Metastatic spread of colorectal cancer (CRC) to the peritoneal cavity is common and difficult to treat, with many patients dying from malignant bowel obstruction. Chimeric antigen receptor T cell (CAR-T) immunotherapy has shown great promise, and we previously reported murine and phase I clinical studies on regional intrahepatic CAR-T infusion for CRC liver metastases. We are now studying intraperitoneal (IP) delivery of CAR-Ts for peritoneal carcinomatosis. Regional IP infusion of CAR-T resulted in superior protection against CEA+ peritoneal tumors, when compared to systemically infused CAR-Ts. IP CAR-Ts also provided prolonged protection against IP tumor re-challenges and demonstrated an increase in effector memory phenotype over time. IP CAR-Ts provided protection against tumor growth at distant subcutaneous (SC) sites in association with increases in serum IFNγ levels. Given the challenges posed by immunoinhibitory pathways in solid tumors, we combined IP CAR-T treatment with suppressor cell targeting. High frequencies of myeloid-derived suppressor cells (MDSC) and regulatory T cells (Treg) were found within the IP tumors, with MDSC expressing high levels of immunosuppressive PD-L1. Combinatorial IP CAR-T treatment with depleting antibodies against MDSC and Treg further improved efficacy against peritoneal metastases. Our data support further development of combinatorial IP CAR-T immunotherapy for peritoneal malignancies.

Assumptions that liver immune cells and immunosuppressive pathways are similar to their counterparts in other spaces have led to gaps in our understanding of intrahepatic neoplasm aggressiveness. Myeloid-derived suppressor cells (MDSCs) are potent inhibitors of antitumor immunity and pose a major obstacle to solid tumor treatment. Liver MDSCs (L-MDSCs) associated with liver metastases (LM) are particularly problematic by contributing to intrahepatic immunosuppression that promotes tumor progression. L-MDSCs have been reported to expand in response to granulocyte-macrophages colony-stimulating factor (GM-CSF) and suppress antitumor immunity in LM. To extend these findings, we examined mechanisms of intrahepatic immunosuppression exploited by L-MDSCs. We found that the majority of L-MDSCs co-expressed GM-CSF receptor (GM-CSF-R), indoleamine 2,3-dioxygenase (IDO) and programmed death ligand 1 (PD-L1), while demonstrating high levels of signal transducer and activator of transcription factor 3 (STAT3) activation. GM-CSF-secreting tumor cells induced STAT3 phosphorylation in L-MDSCs in addition to expression of IDO and PD-L1. GM-CSF or GM-CSF-R blockade markedly reduced L-MDSC IDO and PD-L1 expression, implicating tumor-derived GM-CSF in supporting L-MDSC-immunoinhibitory molecule expression. Small-molecule inhibitors of Janus-activated kinase 2 (JAK2) and STAT3 also dramatically diminished IDO and PD-L1 expression in L-MDSCs. We determined that STAT3 exerts transcriptional control over L-MDSC IDO and PD-L1 expression by binding to the IDO1 and PD-L1 promoters. Our data suggest that the GM-CSF/JAK2/STAT3 axis in L-MDSCs drives immunosuppression in a model of LM and blockade of this pathway may enable rescue of intrahepatic antitumor immunity.

Immunosuppressive myeloid-derived suppressor cells (MDSC) subvert antitumor immunity and limit the efficacy of chimeric antigen receptor T cells (CAR-T). Previously, we reported that the GM-CSF/JAK2/STAT3 axis drives liver-associated MDSC (L-MDSC) proliferation and blockade of this axis rescued antitumor immunity. We extended these findings in our murine liver metastasis (LM) model, by treating tumor-bearing mice with STAT3 inhibitors (STATTIC or BBI608) to further our understanding of how STAT3 drives L-MDSC suppressive function. STAT3 inhibition caused significant reduction of tumor burden as well as L-MDSC frequencies due to decrease in pSTAT3 levels. L-MDSC isolated from STATTIC or BBI608-treated mice had significantly reduced suppressive function. STAT3 inhibition of L-MDSC was associated with enhanced antitumor activity of CAR-T. Further investigation demonstrated activation of apoptotic signaling pathways in L-MDSC following STAT3 inhibition as evidenced by an upregulation of the pro-apoptotic proteins Bax, cleaved caspase-3, and downregulation of the anti-apoptotic protein Bcl-2. Accordingly, there was also a decrease of pro-survival markers, pErk and pAkt, and an increase in pro-death marker, Fas, with activation of downstream JNK and p38 MAPK. These findings represent a previously unrecognized link between STAT3 inhibition and Fas-induced apoptosis of MDSCs. Our findings suggest that inhibiting STAT3 has potential clinical application for enhancing the efficacy of CAR-T cells in LM through modulation of L-MDSC.

