Preoperative granulocyte/macrophage colony-stimulating factor (GM-CSF) increases hepatic dendritic cell numbers and clustering with lymphocytes in colorectal cancer patients.

Despite surgery with curative intent, approximately 30% of colorectal carcinoma patients will develop liver metastases during follow-up. Synchronous occult micrometastases, tumor cell shedding into the portal circulation and postoperative immune impairment have all been suggested to facilitate outgrowth of liver metastases. In experimental models, increases in both number of resident macrophages of the liver, the so-called Kupffer cells (KC), and tumoricidal capacity of KC were observed after pretreatment with granulocyte/macrophage colony-stimulating factor (GM-CSF), a potent immuno-stimulatory agent. Following perioperative recombinant human GM-CSF (rhGM-CSF), we previously showed activation of the systemic immune response in the postoperative period, which is normally transiently down-modulated after surgery. Therefore, in this pilot study, effects of preoperative rhGM-CSF administration on the composition of human liver immune cell population were evaluated in patients undergoing surgery for colorectal cancer. No difference in KC numbers of rhGM-CSF-treated patients was observed. Importantly, however, a 6-fold increase in dendritic cell (DC) numbers was observed compared to control patients, as quantified by immunohistochemistry of liver biopsies, taken during laparotomy. Furthermore, direct contact between liver CD8+ cells and DC was significantly enhanced in rhGM-CSF-treated patients. Both increases in DC numbers and DC interaction with CD8+ T cells suggest enhanced immunological activation, which may reduce liver metastases formation and ultimately improve survival after initial colorectal surgery.

The effects of surgery, with or without rhGM-CSF, on the angiogenic profile of patients treated for colorectal carcinoma.

Wound healing is a process with immunological and angiogenic aspects. rhGM-CSF is known to stimulate the immune system and angiogenesis via multiple pathways. In this study we investigated the combined effects of surgery, with or without rhGM-CSF, on angiogenic parameters in patients with a colorectal carcinoma. In this phase II randomized, placebo-controlled trial, 16 patients were assigned to perioperative rhGM-CSF (2.8 microg/kg body weight) treatment or saline. Patients received subcutaneous injections from three days before surgery until four days after. IL-6, VEGF, endostatin and angiostatin levels were measured perioperatively. rhGM-CSF enhanced the production of IL-6 and VEGF, but had no effect on the antiangiogenic agents endostatin and angiostatin. Surgery induced a transient decrease of endostatin. Two types of angiostatin (kringle 1-3 and kringle 1-4) became visible postoperatively. We conclude that this study demonstrated the immediate initiation of angiogenesis postoperatively, reflected by the increase of VEGF and a transient decrease of endostatin, followed by the appearance of two angiostatin bands, which confirms physiological wound healing in these cancer patients.