Novel VHH-Based Tracers with Variable Plasma Half-Lives for Imaging of CAIX-Expressing Hypoxic Tumor Cells

Abstract

Hypoxic areas are present in the majority of solid tumors, and hypoxia is associated with resistance to therapies and poor outcomes. A transmembrane protein that is upregulated by tumor cells that have adapted to hypoxic conditions is carbonic anhydrase IX (CAIX). Therefore, noninvasive imaging of CAIX could be of prognostic value, and it could steer treatment strategies. The aim of this study was to compare variants of CAIX-binding VHH B9, with and without a C-terminal albumin-binding domain with varying affinity (ABD_low and ABD_high), for SPECT imaging of CAIX expression. The binding affinity and internalization of the various B9-variants were analyzed using SK-RC-52 cells. Biodistribution studies were performed in mice with subcutaneous SCCNij153 human head and neck cancer xenografts. Tracer uptake was determined by ex vivo radioactivity counting and visualized by SPECT/CT imaging. Furthermore, autoradiography images of tumor sections were spatially correlated with CAIX immunohistochemistry. B9-variants demonstrated a similar moderate affinity for CAIX in vitro. Maximal tumor uptake and acceptable tumor-to-blood ratios were found in the SCCNij153 model at 4 h post injection for [¹¹¹In]In-DTPA-B9 (0.51 ± 0.08%ID/g and 8.1 ± 0.85, respectively), 24 h post injection for [¹¹¹In]In-DTPA-B9-ABD_low (2.39 ± 0.44%ID/g and 3.66 ± 0.81, respectively) and at 72 h post injection for [¹¹¹In]In-DTPA-B9-ABD_high (8.7 ± 1.34%ID/g and 2.43 ± 0.15, respectively)_. An excess of unlabeled monoclonal anti-CAIX antibody efficiently inhibited tumor uptake of [¹¹¹In]In-DTPA-B9, while only a partial reduction of [¹¹¹In]In-DTPA-B9-ABD_low and [¹¹¹In]In-DTPA-B9-ABD_high uptake was found. Immunohistochemistry and autoradiography images showed colocalization of all B9-variants with CAIX expression; however, [¹¹¹In]In-DTPA-B9-ABD_low and [¹¹¹In]In-DTPA-B9-ABD_high also accumulated in non-CAIX expressing regions. Tumor uptake of [¹¹¹In]In-DTPA-B9-ABD_low and [¹¹¹In]In-DTPA-B9-ABD_high, but not of [¹¹¹In]In-DTPA-B9, could be visualized with SPECT/CT imaging. In conclusion, [¹¹¹In]In-DTPA-B9 has a high affinity to CAIX and shows specific targeting to CAIX in head and neck cancer xenografts. The addition of ABD prolonged plasma half-life, increased tumor uptake, and enabled SPECT/CT imaging. This uptake was, however, partly CAIX- independent, precluding the ABD-tracers for use in hypoxia quantification in this tumor type.

Keywords: albumin-binding domain (ABD); carbonic anhydrase IX (CAIX); tumor hypoxia; variable domain of heavy chain only antibody (VHH).

Figure 1

Binding of various concentrations of fluorescently labeled B9-ABD_low and B9-ABD_high to coated human and murine serum albumin to determine binding affinity.

Figure 2

In vitro characterization with SKRC-52 cells showing (A) internalization and binding, (B) IC₅₀ analysis, and (C) specific binding and scatchard curves of [¹¹¹In]In-DTPA-B9 (left), [¹¹¹In]In-DTPA-B9-ABD_low (middle), and [¹¹¹In]In-DTPA-B9-ABD_high (right).

Figure 3

Biodistribution analyses showing the percentage of the injected dose per gram of tissue (%ID/g) at 4 h upon intravenous injection of various protein doses of [¹¹¹In]In-DTPA-B9, with or without preinjection of 300 μg of unlabeled B9.

Figure 4

(A) Biodistribution analysis of [¹¹¹In]In-DTPA-B9 and [¹¹¹In]In-DTPA-B9-ABD_low and [¹¹¹In]In-DTPA-B9-ABD_high showing (A) the percentage of injected dose per gram of tissue (%ID/g) at 4, 24, and 72 h after intravenous injection (B) tumor-to-muscle, (C) tumor-to-blood, and (D) tumor-to-kidney ratios.

Figure 5

Biodistribution analysis of [¹¹¹In]In-DTPA-B9 (4 h after injection), [¹¹¹In]In-DTPA-B9-ABD_low (24 h post injection), and [¹¹¹In]In-DTPA-B9-ABD_high (72 h post injection) with or without injection of a molar excess girentuximab at 24 h before tracer injection.

Figure 6

Accumulation of (A) [¹¹¹In]In-DTPA-B9 with and without excess girentuximab (block) at 4 h, (B) [¹¹¹In]In-DTPA-B9-ABD_low with and without block at 24 h, and (C) [¹¹¹In]In-DTPA-B9-ABD_high with and without block at 72 h, as visualized with microSPECT/CT (left panels, lateral whole body MIP, lateral tumor region MIP and axial 2D scan of the tumor region). Note that we only scanned the tumor area for 2 h; this area is indicated with the dotted rectangle in the whole body SPECT images. In the right panel, autoradiography, an immunohistochemistry (IHC) image showing staining for CAIX (red) and tissue perfusion with Hoechst (green) and an HE image are shown. (BL = bladder, TU = tumor, KE = keratinized tumor regions as indicated with circumscribed areas in the HE images).