Résumé
Single-domain antibodies (sdAbs) demonstrate favorable pharmacokinetic profiles for molecular imaging applications. However, their renal excretion and retention are obstacles for applications in targeted radionuclide therapy (TRT). Methods: Using a click-chemistry-based pretargeting approach, we aimed to reduce kidney retention of a fibroblast activation protein a (FAP)-targeted sdAb, 4AH29, for $^{177}$Lu-TRT. Key pretargeting parameters (sdAb-injected mass and lag time) were optimized in healthy mice and U87MG (FAP 1 ) xenografts. A TRT study in a pancreatic ductal adenocarcinoma (PDAC) patient-derived xenograft (PDX) model was performed as a pilot study for sdAb-based pretargeting applications.Results: Modification of 4AH29 with transcyclooctene (TCO) moieties did not modify the sdAb pharmacokinetic profile. A 200-mg injected mass of 4AH29-TCO and an 8-h lag time for the injection of [$^{177}$Lu]Lu-DOTA-PEG 7 -tetrazine resulted in the highest kidney therapeutic index (2.0 ± 0.4), which was 5-fold higher than that of [$^{177}$Lu]Lu-DOTA-4AH29 (0.4 ± 0.1). FAP expression in the tumor microenvironment was validated in a PDAC PDX model with both immunohistochemistry and PET/CT imaging. Mice treated with the pretargeting high-activity approach (4AH29-TCO + [$^{177}$Lu]Lu-DOTA-PEG7-tetrazine; 3 x 88 MBq, 1 injection per week for 3 wk) demonstrated prolonged survival compared with the vehicle control and conventionally treated ([$^{177}$Lu]Lu-DOTA-4AH29; 3 x 37 MBq, 1 injection per week for 3 wk) mice. Mesangial expansion was reported in 7 of 10 mice in the conventional cohort, suggesting treatment-related kidney morphologic changes, but was not observed in the pretargeting cohort.Conclusion: This study validates pretargeting to mitigate sdAbs' kidney retention with no observation of morphologic changes on therapy regimen at early time points. Clinical translation of clickchemistry-based pre-TRT is warranted on the basis of its ability to alleviate toxicities related to biovectors' intrinsic pharmacokinetic profiles. The absence of representative animal models with extensive stroma and high FAP expression on cancer-associated fibroblasts led to a low mean tumor-absorbed dose even with high injected activity and consequently to modest survival benefit in this PDAC PDX.