An Effective Immuno-PET Imaging Method to Monitor CD8-Dependent Responses to Immunotherapy

Abstract

The rapidly advancing field of cancer immunotherapy is currently limited by the scarcity of noninvasive and quantitative technologies capable of monitoring the presence and abundance of CD8(+) T cells and other immune cell subsets. In this study, we describe the generation of (89)Zr-desferrioxamine-labeled anti-CD8 cys-diabody ((89)Zr-malDFO-169 cDb) for noninvasive immuno-PET tracking of endogenous CD8(+) T cells. We demonstrate that anti-CD8 immuno-PET is a sensitive tool for detecting changes in systemic and tumor-infiltrating CD8 expression in preclinical syngeneic tumor immunotherapy models including antigen-specific adoptive T-cell transfer, agonistic antibody therapy (anti-CD137/4-1BB), and checkpoint blockade antibody therapy (anti-PD-L1). The ability of anti-CD8 immuno-PET to provide whole body information regarding therapy-induced alterations of this dynamic T-cell population provides new opportunities to evaluate antitumor immune responses of immunotherapies currently being evaluated in the clinic.

Figure 1. Anti-CD8 169 cDb characterization

(A) Antibody engineering schematic of cys-diabody construction and site-specific conjugation to the engineered thiols. V_L and V_H are variable light and heavy chains, respectively. C_H1-3 are the heavy chain constant domains 1–3 and C_L is the light chain constant domain. (B) SDS/PAGE gel (left panel) shows purified 169 cDb (Lane 1) and reduced and mal488-conjugated 169 cDb (Lane 2) for fluorescent flow cytometry cell binding assays. The UV image (right panel) of the same gel shows mal488 conjugated to 169 cDb. (C) Size exclusion chromatography demonstrated the site-specific conjugation to mal488 has not disrupted the diabody confirmation (Left panel). Site-specific conjugation to malDFO resulted in a similar size exclusion profile (Right panel). Reference arrows indicate albumin (66 kDa) at 20.8 min, carbonic anhydrase (29 kDa) at 24.7 min, and cytochrome C (12.4 kDa) at 27.4 min. (D) Flow cytometry using the mal488-169 cDb of single cell suspensions from the blood, thymus, spleen, and lymph nodes of C57BL/6 (Lyt2.2⁺; left column) and AKR (Lyt2.1⁺; right column) mice.

Figure 2. Anti-CD8 microPET of ⁸⁹Zr-malDFO-169 cDb in wild type and CD8-blocked AKR mice

(A) ⁸⁹Zr-malDFO-169 cDb is injected into wild type (top row) and CD8-blocked (bolus 3 mg/kg GK1.4 cDb; bottom row) AKR mice and imaged at 4, 8, and 22 h post-injection. Images are represented as 25 mm maximum intensity projections (MIPs). (B) Transverse CT and PET/CT images indicated with dashed white bars show specific targeting to the spleen and lymph nodes of wild type AKR mice. (C) Ex vivo biodistribution at 22 h post-injection of ⁸⁹Zr-malDFO-169 cDb from wild type and CD8-blocked AKR mice (n = 3 per group). Abbreviations: CLN – cervical lymph node, ALN – axillary lymph node, Li – liver, Sp – Spleen, K – Kidney, ILN – inguinal lymph node, B – bone, and PLN – popliteal lymph node.

