Pathological tissue changes in brain tumors affect the pH‐sensitivity of the T1‐corrected apparent exchange dependent relaxation (AREX) of the amide protons

Measuring the intracellular pH (pHi) is of interest for brain tumor diagnostics. Common metrics of CEST imaging like the amide proton transfer‐weighted (APTw) MTRasym are pHi sensitive and allow differentiating malignant tumor from healthy tissue. Yet, the image contrast also depends on additional magnetization transfer effects and T1. In contrast, the apparent exchange‐dependent relaxation (AREX) provides a T1 corrected exchange rate of the amide protons. As AREX still depends on amide proton density, its pHi sensitivity remains ambiguous. Hence, we conducted this study to assess the influence of pathologic tissue changes on the pHi sensitivity of AREX in vivo. Patients with newly diagnosed intra‐axial brain tumors were prospectively recruited and underwent conventional MRI, quantitative T1 relaxometry, APT‐CEST and 31P‐MRS on a 3T MRI scanner. Tumors were segmented into contrast‐enhancing tumor (CE), surrounding T2 hyperintensity (T2‐H) and contralateral normal appearing white matter (CNAWM). T1 mapping and APT‐CEST metrics were correlated with 31P‐MRS‐derived pHi maps (Pearson's correlation). Without differentiating tissue subtypes, pHi did not only correlate significantly with MTRasym (r = 0.46) but also with T1 (r = 0.49). Conversely, AREX only correlated poorly with pHi (r = 0.17). Analyzing different tissue subtypes separately revealed a tissue dependency of the pHi sensitivity of AREX with a significant correlation (r = 0.6) in CNAWM and no correlation in T2‐H or CE (r = −0.11/−0.24). CE showed significantly increased MTRasym, pHi, and T1 compared with CNAWM (p