Search

Inhibition of immune checkpoints, including cytotoxic T-lymphocyte antigen-4 (CTLA-4), programmed death-1 (PD-1), and its ligand PD-L1, has demonstrated exciting and durable remissions across a spectrum of malignancies. Combinatorial regimens blocking complementary immune checkpoints further enhance the therapeutic benefit. The activity of these agents for patients with glioblastoma, a generally lethal primary brain tumor associated with significant systemic and microenvironmental immunosuppression, is not known. We therefore systematically evaluated the antitumor efficacy of murine antibodies targeting a broad panel of immune checkpoint molecules, including CTLA-4, PD-1, PD-L1, and PD-L2 when administered as single-agent therapy and in combinatorial regimens against an orthotopic, immunocompetent murine glioblastoma model. In these experiments, we observed long-term tumor-free survival following single-agent anti–PD-1, anti–PD-L1, or anti–CTLA-4 therapy in 50%, 20%, and 15% of treated animals, respectively. Combination therapy of anti–CTLA-4 plus anti–PD-1 cured 75% of the animals, even against advanced, later-stage tumors. In long-term survivors, tumor growth was not seen upon intracranial tumor rechallenge, suggesting that tumor-specific immune memory responses were generated. Inhibitory immune checkpoint blockade quantitatively increased activated CD8+ and natural killer cells and decreased suppressive immune cells in the tumor microenvironment and draining cervical lymph nodes. Our results support prioritizing the clinical evaluation of PD-1, PD-L1, and CTLA-4 single-agent targeted therapy as well as combination therapy of CTLA-4 plus PD-1 blockade for patients with glioblastoma. Cancer Immunol Res; 4(2); 124–35. ©2015 AACR.

Supplemental Figure 1. Western blotting of PD-L1 mAbs. Supplemental Figure 2. Western blotting of PD-L1 mAbs specific for the cytoplasmic domain. Supplemental Figure 3. 300.19 and stably transfected 300.19 expressing human PD-L1 were stained with 9A11 mAb. Supplemental Figure 4. Isotype control staining. Staining of the same nasopharyngeal carcinoma used in Figure 4 with isotype control antibody at the same concentration as its respective mouse or rabbit PD-L1 mAb.

Supplementary Figure S1. Ipi-Bev treatment altered circulating cytokine/chemokine expression. Supplementary Figure S2. Pre-treatment and post-treatment serum levels of IP-10, TNFa, IL-1a, INFa2, GRO, and IL-8 in Ipi-Bev-treated patients.

Ordinal logistic regression analysis to assess the association of tumor PDCD1LG2 (PD-L2) expression (predictor) with the overall lymphocytic reaction score (outcome)

Supplementary Figure Legend

Supplementary Figure S1. Gal-1 antibodies were identified in the posttreatment plasma samples from melanoma patients on Ipi-Bev and pembrolizumab treatment. Supplementary Figure S2. Gal-1 antibody titers in pretreatment and posttreatment plasma samples from melanoma patients receiving ipilimumab and PD-1 blockade. Supplementary Figure S3. Longitudinal analyses of Gal-1 antibody responses to Ipi-Bev in metastatic melanoma patients. Supplementary Figure S4. Purification of recombinant human Gal-1 with His, biotinylated Avi, and SUMO tags at the N terminus (HAS-Gal-1). Supplementary Figure S5. Enrichment of Gal-1 antibodies from patient plasma. Supplementary Figure S6. Binding of Gal-1 to CD45. Supplementary Figure S7. Immunohistochemistry of Gal-1 expression in pretreatment and posttreatment tumor biopsies of patients on Ipi-Bev therapy. Supplementary Figure S8. Kaplan-Meier survival curves of patients based on pretreatment circulating Gal-1 levels. Supplementary Figure S9. Kaplan-Meier survival curves of patients based on circulating Gal-1 fold changes as a result of ipilimumab treatment.

