Effects of surgery and anesthetic choice on immunosuppression and cancer recurrence

1Surgical resection is the most effective method to remove primary tumors and metastatic lymph nodes. However, some cancer cells may remain after surgery, and micro-metastases or tumor dislodged during surgical manipulation may spread via lymphovascular vessels. During the perioperative period, surgery induced stress responses and anesthetic-induced immunosuppression may play a critical role in establishment and growth of metastatic lesions. Because immune responses are regulated by the hypothalamic–pituitary–adrenal (HPA) axis and sympathetic nervous system (SNS), surgery-induced or anesthesia-induced activation of these two systems may facilitate metastasis through several tumor-derived soluble factors.

2HPA-axis and SNS activation suppress cell-mediated immunity (CMI) and release of catecholamines and prostaglandin E2 (PGE2). These factors, in turn, increase immunosuppressive cytokines, soluble factors (e.g., interleukin 4 [IL-4], IL-10, transforming growth factor beta [TGF-β], and vascular endothelial growth factor [VEGF]), and proinflammatory cytokines (e.g., IL-6 and IL-8), which promote tumor angiogenesis and metastasis. Furthermore, volatile anesthetics and opioids suppress CMI and promote cancer cell proliferation and angiogenesis, whereas propofol inhibits tumor angiogenesis and does not suppress CMI. Regional anesthesia (RA) preserves CMI and decreases surgery-induced neuroendocrine responses by attenuating afferent neural transmission activation of the HPA-axis and SNS response. Thus, reduction in opioid and volatile anesthetic use may reduce cancer recurrence.

3Clinically, the key question of whether anesthetic choice affects cancer outcome remains unresolved. Retrospective studies and meta-analyses suggest that particular anesthetic techniques may reduce cancer related mortality and recurrence by decreasing immunosuppression after surgical treatment for certain types of cancer. Several prospective randomized controlled trials (RCTs) to define the effect of anesthesia on cancer recurrence are currently underway. The present study relied on preclinical study review to determine the potential effects of surgery and anesthetic choice on immunosuppression and cancer outcomes to help guide treatment choices by clinicians and cancer surgeons.

4The perioperative period is divided into three phases: the preoperative period (a few preoperative hours), the intraoperative period, and the postoperative period (several days after surgery) (Fig. 1a). During the intraoperative period, general anesthesia consists of administration of intravenous anesthetics (e.g., thiopental or propofol) for induction, followed by muscle relaxants and endotracheal intubation, then volatile anesthetics (e.g., sevoflurane) and opioids for maintenance and pain control. In contrast, RA uses a local anesthetic (e.g., lidocaine or bupivacaine) to block peripheral or spinal nerve transmission to produce a paravertebral or epidural block. Local anesthetics prevent surgical pain and reduce surgery-induced neuroendocrine stress by suppressing afferent neural transmission to the central nervous system. Thus, HPA-axis and SNS responses are avoided. Anesthetic choices during cancer surgery positively or negatively affect immune function during the perioperative period, and this immune balance may play a key role in cancer spread and recurrence (Fig. 1b).

5Tumors release soluble factors into their microenvironments to block CMI surveillance and facilitate tumor growth and metastasis. Soluble factors affect residual cancer cells and pre-existing micro-metastases to promote new metastases, which are the major cause of cancer-related death if not eliminated by immune cells. In these situations, the perioperative period is pivotal in determining cancer outcomes following primary surgical treatment. Surgery, anesthesia, analgesia, and specific agents all influence immune function and tumor metastasis. Immunosuppression arises within a few hours of surgery and lasts for several days, in proportion to the extent of surgical trauma. Although the immune system normally protects against tumor development, surgery-induced stress counteracts the anti-metastatic effects of CMI to allow dissemination and metastasis of cancer cells during and following surgery. The perioperative period may be crucial to residual cancer cell spread, with anesthetic-induced immunosuppression affecting cancer recurrence and long-term prognosis.