No abstract

Chimeric antigen receptor expressing T cells (CAR-T) are a promising form of immunotherapy, but the influence of age-related immune changes on CAR-T production remains poorly understood. We showed that CAR-T cells from geriatric donors (gCAR-T) are functionally impaired relative to CAR-T from younger donors (yCAR-T). Higher transduction efficiencies and improved cell expansion were observed in yCAR-T cells compared with gCAR-T. yCAR-T demonstrated significantly increased levels of proliferation and signaling activation of phosphorylated (p)Erk, pAkt, pStat3, and pStat5. Furthermore, yCAR-T contained higher proportions of CD4 and CD8 effector memory (EM) cells, which are known to have enhanced cytolytic capabilities. Accordingly, yCAR-T demonstrated higher levels of tumor antigen-specific cytotoxicity compared with gCAR-T. Enhanced tumor killing by yCAR-T correlated with increased levels of perforin and granzyme B. yCAR-T had increased α5β1 integrin expression, a known mediator of retroviral transduction. We found that treatment with M-CSF or TGF-β1 rescued the impaired transduction efficiency of the gCAR-T by increasing the α5β1 integrin expression. Neutralization of α5β1 confirmed that this integrin was indispensable for CAR expression. Our study suggests that the increase of α5β1 integrin expression levels enhances CAR expression and thereby improves tumor killing by gCAR-T.

OBJECTIVES: There are no effective treatment options for patients with unresectable CEA+ liver metastases (LM) from gastrointestinal adenocarcinoma refractory to conventional systemic therapy. In the previous Hepatic Immunotherapy for Metastases (HITM) phase I study we demonstrated the safety and biologic activity of anti-CEA CAR-T cell hepatic artery infusions (HAI). HITM-SIR was a single arm phase Ib trial testing anti-CEA CAR-T HAI followed by selective internal radiation therapy (SIRT) in patients with refractory CEA+ LM. METHODOLOGY: We enrolled 8 patients with unresectable, chemotherapy refractory CEA+ LM and 6/8 completed the study. Two patients were withdrawn for disease progression prior to CAR-T infusion and biliary obstruction due to centrally located disease. Limited extrahepatic disease (EHD) was permitted. Subjects received 3 HAI of anti-CEA CAR-T cells (1e10 cells per dose) along with low dose continuous IL-2 infusions (50,000 IU/kg/day). SIRT was administered in standard fashion 2 weeks following the 3rd CAR-T HAI. Primary objective was to establish safety of the CAR-T/SIRT combination. Secondary objectives included response assessed by modified RECIST (mRECIST), immune-related response criteria (irRC), and tumor marker kinetics. RESULTS: The mean age for enrolled subjects was 54.6 years (39-61) with 3 women and 5 men. Histologies (completed patients): 2 colon, 2 rectal, and 2 pancreas. This heavily pre-treated, advanced disease group of patients received an average of 2.3 lines of prior chemotherapy, 3/6 had >10 LM, and the average largest LM size was 7.3 cm. The average transduction efficiency as measured by CAR expression was 60.4%, with 90.9% viability, and an average production time for >3e10 cells of 13.8 days. There were no grade (G) 4/5 events related to the CAR-T, SIRT, or combination. Toxicities included G 1/2 liver function test elevations (n=5/6), fever (n=5/6), hypereosinophilia (n=2/6), and edema (n=2/6). G3 toxicities included colitis (n=2/6), fever (n=2/6), and edema (n=2/6). One patient experienced a hypertensive crisis during a single CAR-T infusion but tolerated 2 infusions without incident. All colitis episodes resolved with IL-2 dose reductions. Post-SIRT serum CEA decreases were noted in 2/6 patients (-40% and -71%) with CA19-9 decreases in 2/5 detectable patients (-31% and -32%). When considering all on-study time points, 5/6 patients had CEA responses (mean decrease 59.7%) and 4/5 patients expressing CA 19-9 decreases (mean 59.6%). At completion of the study, 3/6 patients had stable disease (SD) in the liver by mRECIST and irRC, and 3/6 SD overall by irRC. Target liver lesion decreases in size among patients with SD ranged from 6-28%. Regression of extrahepatic tumors or abscopal effects were noted in 2/6 at lung and bone sites. One patient remains without evidence of viable liver disease at 10.8 months follow-up. The median overall survival time for all patients is 6.9 months (range 3.8-10.8+). CONCLUSIONS: Following this phase Ib trial, the recommended phase 2 dose for anti-CEA CAR-T HITM infusions is 1010 cells with or without SIRT. The favorable safety profile and evidence of biologic activity indicate that CAR-T HITM infusions should be further studied in a phase 2 trial. Clinical trial information: NCT02416466. Citation Format: Steven C. Katz, Ethan Prince, Marissa Cunetta, Prajna Guha, Ashley Moody, Vincent Armenio, Li J. Wang, N. Joseph Espat, Richard P. Junghans. HITM-SIR: Phase Ib trial of CAR-T hepatic artery infusions and selective internal radiation therapy for liver metastases [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2017; 2017 Apr 1-5; Washington, DC. Philadelphia (PA): AACR; Cancer Res 2017;77(13 Suppl):Abstract nr CT109. doi:10.1158/1538-7445.AM2017-CT109

