The value of subtraction MRI in detection of amyloid-related imaging abnormalities with oedema or effusion in Alzheimer’s patients: An interobserver study

Roland M. Martens, Arianne Bechten, Silvia Ingala, Ronald A. van Schijndel, Vania B. Machado, Marcus C. de Jong, Esther Sanchez, Derk Purcell, Michael H. Arrighi, Robert H. Brashear, Mike P. Wattjes, Frederik Barkhof

Research output: Contribution to journalArticleAcademicpeer-review

Abstract

Background: Immunotherapeutic treatments targeting amyloid-β plaques in Alzheimer’s disease (AD) are associated with the presence of amyloid-related imaging abnormalities with oedema or effusion (ARIA-E), whose detection and classification is crucial to evaluate subjects enrolled in clinical trials. Purpose: To investigate the applicability of subtraction MRI in the ARIA-E detection using an established ARIA-E-rating scale. Methods: We included 75 AD patients receiving bapineuzumab treatment, including 29 ARIA-E cases. Five neuroradiologists rated their brain MRI-scans with and without subtraction images. The accuracy of evaluating the presence of ARIA-E, intraclass correlation coefficient (ICC) and specific agreement was calculated. Results: Subtraction resulted in higher sensitivity (0.966) and lower specificity (0.970) than native images (0.959, 0.991, respectively). Individual rater detection was excellent. ICC scores ranged from excellent to good, except for gyral swelling (moderate). Excellent negative and good positive specific agreement among all ARIA-E imaging features was reported in both groups. Combining sulcal hyperintensity and gyral swelling significantly increased positive agreement for subtraction images. Conclusion: Subtraction MRI has potential as a visual aid increasing the sensitivity of ARIA-E assessment. However, in order to improve its usefulness isotropic acquisition and enhanced training are required. The ARIA-E rating scale may benefit from combining sulcal hyperintensity and swelling. Key Points: • Subtraction technique can improve detection amyloid-related imaging-abnormalities with edema/effusion in Alzheimer’s patients.• The value of ARIA-E detection, classification and monitoring using subtraction was assessed.• Validation of an established ARIA-E rating scale, recommendations for improvement are reported.• Complementary statistical methods were employed to measure accuracy, inter-rater-reliability and specific agreement.

Original languageEnglish
Pages (from-to)1-12
Number of pages12
JournalEuropean Radiology
Volume28
Issue number3
DOIs
Publication statusPublished - Mar 2018

Cite this

@article{568ea189d6ae418eb52312e950f28da9,
title = "The value of subtraction MRI in detection of amyloid-related imaging abnormalities with oedema or effusion in Alzheimer’s patients: An interobserver study",
abstract = "Background: Immunotherapeutic treatments targeting amyloid-β plaques in Alzheimer’s disease (AD) are associated with the presence of amyloid-related imaging abnormalities with oedema or effusion (ARIA-E), whose detection and classification is crucial to evaluate subjects enrolled in clinical trials. Purpose: To investigate the applicability of subtraction MRI in the ARIA-E detection using an established ARIA-E-rating scale. Methods: We included 75 AD patients receiving bapineuzumab treatment, including 29 ARIA-E cases. Five neuroradiologists rated their brain MRI-scans with and without subtraction images. The accuracy of evaluating the presence of ARIA-E, intraclass correlation coefficient (ICC) and specific agreement was calculated. Results: Subtraction resulted in higher sensitivity (0.966) and lower specificity (0.970) than native images (0.959, 0.991, respectively). Individual rater detection was excellent. ICC scores ranged from excellent to good, except for gyral swelling (moderate). Excellent negative and good positive specific agreement among all ARIA-E imaging features was reported in both groups. Combining sulcal hyperintensity and gyral swelling significantly increased positive agreement for subtraction images. Conclusion: Subtraction MRI has potential as a visual aid increasing the sensitivity of ARIA-E assessment. However, in order to improve its usefulness isotropic acquisition and enhanced training are required. The ARIA-E rating scale may benefit from combining sulcal hyperintensity and swelling. Key Points: • Subtraction technique can improve detection amyloid-related imaging-abnormalities with edema/effusion in Alzheimer’s patients.• The value of ARIA-E detection, classification and monitoring using subtraction was assessed.• Validation of an established ARIA-E rating scale, recommendations for improvement are reported.• Complementary statistical methods were employed to measure accuracy, inter-rater-reliability and specific agreement.",
keywords = "Alzheimer's disease (AD), Amyloid beta (Aβ), ARIA (amyloid-related imaging abnormalities), Immunotherapy, MRI (magnetic resonance imaging)",
author = "Martens, {Roland M.} and Arianne Bechten and Silvia Ingala and {van Schijndel}, {Ronald A.} and Machado, {Vania B.} and {de Jong}, {Marcus C.} and Esther Sanchez and Derk Purcell and Arrighi, {Michael H.} and Brashear, {Robert H.} and Wattjes, {Mike P.} and Frederik Barkhof",
year = "2018",
month = "3",
doi = "10.1007/s00330-017-5022-6",
language = "English",
volume = "28",
pages = "1--12",
journal = "European Radiology",
issn = "0938-7994",
publisher = "Springer Verlag",
number = "3",

}

The value of subtraction MRI in detection of amyloid-related imaging abnormalities with oedema or effusion in Alzheimer’s patients : An interobserver study. / Martens, Roland M.; Bechten, Arianne; Ingala, Silvia; van Schijndel, Ronald A.; Machado, Vania B.; de Jong, Marcus C.; Sanchez, Esther; Purcell, Derk; Arrighi, Michael H.; Brashear, Robert H.; Wattjes, Mike P.; Barkhof, Frederik.

