Login to MyKarger

New to MyKarger? Click here to sign up.



Login with Facebook

Forgot your password?

Authors, Editors, Reviewers

For Manuscript Submission, Check or Review Login please go to Submission Websites List.

Submission Websites List

Institutional Login
(Shibboleth or Open Athens)

For the academic login, please select your country in the dropdown list. You will be redirected to verify your credentials.

Original Paper

Free Access

Computed Tomography Perfusion Improves Diagnostic Accuracy in Acute Posterior Circulation Stroke

Sporns P.a · Schmidt R.c · Minnerup J.b · Dziewas R.b · Kemmling A.e · Dittrich R.b · Zoubi T.a · Heermann P.a · Cnyrim C.a · Schwindt W.a · Heindel W.a · Niederstadt T.a · Hanning U.a, d

Author affiliations

aDepartment of Clinical Radiology and bDepartment of Neurology, University Hospital of Muenster, cInstitute of Biostatistics and Clinical Research, dDepartment of Epidemiology and Social Medicine, University of Muenster, Muenster, eInstitute of Neuroradiology, University Hospital of Luebeck, Luebeck, Germany

Corresponding Author

Uta Hanning, MD and Peter Sporns, MD

Department of Clinical Radiology and Department of Epidemiology and Social Medicine, University Hospital of Muenster

Albert-Schweitzer-Campus 1, Gebäude D3, DE-48149 Muenster (Germany)

E-Mail uhanning@uni-muenster.de; Peter.Sporns@ukmuenster.de

Related Articles for ""

Cerebrovasc Dis 2016;41:242-247

Abstract

Background and Purpose: Computed tomography perfusion (CTP) has a high diagnostic value in the detection of acute ischemic stroke in the anterior circulation. However, the diagnostic value in suspected posterior circulation (PC) stroke is uncertain, and whole brain volume perfusion is not yet in widespread use. We therefore studied the additional value of whole brain volume perfusion to non-contrast CT (NCCT) and CT angiography source images (CTA-SI) for infarct detection in patients with suspected acute ischemic PC stroke. Methods: This is a retrospective review of patients with suspected stroke in the PC in a database of our stroke center (n = 3,011) who underwent NCCT, CTA and CTP within 9 h after stroke onset and CT or MRI on follow-up. Images were evaluated for signs and pc-ASPECTS locations of ischemia. Three imaging models - A (NCCT), B (NCCT + CTA-SI) and C (NCCT + CTA-SI + CTP) - were compared with regard to the misclassification rate relative to gold standard (infarction in follow-up imaging) using the McNemar's test. Results: Of 3,011 stroke patients, 267 patients had a suspected stroke in the PC and 188 patients (70.4%) evidenced a PC infarct on follow-up imaging. The sensitivity of Model C (76.6%) was higher compared with that of Model A (21.3%) and Model B (43.6%). CTP detected significantly more ischemic lesions, especially in the cerebellum, posterior cerebral artery territory and thalami. Conclusions: Our findings in a large cohort of consecutive patients show that CTP detects significantly more ischemic strokes in the PC than CTA and NCCT alone.

© 2016 S. Karger AG, Basel


Background and Purpose

Computed tomography perfusion (CTP) has a high sensitivity and a very high specificity and, therefore, an additional diagnostic value for detection of acute ischemic stroke (AIS) in the anterior circulation [1,2]. However, no consensus could still be agreed upon imaging of stroke in the posterior circulation (PC). A recent study proposes an increase in diagnostic accuracy by additional CT whole brain volume perfusion to protocols including only CT angiography (CTA) and non-contrast CT (NCCT) [3]. Another blinded study recognized no significant difference in the detection of supratentorial and infratentorial stroke lesions [4].

