Surgical resection of the primary tumour is the mainstay of breast cancer treatment. However, recent clinical findings have raised concerns about possible effects of perioperative medications on cancer recurrence and metastasis. Opioids are the cornerstone of pain management during and after surgery and morphine and fentanyl, as potent agonists of the mu opioid receptor are commonly used for pain management during this period. On the other hand, cyclooxygenase (COX) inhibitors are inexpensive and efficient analgesic and antiinflammatory agents widely used to control postoperative pain. Recently the possible effects of these drugs on tumours have received increasing attention due to mounting evidence on the involvement of the COX system in tumour progression. Plasminogen activator inhibitor antifibrinolytics are another perioperative drug class that may affect tumour invasiveness due to the key role of plasminogen activation system in tumour cell invasion and metastasis. A variety of in vitro and in vivo models have been employed to investigate the possible effects of these drugs on tumour growth and metastasis leading to contradictory results that show either pro-tumour, anti-tumour or no effect. While most of the in vitro studies have focused on tumour cells, little is known about the possible effect of these drugs on the tumour microenvironment. Tumour microenvironment consists of different types of resident and infiltrating cells including fibroblasts, inflammatory and endothelial cells that closely regulate tumour growth and invasion via various mediators. Extracellular matrix (ECM) proteases such as matrix metalloproteinases (MMPs) and urokinase-type plasminogen activator (uPA) are key molecules in the invasion of tumour cells to adjacent tissue and entry to the circulation as well as localization and growth in secondary organs. There is evidence that the tumour microenvironment plays an important part in the regulation of the expression and activity of these enzymes in the tumour tissue.
We hypothesized that the interaction between breast tumour cells and non-malignant cells in the tumour microenvironment leads to a pro-migratory milieu by modulating the level of ECM proteolytic enzymes, which can be altered by commonly used perioperative medications, namely opioid analgesics, COX inhibitors and plasminogen activator inhibitor antifibrinolytics. In a syngeneic model of breast cancer, mice were inoculated with 4T1 breast carcinoma cells either through the tail vein or in the mammary gland. After 14 days, the levels of ECM degrading enzymes matrix metalloproteinase-9 (MMP-9) and uPA wereelevated in the serum as well as breast tumour and its surrounding tissue. In order to determine the role of non-malignant cells of the tumour microenvironment in the upregulation of proteases, we established a coculture model in which breast cancer cells were cultured together with fibroblasts, macrophages or endothelial cells. None of the proteases tested in our studies was changed by coculture of breast cancer cells with fibroblasts compared to cells cultured alone. However, in cocultures of breast cancer cells with macrophages, MMP-9 and TIMP-1 and in cocultures of breast cancer cells with endothelial cells MMP-9, MMP-2 and TIMP-1 as well as uPA were significantly increased. We determined whether these changes in proteolytic profile were affected by the opioid agonists morphine and fentanyl, the non-selective COX inhibitors aspirin and ketorolac, the selective COX-2 inhibitor celecoxib and the antifibrinolytics ε-aminocaproic acid (EACA) and tranexamic acid (TXA). Opioid agonists decreased MMP-9 and increased TIMP-1 in the cocultures of breast cancer cells with macrophages or endothelial cells but had no significant effect on individual cells. Fentanyl also decreased uPA in 4T1 cells. Antifibrinolytics and COX inhibitors exerted a complex effect in both individual cells and cocultures. EACA increased the activity of MMP-9 in 4T1, RAW264.7 and their cocultures; however the MMP- 9/TIMP-1 ratio was not changed in a statistically significant manner. Alternatively, EACA and TXA both attenuated the increase in MMP-2 level in 4T1 and H5V cocultures without affecting it in individual cells. Non-selective COX inhibitors, aspirin and ketorolac decreased uPA in breast cancer cells and endothelial cells grown alone or in cocultures. These drugs also attenuated the increase in MMP-2 in the cocultures of cancer cells with endothelial cells. The selective COX-2 inhibitor celecoxib did not affect any of the proteolytic enzymes. However, it increased the level of TIMP-1 in the cocultures of 4T1 with both macrophages and endothelial cells.
In summary, in this project it is shown for the first time that drugs used in the perioperative period can affect the proteolytic profile of the tumour microenvironment. These findings provide insight into one possible mechanism by which common perioperative medications might affect the course of cancer. A better understanding of these mechanisms is useful in more effective use of perioperative medications during the critical period of cancer surgery.