Safety of targeting ROR1 in primates with chimeric antigen receptor-modified T cells

Abstract

Genetic engineering of T cells for adoptive transfer by introducing a tumor-targeting chimeric antigen receptor (CAR) is a new approach to cancer immunotherapy. A challenge for the field is to define cell surface molecules that are both preferentially expressed on tumor cells and can be safely targeted with T cells. The orphan tyrosine kinase receptor ROR1 is a candidate target for T-cell therapy with CAR-modified T cells (CAR-T cells) because it is expressed on the surface of many lymphatic and epithelial malignancies and has a putative role in tumor cell survival. The cell surface isoform of ROR1 is expressed in embryogenesis but absent in adult tissues except for B-cell precursors and low levels of transcripts in adipocytes, pancreas, and lung. ROR1 is highly conserved between humans and macaques and has a similar pattern of tissue expression. To determine if low-level ROR1 expression on normal cells would result in toxicity or adversely affect CAR-T cell survival and/or function, we adoptively transferred autologous ROR1 CAR-T cells into nonhuman primates. ROR1 CAR-T cells did not cause overt toxicity to normal organs and accumulated in bone marrow and lymph node sites, where ROR1-positive B cells were present. The findings support the clinical evaluation of ROR1 CAR-T cells for ROR1(+) malignancies and demonstrate the utility of nonhuman primates for evaluating the safety of immunotherapy with engineered T cells specific for tumor-associated molecules that are homologous between humans and nonhuman primates.

Figure 1

A, RT-qPCR of ROR1 mRNA expression in B-CLL cells and a panel of human and rhesus macaque tissues. ROR1-expression was normalized to the geometric mean of the housekeeping genes GAPDH and TBP. B-CLL sample #1 was used as the reference control (ROR1-expression=1). B, cytolytic activity of rhesus T cells modified to express a R12-ROR1 CAR with human (hu) or rhesus macaque (rh) 4–1BB and CD3ζ domains against ⁵¹Cr-labeled K562/ROR1 or K562 cells. Results (mean, SEM) are from 3 independent experiments in 3 animals (A11047, A13011, A13002). C, flow cytometric analysis of purity and phenotype of ROR1 CAR-modified CD4⁺ and CD8⁺ T cells showing expression of tCD19 in CD4⁺ (top panels) or CD8⁺ (bottom panels) T cells either mock-transduced, after transduction with the ROR1 CAR (pre-selection), after selection for tCD19-expression (post-selection 1), and after selection for CD4⁺ or CD8⁺ cells, respectively (post-selection 2). All samples are gated on CD3⁺ cells. D, cytolytic activity of CD4⁺ and CD8⁺ ROR1 CAR-T cells against ⁵¹Cr-labeled K562/ROR1 or K562 cells. E, proliferation of CFSE-labeled ROR1 CAR-T cells 72 hours after stimulation with K562/ROR1 cells or unmodified K562 cells. CFSE-dye dilution is shown after gating on CD3⁺CD4⁺tCD19⁺ or CD3⁺CD8⁺tCD19⁺T cells. F, Luminex cytokine assay of supernatants obtained after 24 hours from triplicate co-cultures of 5×10⁴ CD4⁺ or CD8⁺ROR1 CAR-T cells with K562/ROR1 cells or media alone as described in Methods. The asterisk (*) demarks cytokine levels below detection level. Cytokine levels in media controls were below detection level. Data in C-F is shown for macaque A13002 and representative of results in 3 animals.

Figure 2

Toxicity and in vivo persistence of ROR1 CAR-T cells. A, body weight and serum chemistry before and at the indicated days after the T-cell infusion. The grey shaded area demarks the normal range for each parameter in rhesus macaques. B, plasma cytokine levels measured prior to and post-infusion. C, frequency of transferred T cells (%) within the CD3⁺CD4⁺ and CD3⁺CD8⁺ subsets in blood after infusion of ROR1 CAR and control EGFRt⁺ T cells. Stainings are with mAbs specific for CD3, CD4, CD8, and CD19 or EGFRt. D, absolute numbers of ROR1 CAR⁺ and EGFRt⁺T cells in the blood measured by flow cytometry and calculated based on the results of a CBC on the indicated days in an accredited clinical laboratory. E, real-time qPCR for the presence of transgene vector-specific DNA sequences in samples of PBMC obtained before and after T-cell transfer. The arrow indicates the T-cell infusion (D, E).