Van der Hoeven et al. [3] report that the detection rate of ischemic lesions using CTP was significantly higher than by CTA and NCCT alone especially in the cerebellum, posterior cerebral artery territory and the thalami. However, patient populations were relatively small, and the diagnostic accuracy of the different diagnostic models was not tested and isolated for each pc-ASPECT's region. We therefore conducted a research containing a large number of consecutive patients to evaluate the additional diagnostic value of CTA and CTP in acute PC stroke. We hypothesized a significant increase in the detection rate of PC infarcts with an additional application of whole brain volume perfusion (CTP) to NCCT and CTA.

Methods

Patients

We retrospectively evaluated data of consecutive patients with suspected ischemic stroke admitted between January 1, 2011, and March 31, 2015, at a tertiary care center.

The inclusion criteria for this study were (1) suspected ischemic stroke of the PC as defined in the Oxfordshire classification [5]; (2) NCCT, CTA and CTP performed on admission; and (3) CT performed <9 h after symptom onset.

Exclusion criteria were (1) poor imaging quality and (2) not all regions of the PC included in the CT images or (3) missing follow-up imaging.

We evaluated the CT at admission and follow-up imaging, time from symptom onset to imaging and to groin puncture for every patient from Radiology Information System (RIS, Centricity RIS-IC, GE-Healthcare, Munich, Germany). In addition, we obtained the final diagnosis, National Institutes of Health Stroke Scale (NIHSS) scores and intravenous application of tissue plasminogen activator from neurology clinic records.

All patients were admitted to the stroke unit or neurological intensive care unit, and clinical signs for malignant PC infarction were determined at least every 8 h by an experienced neurologist. The NIHSS score was assessed at admission by an experienced stroke neurologist. All thrombectomies were carried out by experienced neuroradiologists; interpretation of CT, MRI and angiographic imaging results were carried out or supervised by experienced radiologists.

Imaging Protocol

NCCT, CTA and CTP at our institution is part of the routine protocol for patients with suspected stroke and admission <6 h after symptom onset or with an unknown time of symptom onset.

In cases of suspected ischemic stroke of the PC, time window can be up to <9 h after symptom onset. Allocation of PC strokes was defined according to pc-ASPECTS [6].

Follow-up imaging consisted of NCCT usually 2 days after admission, if clinically relevant earlier. Primary or additional MRI was performed in cases where only minor lesions in the PC (especially in the brain stem) were suspected.

CT scans were performed on a 128-slice dual-source CT scanner (Somatom Definition Flash; Siemens Medical Solutions, Forchheim, Germany).

Non-contract head images were obtained from the vertex to the skull base (120 kV, 340 mAs, 5.0-mm slice reconstruction, 1.0-mm increment, 0.6-mm collimation, 0.8 pitch and H30s soft kernel). This was followed by CTA (120 kV, 175 mAs, 1.0-mm slice reconstruction, 1-mm increment, 0.6-mm collimation, 0.8 pitch, H20f soft kernel, 80 ml Ultravist 370 and 50 ml NaCl flush at 4 ml/s, scan start 6 s after bolus tracking at the level of the ascending aorta).

Finally, a whole brain volume perfusion (CTP) was performed (4-dimensional adaptive spiral covering 96-mm scan length, 80 kV, 200 mAs, 5.0-mm slice reconstruction, 5-mm increment, 0.6-mm collimation, 0.8 pitch, H20f soft kernel, 30 ml Ultravist 370 and 30 ml NaCl flush at 6 ml/s, 45-second scan) and the cerebral blood volume (CBV), cerebral blood flow (CBF), time to drain (TTD) and mean transit time (MTT) maps were calculated (fig. 1).

Fig. 1

a, b Illustration of multimodal CT in PC stroke.

http://www.karger.com/WebMaterial/ShowPic/492314

MRI was used as follow-up imaging (fig. 1). The MRI standard protocol included fluid attenuated inversion recovery (FLAIR; TR/TE, 8,000/120; FOV 23 cm; 256 × 201 matrix), a time of flight angiography (TOF; TR/TE, 23/23,826; FOV 16 cm; 256 × 196 matrix) and diffusion-weighted imaging (DWI). Diffusion-weighted MRIs were acquired in the transverse plane using the single-shot echo planar imaging technique. The applied sequence parameters for DWI were the following: axial plane 4-mm section thickness, FOV of 240 mm and b-values of 0 and 1,000 s/mm2. MRI was performed on two 1.5-T MRI scanners (Philips Achieva Intera; Philips Medical Systems, Best, The Netherlands).

