Monitoring the efficacy of adoptively transferred prostate cancer-targeted human T lymphocytes with PET and bioluminescence imaging

Abstract

Noninvasive imaging technologies have the potential to enhance the monitoring and improvement of adoptive therapy with tumor-targeted T lymphocytes. We established an imaging methodology for the assessment of spatial and temporal distributions of adoptively transferred genetically modified human T cells in vivo for treatment monitoring and prediction of tumor response in a systemic prostate cancer model.

Methods: RM1 murine prostate carcinoma tumors transduced with human prostate-specific membrane antigen (hPSMA) and a Renilla luciferase reporter gene were established in SCID/beige mice. Human T lymphocytes were transduced with chimeric antigen receptors (CAR) specific for either hPSMA or human carcinoembryonic antigen (hCEA) and with a fusion reporter gene for herpes simplex virus type 1 thymidine kinase (HSV1tk) and green fluorescent protein, with or without click beetle red luciferase. The localization of adoptively transferred T cells in tumor-bearing mice was monitored with 2'-(18)F-fluoro-2'-deoxy-1-beta-d-arabinofuranosyl-5-ethyluracil ((18)F-FEAU) small-animal PET and bioluminescence imaging (BLI).

Results: Cotransduction of CAR-expressing T cells with the reporter gene did not affect CAR-mediated cytotoxicity. BLI of Renilla and click beetle red luciferase expression enabled concurrent imaging of adoptively transferred T cells and systemic tumors in the same animal. hPSMA-specific T lymphocytes persisted longer than control hCEA-targeted T cells in lung hPSMA-positive tumors, as indicated by both PET and BLI. Precise quantification of T-cell distributions at tumor sites by PET revealed that delayed tumor progression was positively correlated with the levels of (18)F-FEAU accumulation in tumor foci in treated animals.

Conclusion: Quantitative noninvasive monitoring of genetically engineered human T lymphocytes by PET provides spatial and temporal information on T-cell trafficking and persistence. PET may be useful for predicting tumor response and for guiding adoptive T-cell therapy.

FIGURE 1

BLI of adoptively transferred T-cell persistence and tumor progression. T-cell distribution and persistence were imaged with CBRluc and d-luciferin. Tumor progression was imaged with Rluc and coelenterazine. Data for 2 representative animals from Pz1-treated and Cz1-treated groups are shown. Note different scales for assessment of T-cell bioluminescence signal intensities on days +1–3 and +6–20 after T-cell injection. Same scales were used for assessment of tumor BLI data.

FIGURE 2

Quantitative analysis by BLI of adoptively transferred T-cell persistence and tumor progression in individual animals. Solid lines represent bioluminescence signal intensity for T-cell BLI; dotted lines represent that for tumor BLI. (A) All animals in Pz1-treated group showed uniform decrease in bioluminescence signal from T cells with pronounced delay by day +3. Mouse 5 in Pz1-treated group showed early tumor progression, similar to that in Cz1-treated group. (B) In all animals in Cz1-treated group, T-cell bioluminescence signal decreased rapidly without any delay. ph = photons; sr = steradian.

FIGURE 3

Small-animal PET imaging of ¹⁸F-FEAU accumulation at different time points after adoptive transfer of CAR-positive, HSV1tk/GFP-positive T lymphocytes. (A) Three mutually perpendicular projections for typical Pz1-treated, Cz1-treated, and nontreated tumor-bearing and Pz1-treated tumor-free animals are presented. On day of T-cell injection (day 0), highly intense small-animal PET signal clearly demonstrated presence of pooled T cells in lung areas of treated animals. By day +3, ¹⁸F-FEAU accumulation was detected only in Pz1-treated tumor-bearing animals. No radiotracer accumulation was detected on day +6. (B) Small-animal PET signal quantitation performed as described in Materials and Methods. Statistically significant differences in levels of ¹⁸F-FEAU accumulation between Pz1- and Cz1-treated tumor-bearing animals were observed on day +3. Columns represent means; bars represent SEMs. (C) Regression analysis of T-cell persistence in and overall survival of Pz1-treated tumor-bearing animals. Ratios of %ID/g in ROI to %ID/g in background for individual animals on day +3 were plotted against time to death. Nearly linear relationship (R² = 0.95) between small-animal PET signal intensity and survival was revealed.

FIGURE 4

Small-animal PET/micro-CT fusion imaging and analysis. Analysis of colocalization of ¹⁸F-FEAU accumulation on small-animal PET images with tumor regions or thoracic structures (vessels and lung parenchyma) on micro-CT images is shown. One representative tumor-bearing animal each from Pz1-treated (A), Cz1-treated (B), and control nontreated (C) groups is shown. (A) In Pz1-treated animals, highest levels of ¹⁸F-FEAU accumulation were detected (boxes 1, 2, 4, and 5) in regions corresponding to tumor foci on micro-CT image. Lower signal intensity was observed in intact lung parenchyma and blood vessel zones (boxes 3 and 6, respectively). (B and C) In contrast, Cz1-treated (B) and nontreated (C) animals had low small-animal PET signal intensities (boxes 1, 2, and 3) in selected anatomic areas on micro-CT images. Heart and blood vessels were enhanced by vasculature contrast agent Fenestra VC. Quantitative analysis of ¹⁸F-FEAU accumulation is shown in Table 1.

FIGURE 5

Histologic analysis of tumor specimens. (A) Anti-hPSMA antibody immunohistochemical staining revealed presence of hPSMA-positive tumor cells in lungs from representative Pz1-treated, Cz1-treated, and nontreated tumor-bearing animals in small-animal PET imaging experiment. Magnification, ×10. (B) Anti-GFP antibody was used for immunofluorescence microscopy of HSV1tk/GFP-expressing T cells (green) infiltrating hPSMA-positive tumors (red; stained with anti-hPSMA antibody). Nuclei were stained with 4,6-diamidino-2-phenylindole (blue). Magnification, ×20. Similar results were obtained in BLI experiment.