Tumor-targeted non-ablative radiation promotes solid tumor CAR T-cell therapy efficacy

Hue Tu Quach; Matthew S Skovgard; Jonathan Villena-Vargas; Rebecca Y Bellis; Navin K Chintala; Alfredo Amador-Molina; Yang Bai; Srijita Banerjee; Jasmeen Saini; Yuquan Xiong; William-Ray Vista; Alexander J Byun; Andreas De Biasi; Masha Zeltsman; Marissa Mayor; Aurore Morello; Vivek Mittal; Daniel R Gomez; Andreas Rimner; David R Jones; Prasad S Adusumilli

doi:10.1158/2326-6066.CIR-22-0840

. Author manuscript; available in PMC: 2024 Apr 4.

Published in final edited form as: Cancer Immunol Res. 2023 Oct 4;11(10):1314–1331. doi: 10.1158/2326-6066.CIR-22-0840

Tumor-targeted non-ablative radiation promotes solid tumor CAR T-cell therapy efficacy

Hue Tu Quach ^1,^†, Matthew S Skovgard ^1,^†, Jonathan Villena-Vargas ^1,^†, Rebecca Y Bellis ¹, Navin K Chintala ¹, Alfredo Amador-Molina ¹, Yang Bai ^2,³, Srijita Banerjee ¹, Jasmeen Saini ¹, Yuquan Xiong ¹, William-Ray Vista ¹, Alexander J Byun ¹, Andreas De Biasi ¹, Masha Zeltsman ¹, Marissa Mayor ¹, Aurore Morello ¹, Vivek Mittal ^2,³, Daniel R Gomez ⁴, Andreas Rimner ⁴, David R Jones ¹, Prasad S Adusumilli ^1,⁵

PMCID: PMC10592183 NIHMSID: NIHMS1924120 PMID: 37540803

Abstract

Infiltration of tumor by T cells is a prerequisite for successful immunotherapy of solid tumors. In this study, we investigate the influence of tumor-targeted radiation on chimeric antigen receptor (CAR) T-cell therapy tumor infiltration, accumulation, and efficacy in clinically relevant models of pleural mesothelioma and non-small cell lung cancers. We use a non-ablative dose of tumor-targeted radiation prior to systemic administration of mesothelin-targeted CAR T cells to assess infiltration, proliferation, anti-tumor efficacy, and functional persistence of CAR T cells at primary and distant sites of tumor. A tumor-targeted, non-ablative dose of radiation promotes early and high infiltration, proliferation, and functional persistence of CAR T cells. Tumor-targeted radiation promotes tumor-chemokine expression and chemokine-receptor expression in infiltrating T cells, and results in a subpopulation of higher-intensity CAR-expressing T cells with high co-expression of chemokine receptors that further infiltrate distant sites of disease, enhancing CAR T-cell anti-tumor efficacy. Enhanced CAR T-cell efficacy is evident in models of both high-mesothelin-expressing mesothelioma and mixed-mesothelin-expressing lung cancer—two thoracic cancers for which radiation therapy is part of the standard of care. Our results strongly suggest that the use of tumor-targeted radiation prior to systemic administration of CAR T cells may substantially improve CAR T-cell therapy efficacy for solid tumors. Building on our observations, we describe a translational strategy of “sandwich” cell therapy for solid tumors that combines sequential metastatic site–targeted radiation and CAR T cells—a regional solution to overcome barriers to systemic delivery of CAR T cells.

Keywords: Chemokines, Chemokine receptors, Radiation therapy, Solid tumors cell therapy, Mesothelin CAR

BACKGROUND

Infiltration of tumor by T cells is the first prerequisite for successful immunotherapy of solid tumors (1,2). Immune checkpoint inhibitor (ICI) agents reinvigorate exhausted tumor-infiltrating lymphocytes in immunologically “hot” tumors but are ineffective in “cold” tumors (3). The use of adoptive cell therapy aims to increase the number of tumor-targeted T cells by transducing them with genetically engineered synthetic receptors—chimeric antigen receptors (CARs) or T-cell receptors (4–6). Despite an increase in the number of antigen-targeted T cells, systemically administered CAR T cells targeting solid tumors are constrained by pulmonary sequestration, ineffective trafficking, and poor tumor infiltration (5,7). Regional administration of CAR T cells, translated by our group and others, bypasses pulmonary sequestration and aids in early antigen activation of CD4 CAR T cells, augmenting their helper function to enhance accumulation of CD8 CAR T cells (8,9). Antigen-activated cytotoxic T lymphocytes and CAR T cells, by diffusion of chemokines, use intrapopulation signaling to drive rapid convergence to the tumor, a phenomenon akin to “swarming” of neutrophils (10). These processes underscore the significance of early antigen activation of CAR T cells. However, the use of a regional delivery strategy alone may not be adequate to treat solid tumors with metastases in multiple organs.

Ablative doses of tumor-targeted radiation therapy (RT), an integral part of multimodality therapy for solid tumors, have been combined with ICI agent immunotherapy with the goal of generating neoantigen responses and promoting immunity (11–13). The inflammatory effects of RT have been described, including vascular remodeling, increased IP10 expression, enhanced T-cell activation from released IFN-γ, and endothelial cells adhesion molecule changes (14,15). In addition, RT is known to induce a chemokine gradient in the tumor that promotes immune cell infiltration (16,17). Ablative, cytotoxic doses of RT can promote an immune-inhibitory infiltrate rich in macrophages and regulatory T cells (Tregs) (18–20). Cyclophosphamide, a commonly used lymphodepleting agent for preconditioning in CAR T-cell therapy, is known to decrease Tregs and macrophages in solid tumors (21–25). We hypothesized that, in the setting of a cyclophosphamide-modulated tumor microenvironment, non-ablative (non-cytotoxic) doses of RT can generate a chemokine gradient that promotes tumor infiltration and accumulation of systemically administered CAR T cells to enhance antitumor efficacy without undue toxicity.

We studied the effects of combining a non-ablative dose of RT with CAR T-cell therapy for the treatment of mesothelin (MSLN)–expressing solid tumors. MSLN is a cell-surface antigen that promotes aggressiveness in solid tumors; MSLN-targeted CAR T-cell therapy is currently being studied in clinical trials by our group (NCT02414269, NCT02792114, NCT04577326) and others (26–28); safety and anti-tumor efficacy in a phase I trial was published (29). Herein, we demonstrate beneficial antitumor immunity in both high-MSLN-expressing malignant pleural mesothelioma (MPM) and mixed-MSLN-expressing non-small cell lung cancer (NSCLC) mouse models, two thoracic cancers for which RT is part of the standard of care (30–32).

MATERIALS AND METHODS

Experimental Design

The objective of the study is to investigate the influence of tumor-targeted radiation on CAR T-cell therapy efficacy in clinically relevant models of pleural mesothelioma and non-small cell lung cancers. The design of the study focused on clinical translation of mechanistic observations. All the reagents and models were previously published. In vitro experiments were conducted with at least 3 replicates and in vivo experiments were repeated at least 3 times for reproducibility.

Tumor Cell Lines

MSTO-211H (human pleural mesothelioma) and A549 (human NSCLC) cells were obtained from American Type Culture Collection. MSTO-211H and A549 cells were retrovirally transduced to express the GFP–firefly luciferase fusion protein (GFP-ffLuc+) to facilitate the performance of noninvasive in vivo bioluminescent imaging (BLI) (33). These cells were then also transduced with the human MSLN-variant 1 (isolated from the human ovarian cancer cell line OVCAR-3) subcloned into an SFG retroviral vector to generate MSTO MSLN+ GFP-ffLuc+ (abbreviated MSTO-GM hereafter) or A549 MSLN+ GFP-ffLuc+ (abbreviated A549-GM hereafter) (33,34).

Gammaretroviral Vector Construction and Viral Production

The construction and validation of MSLN-specific CARs, PD1DNR, and 1XX have been described previously. We inserted an internal ribosomal entry site to facilitate bicistronic expression of CARs with an LNGFR or Myc reporter gene, as illustrated in Fig. 1A. The resultant SFG-M28z and SFG-P28z constructs were then transfected into 293T H29 packaging cell lines, and these viral supernatants were used to transduce and generate stable 293T RD114 cell lines, as previously described (35).

Fig. 1. — **(A)** Mesothelin-targeted CAR construct with a CD28 costimulatory domain (M28z) with LNGFR tag or Myc tag for flow cytometric detection of CAR T cells. A prostate-specific membrane antigen–targeted CAR construct with CD28 costimulatory domain (P28z) served as a negative control. **(B)** M28z CAR T cells, but not P28z, lysed mesothelin-positive tumor cells, as measured by cytotoxicity assay (chromium-release assay). Tumor cell radiation (4 Gy) before CAR T-cell treatment did not enhance cytolysis. **(C)** NOD/SCID gamma mice with pleural mesothelioma monitored for tumor progression or regression by bioluminescence imaging. Mice received tumor-targeted thoracic radiation (4 Gy) followed by a single low dose of intravenous CAR T cells (5×10⁴). Tumors progressed without treatment (control) or with radiation alone. Tumors regressed in 2 of 6 mice after CAR T cells alone and in all 6 mice (up to 60 days) after preconditioning radiation and CAR T cells. **(D)** Radiation with CAR T cells yielded potent early tumor regression, compared with CAR T cells alone (50-fold difference at treatment day 16; P<0.001). **(E-F)** Kaplan-Meier survival analysis demonstrates superior antitumor efficacy of radiation with CAR T cells at multiple doses. **(E)** At a dose of 5×10⁴ CAR T cells, median survival in mice that received radiation with CAR T cells was 94 days, compared with 25 days in mice that received CAR T cells alone (P<0.05). (F) At a dose of 1×10⁵ CAR T cells, median survival was not reached in mice that received radiation with CAR T cells, compared with 57 days in mice treated with CAR T cells alone (P<0.05) and 27 days in control mice (n=6–8 mice/group). Student’s t tests were performed, and survival was analyzed using the log-rank test. *P<0.05; ***P<0.001. Error bars represent ± standard error of the mean. † indicates death.

T-Cell Isolation and Gene Transfer

Peripheral blood leukocytes were isolated from the blood of healthy volunteer donors under a Memorial Sloan Kettering Cancer Center institutional review board–approved protocol. CAR transduction has been described previously. Transduced cells were either used immediately in experiments or cultured in RPMI 1640 supplemented with 10% fetal bovine serum, 100 units/mL penicillin, and 100 μg/mL streptomycin. For in vivo experiments, media were also supplemented with 20 units/mL IL-2.

CAR T-Cell Cytotoxicity Assays

Cytotoxicity of T cells transduced with the MSLN-specific CAR (M28z) or irrelevant-antigen control (P28z) was determined by use of standard ⁵¹Cr-release assays, as previously described. Data are reported as the mean of triplicate measurements ± standard deviation and were analyzed using Microsoft Excel (Microsoft, Redmond, WA) or GraphPad Prism (GraphPad Software, San Diego, CA).

In Vitro Chemokine and Cytokine Detection Assays

Cytokine-release assays were performed as follows on 5×10⁵ MSTO-211H cells plated in triplicate in 200 μL of media in 96-well round-bottom plates. One day after plating, experimental-group MSTO-211H cells were irradiated with 2, 4, or 10 Gy using an X-Rad 250 Biological Irradiator (Precision X-Ray, North Branford, CT). Four days after plating (i.e., 3 days after radiation pretreatment), 50 μL of supernatant was collected for cytokine analysis. Cytokine levels were determined using multiplex bead Human Cytokine Detection kits (EMD Millipore, Burlington, MA) for IL-2, IL-4, IL-6, IL-10, IL-17, MIP-1, MCP-1, RANTES, GM-CSF, TNF-α, and IFN-γ; kits were run on a Luminex IS100 system. Values represent the mean of the triplicate wells ± standard deviation. These data were analyzed using IS 2.3 software (Luminex Software, Inc., Riverside, CA), Microsoft Excel, and GraphPad Prism.

In Vitro CAR T-Cell Accumulation Assay with and without Radiation-Conditioned Media

M28z-transduced T cells were stimulated with either nonirradiated MSTO-GM cells in the presence of media conditioned by 4 Gy–irradiated MSTO-GM cells or with MSTO-GM cells that themselves had been exposed to 4 Gy of radiation; subsequent CAR T-cell accumulation was then quantified. As controls, 1×10⁶ M28z-transduced T cells were suspended in media conditioned by either nonirradiated or 4 Gy–irradiated MSTO-GM cells (conditioned for 48 h) and were cultured in the absence of target MSTO-GM cells. Four days later, the T cells were collected, counted by hemocytometry, analyzed by flow cytometry, and then restimulated using the above-described protocol.

