Reproducibility of 3T APT-CEST in healthy volunteers and brain glioma patients

BACKGROUND
Amide proton transfer (APT) imaging is a chemical exchange saturation transfer (CEST) technique offering potential clinical applications in patients with brain tumors.

PURPOSE
To investigate whether cerebral APT-CEST is sufficiently reproducible in healthy tissue and glioma for clinical use at 3T.

STUDY TYPE
Prospective, longitudinal

SUBJECTS
Twenty-one healthy volunteers (M:F = 10:11; age 39±11 years) and six glioma patients (M:F = 3:3; age 50±17 years: four glioblastomas, one oligodendroglioma, one suspected low-grade glioma).

FIELD STRENGTH/SEQUENCE
3T, SPACE-CEST

ASSESSMENT
APT-CEST measurement reproducibility was assessed within-session (glioma patients, healthy volunteers), and between-sessions and between-days (healthy volunteers). The standard deviation of the within-subject difference (SDdiff) was calculated in tumor regions of interest (ROI), and eight ROIs at relevant locations including a whole-brain ROI.

STATISTICAL TESTS
Brown-Forsythe tests and variance component analyses (VCA) were used to assess the reproducibility of ROIs for the three reproducibility time intervals. Intraclass correlation coefficient (ICC) was used to assess agreement between the ROIs for the three reproducibility time intervals.

RESULTS
APT-CEST magnetization transfer ratio asymmetry (MTRasym) was 0.89±0.96% on average in healthy brain tissue and 1.59±0.67% in tumor tissue. Intratumoral mean MTRasym was significantly higher than MTRasym in healthy-appearing tissue in patients (0.5±0.46%; P< 0.001). The APTCEST difference between GBMs and contralateral tissue was 1.11%. The average within-session, between-sessions, and between-days SDdiff of healthy control brains was 0.2%. The within-session SDdiff of whole-brain was 0.2% in both healthy volunteers and patients, and 0.21% in the segmented tumor. The orbitofrontal gyri were the ROI with the highest within-session SDdiff (0.61%). Within-session reproducibility of ROIs did not differ significantly from between-sessions or between-day reproducibility (0.76>P>0.22) and VCA showed that within-session variance was the most important factor (60%), but differed from between-days reproducibility in putamen and the central brain (P<0.05). ICC within-session agreement was excellent for glioma (ICC=0.97) and healthy brain in volunteers (ICC=0.81). The effect size of the APTCEST between healthy brain and GBM was 5.

CONCLUSION
Cerebral APT-CEST imaging has good scan-rescan reproducibility in healthy tissue and tumors with clinically feasible scan times at 3T. Short-term measurement effects are dominant components in reproducibility.