Key points

Immunotherapy

The development of kirsten rat sarcoma viral oncogene homolog (KRAS)-driven mouse models of PDAC that recapitulate the evolution from low-grade to high-grade pancreatic intraepithelial neoplasia (PanIN) to invasive carcinoma observed in human disease has increased the understanding of the complex immunologic changes involved in the development of PDAC ( Table 1 ). Invasive pancreatic cancer, rather than being poorly immunogenic as was previously considered, is associated with a dynamic immune response; early-stage PanINs promote an inflammatory response that in turn promotes further dysplasia, whereas the established tumor microenvironment is immunosuppressive, facilitating immune evasion. The accumulation of genetic alterations including KRAS and SMAD4 result in the secretion of additional factors, including interleukin 8 and transforming growth factor (TGF) β, which further promote an inflammatory immune response. An established pancreatic tumor therefore typically has an immunosuppressive microenvironment, containing increased number of tumor-associated macrophages with immunosuppressive phenotype, increased numbers of Treg cells and myeloid-derived suppressor cells, and reduced numbers of CD4 ⁺ and CD8 ⁺ T cells. The challenge remains, however, to translate these insights into the complex pathways with many inhibitory and stimulatory signals into the development of effective therapeutic strategies for PDAC.

Immunotherapy

The development of kirsten rat sarcoma viral oncogene homolog (KRAS)-driven mouse models of PDAC that recapitulate the evolution from low-grade to high-grade pancreatic intraepithelial neoplasia (PanIN) to invasive carcinoma observed in human disease has increased the understanding of the complex immunologic changes involved in the development of PDAC ( Table 1 ). Invasive pancreatic cancer, rather than being poorly immunogenic as was previously considered, is associated with a dynamic immune response; early-stage PanINs promote an inflammatory response that in turn promotes further dysplasia, whereas the established tumor microenvironment is immunosuppressive, facilitating immune evasion. The accumulation of genetic alterations including KRAS and SMAD4 result in the secretion of additional factors, including interleukin 8 and transforming growth factor (TGF) β, which further promote an inflammatory immune response. An established pancreatic tumor therefore typically has an immunosuppressive microenvironment, containing increased number of tumor-associated macrophages with immunosuppressive phenotype, increased numbers of Treg cells and myeloid-derived suppressor cells, and reduced numbers of CD4 ⁺ and CD8 ⁺ T cells. The challenge remains, however, to translate these insights into the complex pathways with many inhibitory and stimulatory signals into the development of effective therapeutic strategies for PDAC.

GVAX, CRS207 vaccines

Multiple vaccine-based immunotherapeutic approaches to PDAC have been developed during the last decade, including peptide vaccines, recombinant microorganism-based vaccines, and whole-cell vaccines, with limited activity overall. Encouraging data were presented from a phase 2 trial evaluating the allogeneic whole-cell vaccine, granulocyte-macrophage colony-stimulating factor (GM-CSF) gene-transfected tumor cell vaccine (GVAX), given in combination with low-dose cyclophosphamide alone or followed by live-attenuated listeria monocytogenes–expressing mesothelin (CRS207), an attenuated Listeria strain modified to express mesothelin, an antigen expressed by most PDACs, but with limited expression in normal tissues. Ninety patients with metastatic PDAC with eastern cooperative oncology group performance status (ECOG) 0 to 1 performance status previously treated and some with stable disease at enrollment were randomized 2:1 to either cyclophosphamide/GVAX followed by CRS-207 or cyclophosphamide/GVAX. The primary end point was overall survival. Results demonstrated an improvement of 6 months in survival in the CRS-207 group versus 3.4 months in the cyclophosphamide/GVAX group; P = .0057, hazard ratio (HR) 0.4477, and there was a suggestion of increased benefit in more heavily pretreated patients, possibly representing both tumor biology and patient selection. Both vaccines were well tolerated. An ongoing randomized phase 2 trial (ECLIPSE [Safety and Efficacy of Combination Listeria/GVAX Pancreas Vaccine in the Pancreatic Cancer Setting]) is enrolling patients with previously treated metastatic PDAC to evaluate CRS207 alone or in combination with cyclophosphamide/GVAX, with a third arm randomized to investigators’ choice of several single-agent cytotoxic agents or erlotinib ( NCT02004262 ).