Boyden Chamber CAR T-Cell Migration Experiments

For these experiments, 1×10⁶ M28z-transduced T cells were suspended in 100 μL of serum-free RPMI 1640 and placed into the upper well of a Boyden chamber (pore size, 3 mm); the lower well contained 600 μL of either serum-free RPMI 1640, serum-free RPMI 1640 conditioned by MSTO-GM cells for 24 h, or serum-free RPMI 1640 conditioned for 24 h by MSTO-GM cells that had been exposed to 4 Gy of radiation 3 days prior. T cells were allowed to migrate from the upper to the lower wells for 4 h, at which point the T cells in the lower wells were collected, counted by hemocytometry, and analyzed for CAR expression and CD4 versus CD8 breakdown by flow cytometry. Chemokine blockade experiments were conducted by co-culturing antigen-activated CAR T cells with or without anti-CXCR3 blocking antibody and placed into the upper well of the Boyden chamber, with the lower well containing media conditioned by MGM cells treated with or without 4 Gy radiation. After 18 hours, cells that migrated to the bottom well were collected, enumerated, and analyzed by flow cytometry.

In Vitro CAR T-Cell Proliferation Assays

MSTO-GM cells were grown in 10-cm round plates to 50% confluency; half of these plates were then exposed to 4 Gy of radiation using an X-Rad 250 Biological Irradiator. Three days later, both the nonirradiated and the irradiated cells were harvested; 1×10⁵ nonirradiated or 4 Gy–irradiated MSTO-GM cells were then plated using 24-well plates. One day later, 5×10⁵ M28z-transduced T cells labeled with 5 μM of CellTrace Violet cell proliferation stain (Life Technologies, Carlsbad, CA) were overlayed on the MSTO-GM targets plated the day prior. The CAR T cells were then allowed to proliferate for 3 days, at which point they were harvested and analyzed for CAR positivity and CD4 versus CD8 breakdown by flow cytometry.

In Vitro Chemokine Receptor Profiles

Chemokine receptor profiling was performed using healthy donor peripheral blood mononuclear cells that were isolated and genetically modified as described above. On day 6 after isolation, cells were exposed for 24 h to MGM cells (1:1 ratio) or the supernatant of irradiated or nonirradiated MGM cells. On day 7 after isolation, peripheral blood mononuclear cells were bead-sorted for CD3+, M28z CAR+ into CD4 and CD8 subtypes on a BD FACS Aria III sorter (BD Biosciences, San Jose, CA) in the Sloan Kettering Institute Flow Cytometry Core Facility (Memorial Sloan Kettering Cancer Center, New York, NY). Immediately after sorting, the cell pellets were frozen at −80°C until analysis. RNA isolation was performed using the RNeasy Plus Mini Kit in accordance with the manufacturer’s protocol. RNA concentration and purity were measured using the NanoDrop 2000 kit (ThermoScientific, Waltham, MA). Reverse transcription was performed using the RT2 First Strand kit (Qiagen, Germantown, MD) with 100 to 500 ng per sample, in accordance with the manufacturer’s protocol. Real-time quantitative polymerase chain reaction was performed using RT2 SYBR Green Mastermix (Qiagen) and the premade Human Chemokines and Receptors Profiler Array (Qiagen), in accordance with the manufacturer’s protocol. These data were analyzed using Microsoft Excel and GraphPad Prism.

Orthotopic MPM Mouse Model and Adoptive T-Cell Therapy

The generation of orthotopic MPM mice using female NOD/SCID gamma mice (Taconic, Rensselaer, NY) has been described previously (36). All procedures were approved by the Institutional Animal Care and Use Committee (approval #08-001). Tumor-bearing mice underwent thoracic radiation 9 days after tumor inoculation. Specifically, mice were anesthetized with inhaled isoflurane and oxygen and placed in a X-Rad 320 x-ray unit (Precision X-Ray). The thoracic cavity was isolated with lead shields (Supplementary Fig. S1), and 4 Gy of radiation was delivered. Next, 5×10⁴ to 2×10⁶ CAR T cells were adoptively transferred into tumor-bearing mice at specified time points, as described in the Results section.

Intra-pulmonary NSCLC Mouse Model and Systemic T-Cell Therapy

Human NSCLC mice were developed as described previously (34). High-, low-, and mixed-MSLN-expressing tumors were generated via selective combination of transduced cell lines. Mice were placed into cohorts 18 days after tumor cell administration to ensure equal tumor burden between groups; mice subsequently underwent thoracic radiation with 4 Gy, as described above. Twenty-one days after tumor injection, 5×10⁴ transduced T cells suspended in 200 μL of serum-free media were systemically administered into mice via the dorsal tail vein.

Orthotopic MPM Mice with Flank Tumor Model and CAR T-Cell Therapy

Female NOD/SCID gamma mice at 6 to 10 weeks old underwent injection of 2×10⁶ MGM cells suspended in 200 μL of serum-free media and Matrigel (BD Biosciences, Cat 356234, Franklin Lakes, NJ) in a 1:1 ratio into the right flank caudal to the hind limb. Seven days later, mice were established with orthotopic MPM, as described previously. Mice underwent thoracic-only irradiation 3 days before intrapleural administration of CAR T cells. Flank tumor volumes were measured every 2 days with digital calipers and calculated as follows: tumor size (mm³) = [width (mm)]² × [length (mm)]/2. Mice were sacrificed at predetermined timepoints to harvest tumor and organs for flow cytometric analysis of total T cells and CAR T cells.

Quantitative Tumor and T-Cell BLI

BLI was used in vivo to assess tumor burden in tumor-bearing mice (36). Animals were injected with a single intraperitoneal dose of 150 mg/kg D-Luciferin. Mice were then imaged with the Xenogen IVIS 100 Imaging System (Caliper Life Sciences, Hopkinton, MA) 15 min after D-Luciferin injection. Images were acquired for 5 to 30 s, depending on signal strength. BLI data were analyzed using Living Image (version 4.61, Caliper Life Sciences); the BLI signal is reported as total flux per second (photons/s). Individual mouse signals were determined by averaging the dorsal and ventral BLI signals for each animal using Microsoft Excel; resultant data were analyzed using GraphPad Prism. BLI was also used to visualize in vivo trafficking of adoptively transferred CAR T cells. T cells transduced with an enhanced firefly luciferase reporter gene (provided by P. Hwu, M.D. Anderson Cancer Center) were transferred into mice by a single intrapleural injection. T cells were visualized by injecting the T cell–bearing mice with a single intraperitoneal dose of 150 mg/kg D-Luciferin and imaging the animals, using the Xenogen IVIS 100 Imaging System, for 120 s at 20 min after injection.

In Vivo Serum Chemokine and Cytokine Detection Assay

Both human and mouse serum chemokine and cytokine concentrations were then determined 24 h, 72 h, and 5 days after treatment by use of multiplex bead-based enzyme-linked immunosorbent assay, using multiplex bead Human Cytokine Detection kits (EMD Millipore) for IL-2, IL-4, IL-6, IL-10, IL-17, MIP1, MCP-1, RANTES, GM-CSF, TNF-α, and IFN-γ; kits were run on a Luminex IS100 system. Values represent the mean of triplicate mice ± standard deviation. These data were analyzed using IS 2.3 software (Luminex), Microsoft Excel, and GraphPad Prism.

Harvesting and Preparation of CAR T Cells in Tumor, Spleen, and the Peritoneal Cavity

To analyze tumors, spleens, and the peritoneal cavity for infiltrating CAR T cells, mice were euthanized and necropsied. Tumor and splenic tissues were isolated, weighed, suspended in 30 mL of RPMI 1640 and then strained into single-cell suspensions in 50-mL conical tubes. After centrifugation at 1500 RPM for 5 min, the media was removed, and 1 mL of ACK lysis buffer (Thermo Fisher Scientific, Waltham, MA) was added on top of the cell pellet. Five minutes later, the cells were quenched with 10 mL of fluorescence-activated cell sorting (FACS) buffer and were again spun down, as above; finally, the cells were resuspended in 1 mL of FACS buffer. At this point, the tissue samples were stained for FACS analyses, as detailed below. In experiments involving collection of CAR T cells in the peritoneal cavity, a 7-mm incision was made overlying the spleen, and the spleen was tied off at the pedicle and removed. The abdominal incision was closed, and the peritoneal cavity was isolated. Next, 3 mL of phosphate-buffered saline was instilled into the abdomen under direct visualization. The peritoneal compartment was agitated, and the resultant suspension was aspirated. This was repeated for a total of 3 washes, and cells were isolated and analyzed as described above.

Flow Cytometry and T-Cell Proliferation Labeling

We used fluorochrome-conjugated antibodies to MSLN (anti-MSLN PE-conjugated Rat IgG2A FAB32652P; R&D Systems), CD3 (PE/Cy7 anti-human CD3 Clone HI3A, 300316; Biolegend), CD45 (FITC Mouse anti-human CD45, 555482; BD Pharmingen), CD4 (APC Mouse Anti-Human CD4, 55349; BD Pharmingen), CD8 (FITC Mouse Anti-Human CD8, 561948; BD Pharmingen), CD62L, CD45RA, and CXCR3 (AF700 mouse anti-human CXCR3, 353742; Biolegend) (Supplementary Table S1). In assessing in vitro and in vivo T-cell proliferation and accumulation, cells were labeled with either CFSE or CellTrace Violet cell proliferation stain, in accordance with the manufacturers’ instructions. Cells were quantified using CountBright Absolute Counting Beads (Life Technologies). All flow cytometric analyses were performed on either a BD FACSCalibur or LSR II (BD Biosciences, San Jose, CA); data were analyzed using FlowJo software (version 6.0, TreeStar).

In Vivo Gene Expression Analysis

Gene expression analysis was performed using Nanostring Technologies (Seattle, WA). The tumors or spleens of 4 mice were pooled, and CAR T cells were isolated by flow sorting, followed by isolation of RNA as described above, ensuring that adequate (>100 nanograms) RNA was obtained. Samples were then prepared as outlined by the manufacturer for the Nanostring nCounter Spring cartridge using the CAR T cell Characterization Panel, a 770-plex gene expression panel that profiles CAR T cells. Samples were run in duplicates. Analysis was performed with Nanostring nSolver software (version 3.0, Nanostring Technologies).

HKP1 Cancer Models and Irradiation Treatment

HKP1 lung cancer cells expressing mCherry-luciferase were derived from spontaneous KrasLSL-G12D/+; p53flox/flox (KP) mice as described previously (37). Lung orthotopic tumors were generated via injecting 1.5 × 105 HKP1 cells in 100 μl sterile PBS through tail vein into 7-week-old female C57BL6/J mice. Tumors in the lungs was monitored by bioluminescence imaging twice a week. Images were taken on the IVIS system at day 6, 9, 12, 16 after HKP1 cells administration. For BLI plots, photon counts of each tumor area with the same area of interest were measured and normalized to the day 6 value of each mouse. For RT, mice were anesthetized during the process. Mice were irradiated using the Small Animal Radiation Research Platform Dose Planning System (Xstrahl, Inc., Suwanee, GA) to precisely target each lung. Radiation dosage is 4Gy for three consecutive days from day 10 to 12. The HKP1 tumor-bearing lungs were collected at day 13 for scRNA-seq analyses. Three tumor bearing lungs were pooled to generate single cell suspension for each condition.

scRNA-seq Analysis

Single-cell suspension was generated by physical and enzymatic dissociation of tumor-bearing lungs. Cells were stained with DAPI and fluorescent-labeled CD45 antibody followed by sorting of CD45− and CD45+ cells via flow cytometry (Aria). Cell viability was measured at Weill Cornell genetics core using an automated cell counter (Bio-Rad Laboratories, Inc, Hercules, CA). Samples with higher than 70% viability were submitted for scRNA-seq. Single-cell droplets, barcoding and libraries were generated using 10X Genomics Chromium platform. 6,000 cells from each condition were targeted to sequence. A test-run of all the samples on Mi-seq (Illumina, Inc., San Diego, CA) and analysis on Cell Ranger software (10X Genomics, Pleasanton, CA) were performed to estimate the cell number and evaluate the library quality. All the samples were pooled and sequenced at around 50,000 reads/cells depth on the NovaSeq sequencer (Illumina) to avoid batch effect. The scRNA-seq data were analyzed using the Partek Flow software (Partek, Inc., Chesterfield, MO). The sequencing data were trimmed and aligned to the mouse transcriptome reference (mm10) by STAR alignment method. Following de-duplication of unique molecular identifiers, filtering of background signals and noises, the cell count was quantified of each sample. Single-cell quality assurance/quality control was then controlled by the total number of genes (≤~6,000), unique molecular identifiers (≤~40,000) and the percentage of mitochondria genes (≤10%) of each sample. The top 10 principal components were used for t-SNE/UMAP visualization. Cell clusters were generated using unsupervised clustering and CD8 T cells and tumor cells are manually assigned based on the expression of CD8a and KRT19 respectively. Differential gene analysis was performed between groups using GSA analysis.

