Pancreatic stellate cells (PSC), which are implicated in desmoplasia in pancreatic cancer, enhance the malignancy of cancer cells and confer resistance to established treatments. We investigated whether the antifibrotic agent pirfenidone can suppress desmoplasia and exert antitumor effects against pancreatic cancer. Primary PSCs were established from pancreatic cancer tissue obtained during surgery. In vitro, pirfenidone inhibited the proliferation, invasiveness, and migration of PSCs in a dose-dependent manner. Although supernatants of untreated PSCs increased the proliferation, invasiveness, and migration of pancreatic cancer cells (PCC), supernatants of pirfenidone-treated PSCs decreased these effects. Exposure to PCC supernatant increased the production of platelet-derived growth factor-A, hepatic growth factor, collagen type I, fibronectin, and periostin in PSCs, which was significantly reduced by pirfenidone. Mice were subcutaneously implanted with PCCs (SUIT-2 cells) and PSCs into the right flank and PCCs alone into the left flank. Oral administration of pirfenidone to these mice significantly reduced tumor growth of co-implanted PCCs and PSCs, but not of PCCs alone. Pirfenidone also decreased the proliferation of PSCs and the deposition of collagen type I and periostin in tumors. In mice with orthotopic tumors consisting of PCCs co-implanted with PSCs, pirfenidone suppressed tumor growth, reduced the number of peritoneal disseminated nodules, and reduced the incidence of liver metastasis. Pirfenidone in combination with gemcitabine more effectively suppressed orthotopic tumor growth compared with pirfenidone or gemcitabine alone. In conclusion, our findings indicate that pirfenidone is a promising antitumor agent for pancreatic cancer, owing to its suppression of desmoplasia through regulating PSCs. Cancer Res; 73(7); 2345–56. ©2013 AACR.

Background: Interactions between cancer cells and surrounding cancer-associated fibroblasts (CAFs) play an important role in cancer progression. Invasive ductal carcinoma (IDC) of the pancreas is characterized by abundant fibrous connective tissue called desmoplasia. Podoplanin (PDPN) is a lymphatic vessel marker (D2-40), and expression of PDPN by stromal CAFs has been reported to be a prognostic indicator in various types of cancer. Methods: Expression of PDPN in pancreatic IDCs was assessed by immunohistochemical examination in 105 patients who underwent pancreatic resection. Primary CAFs were established from pancreatic cancer tissue obtained by surgery. Quantitative reverse transcription-polymerase chain reaction and flow cytometric analysis were performed to investigate PDPN expression in CAFs. We sorted CAFs according to PDPN expression, and analyzed the functional differences between PDPN+ CAFs and PDPN– CAFs using indirect co-culture with pancreatic cancer cell lines. We also investigated the culture conditions to regulate PDPN expression in CAFs. Results: PDPN expression in stromal fibroblasts was associated with lymphatic vessel invasion (P = 0.0461), vascular invasion (P = 0.0101), tumor size ≥3 cm (P = 0.0038), histological grade (P = 0.0344), Union for International Cancer Control classification T stage (P = 0.029), and shorter survival time (P < 0.0001). Primary CAFs showed heterogeneous PDPN expression in vitro. Moreover, migration and invasion of pancreatic cancer cell lines (PANC-1 and SUIT-2) were associated with PDPN expression in CAFs (P < 0.01) and expression of CD10, matrix metalloproteinase (MMP) 2, and MMP3. In cultured CAFs, PDPN positivity changed over time under several conditions including co-culture with cancer cells, different culture media, and addition of growth factor. Conclusions: PDPN-expressing CAFs enhance the progression of pancreatic IDC, and a high ratio of PDPN-expressing CAFs is an independent predictor of poor outcome. Understanding the regulation of the tumor microenvironment is an important step towards developing new therapeutic strategies.

Stromal fibrosis is a prominent histologic characteristic of pancreatic ductal adenocarcinoma (PDAC), but how stromal fibroblasts are regulated in the tumor microenvironment (TME) to support tumor growth is largely unknown. Here we show that PDAC cells can induce DNA methylation in cancer-associated fibroblasts (CAF). Upon direct contact with PDAC cells, DNA methylation of SOCS1 and other genes is induced in mesenchymal stem cells or in CAF that lack SOCS1 methylation at baseline. Silencing or decitabine treatment to block the DNA methylation enzyme DNMT1 inhibited methylation of SOCS1. In contrast, SOCS1 gene methylation and downregulation in CAF activated STAT3 and induced insulin-like growth factor-1 expression to support PDAC cell growth. Moreover, CAF facilitated methylation-dependent growth of PDAC tumor xenografts in mice. The ability of patient-derived CAF with SOCS1 methylation to promote PDAC growth was more robust than CAF without SOCS1 methylation. Overall, our results reveal how PDAC cells can reprogram CAF to modify tumor–stromal interactions in the TME, which promote malignant growth and progression. Cancer Res; 76(18); 5395–404. ©2016 AACR.

