Towards miniaturized quality control for 18F-labeled PET tracers: Separation of D-FAC alpha and beta anomers via capillary electrophoresis

Objectives In response to the rising demand for Positron emission tomography (PET), our lab and others have been developing microfluidic technologies to enable de-centralization of PET tracer production from radiopharmacies to imaging centers. This shift is predicted to provide the flexibility of producing PET tracers on demand with respect to time and quantity and also to allow a larger variety of tracers. De-centralization is not possible with present technologies for tracer production and quality control (QC) analysis. Here, we explore the feasibility of using capillary electrophoresis (CE) to substitute some of the current equipment used in conducting QC analysis of PET tracers.

Methods CE is a general separation method [1] that can be miniaturized into microfluidic chips [2, 3]. We have previously shown the separation of 3′-deoxy-3′-fluorothymidine (FLT) from 6 impurities on a home-made CE system [4]. Here, we further show the versatility of CE by separating 1-(2′-Deoxy-2′-fluoro-β-D-arabinofuranosyl)cytosine (β-D-FAC) from the alpha anomer. In addition, we show the ability to separate and detect Kryptofix 2.2.2 (K2.2.2) from FAC. Samples were injected electrokinetically, and detection was performed at 224 nm.

Results The separation of the two FAC anomers can be achieved in < 8.5 min in the presence of SDS and the use of a phosphate buffer (Fig 1). With the use of a carbonate buffer at a pH > 10, FAC could also be separated from K2.2.2.

Conclusions CE can be used to separate a PET tracer (FAC) and its anomeric impurity with good resolution. These results along with previous studies on FLT suggest the feasibility of utilizing CE for several aspects of the analysis of the chemical and radiochemical purity of PET tracers. In the future, we intend to extend the generality of our work to other PET tracers and move into chip-based CE separations to decrease the instrument footprint. We are also actively exploring methods to improve the limits of detection.