Finding the appropriate tool for the accurate prediction of microsphere biodistribution: SIR-Spheres radiolabelled with ¹⁸F, ¹¹¹In and ^99mTc

Objectives: Resin based microspheres labeled with ⁹⁰Y (⁹⁰Y SIR-Spheres) are frequently used to treat patients with primary and metastatic solid liver tumors. For the prediction of patient response to the treatment an estimation of the microsphere biodistrubution is made by means of ^99mTc-labeled macroaggregated albumin (^99mTc-MAA) . However, ^99mTc-MAA does not correlate accurately with patient response. The aim of this work was to explore the direct labeling of SIR-Spheres with different isotopes to serve as a more accurate surrogate for ^99mTc-MAA.

Methods: Previously unused (Not previously labeled with ⁹⁰Y ) SIR-spheres were obtained from Sirtex (gift) and were pretreated with an acid-wash step to remove any unwanted contaminants. The spheres were then stored in an appropriate buffer, depending on the used labeling method. Fresh ¹⁸F-fluoride was cyclotron produced in house by the ¹⁸O(p,n)¹⁸F reaction, ^99mTc-pertechnetate was freshly eluted from a drytec-generator (GE) and ¹¹¹In-chloride was purchased from GE. Radiofluorination of the spheres was performed via aluminum-fluoride (Al¹⁸F) chelation. Labelling of the SIR-spheres with ^99mTc was performed by means of the SnCl₂ method. For the labeling with ¹¹¹In, SIR-spheres were incubated in acetate buffer together with ¹¹¹In-chloride. All reaction mixtures were purified by means of three washing/ultracentrifugation cycles. Shelf-life and stability in human plasma was determinded for the different radiolabelled SIR sphere-formulations.

Results: Good radiolabelling yields were obtained after relatively short incubation times. Labelling yields for ¹⁸F, ^99mTc and ¹¹¹In were respectively 85 % (after 5 min.), 60 % (after 30 min) and 85 % (after 15 minutes). Preparations of ¹⁸F- SIR-spheres were too unstable in their final formulation and in human plasma. Spheres labeled with ¹¹¹In were stable for more than 2 hours in isotonic saline but degraded quickly in plasma (40% loss of label after 30 minutes). In stark contrast with ¹⁸F and ¹¹¹In, ^99mTc-labelled SIR spheres showed an excellent stability in both final formulation and human plasma. No significant loss of the label was observed when the labeled spheres were stored over 2 hours at 20°C or at 37°C. This remarkable stability is likely due to the Hard Lewis Acid -Base interaction of the label with the functional groups on the surface of the microspheres.

Conclusion: Due to the simplicity of the labeling method and the stability of the obtained labeled SIR-spheres in human plasma, ^99mTc-labelled SIR-spheres are a promising alternative to ^99mTc-MAA. Extensive in vitro and in vivo studies are now warranted to explore the potential of ^99mTc-labelled SIR spheres for the prediction of patient response towards ⁹⁰Y SIR spheres. Research Support: Nothing to declare