Reproducible topographies of metabolic brain networks in American and Chinese patients with progressive supranuclear palsy

Objectives We have previously observed that brain network activity for progressive supranuclear palsy (PSP)-related covariance pattern (PSPRP) identified with FDG PET in American patients [1] was similarly expressed in independent Chinese patients scanned on a different PET camera. In this study, we cross-validated the two brain networks produced separately in these cohorts of clinically-confirmed patients with PSP.

Methods The American cohort consisted of 10 PSP patients and 10 age-matched healthy controls scanned on a GE Advance PET camera. The sample size was the same in the Chinese cohort scanned on a Siemens Biogragh 64 PET/CT system in China. Using FDG PET images and covariance analysis we derived PSPRP in each cohort and computed the corresponding network scores prospectively in the other cohort as described previously [2-3].

Results Both cohorts generated similar PSPRPs characterized by metabolic decreases in the medial prefrontal cortex/cingulate, ventrolateral prefrontal cortex, striatum, medial thalamus, and midbrain, along with covarying metabolic increases in the hippocampus and parieto-temporal regions (Figure 1). PSPRP scores were similarly elevated (P < 0.0001) in the patients relative to the controls in the derivation cohort in the USA and in the validation cohort in China or vice versa. PSPRP scores correlated strongly (R > 0.96; P<0.001) in the two corresponding cohorts of patients and healthy controls from the USA and Chinese sites, respectively.

Conclusions We demonstrated that topographies and network scores of disease-related patterns were fully comparable and reproducible in individual cohorts of PSP patients and healthy controls across the centers in USA and China. These findings resembled those published in an independent European study [4]. These results support the notion that this characteristic covariance pattern may provide a viable network biomarker regardless of patient cohorts and imaging systems.

Research Support The work at the US site was supported by the US-China Biomedical Collaborative Research Program (R01 NS083490) and the Morris K Udall Center of Excellence for Parkinson’s Disease Research (P50 NS071675) from the National Institute of Health. The work at the Chinese site was supported by the China-US Biomedical Collaborative Research Program (No. 81361120393) and other grants (No. 81071018, No. 81171189, No. 81301136 and No. 81371413) from the National Natural Science Foundation of China.