz J.V.
      • Adamo J.M.
      • et al.
      Phase-contrast magnetic resonance angiography for the determination of cerebrovascular reserve.
      ,
      • Valikovics A.
      • Olah L.
      • Fulesdi B.
      • et al.
      Cerebrovascular reactivity measured by transcranial Doppler in migraine.
      ,
      • Fulesdi B.
      • Limburg M.
      • Bereczki D.
      • et al.
      Impairment of cerebrovascular reactivity in long-term type 1 diabetes.
      ,
      • Olah L.
      • Valikovics A.
      • Bereczki D.
      • et al.
      Gender-related differences in acetazolamide-induced cerebral vasodilatory response: a transcranial Doppler study.
      ,
      • Schwertfeger N.
      • Neu P.
      • Schlattmann P.
      • et al.
      Cerebrovascular reactivity over time course in healthy subjects.
      ] and patients with UIA [
      • Bothun M.L.
      • Haaland O.A.
      • Logallo N.
      • et al.
      Cerebrovascular reactivity after treatment of unruptured intracranial aneurysms - a transcranial Doppler sonography and acetazolamide study.
      ], we studied CVR as a continuous variable. Effects are reported per percentage point change in CVR. Midline aneurysms were allocated according to chosen side of approach for treatment. Standard t-tests and multiple regression analyses (adjusted for age, sex, hypertension, smoking, aneurysm diameter and treatment modality) were carried out to examine the relationship between CVR and rupture status.
      We then compared clinical, radiographical and sonographical variables of patients with and without DCIclinical, with chi-squared or Fisher's exact test, as appropriate. Next, we assessed the predictive potential of both contralateral and ipsilateral CVR for DCIclinical and DCIinfarction, to decide which measure of CVR was most appropriate to use in patients with multiple measurement (first, lowest, or within-patient average of all exams). A priori, we chose the lowest or first measured CVR as the exposure (more convenient in clinical use). t-test was used to compare CVR in aSAH patients with and without DCIclinical or DCIinfarction, and to assess side differences. Simple logistic regression analyses were conducted for both DCIclinical and DCIinfarction versus both within-patient average CVR and lowest measured contralateral CVR. Further, box plots were constructed, based on the lowest measured iCVR and cCVR for UIA, aSAH without DCIclinical, and aSAH with DCIclinical.
      From this, two prediction models for DCIclinical were created. Model I included WFNS, modified CT Fisher scale, age, sex, smoking, and hydrocephalus, all variables suspected to affect risk of DCIclinical. Model II included all variables from Model I plus contralateral CVR. Model discrimination was assessed with area under the ROC curve (AUC), and model calibration with Hosmer-Lemeshow's C and calibration plots [
      • Steyerberg E.W.
      • Vickers A.J.
      • Cook N.R.
      • et al.
      Assessing the performance of prediction models: a framework for traditional and novel measures.
      ]. Results were adjusted for optimism using the bootstrapping techniques [
      • Harrell F.
      Regression Modeling Strategies With Applications to Linear Models, Logistic and Ordinal Regression, and Survival Analysis.
      ]. Uniformity of fit was evaluated across clinical grade (WFNS I-III vs. IV-V).
      Statistical analysis was performed with R version 3.4.3 [
      • R Core Team
      R: A Language and Environment for Statistical Computing.
      ]. The data that support the findings of this study are available from the corresponding author upon reasonable request.

      3. Results

      3.1 Patients, aneurysm and treatment

      An overview of the source population is available in the online supplement (Table S1). Of 136 patients consecutively recruited during the study period, 57 were excluded, leaving a study population of 79 patients. In total, 37 patients were treated for UIA and 42 patients had aSAH (Table 1).
      Table 1Patient, aneurysm and treatment characteristics.
      aSAH (n = 42)UIA (n = 37)
      Age, years
      Mean ± standard deviation. All other variables are reported as n (%).
      53 ± 1350 ± 11
      Height, cm
      Mean ± standard deviation. All other variables are reported as n (%).
      171 ± 9169 ± 8
      Weight, kg
      Mean ± standard deviation. All other variables are reported as n (%).
      76 ± 1877 ± 16
      BMI, kg/m2a26 ± 527 ± 5
      Female24 (57.1)24 (64.9)
      Hypertension19 (45.2)18 (48.6)
      Smoking
       Current27 (64.3)20 (54.1)
       Previous8 (19.0)13 (35.1)
       Never7 (16.7)4 (10.8)
      Multiple aneurysm9 (21.4)12 (32.4)
      Aneurysm diameter
      Mean ± standard deviation. All other variables are reported as n (%).
      Maximum diameter of the aneurysm dome. For patients with aSAH the size of the ruptured aneurysm is reported. For patients that underwent treatment for multiple UIA during the same procedure, aneurysm size was averaged.
      7 ± 37 ± 3
      Location treated aneurysms
       Middle cerebral artery12 (28.6)17 (45.9)
       ACOM, anterior complex and pericallosa17 (40.5)7 (18.9)
       ICA, incl. ophthalmic artery and PCOM6 (14.3)10 (27.0)
       Basilar top, cerebelli superior, PICA, VB, distal posterior7 (16.7)3 (8.1)
      Treatment modality
       Coil30 (71.4)22 (59.5)
       Clip12 (28.6)15 (40.5)
      Treatment side
      Twenty-four patients with midline aneurysms (ACOM and basilar top) were allocated to the chosen side of approach. One patient treated with combined clipping of an ACOM aneurysm and a right middle cerebral artery aneurysm in one procedure was allocated to the right side. The coiling procedure failed to adequately secure the aneurysm in one patient, so clipping was performed instead.
       Left21 (50.0)19 (51.4)
       Right21 (50.0)18 (48.6)
      ACOM: anterior communicating artery; aSAH: aneurysmal subarachnoid hemorrhage; BMI: body mass index; ICA: internal carotid artery; PCOM: posterior communicating artery; PICA: posterior inferior cerebellar artery; UIA: unruptured intracranial aneurysms; VB: vertebrobasilar arteries.
      a Mean ± standard deviation. All other variables are reported as n (%).
      b Maximum diameter of the aneurysm dome. For patients with aSAH the size of the ruptured aneurysm is reported. For patients that underwent treatment for multiple UIA during the same procedure, aneurysm size was averaged.
      c Twenty-four patients with midline aneurysms (ACOM and basilar top) were allocated to the chosen side of approach. One patient treated with combined clipping of an ACOM aneurysm and a right middle cerebral artery aneurysm in one procedure was allocated to the right side. The coiling procedure failed to adequately secure the aneurysm in one patient, so clipping was performed instead.

      3.2 Cerebrovascular reactivity

      In total, 101 bilateral and 20 unilateral examinations were performed. In patients with aSAH, 84 examinations were done. Median time for first exam was 3.2 (IQR 3.2; range: 1.3–15.0) days after ictus and 2.3 (IQR 2.9; range: 0.8–10.8) days after aneurysm treatment. Median time for the exam with the lowest CVR value was 5.3 (IQR 4.5; range: 1.8–20.2) days after ictus and 4.7 (IQR 3.9; range: 0.8–19.8) days after aneurysm treatment. In patients with UIA, a single examination was performed a median of 2.2 (IQR: 1.8; range: 1.1–6.2) days after aneurysm treatment.