Hyperacute vaccination

Other strategies under evaluation include an allogeneic pancreatic cancer vaccine algenpantucel-L; composed of 2 human PDAC cell lines (HAPa-1 and HAPa-2) that have been genetically modified to express αGal by using retroviral transfer of the murine αGT gene. Antibodies against αGal are produced by human intestinal bacterial flora continuously and are the mediators of hyperacute rejection, thereby potentially inducing an antitumor immune stimulatory response. A recently published single-arm non–phase 2 trial evaluated the addition of algenpantucel-L to standard therapy in patients with resected pancreas cancer and showed encouraging survival in treated patients compared with historical data. Ongoing phase 3 studies are evaluating the efficacy of Algenpantucel-L in patients with locally advanced/borderline resectable and resected PDAC ( NCT01836432 , NCT01072981 ).

Checkpoint blockade

The monoclonal antibody ipilimumab, which blocks binding to cytotoxic T-lymphocyte antigen-4 (CTLA-4), has shown significant activity in clinical trials of patients with advanced melanoma, and thus evaluation has proceeded in other malignancies including PDAC. Binding of CD28 on the T-cell surface to B7-1 and B7-2 receptors on antigen-presenting cells provides a costimulatory signal, which in combination with major histocompatibility complex binding to the T-cell receptor, results in T-cell activation. Inhibition of CTLA-4 therefore prevents the development of immune tolerance, allowing maintenance of T-cell activation in response to tumor-associated antigen presentation. Initial evaluation of ipilimumab in a phase 2 trial of patients with advanced PDAC showed limited activity by conventional response criteria, but 1 patient was observed to have a delayed response. Ongoing trials are evaluating novel checkpoint inhibitors of PD1 alone and in combination with CTLA-4 blockade and inhibitors of PDL1 in patients with advanced pancreatic cancer ( NCT01693562 , NCT01928394 ).

Immune microenvironment

Although these strategies are under early evaluation, it is likely that given the complexity of the PDAC microenvironment, combination strategies of immune checkpoint blockade with stromal depleting agents, vaccines, or agents targeting other components such as macrophages and tumor-associated fibroblasts will be required for significant responses to be observed. Several preclinical studies have shown promising results in preclinical models of PDAC. Feig and colleagues found that depletion of cancer-associated fibroblasts (CAFs) expressing fibroblast activation protein (FAP) allowed control of pancreatic tumors by anti-CTLA 4 and anti-PD-L1. In addition, FAP + CAFs were found to secrete the chemokine CXCL12, and administration of the CXCL12 receptor inhibitor, plerixafor (AMD3100), acted synergistically with anti-PDL1 to induce significant tumor regression. Early phase 1 assessment of this strategy in advanced solid tumor malignancies and PDAC is scheduled to start in the near future.

T-cell manipulation

Another immunotherapeutic strategy with potential promise is utilization of modified T cells that are genetically engineered to express a chimeric antigen receptor, targeting in solid tumor malignancies, mesothelin, a tumor antigen significantly expressed in PDAC. An ongoing phase 1 trial is evaluating the safety and early efficacy signal of this approach in PDAC ( NCT01897415 ).

Immunotherapeutic approaches are also beginning to be investigated in combination with cytotoxic therapy and as a maintenance strategy in PDAC. An example of the former is the evaluation of indoleamine 2,3-dioxygenase inhibitors in conjunction with a standard cytotoxic regimen ( NCT02077881 ), the rationale in part based on preclinical data demonstrating synergy with chemotherapy.

