Ventilation Distribution Studies Comparing Technegas and “Gallgas” Using ⁶⁸GaCl₃ as the Label

Preparation of Gallgas

⁶⁸Ga (half-life of 68 min, decay by positron emission [89%], and electron capture [11%]) was available from a ⁶⁸Ge/⁶⁸Ga generator system in which ⁶⁸Ge (half-life, 270.8 d) was attached to a column of an inorganic matrix using titanium dioxide. The nominal ⁶⁸Ge radioactivity loaded onto the generator column was 1,850 MBq. The specified shelf-life of the generator was 2–3 y. The ⁶⁸Ga was eluted with 6 mL of 0.1 M hydrochloric acid. The whole volume of the eluate was preconcentrated on an anion-exchange cartridge to decrease the volume of the generator eluate (from 6 mL to 200 μL) ( 3).

Gallgas was prepared on the commercially available Technegas Plus Generator (Vita Medical Ltd.). The standard procedure was used, including the prewetting of the crucible with ethanol before the application of the ⁶⁸GaCl₃ solution. The preconcentrated ⁶⁸Ga solution (100–140 μL; 50–100 MBq) was applied to the graphite crucible and evaporated in the furnace at 50°C–70°C. ⁶⁸Ga was then encapsulated by carbon composites. Gallium metal was covered by layers of graphite sheets that isolated the metal from the external environment (performed at 2,500°C–2,700°C for 15–20 s). The imaging agent was used for the experiments via inhalation within 10 min to avoid aggregation of the particles.

Imaging Techniques

The animals were studied supine, first on a CT-equipped γ-camera using inhalation of Technegas and then on a PET/CT camera using inhalation of Gallgas. The radioactivity inhaled by the piglets was 30 and 5–10 MBq, respectively, for ^99mTc and ⁶⁸Ga. A vacuum mattress was used to keep the piglet in the same position during the experiments and when the animal was moved between the cameras.

SPECT Acquisition

SPECT was performed on a dual-head γ-camera (Infinia; GE Healthcare) equipped with all-purpose, low-energy collimators. The SPECT images were acquired with a low-energy general-purpose collimator in 60 projections (30 per head) stored in a 128 matrix. The acquisition time was 30 s per projection. Low-resolution CT (2.5 mA, 140 kV) was performed covering the same volume as the SPECT acquisition to evaluate lung borders and enable attenuation correction. The overall acquisition time was 25 min. Data were then reconstructed on a Xeleris workstation using a reconstruction technique with 4 iterations, 8 subsets, and a Hann 3-dimenstional postreconstruction filter.

PET Acquisition

PET was performed on a Discovery VCT scanner (GE Healthcare). The acquisition mode used a 3-dimensional imaging technique, a scan time of 3 min per bed position, 128 × 128 pixels, and a 70 cm field of view. The VUE Point HD iterative reconstruction algorithm (GE Healthcare) was used to reconstruct images with 2 iterations and 28 subsets. The z-axis filter was a standard box car (1-4-1), and the postprocessing filter was set to 6.00 mm (in full width at half maximum). Images from the CT study were attenuation-corrected 120 kV; 30–80 mA; smart mA and auto mA (GE Healthcare); and rotation time, 0.5 s).

Analysis Techniques: Distribution Volume for Pseudogas

The distribution volume for the pseudogas when labeled either with ⁶⁸Ga or with ^99mTc was calculated from the SPECT and PET images as the sum of all voxels multiplied by the individual voxel volume that contained radioactivity of ⁶⁸Ga or ^99mTc inside the lung. Lungs were delineated by creating a lung mask from the CT image.

The variation in tracer gas radioactivity, reflecting any inhomogeneity in the distribution of the gas, was calculated by applying statistical modeling.

Statistics

PET and SPECT counts were log-transformed in a transversal slice (the one with the largest lung area) to enable a regression analysis of the heterogeneity of the ventilation distribution. For each piglet, the transformed PET and SPECT counts were related to 2-dimensional coordinates (i.e., x, y from that slice) in a regression model using a quadratic function of the coordinates. The quadratic function was fitted for the 2 lungs combined, delineated by the created lung masks from the CT images.

The difference between the coefficients of determination (R²) from the 2 models (the SPECT and PET) was calculated for each piglet, and a Wilcoxon matched-pairs signed-rank test was performed to test the null hypothesis of equal R².

For the other physiologic variables, we used repeated-measures ANOVA for comparisons during the protocol. Significance was defined as a P level (bicaudal) of less than 0.05.