      3.3 Ruptured versus unruptured aneurysms

      Patients with aSAH had lower CVR compared with patients with UIA (Fig. 1). There was no difference in baseline velocities (MFVBASELINE) between aSAH and UIA patients (Table 2). Increase in velocity after AZ (ΔMFVAZ) was however smaller in patients with aSAH, yielding lower both ipsilateral and contralateral CVR. For all patients, regardless of aneurysm rupture status, CVR was on average 5.5 percentage points lower on the ipsilateral side (p = .04). Regression analyses (adjustment for age, sex, hypertension, smoking, aneurysm diameter and treatment modality) confirmed that CVR was lower in patients with aSAH. Patients with UIA had a single CVR measurement, whereas aSAH patients were subject to several measurements. There are thus several approaches to calculate individual CVR: the first exam, the lowest value across exams, or the within-patient average for all exams, on both the ipsilateral and contralateral side. Using the first exam, the adjusted ipsilateral CVR was 19.6 percentage points lower in patients with aSAH (p = .005), and the contralateral CVR was 16.6 percentage points lower (p = .002). Using the lowest value across exams, the adjusted difference between the aSAH and UIA groups was 31.1 percentage points on the ipsilateral side (p < .001), and 23.6 percentage points on the contralateral side (p < .001). Using the within-patient average, CVR was 19.6 percentage points lower among patients with aSAH on the ipsilateral side (p = .002) and 14.7 percentage points lower on the contralateral side (p = .004).
      Fig. 1
      Fig. 1Box plots comparing cerebrovascular reactivity (CVR) in patients with unruptured intracranial aneurysms (UIA), aneurysmal subarachnoid hemorrhage (aSAH) without clinical deterioration due to delayed cerebral ischemia (DCI), and aSAH with clinical deterioration due to DCI. For patients with aSAH serial measurements of CVR were performed, and the lowest measured value is presented. Ipsilateral CVR are shown in dark grey boxes, and contralateral CVR in light grey boxes. Boxes extend from the 25th to the 75th percentile. Horizontal bars represent the median, and whiskers extend to the most extreme point that is <1.5 times the interquartile range from the box. Mean values are marked with white crosses, and outliers are depicted as points.
      Table 2Blood flow velocities and cerebrovascular reactivity in patients with ruptured vs. unruptured aneurysms.
      UIA (n = 37)aSAH (n = 42)
      Singel examFirst examLowest valueAverage of all examsp-values UIA vs. aSAH
      MFVBASELINE (cm/s)
       Ipsi60 ± 1457 ± 2055 ± 1961 ± 180.54, 0.25, 0.79
       Contra58 ± 1459 ± 1853 ± 1860 ± 190.79, 0.18, 0.59
      MFVAZ (cm/s)
       Ipsi94 ± 1880 ± 2876 ± 2786 ± 250.02, 0.003, 0.15
       Contra93 ± 2386 ± 2979 ± 2989 ± 290.25, 0.02, 0.42
      ΔMFVAZ (cm/s)
       Ipsi34 ± 923 ± 2118 ± 1725 ± 170.004, <0.001, 0.01
       Contra36 ± 1128 ± 1922 ± 1529 ± 160.03, <0.001, 0.03
      CVR (%)
       Ipsi59 ± 1943 ± 3533 ± 2845 ± 300.02, <0.001, 0.02
       Contra63 ± 1748 ± 2740 ± 2351 ± 250.008, <0.001, 0.01

      3.4 Delayed cerebral ischemia and cerebral infarctions

      Of 42 aSAH patients, 22 (52.4%) developed DCIclinical. Median time from ictus to diagnosis of DCIclinical was 9 days (IQR 4.8, range 5–19). Furthermore, 19 patients (45.2%) developed DCIinfarction, and five patients (11.9%) developed infarctions from other causes (Table 3). Four patients with severe, refractory radiographic vasospasm and DCIclinical were given nimodipine intrarterially as rescue therapy and survived with DCIinfarction. In all patients, angiographic vasospasm and DCIinfarction were located bilaterally or on the same side as the ruptured aneurysm. No patients had isolated vasospasm or infarctions on the contralateral side of the ruptured aneurysm. The overall prevalence of poor grade patients and thick bleedings was high; with WFNS grade IV-V in 42.9% of all patients with aSAH, and modified Fisher grade 3–4 in 66.7%. The proportion of patients with excellent clinical admission status (WFNS grade I) was lower (p = .02) and hydrocephalus was more prevalent (p = .007) in patients with DCIclinical.
      Table 3Characteristics of patients with aneurysmal subarachnoid hemorrhage.
      Total (n = 42
      Missing information regarding presence or absence of clinical DCI in one patient due to sedation
      )
      Clinical DCI (n = 22)No clinical DCI (n = 19)p-Value
      GCS on admission
      Mean ± standard deviation. All other variables are reported as n (%)
      11 ± 411 ± 412 ± 40.44
      WFNS
       I GCS 158 (19.0)1 (4.5)7 (36.8)0.02
      Due to few observations Fishers exact is used instead of chi-squared test
       II GCS 13–14, without neurological deficit14 (33.3)9 (40.9)5 (26.3)0.51
       III GCS 13–14, with neurological deficit2 (4.8)1 (4.5)1 (5.3)1.00
      Due to few observations Fishers exact is used instead of chi-squared test
       IV GCS 7–1210 (23.8)7 (31.8)3 (15.8)0.29
      Due to few observations Fishers exact is used instead of chi-squared test
       V GCS 3–68 (19.0)4 (18.2)3 (15.8)1.00
      Due to few observations Fishers exact is used instead of chi-squared test
      Modified CT Fisher
       1 Thin SAH without IVH6 (14.3)1 (4.5)5 (26.3)0.08
      Due to few observations Fishers exact is used instead of chi-squared test
       2 Thin SAH with IVH8 (19.0)4 (18.2)3 (15.8)1.00
      Due to few observations Fishers exact is used instead of chi-squared test
       3 Thick SAH without IVH15 (35.7)7 (31.8)8 (42.1)0.72
       4 Thick SAH with IVH13 (31.0)10 (45.5)3 (15.8)0.09
      Hydrocephalus (during primary stay)30 (71.4)20 (90.9)9 (47.4)0.007
      Cerebral infarction due to DCI19 (45.2)13 (59.1)5 (26.3)0.07
      Angiographic vasospasm
       None22 (52.4)6 (27.3)16 (84.2)<0.001
       Mild3 (7.1)2 (9.1)1 (5.3)1.00
      Due to few observations Fishers exact is used instead of chi-squared test
       Moderate12 (28.6)9 (40.9)2 (10.5)0.07
       Severe5 (11.9)5 (22.7)0 (0.0)0.05
      Due to few observations Fishers exact is used instead of chi-squared test
      Angiographic scale ad modum Nathal
      The vascular axes were defined as follows: a) internal carotid artery, b) middle cerebral artery, c) anterior cerebral artery, d) vertebral artery, e) basilar artery, f) posterior cerebral artery, and g) any other arterial territory.
       No vasospasm22 (52.4)6 (27.3)16 (84.2)<0.001
       One axis5 (11.9)2 (9.1)2 (10.5)1.00
      Due to few observations Fishers exact is used instead of chi-squared test
       Two axes1 (2.4)1 (4.5)0 (0.0)1.00
      Due to few observations Fishers exact is used instead of chi-squared test
       Three axes6 (14.3)5 (22.7)1 (5.3)0.19
      Due to few observations Fishers exact is used instead of chi-squared test
       Generalized/diffuse vasospasm8 (19.0)8 (36.4)0 (0.0)0.004
      Due to few observations Fishers exact is used instead of chi-squared test
      Sonographic vasospasm
      Missing sonographic information in one patient.
       None20 (48.8)1 (4.8)18 (94.7)<0.001
       Mild to moderate14 (34.1)13 (61.9)1 (5.3)<0.001
       Severe7 (17.1)7 (33.3)0 (0.0)0.009
      Due to few observations Fishers exact is used instead of chi-squared test
      aSAH: aneurysmal subarachnoid hemorrhage; CT: Computer Tomography; DCI: Delayed cerebral ischemia; GCS: Glasgow Coma Score; IVH: intraventricular hemorrhage; SAH: subarachnoid hemorrhage; WFNS: World Federation of Neurosurgical Societies score.
      a Missing information regarding presence or absence of clinical DCI in one patient due to sedation
      b Mean ± standard deviation. All other variables are reported as n (%)
      c Due to few observations Fishers exact is used instead of chi-squared test
      d The vascular axes were defined as follows: a) internal carotid artery, b) middle cerebral artery, c) anterior cerebral artery, d) vertebral artery, e) basilar artery, f) posterior cerebral artery, and g) any other arterial territory.
      e Missing sonographic information in one patient.
      The different approaches to CVR (first exam, lowest value and within-patient average) were compared for patients with and without DCIclinical or DCIinfarction (Tables S2-S3). For DCIinfarction, there was no difference in ipsilateral or contralateral CVR between groups across all approaches (Table S2). For DCIclinical (Table S3), there was no difference in ipsilateral CVR between groups. However, the lowest contralateral CVR was lower in patients with DCIclinical than without (33.9% vs. 49.2%, p = .05) and the difference in within-patient average CVR was even larger (43.2% vs. 61.9%, p = .02). Differences in first-exam CVR were less apparent (44.0% vs. 56.2%, p = .19). Mean combined values of the right and left side masked side-differences between the groups. Fig. 1 shows box plots illustrating the difference between ipsi- and contralateral sides.
      Because CVR on the ipsilateral side was similar between patients with and without DCIclinical and DCIinfarction, contralateral CVR was selected as the exposure variable in logistic regressions (Fig. 2). CVR did not predict DCIinfarction, but did predict DCIclinical. For DCIclinical, OR was 0.96 for both the lowest value (95% CI 0.93–1.00, p = .05) and within-patient average (95% CI 0.93–1.00, p = .03). This corresponds to a 4% reduction in the odds of developing DCIclinical per percentage point increase in CVR. Hence, DCIclinical risk increased with lower contralateral CVR. The risk of developing DCIclinical is 28.3% when the lowest CVR on the contralateral side is 75%, whereas the risk is 49.0% with CVR 50% and 70.1% with CVR 25%.
      Fig. 2
      Fig. 2Probabilities of developing clinical deterioration (A, B) and cerebral infarction (C, D) due to DCI for different CVR values, as predicted by logistic regression. CVR on the contralateral side of aneurysm treatment is used in the models, and the lowest measured CVR (A, C) and within-patient-average CVR of all exams is presented (B, D). The solid black line shows the estimated regression line. Uncertainty is indicated by the shadowed area (95% confidence band). Two patients, marked with a solid grey and black dot, had a paradoxical velocity reduction after AZ.
      The prediction models using lowest value and within-patient average were similar for DCIclinical, even though the p-value for the OR was slightly lower in the second model. Still, the lowest measured CVR is more convenient to use in clinical practice, and the time of the lowest CVR reflects the typical time period for onset of DCI. The lowest value on the contralateral side was therefore chosen as the CVR predictor for DCIclinical in the following analyses.
      In prediction Model I the area under the receiver operating characteristic curve (AUC) after correction for optimism was 0.82, and in Model II the corresponding AUC was 0.86 (Table 4). Confidence intervals for AUC for the two models were overlapping. Receiver operating characteristic curves and calibration plots are shown in Fig. 3. Hosmer and Lemeshow's C statistics were C = 15.33 (Model I), and C = 9.41 (Model II). Stratification for clinical grade (WFNS I-III vs. IV-V) did not cause any major changes in the prediction models.
      Table 4Multivariable predictors of clinical deterioration due to delayed cerebral ischemia after aneurysmal subarachnoid hemorrhage.
      Odds ratio (95% Confidence Interval)
      Model I (without CVR)Model II (with CVR)
      Candicate predictors
       CVR0.94 (0.89, 1.00)
       Smoking (current vs. former/never)0.33 (0.04, 2.63)0.20 (0.01, 3.2)
       Hydrocephalus0.30 (0.02, 4.40)0.06 (0.00, 4.0)
      Established predictors
       Modified CT Fisher I-IIrefref
       Modified CT Fisher III-IV3.1 (0.40, 23)2.1 (0.13, 31)
       WFNS Irefref
       WFNS II-III24 (0.71, 836)3.2 (0.03, 303)
       WFNS IV69 (1.07, 4450)31 (0.16, 5892)
       WFNS V5.9 (0.14, 251)1.0 (0.01, 162)
      Demographic characteristics
       Male sex13 (1.14, 150)37 (2.0, 666)
       Age0.90 (0.82, 0.99)0.95 (0.85, 1.05)
      AUC, crude0.90 (0.80, 1.00)0.96 (0.90, 1.00)
      AUC, adjusted for optimism
      Adjusted for optimism with bootstrapping techniques
      0.82 (0.69, 0.90)0.86 (0.70, 0.94)
      AUC: area under the receiver operating curve; CT: computed tomography; CVR: cerebrovascular reactivity; ref.: reference; WFNS: World Federation of Neurological Surgeons Scale.
      a Adjusted for optimism with bootstrapping techniques
      Fig. 3
      Fig. 3Discrimination and calibration plots for predictive models with and without CVR. A, The crude receiver operating characteristic (ROC) show the discrimination of the predictive model with and without CVR, not adjusted for optimism. See . B, The plot shows the calibration (actual outcome versus predicted outcome) analyzed in four equal groups for both predictive models.
      Two patients had a paradoxical velocity reduction after AZ. One patient with a single contralateral CVR of −14.9% (grey dot, Fig. 2) developed both DCIclinical and DCIinfarction. Another patient with a negative CVR recovered without ischemic symptoms or infarction (black dot, Fig. 2). Notably, the CVR reduction was transient, with CVR of −15.4% in the first exam and 37.4% when re-tested a week later, making the within-patient average CVR 11.0%.