All images were evaluated by experienced neuroradiologists. At first, we used the original clinical reports for evaluation. If uncertainties occurred, the results were re-evaluated by experienced radiologists.

We assessed all 4 calculated maps (CBF, CBV, TTD, MTT) and if changes occurred in TTD and/or MTT, they were rated as infarcts. We also found TTD to be most specific as previously described in the literature [7]; however, usually changes in our study occurred in all 4 maps.

Diagnostic Models

Based on the imaging modalities NCCT, CTA source images (CTA-SI) and CTP, 3 nested diagnostic models were considered - Model A: NCCT; Model B: NCCT + CTA-SI (claiming infarction if NCCT and/or CTA-SI is positive); Model C: NCCT + CTA-SI + CTP (claiming infarction if NCCT and/or CTA-SI and/or CTP is positive).

Statistical Analysis

Univariable distribution of metric variables is described by median and interquartile range. For categorical data, absolute and relative frequencies are given. For each imaging modality (NCCT, CTA-SI, CTP) and for each diagnostic model (A, B, C), sensitivity, specificity, positive predictive value (PPV) and negative predictive value (NPV) are given with exact 95% CIs. The underlying gold standard is occurrence of infarction in follow-up imaging.

Primary end point for this study is the misclassification rate (MR). A misclassification is present whenever the prediction of an imaging modality or diagnostic model does not coincide with the gold standard. MR was calculated as the ratio of the number of patients with misclassifications and the overall number of patients.

The following null hypotheses were tested by 2-sided McNemar test on a 2-sided multiple significance level of 5%: the MR of Model A does not differ from the MR of Model B (and analogously for the pairwise comparisons Model A vs. C, Model B vs. C). These null hypotheses were tested within each localization (cerebellum, thalamus, PCA territory, pons/midbrain) as well as regarding the overall prediction (infarction in any localization), corresponding to a total of 15 null hypotheses. Adjustment for multiple testing is done by means of the Bonferroni-Holm method [8].

Results

A total of 3,011 patients were diagnosed with AIS from January 1, 2011, until March 31, 2015. Of these patients, 267 fulfilled the inclusion criteria (table 1).

Table 1

Baseline characteristics of the study population

http://www.karger.com/WebMaterial/ShowPic/492318

Admission NCCT detected 40 (21.3%), CTA-SI detected 82 (43.6%) and CTP detected 144 (76.6%) of the 188 patients with an infarct in the PC on follow-up imaging. Follow-up imaging revealed 330 infarcts in 1,104 analyzed regions.

All diagnostic modalities evidenced high PPVs (NCCT 93.0%, CTA-SI 97.6% and CTP 96.0%). NCCT detected 63 (19%), CTA-SI detected 124 (37.6%) and CTP detected 270 (82.0%) infarcts. CTP detected significantly more ischemic lesions in the cerebellum, posterior cerebral artery territory and thalami than NCCT and CTA-SI. The NPV of CTP (62.4%) was higher compared with that of NCCT (33.9%) and CTA-SI (42.1%). Lesion detection was the poorest in pons/midbrain and thalamus for NCCT (pons/midbrain: PPV 75% and thalamus 77.8%; table 2).

Table 2

Detection of infarcts in the PC for each imaging modality

http://www.karger.com/WebMaterial/ShowPic/492317

Sensitivity of Model C (NCCT + CTA-SI + CTP; 76.6%) was higher compared with that of Model A (NCCT; 21.3%) and Model B (NCCT + CTA-SI; 43.6%). Model A (96.2%) and Model B (96.2%) demonstrated a higher specificity than Model C (91.2%).

