Chimeric antigen receptor T cells as adjuvant therapy for unresectable adenocarcinoma

Ugur Uslu; Tong Da; Charles-Antoine Assenmacher; John Scholler; Regina M Young; Julia Tchou; Carl H June

doi:10.1126/sciadv.ade2526

. 2023 Jan 11;9(2):eade2526. doi: 10.1126/sciadv.ade2526

Chimeric antigen receptor T cells as adjuvant therapy for unresectable adenocarcinoma

Ugur Uslu ^1,², Tong Da ^1,², Charles-Antoine Assenmacher ³, John Scholler ^1,², Regina M Young ^1,², Julia Tchou ^2,⁴, Carl H June ^1,^2,^*

PMCID: PMC9833675 PMID: 36630514

Abstract

Incomplete surgery of solid tumors is a risk factor for primary treatment failure. Here, we have investigated whether chimeric antigen receptor T cells (CARTs) could be used as an adjuvant therapy to clear residual cancer cells. We tested the feasibility of this approach in two partial resection xenograft models using mesothelin-specific CARTs. In addition, we developed a previously unexplored in vivo toxicity model to evaluate safety and effects on wound healing in immunocompetent C57BL/6 mice. We found that the local delivery of CARTs in a fibrin glue–based carrier was effective in clearing residual cancer cells following incomplete surgery. This resulted in significantly longer overall survival when compared to mice treated with surgery and CARTs without fibrin glue. On-target off-tumor toxicity was diminished, and wound healing complications were not seen in any of the mice. On the basis of these observations, a clinical trial in patients with locally advanced breast cancer is planned.

CAR T cells can be effectively and safely used as an adjuvant therapy in solid tumors that cannot be completely excised.

INTRODUCTION

Despite recent advances in cancer therapy, pancreatic ductal adenocarcinoma (PDA) and triple-negative breast cancer (TNBC) still represent tumor entities with limited therapeutic options and poor prognosis (1, 2). Surgical excision of the primary tumor is performed if feasible. However, the risk of positive surgical margins (i.e., residual tumor cells at the resection margin) remains high, especially in locally advanced tumors (3, 4), resulting in an increased risk for local recurrences and inferior clinical outcome (4). For those patients, treatment options to clear residual tumor cells following incomplete surgical excision that could be administered promptly and safely in an intraoperative setting would be beneficial. Approaches that could simultaneously clear positive margins without raising safety concerns and interfering with wound healing would be ideal. Here, we have tested the hypothesis that the local, intraoperative use of chimeric antigen receptor (CAR) T cell therapy might be an effective adjuvant therapy based on their emergence as an effective systemic immunotherapy for several hematological malignancies (5).

Our recent clinical data indicate the feasibility of direct intratumoral injection of CAR T cells (6) and that of another group demonstrate intracranial infusion into the resected tumor cavity (7). Appropriate trafficking of CAR T cells is required to ensure that all residual tumor cells can be cleared by CAR T cells. In addition, retention of CAR T cells at the surgery site may increase safety by decreasing systemic toxicity. For this purpose, we have optimized the use of tissue adhesives as a CAR T cell carrier, which could be applied intraoperatively on the wound surface and have evaluated fibrin glue, a biologic tissue adhesive that was found to be an effective sealant and topical hemostatic agent (8). A fibrin clot is produced by mixing two solutions containing fibrinogen in the presence of calcium and thrombin. This material is commercially available [Tisseel (fibrin sealant), Baxter] and has been approved for use in surgery (8). In addition, several preclinical and clinical reports have reported the potential use of fibrin glue–based solutions as a cell delivery technique for tissue regeneration (9–12). The fibrin matrix also supports the survival of T cells (13) and was recently shown to increase the efficacy of CAR T cells after intracranial administration for glioblastoma (14). In this study, a small window in the skull was made using a bone drill for CAR T cell inoculation, which was then closed using a tissue adhesive (14).

To develop a therapy for incompletely resected adenocarcinoma, we have established tumor models by incomplete surgical excision of subcutaneous tumors and the intraoperative application of mesothelin-specific CAR T cells in a fibrin glue–based carrier (fibrin gel) on the residual tumor and wound surface without the need for intratumoral injection. In two xenograft models (PDA and TNBC), the delivery of CAR T cells in fibrin gel applied into the resected tumor cavity cleared residual cancer cells following incomplete surgery of subcutaneous tumors and resulted in improved overall survival when compared to mice treated via direct intracavitary CAR T cell injection without fibrin gel. In a novel in vivo toxicity model using human mesothelin knock-in mice, on-target off-tumor toxicity was diminished. In addition, wound healing complications were not observed in any of the immunocompromised or immunocompetent mice.

RESULTS

CAR T cell antitumor efficacy depends on the fibrin gel dose

As a first step, we optimized the ideal fibrin gel formulation for CAR T cell delivery. This is a dose that retains CAR T cells in the resected tumor cavity but allows CAR T cell migration out of the gel. To find the ideal fibrin gel formulation, human CAR T cells in different concentrations of the human sealer protein solution containing fibrinogen and the human thrombin solution were analyzed in vitro and in vivo.

The in vitro passive transport rate of T cells through the fibrin gel was tested in a transwell assay. Green fluorescent protein (GFP)–labeled T cells (fig. S1A) were mixed in fibrin gel at different concentrations of both glue components and were added into the upper transwell compartment. Twenty-four hours were allowed for T cells to pass from the upper to the lower transwell compartment. The amount of GFP-positive T cells in the lower compartment was then counted using the Incucyte live-cell analysis instrument. T cells in fibrin gel using a fibrinogen solution of 10 mg/ml showed a significantly lower T cell passive transport rate through the membrane (fig. S1B), so that fibrin gel using fibrinogen solution of 5 mg/ml was used for further experiments. When T cell migration at different concentrations of the thrombin solution was analyzed, lower amounts of thrombin resulted in improved in vitro T cell passage rates (fig. S1C). Viability analysis of passaged cells in the lower transwell compartment showed no significant differences throughout all cohorts (fig. S1, B and C).

We next performed in vivo titration of different thrombin concentrations in a xenograft model. Human TNBC cells MDA-MB-231 were injected subcutaneously into the right flank of NOD/scid/IL2rγ^−/− (NSG) mice (fig. S2A). When the tumor size reached approximately 1.0 cm in diameter (tumor volume, ~250 to 300 mm³) (fig. S2B), incomplete surgical excision of the subcutaneous tumor was performed, and mice were treated either with direct intracavity inoculation of mesothelin-specific CAR T cells in medium (CAR Ts + medium) or with CAR T cells in fibrin gel using the indicated thrombin concentrations. Tumor recurrence following incomplete surgery could not be controlled in mice treated with CAR Ts + medium, as well as with CAR T cells in fibrin gel using the lowest and highest thrombin concentrations of 0.3 and 10.0 U/ml, respectively (fig. S2C). In contrast, clearance of residual tumor cells was observed in mice receiving CAR T cells in fibrin gel using intermediate thrombin concentrations (fig. S2C). Faster and more effective clearance of residual tumor cells was seen with thrombin concentration of 3.0 U/ml (fig. S2C). Considering the in vitro and in vivo data, CAR T cells in fibrin gel using fibrinogen solution of 5 mg/ml and thrombin solution of 3.0 U/ml showed the most robust antitumor efficacy, and these conditions were used for subsequent experiments.

Fibrin gel alone does not affect tumor growth and overall survival

To analyze whether the fibrin gel itself (without CAR T cells) affects tumor growth and overall survival of treated mice, NSG mice were inoculated with the human TNBC cell line MDA-MB-231 subcutaneously into the right flank (fig. S2D). Twenty-two days later, mice received either surgery only or surgery followed by direct intracavity inoculation of fibrin gel (fig. S2D). In both cohorts, a decrease followed by a rapid recurrence of tumor volume after treatment was seen (fig. S2E). No differences in overall survival were seen when both groups were compared (fig. S2F). In summary, these data confirmed that there is no fibrin gel–specific impact on tumor growth and overall survival of treated mice.

Local CAR T cells in fibrin gel clear residual tumor cells in a TNBC xenograft model

After determining the ideal fibrin gel composition and ruling out that fibrin gel itself affects tumor growth, we first tested the feasibility of this approach in a xenograft model using the previously described human TNBC cell line MDA-MB-231 (fig. S3). T cells were transduced to express a human mesothelin-specific CAR (CARM5) or a human CD19-specific CAR (CAR19) for control. Human T cells from two donors were used for xenograft studies (ND517 and ND569) to confirm that the observed effects are independent of the T cell donor. Cell surface staining of the transduced T cells of both donors revealed expression of the CAR constructs in approximately 50 to 63% of cells (fig. S4). In vitro, CARM5 T cells from both donors were functional and secreted the cytokines interleukin-2 (IL-2), tumor necrosis factor (TNF), and interferon-γ (IFN-γ) when coincubated with MDA-MB-231 cells (Fig. 1A). In addition, CARM5 T cells from both donors exhibited specific lysis of the tumor cells (Fig. 1B). CARM5 T cells and CAR19 T cells of both donors also showed specific cytokine secretion and cytotoxicity when coincubated with the human leukemia cell line K562 transduced with either human mesothelin or human CD19, respectively (fig. S5).

When used in xenograft models, MDA-MB-231 cells were injected subcutaneously into the right flank of NSG mice (Fig. 1C). Twenty-two days later (= day 0), mice were treated either with incomplete surgery only (surgery), with incomplete surgery followed by direct intracavity inoculation of CAR19 T cells in fibrin gel for control [surgery + CAR19 (fibrin gel)], with incomplete surgery followed by direct intracavity inoculation of CARM5 T cells in medium [surgery + CARM5 (medium)], or with incomplete surgery followed by direct intracavity inoculation of CARM5 T cells in fibrin gel [surgery + CARM5 (fibrin gel)] (Fig. 1C). In both T cell donors, tumor regrowth following incomplete surgery occurred in mice receiving surgery only or surgery + CAR19 (fibrin gel) (Fig. 1D). Three of the 10 mice treated with surgery + CARM5 (medium) showed clearance of residual tumor, while the majority of mice in this cohort (7 of 10) showed tumor progression (Fig. 1D). In contrast, tumor clearance was observed in all mice treated with surgery + CARM5 (fibrin gel) (Fig. 1D). This resulted in significantly longer overall survival of mice receiving surgery + CARM5 (fibrin gel) when compared to mice with surgery + CAR19 (fibrin gel), as well as to mice with surgery + CARM5 (medium) (Fig. 1E). At day 14, no statistical differences were detected in the levels of CD45⁺ cells in peripheral blood between cohorts (Fig. 1F). Transient and most likely surgery-related weight loss occurred in some mice directly following treatment throughout all cohorts, with subsequent resumption of growth (Fig. 1G). Clinically, none of the mice showed postsurgical signs of inflammation or wound healing complications (e.g., no redness, swelling, or dehiscence of surgical clips) at the surgery site (fig. S6).

Residual tumor is also cleared by local CAR T cells in a PDA xenograft model

To confirm that the observed effects are independent of the tumor entity, a second partial resection xenograft model was established using the human PDA cell line AsPC-1 (fig. S3). The same T cell donors (ND517 and ND569) were used for this model. In vitro, CARM5 T cells of both donors were functional and secreted the cytokines IL-2, TNF, and IFN-γ when coincubated with AsPC-1 (Fig. 2A). In addition, AsPC-1 cells were specifically eradicated by CARM5 T cells in vitro (Fig. 2B).