      4. Discussion

      To our knowledge, this is the first study that has used TCD and AZ to compare CVR in patients with aSAH and UIA. The study indicates that contralateral CVR may be an independent predictor for clinical deterioration due to DCI after aSAH, but CVR is not able to predict radiographic infarction due to DCI.
      Various methods are used to assess CVR [
      • Fierstra J.
      • Sobczyk O.
      • Battisti-Charbonney A.
      • et al.
      Measuring cerebrovascular reactivity: what stimulus to use?.
      ]. It is unclear if CVR results are consistent for different methodologies. Comparing results by multiple methods could elucidate potential differences, but has not been done. The TCD-AZ test is widely applied for CVR testing. The method has some advantages compared with other methods (easier set-up, reduced need for patient co-operation, high safety profile), but also some disadvantages (potential side effects, dose-dependent effects, and contraindications for AZ) [
      • Fierstra J.
      • Sobczyk O.
      • Battisti-Charbonney A.
      • et al.
      Measuring cerebrovascular reactivity: what stimulus to use?.
      ]. We found that CVR was considerably lower in patients with ruptured aneurysms compared with patients with unruptured aneurysms and this difference appeared to be bigger on the ipsilateral side of aneurysm treatment. This is in concordance with previous CVR-studies using other vasoactive stimuli and measurement methods [
      • Dernbach P.D.
      • Little J.R.
      • Jones S.C.
      • et al.
      Altered cerebral autoregulation and CO2 reactivity after aneurysmal subarachnoid hemorrhage.
      ,
      • Seiler R.W.
      • Nirkko A.C.
      Effect of nimodipine on cerebrovascular response to CO2 in asymptomatic individuals and patients with subarachnoid hemorrhage: a transcranial Doppler ultrasound study.
      ,
      • Abe K.
      • Nishimura M.
      • Yoshiya I.
      Local cerebral blood flow and CO2 reactivity during prostaglandin E1-induced hypotension in patients undergoing cerebral aneurysm surgery.
      ,
      • Abe K.
      • Demizu A.
      • Mima T.
      • et al.
      Carbon dioxide reactivity during prostaglandin E1 induced hypotension for cerebral aneurysm surgery.
      ,
      • Schmieder K.
      • Jarus-Dziedzic K.
      • Wronski J.
      • et al.
      CO2 reactivity in patients after subarachnoid haemorrhage.
      ].
      Since the 1970s, literature has suggested that impaired CVR may be associated with DCI after aSAH [
      • da Costa L.
      • Houlden D.
      • Rubenfeld G.
      • et al.
      Impaired cerebrovascular reactivity in the early phase of subarachnoid hemorrhage in good clinical grade patients does not predict vasospasm.
      ,
      • Tanaka A.
      • Yoshinaga S.
      • Nakayama Y.
      • et al.
      Cerebral blood flow and the response to acetazolamide during the acute, subacute, and chronic stages of aneurysmal subarachnoid hemorrhage.
      ,
      • Carrera E.
      • Kurtz P.
      • Badjatia N.
      • et al.
      Cerebrovascular carbon dioxide reactivity and delayed cerebral ischemia after subarachnoid hemorrhage.
      ,
      • Ishii R.
      Regional cerebral blood flow in patients with ruptured intracranial aneurysms.
      ,
      • Kimura T.
      • Shinoda J.
      • Funakoshi T.
      Prediction of cerebral infarction due to vasospasm following aneurysmal subarachnoid haemorrhage using acetazolamide-activated 123I-IMP SPECT.
      ,
      • Frontera J.A.
      • Rundek T.
      • Schmidt J.M.
      • et al.
      Cerebrovascular reactivity and vasospasm after subarachnoid hemorrhage: a pilot study.
      ,
      • da Costa L.
      • Fisher J.
      • Mikulis D.J.
      • et al.
      Early identification of brain tissue at risk for delayed cerebral ischemia after aneurysmal subarachnoid hemorrhage.
      ,
      • Voldby B.
      • Enevoldsen E.M.
      • Jensen F.T.
      Cerebrovascular reactivity in patients with ruptured intracranial aneurysms.
      ,
      • Nakagawara J.
      • Wada K.
      • Takeda R.
      • et al.
      Prediction of cerebral ischemia due to cerebral vasospasm in SAH using SPECT and 123I-IMP with acetazolamide test.
      ,
      • Nakagawara J.
      • Wada K.
      • Takeda R.
      • et al.
      Incidence of clinically significant cerebral vasospasm in SAH estimated by local cerebral vasodilatory capacity using acetazolamide-activated 123I-IMP SPECT.
      ,
      • Messeter K.
      • Brandt L.
      • Ljunggren B.
      • et al.
      Prediction and prevention of delayed ischemic dysfunction after aneurysmal subarachnoid hemorrhage and early operation.
      ,
      • Mourier K.L.
      • George B.
      • Raggueneau J.L.
      • et al.
      Value of the measurement of cerebral blood flow before and after diamox injection in predicting clinical vasospasm and final outcome in aneurysmal subarachnoid hemorrhage.
      ,
      • Tran Dinh Y.R.
      • Lot G.
      • Benrabah R.
      • et al.
      Abnormal cerebral vasodilation in aneurysmal subarachnoid hemorrhage: use of serial 133Xe cerebral blood flow measurement plus acetazolamide to assess cerebral vasospasm.
      ,
      • Hassler W.
      • Chioffi F.
      CO2 reactivity of cerebral vasospasm after aneurysmal subarachnoid haemorrhage.
      ]. With the exception of one study [
      • da Costa L.
      • Houlden D.
      • Rubenfeld G.
      • et al.
      Impaired cerebrovascular reactivity in the early phase of subarachnoid hemorrhage in good clinical grade patients does not predict vasospasm.
      ], these findings are consistent across different methodologies, study designs, sample sizes, and definitions of DCI. Study sizes have however been small, and CVR has not been integrated as standard assessment after aSAH [
      • Steiner T.
      • Juvela S.
      • Unterberg A.
      • et al.
      European Stroke Organization guidelines for the management of intracranial aneurysms and subarachnoid haemorrhage.
      ,
      • Sloan M.A.
      • Alexandrov A.V.
      • Tegeler C.H.
      • et al.
      Assessment: transcranial Doppler ultrasonography: report of the Therapeutics and Technology Assessment Subcommittee of the American Academy of Neurology.
      ,
      • Connolly Jr., E.S.
      • Rabinstein A.A.
      • Carhuapoma J.R.
      • et al.
      Guidelines for the management of aneurysmal subarachnoid hemorrhage: a guideline for healthcare professionals from the American Heart Association/american Stroke Association.
      ,
      • Diringer M.N.
      • Bleck T.P.
      • Claude Hemphill 3rd, J.
      • et al.
      Critical care management of patients following aneurysmal subarachnoid hemorrhage: recommendations from the Neurocritical Care Society's Multidisciplinary Consensus Conference.
      ]. In clinical work, a CVR based prediction model of DCI would be valuable.
      In our prediction models, the confidence intervals for the coefficients of the established predictors all included the value 1 by quite some margin, whereas the confidence interval for the coefficient of CVR was 0.89 to 1.00 (Table 4). Note also that 0.94 is the OR per unit change in CVR, which corresponds to an OR of about 0.54 per change of 10, and an OR of 0.05 per change of 50. Among patients in our study, CVR ranges from less than zero to >90. Hence, contralateral CVR appears to be the strongest predictor of DCI in the model. The confidence intervals of the AUCs were overlapping, which indicates that CVR did not have additional value in the prediction of DCI after aSAH. Still, including CVR in the model caused a substantial increase of optimism adjusted AUC from 0.82 to 0.86. This warrants further investigation, especially as a more accurate prediction model for DCI is needed.
      Practical considerations such as ease-of-use and patient comfort are arguments in favor of single CVR testing. Serial testing is time consuming and can be challenging due to reduced patient co-operation, postoperative intracranial air, drains and monitoring equipment in the intensive care unit. Still, DCI has a dynamic nature, the first measured CVR had very low predictive power, and serial CVR-measurements provided added information regarding the dynamic changes in CVR throughout the acute phase after aSAH.
      Overall, 52.4% of patients with aSAH in this study developed DCI. In comparison, the reported incidence of DCI is 20–35% in larger case series [
