Pretargeted PET Imaging with a TCO-Conjugated Anti-CD44v6 Chimeric mAb U36 and [89Zr]Zr-DFO-PEG5-Tz

Abstract

The recent advances in the production of engineered antibodies have facilitated the development and application of tailored, target-specific antibodies. Positron emission tomography (PET) of these antibody-based drug candidates can help to better understand their in vivo behavior. In this study, we report an in vivo proof-of-concept pretargeted immuno-PET study where we compare a pretargeting vs targeted approach using a new ⁸⁹Zr-labeled tetrazine as a bio-orthogonal ligand in an inverse electron demand Diels-Alder (IEDDA) in vivo click reaction. A CD44v6-selective chimeric monoclonal U36 was selected as the targeting antibody because it has potential in immuno-PET imaging of head-and-neck squamous cell carcinoma (HNSCC). Zirconium-89 (t_1/2 = 78.41 h) was selected as the radionuclide of choice to be able to make a head-to-head comparison of the pretargeted and targeted approaches. [⁸⁹Zr]Zr-DFO-PEG₅-Tz ([⁸⁹Zr]Zr-3) was synthesized and used in pretargeted PET imaging of HNSCC xenografts (VU-SCC-OE) at 24 and 48 h after administration of a trans-cyclooctene (TCO)-functionalized U36. The pretargeted approach resulted in lower absolute tumor uptake than the targeted approach (1.5 ± 0.2 vs 17.1 ± 3.0% ID/g at 72 h p.i. U36) but with comparable tumor-to-non-target tissue ratios and significantly lower absorbed doses. In conclusion, anti-CD44v6 monoclonal antibody U36 was successfully used for ⁸⁹Zr-immuno-PET imaging of HNSCC xenograft tumors using both a targeted and pretargeted approach. The results not only support the utility of the pretargeted approach in immuno-PET imaging but also demonstrate the challenges in achieving optimal in vivo IEDDA reaction efficiencies in relation to antibody pharmacokinetics.

Figure 1

Pretargeting method based on an inverse electron demand Diels–Alder (IEDDA) ligation between trans-cyclooctene (TCO) and tetrazine. In the first step (a), a TCO-conjugated antibody is administered and allowed to reach the target, while unbound antibodies are slowly cleared from the circulation. In the second step (b), a radiolabeled tetrazine is administered and it reacts with the TCO-antibody. Unreacted tetrazine molecules are cleared fast from circulation. The radiolabeled antibody (c) is now visible compared to the nontarget tissue since most of the detected radioactivity signals originate from the tumor.

Scheme 1. Schematic Representation of the Chemical Synthesis of 3 and Radiosynthesis of [⁸⁹Zr]Zr-DFO-PEG₅-Tz ([⁸⁹Zr]Zr-3)

Reaction conditions: (i) dimethyl formamide (DMF), Et₃N, hexafluorophosphate (HATU), overnight reaction at room temperature (rt) in dark conditions, (ii) ⁸⁹Zr-oxalate, Na₂CO₃, oxalic acid, 4-(2-hydroxyethyl)-1-piperazineethanesulfonic acid (HEPES) buffer (pH 7) at room temperature.

Scheme 2. Synthetic Scheme of TCO-Functionalized U36 Antibody (TCO–U36)

Reaction conditions: (i) PBS (pH 8.5), room temperature, overnight.

Figure 2

Ex vivo biodistribution of [⁸⁹Zr]Zr-3 (350 ± 50 kBq i.v., in 100 μL of 10% EtOH in saline + 0.1% Tween + 20 mM gentisic acid, pH 5.2) at 24 h p.i. in VU-SCC-OE tumor-bearing mice (n = 4). The results demonstrate fast clearance via the urinary system and low nonspecific tracer accumulation in healthy organs and in the tumor. The results are presented as % ID/g (mean ± standard deviation, SD).

Figure 3

Ex vivo biodistribution of in vitro and in vivo [⁸⁹Zr]Zr-3-labeled TCO–U36 (0.1 mg, 0.66 nmol) at 72 h p.i. cmAb with a TCO-to-U36 ratio of 27:1 in VU-SCC-OE tumor-bearing mice. For the in vivo pretargeting, [⁸⁹Zr]Zr-3 was injected 24 and 48 h p.i. of TCO–U36 (4.1 ± 0.3 and 3.9 ± 0.5 MBq, 0.7 μg, 0.66 nmol, respectively) ([⁸⁹Zr]Zr-3-to-U36 ratio 1:1). The results are presented as % ID/g (mean ± SD, n = 4).

Figure 4

Ex vivo biodistribution of [¹²⁵I]I-U36 (350 ± 50 kBq, 0.1 mg, 0.66 nmol) and in vitro-radiolabeled [⁸⁹Zr]Zr-3–TCO–U36 (150 ± 50 kBq, 0.1 mg, 0.66 nmol) with different TCO-to-U36 ratios 72 h after injection to athymic nude NMRI mice. The results are presented as % ID/g (mean ± SD; n = 4).

Figure 5

Comparison of radioactivity (% ID/g) in liver and blood for ¹²⁵I-labeled U36 and in vitro-radiolabeled [⁸⁹Zr]Zr-3–TCO–U36 with different TCO-to-U36 ratios at 72 h p.i. in athymic nude NMRI mice and in mice bearing VU-SCC-OE xenografts (27:1 TCO-to-U36) (columns denote mean ± SD, n = 4).

Figure 6

Ex vivo biodistribution of (A) in vitro-labeled [⁸⁹Zr]Zr-3–TCO–U36 (3.0 ± 0.3 MBq, 0.1 mg, 0.66 nmol) and (B) in vivo ([⁸⁹Zr]Zr-3) (2.5 ± 0.2 MBq, 0.7 μg, 0.66 nmol)-labeled U36 (0.1 mg, 0.66 nmol, 6:1 TCO-to-U36) at 72 h p.i. of cmAb in VU-SCC-OE xenografts ([⁸⁹Zr]Zr-3-to-U36 ratio 1:1). The results are presented as % ID/g (mean ± SD, n = 4).

Figure 7

Coronal PET/CT images for all groups at 71 h p.i. of the U36 antibody administration in VU-SCC-OE xenografts; [⁸⁹Zr]Zr-3 was injected (a) 24 h or (b) 48 h p.i. of TCO–U36 ([⁸⁹Zr]Zr-3-to-U36 ratio 1:1). The third group (c) was injected with in vitro-labeled [⁸⁹Zr]Zr-3–TCO–U36 at t = 0.

Figure 8

Standardized uptake values (SUVs) in the VU-SCC-OE xenograft tumors for all groups at 1, 24, 48, and 71 h after the U36 injection. The results are presented as SUV (mean ± SD, n = 4).

Scheme 3. Experimental Scheme for the PET Imaging Studies