6Although surgical resection is a major component of cancer treatment, surgery itself suppresses immunity; thus, metastasis is promoted through growth facilitation of pre-existing micro-metastases and dissemination of cancer cells during resection of the primary lesion. Detection of tumor cells in peritoneal blood and fluid after surgery has been associated with significantly shorter disease-free survival in patients with colorectal cancer. Given that surgery modifies neural, endocrine, metabolic, inflammatory, and immunologic microenvironments, surgery-induced stress responses may activate angiogenesis and increase vascularization to promote tumor growth.

7Additionally, surgical resection may promote tumor growth and metastasis through increased matrix metalloproteinase 9 (MMP-9) and VEGF expression, as in one model of breast cancer. Plasma VEGF levels are increased by surgery-induced stress during laparotomy and mastectomy, whereas TGF-β plasma levels decrease in response to lung metastasis in animal models. Acceleration of metastasis after surgical resection through proliferation of distant, dormant micro-metastases has been observed in patients with breast cancer.

8Primary tumor resection may directly stimulate cancer cell spread through metastatic lesion growth. Following surgical resection of a primary tumor, decreased endostatin and angiostatin levels allow new blood vessel growth, which promotes growth or metastatic lesions and uncontrolled proliferation. Following primary colorectal tumor resection, decreased angiostatin and endostatin levels in urine and plasma are associated with increased metabolic activity in liver metastases. Thus, it seems that the primary tumor inhibits angiogenesis for distant metastases, but that primary tumor resection allows for neovascularization and increased metabolic activity in metastases. If surgery-induced immunosuppression occurs, surgery may fail to prolong survival in patients with cancer. Surgery reduces levels of endogenous antiangiogenic factors such as endostatin and angiostatin while weakening the immune surveillance needed to inhibit the growth of metastatic lesions.

9Changes to NK cell activity depend on both the degree of surgical treatment and the intensity of the surgical stress response, which activates the HPA-axis and SNS to release catecholamines and prostaglandins. Laparotomy increases lung tumor retention (LTR), whereas combined β-adrenergic antagonism and cyclooxygenase (COX) inhibition decrease LTR and restore NK cell function in experimental animal models. Clinically, surgery decreases circulating NK and T cells through the programmed death-1 (PD-1) and programmed death–ligand 1 (PD-L1) pathway, due to increased caspase-3 activity in association with PD-1 expression on immune cells. Surgical stress increases Th2 cells and decreases Th1 cells, which decreases the Th1/Th2 ratio and eventually suppresses CMI. During surgical stress, levels of immune stimulating cytokines such as IL-2, IL-12, and interferon-γ (IFN-γ) are decreased, whereas anti-inflammatory cytokines such as IL-10 are increased. The magnitude of immunosuppression is in proportion to the extent of surgical treatment. The overall effect of surgery on immune function and tumor metastasis is summarized in Table 1.

10Intravenous anesthetics such as ketamine and thiopental produce multiple effects on immune system components. Unlike propofol, ketamine and thiopental suppress NK cell activity. Whereas ketamine induces human lymphocyte apoptosis via the mitochondrial pathway and inhibits dendritic cell (DC) functional maturation, whereas thiopental protects against T-lymphocyte apoptosis through induction of heat shock proteins. However, both of these intravenous anesthetics suppress the immune system in other ways: ketamine decreases production of pro-inflammatory cytokines such as IL-6 and tumor necrosis factor-α (TNF-α), and thiopental inhibits neutrophil function and suppresses activation of nuclear factor kappa B (NF-κB). This NF-κB suppression by thiopental is associated with inhibition of NF-κB-driven reporter gene activity, which includes T-lymphocyte activation as well as IL-2, IL-6, IL-8, and IFN-γ expression. Thiopental also inhibits lipopolysaccharide-induced production of IL-1β, TNF-α, and IL-6 by monocytes. Although intraperitoneal injection of midazolam impairs monocyte and neutrophil function, it does not affect cytotoxic T-lymphocyte (CTL) activity in a mouse model.

