APT and GluCEST Imaging at 5.0 T in Patients with Brain Tumors: A Phantom Reproducibility Validation and Clinical Study

Why It Matters

Accurate, non‑invasive glioma grading informs surgical planning and therapy selection; GluCEST offers a reliable imaging biomarker that could reduce dependence on invasive biopsies.

Key Takeaways

• APT and GluCEST at 5 T show ICCs >0.96, CVs <5%.
• GluCEST more sensitive to pH changes than APT.
• In 96 patients, GluCEST distinguishes high‑grade from low‑grade gliomas.
• Both techniques demonstrate high intra‑ and inter‑observer agreement (>0.93 ICC).
• GluCEST emerges as complementary biomarker for glioma grading.

Pulse Analysis

High‑field 5 Tesla MRI has opened new possibilities for metabolic imaging, and this study validates two advanced CEST techniques—APT and GluCEST—through rigorous phantom testing. By varying pH and protein concentration, the investigators demonstrated that both methods achieve intraclass correlation coefficients exceeding 0.96 and coefficients of variation below 5 %, establishing a benchmark for repeatability that is essential for any quantitative imaging biomarker. Such reproducibility ensures that longitudinal studies and multi‑center trials can compare results with confidence, a prerequisite for clinical translation.

In the patient cohort of 96 individuals with diverse brain tumors, both APT and GluCEST produced consistent measurements across readers, with inter‑observer ICCs above 0.93. Importantly, GluCEST showed a stronger correlation with tumor grade: high‑grade gliomas exhibited markedly higher GluCEST-derived difference and change‑ratio metrics than low‑grade lesions, a distinction that APT metrics largely missed after statistical correction. This heightened sensitivity to glutamate‑related biochemical alterations suggests that GluCEST can serve as a non‑invasive surrogate for tumor aggressiveness, potentially guiding neurosurgeons and oncologists in treatment planning.

The implications extend beyond diagnostic accuracy. By providing a reproducible, quantitative readout of tumor metabolism, GluCEST could complement conventional MRI, MR spectroscopy, and PET imaging, offering a faster, radiation‑free alternative for glioma grading. Future work may focus on integrating GluCEST into routine neuro‑oncology protocols, exploring its predictive value for therapy response, and standardizing acquisition across vendors. As the field moves toward precision medicine, reliable metabolic imaging biomarkers like GluCEST are poised to become integral tools in personalized brain‑tumor care.