Statistical Analysis

Data were analyzed using GraphPad Prism software (version 6.0) and are presented as mean ± standard error of the mean, as stated in the figure legends. Results were analyzed using unpaired Student’s t tests (2-tailed), with the Sidak-Bonferroni correction used to correct for multiple comparisons when applicable. Survival curves were analyzed using the log-rank test. Statistical significance was defined as P<0.05. All statistical analyses were performed using GraphPad Prism software.

Data Availability

The data generated in this study are available within the article and its supplementary data files. The gene expression data is available at GSE157881 (Sub-series part of super-series GSE157883)

RESULTS

Non-ablative Tumor-Targeted RT Promotes the Antitumor Efficacy of CAR T Cells

Our CAR construct (NCT02414269, NCT02792114), M28z, incorporates a human MSLN-specific single-chain variable fragment (38), a CD28/CD3ζ signaling domain, an SFG–retroviral vector (8,39), and an LNGFR or Myc tag for identification by flow cytometry (Fig. 1A). A prostate-specific membrane antigen–specific CAR construct (P28z) served as a negative control. MSLN-expressing MSTO-211H (MSLN+) cells served as targets. Non-ablative radiation (4 Gy) administered to tumor cells in vitro 3 days prior to CAR T-cell coculture had no effect on MSLN expression (Supplementary Fig. S1A–B) or cytolysis and did not increase the cytotoxicity of M28z CAR T cells (Fig. 1B). A single dose of non-ablative radiation (4 Gy) was delivered to the thoracic cavity of tumor bearing mice ([Supplementary Fig. S1C]). A low dose (5×10⁴) of M28z or P28z CAR T cells was administered intravenously 3 days after RT (Fig. 1C); no side effects were observed with the combination. On tumor bioluminescence imaging (BLI) (8,33,36,39,40), non-ablative RT alone had no effect on tumor progression (Fig. 1C). Whereas 2 of 6 mice that received a low dose of CAR T cells alone had tumor regression, all 6 mice that received non-ablative tumor-targeted RT before M28z CAR T cells had potent and early tumor regressions (Fig. 1C). By day 16 after treatment, the average tumor burden was 50-fold lower in mice treated with RT and CAR T cells, compared with mice treated with T cells alone (P<0.001) (Fig. 1D). Combination therapy resulted in nearly 4-fold longer survival (94 days vs 25–27 days in the other 3 groups; P<0.05) (Fig. 1E). At a 1×10⁵ dose (effector to target [E:T] ratio, 1:3000), CAR T cells alone modestly improved survival, compared with controls (median survival, 57 vs 26 days) (Fig. 1F). In mice treated with RT before CAR T cells, median survival was not reached at 100 days (P<0.05).

Tumor-Targeted RT Enhances Accumulation of CAR T Cells in the Tumor

Flow cytometric analysis of harvested tumors at 7 days showed that accumulation of CD3+CD45+ T cells increased by 2-fold in mice pretreated with 4 Gy tumor-targeted RT prior to administration of 2×10⁵ CAR T cells compared to mice treated with CAR T cells alone (P<0.05) (Fig. 2A, right).

To validate our results, we doubly transduced T cells with a CAR and an enhanced firefly luciferase reporter gene (Effluc) (Fig. 2B–C). The BLI signal produced by Effluc+ T cells linearly correlated with T-cell quantity (Fig. 2C, right), which allowed for relative quantification. Accumulation of T cells was then serially assessed by imaging in mice bearing MPM that were treated with (4 Gy) or without RT followed by CAR T cells (Fig. 2D). Tumor-targeted RT promoted early (5 days with RT vs 7 days without RT) and robust (5-fold higher; P<0.01) accumulation of T cells in the pleural tumor, predominantly (80%) CAR T cells (Fig. 2E). In contrast, in mice without RT, systemically administered CAR T cells were sequestered in the lungs (Supplementary Fig. S2). Thoracic RT did not result in increased engraftment of CAR T cells in the lungs or other organs (Fig. 2F).

RT Pretreatment Enhances Chemotaxis, Tumor Infiltration, Proliferation, Memory, and Persistence of CAR T Cells

Given that pretreatment with RT promoted early and higher accumulation of systemically administered CAR T cells, we first assessed chemotaxis using a Boyden chamber assay. Chemotaxis of CAR T cells was higher to media from nonradiated tumor than to control media (P<0.05) (Fig. 3A), and the chemotactic response favored CD4 CAR T cells over CD8 CAR T cells (Supplementary Fig. S2A). CAR T-cell chemotaxis was higher with media from irradiated tumor than with media from nonirradiated tumor (Fig. 3A).

Fig. 3. — **(A)** Chemotaxis of CAR T cells was statistically significantly higher (with CD8 predominance, right panel) toward media from irradiated tumor cells than toward media from tumor cells alone, as assessed using a Boyden chamber assay. Unconditioned media served as controls. **(B-C)** Preconditioning radiation facilitated CD8 CAR T-cell infiltration (B) and proliferation in the tumor (C). Following preconditioning radiation (4 Gy), mice bearing orthotopic mesothelioma were administered carboxyfluorescein succinimidyl ester–labelled CAR T cells. Flow-assisted cell sorting analysis at day 7 demonstrated higher infiltration of CD8 CAR T cells (Generation 0) in irradiated tumor, compared with nonirradiated tumor. Higher CD8 CAR T cell proliferation (Generation ≥1) by site of radiation was noted in the tumor but not in the spleen. **(D-E)** Radiation was associated with a preserved central memory phenotype of CAR T cells. Phenotyping, as quantified by flow cytometry (D), demonstrated a higher fraction of central memory cells (P<0.01) and lower fraction of terminal effector cells (P<0.05) in irradiated tumor, compared with nonirradiated tumor, in vivo. (E) Memory-specific gene expression (CD62L, CD127, and CD45RO) was higher in CAR T cells harvested from irradiated tumor, compared with nonirradiated tumor (P<0.05, P<0.01, and P<0.05, respectively). **(F)** Radiation promoted long-term persistence of CAR T cells. In mice treated with or without radiation and 2×10⁶ CAR T cells, tumor was eradicated. **(G)** CAR T cells eradicated pleural tumor by day 18, with no relapse until sacrifice on day 56. Analysis of harvested spleen by flow cytometry showed robust CAR T-cell persistence in mice that received radiation before CAR T cells, compared with mice that received CAR T cells alone (P<0.05; n=4–8 mice/group). *P<0.05; **P<0.01, by Student’s t test with Bonferroni correction. Error bars represent ± standard error of the mean.

In vivo, mice with orthotopic MPM tumors were treated with (4 Gy) or without RT followed by carboxyfluorescein succinimidyl ester (CFSE)–labeled CAR T cells (Fig. 3B). Seven days after treatment, mice were sacrificed, and infiltrating CAR T cells (Generation 0) harvested from tumor were quantified. RT predominantly enhanced infiltration of CD8 CAR T cells, with a nearly 4-fold increase in CD8 CAR T cells (P<0.05) (Fig. 3B, right).

In addition to promoting early and higher infiltration, pretreatment with RT enhanced proliferation of CAR T cells. In irradiated tumors, CAR T cells underwent 3 to 4 proliferation cycles (P<0.05) (Fig. 3C, left; Supplementary Fig. S3A), compared with nonirradiated tumors, in which most CAR T cells remained within 3 generations. Furthermore, proliferation was site specific and limited to irradiated tumor, with no enhancement in the spleen (Fig. 3C, middle). RT promoted proliferation of CD8 CAR T cells (4.5-fold; P<0.05) (Fig. 3C), with no statistically significant difference in proliferation of CD4 CAR T cells. To confirm that these observations were not secondary to the early and higher infiltration of CAR T cells, we performed an in vitro proliferation assay (Supplementary Fig. S3C); CFSE-labeled CAR T-cell proliferation was 3-fold higher against irradiated tumor than nonirradiated tumor (P<0.001).

In addition to increased tumor infiltration and proliferation, CAR T cells were harvested from tumor 12 days after treatment (Fig. 3D) and analyzed by flow cytometry showed the fraction of central memory (CD62L+CD45RA-) CAR T cells was 2- to 3-fold higher (P<0.01) and the fraction of terminally differentiated effector cells (CD62L-CD45RA+; Supplementary Fig. S3D–E) was lower (P<0.05) in irradiated tumors than nonirradiated tumors. Consistent with these observations, markers of memory phenotype (CD62L, CD127, and CD45RO) were higher by mRNA gene expression in irradiated tumors than nonirradiated tumors (P<0.05, P<0.01, and P<0.05, respectively, Fig. 3E).

We questioned whether pretreatment with RT—which was associated with an increase in tumor infiltration, proliferation, and memory phenotype of CAR T cells—would also result in long-term persistence. We administered a single therapeutic dose (2×10⁶) of CAR T cells with and without RT (Fig. 3F–G). Again, there was an earlier decrease in tumor burden with combination therapy (day 5–7). With and without RT, pleural tumor was eradicated by day 18, with no tumor relapse until mice were sacrificed on day 56. Spleens were harvested, and CAR T cells were quantified by flow cytometry (Fig. 3G). Persistence of CAR T cells in the spleen was 2.5-fold higher with thoracic RT than without thoracic RT (P<0.05).

RT Promotes Tumor Secretion of Chemokines

Following the observations of higher chemotaxis in vitro and tumor infiltration of CAR T cells in vivo, we postulated that RT-induced tumor secretion of chemokines may be the driving factor underlying these observations. In vitro, MSTO-211H cells were either sham irradiated or irradiated with 2, 4, or 10 Gy, and media were collected and analyzed using the Luminex assay. We observed a radiation dose–dependent increase in chemokine and cytokine production (72-h collection time point shown; Fig. 4A, Supplementary Fig. S4). In vivo (human and mouse chemokines measured by Luminex assay), serum collected from mice with orthotopic MPM that received thoracic RT (4 Gy) had higher chemokine levels 3 to 5 days after radiation, compared with mice that did not receive RT (Fig. 4B). Aforementioned results of radiation-induced chemokine upregulation were reproduced with mouse mesothelioma tumor, AB12 (Supplementary Fig. S4B). Mouse serum chemokine levels were insignificantly increased three days after RT in control mice without tumor that received thoracic RT (Supplementary Fig. S5), making it unlikely that RT-induced mouse chemokines promoted engraftment of CAR T cells. In addition, no significant increase in mouse serum chemokine levels was observed in mice without tumor three days after thoracic RT followed by CAR T cells (Supplementary Fig. S6).

Fig. 4. — **(A-B)** Tumor-secreted chemokines and cytokines, as quantified by Luminex assay, had dose-dependent (A, in vitro, baseline values normalized to 1) (B, in vivo, serial serum sampling after tumor-targeted radiation) chemokine secretion, with peak secretion 3 to 5 days after radiation. **(C)** Radiation-induced chemokine secretion was observed clinically. A patient with pleural mesothelioma received tumor-targeted palliative radiation. Serum chemokines, particularly corresponding ligands for CXCR3 (CXCL9 and CXCL11), increased after radiation. CAR T cells reappeared in peripheral blood following RT, as measured by vector copy number. **(D-E**) CXCR3 expression in CAR T cells infiltrating irradiated tumors. Tumor-bearing mice that received CAR T cells with or without radiation were sacrificed at the first sign of antitumor efficacy on tumor bioluminescence imaging (day 12; n=8 mice/group). Flow cytometry demonstrated a higher fraction of CXCR3-expressing CAR T cells (percent positive and quantity per gram of tumor; P<0.05 and P<0.01, respectively) in irradiated tumor, compared with nonirradiated tumor. The increase in CXCR3 expression was CAR T cell specific, as the untransduced T cells in the same tumors had no upregulation in CXCR3 expression. In irradiated tumor, the median fluorescent intensity of CXCR3 expression in CAR T cells was associated with an increase in CAR median fluorescence intensity; this was not observed in nonirradiated tumor (P<0.001). **(E)** Two representative flow cytometry plots demonstrate positive correlation between CXCR3 and CAR median fluorescence intensity. **(F)** By mRNA gene expression, CXCR3, CXCR6, CCR5, CCR2, and CX3CR1 were higher in CAR T cells in irradiated tumor, compared with nonirradiated tumor (≥2-fold; n=4 mice/group). **(G)** Furthermore, by mRNA gene expression, CAR T cells in irradiated tumor expressed higher markers of transendothelial migration (trend was noted, not statistically significant), compared with CAR T cells in nonirradiated tumor. *P<0.05; **P<0.01; ***P<0.001, by Student’s t test. Error bars represent ± standard error of the mean.