11In contrast to other intravenous anesthetics, propofol increases CTL activity, decreases pro-inflammatory cytokines, and inhibits COX-2 and PGE2 functions. Furthermore, propofol does not affect Th1/Th2, IL-2/IL-4, or CD4/CD8 T cell ratios, so surgery-induced immunosuppression is mitigated.

12Volatile anesthetics also affect immune response. For example, halothane decreases NK cell activity and increases expression of hypoxia-inducible factor 1α (HIF-1α), and sevoflurane induces T-lymphocyte apoptosis and upregulates HIF-1α expression. Sevoflurane has also been shown to increase levels of pro-tumorigenic cytokines and MMPs in breast cancer surgery. One study comparing desflurane to sevoflurane showed that sevoflurane decreases lymphocytes and NK cells while increasing leukocytes and neutrophils during abdominal surgery. Similarly, isoflurane attenuates NK cell activity, induces T-lymphocyte and B-lymphocyte apoptosis, and decreases the Th1/Th2 ratio. Desflurane does not induce T-lymphocyte apoptosis.

13Opioids usually inhibit T-lymphocyte proliferation. Morphine suppresses NK cell activity and T cell differentiation, promotes lymphocyte apoptosis, and decreases toll-like receptor 4 (TLR4) expression on macrophages. Likewise, fentanyl and sufentanil decrease NK cell activity but increase regulatory T cells. Sufentanil also inhibits leukocyte migration. Alfentanil decreases NK cell activity, and remifentanil has demonstrated suppression of NK cell activity and lymphocyte proliferation in a rat model. A comparison of sufentanil and remifentanil using target-controlled infusion during laparoscopic colorectal cancer resection showed that cortisol and IL-6 increased more in the remifentanil group and that the proportion of T cell subsets decreased more in the sufentanil group.

14COX-2 induction, which is frequently observed in cancer, plays a role in immune evasion and resistance to the immune response. COX-2 inhibitors increase NK cytotoxicity and β-adrenergic antagonism while reducing postoperative LTR. Additionally, combined β-adrenergic antagonism and COX-2 inhibition have been shown to eliminate LTR and decrease metastasis in animal models. A selective COX-2 inhibitor can suppress PGE2 release and promote CTL immune responses that cause ovarian tumor regression. Furthermore, a murine model has shown that celecoxib, a COX-2 inhibitor that reduces PGE2 levels, reduces and suppresses myeloid-derived suppressor cells (MDSCs); this in turn decreases reactive oxygen species and nitric oxide (NO) levels and reverses T cell tolerance. Preoperative treatment with nonsteroidal anti-inflammatory drugs (NSAIDs) increases infiltration of activated immune cells into colorectal cancer tissue. Of interest, a recent study showed that lidocaine at typical clinical concentrations enhanced NK cell activity against cancer cells in vitro via the release of lytic granules. The overall effect of anesthetic agents on immune function is summarized in Table 2.

15Treatment with intravenous anesthetics such as ketamine and thiopental stimulate lung and liver metastases in animal models, with one study showing that ketamine and thiopental increase LTR or lung metastasis via NK cell suppression in a rat model. Similarly, the volatile anesthetic halothane can stimulate lung and liver metastases. In contrast, sevoflurane suppresses hypoxia-inducible growth and metastasis of lung cancer cells by inhibiting HIF-1α, which is involved in the p38 mitogen-activated protein kinase (MAPK) signaling pathway. Another study has shown that sevoflurane increases proliferation, migration, and invasion of estrogen receptor (ER)-positive breast cancer cells, as well as proliferation and migration of ER-negative cells. Furthermore, serum from patients who received sevoflurane and an opioid for breast cancer surgery did not inhibit proliferation of ER-negative breast cancer cells, but serum from those receiving propofol and paravertebral anesthesia did inhibit proliferation.

