Additive value of [18F]PI-2620 perfusion imaging in progressive supranuclear palsy and corticobasal syndrome

Abstract

Purpose: Early after [¹⁸F]PI-2620 PET tracer administration, perfusion imaging has potential for regional assessment of neuronal injury in neurodegenerative diseases. This is while standard late-phase [¹⁸F]PI-2620 tau-PET is able to discriminate the 4-repeat tauopathies progressive supranuclear palsy and corticobasal syndrome (4RTs) from disease controls and healthy controls. Here, we investigated whether early-phase [¹⁸F]PI-2620 PET has an additive value for biomarker based evaluation of 4RTs.

Methods: Seventy-eight patients with 4RTs (71 ± 7 years, 39 female), 79 patients with other neurodegenerative diseases (67 ± 12 years, 35 female) and twelve age-matched controls (69 ± 8 years, 8 female) underwent dynamic (0-60 min) [¹⁸F]PI-2620 PET imaging. Regional perfusion (0.5-2.5 min p.i.) and tau load (20-40 min p.i.) were measured in 246 predefined brain regions [standardized-uptake-value ratios (SUVr), cerebellar reference]. Regional SUVr were compared between 4RTs and controls by an ANOVA including false-discovery-rate (FDR, p < 0.01) correction. Hypoperfusion in resulting 4RT target regions was evaluated at the patient level in all patients (mean value - 2SD threshold). Additionally, perfusion and tau pattern expression levels were explored regarding their potential discriminatory value of 4RTs against other neurodegenerative disorders, including validation in an independent external dataset (n = 37), and correlated with clinical severity in 4RTs (PSP rating scale, MoCA, activities of daily living).

Results: Patients with 4RTs had significant hypoperfusion in 21/246 brain regions, most dominant in thalamus, caudate nucleus, and anterior cingulate cortex, fitting to the topology of the 4RT disease spectrum. However, single region hypoperfusion was not specific regarding the discrimination of patients with 4RTs against patients with other neurodegenerative diseases. In contrast, perfusion pattern expression showed promise for discrimination of patients with 4RTs from other neurodegenerative diseases (AUC: 0.850). Discrimination by the combined perfusion-tau pattern expression (AUC: 0.903) exceeded that of the sole tau pattern expression (AUC: 0.864) and the discriminatory power of the combined perfusion-tau pattern expression was replicated in the external dataset (AUC: 0.917). Perfusion but not tau pattern expression was associated with PSP rating scale (R = 0.402; p = 0.0012) and activities of daily living (R = - 0.431; p = 0.0005).

Conclusion: [¹⁸F]PI-2620 perfusion imaging mirrors known topology of regional hypoperfusion in 4RTs. Single region hypoperfusion is not specific for 4RTs, but perfusion pattern expression may provide an additive value for the discrimination of 4RTs from other neurodegenerative diseases and correlates closer with clinical severity than tau pattern expression.

Keywords: Neuronal injury; PET; Perfusion; Tau; [18F]PI-2620.

Fig. 1

Regional pattern of perfusion alterations in 4RTs when compared to healthy controls. Top: Percentage differences of perfusion imaging are illustrated without significance thresholding for cortical (surface projections) and subcortical (axial slices) Brainnetome regions. Bottom: Percentage differences of late-phase [18F]PI-2620 are illustrated for comparison purpose and reflected previously published data [1, 2]

Fig. 2

Performance of a single region hypoperfusion classifier for detection of 4RTs (n = 78) in contrast to other neurodegenerative disorders (n = 79). A single perfusion image was judged as a 4RT-like pattern when one of 21 target regions (Table 2) was below the mean value – 2 SD of healthy controls. Bootstrapping was performed with three and five regions as well as with − 2.5 and − 3.0 SD thresholds. PPV = positive predictive value; NPV = negative predictive value

Fig. 3

Exploratory discrimination of 4RTs and other neurodegenerative disorders by perfusion and tau pattern expression. The LMU sample consisted of n = 78 4RTs and n = 79 other neurodegenerative disorders (A) whereas the independent validation cohort consisted of n = 21 4RTs and n = 16 other neurodegenerative disorders (B). Area under the curve (AUC) values were determined using a receiver operating curve (ROC) analysis. * p < 0.05 for non-parametric comparison of ROC curves

Fig. 4

Associations of perfusion (top) and tau pattern expression (bottom) with clinical severity of 4RTs. Pearson’s correlation coefficients (R) derived from patients with 4RTs with available PSP rating scale data (n = 65, A, D), Schwab and England activities of daily living (SEADL, n = 66, B, E) and Montreal cognitive assessment (MoCA, n = 69, D, F). Correlations were controlled for age, sex, and disease duration (symptom onset)