      • Roos Y.B.
      • de Haan R.J.
      • Beenen L.F.
      • et al.
      Complications and outcome in patients with aneurysmal subarachnoid haemorrhage: a prospective hospital based cohort study in the Netherlands.
      ,
      • Michael Schmidt J.
      • Wartenberg Katja E.
      • Fernandez Andres
      • et al.
      Frequency and clinical impact of asymptomatic cerebral infarction due to vasospasm after subarachnoid hemorrhage.
      ,
      • Rivero-Arias O.
      • Wolstenholme J.
      • Gray A.
      • et al.
      The costs and prognostic characteristics of ischaemic neurological deficit due to subarachnoid haemorrhage in the United Kingdom. Evidence from the MRC International Subarachnoid Aneurysm Trial.
      ,
      • Hijdra A.
      • Van Gijn J.
      • Stefanko S.
      • et al.
      Delayed cerebral ischemia after aneurysmal subarachnoid hemorrhage: clinicoanatomic correlations.
      ,
      • Charpentier C.
      • Audibert G.
      • Guillemin F.
      • et al.
      Multivariate analysis of predictors of cerebral vasospasm occurrence after aneurysmal subarachnoid hemorrhage.
      ,
      • Claassen J.
      • Bernardini G.L.
      • Kreiter K.
      • et al.
      Effect of cisternal and ventricular blood on risk of delayed cerebral ischemia after subarachnoid hemorrhage: the Fisher scale revisited.
      ,
      • Hop J.W.
      • Rinkel G.J.
      • Algra A.
      • et al.
      Initial loss of consciousness and risk of delayed cerebral ischemia after aneurysmal subarachnoid hemorrhage.
      ,
      • Lasner T.M.
      • Weil R.J.
      • Riina H.A.
      • et al.
      Cigarette smoking-induced increase in the risk of symptomatic vasospasm after aneurysmal subarachnoid hemorrhage.
      ,
      • Murayama Y.
      • Malisch T.
      • Guglielmi G.
      • et al.
      Incidence of cerebral vasospasm after endovascular treatment of acutely ruptured aneurysms: report on 69 cases.
      ,
      • Qureshi A.I.
      • Sung G.Y.
      • Razumovsky A.Y.
      • et al.
      Early identification of patients at risk for symptomatic vasospasm after aneurysmal subarachnoid hemorrhage.
      ]. The high frequency of DCI may reflect the high proportion of poor grade patients in our cohort, as poor clinical grade upon admission is associated with higher risk of developing DCI. Given the high proportion of patients with large bleedings and poor clinical conditions in our study, results may not be fully valid for other populations. Still, stratification for clinical grade (WFNS I-III vs. IV-V) did not cause any major changes in the prediction model.
      In a previous report we argued that ipsi- and contralateral CVR should be assessed separately, as mean values can conceal side-differences [
      • Bothun M.L.
      • Haaland O.A.
      • Logallo N.
      • et al.
      Time course of cerebrovascular reactivity in patients treated for unruptured intracranial aneurysms: a one-year transcranial doppler and acetazolamide follow-up study.
      ]. One-year follow-up study indicated that patients with unruptured aneurysms had a temporary reduction in ipsilateral CVR. The present study indicates that ipsilateral CVR is reduced also after treatment for ruptured aneurysms. Ipsilateral CVR reduction thus seem to occur in all patients after aneurysm treatment, regardless of DCIclinical status. Acknowledging this side-difference enables improved interpretation of CVR after aneurysm treatment. Aneurysm treatment in itself does not induce a substantial CVR reduction on the contralateral side, making the contralateral CVR more suitable as a predictor since any DCI-related CVR reduction will be more pronounced on this side compared with the already reduced CVR on the ipsilateral side.
      In concordance with the pre-existing hypothesis that CVR can predict DCI, we found impaired CVR in patients with DCIclinical. Surprisingly, we did not find an equivalent reduction in CVR in patients with DCIinfarction. This is quite a conundrum as one would expect impaired CVR for both categories of DCI, assuming clinical symptoms and infarctions are two time-points in a continuum of the same disease. The statistical power in our study may be insufficient to detect CVR impairment in patients with DCIinfarction. Still, as DCIinfarction is presumably associated with more advanced ischemia one would also expect CVR impairment to be more advanced in patients with DCIinfarction compared with DCIclinical. However, the pathophysiology of cerebral ischemia after aSAH is multifactorial and complex [
      • Loch Macdonald R.
      Vasospasm: my first 25 years-what worked? what didn't? what next?.
      ], and it is possible that there are differences in the pathogenesis for clinical and radiographical presentations. Furthermore, we used DCIinfarction as a dichotomous variable, not taking into account the size or severity of the infarction. For ethical reasons, moribund patients and patients with increased ICP were excluded (Table S1), this probably reduced the number of patients with severe infarctions that was tested and left a higher proportion of minor infarctions in the studied population. Additional studies with larger study sample and more comprehensive radiographical examinations including assessment of infarction size could help clarify if impaired CVR is associated with DCIinfarction or not.
      There are limitations to our study. The study population is small, although the number of patients (n = 42) is higher than in other neurosonological studies (median n = 27, range 18–34) [
      • da Costa L.
      • Houlden D.
      • Rubenfeld G.
      • et al.
      Impaired cerebrovascular reactivity in the early phase of subarachnoid hemorrhage in good clinical grade patients does not predict vasospasm.
      ,
      • Carrera E.
      • Kurtz P.
      • Badjatia N.
      • et al.
      Cerebrovascular carbon dioxide reactivity and delayed cerebral ischemia after subarachnoid hemorrhage.
      ,
      • Schmieder K.
      • Jarus-Dziedzic K.
      • Wronski J.
      • et al.
      CO2 reactivity in patients after subarachnoid haemorrhage.
      ,
      • Frontera J.A.
      • Rundek T.
      • Schmidt J.M.
      • et al.
      Cerebrovascular reactivity and vasospasm after subarachnoid hemorrhage: a pilot study.
      ,
      • Hassler W.
      • Chioffi F.
      CO2 reactivity of cerebral vasospasm after aneurysmal subarachnoid haemorrhage.
      ,
      • Jarus-Dziedzic K.
      • Glowacki M.
      • Warzecha A.
      • et al.
      Cerebrovascular reactivity evaluated by transcranial doppler sonography in patients after aneurysmal subarachnoid haemorrhage treated with microsurgical clipping or endovascular coiling technique.
      ,
      • Szabo S.
      • Sheth R.N.
      • Novak L.
      • et al.
      Cerebrovascular reserve capacity many years after vasospasm due to aneurysmal subarachnoid hemorrhage. A transcranial Doppler study with acetazolamide test.
      ]. Further limitations are the lack of external validation of the prediction model. We did not examine patients according to a rigid time schedule. To better compare the aSAH and UIA groups, and overcome the dynamic nature of CVR over time, CVR should ideally have been examined in all patients at the same time intervals after treatment. All patients underwent cerebral imaging with CT and/or MRI after aneurysm treatment. Among 42 patients with aSAH, 17 were only evaluated with CT/CTA during their primary hospital stay. Still, MRI was performed in the majority of patients during follow-up. Of the 39 patients alive after one year, 36 underwent MRI; two were tourists and lost to follow-up; and one patient declined to take the exam due to claustrophobia. We used a well-known classification system for cerebral infarction [
      • Rabinstein A.A.
      • Weigand S.
      • Atkinson J.L.
      • et al.
      Patterns of cerebral infarction in aneurysmal subarachnoid hemorrhage.
      ,
      • Naidech A.M.
      • Bendok B.R.
      • Bassin S.L.
      • et al.
      Classification of cerebral infarction after subarachnoid hemorrhage impacts outcome.
      ]. Still, the attribution of infarction etiology after aSAH is difficult and not easily validated [
      • Ibrahim G.M.
      • Weidauer S.
      • Vatter H.
      • et al.
      Attributing hypodensities on CT to angiographic vasospasm is not sensitive and unreliable.