Esophageal adenocarcinoma is an increasingly common disease with a dismal 5-year survival rate of 10% to 15%. In the first systematic evaluation of the PD-1 pathway in esophageal adenocarcinoma, we identify expression of PD-L2 in cancer cells in 51.7% of esophageal adenocarcinomas. Epithelial PD-L1 was expressed on only 2% of cases, although PD-L1+ immune cells were observed in 18% of esophageal adenocarcinomas. We also evaluated expression in the precursor lesion of esophageal adenocarcinoma, Barrett's esophagus, which emerges following gastric reflux–induced esophageal inflammation, and found PD-L2 expression in Barrett's esophagus but not in non–Barrett's esophagus esophagitis. Because the progression from squamous esophagitis to Barrett's esophagus is accompanied by a transition from a TH1 to TH2 immune response, we hypothesized that the TH2 cytokines IL4/IL13 could contribute to PD-L2 induction. We confirmed that these cytokines can augment PD-L2 expression in esophageal adenocarcinoma cell lines. These results suggest that the inflammatory environment in Barrett's esophagus and esophageal adenocarcinoma may contribute to the expression of PD-L2. Furthermore, the potential for PD-1 receptor blockade to be effective in esophageal adenocarcinomas with epithelial PD-L2 or immune cell PD-L1 expression should be evaluated in clinical trials. Cancer Immunol Res; 3(10); 1123–9. ©2015 AACR.

Genes associated with poor immune infiltration and seromic profiles

PD-L1 immunohistochemical staining does not always predict whether a cancer will respond to treatment with PD-1 inhibitors. We sought to characterize immune cell infiltrates and the expression of T-cell inhibitor markers in PD-L1–positive and PD-L1–negative malignant pleural mesothelioma samples. We developed a method for immune cell phenotyping using flow cytometry on solid tumors that have been dissociated into single-cell suspensions and applied this technique to analyze 43 resected malignant pleural mesothelioma specimens. Compared with PD-L1–negative tumors, PD-L1–positive tumors had significantly more infiltrating CD45+ immune cells, a significantly higher proportion of infiltrating CD3+ T cells, and a significantly higher percentage of CD3+ cells displaying the activated HLA-DR+/CD38+ phenotype. PD-L1–positive tumors also had a significantly higher proportion of proliferating CD8+ T cells, a higher fraction of FOXP3+/CD4+ Tregs, and increased expression of PD-1 and TIM-3 on CD4+ and CD8+ T cells. Double-positive PD-1+/TIM-3+ CD8+ T cells were more commonly found on PD-L1–positive tumors. Compared with epithelioid tumors, sarcomatoid and biphasic mesothelioma samples were significantly more likely to be PD-L1 positive and showed more infiltration with CD3+ T cells and PD-1+/TIM-3+ CD8+ T cells. Immunologic phenotypes in mesothelioma differ based on PD-L1 status and histologic subtype. Successful incorporation of comprehensive immune profiling by flow cytometry into prospective clinical trials could refine our ability to predict which patients will respond to specific immune checkpoint blockade strategies. Cancer Immunol Res; 4(12); 1038–48. ©2016 AACR.

PD-L2 expression in patient derived xenografts.

Supplementary Figure S1. Flow diagram of the study population in the Nurses'' Health Study and the Health Professionals Follow-Up Study. Supplementary Figure S2. Immunohistochemistry of a colorectal carcinoma case using anti-PDCD1LG2 (PD-L2) antibody (A) and its isotype control (B). Supplementary Figure S3. Kaplan-Meier curves for colorectal cancer-specific mortality (A) and overall mortality (B) according to tumor PDCD1LG2 (PD-L2) expression.