Myeloid derived suppressor cells (MDSC) subvert anti-tumor immunity. Previously we reported that the GM-CSF/ JAK2/STAT3 axis drives liver MDSC (L-MDSC) proliferation and CAR-T suppression. We hypothesized that STAT3 supports L-MDSC survival and suppressive function by inhibiting apoptosis. We treated liver metastasis (LM) in mice with STAT3 inhibitors (STATTIC or BBI) or with vehicle control. STAT3 inhibition caused a significant reduction in tumor burden (p<0.05) and L-MDSC frequency (DMSO 41±3% vs. STATTIC 29±3%/BBI 20±3%, p<0.0001, n=10) in association with lower pSTAT3 levels (DMSO 33±4% vs. STATTIC 11±3%/BBI 9±2%, p<0.0001, n=10). L-MDSC isolated from STATTIC or BBI treated mice were co-cultured with CAR-Ts and corresponding target tumor cells at 1:1:1 ratio. There was a significant decrease in tumor cell density (DMSO 100% vs. STATTIC 71±5%/BBI 20±3%, p<0.05, n=5) and enhancement of tumor cell killing (DMSO 28±4% vs. STATTIC 54±4%/BBI 52±7%, p<0.01, n=5). Rescue of CAR-T tumor killing function correlated with enhanced L-MDSC apoptosis signaling. We detected upregulation of pro-apoptotic proteins Bax, caspase 3, and Fas. Signaling molecules downstream of Fas, JNK and p38 MAPK (p<0.05), were also activated in L-MDSC. In contrast, L-MDSC pro-survival Bcl2, pErk, and pAkt (p<0.05) were downregulated in response to STAT3 inhibition. Microarray results confirmed the STAT3-induced changes in apoptotic and survival gene expression, which was validated by RT-PCR (p<0.05). Within LM, STAT3 inhibition drove L-MDSC apoptosis via the Fas/Fas- L pathway with downstream pro-apoptotic signaling through p38 MAPK. Blocking STAT3 may have clinical application for enhancing immunotherapy for LM.

Myeloid suppressor cells (MSC) are a tolerogenic population of immune cells that poses a major obstacle to cancer treatment. Intrahepatic MSC associated with liver metastases (LM) are especially detrimental by enhancing immunosuppression. We have recently found that MSC in a mouse model of LM suppress endogenous T cell function and anti-tumor cell-based immunotherapy. Mice with LM induced by a splenic injection of MC38 tumor cells, showed an expansion of MSC (5-fold, p=0.0005), while GM-CSF blockade reduced MSC frequency and increased the overall survival (6-fold, p=0.03). As MC38 and other tumors overexpress GM-CSF, we wanted to further investigate the importance of GM-CSF in MSC-driven immunosuppression. A significant proportion of MSC in LM co-expressed IDO and PD-L1 (34±5%), while maintaining high levels of pSTAT3, a transcription factor important to MSC function. Ex vivo, MSC in culture with MC38 cells or supplemented with GM-CSF maintained high IDO/PD-L1 expression. Conversely, when MSC were co-cultured with tumor cells supplemented with anti-GM-CSF, IDO/PD-L1 levels were significantly reduced (p=0.03), implicating GM-CSF in promoting IDO/PD-L1 expression in MSC. Small molecule STAT3 and JAK2 inhibitors abolished IDO/PD-L1 expression in MSC (p=0.003 and p=0.002), suggesting the importance of STAT3/JAK2 signaling pathway in IDO/PD-L1 expression. STAT3 and JAK2 pathways represent viable targets for blocking GM-CSF induced immunosuppressive pathways in liver MSC.