16Exposure to sevoflurane but not total intravenous anesthesia (TIVA) by propofol results in increased prosurvival proteins such as cytoplasmic HIF-2α and nuclear p38 MAPK in head and neck squamous cell carcinoma. Isoflurane is associated with increased HIF-1α levels and increased prostate cancer cell proliferation and migration. In contrast, isoflurane-induced HIF-1α activation is prevented by propofol, which is associated with partial reduction of malignant activities by cancer cells. Additionally, tumor growth in inoculated in mice is suppressed by propofol, which may have immune-mediated antitumor effects. Isoflurane increases the malignant potential of ovarian cancer cells through the upregulation of insulin-like growth factor (IGF)-1 and its receptor IGF-1R, as well as VEGF, angiopoietin-1, MMP-2, and MMP-9. Furthermore, isoflurane exposure leads to apoptotic resistance in human colon cancer cells through a caveolin-1-dependent mechanism. Nitrous oxide (N2O) impairs DNA, purine, and thymidylate synthesis, which can itself cause of oncogenesis. A tumor-bearing mouse model has shown that N2O suppresses chemotaxis, which may be the most potent stimulator of postsurgical lung and liver metastasis development. However, it is unlikely that N2O increases the risk of cancer recurrence compared to that of nitrogen after colorectal surgery.

17Commonly used opioid analgesics may affect tumor development through their modulation of cell proliferation and cell death. It has been suggested that opioids suppress the immune response because various immune competent cells express opioid receptors and induce apoptosis during opioid alkaloid treatment. Tumor growth promotion is mediated through AKT and extracellular signal–regulated kinase (ERK) signaling cascades, whereas death-promoting effects are mediated through NF-κB inhibition, increased Fas expression, p53 stabilization, activation of p38, and c-Jun-N-terminal kinase (JNK). It is likely that opioid-induced cell proliferation and cell death depend on opioid concentration or exposure duration. Tumor growth promotion occurs with low concentrations or single doses of opioids, whereas growth inhibition occurs with chronic opioid use or relatively high drug concentrations.

18Breast cancer cells treated with low morphine concentrations induce naloxone (NX)-sensitive, concentration-dependent increases in GTPase activity, with morphine signals being transmitted by opioid receptors via a G protein. In contrast, the anti-proliferative effects of morphine are not eliminated by NX. Morphine-induced p53 phosphorylation and stabilization in breast cancer cells expressing wild type p53 causes increased production of p53-dependent proteins, including p21, Bax, and Fas. These findings suggest that morphine may reduce growth of certain cancer cells through p53 activation. Additionally, morphine has been shown to inhibit expression and secretion of MMP-2 and MMP-9 in breast cancer cells in a time-dependent and concentration-dependent manner. This MMP activity is not reversible with NX, indicating that attenuation of MMP secretion by morphine is not mediated by opioid receptors, but is controlled by the NO system.

19Based on preclinical and clinical studies, differences in recurrence rates for certain cancers may be due to immune suppression and direct effects of volatile anesthesia and opioids on cancer growth. Overexpression of the μ-opioid receptor (MOR), which promotes tumor growth and metastasis, is observed in several human cancers. AKT and mTOR activation, cell proliferation, and extravasation are all related to MOR overexpression in a nude mouse model of non-small cell lung cancer (NSCLC). In addition, a potential direct effect of opiates has been observed in animal models that show MOR regulating tumorigenicity in Lewis lung carcinoma (LLC). Similarly, a study has shown the potential direct effect of opioids on MOR through growth factor-signaling proliferation, migration, and epithelial–mesenchymal transition during lung cancer progression. Treatment with methylnaltrexone (MNTX), a peripheral opioid antagonist, inhibits LLC invasion and anchorage-independent growth, whereas continuous MNTX infusion decreases primary LLC tumor growth and lung metastasis. Further, MNTX inhibits opioid-induced proliferation and migration of pulmonary microvascular endothelial cells through its effects on VEGF receptor phosphorylation and transactivation and inhibition of Rho A activation. Clinically, MNTX treatment is associated with increased overall survival in patients with advanced cancer; this finding supports the hypothesis that MOR is involved in tumor progression and that MNTX may target MOR. Because morphine reciprocally transactivates MOR and VEGF receptors, MOR-knockout mice do not grow significant lung cancer tumors; MNTX treatment markedly decreases tumor growth in experimental mouse models.