In xenograft models, AsPC-1 cells were injected subcutaneously into the right flank of NSG mice (Fig. 2C). Twenty-two days later (= day 0), mice received the same treatments at the same doses as previously described for the TNBC xenograft model (Fig. 2C). Similar to the TNBC model, tumor regrowth occurred in mice with PDA following incomplete surgery in mice receiving surgery only or surgery + CAR19 (fibrin gel) (Fig. 2D). One of the 10 mice treated with surgery + CARM5 (medium) showed clearance of residual tumor, while the other mice in this cohort showed tumor progression and subsequently required euthanasia due to tumor progression (Fig. 2D). However, eradication of residual tumor was observed in 9 of the 10 mice treated with surgery + CARM5 (fibrin gel) (Fig. 2D), which resulted in significantly longer overall survival in mice receiving surgery + CARM5 (fibrin gel) when compared to mice with surgery + CAR19 (fibrin gel), as well as to mice with surgery + CARM5 (medium) (Fig. 2E). Staining of peripheral blood could detect significantly more CD45⁺ cells in mice treated with surgery + CARM5 (fibrin gel) at 14 days after treatment when compared to the control cohorts (Fig. 2F). As observed in the TNBC xenograft model, transient and most likely surgery-related weight loss was also seen in some mice directly following treatment throughout all cohorts, with subsequent increase and stabilization of the body weight (Fig. 2G). Clinically, none of the mice showed postsurgical signs of inflammation or wound healing complications at the surgery site.

Taking the data of both xenograft models together (Figs. 1 and 2 and supplementary figures), we concluded that mesothelin-specific CAR T cells applied in fibrin gel within the resection cavity cleared residual tumor cells following incomplete surgical excision in all mice, unlike CAR T cells delivered at the same dose but without fibrin gel. These results suggest that regional delivery of CAR T cells in fibrin gel provides antitumor activity over an extended period of time after surgery. Systemic trafficking of the T cells in the peripheral blood was detected in some mice after fibrin gel implantation. Clinically, wound healing complications following treatment were not observed in any of the mice.

Reduction of on-target off-tumor toxicity with local CAR T cell treatment

To evaluate the potential of systemic on-target off-tumor toxicity of this treatment approach, we developed a novel in vivo toxicity model using human mesothelin knock-in NSG (huMeso-KI-NSG) mice. These mice express human mesothelin knocked into the mouse msln locus on the NSG background. In this model, the same anti-human mesothelin CAR can be tested; however, unlike in the standard NSG xenograft model, healthy tissue can also be attacked by human mesothelin-specific CAR T cells, which can result in immunopathology and toxicity-related death. For the toxicity studies, human T cells were transduced with a lentiviral vector coencoding click beetle red (CBR) luciferase and the mesothelin-specific CAR construct (CARM5-CBR), allowing the T cells to be tracked in vivo via bioluminescence imaging (BLI). For a specificity control, CBR-labeled CD19-specific CAR T cells were used (CAR19-CBR). Cell surface staining of the transduced T cells revealed CAR-CBR expression in approximately 47 and 67% of mesothelin-specific and CD19-specific cells, respectively (fig. S7A). In vitro, CARM5-CBR T cells and CAR19-CBR T cells were luciferase active (fig. S7B), and CARM5-CBR T cells were functional and specifically secreted the cytokines IL-2, TNF, and IFN-γ when coincubated with MDA-MB-231 (Fig. 3A). In addition, specific tumor cell killing by CARM5-CBR T cells was seen after tumor cell coincubation (Fig. 3B).

For the in vivo toxicity model, MDA-MB-231 cells were injected subcutaneously into the right flank of huMeso-KI-NSG mice (Fig. 3C). Twenty-two days later (= day 0), mice were treated either with incomplete surgery followed by direct intracavity inoculation of CAR19-CBR T cells in fibrin gel for control [surgery + CAR19-CBR (fibrin gel)] or with incomplete surgery followed by direct intracavity inoculation of CARM5-CBR T cells in fibrin gel [surgery + CARM5-CBR (fibrin gel)] (Fig. 3C). Luciferase-based biodistribution and total flux analyses of CARM5-CBR T cells showed that T cells migrated out of the gel and engrafted into the residual tumor following incomplete surgical excision within 7 days in all mice (Fig. 3, D and F). However, in four of the 10 mice, CARM5 T cells trafficked out of the flank tumor site beginning at week 4 and expanded at other anatomical sites within healthy tissues (on-target off-tumor) (Fig. 3D; mouse nos. 2, 3, 5, and 7). This resulted in therapy-related death in three of these mice within 12 weeks after treatment (Fig. 3D; mouse nos. 3, 5, and 7). In 2 of the 10 treated mice, T cells migrated from the local site but did not have continued expansion detected (Fig. 3D; mouse nos. 4 and 6), while in the remaining 4 of the 10 mice, T cells did not leave local site at detectable levels (Fig. 3D; mouse nos. 1, 8, 9, and 10). The luciferase signal of CAR T cells in the latter four mice showed a continual decrease and became undetectable after clearance of the residual tumor in two mice (Fig. 3E; mouse nos. 1 and 9 in Fig. 3D).

In the control group of huMeso-KI-NSG mice receiving CAR19-CBR T cells, none of the mice showed T cell engraftment into the tumor, and no luciferase signal could be detected (Fig. 3, D and E). All of the mice in the control cohort were euthanized because of tumor progression, measured by caliper, within 8 weeks (fig. S7C). In contrast, 9 of the 10 mice treated with CARM5-CBR exhibited clearance of the residual tumor (fig. S7C). Despite the death of three mice due to on-target off-tumor toxicity, significantly longer overall survival was observed in mice receiving surgery + CARM5-CBR (fibrin gel) when compared to the control cohort (Fig. 3F). At days 15 and 36, no statistical differences were detected in the levels of CD45⁺ cells in peripheral blood between cohorts (Fig. 3G). In mouse serum, significantly higher concentrations of IFN-γ were observed at both time points after treatment, while significantly higher concentrations of TNF were only seen at day 15 after treatment (Fig. 3H). No differences could be detected in IL-2 concentrations (Fig. 3H). Transient and most likely surgery-related weight loss in some mice directly following treatment was seen in both cohorts, with subsequent resumption of growth (fig. S7D). However, delayed toxicity-related weight loss was observed in two animals treated with CARM5 (fibrin gel) (fig. S7D; mouse nos. 2 and 7 in Fig. 3D). Clinically, none of the mice showed postsurgical signs of inflammation or wound healing complications.

No or low T cell infiltration into lungs is seen in locally treated mice

We next sought to evaluate CAR T cell infiltration into the lungs as a result of on-target off-tumor toxicity in huMeso-KI-NSG mice. Therefore MDA-MB-231 cells were injected subcutaneously into the right flanks of huMeso-KI-NSG mice (Fig. 4A). Twenty-two days later (i.e., day 0), mice were treated either with incomplete surgery followed by direct intracavity inoculation of trackable CAR19-CBR T cells for control [surgery + CAR19-CBR (fibrin gel)] or with incomplete surgery followed by direct intracavity inoculation of trackable CARM5-CBR T cells [surgery + CARM5-CBR (fibrin gel)]. In this study, we used two more cohorts: Healthy (non–tumor-bearing) mice were treated with an equal dose of intravenously injected mesothelin-specific CAR T cells (intravenous CARM5-CBR) as a positive control for on-target off-tumor–associated CAR T cell infiltration and expansion in the lung, as well as untreated and healthy (non–tumor-bearing) mice as negative control where no T cell detection was expected (Fig. 4A). Twenty-one days after treatment, mice were euthanized following terminal bleeding, and lung tissues were analyzed for hematoxylin and eosin (H&E) and immunohistochemistry (IHC) staining for human mesothelin and human CD45.

Fig. 4. — (A) MDA-MB-231 cells were injected subcutaneously into the right flank of huMeso-KI-NSG mice. Twenty-two days later, mice were treated as indicated. As positive control, healthy (non–tumor-bearing) mice were intravenously treated with CAR T cells (intravenous CARM5). Twenty-one days after treatment, mice were euthanized following terminal bleeding, and lung tissues were analyzed. Tissues of healthy (non–tumor-bearing) and nontreated mice were also analyzed for control. n = 5 mice per cohort. (B) Biodistribution of CBR-labeled CAR T cells 1 day before organ harvest. Luciferase signal was tracked via BLI. (C) Total flux values of CBR-CAR T cells in BLI 1 day before organ harvest. Average and individual values are shown. Gray area below horizontal dotted lines represents background luminescence. Kruskal-Wallis one-way analysis of variance was used for statistical analysis. ns, P > 0.05; **P ≤ 0.01. (D) Tumor volume 1 day before organ harvest. Mann-Whitney U test was used for statistical analysis. **P ≤ 0.01. (E) CD45⁺ cell count in peripheral blood at the day of organ harvest. Kruskal-Wallis one-way analysis of variance was used for statistical analysis. ns, P > 0.05. (F) IL-2, TNF, and IFN-γ concentration in serum at the day of organ harvest. Kruskal-Wallis one-way analysis of variance was used for statistical analysis. ns, P > 0.05; *P ≤ 0.05. (G) Heatmap summarizing the pathological assessment and IHC for CD45. Top box shows results of individual mice. Bottom box shows the grading legend. (H) Photomicrographs (×40) of hematoxylin and eosin (H&E; top images) and immunohistochemical (IHC) staining for human CD45⁺ cells (hCD45; in brown, bottom images) of lung tissues and one representative mouse per cohort are shown. Arrows indicate infiltration of lymphocytes.

Robust engraftment and expansion, as measured by BLI, were detected in all intravenously dosed CARM5-CBR T cell tumor-free mice and in tumor-bearing mice treated with surgery + CARM5-CBR T cells until the day of organ harvest, while no engraftment and expansion as judged by the luciferase signal were seen in surgery + CAR19-CBR T cell–treated and in untreated healthy mice (Fig. 4, B and C, and fig. S8A). No tumor regrowth following incomplete surgery was seen in CARM5-CBR–treated mice following surgery, while all mice in the control cohort showed subsequent tumor progression (Fig. 4D and fig. S8B). At the day of organ harvest, CD45⁺ cells could be detected in the peripheral blood of intravenously dosed CARM5-CBR–treated mice and in mice treated with surgery + CARM5-CBR T cells, while no statistical significance between these two cohorts was observed (Fig. 4E). In addition, high concentrations of IFN-γ could be detected in the serum of intravenously dosed CARM5-CBR–treated mice and in surgery + CARM5-CBR T cell–treated mice at the day of organ harvest, while low TNF and IL-2 concentrations were seen (Fig. 4F).

IHC staining for human mesothelin revealed intravascular and interstitial mesothelin expression within lung tissues as well as pronounced expression in the pleural layer of mice throughout all cohorts, confirming the potential for on-target off-tumor toxicity by human mesothelin-specific CAR T cells (fig. S8C). H&E staining of lung tissues in conjunction with IHC staining for human CD45 showed mild to moderate subpleural and interstitial infiltration or intravascular circulation of CD45⁺ cells in positive control mice (intravenous CARM5-CBR), while mice receiving surgery + locally implanted CARM5-CBR showed minimal CD45⁺ cell infiltrates (Fig. 4, G and H). In one mouse treated with surgery + CARM5-CBR, no CD45⁺ cells were seen in IHC, while in another mouse, mild CD45⁺ cell infiltration was observed (Fig. 4G). No CD45⁺ cell infiltration was seen in the control cohorts (surgery + CAR19 and untreated mice; Fig. 4, G and H), indicating that T cell expansion and infiltration were dependent on mesothelin recognition. Taking the huMeso-KI-NSG data together (Figs. 3 and 4 and supplementary figures), we concluded that local mesothelin-specific CAR T cell delivery reduces the incidence and severity of on-target off-tumor toxicity. Local CAR T cells cleared the residual tumor and remained at the surgery site in some mice, which may increase the therapeutic index when targeting antigens with shared expression on tumor and healthy tissues.