      ], and one cannot be certain whether infarctions are related to DCI or not. To avoid misclassifying procedure-related infarcts as vasospasm-induced, CT and MRI scans should be performed between 24 and 48 h after aneurysm occlusion. This was only done in 33 of 42 patients. The diagnosis of sonographic vasospasm and clinical deterioration due to DCI was set without blinding of CVR-results. Ideally, the AZ doses should have been bodyweight-based in all patients in the study, not only in patients weighing ≥80 kg. Still, the recommended AZ dose of 13 to 18 mg/kg [
      • Dahl A.
      • Russell D.
      • Rootwelt K.
      • et al.
      Cerebral vasoreactivity assessed with transcranial Doppler and regional cerebral blood flow measurements. Dose, serum concentration, and time course of the response to acetazolamide.
      ,
      • Grossmann W.M.
      • Koeberle B.
      The dose-response relationship of acetazolamide on the cerebral blood flow in normal subjects.
      ] was achieved in 91.7% of CVR tests. We used different brands of AZ, yet no difference in CVR has been found when the three manufacturers were compared [
      • Bothun M.L.
      • Haaland O.A.
      • Logallo N.
      • et al.
      Time course of cerebrovascular reactivity in patients treated for unruptured intracranial aneurysms: a one-year transcranial doppler and acetazolamide follow-up study.
      ].
      A strength of our study is the methodology. Unlike the majority of previous studies [
      • da Costa L.
      • Houlden D.
      • Rubenfeld G.
      • et al.
      Impaired cerebrovascular reactivity in the early phase of subarachnoid hemorrhage in good clinical grade patients does not predict vasospasm.
      ,
      • Tanaka A.
      • Yoshinaga S.
      • Nakayama Y.
      • et al.
      Cerebral blood flow and the response to acetazolamide during the acute, subacute, and chronic stages of aneurysmal subarachnoid hemorrhage.
      ,
      • Ishii R.
      Regional cerebral blood flow in patients with ruptured intracranial aneurysms.
      ,
      • Kimura T.
      • Shinoda J.
      • Funakoshi T.
      Prediction of cerebral infarction due to vasospasm following aneurysmal subarachnoid haemorrhage using acetazolamide-activated 123I-IMP SPECT.
      ,
      • Frontera J.A.
      • Rundek T.
      • Schmidt J.M.
      • et al.
      Cerebrovascular reactivity and vasospasm after subarachnoid hemorrhage: a pilot study.
      ,
      • da Costa L.
      • Fisher J.
      • Mikulis D.J.
      • et al.
      Early identification of brain tissue at risk for delayed cerebral ischemia after aneurysmal subarachnoid hemorrhage.
      ,
      • Voldby B.
      • Enevoldsen E.M.
      • Jensen F.T.
      Cerebrovascular reactivity in patients with ruptured intracranial aneurysms.
      ,
      • Nakagawara J.
      • Wada K.
      • Takeda R.
      • et al.
      Prediction of cerebral ischemia due to cerebral vasospasm in SAH using SPECT and 123I-IMP with acetazolamide test.
      ,
      • Nakagawara J.
      • Wada K.
      • Takeda R.
      • et al.
      Incidence of clinically significant cerebral vasospasm in SAH estimated by local cerebral vasodilatory capacity using acetazolamide-activated 123I-IMP SPECT.
      ,
      • Messeter K.
      • Brandt L.
      • Ljunggren B.
      • et al.
      Prediction and prevention of delayed ischemic dysfunction after aneurysmal subarachnoid hemorrhage and early operation.
      ,
      • Mourier K.L.
      • George B.
      • Raggueneau J.L.
      • et al.
      Value of the measurement of cerebral blood flow before and after diamox injection in predicting clinical vasospasm and final outcome in aneurysmal subarachnoid hemorrhage.
      ,
      • Tran Dinh Y.R.
      • Lot G.
      • Benrabah R.
      • et al.
      Abnormal cerebral vasodilation in aneurysmal subarachnoid hemorrhage: use of serial 133Xe cerebral blood flow measurement plus acetazolamide to assess cerebral vasospasm.
      ], we used standardized definitions of DCI separated into clinical and radiographic findings [
      • Vergouwen M.D.
      • Vermeulen M.
      • van Gijn J.
      • et al.
      Definition of delayed cerebral ischemia after aneurysmal subarachnoid hemorrhage as an outcome event in clinical trials and observational studies: proposal of a multidisciplinary research group.
      ,
      • Vergouwen MD and Participants in the International Multi-Disciplinary Consensus Conference on the Critical Care Management of Subarachnoid H
      Vasospasm versus delayed cerebral ischemia as an outcome event in clinical trials and observational studies.
      ]. We performed serial testing, which is better adjusted to the dynamic nature of vasospasm and DCI. We also tested CVR in patients treated for UIA to increase our general understanding of CVR in patients with intracranial aneurysms. One sonographer (MLB) performed all ultrasound examinations and CVR-tests to reduce operator variability. One neuroradiologist (GM) diagnosed angiographic vasospasm and cerebral infarction, unaware of the patient's clinical and sonographic status.
      Central drawbacks with regards to clinical use is that CVR testing is time consuming and that a number of patients experience side effects related to AZ, like headache, flushing, nausea or paresthesia [
      • Bothun M.L.
      • Haaland O.A.
      • Logallo N.
      • et al.
      Cerebrovascular reactivity after treatment of unruptured intracranial aneurysms - a transcranial Doppler sonography and acetazolamide study.
      ]. Still, potential side effects are usually transient and well tolerated [
      • Settakis G.
      • Molnar C.
      • Kerenyi L.
      • et al.
      Acetazolamide as a vasodilatory stimulus in cerebrovascular diseases and in conditions affecting the cerebral vasculature.
      ]. Results are promising, but does not provide firm confirmation that impaired CVR predicts DCI. There are still unresolved issues, and it is too soon to recommend routine use of CVR in clinical practice. Results needs to be externally validated, and the conundrum regarding CVRs relation to DCIinfarction needs to be investigated. If validation studies confirm our findings, CVR testing may assist clinicians in early identification of patients who may benefit from aggressive prophylactic treatment, closer monitoring, or repeat vascular imaging. Determining those at greatest risk can help reduce DCI-related poor outcomes while minimizing treatment complications and titrate length of stay in an intensive care unit.
      Pretreatment CVR testing could assess potential effects on CVR caused by aneurysm treatment itself. Still, patients were not examined prior to treatment because of the potential risk of causing aneurysm (re-)rupture, and to avoid delay in the acute treatment of a ruptured aneurysm. CVR can be assessed at an earlier time in patients with UIA with assumed low rupture risk and no planned treatment, but this will not give any information about the status of cerebrovascular integrity immediately before aneurysm treatment or rupture. We did not take into account variations of intracranial pressure (ICP) or systemic blood pressure in patients with aSAH. AZ does not cause major changes in systemic blood pressure [
      • Piepgras A.
      • Schmiedek P.
      • Leinsinger G.
      • et al.
      A simple test to assess cerebrovascular reserve capacity using transcranial Doppler sonography and acetazolamide.
      ,
      • Vagal A.S.
      • Leach J.L.
      • Fernandez-Ulloa M.
      • et al.
      The acetazolamide challenge: techniques and applications in the evaluation of chronic cerebral ischemia.
      ], but may lower ICP by reducing cerebrospinal fluid secretion [
      • Piper R.J.
      • Kalyvas A.V.
      • Young A.M.
      • et al.
      Interventions for idiopathic intracranial hypertension.
      ]. AZ may also increase ICP by cerebral vasodilation [
      • Demolis P.
      • Florence G.
      • Thomas L.
      • et al.
      Is the acetazolamide test valid for quantitative assessment of maximal cerebral autoregulatory vasodilation? An experimental study.
      ] and may not be well suited for patients with increased ICP. These high-risk patients are closely monitored for development of DCI and AZ testing might be of less importance. In less severe cases, however, improved prediction of DCI is highly useful, and can assist medical and logistic decision making.