20Morphine at clinical blood concentrations stimulates proliferation and angiogenesis of microvascular endothelial cells by activating MAPK/ERK phosphorylation using Gi/Go-coupled G protein receptors and NO. Effects include apoptotic inhibition of apoptosis through AKT activation and promotion of cell cycle progression through increased cyclin D1. Morphine at clinically useful doses promotes tumor neovascularization and progression in a xenograft model of a human breast tumor. Similarly, clinical doses of morphine promote angiogenesis and tumor progression in ER-negative breast cancer cells in vitro and in vivo. Morphine is also able to stimulate in vitro vascular endothelial cell proliferation, which is mediated by the MAPK pathway. It is likely that MOR has an important role in angiogenesis and oncogenic signaling.

21Preoperative and postoperative morphine administration for analgesia decreases the tumor promotion surgical effects in a rat model. Preoperative and postoperative morphine treatment in rats significantly reduces surgery-induced corticosterone increases. This finding suggests that preoperative morphine may play a key role in protecting against surgery-induced metastasis. Intraoperative opioid use has been associated with increased overall survival in patients with stage I but not stage II or III NSCLC.

22Fentanyl has demonstrated antitumor-like effects in colorectal cancer cells in vitro. Its use is associated with decreased cell clone formation, and inhibition of cell migration and invasion through inhibition of negative regulation of E26 transformation–specific sequence-1 on serine/threonine kinase protein kinase B-raf (BRAF)-activated lncRNA. Another study has shown that fentanyl inhibits tumor growth and cell invasion in colorectal cancer by downregulating miR-182 and MMP-9 expression using β-catenin. A recent study showed that sufentanil does not affect the apoptosis rate or cell cycle distribution of colon and pancreatic cancer cells at clinical concentrations in vitro.

23Although benefits of using RA to avoid opioids have been suggested by clinical trials, it is unclear whether benefits result from withholding opioids or adding RA. Morphine administration may be beneficial for pain control, but MOR is involved in tumor progression for certain cancer cell types. Opioids may play a crucial role in cancer metastasis and recurrence, but this effect varies by cancer cell type. Prostaglandin E2, a soluble, tumor-derived angiogenic factor, is associated with VEGF-independent angiogenesis. PGE2 production in preclinical breast and colon cancer models is controlled by COX-2 expression, and COX-2 inhibition enhances VEGF blockade to inhibit angiogenesis, tumor growth, and metastasis to increase overall survival. Previous case control studies show that selective COX-2 inhibitors reduce breast and colorectal cancer risk, with the NSAID analgesic ketorolac being associated with a five-fold reduction in cancer relapse in the first few years after breast surgery. Because transient and systemic inflammation following surgery may be involved in metastatic tumor seeding and angiogenesis, perioperative antiinflammatory agents may be used to block those effects.

24Although local anesthetics suppress proliferation of several cancer cell types, their mechanism is unknown. Local anesthetics block voltage-gated sodium channels (VGSC), which are transmembrane proteins composed of one pore-forming α-unit and one or more auxiliary β-units. Cancer cells express an array of ion channels that their terminally differentiated counterparts do not. VGSCs are highly expressed and active in breast, colon, and lung cancers, and local anesthetics that cause channel blockade may inhibit tumor growth. In fact, lidocaine, ropivacaine, and bupivacaine, which inhibit proliferation and differentiation, are cytotoxic to mesenchymal stem cells (MSCs) in vitro, and have key functions for tumor growth and metastatic formation in cancer cells.