Model C achieved the greatest sensitivity in the pc-ASPECTS region cerebellum (87.2%), followed by PCA territory (88.5%), thalamus (69.7%) and pons/midbrain (60.0%) (table 3).

Table 3

Detection of infarcts in the PC for each diagnostic model

http://www.karger.com/WebMaterial/ShowPic/492316

Differences in the misclassification of infarction in follow-up imaging were significant in all pairwise comparisons of Models A, B and C, proving that the MR of Model C is significantly lower than that of Models A and B (p < 0.0001), and the MR of Model B is significantly lower than that of Model A (p < 0.0001; table 4).

Table 4

Pairwise comparison of diagnostic Model A (NCCT), Model B (NCCT + CTA-SI) and Model C (NCCT + CTA-SI + CTP)

http://www.karger.com/WebMaterial/ShowPic/492315

In the pairwise comparison of the imaging Models B and C, 57.7% infarcts were classified correctly by Model B and Model C, and 23.2% of the cases were classified correctly in Model C and misclassified in Model B.

A total of 41.9% of the infarcts were classified correctly by Models A and C, whereas 39.0% of the infarcts in follow-up imaging were predicted correctly by Model C and misclassified by Model A.

The pc-ASPECTS region pons/midbrain evidenced the largest rate of misclassifications in the pairwise comparison of all imaging models (AB: 25.4%, AC: 13.5% and BC: 13.5%; table 4).

Discussion

Our findings in a large cohort of stroke patients show that multimodal CT including whole brain volume perfusion detects significantly more ischemic strokes in the PC than the combined imaging model NCCT and CTA and NCCT alone.

To the best of our knowledge, this is so far the largest study investigating the detection of different pc-ASPECTS infarct regions by varied CT imaging models. We illustrated for the first time the diagnostic superiority of multimodal CT including CTP with the implementation of MRs. A misclassification was present whenever the prediction of an imaging modality or diagnostic model does not coincide with the gold standard: the occurrence of infarction in follow-up imaging. In our study, multimodal CT including CTP revealed the lowest MRs in all pc-ASPECTS regions.

Improving the detection rate of PC strokes is important, as every 5th ischemic stroke is located in the PC. Additionally, clinical detection of PC strokes is challenging because clinical signs of anterior and PC strokes do not have any obvious distinction [9]. Therefore, there is a need for the best outcome predicting neuroimaging model to obtain the optimal therapy.

Concordant to other, smaller studies, the infarct detection rate of multimodal CT including CTP was significantly higher in all regions of the PC according to pc-ASPECTS (cerebellum, posterior cerebral artery territory, thalamus, brain stem/midbrain) and, therefore, provides the best outcome prediction [6,10].

Detection of ischemic lesions in the brain stem is more challenging. In our study, the pc-ASPECTS region pons/midbrain evidenced the largest rate of misclassifications in all imaging models. However, the rate for detection of brain stem infarcts was slightly higher than in comparable studies [3,4]. This might be according to basilar occlusions (with a higher rate of CTP alterations in the brain stem), which have been moved to our hospital as a regional stroke center.

A possible limitation of our study is the retrospective design. False-positive results may have been caused by intravenous thrombolytic therapy [9] and endovascular thrombectomy as tissues at risk may not have progressed to definite infarctions.

Intravenous thrombolysis and thrombectomy may also have influenced sensitivity and specificity of the imaging models as it can cause a temporarily reperfusion. The higher rate of detected infarctions on follow-up imaging, especially in the brain stem, is best explained by the much higher rate of MRI as follow-up examination [11] compared to the other studies [3,4].

Conclusions

Our findings in a large cohort of consecutive patients show that CTP detects significantly more ischemic strokes in the PC than CTA and NCCT alone.

Clinical Relevance

Whole brain perfusion should be added to standard CT protocols for detection of ischemic stroke in the PC.