These findings led us to consider the experience of a MPM patient enrolled in our ongoing CAR T-cell clinical trial (NCT02414269) (27,29). Six weeks following intrapleural M28z CAR T cells administration, patients received palliative thoracic RT, serum chemokine levels increased, including those corresponding to CXCR3: CXCL9 and CXCL11 (Fig. 4C). Furthermore, CAR T cells reappeared in peripheral blood after RT.

Given that RT predominantly enhanced accumulation of CD8 CAR T cells, rather than CD4 CAR T cells, we quantitatively profiled the chemokine receptors of CD8 and CD4 T cells by use of an mRNA gene expression array (n=4 donors; Supplementary Fig. S7). Chemokine receptors with a >2-fold difference across all donors included CXCR3, CCR5, and CCR4 (Supplementary Fig. S7A). Consistent with the published literature, CXCR3 was predominant on CD8 CAR T cells, and CCR4 and CCR5 were predominant on CD4 CAR T cells; other chemokine receptors had heterogeneity between donors (Supplementary Fig. S7B). Given that CXCR3 is predominantly expressed by CD8 CAR T cells (which had higher chemotaxis, infiltration, and proliferation against irradiated tumors), and that tumor-targeted RT increased the expression of CXCR3 ligands (CXCL9, CXCL10, and CXCL11), we turned our focus to CXCR3 as an example rather than the sole underlying mechanism. In mice bearing pleural tumor treated with RT, serially harvested tumor lysates showed upregulated CXCR3 ligands (Supplementary Fig. S8).

Upregulated Chemokines in Irradiated Tumor Promote Corresponding Expression of Chemokine Receptors in Tumor-Infiltrating CAR T Cells

Tumors were harvested from mice with MPM tumors with or without RT pretreatment (Fig. 4D–G) 12 days after treatment, RT pretreatment increased antitumor efficacy (Fig. 4D, middle panel). In irradiated tumor, CAR T cells had higher proportion of CXCR3 positive cells (Fig. 4D, middle panel) (49% vs 38% in nonirradiated tumor; P<0.05), a higher quantity of CXCR3+ CAR T cells (5.5-fold increase; P<0.01), and a higher median fluorescence intensity (MFI) (Fig. 4D, lower panel). The increase in expression of CXCR3 was CAR T-cell specific, as we did not observe any differences in expression of CXCR3 among the untransduced (non-CAR) T-cell populations between irradiated and nonirradiated tumors (Fig. 4D, middle panel). Furthermore, simply exposing CAR T cells to the media of irradiated or nonirradiated tumor did not affect the expression of chemokine receptors (Supplementary Fig. S9).

Interestingly, the CAR (LNGFR) MFI was 2- to 2.5-fold higher in irradiated tumor than in nonirradiated tumor (P<0.001) (Fig. 4D). In irradiated tumor, the MFI of CXCR3 and CAR were correlated; CAR^high T cells had higher expression of CXCR3, a trend unique to CAR T cells harvested from irradiated tumor, as expression of CXCR3 on CAR^high T cells was not altered in nonirradiated tumor (Fig. 4E). CAR^high T cells have been proposed to be functionally superior to CAR T cells with lower levels of expression (41). Tumor-infiltrated CAR T cells were pooled and profiled by use of gene expression networks (Fig. 4F–G). Expression of chemokine receptors, including CXCR3, CXCR6, CCR2, CCR5, and CX3CR1, after infiltration of CAR T cells into irradiated tumor was higher than after infiltration into nonirradiated tumor and was also higher than preinjection levels (Fig. 4F). Given that expression of CXCR3, CCR2, and CCR5, which facilitate transendothelial migration, was increased, it is not surprising that the levels of multiple transendothelial migration markers were increased as well (Fig. 4G). We further confirmed the role of CXCR3 expression on CAR T cells migrating towards CXCR3 ligands by blockade of CXCR3 on antigen-activated CAR T cells (Supplementary Fig. S10).

Tumor-infiltrating CAR T Cells Following Tumor-targeted Radiation Promote Efficacy Against Distant Tumors

Tumor-targeted RT promoted CAR T cells with increased chemotaxis and tumor infiltration (Fig. 5A). To test whether CAR T cells with upregulated chemokine receptor expression are functionally superior at infiltrating tumor and promoting efficacy at distal nonirradiated sites, mice were established with flank and pleural MSLN+ tumors and treated with (4 Gy) or without RT to the thoracic cavity (Fig. 5B). Flank tumors were guarded with lead shields to ensure no exposure to radiation. A single dose of CAR T cells (1×10⁶) was regionally administered by direct intrapleural inoculation 3 days after RT. This relatively high dose was chosen to eradicate pleural tumor without relapse and with long-term survival of mice in both experimental groups, to allow monitoring of antitumor efficacy at the distal tumor in the flank. Furthermore, CAR T cells were administered intrapleurally to ensure that they trafficked through irradiated or nonirradiated tumor microenvironments first. Mice were sacrificed 11 days after treatment, and flank tumors were analyzed for CAR T cells by use of flow cytometry. We observed that, CD8 CAR T cells increased in the flank tumor, both CD4 and CD8 CAR T cells had increased expression of CXCR3 (Fig. 5B). These findings translated to superior antitumor efficacy at flank tumor sites, as measured by tumor volume and imaging (Fig. 5B, right).

Irradiated Tumor-infiltrating CAR T Cells Promote Functional Persistence

Whereas irradiated tumor-infiltrating CAR T cells were shown to have enhanced quantitative persistence, their functional persistence was not known. Keeping in focus our goal for translation, for the next set of experiments we used the CAR T-cell construct currently in use in our clinical trial (NCT04577326): an M28z1XXPD1DNR (Fig. 5C–F) CAR that co-expresses a PD-1 dominant negative receptor lacking an inhibitory signaling domain and a CD3ζ domain containing a single functional immunoreceptor tyrosine-based activation motif, termed 1XX (39,42,43). To assess the functional persistence of the CAR T cells, we developed and used a novel “tumor treadmill test,” which consisted of a series of escalating doses of tumor rechallenge at a distal site (Fig. 5D) to “stress” CAR T cells with a high tumor burden. In mice established with a single large tumor burden, long-term assessment is not possible, as a low dose of CAR T cells results in death of mice and a high dose result in tumor eradication, which alleviates antigen stress. Mimicking clinical conditions, the tumor treadmill test exposed CAR T cells to larger doses of tumor over a longer period, allowing for long-term assessment of antigen stress. In contrast to M28z CAR T cells, M28z1XXPD1DNR CAR T cells systemically administered at a single low dose (5×10⁴) are effective at eradicating pleural tumor burden with or without radiation. Six months after eradication of pleural tumor without relapse, escalating doses of tumor cells were administered into the peritoneal cavity every 3 to 6 days. Initially, at doses of 1–5×10⁶ tumor cells, there was no difference in peritoneal tumor regression on tumor imaging. At higher doses of 20–40×10⁶ tumor cells, superior efficacy was observed in mice that received a single dose of 4Gy thoracic RT (peritoneum shielded with lead) and CAR T cells, compared to mice that received CAR T cells alone. Fold difference in tumor burden 3 days after tumor administration (Fig. 5D, inset) demonstrated progressively improving distant site antitumor immunity in the RT group, culminating in a 12-fold difference in tumor burden after the sixth challenge (P<0.01).

At this time, the spleen of irradiated mice (Fig. 5E), CAR T cells had increased central memory (2-fold difference) and increased expression of CXCR3, compared with nonirradiated mice (P<0.05). In the peritoneal cavity, where antigen-expressing tumor cells were administered, terminal effector CD8 CAR T-cell expression of CXCR3 was higher in mice that received thoracic RT than in mice that did not (P<0.01) (Fig. 4F). Interestingly, there was no difference in the quantity of CAR T cells in the peritoneum, suggesting that the differences in tumor regression were not necessarily quantitative (secondary to higher E:T ratios) in the radiation cohort but rather were attributable to potential functional superiority.

Radiation Potentiates CAR T-Cell Efficacy in Mixed -Antigen-Expressing NSCLC

Our published clinical data show that, whereas MPM tumors demonstrate uniform and strong expression of MSLN on the cancer cell surface, NSCLC tumors have mixed expression of MSLN (Fig. 6A), a known hurdle in the treatment of solid tumors (26,34). To ensure the reproducibility of our concept in mixed-antigen-expressing NSCLC, we developed and used a model of intra-pulmonary human lung adenocarcinoma with high, low, and mixed levels of MSLN expression (Fig. 6B). NSCLC cells, A549 promote chemokine secretion following treatment with RT (Supplementary Fig. S11). In the tumor generated by mixing 50% high- and 50% low-MSLN-expressing tumor cells (A549), only low-MSLN-expressing cells were transduced with green fluorescent protein (GFP)–luciferase, so that tumor progression or regression of low-antigen-expressing tumor cells could be observed, as they are the most likely source of tumor relapse in the heterogeneous setting. Indeed, CAR T-cell cytotoxicity was dependent on antigen expression in a density-dependent manner (Fig. 6C); only higher E:T ratios showed cytotoxicity against low-MSLN-expressing cells. As in MPM, RT increased tumor secretion of chemokines and cytokines in NSCLC (Supplementary Fig. S11).

In mice with high-MSLN-expressing NSCLC (Fig. 6D–F), treatment with non-ablative thoracic RT (4 Gy) followed by a low dose (5×10⁴) of M28z CAR T cells was associated with early increased antitumor efficacy, compared with CAR T cells alone (day 8, 2.5-fold difference). Furthermore, long-term tumor eradication was seen for >100 days in 6 of 8 mice that received RT and CAR T cells along with survival than in mice treated with CAR T cells alone (median survival, not reached at day 160 vs 107 days [P< 0.01] vs 21 days in controls vs 25 days with RT alone).

We postulated that, in mixed-antigen-MSLN-expressing NSCLC, increased tumor infiltration and subsequent proliferation in irradiated tumor would generate higher E:T ratios, resulting in augmented antitumor efficacy against low-MSLN-expressing cells. Mice established with mixed-antigen-MSLN-expressing NSCLC (only low-MSLN-expressing cells were transduced with GFP-luciferase) received a single dose of 5×10⁴ CAR T cells (Fig. 6G). The low-MSLN-expressing tumor burden was reduced 6-fold with RT versus a transient 2-fold reduction without RT. Mice treated with RT and CAR T cells had longer survival than mice treated with CAR T cells alone (median survival, 63 vs 43 days; P<0.001) (Fig. 6H).

In mice with NSCLC tumor with only low MSLN expression, RT again potentiated early antitumor efficacy (Supplementary Fig. S12A–C). However, at 5×10⁴ dose of CAR T cells (Supplementary Fig. S12D), only when mice were further treated with an anti–PD-1 ICI agent (3 mg/kg, dosed weekly beginning 1 week after CAR T-cell therapy) to prevent functional exhaustion of CAR T cells, survival in mice that received RT and CAR T cells was longer than in mice treated with CAR T cells and ICI agent (median survival, 77 vs 46 days; P<0.001).

With the observation that the addition of an ICI agent confers a survival benefit, we repeated the above experiments with the M28z1XXPD1DNR CAR construct currently in clinical trial. Mice established with high-, low-, and mixed-MSLN-expressing tumors treated with or without RT received a low dose (5×10⁴) of CAR T cells (Fig. 6I–J). In mice with high-MSLN-expressing tumors, CAR T cells eradicated tumor with or without RT (Fig. 6I, left). Against low-MSLN-expressing or mixed-antigen-MSLN-expressing tumor (Fig. 6I, center and right), killing of low-MSLN-expressing tumor was nearly 6-fold higher with RT than without RT. At this dose of CAR T cells, mice with mixed-MSLN-expressing NSCLC that were treated with preconditioning RT and CAR T cells had longer survival than mice treated with CAR T cells alone (median survival, not reached at 100 days vs 72 days; P<0.05).