25Locally administered lidocaine directly inhibits epidermal growth factor receptor (EGFR), which is a potential target for anticancer drugs. Clinical concentrations of lidocaine have been shown to inhibit serum-induced and EGF-induced proliferation in human tongue cancer cells in association with tyrosine kinase activity of EGFR. One study that assessed the direct effect of local anesthetics showed that clinically useful concentrations of lidocaine and bupivacaine induce apoptosis in breast cancer cells in vitro and in vivo, suggesting a potential benefit of local anesthetics for breast cancer surgery. Lidocaine and tetracaine, which both inhibit kinesin motor proteins, reduce formation and function of tubulin micro-tentacles; thus, these drugs may have a novel ability to decrease metastatic spread in breast cancer cells. Lidocaine use at clinical concentrations results in DNA demethylation from ER-positive and ER-negative breast cancer cells in vitro. Although infiltrative anesthetics have the same membrane-stabilizing activity as lidocaine, they effectively inhibit the invasive ability of human cancer cells at the 5 mM to 20 mM concentrations used in surgery. Lidocaine additionally blocks human cancer cell invasion through modulation of intracellular Ca2+ concentrations and inhibition of ectodomain shedding of heparin-binding epidermal growth factor from cell surfaces. Furthermore, lidocaine, ropivacaine, and bupivacaine all reduce MSC proliferation at 100 μM concentrations by causing cell cycle delay or arrest at the G0/1-S phase; this feature is the reason why local anesthetics are used perioperatively for treatment of patients with cancer. In contrast, ropivacaine and bupivacaine do cause apoptosis and cell cycle distribution at clinical concentrations for colon and pancreatic cancer cells in vitro; their only antitumor growth activity occurs at high concentrations. Based on these findings, it is unlikely that the observed protective effects of RA on CMI result from direct effects on cancer cells. The overall effect of anesthetic agents on tumor development is summarized in Table 3.

26In general, cancer is considered as a systemic disease with circulating tumor cells and micro-metastases present at initial diagnosis. Surgery and anesthetic-induced immunosuppression activate HPA-axis and SNS responses, which in turn increase neuroendocrine mediators. These mediators promote metastasis to regional lymph nodes and distant sites from residual or circulating tumor cells, and stimulate growth of preexisting, dormant micro-metastases through immunosuppression (Fig. 2). During this process, cancer cells must escape immunoediting by NK cells and CTLs to establish themselves at distant sites and proceed to angiogenesis.

27Tumor dormancy, often described as “cancer without disease,” is the poorly understood phenomenon by which quiescent cancer cells exist but do not produce clinical disease. Distant recurrence appearing months or years after surgical resection have been described as dormant metastases, which are clinically undetectable, pre-existing disease foci that then become clinically detectable. Two potential explanations for tumor dormancy are 1) lack of angiogenic activity; and 2) immunologic equilibrium between tumor and host immunity, which prevents further tumor growth in the microenvironment. Because neuroendocrine mediators regulate tumor progression biology and act as endogenous angiogenesis modulators of reactivation from dormancy, HPA-axis and SNS neuroendocrine dynamics may be responsible for loss of tumor dormancy. Thus, surgery-induced and anesthetic-induced immunosuppression may promote cancer recurrence through HPA-axis and SNS activation during the perioperative period in patients with cancer.

28Currently available preclinical studies suggest that anesthetic-induced immunosuppression may promote cancer recurrence in patients with certain types of cancer. Volatile anesthetic agents and morphine or synthetic opioids produce diverse effects on cancer cells that depend on dose, duration, and timing of use. Nevertheless, locoregional anesthesia and propofol-based anesthesia seem to reduce surgical stress, perioperative immunosuppression, and angiogenesis compared to general anesthesia with volatile anesthetics and opioids. Although a causal link between anesthetics, immune function, survival, and residual disease remains to be elucidated, several ongoing prospective RCTs should provide more definitive information about the effects of anesthesia on cancer recurrence after surgery.