In vivo evaluation of [18F]fluoroetanidazole as a new marker for imaging tumour hypoxia with positron emission tomography

Abstract

Development of hypoxia-targeted therapies has stimulated the search for clinically applicable noninvasive markers of tumour hypoxia. Here, we describe the validation of [(18)F]fluoroetanidazole ([(18)F]FETA) as a tumour hypoxia marker by positron emission tomography (PET). Cellular transport and retention of [(18)F]FETA were determined in vitro under air vs nitrogen. Biodistribution and metabolism of the radiotracer were determined in mice bearing MCF-7, RIF-1, EMT6, HT1080/26.6, and HT1080/1-3C xenografts. Dynamic PET imaging was performed on a dedicated small animal scanner. [(18)F]FETA, with an octanol-water partition coefficient of 0.16+/-0.01, was selectively retained by RIF-1 cells under hypoxia compared to air (3.4- to 4.3-fold at 60-120 min). The radiotracer was stable in the plasma and distributed well to all the tissues studied. The 60-min tumour/muscle ratios positively correlated with the percentage of pO(2) values <5 mmHg (r=0.805, P=0.027) and carbogen breathing decreased [(18)F]FETA-derived radioactivity levels (P=0.028). In contrast, nitroreductase activity did not influence accumulation. Tumours were sufficiently visualised by PET imaging within 30-60 min. Higher fractional retention of [(18)F]FETA in HT1080/1-3C vs HT1080/26.6 tumours determined by dynamic PET imaging (P=0.05) reflected higher percentage of pO(2) values <1 mmHg (P=0.023), lower vessel density (P=0.026), and higher radiobiological hypoxic fraction (P=0.008) of the HT1080/1-3C tumours. In conclusion, [(18)F]FETA shows hypoxia-dependent tumour retention and is, thus, a promising PET marker that warrants clinical evaluation.

Figure 1

Radiochemical synthesis of [¹⁸F]FETA. I, II, III, and IV represent N-[2-(toluene-4-sulphonyloxy)-ethyl]-phthalimide, 2,3,5,6-Tetrafluorophenyl 2-(2-nitroimidazol-1-yl) acetate, [¹⁸F]Fluoroetanidazole, and [¹⁸F]Fluoroethylamine.

Figure 2

In vitro binding of [¹⁸F]FETA to RIF-1 cells under hypoxic (-•-) and normoxic (-○-) conditions. The cells were incubated with [¹⁸F]FETA under nitrogen gas or air for 0–120 min and washed to remove unbound radioactivity. The percentage of bound radioactivity was calculated as (bound radioactivity in 10⁶ cells × 100)/total radioactivity. Data are mean±s.e.m. (n=3), ^*P⩽0.05.

Figure 3

Tumour and normal tissue distribution of [¹⁸F]FETA at 1 h p.i. in HT1080/26.6 tumour-bearing male Balb/c nu/nu mice (n=10). Inset: radiotracer uptake in the different tumour types (n=5–10). Data are mean % ID g⁻¹±s.e.m.

Figure 4

Reversed-phase high-performance liquid chromatograms of [¹⁸F]FETA and its putative metabolites at an early time-point (10 min p.i.) in plasma (A), liver (D), and tumour (G), and a late time-point (60 min p.i.) in plasma (B), muscle (C), liver (E), gall bladder (F), tumour (H), and urine (I). Filled arrows are parent [¹⁸F]FETA compound, and double-lined arrows are metabolites.

Figure 5

Association between tumour oxygenation as measured by OxyLite probes and [¹⁸F]FETA tumour uptake determined by ex vivo biodistribution studies. (A) Relative [¹⁸F]FETA uptake expressed as tumour-to-muscle ratio vs relative frequency of pO₂ values <5 mmHg. (B) [¹⁸F]FETA uptake vs relative frequency of pO₂ values <2.5 mmHg.

Figure 6

[¹⁸F]FETA-PET images of HT1080 tumour-bearing mice acquired on the small animal quad-HIDAC scanner (pixel size 0.5 × 0.5 × 0.5 mm³). (A) Three-dimensional (volume-rendered) image of an HT1080/1-3C tumour-bearing mouse (30–60 min p.i. summed) showing a dorsal view of the mouse. Here, pixel values are defined by the maximum voxel value in corresponding lines in the z-axis. Arrows point to tumour (Tm), Kidneys (Ki), small intestine (In), and urinary bladder (Bl). (B) Sagittal (0.5 mm) slice of 30–60 min p.i. summed [¹⁸F]FETA-PET images from the same mouse as in (A) at the midplane level, showing low radiotracer uptake in brain (Br) and spinal cord (Co), as well as high accumulation in urinary bladder. (C) Transverse (0.5 mm) slice of 30–60 min p.i. summed [¹⁸F]FETA-PET images from the same HT1080/1-3C tumour-bearing mouse as in (A) at the level of the maximal tumour diameter. (D) Corresponding transverse slice from an HT1080/26.6 tumour-bearing mouse, exhibiting lower tumour radiotracer uptake.

Figure 7

Tumour time–activity curves (TACs), normalised to the integral of the heart cavity TAC. The curves were obtained from region of interest analysis of dynamic [¹⁸F]FETA-PET images from HT1080/26.6 (-•-) and HT1080/1-3C (-○-) tumour-bearing mice. Data are mean±s.e.m. (n=3–4 per group).

Figure 8

Vessel density, pO₂, and radiation sensitivity of HT1080/26.6 and HT1080/1-3 tumours. (A–B) Typical 5 μm H&E-stained histological sections of HT1080/26.6 (A) and HT1080/1-3C (B) tumours, with arrows pointing to vessels. (C) Summary data for vessel density depicting average number of vessels from five randomly selected fields of view (0.23 mm²; × 200 magnification) per section; three sections per tumour. (D) OxyLite pO₂ measurements. (E–F) Radiation sensitivity determined in clonogenic assays and expressed as radiobiological hypoxic fraction of HT1080/26.6 (E) and HT1080/1.3C (F) tumours. Data are mean survival (n=4). RHF, radiobiological hypoxic fraction. There was a statistically significant difference between the RHF of 1-3C and 26.6 tumours (P=0.008).