BRCA-related therapies and poly-ADP ribose polymerase inhibitors

Although most cases of PDAC are sporadic, up to 1 in 10 cases occur in the setting of a hereditary cancer predisposition syndrome, the most common of which is a germline BRCA1 or 2 mutation. The lifetime risk of PDAC in a BRCA1/2 mutation carrier is estimated at between 2 and 3.5 times that of the general population. Among patients with more than 3 family members affected with pancreatic cancer, the prevalence of germline breast cancer 2 early onset ( BRCA2 ) mutations was found to be 17%, whereas in a study of European families with at least 2 first-degree relatives with pancreatic cancer, 12% were found to harbor a BRCA2 mutation. In Ashkenazi Jewish patients, mutations in BRCA1/2 are more prevalent and are usually 1 of 3 founder mutations; breast cancer 1 early onset ( BRCA1 ) 185delAG and 5382insC and BRCA2 6174delT. In a series of patients of Ashkenazi ancestry unselected for family history from Memorial Sloan Kettering Cancer Center (MSKCC) with resected pancreatic cancer, 5.5% were found to be BRCA1 / 2 mutation positive. In a subsequent series evaluating 211 Ashkenazi Jewish patients with a personal history of breast cancer and a family history of PDAC, 30 (14.2%) were found to harbor a mutation in BRCA , 14 (47%) in BRCA1 , and 16 (53%) in BRCA2 . Germline mutations in the gene PALB2 , which encodes a protein critical for the initiation of homologous recombination, have also been identified in patients with pancreatic cancer and a personal or family history of breast cancer.

Overall, this population represents a small but significant number of patients with PDAC, in whom the identification of an inherited cancer predisposition syndrome may be potentially exploited for therapeutic benefit. DNA double strand breaks (DSBs) may be induced during cell replication by stalled replication forks and are optimally repaired by homologous recombination; in the absence of functional BRCA1/BRCA2 or PALB2, however, these DSBs are repaired by the error-prone nonhomologous end joining pathway leading to genetic instability. Poly ADP-ribose polymerase (PARP) inhibitors (PARPis) block repair of single strand breaks and induce DSBs. They have been developed for treatment of cancers deficient in homologous recombination, which are unable to accurately repair DSBs and so maintain genomic integrity, resulting in cell death via a synthetic lethal effect. Similarly, BRCA-deficient cells demonstrate increased susceptibility to platinum drugs because of induction of interstrand cross-links, which require repair by the homologous repair pathway. Several retrospective series have described significant response to platinum agents and PARPis in patients with known BRCA1/2 mutations and pancreatic cancer. A prospective phase 2 study of the PARPi olaparib in patients with BRCA1/2-associated malignancy enrolled 23 patients with pancreatic cancer, of whom 22% showed either complete or partial response to treatment with single agent olaparib. In addition, 35% of patients with pancreatic cancer demonstrated stable disease for at least 8 weeks. Overall survival after 1 year was 41% for patients with BRCA1/2 mutation-associated pancreatic cancer treated in the study.

At MSKCC, in collaboration with the National Cancer Institute (NCI) and other sites in the United States, Israel, and Canada, the authors are conducting several trials evaluating the safety and activity of platinum drugs and PARP inhibition in this subpopulation of BRCA and PALB2 mutated PDAC by combining the PARPi veliparib with the cytotoxic backbone of cisplatin and gemcitabine. Results of the phase 1b portion of this study were presented at the American Society of Clinical Oncology in 2014; 17 patients were enrolled, 9 of whom had known BRCA1/2 mutations. The main grade 3 to 4 toxicities were fatigue and hematologic events. The recommended phase 2 dose of veliparib in combination with gemcitabine and cisplatin was identified to be 80 mg by mouth twice a day from day 1 to 12 every 3 weeks. Significant activity was observed in all BRCA/12 mutation carriers treated. A randomized phase 2 trial evaluating cisplatin and gemcitabine with or without the addition of veliparib in BRCA1/2 or PALB2 mutation carriers is underway. An additional phase 2 protocol evaluated the use of single agent veliparib, at a dose of 400 mg daily, as second-line or third-line therapy in patients with previously treated BRCA- or PALB2-mutated pancreas cancer is complete and results are pending ( NCT01585805 ).