No impact on wound healing is observed in locally treated immunocompetent mice

To further analyze potential wound healing complications following local use of CAR T cells in fibrin gel, a syngeneic partial tumor resection model was established using the murine pancreatic cancer cell line PDA7940b (fig. S9A). Murine T cells were transduced to express a mouse mesothelin-specific CAR construct (CARmMeso) or a human CD19-specific CAR construct for control (CARh19). At the time of in vivo and in vitro use, cell surface staining of transduced T cells revealed CAR expression in approximately 82 and 72% of mouse mesothelin-specific and human CD19-specific cells, respectively (fig. S9B). In vitro, these mouse CAR T cells showed functionality by specifically secreting TNF and IFN-γ when coincubated with the murine KPC (Kras^LSL.G12D/+p53^R172H/+) pancreatic cancer cells PDA7940b, while specific IL-2 secretion was not detected (Fig. 5A).

For in vivo studies, C57BL/6 mice were inoculated with PDA7940b cells subcutaneously into the right flank (Fig. 5B). At day 10 (1 day before treatment), mice underwent lymphodepletion with cyclophosphamide. Surgery with or without local use of CAR T cell in fibrin gel was performed 1 day following lymphodepletion (= day 0), when the tumor size was approximately 0.7 cm in diameter (~100 to 150 mm³) (fig. S9C). Mice received either incomplete surgery only (surgery), incomplete surgery followed by direct intracavity inoculation of CARhu19 T cells in fibrin gel for control [surgery + CARhu19 (fibrin gel)], or incomplete surgery followed by direct intracavity inoculation of CARmMeso T cells in fibrin gel [surgery + CARmMeso (fibrin gel)] (Fig. 5B). A higher CAR T cell dose was chosen for this syngeneic animal experiment as a stress test to investigate the possibility of wound healing complications (Fig. 5B).

Progressive tumor regrowth after treatment was observed in mice receiving incomplete surgery only or surgery + CARhu19 (fibrin gel) (Fig. 5C). In contrast, incomplete surgery followed by CARmMeso T cells in fibrin gel showed clearance of residual tumor cells in five of the six mice (Fig. 5C), which resulted in significantly longer overall survival when compared to mice receiving surgery + CARhu19 (fibrin gel) (Fig. 5E). Transient and most likely surgery-related weight loss with subsequent stabilization within 7 days was observed in all cohorts (Fig. 5D). Postsurgery, detailed clinical follow-up for potential wound-healing complications was performed on a daily basis for a total of 10 days and then twice weekly. None of the treated mice showed clinical signs of wound healing complications at the surgery site (e.g., no redness, swelling, exudate, or dehiscence of surgical clips) (Fig. 5F).

We next sought to evaluate wound healing following intraoperative CAR T cell administration on histopathological levels. Therefore, C57BL/6 mice were inoculated with PDA7940b cells subcutaneously into the right flank (fig. S10A). Ten days later, mice underwent lymphodepletion with cyclophosphamide and received surgery + CARmMeso T cells in fibrin gel 1 day following lymphodepletion. In this experiment, we used three more cohorts: non–tumor-bearing mice were either left untreated (healthy skin), underwent surgical incision only, or underwent surgical incision followed by local intracavitary application of CARmMeso T cells in fibrin gel for control (fig. S10A). Ten days after treatment, mice were euthanized, and skin tissues at resection site were analyzed for wound healing on histopathological levels via H&E staining.

Physiologic wound healing was seen in all mice throughout all cohorts as per the histopathological assessment (Fig. 5G and fig. S10B). All animals that underwent a cutaneous incision displayed variable levels of dermal and subcutaneous fibrosis and granulation tissue formation, inflammation with intralesional foreign material (hair shafts), and epidermal hyperplasia as expected (fig. S10C). Although levels of fibrosis and granulation tissue and inflammation were slightly higher in the non–tumor-bearing animals receiving CAR T cells (fig. S10C), these were expected changes in normal wound healing. Together, these results confirmed that CAR T cells in fibrin gel can be used as an adjuvant to surgery, clearing residual tumor cells in immunocompetent C57BL/6 mice without inducing wound healing complications.

DISCUSSION

Here, we found that CAR T cells can be safely and effectively used to clear residual cancer cells following incomplete surgical excision of adenocarcinomas in two animal models. Our studies confirm and extend the results in the study of Ogunnaike et al. (14), showing that fibrin gels could be used to implant CAR T cells in the resection cavity of gliomas. On-target off-tumor toxicity was diminished compared to mice treated with systemically administered CAR T cells. Wound healing complications were not observed in immunocompromised or immunocompetent mice.

A major limitation of the use of CAR and TCR T cells for solid tumors has been on-target off-tumor toxicity. For example, adverse effects in clinical trials have been reported for human epidermal growth factor receptor 2 (her2) and carcinoembryonic antigen (CEA) (15, 16). Mesothelin is a promising target for CAR and TCR T cells for a variety of solid tumors (17), and dose-limiting pulmonary toxicities have been observed after systemic administration of mesothelin-specific CAR T cells (NCT03054298 and NCT03907852). Although human CAR T cells may not persist over a period of several weeks or months in mice, which reduces the likelihood of clinically evident on-target off-tumor toxicities, our results suggest that local delivery of CAR T cells would be expected to reduce the risk of such toxicities, as local CAR T cells did not efficiently leave the surgical site in huMeso-KI-NSG mice. Furthermore, there was a significant reduction in lung and pleural immunopathology in these mice after local injection compared to intravenous injection of CAR T cells.

A concern with the use of CAR T cells for local control of tumors is whether inflammation and tissue damage would create an unhealing wound. The mechanism of tumor killing by CAR T cells has not been fully understood and may include direct cytotoxic killing by CAR T cells as well as recruitment of innate effector cells such as macrophages. The lack of inflammatory tissue damage that we observed is consistent with previous studies of cytotoxic T cells and tumor eradication (18). Furthermore, these studies are consistent with the lack of inflammation observed in allophenic mice undergoing skin allograft rejection (19).

Head and neck cancers have the highest incidence of positive surgical margins for both genders, while ovarian and prostate cancer have the highest gender-specific incidence of positive margins (3). Approximately 25% of patients with pancreatic cancer experience local recurrence after the initial resection (20). In addition, it is known that breast-conserving surgery can also increase the risk of positive surgical margins and adverse outcomes (21). Given that systemic CAR T cells are being tested for advanced disease in all these histologies, the use of locally administered CAR T cells during initial surgery may be useful. On the basis of our observations in this study, a clinical trial in patients with locally advanced breast cancer is now planned at our center. Fibrin glue is already U.S. Food and Drug Administration–approved, and the human mesothelin-specific CAR T cell construct is now being tested after systemic administration in different clinical trials with known toxicity profiles, which further facilitates the clinical translation of this approach. In the case of breast cancer, CAR target antigens, besides mesothelin, could include HER2 (lacking in TNBC), tyrosine-protein kinase Met, CEA, and cancer-associated Tn glycoform of Mucin 1 (MUC1), which were or now are being clinically tested in different tumor entities (17, 22). Beside breast cancer where mesothelin is frequently expressed (17, 23), the use of this approach might be a promising tool for a variety of other mesothelin-expressing tumor entities where complete surgical excision might not be always possible, including pancreatic cancer (17). After tumor resection, albeit incomplete, the overall tumor burden is expected to be much lower and a smaller CAR T cell dose may be sufficient to eradicate residual tumor cells, which might further decrease possible systemic toxicities in a clinical setting.

There are several potential limitations of local administration of CAR T cells during surgery. In cases of planned resection where positive surgical margins are anticipated, autologous and allogeneic CAR T can be prepared. For clinical translation of this approach, autologous cells will be used in the first step, and recent approaches developed at our center and by others could minimize the manufacturing time of autologous CAR T cells to less than 24 hours (24, 25). The use of an “off-the-shelf” approach with allogeneic CAR T cells generated by using T cells from healthy donors (26) could further facilitate rapid intraoperative availability of the CAR T cell. A limitation of allogeneic CAR T cells is short duration of persistence, and this may not be a drawback with local eradication of tumor. A future direction would be to use CAR T cells that are designed to activate and prime endogenous antitumor immunity. It is possible to design CAR T cells that recruit dendritic cells to the tumor microenvironment (27) or to combine CAR T cells with other agents such as cytokines or oncolytic viruses that may recruit endogenous immunity and therefore accomplish local tumor control while preventing systemic recurrence (28, 29). From a theoretic perspective, it is worth considering that local irradiation is often used for therapy of locally recurrent tumors and that the irradiation may be partially antagonistic through toxic effects on potentially beneficial tumor infiltrating lymphocytes.

In summary, CAR T cells can be effectively and safely used as a surgical adjuvant therapy for solid tumors that cannot be completely excised. The use of a fibrin glue–based carrier in this context may represent an effective and translatable tool to maximize CAR T cell distribution and antitumor efficacy at the tumor resection site. This approach can be broadened to additionally deliver cytokines, oncolytic viruses, checkpoint antibodies, or other drugs/biological agents, which together might further increase the overall antitumor response.

MATERIALS AND METHODS

Study design

The purpose of this study was to use CAR T cells as an adjuvant therapy in unresectable solid tumors as a novel approach, applied intraoperatively in a fibrin glue–based carrier into the resection cavity to clear residual cancer cells following incomplete surgical excision. To analyze this approach, partial excision animal models were established to mimic positive surgical margins. Four mouse models were used: two xenograft models, a toxicity model, and a syngeneic model. The feasibility of this approach was shown in two xenograft models (human PDA and human TNBC) using NSG mice and two different human T cell donors to confirm that the observed effects are independent of the tumor entity and T cell donors. The potential of on-target off-tumor toxicities was analyzed in a novel toxicity model using huMeso-KI-NSG mice, while immunocompetent C57BL/6 mice were used to rule out wound healing complications.

Human tumor cells

The human PDA cell line AsPC-1, the human TNBC cell line MDA-MB-231, and the human leukemia cell line K562 were obtained from the American Type Culture Collection and routinely authenticated by the University of Arizona Genetics Core. MDA-MB-231 cells were lentivirally transduced and fluorescence-activated cell sorting (FACS)–sorted to express human mesothelin, which served as the target antigen for our CAR construct, while AsPC-1 endogenously expresses mesothelin. For in vitro imaging–based cytotoxicity assays using the Incucyte SX5 live-cell analysis instrument (Sartorius), tumor cell lines were additionally transduced with a lentiviral vector coencoding click beetle green (CBG) luciferase and GFP under an elongation factor 1 alpha (EF1α) promoter separated by a P2A sequence. Tumor cell lines were then FACS-sorted for ~100% GFP positivity before use as targets in experiments. The purity of mesothelin and GFP expression was routinely validated by flow cytometry. K562 cells, transduced with CBG-GFP as well as either human mesothelin or human CD19, were used for in vitro functionality assays only. All tumor cell lines were tested in regular intervals in the presence of mycoplasma contamination by the Department of Genetics at University of Pennsylvania (MycoAlert Mycoplasma Detection Kit, Lonza). AsPC-1 cells were maintained in culture with R20 medium: RPMI 1640 (Gibco) supplemented with 20% heat-inactivated fetal bovine serum (FBS; Seradigm), 2% 1 M Hepes buffer solution (Gibco), 1% 100× GlutaMAX (Gibco), and 1% penicillin (10,000 U/ml) + streptomycin (10,000 μg/ml) (Gibco). MDA-MB-231 cells and K652 cells were maintained in culture with R10 medium: RPMI 1640 (Gibco) supplemented with 10% heat-inactivated FBS (Seradigm), 2% 1 M Hepes buffer solution (Gibco), 1% 100× GlutaMAX (Gibco), and 1% penicillin (10,000 U/ml) + streptomycin (10,000 μg/ml) (Gibco).

