Research ArticleNeurology

Determining Amyloid-β Positivity Using 18F-AZD4694 PET Imaging

Joseph Therriault, Andrea L. Benedet, Tharick A. Pascoal, Melissa Savard, Nicholas J. Ashton, Mira Chamoun, Cecile Tissot, Firoza Lussier, Min Su Kang, Gleb Bezgin, Tina Wang, Jaime Fernandes-Arias, Gassan Massarweh, Paolo Vitali, Henrik Zetterberg, Kaj Blennow, Paramita Saha-Chaudhuri, Jean-Paul Soucy, Serge Gauthier and Pedro Rosa-Neto
Journal of Nuclear Medicine February 2021, 62 (2) 247-252; DOI: https://doi.org/10.2967/jnumed.120.245209

Article Figures & Data

Figures

  • FIGURE 1.
    FIGURE 1.

    Transaxial (top) and midsagittal (bottom) representative 18F-AZD4694 SUVR PET images of 4 subjects representing range of binding patterns in present study. All images are presented in template space. MNI coordinates: x = 2, y = −59, z = 15. CDR = Clinical Dementia Rating; MMSE = Mini-Mental State Examination.

  • FIGURE 2.
    FIGURE 2.

    Means and SDs (error bars) in 18F-AZD4694 PET SUVR for CU young adults (age < 25 y), CU elderly, and CI groups. Young adults displayed minimal amyloid PET uptake (mean, 1.14; SD, 0.09). Dashed line represents 2 SDs above mean of young adults, at 1.33 18F-AZD4694 SUVR.

  • FIGURE 3.
    FIGURE 3.

    ROC curves contrasting visually positive vs. negative cases. (A) When contrasting CU elderly with AD dementia groups, we observed good AUC (82.5%; sensitivity, 85%; specificity, 73%). (B) Optimal threshold at this point was 1.56 SUVR, represented by dashed line. (C) Area under ROC curve contrasting visually negative vs. visually positive cases displayed excellent AUC (97%; sensitivity, 90.91%; specificity, 95%). (D) 18F-AZD4694 PET means are shown for visually positive (red) and visually negative (blue) groups, with dashed line representing optimal threshold derived from ROC curve (1.55 SUVR).

  • FIGURE 4.
    FIGURE 4.

    Gaussian mixture modeling representing 2 distributions. Low 18F-AZD4694 (red) and high 18F-AZD4694 (green) gaussian distributions are superimposed on subject density histogram for all 18F-AZD4694 PET SUVRs from CU elderly and CI populations. Optimal cut point from gaussian mixture modeling was 1.55 SUVR.

  • FIGURE 5.
    FIGURE 5.

    ROC curves contrasting CSF-positive vs. -negative individuals. (A) Area under ROC curve contrasting individuals dichotomized on basis of their cerebrospinal measure of Aβ42/Aβ40 ratio. This method resulted in area under ROC curve of 95% (sensitivity, 88.9%; specificity, 91.4%). (B) AZD4694 PET means are shown for CSF-negative (blue) and CSF-positive (red) individuals, with dashed line representing optimal threshold derived from ROC curve (1.51 SUVR).

Tables

  • TABLE 1

    Demographic, Clinical, and Biomarker Characteristics of Sample

    P*
    CharacteristicCU youngCU elderlyCICU youngvs.CU elderlyCU elderlyvs.CI
    Total patients (n)228965——
    Mean age (y)22.7 (SD, 1.3)72.33 (SD, 5.88)67.91 (SD, 8.97)<0.00010.0004
    Female (n)14 (63%)51 (57%)36 (55%)0.590.81
    Mean education (y)16.61 (SD, 1.33)15.06 (SD, 3.81)15.1 (SD, 3.34)0.060.94
    APOEε4 carriers (n)6 (27%)33 (37%)41 (63%)0.58<0.0001
    Mean MMSE29.77 (SD, 0.53)29.12 (SD, 1.07)24.03 (SD, 6.07)0.009<0.0001
    Mean neocortical 18F-AZD4694 SUVR1.14 (SD, 0.09)1.48 (0.38)2.04 (SD, 0.57)<0.0001<0.0001
    Mean CSF Aβ42/Aβ400.09 (SD, 0.006)0.07 (SD, 0.02)0.05 (SD, 0.02)<0.0001<0.0001

    • ↵* Assessed with 2-sided independent-samples t tests for each variable except sex and APOEε4 status, for which contingency χ2 tests were performed.

    • MMSE = Mini-Mental State Examination.

Additional Files

Back to top

In this issue

Print
Download PDF
Article Alerts
Email Article
Citation Tools
Bookmark this article

Jump to section

Keywords