Immune checkpoint therapies targeting CTLA-4 and PD-1 have proven effective in cancer treatment. However, the identification of biomarkers for predicting clinical outcomes and mechanisms to overcome resistance remain as critical needs. Angiogenesis is increasingly appreciated as an immune modulator with potential for combinatorial use with checkpoint blockade. Angiopoietin-2 (ANGPT2) is an immune target in patients and is involved in resistance to anti-VEGF treatment with the monoclonal antibody bevacizumab. We investigated the predictive and prognostic value of circulating ANGPT2 in metastatic melanoma patients receiving immune checkpoint therapy. High pretreatment serum ANGPT2 was associated with reduced overall survival in CTLA-4 and PD-1 blockade–treated patients. These treatments also increased serum ANGPT2 in many patients early after treatment initiation, whereas ipilimumab plus bevacizumab treatment decreased serum concentrations. ANGPT2 increases were associated with reduced response and/or overall survival. Ipilimumab increased, and ipilimumab plus bevacizumab decreased, tumor vascular ANGPT2 expression in a subset of patients, which was associated with increased and decreased tumor infiltration by CD68+ and CD163+ macrophages, respectively. In vitro, bevacizumab blocked VEGF-induced ANGPT2 expression in tumor-associated endothelial cells, whereas ANGPT2 increased PD-L1 expression on M2-polarized macrophages. Treatments elicited long-lasting and functional antibody responses to ANGPT2 in a subset of patients receiving clinical benefit. Our findings suggest that serum ANGPT2 may be considered as a predictive and prognostic biomarker for immune checkpoint therapy and may contribute to treatment resistance via increasing proangiogenic and immunosuppressive activities in the tumor microenvironment. Targeting ANGPT2 provides a rational combinatorial approach to improve the efficacy of immune therapy. Cancer Immunol Res; 5(1); 17–28. ©2016 AACR.

Blockade of the pathway including programmed death-ligand 1 (PD-L1) and its receptor programmed cell death protein 1 (PD-1) has produced clinical benefits in patients with a variety of cancers. Elevated levels of soluble PD-L1 (sPD-L1) have been associated with worse prognosis in renal cell carcinoma and multiple myeloma. However, the regulatory roles and function of sPD-L1 particularly in connection with immune checkpoint blockade treatment are not fully understood. We identified four splice variants of PD-L1 in melanoma cells, and all of them are secreted. Secretion of sPD-L1 resulted from alternate splicing activities, cytokine induction, cell stress, cell injury, and cell death in melanoma cells. Pretreatment levels of sPD-L1 were elevated in stage IV melanoma patient sera compared with healthy donors. High pretreatment levels of sPD-L1 were associated with increased likelihood of progressive disease in patients treated by CTLA-4 or PD-1 blockade. Although changes in circulating sPD-L1 early after treatment could not distinguish responders from those with progressive disease, after five months of treatment by CTLA-4 or PD-1 blockade patients who had increased circulating sPD-L1 had greater likelihood of developing a partial response. Induction of sPD-L1 was associated with increased circulating cytokines after CTLA-4 blockade but not following PD-1 blockade. Circulating sPD-L1 is a prognostic biomarker that may predict outcomes for subgroups of patients receiving checkpoint inhibitors. Cancer Immunol Res; 5(6); 480–92. ©2017 AACR.

Supplementary Figure S4. IL-1a, TNFa and IL-1b induce adhesion molecule expression and promote T cell adhesion onto human dermal microvascular endothelial cells (HDMEC). Supplementary Figure S5. VEGF inhibits TNFa-mediated ICAM-1expression in TEC and HUVEC. Supplementary Figure S6. IP-10 induces migration of activated T cells across HDMEC. Supplementary Figure S7. Kaplan-Meier survival curves based on fold-changes of circulating IL-1a, TNFa, IFNa2, IL-8, IP-10, and GRO in response to Ipi-Bev treatment. Supplementary Figure S8. Cytokines/chemokines and clinical outcomes in Ipi-Bev treated patients. Supplementary Figure S9. Ipi-Bev therapy influenced circulating endothelial cells (CEC) in melanoma patients.

