Smoking is highly associated with pancreatic cancer. Nicotine, the addictive component of tobacco, is involved in pancreatic cancer tumorigenesis, metastasis, and chemoresistance. This work aimed to describe the role of nicotine within the pancreatic cancer tumor microenvironment. Nicotine treatment was used in vitro to assess its effect on tumor-associated stromal cells and pancreatic cancer cells. Nicotine treatment was then used in a pancreatic cancer patient-derived xenograft model to study the effects in vivo. Nicotine induced secretion of interleukin 8 (IL-8) by tumor-associated stroma cells in an extracellular signal-regulated kinase (ERK)-dependent fashion. The secreted IL-8 and nicotine acted on the pancreatic cancer cell, resulting in upregulation of IL-8 receptor. Nicotine treatment of mice bearing pancreatic cancer patient-derived xenografts had significantly increased tumor mass, increased tumor-free weight loss, and decreased muscle mass. These represent important pathways through which nicotine acts within the tumor microenvironment and worsens pancreatic cancer-induced cachexia, potentially representing future therapeutic targets.

Cancer cachexia is a debilitating condition seen frequently in patients with pancreatic ductal adenocarcinoma (PDAC). The underlying mechanisms driving cancer cachexia are not fully understood but are related, at least in part, to the immune response to the tumor both locally and systemically. We hypothesize that there are unique differences in cytokine levels in the tumor microenvironment and systemic circulation between PDAC tumors and that these varying profiles affect the degree of cancer cachexia observed. Patient demographics, operative factors, oncologic factors, and perioperative data were collected for the two patients in the patient derived xenograft (PDX) model. Human pancreatic cancer PDX were created by implanting fresh surgical pancreatic cancer tissues directly into immunodeficient mice. At PDX end point, mouse tumor, spleen and muscle tissues were collected and weighed, muscle atrophy related gene expression measured, and tumor and splenic soluble proteins were analyzed. PDX models were created from surgically resected patients who presented with different degrees of cachexia. Tumor free body weight and triceps surae weight differed significantly between the PDX models and control (P < 0.05). Both PDX groups had increased atrophy related gene expression in muscle compared to control (FoxO1, Socs3, STAT3, Acvr2b, Atrogin-1, MuRF1; P < 0.05). Significant differences were noted in splenic soluble protein concentrations in 14 of 15 detected proteins in tumor bearing mice when compared to controls. Eight splenic soluble proteins were significantly different between PDX groups (P < 0.05). Tumor soluble proteins were significantly different between the two PDX groups in 15 of 24 detected proteins (P < 0.05). PDX models preserve the cachectic heterogeneity found in patients and are associated with unique cytokine profiles in both the spleen and tumor between different PDX. These data support the use of PDX as a strategy to study soluble cachexia protein markers and also further efforts to elucidate which cytokines are most related to cachexia in order to provide potential targets for immunotherapy.