Figure 3. Anti-CD8 immuno-PET of ⁸⁹Zr-malDFO-169 cDb in the OT-I adoptive T cell therapy model

(A) C57BL/6 mice bearing s.c. EL4 and EL4-Ova tumors received the full protocol of myelodepletion (900 cGy) with hematopoietic stem cell (HSC)/bone marrow (BM) transplantation, adoptive OT-I T cell transfer followed by OVA-pulsed dendritic cell (DC) vaccination and high dose IL-2. (B) Representative immuno-PET images at 22 h post-injection of ⁸⁹Zr-malDFO-169 cDb of unblocked and CD8-blocked mice bearing EL4-Ova and EL4 tumors five days post adoptive OT-I T cell transfer. Coronal images are presented as 25 mm MIPs and transverse images are presented as 2 mm MIPs. (C) Ex vivo biodistribution at 22 h post-injection of ⁸⁹Zr-malDFO-169 cDb of unblocked and CD8-blocked mice bearing EL4-Ova and EL4 tumors (n = 3 per group). Full biodistribution analysis is in Table S2. (D) Ratios of EL4-Ova:EL4, EL4-Ova:blood, and EL4:blood of unblocked and CD8-blocked mice from ex vivo biodistribution analysis. (E) Flow cytometry analysis of tumors harvested five days post-adoptive T cell therapy was used to determine the percent of CD45⁺/CD3⁺/CD8⁺ T cells in the antigen-positive EL4-Ova tumor versus the antigen-negative EL4 tumor (n=3). * indicates p<0.05, † indicates p<0.005.

Figure 4. Anti-CD8 immuno-PET of ⁸⁹Zr-malDFO-169 cDb in the CT26/anti-CD137 immunotherapy model

(A) Balb/c mice bearing s.c. CT26 tumors were treated with anti-CD137 therapy every other day for four treatments and anti-CD8 immuno-PET was acquired on day 16 post-tumor implantation. (B) Tumor growth curves of CD137 treated and untreated mice (average tumor diameter). (C) On day 8 post-immunotherapy initiation, CD137-treated mice, CD137-treated/CD8-blocked mice, and control mice (no anti-CD137 therapy) were injected with ⁸⁹Zr-malDFO-169 cDb and immuno-PET images were acquired at 22 h post-injection. (D) CD8 IHC of untreated CT26 tumors or anti-CD137-treated CT26 tumors indicate the presence of increased CD8⁺ tumor-infiltrating lymphocytes. (E) Ex vivo biodistribution at 22 h post-injection of ⁸⁹Zr-malDFO-169 cDb of CD137-treated mice, CD137-treated/CD8-blocked mice, and control mice (n = 3 per group). Full biodistribution analysis is in Table S3. (F) Tumor-to-blood ratios of CD137-treated mice, CD137-treated/CD8-blocked mice, and control mice. (G) Flow cytometry analysis of tumors harvested on day 15 was used to determine the percent of CD45⁺/CD8⁺ T cells in the CT26 tumors (n=3). † indicates p<0.005, ‡ indicates p<0.0005.

Figure 5. Anti-CD8 immuno-PET of ⁸⁹Zr-malDFO-169 cDb in the CT26/anti-PD-L1 immunotherapy model

(A) Balb/c mice bearing s.c. CT26 tumors were treated with anti-PD-L1 therapy every other day for four treatments and anti-CD8 immuno-PET was acquired on day 16 post-tumor implantation. (B) Tumor growth curves of control (no anti-PD-L1 therapy), partial responders to anti-PD-L1 therapy (tumor < 8 mm average diameter), and non-responders to anti-PD-L1 therapy (tumor > 8 mm average diameter) treated and untreated mice (average tumor diameter). (C) Representative ⁸⁹Zr-malDFO-169 cDb immuno-PET images acquired at 22 h post-injection of anti-PD-L1 responding and non-responding mice. (D) Ex vivo biodistribution at 22 h post-injection of ⁸⁹Zr-malDFO-169 cDb of anti-PD-L1 responding and non-responding mice (n = 3–4 per group). Full biodistribution analysis is in Table S4. (E) Tumor-to-blood ratios of anti-PD-L1 responding and non-responding mice. (F) Flow cytometry analysis of tumors harvested on day 15 was used to determine the percent of CD45⁺/CD8⁺ T cells in the CT26 tumors in control, anti-PD-L1 responding, and anti-PD-L1 non-responding mice (n=4–5). * indicates p<0.05.