Fig. S1. Kaplan-Meier survival curves of patients receiving ipilimumab or ipilimumab plus bevacizumab. Fig. S2. Clinical responses and survival of PD-1 blockade-treated patients based on serum ANGPT2 fold changes. Fig. S3. Hypoxia upregulated ANGPT2 expression in melanoma cells. Fig. S4. ANGPT2 enhanced PD-L1 expression on CSF1-activated MDM from healthy donors. Fig. S5. Antibody responses to ANGPT2 and clinical responses of melanoma patients receiving immune checkpoint therapy. Fig. S6. Enriched endogenous anti-ANGPT2 antibodies inhibited ANGPT2-mediated Erk1/2 phosphorylation.

Culture and treatment of endothelial cells and melanoma cells. Generation and polarization of monocyte derived macrophages (MDM). Purification of ANGPT2 antibodies from plasma. Immunoblot analyses.

The combination of anti-VEGF blockade (bevacizumab) with immune checkpoint anti–CTLA-4 blockade (ipilimumab) in a phase I study showed tumor endothelial activation and immune cell infiltration that were associated with favorable clinical outcomes in patients with metastatic melanoma. To identify potential immune targets responsible for these observations, posttreatment plasma from long-term responding patients were used to screen human protein arrays. We reported that ipilimumab plus bevacizumab therapy elicited humoral immune responses to galectin-1 (Gal-1), which exhibits protumor, proangiogenesis, and immunosuppressive activities in 37.2% of treated patients. Gal-1 antibodies purified from posttreatment plasma suppressed the binding of Gal-1 to CD45, a T-cell surface receptor that transduces apoptotic signals upon binding to extracellular Gal-1. Antibody responses to Gal-1 were found more frequently in the group of patients with therapeutic responses and correlated with improved overall survival. In contrast, another subgroup of treated patients had increased circulating Gal-1 protein instead, and they had reduced overall survival. Our findings suggest that humoral immunity to Gal-1 may contribute to the efficacy of anti-VEGF and anti–CTLA-4 combination therapy. Gal-1 may offer an additional therapeutic target linking anti-angiogenesis and immune checkpoint blockade. Cancer Immunol Res; 5(6); 446–54. ©2017 AACR.

Supplemental Table I. Description and quantitation of PD-L1 IHC in tumor specimens.

Supplementary Methods

Blocking the programmed death-1 (PD-1) pathway has clinical benefit in metastatic cancer and has led to the approval of the mAbs pembrolizumab and nivolumab to treat melanoma and nivolumab for non–small cell lung cancer. Expression of PD-L1 on the cell surface of either tumor cells or infiltrating immune cells is associated with a higher likelihood of response to PD-1 blockade in multiple studies. Most mAbs to PD-L1 in use are directed to its extracellular domain and immunohistochemically stain tumor tissue with a mixture of cytoplasmic and membrane staining. Cytoplasmic staining obscures the interpretation of a positive reaction on the tumor cell membrane, and thus affects the accuracy of PD-L1 scoring systems. We developed a mAb to the cytoplasmic domain of PD-L1, 405.9A11 (9A11), which is both more selective for membranous PD-L1 and more sensitive in IHC and Western blotting, compared with previous mAbs specific for the PD-L1 extracellular domain. Here, we compare immunohistochemical staining patterns of PD-L1 expression in five types of tumors, using five PD-L1 mAbs: 9A11, 7G11, and three commercially available mAbs. We demonstrate that 9A11, as well as two other cytoplasmic domain-specific mAbs, E1L3N and SP142, can clearly delineate the membrane of PD-L1–positive cells in formalin-fixed paraffin-embedded tissue and facilitate interpretation of staining results. Cancer Immunol Res; 3(12); 1308–15. ©2015 AACR.

Supplementary Table S1. Clinical responses and survival of patients with durable clinical benefit. Supplementary Table S2. Time between pre-treatment and post-treatment samples and the fold changes for Gal-1 antibodies. Supplementary Table S3. Distribution of patients with humoral immune responses to Gal-1. Supplementary Table S4. The relationship between response and Gal-1 antibody change. Supplementary Table S5. H-scores of Gal-1 staining in pre-treatment and post-treatment tumor biopsies.