Murine tumor cells

The murine PDA cell line PDA7940b was established from the KPC mouse pancreatic tumor model (30) and was a gift from G. Beatty (University of Pennsylvania). Tumor cell lines were tested in regular intervals in the presence of mycoplasma contamination by the Department of Genetics at University of Pennsylvania (MycoAlert Mycoplasma Detection Kit, Lonza) and were maintained in culture with R10 medium (see above).

Human CAR T cells

Lentiviral vectors and human CAR T cells were produced as previously described (22). For human CAR T cell production, healthy donor T cells (ND517, ND569, ND500, and ND502) were purchased from Human Immunology Core at University of Pennsylvania. T cells were stimulated with anti-human CD3/CD28 antibody-coated beads (Dynabeads, Gibco) at a bead:T cell ratio of 3:1. Twenty-four hours later, T cells were lentivirally transduced to express a second-generation human mesothelin-specific CAR (CARM5) (31) or a second-generation human CD19-specific CAR (CAR19) for control (32) (donors ND517 and ND569). For generation of BLI-trackable CAR T cells, T cells were transduced with a lentiviral vector coencoding click beetle red (CBR) luciferase and the second-generation human mesothelin/CD19 CAR constructs (CARM5-CBR and CAR19-CBR; donor ND500). For in vitro migration assays using the Incucyte SX5 live-cell analysis instrument (Sartorius), T cells were transduced with a lentiviral vector coencoding GFP and the second-generation human mesothelin-specific CAR construct (CARM5-GFP; donor ND502). A multiplicity of infection of 4 was used for T cell transduction. At day 6 of T cell stimulation, beads were removed from culture, and cells were allowed to rest down to a volume of ~350 fl before cryopreservation. For cryopreservation of T cells, freezing medium consisting of 90% heat-inactivated FBS (Seradigm) and 10% dimethyl sulfoxide (Sigma-Aldrich) was used. CAR expression of transduced T cells were quantified by anti-idiotype staining (CAR19 and CAR19-CBR), by anti-Fab(2)′ staining (CARM5 and CARM5-CBR), or by GFP expression (CARM5-GFP) and flow cytometry 1 day before or at day of cryopreservation (for flow cytometry and antibodies, see below). Human T cells were maintained in culture with R10 medium (see above).

Murine CAR T cells

Murine CAR T cell production using retroviral vectors was previously described (29). Briefly, spleens from C57BL/6 mice were harvested, and T cells were purified with mouse T cell isolation beads (STEMCELL Technologies). Purified mouse T cells were activated with anti-mouse CD3/CD28 antibody-coated beads (Dynabeads, Gibco) at a bead:T cell ratio of 2:1. T cells were then retrovirally transduced to express a mouse mesothelin-specific CAR construct (MSGV-mMesoBBz; CARmMeso) or a human CD19-specific CAR construct for control (MSGV-hu19BBz; CARhu19) on recombinant human fibronectin-coated plates (Retronectin, TaKaRa) 2 days after bead stimulation. Recombinant mouse IL-2 (50 U/ml) was supplemented at day 1 and then supplemented with fresh mouse T cell medium [= RPMI 1640 (Gibco) supplemented with 10% heat-inactivated FBS (Seradigm), 1% 100× GlutaMAX (Gibco), 1% penicillin (10,000 U/ml) + streptomycin (10,000 μg/ml) (Gibco), 1× 100 mM sodium pyruvate (Gibco), and 50 μM β-mercaptoethanol (Sigma-Aldrich)] containing IL-2 (50 U/ml) every day. At day 5 of stimulation, mouse CAR T cells were harvested, debeaded, analyzed for CAR expression by anti-Fab(2)′ staining and flow cytometry (for flow cytometry and antibodies, see below), and used for in vitro and in vivo experiments the same day.

Preparation of fibrin gel components

Ready-to-use 2-ml Tisseel (fibrin sealant) prefilled syringes were purchased from the company Baxter. The double-chamber syringes contain (i) the human sealer protein/fibrinogen component (concentration of human fibrinogen according to package insert, ~ 91 mg/ml) and (ii) the human thrombin component (concentration of human thrombin according to package insert, ~ 500 U/ml). The fibrinogen component was diluted to the appropriate end concentration using 1× tris-buffered saline (TBS; Bio-Rad). The human thrombin solution was diluted to the desired end concentration using 30 mM calcium chloride (Sigma-Aldrich) in 1× TBS (Bio-Rad).

Preparation of CAR T cells in fibrin gel

CAR T cells in fibrin gel solution were prepared in a total volume of 100 μl as follows: 25 μl of R10 medium/mouse T cell medium supplemented with human/murine IL-7 (10 ng/ml; Miltenyi Biotec) and human/murine IL-15 (10 ng/ml; Miltenyi Biotec) containing the required amount of CAR T cells were added to 37.5 μl of fibrinogen solution. The thrombin solution (37.5 μl) was then added to the T cell/fibrinogen solution, mixed, and immediately used for experiments.

In vitro passive migration and viability of fibrin gel–embedded CAR T cells

GFP-labeled human CAR T cells (CARM5-GFP; donor ND502) were used to analyze ability of T cells to passively move out of the fibrin gel matrix in a transwell assay. A 5.0-μm semipermeable polycarbonate membrane separates the upper compartment of the transwell plate (costar) to the lower compartment. For the assay, 1E6 CARM5-GFP T cells were mixed in fibrin gel at different concentrations of fibrinogen solution (1, 5, 10, and 20 mg/ml) or thrombin solution (0.3, 1, 3, and 10 U/ml). The T cell/fibrin gel mix was then added into the upper component of the transwell plate. As control, CAR T cells in R10 medium (= no fibrin gel) and fibrin gel only (= no CARM5-GFP T cells) were used. The lower compartment of the 24 transwell plate contained R10 medium. Twenty-four hours later, the amount of GFP-positive T cells in the lower compartment was counted using the Incucyte SX5 live-cell analysis instrument (Sartorius). For data analysis, the amount of migrated CARM5-GFP T cells in R10 medium (= no fibrin gel) was considered as 100%. Viability of T cells in the lower transwell compartment was analyzed using trypan blue stain (Invitrogen) and Countess automated cell counter device (Invitrogen).

In vitro cytokine secretion assay

CAR T cells were stimulated overnight with the tumor cell lines MDA-MB-231, AsPC-1, K562, or PDA7940b at a 1:1 effector:target cell ratio. Concentrations of the cytokines IL-2, TNF, and IFN-γ in the supernatants were analyzed using the human T helper 1 (T_H1)/T_H2 cytokine cytometric bead array (CBA) kit II for human T cells (BD Biosciences) or the mouse T_H1/T_H2 cytokine CBA kit for mouse T cells (BD Biosciences) according to the manufacturer’s instructions. Following acquisition of sample data using Fortessa LSR II flow cytometer (BD Biosciences) equipped with the FACSDiva software (BD Biosciences), the sample results were generated using the BD CBA analysis software (BD Biosciences).

In vitro cytotoxicity assay using Incucyte SX5

In vitro analysis of CAR T cell killing was performed using the Incucyte SX5 live-cell analysis instrument (Sartorius). Therefore, GFP-expressing tumor cells MDA-MB-231 and AsPC-1 were seeded in 96-well plates. After approximately 24 hours, CAR T cells were added at an effector:target cell ratio of 3:1 (T cell donor ND500) or 1:1 (T cell donors ND517 and 569). The amount of GFP-positive tumor cells per image was counted by the Incucyte SX5 live-cell imager every 3 hours over 5 to 7 days. For data analysis, cell counts were normalized to the time point of CAR T cell addition (approximately 24 hours after start of experiment).

In vitro luciferase assays

Luciferase activity of CBR-labeled CAR T cells (CARM5-CBR and CAR19-CBR; donor ND500) was tested before use in in vitro and in vivo experiments. Therefore, cells were seeded in a 96-well plate, and relative light units were measured using Synergy H4 hybrid multimode microplate reader (BioTek). Unlabeled CAR T cells (CARM5 and CAR19; donor ND517) and R10 medium only were used for control. For luciferase-based killing assay using the luciferase-labeled human leukemia cell line K562-meso/K562-CD19, CAR T cells were added to tumor cells at an effector:target cell ratio of 60:1, 20:1, 6:1, and 2:1, as indicated in the figure legends. Twenty-four hours later, luminescence was measured using Synergy H4 hybrid multimode microplate reader (BioTek). Tumor cells only and medium only were used for controls. Percentage of specific lysis was calculated using the following formula: % specific lysis = 100 × [(experimental data − spontaneous cell death)/(maximum cell death − spontaneous cell death)].

Mice

The University of Pennsylvania Institutional Animal Care and Use Committee (IACUC) approved all animal experiments (protocol number 804226), and all animal procedures were performed in the animal facility at the University of Pennsylvania in accordance with Federal and Institutional IACUC requirements. NSG and huMeso-KI-NSG mice were originally procured from the Jackson Laboratory and bred by the Stem Cell and Xenograft Core (SCXC) at the University of Pennsylvania. For syngeneic mouse experiments, C57BL/6 mice were obtained from the Jackson Laboratory. Six- to 8-week-old female NSG mice, huMeso-KI-NSG mice, and C57BL/6 mice were used for in vivo experiments. Mice were maintained under pathogen-free conditions. Schemas of the used mouse models are shown in detail in each relevant main and supplementary figure. One day before treatment, tumor size was measured in all in vivo experiments to calculate tumor volume. Mice were then categorized into cohorts with equal average tumor sizes. Animals with rapid or slow tumor growth were excluded from the study before surgery/CAR T cell administration. A maximum of six animals per cohort were used per experiment so that qualitatively equivalent surgical excisions could be guaranteed in all treated animals. Data analysis was based on objectively measurable data in an unblinded fashion. Mice were subjected to routine veterinary assessment for signs of overt illness and were killed at experimental termination or when predetermined IACUC rodent health endpoints were reached.