Source of Funding

Dr. U. Hanning receives a research grant of the Medical Faculty of the University of Muenster.

Disclosure Statement

All authors have no disclosures.


References

  1. Biesbroek JM, Niesten JM, Dankbaar JW, Biessels GJ, Velthuis BK, Reitsma JB, et al: Diagnostic accuracy of CT perfusion imaging for detecting acute ischemic stroke: a systematic review and meta-analysis. Cerebrovasc Dis 2013;35:493-501.
  2. Campbell BC, Christensen S, Levi CR, Desmond PM, Donnan GA, Davis SM, Parsons MW: Comparison of computed tomography perfusion and magnetic resonance imaging perfusion-diffusion mismatch in ischemic stroke. Stroke 2012;43:2648-2653.
  3. van der Hoeven EJ, Dankbaar JW, Algra A, Vos JA, Niesten JM, van Seeters T, et al: Additional diagnostic value of computed tomography perfusion for detection of acute ischemic stroke in the posterior circulation. Stroke 2015;46:1113-1115.
  4. Lee IH, You JH, Lee JY, Whang K, Kim MS, Kim YJ, et al: Accuracy of the detection of infratentorial stroke lesions using perfusion CT: an experimenter-blinded study. Neuroradiology 2010;52:1095-1100.
  5. Bamford J, Sandercock P, Dennis M, Burn J, Warlow C: Classification and natural history of clinically identifiable subtypes of cerebral infarction. Lancet 1991;337:1521-1526.
  6. Puetz V, Sylaja PN, Coutts SB, Hill MD, Dzialowski I, Mueller P, et al: Extent of hypoattenuation on CT angiography source images predicts functional outcome in patients with basilar artery occlusion. Stroke 2008;39:2485-2490.
  7. Thierfelder KM, Sommer WH, Baumann AB, Klotz E, Meinel FG, Strobl FF, et al: Whole-brain CT perfusion: reliability and reproducibility of volumetric perfusion deficit assessment in patients with acute ischemic stroke. Neuroradiology 2013;55:827-835.
  8. Holm S: A simple sequentially rejective multiple test procedure. Scand J Stat 1979;6:65-70.
  9. McKinney A, Truwit CL, Kieffer S: Reversibility of an ‘apparent' infarct on dynamic perfusion CT after lytic therapy: comment regarding cerebral blood flow and blood volume thresholds. AJNR Am J Neuroradiol 2006;27:1391-1392; author reply 1392-1393.
  10. Puetz V, Sylaja PN, Hill MD, Coutts SB, Dzialowski I, Becker U, et al: CT angiography source images predict final infarct extent in patients with basilar artery occlusion. AJNR Am J Neuroradiol 2009;30:1877-1883.
  11. Moreau F, Asdaghi N, Modi J, Goyal M, Coutts SB: Magnetic resonance imaging versus CT in transient ischemic attack and minor stroke: the more you see the more you know. Cerebrovasc Dis Extra 2013;3:130-136.

Author Contacts

Uta Hanning, MD and Peter Sporns, MD

Department of Clinical Radiology and Department of Epidemiology and Social Medicine, University Hospital of Muenster

Albert-Schweitzer-Campus 1, Gebäude D3, DE-48149 Muenster (Germany)

E-Mail uhanning@uni-muenster.de; Peter.Sporns@ukmuenster.de


Article / Publication Details

First-Page Preview
Abstract of Original Paper

Received: July 28, 2015
Accepted: December 21, 2015
Published online: January 29, 2016
Issue release date: April 2016

Number of Print Pages: 6
Number of Figures: 1
Number of Tables: 4

ISSN: 1015-9770 (Print)
eISSN: 1421-9786 (Online)