Figure 3

In vivo migration and function of ROR1 CAR-T cells. A, frequency of ROR1 CART cells in BM and LN samples obtained on day 5 after the T-cell infusion. Stainings were performed with mAbs specific for CD3, CD4, CD8, and CD19 or EGFRt, and gating on CD3⁺CD4⁺ or CD3⁺CD8⁺T cells. B, frequency of ROR1⁺ B cells in the BM before and on day 5 after ROR1 CAR-T cell infusion. The gating strategy for CD19⁺CD45^intermediate B cells is shown in Supplementary Fig. S2. C, absolute number of CD19⁺ B cells in blood samples obtained before and after the T-cell infusion determined by staining with mAbs specific for CD19, CD3, CD4, and CD8 and flow cytometry to detect CD19⁺CD3- B cells. Absolute numbers were determined based on the lymphocyte count of a CBC obtained at the same time and determined in an accredited clinical laboratory. D, CD107A-degranulation assay on PBMC obtained before and at day 7 post-infusion, stimulated ex vivo with K562/ROR1 cells or PMA and Ionomycin. Expression of CD107A was determined by flow cytometry. Cultured ROR1 CAR-T cells served as positive control. E, detection of a transgene product-specific T-cell response after transfer of ROR1 CAR-T cells. Pre- and post-infusion PBMC were stimulated twice one week apart with γ-irradiated ROR1 CAR-T cells and autologous feeder cells. Aliquots of these cultures were evaluated in a cytotoxicity assay for recognition of autologous ⁵¹Cr-labeled ROR1 CAR-T cells or unmodified T cells.

Figure 4

Persistence, migration, and safety of a high dose of ROR1 CAR-T cells. A, schematic design of the T-cell infusions. B, frequency of transferred T cells (%) within the CD3⁺CD4⁺ and CD3⁺CD8⁺ subsets in blood on day 1 after the infusion of ROR1 CAR and control tCD34⁺ T cells. Staining is performed with mAbs specific for CD3, CD4, CD8, and CD19 or CD34. C, frequency of ROR1 CAR and control tCD34⁺ T cells in PBMC, BM, and LN samples obtained on day 3 after the T-cell infusion. Staining was performed with mAbs specific for CD3, CD4, CD8, and CD19 or CD34, and gating on CD3⁺CD4⁺ or CD3⁺CD8⁺T cells. The frequency of ROR1 CAR and control tCD34⁺ T cells in each subset is shown in the bar graphs for each animal. D, frequency of ROR1⁺ B cells in the BM before and 3 days after infusion of ROR1 CAR-T cells. Shown are representative staining of macaque A13011 gated on CD19⁺ cells. The staining and gating strategy is shown in Supplementary Fig. S2. E, frequency of ROR1⁺ B cells in the LN before and 3 days after infusion of ROR1 CAR-T cells. The gating strategy is shown in Supplementary Fig. S3. Shown are data from macaque A13011.

Figure 5

Effect of T-APC challenge on transferred ROR1 CAR-T cells in vivo. A, absolute numbers of ROR1 CAR and tCD34⁺ T cells in CD3⁺CD4⁺ and CD3⁺CD8⁺ subsets of PBMC in A13011 and A13002 at the indicated days after T-cell transfer. Staining was performed with mAbs specific for CD3, CD4, CD8, and CD19 or CD34. The CBC was determined in an accredited clinical laboratory. The arrows indicate the day of the T-cell infusion. B, fold-change in the absolute numbers of ROR1 CAR-T cells in the blood after T-APC challenge. The number of endogenous or transferred CD3⁺CD4⁺ and CD3⁺CD8⁺ T cells/μL of blood was measured by flow cytometry before and on indicated days after the T-APC administration. C, presence of tROR1⁺ T cells before and after the T-APC infusion in PBMC obtained at days 1, 5, and 7 from an animal (A13011) with persisting ROR1 CAR-T cells. PBMC were stained with anti-CD3 and anti-ROR1 mAbs. D, presence of tROR1⁺ T cells in PBMC samples obtained before and at the indicated days after T-APC infusion in a control animal (A12022) without ROR1 CAR-T cells.

Figure 6

Toxicity monitoring after adoptive transfer of a high dose of ROR1 CAR-T cells. A, serum chemistry of macaque A13011 and A13002 before and at the indicated days after infusion of ROR1 CAR-T cells (5×10⁸/kg) and tROR1⁺T-APC challenge. Pancreatic, liver, and muscle enzymes, and glucose were measured in an accredited laboratory. The grey shaded area demarks the macaque specific normal range for each parameter. B, plasma IFNγ and IL6 levels before and at the indicated days after the T-cell infusion. Shown are data from macaque A13011. C, plasma adiponectin levels in animals A11047, A13011, and A13002 on the indicated days after the ROR1 CAR-T cell infusion. D, plasma adiponectin levels in control animals undergoing blood draw procedures alone (A12012) or receiving control gene-modified T cells (A12022, A13003).