NSG mouse experiments

MDA-MB-231 and AsPC-1 cells were used for in vivo experiments with NSG mice. 4E6 or 2E6 of MDA-MB-231 cells or AsPC-1 cells, respectively, in a total volume of 100 μl 1:1 matrigel (Corning):1× Dulbecco's phosphate-buffered saline (DPBS) (Gibco) mix were implanted subcutaneously into the right flank of NSG mice. Tumor size was allowed to reach approximately 1 cm in diameter before treatment (tumor volume, ~250 to 300 mm³). At day 22, approximately 75% of the subcutaneous tumor was surgically removed (for description of the surgical procedure, see below). Directly following incomplete surgical excision, 1.5E5 human mesothelin-specific CAR T cells (CARM5; donors ND517 and ND569) in fibrin gel using fibrinogen solution (5 mg/ml) and thrombin solution (3 U/ml) were inoculated into the resected surgical cavity. For control, mice were treated with incomplete surgery only, with human CD19-specific CAR T cells (CAR19) in fibrin gel, or with CARM5 T cells in medium (= no fibrin gel) following incomplete surgical excision. For in vivo fibrin gel titration experiments, 1.5E5 CARM5 T cells (donor ND517) in fibrin gel using fibrinogen solution (5 mg/ml) and the following thrombin solutions were analyzed: 0.3 U/ml, 1 U/ml, 3 U/ml, 10 U/ml, or medium only (= no fibrin gel). Caliper and mouse body weight measurements were performed at least once a week. Values for tumor size and body weight were normalized to day −1 in the main and supplementary figures. Staining for CD45⁺ cells in the peripheral blood was performed at indicated time points. Health monitoring followed the IACUC body scoring system guidelines.

huMeso-KI-NSG mouse experiments (in vivo toxicity model)

To analyze potential on-target off-tumor toxicities, a novel toxicity in vivo model was developed. These NSG-background mice express human mesothelin knocked into the mouse msln locus (huMeso-KI-NSG). Unlike in the standard NSG xenograft model, healthy tissue can be attacked by human mesothelin-specific CAR T cells, which can result in immunopathology and toxicity-related death. To establish the partial excision huMeso-KI-NSG model, 4E6 MDA-MB-231 cells at a total volume of 100 μl 1:1 matrigel (Corning):1× DPBS (Gibco) mix were injected subcutaneously into the right flank of mice. At day 22 when tumor size was approximately 1 cm in diameter (tumor volume, ~250 to 300 mm³), mice underwent incomplete surgery (removal of approximately 75% of the subcutaneous tumor; for description of the surgical procedure, see below) with or without local use of trackable CBR-CAR T cells (donor ND500) in fibrin gel [fibrinogen solution (5 mg/ml) and thrombin solution (3 U/ml)]. Mice were treated either with incomplete surgery followed by direct intracavity inoculation of CARM5-CBR T cells in fibrin gel or with incomplete surgery followed by direct intracavity inoculation of CAR19-CBR T cells in fibrin gel for control. For positive control of pathology samples, healthy (non–tumor-bearing) huMeso-KI-NSG mice were treated with the equal amount of intravenously injected CARM5-CBR T cells to induce on-target off-tumor T cell infiltration and expansion. BLI to track CBR-labeled CAR T cells was performed using IVIS Lumina III (PerkinElmer) once weekly after treatment. BLI images were analyzed with the Living Image software (Caliper Life Sciences). In addition, tumor size measurements by caliper and mouse body weight measurements were performed at least once a week following treatment. Values for tumor size and body weight were normalized to day −1 in the main and supplementary figures. Staining for CD45⁺ cells in the peripheral blood and cytokine analysis from mouse serum was performed at indicated time points. Twenty-one days following treatment, some mice were euthanized to perform pathological assessment of lung tissue (see below).

Syngeneic mouse experiments

C57BL/6 mice were inoculated with 5E5 PDA7940b tumor cells in a total volume of 100 μl 1× DPBS (Gibco) subcutaneously into the right flank. At day 10 (1 day before treatment), mice underwent lymphodepletion with intraperitoneal cyclophosphamide (Sigma-Aldrich) at a dose of 120 mg/kg. Incomplete surgery by removal of approximately 75% with or without local use of CAR T cells in fibrin gel [fibrinogen solution (5 mg/ml) and thrombin solution (3 U/ml)] was performed at day 11, when the tumor size was approximately 0.7 cm in diameter (tumor volume, ~100 to 150 mm³). Mice received incomplete surgery only (removal of approximately 75% of the subcutaneous tumor; for description of the surgical procedure, see below), incomplete surgery followed by direct intracavity inoculation of mouse mesothelin-specific CAR T cells in fibrin gel (CARmMeso), or incomplete surgery followed by direct intracavity inoculation of human CD19-specific CAR T cells in fibrin gel for control (CARhu19). Caliper and mouse body weight measurements were performed at least twice weekly. Values for tumor size and body weight were normalized to day −1 in the main and supplementary figures. Mice were closely monitored following treatment on a daily basis for 10 days and then twice weekly. Special attention was given to the following clinical signs: development of redness, swelling, or pus formation at surgery site, as well as dehiscence of surgical clips. Clinical images were taken from all mice at day 1, day 6 (removal of surgical clips), and day 10 after treatment to objectivize potential wound healing complications. Ten days following treatment, some mice were euthanized to perform pathological assessment of skin tissue at surgical resection site (see below).

Incomplete surgical excision of subcutaneous tumors

Surgery was performed aseptically following the IACUC guidelines by surgically trained personnel when tumor size reached approximately 1 cm in diameter in NSG and huMeso-KI-NSG mice (tumor volume, ~250 to 300 mm³) or 0.7 cm in C57BL/6 mice (tumor volume, ~100 to 150 mm³). Tumor size was measured 1 day before surgery in all mice to calculate tumor volume and to assign mice into cohorts such that the average tumor volume of mice in each cohort was approximately equal. For partial surgical excision of the tumor, a previously published protocol that included a video showing the detailed procedure was followed (33). In short, an approximately 1.5-cm straight incision along the dorsal side of the mouse approximately 3 mm away from the tumor edge was made. The wound was opened by gently holding the skin on the tumor bearing side, which is usually attached to the skin side. The skin containing the tumor was then inverted so that the tumor was visible from the outside. Approximately 75% of the capsule containing the tumor was carefully removed from the skin side, so that the resection cavity was still intact. The remaining 25% of tumor including the capsule was not removed, and it was made sure that it was still attached to the skin to maintain existing vasculature. To close the surgical site, the remaining tumor was placed back underneath the skin. If needed, CAR T cells were prepared in fibrin gel or in medium and were applied into the intact resected tumor cavity. The skin sides of the incision were pulled together, and the skin along the wound was lined up. The incision was then closed with wound clips, which were removed at day 6 or 7 after surgery.

Caliper measurements of subcutaneous tumors

Tumor size was measured with calipers, and the volumes were calculated as follows: volume = (length in millimeters × width² in millimeters)/2. Values for tumor size were normalized to day −1 in the main and supplementary figures.

Peripheral blood stain for T cells

Peripheral blood of NSG mice and huMeso-KI-NSG mice was obtained by retro-orbital bleeding or cardiac puncture, and cell numbers of CD45⁺ T cells were quantified using TruCount tubes (BD Biosciences) according to the manufacturer’s instructions. Brilliant Violet 605 anti-human CD45 (BioLegend, catalog no. 368524) and flow cytometry (see below) were used for detection of T cells. Following acquisition of sample data using Fortessa LSR II flow cytometer (BD Biosciences) equipped with the FACSDiva software (BD Biosciences), sample results were analyzed using FlowJo version 10 software (BD Biosciences).

Cytokine analysis from mouse serum

Concentrations of the cytokines IL-2, TNF, and IFN-γ in the serum of huMeso-KI-NSG mice were analyzed using a human T_H1/T_H2 cytokine CBA kit II for human T cells (BD Biosciences) according to the manufacturer’s instructions. Following acquisition of sample data using Fortessa LSR II flow cytometer (BD Biosciences) equipped with the FACSDiva software (BD Biosciences), sample results were analyzed using the BD CBA analysis software (BD Biosciences).

Standard pathological and immunohistochemistry (IHC) analyses of lung and skin tissue

Following euthanasia by CO₂ asphyxiation and terminal cardiac blood collection, the lungs or skin of huMeso-KI-NSG or C57BL/6 mice, respectively, was harvested and prepared for the standard pathological and/or immunohistochemistry (IHC) analysis by the Comparative Pathology Core at the School of Veterinary Medicine of the University of Pennsylvania. Formalin-fixed tissues were routinely processed for paraffin embedding, sectioning, and staining H&E. For immunohistochemistry, 5-μm-thick paraffin sections were mounted on ProbeOn slides (Thermo Fisher Scientific). The immunostaining procedure was performed using a Leica BOND RXm automated platform combined with the Bond Polymer Refine Detection Kit (Leica). Briefly, after dewaxing and rehydration, sections were pretreated with the epitope retrieval BOND ER2 high-pH buffer (Leica) for 20 min at 98°C. Endogenous peroxidase was inactivated with 3% H₂O₂ for 10 min at room temperature (RT). Nonspecific tissue-antibody interactions were blocked with a Leica PowerVision IHC/ISH Super Blocking solution (PV6122) for 30 min at RT. The same blocking solution also served as diluent for the primary antibodies. Rabbit monoclonal primary antibodies against human CD45 (CD45-LCA; Cell Signaling Technology, catalog no. 13917) and human mesothelin (Mesothelin; Invitrogen, catalog no. MA5-16378) were used at a concentration of 1:300 and 1:400, respectively. Antibodies were incubated on the sections for 45 min at RT. A biotin-free polymeric IHC detection system consisting of horseradish peroxidase–conjugated anti-rabbit immunoglobulin G (IgG) was then applied for 25 min at RT. Immunoreactivity was revealed with the diaminobenzidine chromogen reaction. Slides were lastly counterstained in hematoxylin, dehydrated in an ethanol series, cleared in xylene, and permanently mounted with a resinous mounting medium (Thermo Fisher Scientific ClearVue coverslipper). Sections of human tonsil and mesothelioma were included as positive controls. Negative controls were obtained by replacing the primary antibody with an irrelevant isotype-matched rabbit monoclonal antibody. Tissue sections were analyzed by a board-certified veterinary pathologist in a blinded fashion, and details concerning experimental design and tested compounds were revealed only at the end of study. The severity of CD45⁺ cell infiltration and pleural involvement was classified in a semiquantitative way by the pathologist, and results were presented in the form of a heatmap. The scoring of CD45⁺ cell infiltration was performed as follows: (i) no, (ii) minimal, (iii) mild, (iv) moderate, and (v) severe infiltration, with levels of positive cell infiltration ranging from rare positive cells in case of minimal infiltration to high numbers of positive cells within the pulmonary interstitium, subpleural space, or circulating cells in case of severe infiltration. Photomicrographs at a ×40 magnification were acquired using the Olympus DP22 digital cameras with the Olympus cellSens digital imaging software.

Flow cytometry and antibodies

All markers were stained in 1× DPBS (Gibco) containing 3% heat-inactivated FBS (Seradigm). Human mesothelin expression on the human tumor cells MDA-MB-231, AsPC-1, and K562 was detected with biotin anti-human mesothelin (BioLegend, catalog no. 530203). Biotin mouse IgG2a, κ isotype control antibody (BioLegend, catalog no. 400203), was used for isotype control. CD19 expression on K562 cells was analyzed using phycoerythrin (PE)/Dazzle 594 anti-human CD19 antibody (BioLegend, catalog no. 302252). Monoclonal rat anti-mouse MSLN/Mesothelin antibody (LSBio, catalog no. LS-C179484), monoclonal rat IgG2a isotype control antibody (LSBio, catalog no. LS-C292311-1), and PE mouse anti-rat IgG2a antibody (Invitrogen, catalog no. 12-4817-82) were used to stain for mouse mesothelin expression of the murine tumor cells PDA7940b. Cells were stained for viability using the Invitrogen LIVE/DEAD Fixable Near-IR Dead Cell Stain Kit (Invitrogen, catalog no. L10119) according to the manufacturer’s instructions. Biotinylated monoclonal anti-FMC63 scFv antibody, mouse IgG1 (ACROBiosystems, catalog no. 50-201-9662), was used to detect CAR19 T cells. Biotin-SP (long spacer) AffiniPure F(ab′)₂ fragment goat anti-human IgG (Jackson ImmunoResearch, catalog no. 109-066-006) was used for CARM5 T cells and for mouse CAR T cells (CARmMeso and CARhu19). Brilliant Violet 605 anti-human CD45 (BioLegend, catalog no. 368524) was used for detection of T cells in the peripheral blood of NSG/huMeso-KI-NSG mice. For all biotinylated antibodies, PE streptavidin (BioLegend, catalog no. 554061) was used before flow cytometer analysis. All data were collected by a Fortessa LSR II cytometer (BD Biosciences) equipped with the FACSDiva software (BD Biosciences) and analyzed using FlowJo version 10 software (BD Biosciences).

