Performance of nanoScan PET/CT and PET/MR for quantitative imaging of 18F and 89Zr as compared with ex vivo biodistribution in tumor-bearing mice

Abstract

Introduction: The assessment of ex vivo biodistribution is the preferred method for quantification of radiotracers biodistribution in preclinical models, but is not in line with current ethics on animal research. PET imaging allows for noninvasive longitudinal evaluation of tracer distribution in the same animals, but systemic comparison with ex vivo biodistribution is lacking. Our aim was to evaluate the potential of preclinical PET imaging for accurate tracer quantification, especially in tumor models.

Methods: NEMA NU 4-2008 phantoms were filled with ¹¹C, ⁶⁸Ga, ¹⁸F, or ⁸⁹Zr solutions and scanned in Mediso nanoPET/CT and PET/MR scanners until decay. N87 tumor-bearing mice were i.v. injected with either [¹⁸F]FDG (~ 14 MBq), kept 50 min under anesthesia followed by imaging for 20 min, or with [⁸⁹Zr]Zr-DFO-NCS-trastuzumab (~ 5 MBq) and imaged 3 days post-injection for 45 min. After PET acquisition, animals were killed and organs of interest were collected and measured in a γ-counter to determine tracer uptake levels. PET data were reconstructed using TeraTomo reconstruction algorithm with attenuation and scatter correction and regions of interest were drawn using Vivoquant software. PET imaging and ex vivo biodistribution were compared using Bland-Altman plots.

Results: In phantoms, the highest recovery coefficient, thus the smallest partial volume effect, was obtained with ¹⁸F for both PET/CT and PET/MR. Recovery was slightly lower for ¹¹C and ⁸⁹Zr, while the lowest recovery was obtained with ⁶⁸Ga in both scanners. In vivo, tumor uptake of the ¹⁸F- or ⁸⁹Zr-labeled tracer proved to be similar irrespective whether quantified by either PET/CT and PET/MR or ex vivo biodistribution with average PET/ex vivo ratios of 0.8-0.9 and a deviation of 10% or less. Both methods appeared less congruent in the quantification of tracer uptake in healthy organs such as brain, kidney, and liver, and depended on the organ evaluated and the radionuclide used.

Conclusions: Our study suggests that PET quantification of ¹⁸F- and ⁸⁹Zr-labeled tracers is reliable for the evaluation of tumor uptake in preclinical models and a valuable alternative technique for ex vivo biodistribution. However, PET and ex vivo quantification require fully described experimental and analytical procedures for reliability and reproducibility.

Keywords: Ex vivo biodistribution; PET-CT; PET-MRI; Preclinical imaging; Quantification.

Fig. 1

Linear recovery of the PET/CT and PET/MR scanners as assessed with ¹¹C (n = 6 phantoms) (A), ⁶⁸Ga (B), ¹⁸F (C) and ⁸⁹Zr (D). Note the impaired linearity at doses < 0.05 MBq/mL for each of the radionuclides

Fig. 2

NanoScan PET/CT recovery coefficients for ¹¹C (A), ⁶⁸Ga (B), ¹⁸F (C), and ⁸⁹Zr (D) with and without scatter and attenuation correction (see also Additional file 1: Table S2) using the TeraTomo reconstruction algorithm

Fig. 3

Bland–Altman plots comparing tumor uptake of [¹⁸F]FDG (A, B) and [⁸⁹Zr]Zr-DFO-NCS-trastuzumab (C, D) assessed by PET imaging (PET/CT: A, C; PET/MR: B, D) or by ex vivo biodistribution. The middle-dotted line shows the Bias (mean of the ratios) and the upper and lower dotted lines show the 95% limits of agreement. Average (%IA/g) corresponds to the average uptake value per animal between PET and ex vivo biodistribution

Fig. 4

Linear regression plots showing the correlation between tumor uptake assessed by PET imaging (PET/CT: A, C; PET/MR: B, D) and ex vivo biodistribution for [¹⁸F]FDG (A, B) and [⁸⁹Zr]Zr-DFO-NCS-trastuzumab (C, D)