      5. Conclusions

      Impaired CVR on the contralateral side may be an independent predictor of DCIclinical, and may assist in identifying patients in need of closer observation after aSAH. An ipsilateral CVR reduction occurs in all patients after aneurysm treatment, regardless of DCIclinical status, highlighting the need to consider ipsilateral and contralateral CVR separately. This study underscores the value of serial CVR measurements, and future studies aiming to investigate the relationship between CVR and DCI after aSAH should take this into account. Our prediction model can be useful in clinical practice, but needs to be validated.

      Author contribution statement

      Study conception and design: Bøthun, Helland, Thomassen. Acquisition of data: Bøthun, Helland, Moen. Analysis and interpretation of data: All. Drafting of manuscript: Bøthun, Haaland. Critically revising the article: All. Reviewed submitted version of manuscript: All. Statistical analysis: Haaland. Study supervision: Bøthun, Helland, Thomassen.

      Declaration of Competing Interest

      The authors declare that they have no conflict of interest.

      Acknowledgements

      The study was funded by scholarships and grants from the University of Bergen and the Western Norway Regional Health Authority (“Helse Vest”).

      Appendix A. Supplementary data

      References

        • Roos Y.B.
        • de Haan R.J.
        • Beenen L.F.
        • et al.
        Complications and outcome in patients with aneurysmal subarachnoid haemorrhage: a prospective hospital based cohort study in the Netherlands.
        J. Neurol. Neurosurg. Psychiatry. 2000; 68 (2000/02/16): 337-341
        • Kassell N.F.
        • Torner J.C.
        • Haley Jr., E.C.
        • et al.
        The international cooperative study on the timing of aneurysm surgery. Part 1: overall management results.
        J. Neurosurg. 1990; 73: 18-36https://doi.org/10.3171/jns.1990.73.1.0018
        • Adams Jr., H.P.
        • Kassell N.F.
        • Torner J.C.
        • et al.
        Predicting cerebral ischemia after aneurysmal subarachnoid hemorrhage: influences of clinical condition, CT results, and antifibrinolytic therapy. A report of the Cooperative Aneurysm Study.
        Neurology. 1987; 37 (1987/10/01): 1586-1591
        • Fisher C.M.
        • Kistler J.P.
        • Davis J.M.
        Relation of cerebral vasospasm to subarachnoid hemorrhage visualized by computerized tomographic scanning.
        Neurosurgery. 1980; 6 (1980/01/01): 1-9
        • van Gijn J.
        • Kerr R.S.
        • Rinkel G.J.
        Subarachnoid haemorrhage.
        Lancet. 2007; 369 (2007/01/30): 306-318
        • de Oliveira Manoel A.L.
        • Jaja B.N.
        • Germans M.R.
        • et al.
        The VASOGRADE: a simple grading scale for prediction of delayed cerebral ischemia after subarachnoid hemorrhage.
        Stroke. 2015; 46: 1826-1831https://doi.org/10.1161/STROKEAHA.115.008728
        • Crobeddu E.
        • Mittal M.K.
        • Dupont S.
        • et al.
        Predicting the lack of development of delayed cerebral ischemia after aneurysmal subarachnoid hemorrhage.
        Stroke. 2012; 43: 697-701https://doi.org/10.1161/STROKEAHA.111.638403
        • de Rooij N.K.
        • Rinkel G.J.
        • Dankbaar J.W.
        • et al.
        Delayed cerebral ischemia after subarachnoid hemorrhage: a systematic review of clinical, laboratory, and radiological predictors.
        Stroke. 2013; 44 (2012/12/20): 43-54
        • Dernbach P.D.
        • Little J.R.
        • Jones S.C.
        • et al.
        Altered cerebral autoregulation and CO2 reactivity after aneurysmal subarachnoid hemorrhage.
        Neurosurgery. 1988; 22: 822-826
        • Seiler R.W.
        • Nirkko A.C.
        Effect of nimodipine on cerebrovascular response to CO2 in asymptomatic individuals and patients with subarachnoid hemorrhage: a transcranial Doppler ultrasound study.
        Neurosurgery. 1990; 27: 247-251
        • Abe K.
        • Nishimura M.
        • Yoshiya I.
        Local cerebral blood flow and CO2 reactivity during prostaglandin E1-induced hypotension in patients undergoing cerebral aneurysm surgery.
        Eur. J. Anaesthesiol. 1992; 9: 485-491
        • Abe K.
        • Demizu A.
        • Mima T.
        • et al.
        Carbon dioxide reactivity during prostaglandin E1 induced hypotension for cerebral aneurysm surgery.
        Can. J. Anaesthesia. 1992; 39: 253-259https://doi.org/10.1007/BF03008786
        • Shinoda J.
        • Kimura T.
        • Funakoshi T.
        • et al.
        Acetazolamide reactivity on cerebral blood flow in patients with subarachnoid haemorrhage.
        Acta Neurochir. 1991; 109: 102-108
        • da Costa L.
        • Houlden D.
        • Rubenfeld G.
        • et al.
        Impaired cerebrovascular reactivity in the early phase of subarachnoid hemorrhage in good clinical grade patients does not predict vasospasm.
        Acta Neurochir. Suppl. 2015; 120: 249-253https://doi.org/10.1007/978-3-319-04981-6_42
        • Tanaka A.
        • Yoshinaga S.
        • Nakayama Y.
        • et al.
        Cerebral blood flow and the response to acetazolamide during the acute, subacute, and chronic stages of aneurysmal subarachnoid hemorrhage.
        Neurol. Med. Chir. 1998; 38 (discussion 630–622): 623-630
        • Carrera E.
        • Kurtz P.
        • Badjatia N.
        • et al.
        Cerebrovascular carbon dioxide reactivity and delayed cerebral ischemia after subarachnoid hemorrhage.
        Arch. Neurol. 2010; 67: 434-439https://doi.org/10.1001/archneurol.2010.43
        • Ishii R.
        Regional cerebral blood flow in patients with ruptured intracranial aneurysms.
        J. Neurosurg. 1979; 50: 587-594https://doi.org/10.3171/jns.1979.50.5.0587
        • Abe K.
        • Demizu A.
        • Kamada K.
        • et al.
        Prostaglandin E1 and carbon dioxide reactivity during cerebral aneurysm surgery.
        Can. J. Anaesthesia. 1992; 39: 247-252https://doi.org/10.1007/BF03008785
        • Schmieder K.
        • Jarus-Dziedzic K.
        • Wronski J.
        • et al.
        CO2 reactivity in patients after subarachnoid haemorrhage.
        Acta Neurochir. 1997; 139: 1038-1041
        • Bothun M.L.
        • Haaland O.A.
        • Logallo N.
        • et al.
        Cerebrovascular reactivity after treatment of unruptured intracranial aneurysms - a transcranial Doppler sonography and acetazolamide study.
        J. Neurol. Sci. 2016; 363: 97-103https://doi.org/10.1016/j.jns.2015.12.024
        • Kimura T.
        • Shinoda J.
        • Funakoshi T.
        Prediction of cerebral infarction due to vasospasm following aneurysmal subarachnoid haemorrhage using acetazolamide-activated 123I-IMP SPECT.
        Acta Neurochir. 1993; 123: 125-128
        • Frontera J.A.
        • Rundek T.
        • Schmidt J.M.
        • et al.
        Cerebrovascular reactivity and vasospasm after subarachnoid hemorrhage: a pilot study.
        Neurology. 2006; 66: 727-729https://doi.org/10.1212/01.wnl.0000200777.96896.3d
        • da Costa L.
        • Fisher J.
        • Mikulis D.J.
        • et al.
        Early identification of brain tissue at risk for delayed cerebral ischemia after aneurysmal subarachnoid hemorrhage.
        Acta Neurochir. Suppl. 2015; 120 (2014/11/05): 105-109
        • Voldby B.