Statistical analysis

Statistical analysis was performed with Prism version 9 (GraphPad Software). Each figure legend denotes the statistical test used. All central tendencies indicate the mean, and all error bars indicate SD. Survival curves were drawn using the Kaplan-Meier method, and the differences of the two curves were compared with the log-rank Mantel-Cox test. Mann-Whitney U test was used for comparisons between two groups, and Kruskal-Wallis one-way analysis of variance was used to compare three or more groups. For all figures, ns (nonsignificant) indicates P > 0.05, * indicates P ≤ 0.05, ** indicates P ≤ 0.01, *** indicates P ≤ 0.001, and **** indicates P ≤ 0.0001. Graphs were created by Prism version 9 (GraphPad Software), BioRender under paid license (https://biorender.com), and PowerPoint (Microsoft).

Acknowledgments

We thank D. Song for technical assistance and the following facilities at the University of Pennsylvania: the Stem Cell and Xenograft Core (SCXC) for providing equipment for the surgical procedures; the Penn Cytomics and Cell Sorting Resource Laboratory at the University of Pennsylvania for assistance in tumor cell sorting; the Human Immunology Core (HIC) at the University of Pennsylvania for reliable supply of healthy human T cells; and the Comparative Pathology Core (CPC) at School of Veterinary Medicine for pathological assessment, immunohistochemistry, image acquisition, and analysis.

Funding: This work was supported by Mildred-Scheel-Postdoctoral Fellowship of the German Cancer Aid (to U.U.) and NIH grant P01CA214278 (to J.S., R.M.Y., and C.H.J.). The Penn Vet Comparative Pathology Core is supported by the Abramson Cancer Center Support Grant (P30CA016520). The scanner used for whole slide imaging and the image acquisition software was supported by an NIH Shared Instrumentation Grant (S10OD023465-01A1).

Author contributions: Conceptualization: U.U., J.T., and C.H.J. Methodology: U.U., T.D., C.-A.A., and J.S. Investigation: U.U., T.D., and C.-A.A. Visualization: U.U. Funding acquisition: U.U., J.T., and C.H.J. Project administration: U.U. and R.M.Y. Supervision: J.S., R.M.Y., J.T., and C.H.J. Writing—original draft: U.U. Writing—review and editing: T.D., C.-A.A., J.S., R.M.Y., J.T., and C.H.J.

Competing interests: R.M.Y. and C.H.J. are inventors on patents and/or patent applications licensed to Novartis Institutes of Biomedical Research and receive license revenue from such licenses. C.H.J. is an inventor on patents and/or patent applications licensed to Tmunity Therapeutics and Capstan Therapeutics. C.H.J. is a scientific cofounder of Tmunity Therapeutics and Capstan Therapeutics and is a member of the scientific advisory boards of AC Immune, Alaunos, BluesphereBio, Cabaletta, Carisma, Cartography, Cellares, Celldex, Decheng, Poseida, Verismo, and WIRB-Copernicus. The other authors declare that they have no competing interests.

Data and materials availability: All data needed to evaluate the conclusions in the paper are present in the paper and/or the Supplementary Materials. For sharing of materials used in this study, a material transfer agreement will be required. Requests for materials should be sent to C.H.J. (cjune@upenn.edu).

Supplementary Materials

This PDF file includes:

Figs. S1 to S10

Click here for additional data file.^{(6.9MB, pdf)}

View/request a protocol for this paper from Bio-protocol.

REFERENCES AND NOTES

1.J. Kleeff, M. Korc, M. Apte, C. La Vecchia, C. D. Johnson, A. V. Biankin, R. E. Neale, M. Tempero, D. A. Tuveson, R. H. Hruban, J. P. Neoptolemos, Pancreatic cancer. Nat. Rev. Dis. Primers 2, 16022 (2016). [DOI] [PubMed] [Google Scholar]
2.G. Bianchini, C. De Angelis, L. Licata, L. Gianni, Treatment landscape of triple-negative breast cancer—Expanded options, evolving needs. Nat. Rev. Clin. Oncol. 2, 91–113 (2022). [DOI] [PubMed] [Google Scholar]
3.R. K. Orosco, V. J. Tapia, J. A. Califano, B. Clary, E. E. W. Cohen, C. Kane, S. M. Lippman, K. Messer, A. Molinolo, J. D. Murphy, J. Pang, A. Sacco, K. R. Tringale, A. Wallace, Q. T. Nguyen, Positive surgical margins in the 10 most common solid cancers. Sci. Rep. 8, 5686 (2018). [DOI] [PMC free article] [PubMed] [Google Scholar]
4.J. L. Gnerlich, S. R. Luka, A. D. Deshpande, B. J. Dubray, J. S. Weir, D. H. Carpenter, E. M. Brunt, S. M. Strasberg, W. G. Hawkins, D. C. Linehan, Microscopic margins and patterns of treatment failure in resected pancreatic adenocarcinoma. Arch. Surg. 147, 753–760 (2012). [DOI] [PubMed] [Google Scholar]
5.C. H. June, M. Sadelain, Chimeric antigen receptor therapy. N. Engl. J. Med. 379, 64–73 (2018). [DOI] [PMC free article] [PubMed] [Google Scholar]
6.J. Tchou, Y. Zhao, B. L. Levine, P. J. Zhang, M. M. Davis, J. J. Melenhorst, I. Kulikovskaya, A. L. Brennan, X. Liu, S. F. Lacey, A. D. Posey Jr., A. D. Williams, A. So, J. R. Conejo-Garcia, G. Plesa, R. M. Young, S. McGettigan, J. Campbell, R. H. Pierce, J. M. Matro, A. M. DeMichele, A. S. Clark, L. J. Cooper, L. M. Schuchter, R. H. Vonderheide, C. H. June, Safety and efficacy of intratumoral injections of chimeric antigen receptor (CAR) T cells in metastatic breast cancer. Cancer Immunol. Res. 5, 1152–1161 (2017). [DOI] [PMC free article] [PubMed] [Google Scholar]
7.N. A. Vitanza, A. J. Johnson, A. L. Wilson, C. Brown, J. K. Yokoyama, A. Künkele, C. A. Chang, S. Rawlings-Rhea, W. Huang, K. Seidel, C. M. Albert, N. Pinto, J. Gust, L. S. Finn, J. G. Ojemann, J. Wright, R. J. Orentas, M. Baldwin, R. A. Gardner, M. C. Jensen, J. R. Park, Locoregional infusion of HER2-specific CAR T cells in children and young adults with recurrent or refractory CNS tumors: An interim analysis. Nat. Med. 27, 1544–1552 (2021). [DOI] [PubMed] [Google Scholar]
8.W. D. Spotnitz, Fibrin sealant: The only approved hemostat, sealant, and adhesive-a laboratory and clinical perspective. ISRN Surg. 2014, 2014, 203943 [DOI] [PMC free article] [PubMed] [Google Scholar]
9.K. L. Christman, A. J. Vardanian, Q. Fang, R. E. Sievers, H. H. Fok, R. J. Lee, Injectable fibrin scaffold improves cell transplant survival, reduces infarct expansion, and induces neovasculature formation in ischemic myocardium. J. Am. Coll. Cardiol. 44, 654–660 (2004). [DOI] [PubMed] [Google Scholar]
10.V. Falanga, S. Iwamoto, M. Chartier, T. Yufit, J. Butmarc, N. Kouttab, D. Shrayer, P. Carson, Autologous bone marrow-derived cultured mesenchymal stem cells delivered in a fibrin spray accelerate healing in murine and human cutaneous wounds. Tissue Eng. 13, 1299–1312 (2007). [DOI] [PubMed] [Google Scholar]
11.X. Wu, J. Ren, J. Li, Fibrin glue as the cell-delivery vehicle for mesenchymal stromal cells in regenerative medicine. Cytotherapy 14, 555–562 (2012). [DOI] [PubMed] [Google Scholar]
12.K. Kobayashi, Y. Ichihara, N. Tano, L. Fields, N. Murugesu, T. Ito, C. Ikebe, F. Lewis, K. Yashiro, Y. Shintani, R. Uppal, K. Suzuki, Fibrin glue-aided, instant epicardial placement enhances the efficacy of mesenchymal stromal cell-based therapy for heart failure. Sci. Rep. 8, 9448 (2018). [DOI] [PMC free article] [PubMed] [Google Scholar]
13.Z. Zou, E. Denny, C. E. Brown, M. C. Jensen, G. Li, T. Fujii, J. Neman, R. Jandial, M. Chen, Cytotoxic T lymphocyte trafficking and survival in an augmented fibrin matrix carrier. PLOS ONE 7, e34652 (2012). [DOI] [PMC free article] [PubMed] [Google Scholar]
14.E. A. Ogunnaike, A. Valdivia, M. Yazdimamaghani, E. Leon, S. Nandi, H. Hudson, H. Du, S. Khagi, Z. Gu, B. Savoldo, F. S. Ligler, S. Hingtgen, G. Dotti, Fibrin gel enhances the antitumor effects of chimeric antigen receptor T cells in glioblastoma. Sci. Adv. 7, eabg5841 (2021). [DOI] [PMC free article] [PubMed] [Google Scholar]
15.R. A. Morgan, J. C. Yang, M. Kitano, M. E. Dudley, C. M. Laurencot, S. A. Rosenberg, Case report of a serious adverse event following the administration of T cells transduced with a chimeric antigen receptor recognizing ERBB2. Mol. Ther. 18, 843–851 (2010). [DOI] [PMC free article] [PubMed] [Google Scholar]
16.M. R. Parkhurst, J. C. Yang, R. C. Langan, M. E. Dudley, D. A. Nathan, S. A. Feldman, J. L. Davis, R. A. Morgan, M. J. Merino, R. M. Sherry, M. S. Hughes, U. S. Kammula, G. Q. Phan, R. M. Lim, S. A. Wank, N. P. Restifo, P. F. Robbins, C. M. Laurencot, S. A. Rosenberg, T cells targeting carcinoembryonic antigen can mediate regression of metastatic colorectal cancer but induce severe transient colitis. Mol. Ther. 19, 620–626 (2011). [DOI] [PMC free article] [PubMed] [Google Scholar]
17.A. Morello, M. Sadelain, P. S. Adusumilli, Mesothelin-targeted CARs: Driving T cells to solid tumors. Cancer Discov. 6, 133–146 (2016). [DOI] [PMC free article] [PubMed] [Google Scholar]
18.B. Breart, F. Lemaître, S. Celli, P. Bousso, Two-photon imaging of intratumoral CD8⁺ T cell cytotoxic activity during adoptive T cell therapy in mice. J. Clin. Invest. 118, 1390–1397 (2008). [DOI] [PMC free article] [PubMed] [Google Scholar]
19.A. S. Rosenberg, A. Singer, Evidence that the effector mechanism of skin allograft rejection is antigen-specific. Proc. Natl. Acad. Sci. U.S.A. 85, 7739–7742 (1988). [DOI] [PMC free article] [PubMed] [Google Scholar]
20.V. P. Groot, N. Rezaee, W. Wu, J. L. Cameron, E. K. Fishman, R. H. Hruban, M. J. Weiss, L. Zheng, C. L. Wolfgang, J. He, Patterns, timing, and predictors of recurrence following pancreatectomy for pancreatic ductal adenocarcinoma. Ann. Surg. 267, 936–945 (2018). [DOI] [PubMed] [Google Scholar]
21.R. G. Pleijhuis, M. Graafland, J. de Vries, J. Bart, J. S. de Jong, G. M. van Dam, Obtaining adequate surgical margins in breast-conserving therapy for patients with early-stage breast cancer: Current modalities and future directions. Ann. Surg. Oncol. 16, 2717–2730 (2009). [DOI] [PMC free article] [PubMed] [Google Scholar]
22.A. D. Posey Jr., R. D. Schwab, A. C. Boesteanu, C. Steentoft, U. Mandel, B. Engels, J. D. Stone, T. D. Madsen, K. Schreiber, K. M. Haines, A. P. Cogdill, T. J. Chen, D. Song, J. Scholler, D. M. Kranz, M. D. Feldman, R. Young, B. Keith, H. Schreiber, H. Clausen, L. A. Johnson, C. H. June, Engineered CAR T cells targeting the cancer-associated Tn-glycoform of the membrane mucin MUC1 control adenocarcinoma. Immunity 44, 1444–1454 (2016). [DOI] [PMC free article] [PubMed] [Google Scholar]
23.J. Tchou, L.-C. Wang, B. Selven, H. Zhang, J. Conejo-Garcia, H. Borghaei, M. Kalos, R. H. Vondeheide, S. M. Albelda, C. H. June, P. J. Zhang, Mesothelin, a novel immunotherapy target for triple negative breast cancer. Breast Cancer Res. Treat. 133, 799–804 (2012). [DOI] [PMC free article] [PubMed] [Google Scholar]
24.S. Ghassemi, J. S. Durgin, S. Nunez-Cruz, J. Patel, J. Leferovich, M. Pinzone, F. Shen, K. D. Cummins, G. Plesa, V. A. Cantu, S. Reddy, F. D. Bushman, S. I. Gill, U. O’Doherty, R. S. O’Connor, M. C. Milone, Rapid manufacturing of non-activated potent CAR T cells. Nat. Biomed. Eng. 6, 118–128 (2022). [DOI] [PMC free article] [PubMed] [Google Scholar]
25.J. Yang, J. He, X. Zhang, J. Li, Z. Wang, Y. Zhang, L. Qiu, Q. Wu, Z. Sun, X. Ye, W. Yin, W. Cao, L. Shen, M. Sersch, P. Lu, Next-day manufacture of a novel anti-CD19 CAR-T therapy for B-cell acute lymphoblastic leukemia: First-in-human clinical study. Blood Cancer J. 12, 104 (2022). [DOI] [PMC free article] [PubMed] [Google Scholar]
26.S. Depil, P. Duchateau, S. A. Grupp, G. Mufti, L. Poirot, ‘Off-the-shelf’ allogeneic CAR T cells: Development and challenges. Nat. Rev. Drug Discov. 19, 185–199 (2020). [DOI] [PubMed] [Google Scholar]
27.L. R. Johnson, D. Y. Lee, J. S. Eacret, D. Ye, C. H. June, A. J. Minn, The immunostimulatory RNA RN7SL1 enables CAR-T cells to enhance autonomous and endogenous immune function. Cell 184, 4981–4995.e14 (2021). [DOI] [PMC free article] [PubMed] [Google Scholar]
28.N. Nishio, I. Diaconu, H. Liu, V. Cerullo, I. Caruana, V. Hoyos, L. Bouchier-Hayes, B. Savoldo, G. Dotti, Armed oncolytic virus enhances immune functions of chimeric antigen receptor-modified T cells in solid tumors. Cancer Res. 74, 5195–5205 (2014). [DOI] [PMC free article] [PubMed] [Google Scholar]
29.K. Watanabe, Y. Luo, T. Da, S. Guedan, M. Ruella, J. Scholler, B. Keith, R. M. Young, B. Engels, S. Sorsa, M. Siurala, R. Havunen, S. Tahtinen, A. Hemminki, C. H. June, Pancreatic cancer therapy with combined mesothelin-redirected chimeric antigen receptor T cells and cytokine-armed oncolytic adenoviruses. JCI Insight 3, e99573 (2018). [DOI] [PMC free article] [PubMed] [Google Scholar]
30.S. R. Hingorani, L. Wang, A. S. Multani, C. Combs, T. B. Deramaudt, R. H. Hruban, A. K. Rustgi, S. Chang, D. A. Tuveson, Trp53R172H and KrasG12D cooperate to promote chromosomal instability and widely metastatic pancreatic ductal adenocarcinoma in mice. Cancer Cell 7, 469–483 (2005). [DOI] [PubMed] [Google Scholar]
31.C. R. Good, M. A. Aznar, S. Kuramitsu, P. Samareh, S. Agarwal, G. Donahue, K. Ishiyama, N. Wellhausen, A. K. Rennels, Y. Ma, L. Tian, S. Guedan, K. A. Alexander, Z. Zhang, P. C. Rommel, N. Singh, K. M. Glastad, M. W. Richardson, K. Watanabe, J. L. Tanyi, M. H. O’Hara, M. Ruella, S. F. Lacey, E. K. Moon, S. J. Schuster, S. M. Albelda, L. L. Lanier, R. M. Young, S. L. Berger, C. H. June, An NK-like CAR T cell transition in CAR T cell dysfunction. Cell 184, 6081–6100.e26 (2021). [DOI] [PMC free article] [PubMed] [Google Scholar]
32.S. A. Grupp, M. Kalos, D. Barrett, R. Aplenc, D. L. Porter, S. R. Rheingold, D. T. Teachey, A. Chew, B. Hauck, J. F. Wright, M. C. Milone, B. L. Levine, C. H. June, Chimeric antigen receptor-modified T cells for acute lymphoid leukemia. N. Engl. J. Med. 368, 1509–1518 (2013). [DOI] [PMC free article] [PubMed] [Google Scholar]
33.F. X. Rwandamuriye, B. J. Weston, T. G. Johns, W. J. Lesterhuis, R. M. Zemek, A mouse model of incompletely resected soft tissue sarcoma for testing (Neo)adjuvant therapies. J. Vis. Exp., 10.3791/60882 (2020). [DOI] [PubMed] [Google Scholar]