In: European Radiology, Vol. 28, No. 3, 03.2018, p. 1-12.

Research output: Contribution to journalArticleAcademicpeer-review

TY - JOUR

T1 - The value of subtraction MRI in detection of amyloid-related imaging abnormalities with oedema or effusion in Alzheimer’s patients

T2 - An interobserver study

AU - Martens, Roland M.

AU - Bechten, Arianne

AU - Ingala, Silvia

AU - van Schijndel, Ronald A.

AU - Machado, Vania B.

AU - de Jong, Marcus C.

AU - Sanchez, Esther

AU - Purcell, Derk

AU - Arrighi, Michael H.

AU - Brashear, Robert H.

AU - Wattjes, Mike P.

AU - Barkhof, Frederik

PY - 2018/3

Y1 - 2018/3

N2 - Background: Immunotherapeutic treatments targeting amyloid-β plaques in Alzheimer’s disease (AD) are associated with the presence of amyloid-related imaging abnormalities with oedema or effusion (ARIA-E), whose detection and classification is crucial to evaluate subjects enrolled in clinical trials. Purpose: To investigate the applicability of subtraction MRI in the ARIA-E detection using an established ARIA-E-rating scale. Methods: We included 75 AD patients receiving bapineuzumab treatment, including 29 ARIA-E cases. Five neuroradiologists rated their brain MRI-scans with and without subtraction images. The accuracy of evaluating the presence of ARIA-E, intraclass correlation coefficient (ICC) and specific agreement was calculated. Results: Subtraction resulted in higher sensitivity (0.966) and lower specificity (0.970) than native images (0.959, 0.991, respectively). Individual rater detection was excellent. ICC scores ranged from excellent to good, except for gyral swelling (moderate). Excellent negative and good positive specific agreement among all ARIA-E imaging features was reported in both groups. Combining sulcal hyperintensity and gyral swelling significantly increased positive agreement for subtraction images. Conclusion: Subtraction MRI has potential as a visual aid increasing the sensitivity of ARIA-E assessment. However, in order to improve its usefulness isotropic acquisition and enhanced training are required. The ARIA-E rating scale may benefit from combining sulcal hyperintensity and swelling. Key Points: • Subtraction technique can improve detection amyloid-related imaging-abnormalities with edema/effusion in Alzheimer’s patients.• The value of ARIA-E detection, classification and monitoring using subtraction was assessed.• Validation of an established ARIA-E rating scale, recommendations for improvement are reported.• Complementary statistical methods were employed to measure accuracy, inter-rater-reliability and specific agreement.

AB - Background: Immunotherapeutic treatments targeting amyloid-β plaques in Alzheimer’s disease (AD) are associated with the presence of amyloid-related imaging abnormalities with oedema or effusion (ARIA-E), whose detection and classification is crucial to evaluate subjects enrolled in clinical trials. Purpose: To investigate the applicability of subtraction MRI in the ARIA-E detection using an established ARIA-E-rating scale. Methods: We included 75 AD patients receiving bapineuzumab treatment, including 29 ARIA-E cases. Five neuroradiologists rated their brain MRI-scans with and without subtraction images. The accuracy of evaluating the presence of ARIA-E, intraclass correlation coefficient (ICC) and specific agreement was calculated. Results: Subtraction resulted in higher sensitivity (0.966) and lower specificity (0.970) than native images (0.959, 0.991, respectively). Individual rater detection was excellent. ICC scores ranged from excellent to good, except for gyral swelling (moderate). Excellent negative and good positive specific agreement among all ARIA-E imaging features was reported in both groups. Combining sulcal hyperintensity and gyral swelling significantly increased positive agreement for subtraction images. Conclusion: Subtraction MRI has potential as a visual aid increasing the sensitivity of ARIA-E assessment. However, in order to improve its usefulness isotropic acquisition and enhanced training are required. The ARIA-E rating scale may benefit from combining sulcal hyperintensity and swelling. Key Points: • Subtraction technique can improve detection amyloid-related imaging-abnormalities with edema/effusion in Alzheimer’s patients.• The value of ARIA-E detection, classification and monitoring using subtraction was assessed.• Validation of an established ARIA-E rating scale, recommendations for improvement are reported.• Complementary statistical methods were employed to measure accuracy, inter-rater-reliability and specific agreement.

KW - Alzheimer's disease (AD)

KW - Amyloid beta (Aβ)

KW - ARIA (amyloid-related imaging abnormalities)

KW - Immunotherapy

KW - MRI (magnetic resonance imaging)

UR - http://www.scopus.com/inward/record.url?scp=85030118094&partnerID=8YFLogxK

U2 - 10.1007/s00330-017-5022-6

DO - 10.1007/s00330-017-5022-6

M3 - Article

VL - 28

SP - 1

EP - 12

JO - European Radiology

JF - European Radiology

SN - 0938-7994

IS - 3

ER -