        • Enevoldsen E.M.
        • Jensen F.T.
        Cerebrovascular reactivity in patients with ruptured intracranial aneurysms.
        J. Neurosurg. 1985; 62: 59-67https://doi.org/10.3171/jns.1985.62.1.0059
        • Nakagawara J.
        • Wada K.
        • Takeda R.
        • et al.
        Prediction of cerebral ischemia due to cerebral vasospasm in SAH using SPECT and 123I-IMP with acetazolamide test.
        Surg. Cereb. Stroke. 1989; 17: 301-307https://doi.org/10.2335/scs1987.17.4_301
        • Nakagawara J.
        • Wada K.
        • Takeda R.
        • et al.
        Incidence of clinically significant cerebral vasospasm in SAH estimated by local cerebral vasodilatory capacity using acetazolamide-activated 123I-IMP SPECT.
        Surg. Cereb. Stroke. 1991; 19: 178-182https://doi.org/10.2335/scs1987.19.2_178
        • Messeter K.
        • Brandt L.
        • Ljunggren B.
        • et al.
        Prediction and prevention of delayed ischemic dysfunction after aneurysmal subarachnoid hemorrhage and early operation.
        Neurosurgery. 1987; 20 (1987/04/01): 548-553
        • Mourier K.L.
        • George B.
        • Raggueneau J.L.
        • et al.
        Value of the measurement of cerebral blood flow before and after diamox injection in predicting clinical vasospasm and final outcome in aneurysmal subarachnoid hemorrhage.
        Neuro-Chirurgie. 1991; 37 (1991/01/01): 318-322
        • Tran Dinh Y.R.
        • Lot G.
        • Benrabah R.
        • et al.
        Abnormal cerebral vasodilation in aneurysmal subarachnoid hemorrhage: use of serial 133Xe cerebral blood flow measurement plus acetazolamide to assess cerebral vasospasm.
        J. Neurosurg. 1993; 79 (1993/10/01): 490-493
        • Hassler W.
        • Chioffi F.
        CO2 reactivity of cerebral vasospasm after aneurysmal subarachnoid haemorrhage.
        Acta Neurochir. 1989; 98: 167-175
        • Teasdale G.
        • Jennett B.
        Assessment of coma and impaired consciousness. A practical scale.
        Lancet. 1974; 2: 81-84
        • Report of World Federation of Neurological Surgeons Committee on a Universal Subarachnoid Hemorrhage Grading Scale
        J. Neurosurg. 1988; 68: 985-986
        • Frontera J.A.
        • Claassen J.
        • Schmidt J.M.
        • et al.
        Prediction of symptomatic vasospasm after subarachnoid hemorrhage: the modified fisher scale.
        Neurosurgery. 2006; 59 (discussion 21–27): 21-27
        • Lindegaard K.F.
        • Nornes H.
        • Bakke S.J.
        • et al.
        Cerebral vasospasm diagnosis by means of angiography and blood velocity measurements.
        Acta Neurochir. 1989; 100: 12-24
        • Steiner T.
        • Juvela S.
        • Unterberg A.
        • et al.
        European Stroke Organization guidelines for the management of intracranial aneurysms and subarachnoid haemorrhage.
        Cerebrovasc. Dis. 2013; 35 (2013/02/15): 93-112
        • Dorhout Mees S.M.
        • Rinkel G.J.
        • Feigin V.L.
        • et al.
        Calcium antagonists for aneurysmal subarachnoid haemorrhage.
        Cochrane Database Syst. Rev. 2007; (2007/07/20): CD000277
        • Biondi A.
        • Ricciardi G.K.
        • Puybasset L.
        • et al.
        Intra-arterial nimodipine for the treatment of symptomatic cerebral vasospasm after aneurysmal subarachnoid hemorrhage: preliminary results.
        AJNR Am. J. Neuroradiol. 2004; 25 (2004/06/19): 1067-1076
        • Vergouwen M.D.
        • Vermeulen M.
        • van Gijn J.
        • et al.
        Definition of delayed cerebral ischemia after aneurysmal subarachnoid hemorrhage as an outcome event in clinical trials and observational studies: proposal of a multidisciplinary research group.
        Stroke. 2010; 41: 2391-2395https://doi.org/10.1161/STROKEAHA.110.589275
        • Vergouwen MD and Participants in the International Multi-Disciplinary Consensus Conference on the Critical Care Management of Subarachnoid H
        Vasospasm versus delayed cerebral ischemia as an outcome event in clinical trials and observational studies.
        Neurocrit. Care. 2011; 15: 308-311https://doi.org/10.1007/s12028-011-9586-8
        • Nathal E.
        • Lopez-Gonzalez F.
        • Rios C.
        Angiographic scale for evaluation of cerebral vasospasm.
        Acta Neurochir. Suppl. 2007; 000: 1-4
        • Bothun M.L.
        • Haaland O.A.
        • Logallo N.
        • et al.
        Time course of cerebrovascular reactivity in patients treated for unruptured intracranial aneurysms: a one-year transcranial doppler and acetazolamide follow-up study.
        Biomed. Res. Int. 2018; 2018 (2018/06/02): 6489276
        • Sorteberg W.
        • Lindegaard K.F.
        • Rootwelt K.
        • et al.
        Effect of acetazolamide on cerebral artery blood velocity and regional cerebral blood flow in normal subjects.
        Acta Neurochir. 1989; 97: 139-145
        • Piepgras A.
        • Schmiedek P.
        • Leinsinger G.
        • et al.
        A simple test to assess cerebrovascular reserve capacity using transcranial Doppler sonography and acetazolamide.
        Stroke. 1990; 21: 1306-1311
        • Mancini M.
        • De Chiara S.
        • Postiglione A.
        • et al.
        Transcranial Doppler evaluation of cerebrovascular reactivity to acetazolamide in normal subjects.
        Artery. 1993; 20: 231-241
        • Dahl A.
        • Russell D.
        • Nyberg-Hansen R.
        • et al.
        Simultaneous assessment of vasoreactivity using transcranial Doppler ultrasound and cerebral blood flow in healthy subjects.
        J. Cereb. Blood Flow Metab. 1994; 14: 974-981https://doi.org/10.1038/jcbfm.1994.130
        • Dahl A.
        • Russell D.
        • Rootwelt K.
        • et al.
        Cerebral vasoreactivity assessed with transcranial Doppler and regional cerebral blood flow measurements. Dose, serum concentration, and time course of the response to acetazolamide.
        Stroke. 1995; 26: 2302-2306
        • Ulrich P.T.
        • Becker T.
        • Kempski O.S.
        Correlation of cerebral blood flow and MCA flow velocity measured in healthy volunteers during acetazolamide and CO2 stimulation.
        J. Neurol. Sci. 1995; 129: 120-130
        • Hamann G.F.
        • Stoll M.
        • Jost V.
        • et al.
        Time course of acetazolamide effect in normal persons.
        J. Neuroimaging. 1996; 6: 29-31
        • Karnik R.
        • Valentin A.
        • Winkler W.B.
        • et al.
        Sex-related differences in acetazolamide-induced cerebral vasomotor reactivity.
        Stroke. 1996; 27: 56-58
        • Patrick J.T.
        • Fritz J.V.
        • Adamo J.M.
        • et al.
        Phase-contrast magnetic resonance angiography for the determination of cerebrovascular reserve.
        J. Neuroimaging. 1996; 6: 137-143
        • Valikovics A.
        • Olah L.
        • Fulesdi B.
        • et al.
        Cerebrovascular reactivity measured by transcranial Doppler in migraine.
        Headache. 1996; 36: 323-328
        • Fulesdi B.
        • Limburg M.
        • Bereczki D.
        • et al.
        Impairment of cerebrovascular reactivity in long-term type 1 diabetes.
        Diabetes. 1997; 46: 1840-1845
        • Olah L.
        • Valikovics A.
        • Bereczki D.
        • et al.