Supplementary Methods

S1. Schematic diagram of specific primers for the detection of PD-L1 variants. S2. Generation of recombinant fusion proteins of PD-L1 variants. S3. Inhibitory effects of membrane-bound PD-L1 on proliferation of antigen specific CD8+ T cells. S4. Effects of cytokines on expression of PD-L1 and cell proliferation in melanoma cell lines. S5. Associations between sPD-L1 secretion and PD-L1 expression in presence or absence of cytokine. S6. Secretion of sPD-L1 and expression of membrane-bound PD-L1 in response to sodium azide. S7. Specific and different recognitions of sPD-L1 variants by anti-PD-L1 antibody clone 29E.12B1 and 130021. S8. A. Schema on recognition regions of anti-PD-L1 antibodies. B. Detection of PD-L1-HIS (Amino acid 19-239) and PD-L1-3/Ig (Amino acid 19-178) by SDS-PAGE and immunoblotting assay with 130021 antibody. S9. Impact of serum or plasma matrices in quantitative detection of sPD-L1. S10. Kinetic changes of sPD-L1all and sPD-L1L in plasma of melanoma patients receiving ipilimumab plus bevacizumab. S11. Kinetic changes of cytokines in plasma of melanoma patients receiving ipilimumab plus bevacizumab. S12. sPD-L1 in plasma of melanoma patients receiving ipilimumab. S13. Kinetic changes of sPD-L1all and sPD-L1L in sera of melanoma patients receiving ipilimumab in ECOG 1608 trial. S14. Kinetic changes of sPD-L1all and sPD-L1L in plasma of melanoma patients receiving anti-PD-1 antibody. S15. Kinetic changes of cytokines in plasma of melanoma patients receiving anti-PD-1 antibody. S16. Associations of pretreatment levels of sPD-L1 with survival after check-point blockade. S17. Associations of long term or delayed increases in sPD-L1 with survival after check-point blockade. S18. Effects of cytokines on secretions of PD-L1 variants in DC. Supplemental Table 1. A. Amino acid regions of sPD-L1 variants. B. Amino acid regions of recombinant PD-L1 proteins. Supplemental Table 2. Associations of pretreatment levels of sPD-L1 with variables.

Adenoid cystic carcinoma (ACC) is among the most lethal salivary gland tumors, with no treatments for metastatic disease that prolong survival. We examined tissue from 28 primary and metastatic ACC deposits obtained from 21 patients for infiltrating immune cells and PD-L1/PD-L2 expression and determined mRNA profiles of over 1,400 oncogenic and immune-related genes. We also assessed the effect of chemoradiation on immune mediators in a patient who had serial biopsies available. Most tumors expressed PD-L2 but had few infiltrating immune cells. Lack of immune-cell infiltrate was associated with expression of genes in the β-catenin/Wnt and PI3K pathways. Additionally, certain transcripts linked to growth and invasion were differentially expressed among primary and metastatic deposits. Chemoradiation appeared to increase CD8+ effector T cells, decrease regulatory T cells, and promote a systemic humoral response. These data suggest a potential role for PD-L2 inhibition and immune modulation as treatment for patients with ACC. Cancer Immunol Res; 4(8); 679–87. ©2016 AACR.