Associated Data

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

Supplementary Materials

Figs. S1 to S10

Click here for additional data file.^{(6.9MB, pdf)}

[R1] 1.J. Kleeff, M. Korc, M. Apte, C. La Vecchia, C. D. Johnson, A. V. Biankin, R. E. Neale, M. Tempero, D. A. Tuveson, R. H. Hruban, J. P. Neoptolemos, Pancreatic cancer. Nat. Rev. Dis. Primers 2, 16022 (2016). [DOI] [PubMed] [Google Scholar]

[R2] 2.G. Bianchini, C. De Angelis, L. Licata, L. Gianni, Treatment landscape of triple-negative breast cancer—Expanded options, evolving needs. Nat. Rev. Clin. Oncol. 2, 91–113 (2022). [DOI] [PubMed] [Google Scholar]

[R3] 3.R. K. Orosco, V. J. Tapia, J. A. Califano, B. Clary, E. E. W. Cohen, C. Kane, S. M. Lippman, K. Messer, A. Molinolo, J. D. Murphy, J. Pang, A. Sacco, K. R. Tringale, A. Wallace, Q. T. Nguyen, Positive surgical margins in the 10 most common solid cancers. Sci. Rep. 8, 5686 (2018). [DOI] [PMC free article] [PubMed] [Google Scholar]

[R4] 4.J. L. Gnerlich, S. R. Luka, A. D. Deshpande, B. J. Dubray, J. S. Weir, D. H. Carpenter, E. M. Brunt, S. M. Strasberg, W. G. Hawkins, D. C. Linehan, Microscopic margins and patterns of treatment failure in resected pancreatic adenocarcinoma. Arch. Surg. 147, 753–760 (2012). [DOI] [PubMed] [Google Scholar]

[R5] 5.C. H. June, M. Sadelain, Chimeric antigen receptor therapy. N. Engl. J. Med. 379, 64–73 (2018). [DOI] [PMC free article] [PubMed] [Google Scholar]

[R6] 6.J. Tchou, Y. Zhao, B. L. Levine, P. J. Zhang, M. M. Davis, J. J. Melenhorst, I. Kulikovskaya, A. L. Brennan, X. Liu, S. F. Lacey, A. D. Posey Jr., A. D. Williams, A. So, J. R. Conejo-Garcia, G. Plesa, R. M. Young, S. McGettigan, J. Campbell, R. H. Pierce, J. M. Matro, A. M. DeMichele, A. S. Clark, L. J. Cooper, L. M. Schuchter, R. H. Vonderheide, C. H. June, Safety and efficacy of intratumoral injections of chimeric antigen receptor (CAR) T cells in metastatic breast cancer. Cancer Immunol. Res. 5, 1152–1161 (2017). [DOI] [PMC free article] [PubMed] [Google Scholar]

[R7] 7.N. A. Vitanza, A. J. Johnson, A. L. Wilson, C. Brown, J. K. Yokoyama, A. Künkele, C. A. Chang, S. Rawlings-Rhea, W. Huang, K. Seidel, C. M. Albert, N. Pinto, J. Gust, L. S. Finn, J. G. Ojemann, J. Wright, R. J. Orentas, M. Baldwin, R. A. Gardner, M. C. Jensen, J. R. Park, Locoregional infusion of HER2-specific CAR T cells in children and young adults with recurrent or refractory CNS tumors: An interim analysis. Nat. Med. 27, 1544–1552 (2021). [DOI] [PubMed] [Google Scholar]

[R8] 8.W. D. Spotnitz, Fibrin sealant: The only approved hemostat, sealant, and adhesive-a laboratory and clinical perspective. ISRN Surg. 2014, 2014, 203943 [DOI] [PMC free article] [PubMed] [Google Scholar]

[R9] 9.K. L. Christman, A. J. Vardanian, Q. Fang, R. E. Sievers, H. H. Fok, R. J. Lee, Injectable fibrin scaffold improves cell transplant survival, reduces infarct expansion, and induces neovasculature formation in ischemic myocardium. J. Am. Coll. Cardiol. 44, 654–660 (2004). [DOI] [PubMed] [Google Scholar]

[R10] 10.V. Falanga, S. Iwamoto, M. Chartier, T. Yufit, J. Butmarc, N. Kouttab, D. Shrayer, P. Carson, Autologous bone marrow-derived cultured mesenchymal stem cells delivered in a fibrin spray accelerate healing in murine and human cutaneous wounds. Tissue Eng. 13, 1299–1312 (2007). [DOI] [PubMed] [Google Scholar]

[R11] 11.X. Wu, J. Ren, J. Li, Fibrin glue as the cell-delivery vehicle for mesenchymal stromal cells in regenerative medicine. Cytotherapy 14, 555–562 (2012). [DOI] [PubMed] [Google Scholar]

[R12] 12.K. Kobayashi, Y. Ichihara, N. Tano, L. Fields, N. Murugesu, T. Ito, C. Ikebe, F. Lewis, K. Yashiro, Y. Shintani, R. Uppal, K. Suzuki, Fibrin glue-aided, instant epicardial placement enhances the efficacy of mesenchymal stromal cell-based therapy for heart failure. Sci. Rep. 8, 9448 (2018). [DOI] [PMC free article] [PubMed] [Google Scholar]

[R13] 13.Z. Zou, E. Denny, C. E. Brown, M. C. Jensen, G. Li, T. Fujii, J. Neman, R. Jandial, M. Chen, Cytotoxic T lymphocyte trafficking and survival in an augmented fibrin matrix carrier. PLOS ONE 7, e34652 (2012). [DOI] [PMC free article] [PubMed] [Google Scholar]