Upregulated Chemokines in Irradiated Tumors and Corresponding Chemokine Receptor Expression in Tumor-infiltrating T Cells in a Syngeneic Lung Cancer Model

To validate our observations of upregulated chemokines in irradiated tumors and corresponding chemokine receptor upregulation in tumor-infiltrating T cells in an immunocompetent mouse model, we established lung tumors with HKP1 lung cancer cells expressing mCherry-luciferase derived from spontaneous KrasLSL-G12D/+; p53flox/flox (KP) mice as described previously (37). KP1 lung orthotopic tumors with or without RT pretreatment (44), were harvested 13 days after thoracic radiation, antitumor efficacy is noted by BLI (Fig. 7A). We observed a radiation dependent increase in chemokine and cytokine expression by the tumor (Fig. 7A). There were more T cells infiltrating irradiated tumors compared to non-irirradiated mice, and in addition showed enhanced CXCR3 expression (Fig. 7B) (44). There was an increase in lymphocyte chemokine receptor expression including CXCR3, CCR2, and CCR5 within the irradiated compared to nonirradiated tumors (Fig. 7B). Furthermore, we evaluated the chemokine expression in the tumor microenvironment (macrophages, dendritic cells, NK T cells, B cells, fibroblasts, and endothelial cells) including CD45+ cells as shown (Fig. 7C–D). Radiation upregulated chemokines in the tumor microenvironment as well as corresponding chemokine receptors on T cells compared to nonirradiated tumors. We also present chemokine upregulation following radiation in AB12 mouse mesothelioma cells (Supplementary Fig. S4B).

Fig. 7. — Lung orthotopic tumors generated via injecting 1.5 × 105 HKP1 cells through tail vein into 7-week-old female C57BL6/J mice were monitored for tumor progression or regression by bioluminescence imaging (BLI). Mice received tumor-targeted thoracic radiation (4 Gy) or mock radiation. Tumors progressed without treatment (control) and with radiation alone. GSA analysis of mRNA expression (day 13, around 5000 cells per group) shows **(A)** increased tumor specific cheekiness CXCL9 and CXCL11 in CD45- tumor cells and **(B)** higher expression of CXCR3, CCR2, and CCR5 occurred in T cells within irradiated tumor relative to non-irradiated tumor. **(C)** The TME (including CD45+ immune cells and CD45- cells) harvested from the lungs of mice at day 13. RT promotes increased chemokine expression in the tumor microenvironment by GSA analysis of single cell gene expression for all CD45+ cells. **(D)** Macrophages express increased chemokines (CCL12 and CCL2) and markers of transendothelial migration. GSA: Gene-set analysis, TME: tumor microenvironment, t-SNE: t-distributed stochastic neighbor embedding. Data represents the means of triplicates ± SEM. ****p<0.0001, ***p<0.001, **p<0.01, *p<0.05, n.s. not significant (One-way ANOVA).

DISCUSSION

Our study provides strong support for combination therapy comprising non-ablative tumor-targeted RT and systemic CAR T cells to overcome an important barrier in solid tumor immunotherapy: inefficient infiltration of tumor by T cells (5,7). In addition to quantitatively enhancing tumor-infiltrating CAR T cells, RT promotes a central memory phenotype and upregulation of chemokine receptors, which increases anti-tumor efficacy in solid tumors.

An important translational feature of our investigation is the use of a non-ablative dose of tumor-targeted RT. Unlike total-body irradiation used for preconditioning lymphodepletion or ablative-dose tumor-targeted RT used for cytotoxic purposes, non-ablative RT is likely to be better tolerated, an important necessity in a combination therapy. Cyclophosphamide, a common lymphodepleting agent used for preconditioning before adoptive cell therapy, depletes Tregs and macrophages (21–23,25), which may provide a therapeutic window of opportunity to administer non-ablative RT (18,45) to enhance infiltration, engraftment, proliferation and functionally superior phenotype of CAR T cells in the tumor, thus tilting the balance toward an effector response. Radiation-enhanced CAR T-cell infiltration, proliferation and phenotypic change are also antigen specific and not secondary to nonspecific T-cell activation, as we observed no enhancement of tumor-infiltrating non-CAR or control CAR T cells targeting an irrelevant antigen. However, radiation-induced immunogenic cell death and subsequent release of type 1 IFN might have influenced the observed CAR T-cell efficacy (46).

Upregulated expression of chemokine receptors on antigen-activated CAR T cells, a tumor-secreted chemokine gradient may itself serve as an attractant for T cells. To promote tumor infiltration—rather than depend on the expression of radiation-driven tumor-secreted chemokines—other investigators have adopted strategies of overexpressing a specific chemokine receptor, such as CCR2, CXCR1, CXCR2, or CXCR6 on CAR T cells (47–49). One inherent limitation of this approach, however, is the reliance on the secretion of specific chemokine ligands that are compatible with a particular chemokine receptor (50). Furthermore, it has been established that overexpression of a chemokine receptor alone does not necessarily result in improved function of CAR T cells (51). Whereas a soluble chemokine gradient may attract T cells to tumor, downstream chemokine receptor signaling pathways are a function of T-cell adhesion, activation status, and phenotype (51). In fact, in our investigation, the coculturing of resting CAR T cells with irradiated tumor supernatant did not increase chemokine receptor gene expression, and non-CAR T cells in irradiated tumor did not upregulate chemokine receptor expression. Exposure to irradiated tumor, on the other hand, generated CAR T cells that highly expressed multiple chemokine receptors, including CXCR3, CXCR6, CCR2, CCR5, and CX3CR1. Although we investigated one chemokine CXCR3 as an example, the anti-tumor efficacy might be a result of multiple factors including chemokine gradient, chemokine receptor upregulation, release of type 1 IFN and radiation-induced tumor vascular changes, which we did not investigate.

A portion of CAR T cells, as we observed, had increased expression of CARs on the cell surface as well as increased expression of chemokine receptors after infiltration of irradiated tumor. This subpopulation, termed CAR^high, which others have generated by activation of pure central memory T cells, also expressed increased markers of transendothelial migration and had enhanced persistence in secondary lymphoid organs. These phenotypic changes facilitated antitumor efficacy at distal tumor sites. Interestingly, generation of CAR^high T cells required contact with irradiated tumor, as exposure to the chemokine or cytokine gradient alone was insufficient to produce tumor-infiltrating CAR T cells. The underlying mechanism/s are the focus of our ongoing investigation.

Our results were reproduced in an NSCLC model with mixed-antigen (high and low) expression, and different chemokine expression. Additional tumor–T cell synapses from chemokine receptor or ligand adherence, and high E:T ratios resulting from augmented tumor-infiltration and proliferation of CAR T cells surpassing the lytic threshold of low-antigen-expressing target cells might have contributed to the additive cytotoxicity (52). A third potential mechanism may lie in the generation of CAR^high T cells. Activation of CAR T cells requires the tumor-antigen density to be above a certain threshold, but a higher expression of CAR combined with additional chemokine synapses can allow activation against low-antigen-expressing target cells. CAR^high T cells can initiate an activation cascade that induces a swarming effect of potent tumor killing; activation results in cytokine production, cell proliferation, and peripheral CAR T-cell recruitment, thereby achieving higher E:T ratios.

On the basis of these results, we propose a clinical trial of non-ablative radiation–guided sandwich cell therapy to treat metastatic cancers (Supplementary Fig. S13). Patients with metastatic thoracic cancers can first be conditioned with lymphodepleting cyclophosphamide. Tumor-targeted radiation can be “sandwiched” with systemically infused CAR T cells. Although it is not feasible to administer repeat doses of RT to the same tumor site, for fear of killing accumulated CAR T cells in the tumor, sequential radiation to metastatic sites will guide CAR T cells from one site to the next without impeding CAR T-cell function. Our proposed strategy is translational to increase the potency of CAR T cells against solid tumors. However, the proposed strategy is hypothetical, need to be tested in a phased fashion in the trial with RT to one site in the beginning and subsequently increasing the number of sites. Such an approach, if proven to be beneficial in our upcoming trial, can be adopted to other solid tumors.

Our study has limitations. We did not use a mouse CAR in an immunocompetent model to assess the benefit of neoantigen responses or suppressor immune cell inhibition, we did not assess the toxicity of RT when combined with cyclophosphamide, we did not perform investigations in additional tumor models, and we did not characterize chemokine profiles from multiple metastatic sites.

We have shown that tumor-targeted RT stimulates CAR T cells to provide potent and long-term antitumor immunity against solid cancers. These results—from potentiating T cells at the right place and the right time—highlight the importance of early antigen activation, as a single low dose of radiation results in CAR T cells that maintain functional superiority for as long as 6 months after initial tumor exposure. Our results strongly suggest that the use of locoregional radiation before administration of CAR T cells may substantially improve CAR T-cell therapy for solid tumors, warranting further clinical studies.

Supplementary Material

NIHMS1924120-supplement-1.pdf^{(6.2MB, pdf)}

NIHMS1924120-supplement-2.pdf^{(90.1KB, pdf)}

Synopsis.

Our study demonstrates that non-ablative, tumor-targeted radiation generates a tumor-secreted chemokine gradient that facilitates CAR T-cell chemotaxis, early and higher tumor infiltration, proliferation, and superior functional persistence. In the irradiated tumor, we identify a subpopulation of CAR T cells with higher intensity of expression of CAR and co-expression of chemokine receptors that infiltrate metastatic tumors efficiently and promote CAR T-cell, anti-tumor efficacy.

Acknowledgements

We acknowledge excellent editorial assistance from Summer Koop and David B. Sewell of the Memorial Sloan Kettering Cancer Center Thoracic Surgery Service.

Financial Support:

D.R.J.’s laboratory work is supported by grants from the National Institutes of Health (R01 CA217169 and R01 CA240472). P.S.A.’s laboratory work is supported by grants from the National Institutes of Health (P30 CA008748, R01 CA236615-01, and R01 CA235667), the U.S. Department of Defense (BC132124, LC160212, CA170630, CA180889, and CA200437), the Batishwa Fellowship, the Comedy vs Cancer Award, the Dalle Pezze Foundation, the Derfner Foundation, the Esophageal Cancer Education Fund, the Geoffrey Beene Foundation, the Memorial Sloan Kettering Technology Development Fund, the Miner Fund for Mesothelioma Research, the Mr. William H. Goodwin and Alice Goodwin, the Commonwealth Foundation for Cancer Research, and the Experimental Therapeutics Center of Memorial Sloan Kettering Cancer Center. P.S.A.’s laboratory receives research support from Atara Biotherapeutics. M.S.S. is supported in part by grant T32CA009501.

Footnotes

Conflict of Interest Disclosure Statement: D.R.J. serves as a consultant for AstraZeneca and is on a Clinical Trial Steering Committee for Merck. P.S.A. declares research funding from ATARA Biotherapeutics; Scientific Advisory Board Member and Consultant for ATARA Biotherapeutics, Abound Bio, Bio4T2, Carisma Therapeutics, Imugene, ImmPactBio, Johnston & Johnston, Orion pharma, Link Immunotherapeutics, Outpace Bio, Pluri-Biotech, Verismo Therapeutics; Patents, royalties and intellectual property on mesothelin-targeted CAR and other T-cell therapies, which have been licensed to ATARA Biotherapeutics, issued patent method for detection of cancer cells using virus, and pending patent applications on PD-1 dominant negative receptor, wireless pulse-oximetry device, and on an ex vivo malignant pleural effusion culture system.

Memorial Sloan Kettering Cancer Center (MSK) has licensed intellectual property related to mesothelin-targeted CARs and T-cell therapies to ATARA Biotherapeutics, and has associated financial interests.