Supplemental Figure 1. Bioluminescence plots of individual animals following treatment with immune checkpoint blocking mAbs. Supplemental Figure 2. Immune cell infiltrates within intracranial GL261-luc2 tumor following treatment with isotype control IgG or combination therapy with CTLA-4 plus PD-1 mAb therapy. Supplemental Figure 3: Treatment antibody and detection antibody for PD-1 do not share same the epitope. Supplemental Figure 4. Representative immunohistochemistry results of intratumoral CD3+, CD4+ and CD8+ T cell infiltrates following immune checkpoint therapy with mAbs against CTLA-4, PD-1 and both CTLA-4 and PD-1. Supplemental Figure 5. Immune cell infiltration within the draining cervical lymph node (cLN) of GL261-luc2 tumor-bearing mice following treatment with isotype control IgG or combination therapy with mAbs against CTLA-4 plus PD-1. Supplemental Figure 6. Comparisons of chemokine expression levels in sera of mice treated with checkpoint blockade. Supplemental Figure 7: Expression of checkpoint ligands and receptors on GL261-Luc cells in vitro. Supplementary Table 1. Arrangement of chemokines on the Mouse Chemokine Antibody Array.

Immune recognition of tumor targets by specific cytotoxic lymphocytes is essential for the effective rejection of tumors. A phase I clinical trial of ipilimumab (an antibody that blocks CTLA-4 function) in combination with bevacizumab (an antibody that inhibits angiogenesis) in patients with metastatic melanoma found favorable clinical outcomes were associated with increased tumor endothelial activation and lymphocyte infiltration. To better understand the underlying mechanisms, we sought features and factors that changed as a function of treatment in patients. Ipilimumab plus bevacizumab (Ipi-Bev) increased tumor vascular expression of ICAM1 and VCAM1. Treatment also altered concentrations of many circulating cytokines and chemokines, including increases of CXCL10, IL1α, TNFα, CXCL1, IFNα2, and IL8, with decreases in VEGF-A in most patients. IL1α and TNFα induced expression of E-selectin, CXCL1, and VCAM1 on melanoma tumor-associated endothelial cells (TEC) in vitro and promoted adhesion of activated T cells onto TEC. VEGFA inhibited TNFα-induced expression of ICAM1 and VCAM1 and T-cell adhesion, which was blocked by bevacizumab. CXCL10 promoted T-cell migration across TEC in vitro, was frequently expressed by melanoma cells, and was upregulated in a subset of tumors in treated patients. Robust upregulation of CXCL10 in tumors was accompanied by increased T-cell infiltration. Ipi-Bev also augmented humoral immune responses recognizing targets in melanoma, tumor endothelial, and tumor mesenchymal stem cells. Our findings suggest that Ipi-Bev therapy augments immune recognition in the tumor microenvironment through enhancing lymphocyte infiltration and antibody responses. IL1α, TNFα, and CXCL10, together with VEGF neutralization, contribute to Ipi-Bev–induced melanoma immune recognition. Cancer Immunol Res; 4(10); 858–68. ©2016 AACR.

Expression of the immune checkpoint ligand CD274 (programmed cell death 1 ligand 1, PD-L1, from gene CD274) contributes to suppression of antitumor T cell–mediated immune response in various tumor types. However, the role of PDCD1LG2 (PD-L2, CD273, from gene PDCD1LG2) in the tumor microenvironment remains unclear. We hypothesized that tumor PDCD1LG2 expression might be inversely associated with lymphocytic reactions to colorectal cancer. We examined tumor PDCD1LG2 expression by IHC in 823 colon and rectal carcinoma cases within two U.S.-nationwide cohort studies and categorized tumors into quartiles according to the percentage of PDCD1LG2–expressing carcinoma cells. We conducted multivariable ordinal logistic regression analysis to assess the associations of tumor PDCD1LG2 expression with Crohn-like lymphoid reaction, peritumoral lymphocytic reaction, intratumoral periglandular reaction, or tumor-infiltrating lymphocytes, controlling for potential confounders, including microsatellite instability, CpG island methylator phenotype, long-interspersed nucleotide element-1 methylation, and KRAS, BRAF, and PIK3CA mutations. Tumor PDCD1LG2 expression was inversely associated with Crohn-like lymphoid reaction (Ptrend = 0.0003). For a unit increase in the three-tiered ordinal categories of Crohn-like lymphoid reaction, a multivariable OR in the highest (vs. lowest) quartile of the percentage of PDCD1LG2–expressing tumor cells was 0.38 (95% confidence interval, 0.22–0.67). Tumor PDCD1LG2 expression was not associated with peritumoral lymphocytic reaction, intratumoral periglandular reaction, tumor-infiltrating lymphocytes, or patient survival (Ptrend > 0.13). Thus, tumor PDCD1LG2 expression is inversely associated with Crohn-like lymphoid reaction to colorectal cancer, suggesting a possible role of PDCD1LG2-expressing tumor cells in inhibiting the development of tertiary lymphoid tissues during colorectal carcinogenesis. Cancer Immunol Res; 5(11); 1046–55. ©2017 AACR.