To identify a high-risk group of patients with pancreatic ductal adenocarcinoma (PDAC), independently arising in the pancreas with intraductal papillary mucinous neoplasm (IPMN), using histopathologic subtypes. Pathologic features of IPMN with distinct PDAC, including histopathologic subtypes of IPMN and PDAC phenotypes, have not been well characterized. Mucin expression patterns and the mutational status of GNAS and KRAS are useful to explore the relationship between these 2 lesion types. Clinicopathologic data of 179 resected IPMNs and 180 resected PDACs without IPMNs as a control group were reviewed. IPMNs were classified into 4 grades (low-grade, intermediate-grade, high-grade dysplasia, and an associated invasive carcinoma) and 4 subtypes (gastric, intestinal, pancreatobiliary, and oncocytic). The expression of MUC1, MUC2, MUC5AC, MUC6, and CDX2 was investigated by immunohistochemistry in IPMNs and PDACs with and without IPMNs. The mutational status of GNAS and KRAS was evaluated by cycle sequencing in PDACs and pre-/coexisting IPMNs. Twenty-six synchronous or metachronous PDACs were identified in 20 patients (11.2%) with IPMNs. Occurrence of concomitant PDACs was more frequently observed in gastric-type IPMNs (18/110, 16.4%) compared with intestinal (1/49, 2.0%), pancreatobiliary (1/17, 5.9%), or oncocytic-type (0/3, 0%) (P = 0.047). Both PDACs with and without IPMNs were frequently positive for MUC1, MUC5AC, and MUC6 expression, as assessed by immunohistochemistry, but were negative for MUC2 and CDX2. The mucin-staining patterns were similar to those of invasive tubular adenocarcinoma arising from gastric-type IPMNs. Mutation of GNAS within codon 201 was not detected in PDACs and gastric-type IPMNs, whereas most of these exhibited KRAS mutations. However, the R201H GNAS mutation was detected in 1 intestinal-type IPMN with distinct PDAC. Mucin expression patterns demonstrate that PDAC without GNAS mutations of an aggressive phenotype frequently arise in the pancreas with benign gastric-type IPMN in the absence of GNAS mutations.

Background: Immunotherapy has demonstrated a limited clinical efficacy in approximately 5% of cholangiocarcinoma. The main challenges for an effective immunotherapy response in cholangiocarcinoma arise from the tumor microenvironment, which is poorly understood. Methods: For a comprehensive analysis of the tumor microenvironment in cholangiocarcinoma, we performed multiplex immunohistochemistry with two 15-marker immune panels and Nanostring assays for a comprehensive analysis of 104 surgically resected cholangiocarcinomas including intrahepatic, hilar, and distal cholangiocarcinoma. We also validated some key findings with a batch integration analysis of published single cell RNA sequencing data. Results: This study found that natural killer cells occupy the largest immune cell compartment in cholangiocarcinoma. Granzyme-B⁺CD8⁺ effector T cells are significantly associated with better overall survival in both intrahepatic and distal cholangiocarcinoma. Above 85% of intrahepatic cholangiocarcinomas with higher density of PD-1⁻EOMES⁻CD8⁺ effector T cells are associated with long-term survival. However, only the density of PD-1⁻EOMES⁻CD8⁺ T cells in the tumor areas, but not in the peripheries of the tumors, is prognostic. In all three cholangiocarcinoma subtypes, T regulator cells are significantly associated with a poor prognosis; however, M1 and M2 tumor-associated macrophages or PD-L1⁺ tumor-associated macrophage demonstrate different prognostic values. Combining PD-L1⁺ M1 or M2, PD-L1⁻ M1 or M2 tumor-associated macrophages, and T regulator cells to subgroup intrahepatic and distal cholangiocarcinoma, the prognosis is significantly better distinguished. Moreover, PD-L1⁻ M2 tumor-associated macrophages is associated with a good prognosis in intrahepatic and distal cholangiocarcinoma, suggesting this subtype of M2 tumor-associated macrophages may be antitumoral. Interestingly, lower densities of various types of immunosuppressive cells are associated with decreased infiltration of effector T cells in distal and hilar cholangiocarcinoma, but not in intrahepatic cholangiocarcinoma. In intrahepatic cholangiocarcinoma, PD-L1⁺ tumor-associated macrophages exert their immunosuppressive function likely through promoting T cell exhaustion. Conclusions: This study suggests that the densities of Granzyme-B⁺CD8⁺ effector T cells and non-exhausted PD-1⁻EOMES⁻CD8⁺ T cells and the PD-L1 status in the tumor-associated macrophages are prognostic makers in cholangiocarcinomas. The study also supports targeting PD-L1⁺ tumor-associated macrophages as the immunotherapy for cholangiocarcinoma.