[R14] 14.E. A. Ogunnaike, A. Valdivia, M. Yazdimamaghani, E. Leon, S. Nandi, H. Hudson, H. Du, S. Khagi, Z. Gu, B. Savoldo, F. S. Ligler, S. Hingtgen, G. Dotti, Fibrin gel enhances the antitumor effects of chimeric antigen receptor T cells in glioblastoma. Sci. Adv. 7, eabg5841 (2021). [DOI] [PMC free article] [PubMed] [Google Scholar]

[R15] 15.R. A. Morgan, J. C. Yang, M. Kitano, M. E. Dudley, C. M. Laurencot, S. A. Rosenberg, Case report of a serious adverse event following the administration of T cells transduced with a chimeric antigen receptor recognizing ERBB2. Mol. Ther. 18, 843–851 (2010). [DOI] [PMC free article] [PubMed] [Google Scholar]

[R16] 16.M. R. Parkhurst, J. C. Yang, R. C. Langan, M. E. Dudley, D. A. Nathan, S. A. Feldman, J. L. Davis, R. A. Morgan, M. J. Merino, R. M. Sherry, M. S. Hughes, U. S. Kammula, G. Q. Phan, R. M. Lim, S. A. Wank, N. P. Restifo, P. F. Robbins, C. M. Laurencot, S. A. Rosenberg, T cells targeting carcinoembryonic antigen can mediate regression of metastatic colorectal cancer but induce severe transient colitis. Mol. Ther. 19, 620–626 (2011). [DOI] [PMC free article] [PubMed] [Google Scholar]

[R17] 17.A. Morello, M. Sadelain, P. S. Adusumilli, Mesothelin-targeted CARs: Driving T cells to solid tumors. Cancer Discov. 6, 133–146 (2016). [DOI] [PMC free article] [PubMed] [Google Scholar]

[R18] 18.B. Breart, F. Lemaître, S. Celli, P. Bousso, Two-photon imaging of intratumoral CD8⁺ T cell cytotoxic activity during adoptive T cell therapy in mice. J. Clin. Invest. 118, 1390–1397 (2008). [DOI] [PMC free article] [PubMed] [Google Scholar]

[R19] 19.A. S. Rosenberg, A. Singer, Evidence that the effector mechanism of skin allograft rejection is antigen-specific. Proc. Natl. Acad. Sci. U.S.A. 85, 7739–7742 (1988). [DOI] [PMC free article] [PubMed] [Google Scholar]

[R20] 20.V. P. Groot, N. Rezaee, W. Wu, J. L. Cameron, E. K. Fishman, R. H. Hruban, M. J. Weiss, L. Zheng, C. L. Wolfgang, J. He, Patterns, timing, and predictors of recurrence following pancreatectomy for pancreatic ductal adenocarcinoma. Ann. Surg. 267, 936–945 (2018). [DOI] [PubMed] [Google Scholar]

[R21] 21.R. G. Pleijhuis, M. Graafland, J. de Vries, J. Bart, J. S. de Jong, G. M. van Dam, Obtaining adequate surgical margins in breast-conserving therapy for patients with early-stage breast cancer: Current modalities and future directions. Ann. Surg. Oncol. 16, 2717–2730 (2009). [DOI] [PMC free article] [PubMed] [Google Scholar]

[R22] 22.A. D. Posey Jr., R. D. Schwab, A. C. Boesteanu, C. Steentoft, U. Mandel, B. Engels, J. D. Stone, T. D. Madsen, K. Schreiber, K. M. Haines, A. P. Cogdill, T. J. Chen, D. Song, J. Scholler, D. M. Kranz, M. D. Feldman, R. Young, B. Keith, H. Schreiber, H. Clausen, L. A. Johnson, C. H. June, Engineered CAR T cells targeting the cancer-associated Tn-glycoform of the membrane mucin MUC1 control adenocarcinoma. Immunity 44, 1444–1454 (2016). [DOI] [PMC free article] [PubMed] [Google Scholar]

[R23] 23.J. Tchou, L.-C. Wang, B. Selven, H. Zhang, J. Conejo-Garcia, H. Borghaei, M. Kalos, R. H. Vondeheide, S. M. Albelda, C. H. June, P. J. Zhang, Mesothelin, a novel immunotherapy target for triple negative breast cancer. Breast Cancer Res. Treat. 133, 799–804 (2012). [DOI] [PMC free article] [PubMed] [Google Scholar]

[R24] 24.S. Ghassemi, J. S. Durgin, S. Nunez-Cruz, J. Patel, J. Leferovich, M. Pinzone, F. Shen, K. D. Cummins, G. Plesa, V. A. Cantu, S. Reddy, F. D. Bushman, S. I. Gill, U. O’Doherty, R. S. O’Connor, M. C. Milone, Rapid manufacturing of non-activated potent CAR T cells. Nat. Biomed. Eng. 6, 118–128 (2022). [DOI] [PMC free article] [PubMed] [Google Scholar]

[R25] 25.J. Yang, J. He, X. Zhang, J. Li, Z. Wang, Y. Zhang, L. Qiu, Q. Wu, Z. Sun, X. Ye, W. Yin, W. Cao, L. Shen, M. Sersch, P. Lu, Next-day manufacture of a novel anti-CD19 CAR-T therapy for B-cell acute lymphoblastic leukemia: First-in-human clinical study. Blood Cancer J. 12, 104 (2022). [DOI] [PMC free article] [PubMed] [Google Scholar]

[R26] 26.S. Depil, P. Duchateau, S. A. Grupp, G. Mufti, L. Poirot, ‘Off-the-shelf’ allogeneic CAR T cells: Development and challenges. Nat. Rev. Drug Discov. 19, 185–199 (2020). [DOI] [PubMed] [Google Scholar]

[R27] 27.L. R. Johnson, D. Y. Lee, J. S. Eacret, D. Ye, C. H. June, A. J. Minn, The immunostimulatory RNA RN7SL1 enables CAR-T cells to enhance autonomous and endogenous immune function. Cell 184, 4981–4995.e14 (2021). [DOI] [PMC free article] [PubMed] [Google Scholar]

[R28] 28.N. Nishio, I. Diaconu, H. Liu, V. Cerullo, I. Caruana, V. Hoyos, L. Bouchier-Hayes, B. Savoldo, G. Dotti, Armed oncolytic virus enhances immune functions of chimeric antigen receptor-modified T cells in solid tumors. Cancer Res. 74, 5195–5205 (2014). [DOI] [PMC free article] [PubMed] [Google Scholar]

[R29] 29.K. Watanabe, Y. Luo, T. Da, S. Guedan, M. Ruella, J. Scholler, B. Keith, R. M. Young, B. Engels, S. Sorsa, M. Siurala, R. Havunen, S. Tahtinen, A. Hemminki, C. H. June, Pancreatic cancer therapy with combined mesothelin-redirected chimeric antigen receptor T cells and cytokine-armed oncolytic adenoviruses. JCI Insight 3, e99573 (2018). [DOI] [PMC free article] [PubMed] [Google Scholar]

[R30] 30.S. R. Hingorani, L. Wang, A. S. Multani, C. Combs, T. B. Deramaudt, R. H. Hruban, A. K. Rustgi, S. Chang, D. A. Tuveson, Trp53R172H and KrasG12D cooperate to promote chromosomal instability and widely metastatic pancreatic ductal adenocarcinoma in mice. Cancer Cell 7, 469–483 (2005). [DOI] [PubMed] [Google Scholar]

[R31] 31.C. R. Good, M. A. Aznar, S. Kuramitsu, P. Samareh, S. Agarwal, G. Donahue, K. Ishiyama, N. Wellhausen, A. K. Rennels, Y. Ma, L. Tian, S. Guedan, K. A. Alexander, Z. Zhang, P. C. Rommel, N. Singh, K. M. Glastad, M. W. Richardson, K. Watanabe, J. L. Tanyi, M. H. O’Hara, M. Ruella, S. F. Lacey, E. K. Moon, S. J. Schuster, S. M. Albelda, L. L. Lanier, R. M. Young, S. L. Berger, C. H. June, An NK-like CAR T cell transition in CAR T cell dysfunction. Cell 184, 6081–6100.e26 (2021). [DOI] [PMC free article] [PubMed] [Google Scholar]

[R32] 32.S. A. Grupp, M. Kalos, D. Barrett, R. Aplenc, D. L. Porter, S. R. Rheingold, D. T. Teachey, A. Chew, B. Hauck, J. F. Wright, M. C. Milone, B. L. Levine, C. H. June, Chimeric antigen receptor-modified T cells for acute lymphoid leukemia. N. Engl. J. Med. 368, 1509–1518 (2013). [DOI] [PMC free article] [PubMed] [Google Scholar]

[R33] 33.F. X. Rwandamuriye, B. J. Weston, T. G. Johns, W. J. Lesterhuis, R. M. Zemek, A mouse model of incompletely resected soft tissue sarcoma for testing (Neo)adjuvant therapies. J. Vis. Exp., 10.3791/60882 (2020). [DOI] [PubMed] [Google Scholar]

PERMALINK

Chimeric antigen receptor T cells as adjuvant therapy for unresectable adenocarcinoma

Ugur Uslu

Tong Da

Charles-Antoine Assenmacher

John Scholler

Regina M Young

Julia Tchou

Carl H June

Roles

Abstract

INTRODUCTION

RESULTS

CAR T cell antitumor efficacy depends on the fibrin gel dose

Fibrin gel alone does not affect tumor growth and overall survival

Local CAR T cells in fibrin gel clear residual tumor cells in a TNBC xenograft model

Fig. 1. CAR T cells in fibrin gel clear residual tumor cells following incomplete surgery in a TNBC xenograft model.

Residual tumor is also cleared by local CAR T cells in a PDA xenograft model

Fig. 2. CAR T cells in fibrin gel clear residual tumor cells following incomplete surgery in a PDA xenograft model.

Reduction of on-target off-tumor toxicity with local CAR T cell treatment

Fig. 3. Incidence of on-target off-tumor toxicity after local CAR T cells in fibrin gel.

No or low T cell infiltration into lungs is seen in locally treated mice

Fig. 4. Reduction of on-target off-tumor toxicity in locally treated mice.

No impact on wound healing is observed in locally treated immunocompetent mice

Fig. 5. CAR T cells in fibrin gel does not cause wound healing complications in immunocompetent C57BL/6 mice.

DISCUSSION

MATERIALS AND METHODS

Study design

Human tumor cells

Murine tumor cells

Human CAR T cells

Murine CAR T cells

Preparation of fibrin gel components

Preparation of CAR T cells in fibrin gel

In vitro passive migration and viability of fibrin gel–embedded CAR T cells

In vitro cytokine secretion assay

In vitro cytotoxicity assay using Incucyte SX5

In vitro luciferase assays

Mice

NSG mouse experiments

huMeso-KI-NSG mouse experiments (in vivo toxicity model)

Syngeneic mouse experiments

Incomplete surgical excision of subcutaneous tumors

Caliper measurements of subcutaneous tumors

Peripheral blood stain for T cells

Cytokine analysis from mouse serum

Standard pathological and immunohistochemistry (IHC) analyses of lung and skin tissue

Flow cytometry and antibodies

Statistical analysis

Acknowledgments

Supplementary Materials

This PDF file includes:

REFERENCES AND NOTES

Associated Data

Supplementary Materials

ACTIONS

PERMALINK

RESOURCES

Similar articles

Cited by other articles

Links to NCBI Databases