Introduction

Osteoarthritis (OA) has a substantial impact on affected individuals and on the healthcare system as a whole. While diagnostic imaging of OA is widely practiced, structural degradation is often poorly correlated with pain reported by individuals. Acidification is a major component in the progression of pain,^1,2 which could help localize and potentially be corrected to treat patient discomfort.

Non-invasive, pH-sensitive imaging is possible with chemical exchange saturation transfer (acidoCEST) MRI. Recently, Ma et al.³ modified acidoCEST MRI for use in short T2 musculoskeletal tissues such as cartilage and meniscus by adding a 3D ultrashort echo time (UTE) readout to the acidoCEST sequence. This method has been used successfully to measure pH in ex vivo tissue phantoms as well as in vivo.⁴ Due to the difficulty of validating pH in vivo, we have conducted repeated measurements in ex vivo tissue to test the stability and accuracy of acidoCEST-UTE MRI.

The purpose of this study was to verify the reproducibility of the acidoCEST-UTE sequences in musculoskeletal tissue for pH measurement.

Methods

Phantom Preparation:

Thirty-three syringes were filled with either iohexol or iopamidol at 200 mM concentration and PBS at 10 mM and pH-adjusted in a range from 6.2 to 7.8. Coins of pure cartilage and meniscus were harvested from grossly normal tissues from ten donors (nine male, one female; age range 25-80), soaked for 24 hours in the same buffered contrast agent solutions as described above, and included in the syringes. Syringes were grouped into three sets and each set was placed in an agar-filled container (henceforth referred to as Phantom 1-3) and scanned parallel to B₀ at isocenter. All three Phantoms were imaged with the same sequence on three different days.

MRI and Analysis:

Liquid and tissue phantoms were imaged on the same 3T clinical scanner (MR750, GE Healthcare, Milwaukee, WI) using an 8-channel transmit/receive knee coil.

The 3D acidoCEST-UTE-Cones sequence³ parameters are shown as follows: TR=62ms, TE=0.032ms, Nsp=5, spoke interval τ=5 ms, FA=5°, BW=166 kHz, FOV=12×12×11.2cm³, matrix=160×160×28. CEST contrast was created with a Fermi pulse: duration=32ms, bandwidth=40Hz, two average B1s=5.4/1.1µT, 61 frequency offsets from -1200 to 1200 Hz in 40 Hz increments, with 1min 10s scan time per frequency. Steady-state before acquisition was achieved using an 8s dummy scan.

AcidoCEST MRI data was analyzed using ratio of power mismatch (RPM) based on distinct RF-powers (0.54 and 1.10 µT)⁵ at the 4.3 ppm peak for iohexol or the 4.2 ppm peak for iopamidol. For individual phantoms, correlations were preformed between pH and RPM for each contrast and tissue. In addition, pH was modeled as a function of RPM for the three phantoms combined, goodness of fit was measured, and a 95% confidence ellipse plot was generated.

Results

The relationship between RPM and pH shows good agreement across all three phantoms for both iohexol and iopamidol (Figure 1). R² values for each phantom are given in Figure 2. For the three phantoms combined, excellent goodness of fit was demonstrated (iohexol R² range 0.854-0.946, iopamidol R² range 0.792-0.934), indicating strong reproducibility. The 95% confidence ellipses are shown in Figure 3. For both iohexol and iopamidol, confidence intervals overlap between liquid, cartilage, and meniscus, indicating no statistically significant difference.

Discussion and Conclusions

In vivo pH measurements are problematic to accurately verify. Imaging methods such as acidoCEST MRI must therefore first be validated ex vivo to demonstrate reproducibility and accuracy of their measurements. In this study, we show the repeatability of acidoCEST-UTE MRI across three individually made phantoms containing cartilage, meniscus, and iohexol or iopamidol.

Acknowledgements

The authors gratefully acknowledge grant support from the VA Rehabilitation R&D Service (I01RX002604), VA Clinical Science R&D Service (I01CX001388) and NIH (R21AR073496, R01AR062581, R01AR075825, and 1R01AR068987).