Table S1. Patient characteristics. Table S2. Patient characteristics by pre-treatment serum ANGPT2 levels. Table S3. Serum ANGPT2 fold-changes by immune checkpoint therapy. Table S4. ANGPT2, CD68 and CD163 staining of melanoma tissues.

Isolation and culture of tumor associated endothelial cells and mesenchymal stem cells. T cell adhesion assay and transendothelial migration assay. Phenotyping and enumeration of circulating endothelial cells (CEC) in whole blood sample.

Supplementary Figure Legends

Supplemental Figure 1. Difference in CD3+ immune cell infiltrates by histologic subtype of mesothelioma. Supplemental Figure 2. Comparison of immune cell infiltrates among patients who received or did not receive neoadjuvant chemotherapy prior to surgical resection of their tumor. Supplemental Figure 3. Quantification of monocyte subtypes in tumor samples. Supplemental Figure 4. Comparison of immune cell infiltrates by mutational subtype in mesothelioma samples. Supplemental Figure 5. Statistical analysis of t-SNE maps from Figure 6. Supplemental Figure 6. Kaplan-Meier overall survival analysis for patients according to histologic subtype, PD-L1 immunohistochemical status, and CD3 infiltration as assessed by either flow cytometry or immunohistochemistry.

Supplementary Table S1. H-scores of ICAM-1, VCAM-1 and IP-10 expression in tumors of melanoma patients before and after ipilimumab or Ipi-Bev treatment. Supplementary Table S2. CXCR3 expression on circulating CD4+ and CD8+ T cells before and after Ipi-Bev treatment. Supplementary Table S3. Pre-treatment serum levels of cytokines/chemokines were not associated with response (CR/PR vs. SD/PD/Uneval) to Ipi-Bev. Supplementary Table S4. Pre-treatment serum levels of cytokines/chemokines were not associated with disease control (CR/PR/SD vs. PD/Uneval) in Ipi-Bev-treated patients. Supplementary Table S5. Fold-changes of serum cytokines/chemokines were not associated with response (CR/PR vs. SD/PD/Uneval) to Ipi-Bev. Supplementary Table S6. Fold-changes of serum cytokines/chemokines were not associated with disease control (CR/PR/SD vs. PD/Uneval) in Ipi-Bev-treated patients. Supplementary Table S7. Response (CR/PR vs. SD/PD/Uneval) to Ipi-Bev by cytokine expression category. Supplementary Table S8. Disease control (CR/PR/SD vs. PD/Uneval) in Ipi-Bev-treated patients by cytokine expression category. Supplementary Table S9. Pre-treatment cytokine levels according to immune related adverse events (irAEs). Supplementary Table S10. Relationship between cytokine fold-changes according to irAEs. Supplementary Table S11. Clinical responses and survival of patients whose plasma samples were used in antibody response analyses. Supplementary Table S12. Effects of ipilimumab alone and Ipi-Bev on circulating cytokine/chemokine levels.