References

1.Blank CU, Haanen JB, Ribas A, Schumacher TN. CANCER IMMUNOLOGY. The “cancer immunogram”. Science 2016;352(6286):658–60. [DOI] [PubMed] [Google Scholar]
2.Wong PF, Wei W, Smithy JW, Acs B, Toki MI, Blenman KRM, et al. Multiplex quantitative analysis of tumor-infiltrating lymphocytes and immunotherapy outcome in metastatic melanoma. Clin Cancer Res 2019;25(8):2442–9. [DOI] [PMC free article] [PubMed] [Google Scholar]
3.Teng MW, Ngiow SF, Ribas A, Smyth MJ. Classifying cancers based on T-cell infiltration and PD-L1. Cancer Res 2015;75(11):2139–45. [DOI] [PMC free article] [PubMed] [Google Scholar]
4.June CH, Sadelain M. Chimeric antigen receptor therapy. N Engl J Med 2018;379(1):64–73. [DOI] [PMC free article] [PubMed] [Google Scholar]
5.Kiesgen S, Chicaybam L, Chintala NK, Adusumilli PS. Chimeric antigen receptor (CAR) T-cell therapy for thoracic malignancies. J Thorac Oncol 2018;13(1):16–26. [DOI] [PMC free article] [PubMed] [Google Scholar]
6.Clay TM, Custer MC, Sachs J, Hwu P, Rosenberg SA, Nishimura MI. Efficient transfer of a tumor antigen-reactive TCR to human peripheral blood lymphocytes confers anti-tumor reactivity. J Immunol 1999;163(1):507–13. [PubMed] [Google Scholar]
7.Majzner RG, Mackall CL. Clinical lessons learned from the first leg of the CAR T cell journey. Nat Med 2019;25(9):1341–55. [DOI] [PubMed] [Google Scholar]
8.Adusumilli PS, Cherkassky L, Villena-Vargas J, Colovos C, Servais E, Plotkin J, et al. Regional delivery of mesothelin-targeted CAR T cell therapy generates potent and long-lasting CD4-dependent tumor immunity. Sci Transl Med 2014;6(261):261ra151. [DOI] [PMC free article] [PubMed] [Google Scholar]
9.Mulazzani M, Frassle SP, von Mucke-Heim I, Langer S, Zhou X, Ishikawa-Ankerhold H, et al. Long-term in vivo microscopy of CAR T cell dynamics during eradication of CNS lymphoma in mice. Proc Natl Acad Sci U S A 2019;116(48):24275–84. [DOI] [PMC free article] [PubMed] [Google Scholar]
10.Galeano Nino JL, Pageon SV, Tay SS, Colakoglu F, Kempe D, Hywood J, et al. Cytotoxic T cells swarm by homotypic chemokine signalling. Elife 2020;9:e56554. [DOI] [PMC free article] [PubMed] [Google Scholar]
11.Kwon ED, Drake CG, Scher HI, Fizazi K, Bossi A, van den Eertwegh AJ, et al. Ipilimumab versus placebo after radiotherapy in patients with metastatic castration-resistant prostate cancer that had progressed after docetaxel chemotherapy (CA184-043): a multicentre, randomised, double-blind, phase 3 trial. Lancet Oncol 2014;15(7):700–12. [DOI] [PMC free article] [PubMed] [Google Scholar]
12.Theelen W, Peulen HMU, Lalezari F, van der Noort V, de Vries JF, Aerts J, et al. Effect of pembrolizumab after stereotactic body radiotherapy vs pembrolizumab alone on tumor response in patients with advanced non-small cell lung cancer: Results of the PEMBRO-RT phase 2 randomized clinical trial. JAMA Oncol 2019;5(9):1276–82. [DOI] [PMC free article] [PubMed] [Google Scholar]
13.McBride SM, Sherman EJ, Tsai CJ, Baxi SS, Aghalar J, Eng J, et al. A phase II randomized trial of nivolumab with stereotactic body radiotherapy (SBRT) versus nivolumab alone in metastatic (M1) head and neck squamous cell carcinoma (HNSCC). J Clin Oncol 2018;36(15_suppl):6009. [DOI] [PMC free article] [PubMed] [Google Scholar]
14.Viaud S, Ma JSY, Hardy IR, Hampton EN, Benish B, Sherwood L, et al. Switchable control over in vivo CAR T expansion, B cell depletion, and induction of memory. Proc Natl Acad Sci U S A 2018;115(46):E10898–E906. [DOI] [PMC free article] [PubMed] [Google Scholar]
15.Tantalo DG, Oliver AJ, von Scheidt B, Harrison AJ, Mueller SN, Kershaw MH, et al. Understanding T cell phenotype for the design of effective chimeric antigen receptor T cell therapies. J Immunother Cancer 2021;9(5):e002555. [DOI] [PMC free article] [PubMed] [Google Scholar]
16.Matsumura S, Wang B, Kawashima N, Braunstein S, Badura M, Cameron TO, et al. Radiation-induced CXCL16 release by breast cancer cells attracts effector T cells. J Immunol 2008;181(5):3099–107. [DOI] [PMC free article] [PubMed] [Google Scholar]
17.Colton M, Cheadle EJ, Honeychurch J, Illidge TM. Reprogramming the tumour microenvironment by radiotherapy: implications for radiotherapy and immunotherapy combinations. Radiat Oncol 2020;15(1):254. [DOI] [PMC free article] [PubMed] [Google Scholar]
18.Murakami J, Wu L, Kohno M, Chan ML, Zhao Y, Yun Z, et al. Triple-modality therapy maximizes antitumor immune responses in a mouse model of mesothelioma. Sci Transl Med 2021;13(589):eabd988. [DOI] [PubMed] [Google Scholar]
19.Vatner RE, Formenti SC. Myeloid-derived cells in tumors: effects of radiation. Semin Radiat Oncol 2015;25(1):18–27. [DOI] [PubMed] [Google Scholar]
20.Oweida A, Hararah MK, Phan A, Binder D, Bhatia S, Lennon S, et al. Resistance to radiotherapy and PD-L1 blockade is mediated by TIM-3 upregulation and regulatory T-cell infiltration. Clin Cancer Res 2018;24(21):5368–80. [DOI] [PMC free article] [PubMed] [Google Scholar]
21.Cook AM, McDonnell A, Millward MJ, Creaney J, Hasani A, McMullen M, et al. A phase 1b clinical trial optimizing regulatory T cell depletion in combination with platinum-based chemotherapy in thoracic cancers. Expert Rev Anticancer Ther 2021;21(5):465–74. [DOI] [PubMed] [Google Scholar]
22.Zhong H, Lai Y, Zhang R, Daoud A, Feng Q, Zhou J, et al. Low dose cyclophosphamide modulates tumor microenvironment by TGF-beta signaling pathway. Int J Mol Sci 2020;21(3):957. [DOI] [PMC free article] [PubMed] [Google Scholar]
23.Ghiringhelli F, Menard C, Puig PE, Ladoire S, Roux S, Martin F, et al. Metronomic cyclophosphamide regimen selectively depletes CD4+CD25+ regulatory T cells and restores T and NK effector functions in end stage cancer patients. Cancer Immunol Immunother 2007;56(5):641–8. [DOI] [PMC free article] [PubMed] [Google Scholar]
24.Noordam L, Kaijen MEH, Bezemer K, Cornelissen R, Maat L, Hoogsteden HC, et al. Low-dose cyclophosphamide depletes circulating naive and activated regulatory T cells in malignant pleural mesothelioma patients synergistically treated with dendritic cell-based immunotherapy. Oncoimmunology 2018;7(12):e1474318. [DOI] [PMC free article] [PubMed] [Google Scholar]
25.Murad JP, Tilakawardane D, Park AK, Lopez LS, Young CA, Gibson J, et al. Preconditioning modifies the TME to enhance solid tumor CAR T cell efficacy and endogenous protective immunity. Mol Ther 2021;29(7):2335–49. [DOI] [PMC free article] [PubMed] [Google Scholar]
26.Morello A, Sadelain M, Adusumilli PS. Mesothelin-targeted CARs: Driving T cells to solid tumors. Cancer Discov 2016;6(2):133–46. [DOI] [PMC free article] [PubMed] [Google Scholar]
27.Adusumilli PS, Zauderer MG, Rusch VW, O’Cearbhaill R, Zhu A, Ngai D, et al. Regional delivery of mesothelin-targeted CAR T cells for pleural cancers: Safety and preliminary efficacy in combination with anti-PD-1 agent. J Clin Oncol 2019;37(15_suppl):2511.31154918 [Google Scholar]
28.Haas AR, Tanyi JL, O’Hara MH, Gladney WL, Lacey SF, Torigian DA, et al. Phase I study of lentiviral-transduced chimeric antigen receptor-modified T cells recognizing mesothelin in advanced solid cancers. Mol Ther 2019;27(11):1919–29. [DOI] [PMC free article] [PubMed] [Google Scholar]
29.Adusumilli PS, Zauderer MG, Riviere I, Solomon SB, Rusch VW, O’Cearbhaill RE, et al. A phase I trial of regional mesothelin-targeted CAR T-cell therapy in patients with malignant pleural disease, in combination with the anti-PD-1 agent pembrolizumab. Cancer Discov 2021;11(11):2748–63. [DOI] [PMC free article] [PubMed] [Google Scholar]
30.Gomez DR, Rimner A, Simone CB 2nd, Cho BCJ, de Perrot M, Adjei AA, et al. The use of radiation therapy for the treatment of malignant pleural mesothelioma: Expert opinion from the National Cancer Institute Thoracic Malignancy Steering Committee, International Association for the Study of Lung Cancer, and Mesothelioma Applied Research Foundation. J Thorac Oncol 2019;14(7):1172–83. [DOI] [PubMed] [Google Scholar]
31.Cho BCJ, Donahoe L, Bradbury PA, Leighl N, Keshavjee S, Hope A, et al. Surgery for malignant pleural mesothelioma after radiotherapy (SMART): final results from a single-centre, phase 2 trial. Lancet Oncol 2021;22(2):190–7. [DOI] [PubMed] [Google Scholar]
32.Rimner A, Zauderer MG, Gomez DR, Adusumilli PS, Parhar PK, Wu AJ, et al. Phase II study of hemithoracic intensity-modulated pleural radiation therapy (IMPRINT) as part of lung-sparing multimodality therapy in patients with malignant pleural mesothelioma. J Clin Oncol 2016;34(23):2761–8. [DOI] [PMC free article] [PubMed] [Google Scholar]
33.Servais EL, Colovos C, Rodriguez L, Bograd AJ, Nitadori J, Sima C, et al. Mesothelin overexpression promotes mesothelioma cell invasion and MMP-9 secretion in an orthotopic mouse model and in epithelioid pleural mesothelioma patients. Clin Cancer Res 2012;18(9):2478–89. [DOI] [PMC free article] [PubMed] [Google Scholar]
34.Kachala SS, Bograd AJ, Villena-Vargas J, Suzuki K, Servais EL, Kadota K, et al. Mesothelin overexpression is a marker of tumor aggressiveness and is associated with reduced recurrence-free and overall survival in early-stage lung adenocarcinoma. Clin Cancer Res 2014;20(4):1020–8. [DOI] [PMC free article] [PubMed] [Google Scholar]
35.Hollyman D, Stefanski J, Przybylowski M, Bartido S, Borquez-Ojeda O, Taylor C, et al. Manufacturing validation of biologically functional T cells targeted to CD19 antigen for autologous adoptive cell therapy. J Immunother 2009;32(2):169–80. [DOI] [PMC free article] [PubMed] [Google Scholar]
36.Servais EL, Colovos C, Kachala SS, Adusumilli PS. Pre-clinical mouse models of primary and metastatic pleural cancers of the lung and breast and the use of bioluminescent imaging to monitor pleural tumor burden. Curr Protoc Pharmacol 2011;Chapter 14:Unit14 21. [DOI] [PMC free article] [PubMed] [Google Scholar]
37.Choi H, Sheng J, Gao D, Li F, Durrans A, Ryu S, et al. Transcriptome analysis of individual stromal cell populations identifies stroma-tumor crosstalk in mouse lung cancer model. Cell Rep 2015;10(7):1187–201. [DOI] [PubMed] [Google Scholar]
38.Feng Y, Xiao X, Zhu Z, Streaker E, Ho M, Pastan I, et al. A novel human monoclonal antibody that binds with high affinity to mesothelin-expressing cells and kills them by antibody-dependent cell-mediated cytotoxicity. Mol Cancer Ther 2009;8(5):1113–8. [DOI] [PMC free article] [PubMed] [Google Scholar]
39.Cherkassky L, Morello A, Villena-Vargas J, Feng Y, Dimitrov DS, Jones DR, et al. Human CAR T cells with cell-intrinsic PD-1 checkpoint blockade resist tumor-mediated inhibition. J Clin Invest 2016;126(8):3130–44. [DOI] [PMC free article] [PubMed] [Google Scholar]
40.Servais EL, Suzuki K, Colovos C, Rodriguez L, Sima C, Fleisher M, et al. An in vivo platform for tumor biomarker assessment. PLoS One 2011;6(10):e26722. [DOI] [PMC free article] [PubMed] [Google Scholar]
41.Chang ZL, Silver PA, Chen YY. Identification and selective expansion of functionally superior T cells expressing chimeric antigen receptors. J Transl Med 2015;13:161. [DOI] [PMC free article] [PubMed] [Google Scholar]
42.Feucht J, Sun J, Eyquem J, Ho YJ, Zhao Z, Leibold J, et al. Calibration of CAR activation potential directs alternative T cell fates and therapeutic potency. Nat Med 2019;25(1):82–8. [DOI] [PMC free article] [PubMed] [Google Scholar]
43.Kiesgen S, Linot C, Quach HT, Saini J, Bellis R, Banerjee S, et al. Abstract LB-378: Regional delivery of clinical-grade mesothelin-targeted CAR T cells with cell-intrinsic PD-1 checkpoint blockade: Translation to a phase I trial. Cancer Res 2020;80(16_Supplement):LB–378. [Google Scholar]
44.Ban Y, Markowitz GJ, Zou Y, Ramchandani D, Kraynak J, Sheng J, et al. Radiation-activated secretory proteins of Scgb1a1 (+) club cells increase the efficacy of immune checkpoint blockade in lung cancer. Nat Cancer 2021;2(9):919–31. [DOI] [PMC free article] [PubMed] [Google Scholar]
45.Rudqvist NP, Pilones KA, Lhuillier C, Wennerberg E, Sidhom JW, Emerson RO, et al. Radiotherapy and CTLA-4 blockade shape the TCR repertoire of tumor-infiltrating T cells. Cancer Immunol Res 2018;6(2):139–50. [DOI] [PMC free article] [PubMed] [Google Scholar]
46.Kantari-Mimoun C, Barrin S, Vimeux L, Haghiri S, Gervais C, Joaquina S, et al. CAR T-cell entry into tumor islets is a two-step process dependent on IFNgamma and ICAM-1. Cancer Immunol Res 2021;9(12):1425–38. [DOI] [PubMed] [Google Scholar]
47.Jin L, Tao H, Karachi A, Long Y, Hou AY, Na M, et al. CXCR1- or CXCR2-modified CAR T cells co-opt IL-8 for maximal antitumor efficacy in solid tumors. Nature Commun 2019;10(1):4016. [DOI] [PMC free article] [PubMed] [Google Scholar]
48.Moon EK, Carpenito C, Sun J, Wang LC, Kapoor V, Predina J, et al. Expression of a functional CCR2 receptor enhances tumor localization and tumor eradication by retargeted human T cells expressing a mesothelin-specific chimeric antibody receptor. Clin Cancer Res 2011;17(14):4719–30. [DOI] [PMC free article] [PubMed] [Google Scholar]
49.Lesch S, Blumenberg V, Stoiber S, Gottschlich A, Ogonek J, Cadilha BL, et al. T cells armed with C-X-C chemokine receptor type 6 enhance adoptive cell therapy for pancreatic tumours. Nat Biomed Eng 2021;5(11):1246–60. [DOI] [PMC free article] [PubMed] [Google Scholar]
50.Cherkassky L, Adusumilli PS. Arming T cells to infiltrate pancreatic tumours. Nat Biomed Eng 2021;5(11):1243–5. [DOI] [PubMed] [Google Scholar]
51.Rabin RL, Park MK, Liao F, Swofford R, Stephany D, Farber JM. Chemokine receptor responses on T cells are achieved through regulation of both receptor expression and signaling. J Immunol 1999;162(7):3840–50. [PubMed] [Google Scholar]
52.Martin-Cofreces NB, Vicente-Manzanares M, Sanchez-Madrid F. Adhesive interactions delineate the topography of the immune synapse. Front Cell Dev Biol 2018;6:149. [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