        Gender-related differences in acetazolamide-induced cerebral vasodilatory response: a transcranial Doppler study.
        J. Neuroimaging. 2000; 10: 151-156
        • Schwertfeger N.
        • Neu P.
        • Schlattmann P.
        • et al.
        Cerebrovascular reactivity over time course in healthy subjects.
        J. Neurol. Sci. 2006; 249: 135-139https://doi.org/10.1016/j.jns.2006.06.009
        • Steyerberg E.W.
        • Vickers A.J.
        • Cook N.R.
        • et al.
        Assessing the performance of prediction models: a framework for traditional and novel measures.
        Epidemiology. 2010; 21 (2009/12/17): 128-138
        • Harrell F.
        Regression Modeling Strategies With Applications to Linear Models, Logistic and Ordinal Regression, and Survival Analysis.
        2 ed. Springer International Publishing, 2015
        • R Core Team
        R: A Language and Environment for Statistical Computing.
        R Foundation for Statistical Computing, Vienna, Austria2017
        • Fierstra J.
        • Sobczyk O.
        • Battisti-Charbonney A.
        • et al.
        Measuring cerebrovascular reactivity: what stimulus to use?.
        J. Physiol. 2013; 591: 5809-5821https://doi.org/10.1113/jphysiol.2013.259150
        • Sloan M.A.
        • Alexandrov A.V.
        • Tegeler C.H.
        • et al.
        Assessment: transcranial Doppler ultrasonography: report of the Therapeutics and Technology Assessment Subcommittee of the American Academy of Neurology.
        Neurology. 2004; 62 (2004/05/12): 1468-1481
        • Connolly Jr., E.S.
        • Rabinstein A.A.
        • Carhuapoma J.R.
        • et al.
        Guidelines for the management of aneurysmal subarachnoid hemorrhage: a guideline for healthcare professionals from the American Heart Association/american Stroke Association.
        Stroke. 2012; 43 (2012/05/05): 1711-1737
        • Diringer M.N.
        • Bleck T.P.
        • Claude Hemphill 3rd, J.
        • et al.
        Critical care management of patients following aneurysmal subarachnoid hemorrhage: recommendations from the Neurocritical Care Society's Multidisciplinary Consensus Conference.
        Neurocrit. Care. 2011; 15 (2011/07/21): 211-240
        • Michael Schmidt J.
        • Wartenberg Katja E.
        • Fernandez Andres
        • et al.
        Frequency and clinical impact of asymptomatic cerebral infarction due to vasospasm after subarachnoid hemorrhage.
        J. Neurosurg. 2008; 109: 1052-1059https://doi.org/10.3171/jns.2008.109.12.1052
        • Rivero-Arias O.
        • Wolstenholme J.
        • Gray A.
        • et al.
        The costs and prognostic characteristics of ischaemic neurological deficit due to subarachnoid haemorrhage in the United Kingdom. Evidence from the MRC International Subarachnoid Aneurysm Trial.
        J. Neurol. 2009; 256 (2009/02/18): 364-373
        • Hijdra A.
        • Van Gijn J.
        • Stefanko S.
        • et al.
        Delayed cerebral ischemia after aneurysmal subarachnoid hemorrhage: clinicoanatomic correlations.
        Neurology. 1986; 36 (1986/03/01): 329-333
        • Charpentier C.
        • Audibert G.
        • Guillemin F.
        • et al.
        Multivariate analysis of predictors of cerebral vasospasm occurrence after aneurysmal subarachnoid hemorrhage.
        Stroke. 1999; 30 (1999/07/02): 1402-1408
        • Claassen J.
        • Bernardini G.L.
        • Kreiter K.
        • et al.
        Effect of cisternal and ventricular blood on risk of delayed cerebral ischemia after subarachnoid hemorrhage: the Fisher scale revisited.
        Stroke. 2001; 32 (2001/09/08): 2012-2020
        • Hop J.W.
        • Rinkel G.J.
        • Algra A.
        • et al.
        Initial loss of consciousness and risk of delayed cerebral ischemia after aneurysmal subarachnoid hemorrhage.
        Stroke. 1999; 30 (1999/11/05): 2268-2271
        • Lasner T.M.
        • Weil R.J.
        • Riina H.A.
        • et al.
        Cigarette smoking-induced increase in the risk of symptomatic vasospasm after aneurysmal subarachnoid hemorrhage.
        J. Neurosurg. 1997; 87 (1997/09/01): 381-384
        • Murayama Y.
        • Malisch T.
        • Guglielmi G.
        • et al.
        Incidence of cerebral vasospasm after endovascular treatment of acutely ruptured aneurysms: report on 69 cases.
        J. Neurosurg. 1997; 87 (1997/12/31): 830-835
        • Qureshi A.I.
        • Sung G.Y.
        • Razumovsky A.Y.
        • et al.
        Early identification of patients at risk for symptomatic vasospasm after aneurysmal subarachnoid hemorrhage.
        Crit. Care Med. 2000; 28 (2000/05/16): 984-990
        • Loch Macdonald R.
        Vasospasm: my first 25 years-what worked? what didn't? what next?.
        Acta Neurochir. Suppl. 2015; 120 (2014/11/05): 1-10
        • Jarus-Dziedzic K.
        • Glowacki M.
        • Warzecha A.
        • et al.
        Cerebrovascular reactivity evaluated by transcranial doppler sonography in patients after aneurysmal subarachnoid haemorrhage treated with microsurgical clipping or endovascular coiling technique.
        Neurol. Res. 2011; 33: 18-23https://doi.org/10.1179/016164110X12700393823534
        • Szabo S.
        • Sheth R.N.
        • Novak L.
        • et al.
        Cerebrovascular reserve capacity many years after vasospasm due to aneurysmal subarachnoid hemorrhage. A transcranial Doppler study with acetazolamide test.
        Stroke. 1997; 28 (1997/12/31): 2479-2482
        • Rabinstein A.A.
        • Weigand S.
        • Atkinson J.L.
        • et al.
        Patterns of cerebral infarction in aneurysmal subarachnoid hemorrhage.
        Stroke. 2005; 36: 992-997https://doi.org/10.1161/01.STR.0000163090.59350.5a
        • Naidech A.M.
        • Bendok B.R.
        • Bassin S.L.
        • et al.
        Classification of cerebral infarction after subarachnoid hemorrhage impacts outcome.
        Neurosurgery. 2009; 64 (discussion 1057–1058): 1052-1057
        • Ibrahim G.M.
        • Weidauer S.
        • Vatter H.
        • et al.
        Attributing hypodensities on CT to angiographic vasospasm is not sensitive and unreliable.
        Stroke. 2012; 43: 109-112https://doi.org/10.1161/STROKEAHA.111.632745
        • Grossmann W.M.
        • Koeberle B.
        The dose-response relationship of acetazolamide on the cerebral blood flow in normal subjects.
        Cerebrovasc. Dis. 2000; 10: 65-69
        • Settakis G.
        • Molnar C.
        • Kerenyi L.
        • et al.
        Acetazolamide as a vasodilatory stimulus in cerebrovascular diseases and in conditions affecting the cerebral vasculature.
        Eur. J. Neurol. 2003; 10: 609-620
        • Vagal A.S.
        • Leach J.L.
        • Fernandez-Ulloa M.
        • et al.
        The acetazolamide challenge: techniques and applications in the evaluation of chronic cerebral ischemia.
        AJNR Am. J. Neuroradiol. 2009; 30 (2009/02/28): 876-884
        • Piper R.J.
        • Kalyvas A.V.
        • Young A.M.
        • et al.
        Interventions for idiopathic intracranial hypertension.
        Cochrane Database Syst. Rev. 2015; (2015/08/08): CD003434
        • Demolis P.
        • Florence G.
        • Thomas L.
        • et al.
        Is the acetazolamide test valid for quantitative assessment of maximal cerebral autoregulatory vasodilation? An experimental study.
        Stroke. 2000; 31 (2000/02/05): 508-515