Acidosis of the tumor microenvironment is an essential characteristic that promotes tumor development and progression and affects treatment response.[1, 2] There is limited application of the published 31P MRS/MRIS, 1H MRSI, and other methods to measure the pHe. [3] The pH-dependent protonation of a T1-based MRI contrast agent can be used to measure the pHe.[4, 5] However, the T1-based MR contrast is also dependent on agent concentration, confounding the measurement of pHe. We developed PET/MRI co-agents where the PET imaging can determine concentration of the co-agents, which is used to remove this confound when measuring pHe with the MRI agent Also, our next-generation co-agents are more hydrophilic and have higher chelation stability in vivo than our previous contrast reagents thus offering a promising tool for clinical applications. [6]
For the synthesis of both reagents, the amide group was introduced by substitution of 4-methoxybenzenesulfonyl chloride with 2-bromoethylamine hydrobromide. The amide arm was added onto the unprotected amide of hexamethyl 2,2',2''-(1,4,7,10-tetraazacyclododecane-1,4,7-triyl) triadipate followed by the deprotection of the carboxyl methyl ester with lithium hydroxide (Figure 1). The resulting free acid was used to chelate gadolinium (MRI) and 68Ga (PET) separately to obtain the MRI/PET co-agent pair. The overall yield is around 42%. We performed the kinetic PET studies in a 4T1 breast cancer model (Figure 4). The PET reagent was stable for one hour, enough for our PET/MRI co-agent imaging protocol.
Our MRI agent can detect a pH difference of 0.2 pH units in the phantom with relatively high sensitivity (Figure 2). The pHe of 4T1 breast cancer with and without bicarbonate treatment was measured by our PET/MRI contrast reagent. The r1 relaxivity of the MRI co-agent was determined by the ratio of the measurement of PET radioactivity and MRI ∆R1 relaxation rate. The dynamic changes in ∆R1 relaxation rate and PET-derived MRI co-agent concentration were monitored at 7.6 sec and 2-min temporal resolution, respectively as averaged for the tumor region in the image. The r1 relaxivity was determined from ∆R1 and concentration from each two-minute time interval and converted to pHe based on the pH-relaxivity calibration (Figure 3). This pHe was found to be consistent for five measurements, made 20 minutes after s.c. injection, presumably when the MRI co-agent had accumulated in the tumor to a sufficient concentration for areliable measurement. The average of these five pHe measurements after 20 minutes post injection was 6.84 ± 0.05 pH units which was similar to the gold standard pH microsensor 6.76 ± 0.10 pH units.
These preliminary results showed that tumor acidosis can be evaluated with simultaneous ⁶⁸Ga-PET/ Gd-MRI co-agents.
We are currently synthesizing other PET reagents with different isotopes for radiolabeling, such as ⁸⁶Y and ¹⁸F. The best candidate showing the best pharmacokinetic stability and highest accuracy in pH measurement will be tested with other cancer treatment. Our PET/MRI co-inject reagent will become the indicator for tumor development, progression and treatment response.

Acknowledgements

We would like to thank the Small Animal Imaging Facility of the MD Anderson Cancer Center. This work was funded by NIH/NCI R21EB027197 and NIH/NCI P30 CA016672.