NIHMS1924120-supplement-1.pdf^{(6.2MB, pdf)}

NIHMS1924120-supplement-2.pdf^{(90.1KB, pdf)}

Data Availability Statement

The data generated in this study are available within the article and its supplementary data files. The gene expression data is available at GSE157881 (Sub-series part of super-series GSE157883)

[R1] 1.Blank CU, Haanen JB, Ribas A, Schumacher TN. CANCER IMMUNOLOGY. The “cancer immunogram”. Science 2016;352(6286):658–60. [DOI] [PubMed] [Google Scholar]

[R2] 2.Wong PF, Wei W, Smithy JW, Acs B, Toki MI, Blenman KRM, et al. Multiplex quantitative analysis of tumor-infiltrating lymphocytes and immunotherapy outcome in metastatic melanoma. Clin Cancer Res 2019;25(8):2442–9. [DOI] [PMC free article] [PubMed] [Google Scholar]

[R3] 3.Teng MW, Ngiow SF, Ribas A, Smyth MJ. Classifying cancers based on T-cell infiltration and PD-L1. Cancer Res 2015;75(11):2139–45. [DOI] [PMC free article] [PubMed] [Google Scholar]

[R4] 4.June CH, Sadelain M. Chimeric antigen receptor therapy. N Engl J Med 2018;379(1):64–73. [DOI] [PMC free article] [PubMed] [Google Scholar]

[R5] 5.Kiesgen S, Chicaybam L, Chintala NK, Adusumilli PS. Chimeric antigen receptor (CAR) T-cell therapy for thoracic malignancies. J Thorac Oncol 2018;13(1):16–26. [DOI] [PMC free article] [PubMed] [Google Scholar]

[R6] 6.Clay TM, Custer MC, Sachs J, Hwu P, Rosenberg SA, Nishimura MI. Efficient transfer of a tumor antigen-reactive TCR to human peripheral blood lymphocytes confers anti-tumor reactivity. J Immunol 1999;163(1):507–13. [PubMed] [Google Scholar]

[R7] 7.Majzner RG, Mackall CL. Clinical lessons learned from the first leg of the CAR T cell journey. Nat Med 2019;25(9):1341–55. [DOI] [PubMed] [Google Scholar]

[R8] 8.Adusumilli PS, Cherkassky L, Villena-Vargas J, Colovos C, Servais E, Plotkin J, et al. Regional delivery of mesothelin-targeted CAR T cell therapy generates potent and long-lasting CD4-dependent tumor immunity. Sci Transl Med 2014;6(261):261ra151. [DOI] [PMC free article] [PubMed] [Google Scholar]

[R9] 9.Mulazzani M, Frassle SP, von Mucke-Heim I, Langer S, Zhou X, Ishikawa-Ankerhold H, et al. Long-term in vivo microscopy of CAR T cell dynamics during eradication of CNS lymphoma in mice. Proc Natl Acad Sci U S A 2019;116(48):24275–84. [DOI] [PMC free article] [PubMed] [Google Scholar]

[R10] 10.Galeano Nino JL, Pageon SV, Tay SS, Colakoglu F, Kempe D, Hywood J, et al. Cytotoxic T cells swarm by homotypic chemokine signalling. Elife 2020;9:e56554. [DOI] [PMC free article] [PubMed] [Google Scholar]

[R11] 11.Kwon ED, Drake CG, Scher HI, Fizazi K, Bossi A, van den Eertwegh AJ, et al. Ipilimumab versus placebo after radiotherapy in patients with metastatic castration-resistant prostate cancer that had progressed after docetaxel chemotherapy (CA184-043): a multicentre, randomised, double-blind, phase 3 trial. Lancet Oncol 2014;15(7):700–12. [DOI] [PMC free article] [PubMed] [Google Scholar]

[R12] 12.Theelen W, Peulen HMU, Lalezari F, van der Noort V, de Vries JF, Aerts J, et al. Effect of pembrolizumab after stereotactic body radiotherapy vs pembrolizumab alone on tumor response in patients with advanced non-small cell lung cancer: Results of the PEMBRO-RT phase 2 randomized clinical trial. JAMA Oncol 2019;5(9):1276–82. [DOI] [PMC free article] [PubMed] [Google Scholar]

[R13] 13.McBride SM, Sherman EJ, Tsai CJ, Baxi SS, Aghalar J, Eng J, et al. A phase II randomized trial of nivolumab with stereotactic body radiotherapy (SBRT) versus nivolumab alone in metastatic (M1) head and neck squamous cell carcinoma (HNSCC). J Clin Oncol 2018;36(15_suppl):6009. [DOI] [PMC free article] [PubMed] [Google Scholar]

[R14] 14.Viaud S, Ma JSY, Hardy IR, Hampton EN, Benish B, Sherwood L, et al. Switchable control over in vivo CAR T expansion, B cell depletion, and induction of memory. Proc Natl Acad Sci U S A 2018;115(46):E10898–E906. [DOI] [PMC free article] [PubMed] [Google Scholar]

[R15] 15.Tantalo DG, Oliver AJ, von Scheidt B, Harrison AJ, Mueller SN, Kershaw MH, et al. Understanding T cell phenotype for the design of effective chimeric antigen receptor T cell therapies. J Immunother Cancer 2021;9(5):e002555. [DOI] [PMC free article] [PubMed] [Google Scholar]

[R16] 16.Matsumura S, Wang B, Kawashima N, Braunstein S, Badura M, Cameron TO, et al. Radiation-induced CXCL16 release by breast cancer cells attracts effector T cells. J Immunol 2008;181(5):3099–107. [DOI] [PMC free article] [PubMed] [Google Scholar]

[R17] 17.Colton M, Cheadle EJ, Honeychurch J, Illidge TM. Reprogramming the tumour microenvironment by radiotherapy: implications for radiotherapy and immunotherapy combinations. Radiat Oncol 2020;15(1):254. [DOI] [PMC free article] [PubMed] [Google Scholar]

[R18] 18.Murakami J, Wu L, Kohno M, Chan ML, Zhao Y, Yun Z, et al. Triple-modality therapy maximizes antitumor immune responses in a mouse model of mesothelioma. Sci Transl Med 2021;13(589):eabd988. [DOI] [PubMed] [Google Scholar]

[R19] 19.Vatner RE, Formenti SC. Myeloid-derived cells in tumors: effects of radiation. Semin Radiat Oncol 2015;25(1):18–27. [DOI] [PubMed] [Google Scholar]

[R20] 20.Oweida A, Hararah MK, Phan A, Binder D, Bhatia S, Lennon S, et al. Resistance to radiotherapy and PD-L1 blockade is mediated by TIM-3 upregulation and regulatory T-cell infiltration. Clin Cancer Res 2018;24(21):5368–80. [DOI] [PMC free article] [PubMed] [Google Scholar]

[R21] 21.Cook AM, McDonnell A, Millward MJ, Creaney J, Hasani A, McMullen M, et al. A phase 1b clinical trial optimizing regulatory T cell depletion in combination with platinum-based chemotherapy in thoracic cancers. Expert Rev Anticancer Ther 2021;21(5):465–74. [DOI] [PubMed] [Google Scholar]

[R22] 22.Zhong H, Lai Y, Zhang R, Daoud A, Feng Q, Zhou J, et al. Low dose cyclophosphamide modulates tumor microenvironment by TGF-beta signaling pathway. Int J Mol Sci 2020;21(3):957. [DOI] [PMC free article] [PubMed] [Google Scholar]

[R23] 23.Ghiringhelli F, Menard C, Puig PE, Ladoire S, Roux S, Martin F, et al. Metronomic cyclophosphamide regimen selectively depletes CD4+CD25+ regulatory T cells and restores T and NK effector functions in end stage cancer patients. Cancer Immunol Immunother 2007;56(5):641–8. [DOI] [PMC free article] [PubMed] [Google Scholar]

[R24] 24.Noordam L, Kaijen MEH, Bezemer K, Cornelissen R, Maat L, Hoogsteden HC, et al. Low-dose cyclophosphamide depletes circulating naive and activated regulatory T cells in malignant pleural mesothelioma patients synergistically treated with dendritic cell-based immunotherapy. Oncoimmunology 2018;7(12):e1474318. [DOI] [PMC free article] [PubMed] [Google Scholar]

[R25] 25.Murad JP, Tilakawardane D, Park AK, Lopez LS, Young CA, Gibson J, et al. Preconditioning modifies the TME to enhance solid tumor CAR T cell efficacy and endogenous protective immunity. Mol Ther 2021;29(7):2335–49. [DOI] [PMC free article] [PubMed] [Google Scholar]

[R26] 26.Morello A, Sadelain M, Adusumilli PS. Mesothelin-targeted CARs: Driving T cells to solid tumors. Cancer Discov 2016;6(2):133–46. [DOI] [PMC free article] [PubMed] [Google Scholar]

[R27] 27.Adusumilli PS, Zauderer MG, Rusch VW, O’Cearbhaill R, Zhu A, Ngai D, et al. Regional delivery of mesothelin-targeted CAR T cells for pleural cancers: Safety and preliminary efficacy in combination with anti-PD-1 agent. J Clin Oncol 2019;37(15_suppl):2511.31154918 [Google Scholar]

[R28] 28.Haas AR, Tanyi JL, O’Hara MH, Gladney WL, Lacey SF, Torigian DA, et al. Phase I study of lentiviral-transduced chimeric antigen receptor-modified T cells recognizing mesothelin in advanced solid cancers. Mol Ther 2019;27(11):1919–29. [DOI] [PMC free article] [PubMed] [Google Scholar]