For additional information: https://www.karger.com/CED


Copyright / Drug Dosage / Disclaimer

Copyright: All rights reserved. No part of this publication may be translated into other languages, reproduced or utilized in any form or by any means, electronic or mechanical, including photocopying, recording, microcopying, or by any information storage and retrieval system, without permission in writing from the publisher.
Drug Dosage: The authors and the publisher have exerted every effort to ensure that drug selection and dosage set forth in this text are in accord with current recommendations and practice at the time of publication. However, in view of ongoing research, changes in government regulations, and the constant flow of information relating to drug therapy and drug reactions, the reader is urged to check the package insert for each drug for any changes in indications and dosage and for added warnings and precautions. This is particularly important when the recommended agent is a new and/or infrequently employed drug.
Disclaimer: The statements, opinions and data contained in this publication are solely those of the individual authors and contributors and not of the publishers and the editor(s). The appearance of advertisements or/and product references in the publication is not a warranty, endorsement, or approval of the products or services advertised or of their effectiveness, quality or safety. The publisher and the editor(s) disclaim responsibility for any injury to persons or property resulting from any ideas, methods, instructions or products referred to in the content or advertisements.

References

  1. Biesbroek JM, Niesten JM, Dankbaar JW, Biessels GJ, Velthuis BK, Reitsma JB, et al: Diagnostic accuracy of CT perfusion imaging for detecting acute ischemic stroke: a systematic review and meta-analysis. Cerebrovasc Dis 2013;35:493-501.
  2. Campbell BC, Christensen S, Levi CR, Desmond PM, Donnan GA, Davis SM, Parsons MW: Comparison of computed tomography perfusion and magnetic resonance imaging perfusion-diffusion mismatch in ischemic stroke. Stroke 2012;43:2648-2653.
  3. van der Hoeven EJ, Dankbaar JW, Algra A, Vos JA, Niesten JM, van Seeters T, et al: Additional diagnostic value of computed tomography perfusion for detection of acute ischemic stroke in the posterior circulation. Stroke 2015;46:1113-1115.
  4. Lee IH, You JH, Lee JY, Whang K, Kim MS, Kim YJ, et al: Accuracy of the detection of infratentorial stroke lesions using perfusion CT: an experimenter-blinded study. Neuroradiology 2010;52:1095-1100.
  5. Bamford J, Sandercock P, Dennis M, Burn J, Warlow C: Classification and natural history of clinically identifiable subtypes of cerebral infarction. Lancet 1991;337:1521-1526.
  6. Puetz V, Sylaja PN, Coutts SB, Hill MD, Dzialowski I, Mueller P, et al: Extent of hypoattenuation on CT angiography source images predicts functional outcome in patients with basilar artery occlusion. Stroke 2008;39:2485-2490.
  7. Thierfelder KM, Sommer WH, Baumann AB, Klotz E, Meinel FG, Strobl FF, et al: Whole-brain CT perfusion: reliability and reproducibility of volumetric perfusion deficit assessment in patients with acute ischemic stroke. Neuroradiology 2013;55:827-835.
  8. Holm S: A simple sequentially rejective multiple test procedure. Scand J Stat 1979;6:65-70.
  9. McKinney A, Truwit CL, Kieffer S: Reversibility of an ‘apparent' infarct on dynamic perfusion CT after lytic therapy: comment regarding cerebral blood flow and blood volume thresholds. AJNR Am J Neuroradiol 2006;27:1391-1392; author reply 1392-1393.
  10. Puetz V, Sylaja PN, Hill MD, Coutts SB, Dzialowski I, Becker U, et al: CT angiography source images predict final infarct extent in patients with basilar artery occlusion. AJNR Am J Neuroradiol 2009;30:1877-1883.
  11. Moreau F, Asdaghi N, Modi J, Goyal M, Coutts SB: Magnetic resonance imaging versus CT in transient ischemic attack and minor stroke: the more you see the more you know. Cerebrovasc Dis Extra 2013;3:130-136.
ppt logo Download Images (.pptx)


Figures
Thumbnail

Tables
Thumbnail
Thumbnail
Thumbnail
Thumbnail