Purpose:Advanced-stage gastrointestinal cancers represent a high unmet need requiring new effective therapies. We investigated the antitumor activity of a novel T cell–engaging antibody (B7-H6/CD3 ITE) targeting B7-H6, a tumor-associated antigen that is expressed in gastrointestinal tumors.Experimental Design:Membrane proteomics and IHC analysis identified B7-H6 as a tumor-associated antigen in gastrointestinal tumor tissues with no to very little expression in normal tissues. The antitumor activity and mode of action of B7-H6/CD3 ITE was evaluated in in vitro coculture assays, in humanized mouse tumor models, and in colorectal cancer precision cut tumor slice cultures.Results:B7-H6 expression was detected in 98% of colorectal cancer, 77% of gastric cancer, and 63% of pancreatic cancer tissue samples. B7-H6/CD3 ITE-mediated redirection of T cells toward B7-H6–positive tumor cells resulted in B7-H6–dependent lysis of tumor cells, activation and proliferation of T cells, and cytokine secretion in in vitro coculture assays, and infiltration of T cells into tumor tissues associated with tumor regression in in vivo colorectal cancer models. In primary patient-derived colorectal cancer precision-cut tumor slice cultures, treatment with B7-H6/CD3 ITE elicited cytokine secretion by endogenous tumor-infiltrating immune cells. Combination with anti-PD-1 further enhanced the activity of the B7-H6/CD3 ITE.Conclusion:These data highlight the potential of the B7-H6/CD3 ITE to induce T cell–redirected lysis of tumor cells and recruitment of T cells into noninflamed tumor tissues, leading to antitumor activity in in vitro, in vivo, and human tumor slice cultures, which supports further evaluation in a clinical study.

Supplementary Figure 1 (A) OGAP® Screen of tumor and non-tumorous (normal, non-diseased tissues). (B) B7-H6 mRNA expression in gastrointestinal cancer tissues (TCGA). Supplementary Figure 2 B7-H6 expression on cell lines Supplementary Figure 3 Influence of B7-H6/CD3 ITE on the B7-H6-induced activation of NK-92® MI cells. Supplementary Figure 4 Pharmacokinetic profile in NOG mice.

PDF FIle - 95K, Characteristics of colorectal cancer patients according to PTGS2 (COX-2) status.

Purpose: To assess prognostic roles of various KRAS oncogene mutations in colorectal cancer, BRAF mutation status must be controlled for because BRAF mutation is associated with poor prognosis, and almost all BRAF mutants are present among KRAS wild-type tumors. Taking into account experimental data supporting a greater oncogenic effect of codon 12 mutations compared with codon 13 mutations, we hypothesized that KRAS codon 12–mutated colorectal cancers might behave more aggressively than KRAS wild-type tumors and codon 13 mutants.Experimental design: Using molecular pathological epidemiology database of 1,261 rectal and colon cancers, we examined clinical outcome and tumor biomarkers of KRAS codon 12 and 13 mutations in 1,075 BRAF wild-type cancers (i.e., controlling for BRAF status). Cox proportional hazards model was used to compute mortality HR, adjusting for potential confounders, including stage, PIK3CA mutations, microsatellite instability, CpG island methylator phenotype, and LINE-1 methylation.Results: Compared with patients with KRAS wild-type/BRAF wild-type cancers (N = 635), those with KRAS codon 12 mutations (N = 332) experienced significantly higher colorectal cancer–specific mortality [log-rank P = 0.0001; multivariate HR, 1.30; 95% confidence interval (CI), 1.02–1.67; P = 0.037], whereas KRAS codon 13–mutated cases (N = 108) were not significantly associated with prognosis. Among the seven most common KRAS mutations, c.35G>T (p.G12V; N = 93) was associated with significantly higher colorectal cancer–specific mortality (log-rank P = 0.0007; multivariate HR, 2.00; 95% CI, 1.38–2.90, P = 0.0003) compared with KRAS wild-type/BRAF wild-type cases.Conclusions:KRAS codon 12 mutations (in particular, c.35G>T), but not codon 13 mutations, are associated with inferior survival in BRAF wild-type colorectal cancer. Our data highlight the importance of accurate molecular characterization in colorectal cancer. Clin Cancer Res; 18(17); 4753–63. ©2012 AACR.