[R29] 29.Adusumilli PS, Zauderer MG, Riviere I, Solomon SB, Rusch VW, O’Cearbhaill RE, et al. A phase I trial of regional mesothelin-targeted CAR T-cell therapy in patients with malignant pleural disease, in combination with the anti-PD-1 agent pembrolizumab. Cancer Discov 2021;11(11):2748–63. [DOI] [PMC free article] [PubMed] [Google Scholar]

[R30] 30.Gomez DR, Rimner A, Simone CB 2nd, Cho BCJ, de Perrot M, Adjei AA, et al. The use of radiation therapy for the treatment of malignant pleural mesothelioma: Expert opinion from the National Cancer Institute Thoracic Malignancy Steering Committee, International Association for the Study of Lung Cancer, and Mesothelioma Applied Research Foundation. J Thorac Oncol 2019;14(7):1172–83. [DOI] [PubMed] [Google Scholar]

[R31] 31.Cho BCJ, Donahoe L, Bradbury PA, Leighl N, Keshavjee S, Hope A, et al. Surgery for malignant pleural mesothelioma after radiotherapy (SMART): final results from a single-centre, phase 2 trial. Lancet Oncol 2021;22(2):190–7. [DOI] [PubMed] [Google Scholar]

[R32] 32.Rimner A, Zauderer MG, Gomez DR, Adusumilli PS, Parhar PK, Wu AJ, et al. Phase II study of hemithoracic intensity-modulated pleural radiation therapy (IMPRINT) as part of lung-sparing multimodality therapy in patients with malignant pleural mesothelioma. J Clin Oncol 2016;34(23):2761–8. [DOI] [PMC free article] [PubMed] [Google Scholar]

[R33] 33.Servais EL, Colovos C, Rodriguez L, Bograd AJ, Nitadori J, Sima C, et al. Mesothelin overexpression promotes mesothelioma cell invasion and MMP-9 secretion in an orthotopic mouse model and in epithelioid pleural mesothelioma patients. Clin Cancer Res 2012;18(9):2478–89. [DOI] [PMC free article] [PubMed] [Google Scholar]

[R34] 34.Kachala SS, Bograd AJ, Villena-Vargas J, Suzuki K, Servais EL, Kadota K, et al. Mesothelin overexpression is a marker of tumor aggressiveness and is associated with reduced recurrence-free and overall survival in early-stage lung adenocarcinoma. Clin Cancer Res 2014;20(4):1020–8. [DOI] [PMC free article] [PubMed] [Google Scholar]

[R35] 35.Hollyman D, Stefanski J, Przybylowski M, Bartido S, Borquez-Ojeda O, Taylor C, et al. Manufacturing validation of biologically functional T cells targeted to CD19 antigen for autologous adoptive cell therapy. J Immunother 2009;32(2):169–80. [DOI] [PMC free article] [PubMed] [Google Scholar]

[R36] 36.Servais EL, Colovos C, Kachala SS, Adusumilli PS. Pre-clinical mouse models of primary and metastatic pleural cancers of the lung and breast and the use of bioluminescent imaging to monitor pleural tumor burden. Curr Protoc Pharmacol 2011;Chapter 14:Unit14 21. [DOI] [PMC free article] [PubMed] [Google Scholar]

[R37] 37.Choi H, Sheng J, Gao D, Li F, Durrans A, Ryu S, et al. Transcriptome analysis of individual stromal cell populations identifies stroma-tumor crosstalk in mouse lung cancer model. Cell Rep 2015;10(7):1187–201. [DOI] [PubMed] [Google Scholar]

[R38] 38.Feng Y, Xiao X, Zhu Z, Streaker E, Ho M, Pastan I, et al. A novel human monoclonal antibody that binds with high affinity to mesothelin-expressing cells and kills them by antibody-dependent cell-mediated cytotoxicity. Mol Cancer Ther 2009;8(5):1113–8. [DOI] [PMC free article] [PubMed] [Google Scholar]

[R39] 39.Cherkassky L, Morello A, Villena-Vargas J, Feng Y, Dimitrov DS, Jones DR, et al. Human CAR T cells with cell-intrinsic PD-1 checkpoint blockade resist tumor-mediated inhibition. J Clin Invest 2016;126(8):3130–44. [DOI] [PMC free article] [PubMed] [Google Scholar]

[R40] 40.Servais EL, Suzuki K, Colovos C, Rodriguez L, Sima C, Fleisher M, et al. An in vivo platform for tumor biomarker assessment. PLoS One 2011;6(10):e26722. [DOI] [PMC free article] [PubMed] [Google Scholar]

[R41] 41.Chang ZL, Silver PA, Chen YY. Identification and selective expansion of functionally superior T cells expressing chimeric antigen receptors. J Transl Med 2015;13:161. [DOI] [PMC free article] [PubMed] [Google Scholar]

[R42] 42.Feucht J, Sun J, Eyquem J, Ho YJ, Zhao Z, Leibold J, et al. Calibration of CAR activation potential directs alternative T cell fates and therapeutic potency. Nat Med 2019;25(1):82–8. [DOI] [PMC free article] [PubMed] [Google Scholar]

[R43] 43.Kiesgen S, Linot C, Quach HT, Saini J, Bellis R, Banerjee S, et al. Abstract LB-378: Regional delivery of clinical-grade mesothelin-targeted CAR T cells with cell-intrinsic PD-1 checkpoint blockade: Translation to a phase I trial. Cancer Res 2020;80(16_Supplement):LB–378. [Google Scholar]

[R44] 44.Ban Y, Markowitz GJ, Zou Y, Ramchandani D, Kraynak J, Sheng J, et al. Radiation-activated secretory proteins of Scgb1a1 (+) club cells increase the efficacy of immune checkpoint blockade in lung cancer. Nat Cancer 2021;2(9):919–31. [DOI] [PMC free article] [PubMed] [Google Scholar]

[R45] 45.Rudqvist NP, Pilones KA, Lhuillier C, Wennerberg E, Sidhom JW, Emerson RO, et al. Radiotherapy and CTLA-4 blockade shape the TCR repertoire of tumor-infiltrating T cells. Cancer Immunol Res 2018;6(2):139–50. [DOI] [PMC free article] [PubMed] [Google Scholar]

[R46] 46.Kantari-Mimoun C, Barrin S, Vimeux L, Haghiri S, Gervais C, Joaquina S, et al. CAR T-cell entry into tumor islets is a two-step process dependent on IFNgamma and ICAM-1. Cancer Immunol Res 2021;9(12):1425–38. [DOI] [PubMed] [Google Scholar]

[R47] 47.Jin L, Tao H, Karachi A, Long Y, Hou AY, Na M, et al. CXCR1- or CXCR2-modified CAR T cells co-opt IL-8 for maximal antitumor efficacy in solid tumors. Nature Commun 2019;10(1):4016. [DOI] [PMC free article] [PubMed] [Google Scholar]

[R48] 48.Moon EK, Carpenito C, Sun J, Wang LC, Kapoor V, Predina J, et al. Expression of a functional CCR2 receptor enhances tumor localization and tumor eradication by retargeted human T cells expressing a mesothelin-specific chimeric antibody receptor. Clin Cancer Res 2011;17(14):4719–30. [DOI] [PMC free article] [PubMed] [Google Scholar]

[R49] 49.Lesch S, Blumenberg V, Stoiber S, Gottschlich A, Ogonek J, Cadilha BL, et al. T cells armed with C-X-C chemokine receptor type 6 enhance adoptive cell therapy for pancreatic tumours. Nat Biomed Eng 2021;5(11):1246–60. [DOI] [PMC free article] [PubMed] [Google Scholar]

[R50] 50.Cherkassky L, Adusumilli PS. Arming T cells to infiltrate pancreatic tumours. Nat Biomed Eng 2021;5(11):1243–5. [DOI] [PubMed] [Google Scholar]

[R51] 51.Rabin RL, Park MK, Liao F, Swofford R, Stephany D, Farber JM. Chemokine receptor responses on T cells are achieved through regulation of both receptor expression and signaling. J Immunol 1999;162(7):3840–50. [PubMed] [Google Scholar]

[R52] 52.Martin-Cofreces NB, Vicente-Manzanares M, Sanchez-Madrid F. Adhesive interactions delineate the topography of the immune synapse. Front Cell Dev Biol 2018;6:149. [DOI] [PMC free article] [PubMed] [Google Scholar]

PERMALINK

Tumor-targeted non-ablative radiation promotes solid tumor CAR T-cell therapy efficacy

Hue Tu Quach

Matthew S Skovgard

Jonathan Villena-Vargas

Rebecca Y Bellis

Navin K Chintala

Alfredo Amador-Molina

Yang Bai

Srijita Banerjee

Jasmeen Saini

Yuquan Xiong

William-Ray Vista

Alexander J Byun

Andreas De Biasi

Masha Zeltsman

Marissa Mayor

Aurore Morello

Vivek Mittal

Daniel R Gomez

Andreas Rimner

David R Jones

Prasad S Adusumilli

Abstract

BACKGROUND

MATERIALS AND METHODS

Experimental Design

Tumor Cell Lines

Gammaretroviral Vector Construction and Viral Production

Fig. 1. Tumor-targeted radiation prior to chimeric antigen receptor (CAR) T-cell administration results in superior antitumor efficacy.

T-Cell Isolation and Gene Transfer

CAR T-Cell Cytotoxicity Assays

In Vitro Chemokine and Cytokine Detection Assays

In Vitro CAR T-Cell Accumulation Assay with and without Radiation-Conditioned Media

Boyden Chamber CAR T-Cell Migration Experiments

In Vitro CAR T-Cell Proliferation Assays

In Vitro Chemokine Receptor Profiles

Orthotopic MPM Mouse Model and Adoptive T-Cell Therapy

Intra-pulmonary NSCLC Mouse Model and Systemic T-Cell Therapy

Orthotopic MPM Mice with Flank Tumor Model and CAR T-Cell Therapy

Quantitative Tumor and T-Cell BLI

In Vivo Serum Chemokine and Cytokine Detection Assay

Harvesting and Preparation of CAR T Cells in Tumor, Spleen, and the Peritoneal Cavity

Flow Cytometry and T-Cell Proliferation Labeling

In Vivo Gene Expression Analysis

HKP1 Cancer Models and Irradiation Treatment

scRNA-seq Analysis

Statistical Analysis

Data Availability

RESULTS

Non-ablative Tumor-Targeted RT Promotes the Antitumor Efficacy of CAR T Cells

Tumor-Targeted RT Enhances Accumulation of CAR T Cells in the Tumor

Fig. 2. Tumor-targeted radiation facilitates earlier and higher accumulation of chimeric antigen receptor (CAR) T cells.

RT Pretreatment Enhances Chemotaxis, Tumor Infiltration, Proliferation, Memory, and Persistence of CAR T Cells

Fig. 3. Preconditioning radiation promotes migration, infiltration, proliferation, memory, and persistence of chimeric antigen receptor (CAR) T cells.

RT Promotes Tumor Secretion of Chemokines

Fig. 4. Radiation promotes tumor secretion of chemokines, and chimeric antigen receptor (CAR) T cells infiltrating irradiated tumor upregulate chemokine receptors.

Upregulated Chemokines in Irradiated Tumor Promote Corresponding Expression of Chemokine Receptors in Tumor-Infiltrating CAR T Cells

Tumor-infiltrating CAR T Cells Following Tumor-targeted Radiation Promote Efficacy Against Distant Tumors

Fig. 5. Infiltrating chimeric antigen receptor (CAR) T cells from irradiated tumor show superior abscopal antitumor efficacy and tumor immunity after repeated tumor rechallenge using the tumor treadmill test.

Irradiated Tumor-infiltrating CAR T Cells Promote Functional Persistence

Radiation Potentiates CAR T-Cell Efficacy in Mixed -Antigen-Expressing NSCLC

Fig. 6. Preconditioning radiation enhances the efficacy of chimeric antigen receptor (CAR) T cells in non-small cell lung cancer (NSCLC).

Upregulated Chemokines in Irradiated Tumors and Corresponding Chemokine Receptor Expression in Tumor-infiltrating T Cells in a Syngeneic Lung Cancer Model

Fig. 7. Tumor-targeted radiation promotes chemokine expression in tumor, and irradiated tumor-infiltrating T cells upregulate chemokine receptors in a syngeneic model of lung cancer.

DISCUSSION

Supplementary Material

Synopsis.

Acknowledgements

Financial Support:

Footnotes

References

Associated Data

Supplementary Materials

Data Availability Statement

ACTIONS

PERMALINK

RESOURCES

Similar articles

Cited by other articles