Imaging CTLA-4+T-cells in Humanized Mouse Models
Emily Ehlerding,Hye Jin Lee,Dawei Jiang,Carolina Ferreira,Jonathan Engle,Douglas McNeel,Weibo Cai
2019-01-01
Abstract:2 Objectives: Checkpoint immunotherapy treatments are showing great promise for cancer patients; however, the mechanisms and predictors of response are still largely mysteries, especially for CTLA-4-targeted treatments, which have particularly heterogeneous response patterns. Imaging of these immunotherapy targets using positron emission tomography (PET) can allow for noninvasive monitoring of their dynamic expression and biodistribution, and may allow for patient stratification in the future. Therefore, two tracers targeting human CTLA-4, one a full antibody and the other a F(ab’)2 fragment, were radiolabeled with 64Cu and validated in humanized mouse models. Methods: Ipilimumab was enzymatically digested to develop ipilimumab-F(ab’)2, and both the intact antibody and the fragment were conjugated with NOTA to chelate 64Cu for PET. Both tracers maintained their specificity for activated CTLA-4+ T-cells in vitro. The tracers were administered to both control NBSGW mice and humanized mice (PBL mice, which were engrafted with human peripheral blood lymphocytes), and PET was conducted out to 48 h post-injection. The onset of graft-versus-host disease in these humanized mice causes their salivary glands to dysfunction and sequestration of the foreign cells there, providing a localized site for exploration of CTLA-4+ tissue uptake. PET region-of-interest analysis, ex vivo biodistribution studies, and tissue staining were used to confirm that the tracers specifically accumulated in CTLA-4+ tissues. Results: Following intravenous injection of the respective tracers (n=3-5 per group), specific uptake was noted in the salivary gland tissues of the humanized mice. This uptake, as a result of graft-versus-host disease onset, was proven to be due to the presence of human T-cells through staining of the tissues for human CD3 and CTLA-4. 64Cu-NOTA-ipilimumab demonstrated the highest absolute uptake in the salivary glands of PBL mice, peaking at 7.00 ± 2.19 %ID/g, 24 h post-injection. In contrast, 64Cu-NOTA-ipilimumab-F(ab’)2 uptake was at 2.40 ± 0.86 %ID/g at the same time point. However, the F(ab’)2 agent cleared from circulation much more quickly than the intact antibody, providing higher salivary gland-to-blood ratios, which reached 1.78 ± 0.72 at 48 h post-injection, compared to 64Cu-NOTA-ipilimumab at 1.19 ± 0.49. Comparison with a nonspecific IgG, with a salivary gland-to-blood ratio of only 0.34 ± 0.08, verified the specificity of the salivary gland uptake with the targeted tracers. Uptake of the tracers in the salivary glands of control mice, and the nonspecific tracer in the PBL mice, was lower than that of the related targeted tracers at all imaging time points. Ex vivo studies verified all the trends found through PET analysis. Conclusions: PET imaging with both 64Cu-NOTA-ipilimumab and 64Cu-NOTA-ipilimumab-F(ab’)2 was able to localize CTLA-4+ tissues. These tracers may thus help elucidate the mechanisms of response to, and possibly toxicity from, CTLA-4-targeted checkpoint immunotherapy treatments.