As an essential organ of the urinary system, kidney plays an important role in regulating pH homeostasis and removing waste products of metabolism. Renal pH alters in pathological states, such as ischemia, inflammation, infection, cancer and so on, which therefore can serve as a sensitive micro-environmental marker for noninvasive detection and assessment of renal injuries and diseases. Recently, Iopamidol, a clinically approved CT contrast agent carrying two exchangeable groups of amide protons at chemical shifts of 4.3 and 5.5 ppm, has been utilized for renal pH measurement [1]. Ratiometric pH imaging shows accurate pH quantification between the pH ranges of 5.5-7.4 at 7 T, which provides a minimally-invasive way for renal pH characterization [2]. However, the saturation transfer (ST) effects of the two exchangeable proton groups would substantially overlap at lower magnetic fields (e.g., 3 and 4.7 T), making it challenging to translate renal pH measurement to the clinical setting. In this study, we investigated a Lorentzian-based decoupling algorithm to resolve the two ST effects for enhanced ratiometric pH measurement and preliminarily evaluated renal pH imaging at 4.7 T in rodents.

MRI: All experiments have been approved by the institutional animal care and use committee. A phantom containing six iopamidol solution vials (40 mM) with pH titrated to 5.5, 6.0, 6.5, 7.0, 7.5 and 8.0 was prepared and imaged at body temperature. Three adult Wistar rats were anesthetized and their kidneys were scanned along the long axis with respiratory triggering. Iopamidol (Isovue 370, 1.5 mg I/g) was injected via tail vein at a rate of 18 mL/hr. Z-spectra were acquired at frequency offsets ranging between -7 and 7 ppm, using a single-shot SE-EPI sequence preceded by a 2 μT saturation pulse. The imaging parameters were: matrix size = 48×48, FOV = 2×2 cm², TR/TE and the frequency step were 10 s/100 ms and 0.25 ppm for the phantom, and 3 s/18 ms and 0.125 ppm for the in vivo scan, respectively.

Data analysis: After B0 inhomogeneity correction, the Z-spectra were flipped to be 100*(1-M_z/M₀) and then fitted with Lorentizan function on a pixel basis using a model of six pools (i.e., NOE, MT, water, -OH, and two exchangeable amide proton groups of Iopamidol) as described in [3]. Ratiometric measurement was performed by taking the amplitude ratio of CEST peaks at 5.5 and 4.3 ppm (i.e., ST_{5.5 ppm}/ST_{4.3 ppm}). pH calibration curve was obtained by two-order polynomial fitting of pH dependence of ratiometric indices from the six phantoms [4], based on which kidney pH was quantified pixel-by-pixel. Three layers of cortex, medulla and calyx, with respectively denoted as region 1, 2 and 3, were outlined manually on a CEST-weighted image for each kidney (Fig. 1). Mean pH value was measured for each layer and one-way ANOVA with Bonferroni’s correction was performed to test pH difference across the three regions with P<0.05 regarded as statistically different.

Fig. 2 shows that the routine ratiometric analysis of iopamidol phantom without resolving two ST effects resulted in a much smaller dynamic pH range, between 5.5 and 7.0 (black line). In comparison, good polynomial relationship between the ratiometric index and pH was exhibited for pH range of 5.5-7.5 using the decoupling analysis (red line). The increase of pH sensitive range is critical for mapping renal pH, which could vary from 6 to 7.5. Fig. 3 illustrates a representative renal pH map, and the pH values were found to gradually decrease from cortex, medulla to calyx. Specifically, mean pH values were 7.29±0.03, 7.17±0.02 and 6.56±0.02 for cortex, medulla and calyx, respectively. Significantly difference was exhibited across all the three layers (P<0.05).In this study, we demonstrated that ratiometric analysis of resolved CEST effects of Iopamidol presents substantially enhanced range and sensitivity of pH measurements, which is crucial for accurate renal pH mapping. Renal pH maps obtained with this method are in good agreement with the results measured using contrast-enhanced MRI technique [5], showing its feasibility in pH quantification. In conclusion, our proposed method provides a novel way for reliable renal pH mapping, which especially benefits pH quantification at clinical field strengths.