A Novel Technique for Performing Simultaneous Dual Radiotracer Imaging Studies in PET:A Proof of Principle Using¹⁸FMISO and¹⁸FLT

Objectives: Simultaneous ¹⁸FMISO and ¹⁸FLT image acquisition would allow the generation of image maps corresponding to tumor hypoxia and proliferation that are (1) automatically registered temporally and spatially, (2) avoid ambiguities associated with images acquired at different times, and (3) avoid significant short-term changes in animal and tumor physiology. The objective of this work is twofold. First, to demonstrate that simultaneous dual ¹⁸F-radiotracer imaging using ¹⁸FMISO and ¹⁸FLT is feasible in a murine model of prostate cancer. Secondly, to show via mathematical (i.e., compartmental) analysis (CA) of dynamic PET (dPET) images (rather than single-time point static) can be used to resolve the total PET imaging signal into the distinct FMISO and by FLT components images, a novel refinement of PET CA.

Methods: Nine nude rats bearing subcutaneous rat prostate tumor model R3327-AT xenografts in the hind limb (minimizing respiration and motion effects) were included in this study. The rats were anesthetized with 1.5% isoflurane/compressed air and injected via tail vein with ~0.6 mCi ¹⁸FMISO. dPET MicroPET image acquisition, using Focus 120 microPET (Siemens Medical Systems), started simultaneously with ¹⁸FMISO injection. Subjects were subsequently injected with ~0.6 mCi ¹⁸FLT 40min post-FMISO injection. Venous blood samples were drawn immediately before ¹⁸FLT and 90min post-FMISO injections respectively (thus with both radiotracers present in the blood). dPET data were reconstructed into 128x128x96 matrices for each of 37 time bins (20 x15 sec + 35x300 sec), and corrected for randoms and scatter (voxel size 0.87 x 0.87 x 0.80 mm³). A first input function (IF1) for t&lt40min, i.e. when only ¹⁸FMISO was present, was deduced from the hottest 20 aortic voxels. Target volumes were drawn over the tumors and corresponding FMISO output function (OF1) time-activity curve were deduced. A 3-compartment 2-tissue (k4=0) model was used to determine the corresponding kinetic rate constants (KRCs) for ¹⁸FMISO. These KRCs and IF1 were used to model OF1 over the full acquisition (90min). The modeled¹⁸FMISO OF1 for 40-90min was then subtracted from the measured combined ¹⁸FMISO/¹⁸FLT (40-90min) TAC to extract the ¹⁸FLT OF2 and IF2a. FLT metabolite fraction was determined from each 90 min blood sample by Sep-Pak column chromatography to obtain the corrected FLT IF2. CA was then performed for ¹⁸FLT using the same model as for ¹⁸FMISO but with k4≠0, using IF2 and OF2, and ¹⁸FLT KRCs were deduced.

Results: The figure shows extracted FMISO/FLT OFs and best fit (black and red squares, respectively) for (1) a&b normal tissue and (2) c&d tumor. CA showed: (1) for normal tissue, as expected, non-significant ¹⁸FMISO trapping (k₃~0) and k₁=k₂, but significant ¹⁸FLT retention (k₃~0.0004 min^-1) and (2) for tumor trapping is mainly due to ¹⁸FMISO (k₃~0.0004 min^-1) and little ¹⁸FLT (k₃~0.00008 min^-1).

Conclusions: To our knowledge, this is the first example of the kinetic PET analysis of serially injected, co-localizing ¹⁸F tracers that report two orthogonal molecular imaging paradigms simultaneously (hypoxia and proliferation) in a single imaging session. This work demonstrates a novel technique that could dramatically enhance the clinical use of FMISO and FLT and other potentially useful PET radiotracer combinations.

• Download figure
• Open in new tab
• Download powerpoint