Initial studies were performed in vitro to assess the feasibility for trimodal PET/SPECT/CT instrumentation to achieve dual tracer imaging with PET and SPECT isotopes. The radionuclides, 99m
Tc and 18
F, were utilized because they represent two of the most common isotopes employed for SPECT and PET imaging, and they are incorporated into numerous commercially available tracers. Overlapping serial dilutions of these SPECT and PET radionuclides were imaged in a 96-well plate to evaluate crosstalk within each detection platform (). The SPECT module employed here was set to a 15% window for discrimination of gamma rays emitted at 140 keV by 99m
Tc. In theory, this should be sufficient to exclude the 511 keV emission stimulated by a PET tracer. However, the SPECT images of the 96-well plate showed significant artifacts in the presence of PET isotope 18
F. This is most likely caused by down-scatter of the high energy PET gamma rays that penetrate the SPECT collimator and scatter into the detector, potentially to within its SPECT mode 15% discrimination window at 140 keV. Shao et al. and Bartoli et al. have reported previously on the complications of down-scatter on SPECT mode detections performed in the presence of the PET radionuclides [13
]. Furthermore, SPECT mode detection in this system does not have a complementary variable with which to exclude spurious counts resulting from PET downscatter, in the way that PET mode excludes any signal lacking coincident detections. It is therefore not surprising that when 18
F-FDG was imaged in SPECT mode, we noted significant artifacts upon reconstructions that negated this approach from further consideration. Meanwhile, when a PET acquisition of 18
F-FDG was gathered in the presence of a 20-fold higher activity of 99m
Tc, no disruptions of the PET reconstruction or data linearity were noted in the resulting image. These in vitro
results guided the implementation of an in vivo
dosing and imaging method that would allow for SPECT mode acquisition in the absence of a PET tracer. By performing a sequential protocol of SPECT tracer dosing, SPECT imaging, CT imaging, PET tracer dosing, and PET imaging (), we were able to construct a strategy that would potentially enable all three modalities to be utilized on a single living specimen.
Subsequent in vivo studies were performed in which combinations of SPECT and PET tracers were utilized in conjunction with the strategy outlined in . Two different sets of SPECT and PET probes were utilized for in vivo imaging and protocol validation: 99mTc-MAA with Na18F (), and 99mTc-Pentetate with Na18F (). Each of these probes targets orthogonal anatomical structures, such that a straightforward analysis of the data would detect the existence of crosstalk. Since SPECT images were strategically acquired in the absence of PET isotope, they yielded straightforward perfusion images of the intended lung and kidney targets. When the opposite case was tested, no bleed through of SPECT signal in the lungs or kidneys was visually noted in the reconstructed PET images given in and . When VOI analysis was performed on the PET data from the experimental lung and kidney trimodal cohorts, the values obtained were statistically no different than corresponding control animals. The background signals from these tissues resulted from non-specific uptake of PET isotope, and excluded the SPECT signal despite the fact a 10-fold higher activity of 99mTc tracer was injected and concentrating in these anatomical regions. This provided critical evidence that the PET modality would reliably reject SPECT events from its reconstructed image, and serve as the cornerstone for sequential trimodal imaging.
While the in vitro and in vivo data were certainly compelling, the fidelity of PET was further evaluated in the context of the current experimental protocol. The four most important variables affecting the reliability of PET mode acquisition in the presence of SPECT tracer (99mTc) pertain to: 1) the energy resolution of the PET detector, 2) The ability to exclude spurious detections lacking a coincident event at 180 degrees, 3) The dead time of the PET detectors, and 4) The maximum achievable count rate of the PET detector. First, in the case of the Albira PET system, the average achievable energy resolution is about 15%, allowing the system to properly exclude 99mTc 140 keV gamma rays yet still provide adequate sensitivity to the 511 keV gamma rays resulting from positron annihilation. Indeed, the energy resolution of approximately 400 to 600 keV (20%) employed in this study provides sufficient resolution to exclude 99mTc 140 keV gamma rays. Further, commercial PET systems will be guarded from spurious detections due to a requirement for coincident detection within a narrow time window, in this case 5 ns. It is important to note that while certain energies may be excluded from use for reconstruction, they are still detected and recorded in the list mode data. Thus, the SPECT isotope may still have deleterious effects on the ability of the PET detectors to cope with the increase in dead time resulting from increased count rates. As noted in , the total count rate and dead times calculated for the in vivo sequential experimental conditions are well within the tolerances of the PET/SPECT/CT image station used for this study. We expect that when similar analyses are performed on the trimodal platforms offered by other vendors, this sequential protocol will be readily applicable.
In addition to the efforts made to facilitate distinct detection of SPECT and PET tracers, there were other essential factors to this imaging method that pertained mostly to conserving animal position. First, the placement and use of a catheter for dosing was critical for ensuring that the spatial position was fixed between SPECT mode and PET mode imaging. Additionally, the bimodal SPECT/PET system, calibrated for registration to a shared modality (CT), was essential for subsequent automated PET/SPECT/CT fusions. Without automated modality fusions, manual manipulations may have been required here and for any future studies, significantly adding to the labor required to obtain useful data.
Many commercial and custom SPECT and PET tracers utilize the two specific radionuclides (99mTc and 18F) employed throughout this study. Using the imaging protocol described here, imaging of any combination of a 99mTc tracer with an 18F tracer should be possible and should allow for multifunctional SPECT/PET imaging. provides a partial list of some of the potential 99mTc and 18F tracers that could be employed with this straightforward method. Additionally, with an aging reactor infrastructure and recent experience with shortages in 99mTc, this protocol may prove valuable if efforts to convert tracers to alternative isotopes become necessary. We have not evaluated other combinations of radionuclides other than 99mTc and 18F, but based on the results obtained here and the defined energy resolutions of our system and most commercial imaging equipment, we believe that additional permutations of SPECT and PET isotopes should be feasible.
Common radiotracers that incorporate 99mTc or 18F that may be used for dual tracer SPECT/PET imaging.
Many valuable dual SPECT/PET tracer experiments may conceivably be conducted that would warrant the use of the dual SPECT/PET imaging protocol described here. Future improvements in dual SPECT/PET imaging could theoretically allow for improvements in throughput and a reduction in labor. Simultaneous SPECT/PET imaging would potentially provide improved efficiency and time savings, and there are reports of imagers that could potentially allow for simultaneous imaging [12
]. These reports are promising, although there appears to be some compromise with simultaneous SPECT/PET systems developed to date, particularly with regard to complications associated with down-scatter and reported PET sensitivities in comparison to the existing leading single-modality PET imagers. (The Vector system from MILabs is a commercially available preclinical system capable of simultaneous dual-tracer scanning; however to our knowledge, complete performance characteristics of this system for simultaneous scanning have not been published). Until further improvements in simultaneous SPECT/PET imaging systems are made, the methods described here provide a straightforward approach for achieving sensitive dual tracer imaging. It should be possible to adapt the methods described here to most integrated PET/SPECT systems, including the Siemens Inveon or GE Triumph, given sufficient considerations to the detector properties of each instrument.
Dual SPECT/PET tracer imaging protocols may further extend the value of multimodality suites. However, it will be particularly important to factor the animal radiation exposures with dual SPECT/PET dosing, especially when combined with CT imaging. Some estimates indicate that animals could receive approximately 15% of the reported total lethal radiation dose for mice during the course of a typical longitudinal PET/CT study [20
]. These estimates are for studies that utilize a single tracer only, and the addition of a second tracer would further increase the radiation dose. Continued increases in instrument sensitivity will enable lower injected-probe activities and help to mitigate the increased exposure from the use of multiple probes.