This essay has been submitted by a student. This is not an example of the work written by our professional essay writers.
There are a high number of patients with cancer seeking surgery for various reasons, including potential cure. However, the surgical stress response, seeding caused by surgery, and anesthesia medications can potentially promote cancer recurrence or metastasis. Here are presented the current literature on various anesthesia related drugs and techniques, and their potential influences on the immune system and/or cancer recurrence. Including are opioids, cyclo- oxygenase inhibitors, induction and volatile agents, and regional techniques. Through attention to the progressing database of information we can start to make alterations to effect a positive change in outcome for our cancer patients.
Introduction and Physiology
Cancers of all types occur with staggering numbers within the patient population presenting for surgery. Some patients need surgery for curative or palliative reasons related to their malignant pathology, while others requiring surgery and anesthesia have cancer as an incidental diagnosis. As the second leading cause of death in United States1, and with an ever-aging population, the number of patients with cancer who will require anesthesia care will only grow. As anesthetists, it is important to explore the implications our practice may have on the outcomes of these patients' cancer. Perhaps we can employ techniques that lessen the risk of cancer metastasis or recurrence during the perioperative period.
Manipulation of a tumor during resection, while the aim may be to cure the patient, can cause seeding into the bloodstream, serving as an impetus for metastasis.2 Even with surgery remote to a malignancy's location, the surgical stress response and associated immunosuppression can ostensibly give malignant cells an advantage over the weakened immune system, which otherwise might be able to eliminate them. The depression of the immune system after surgery occurs within hours, lasts several days, and occurs in a degree mirroring the extent of tissue trauma.3 Natural killer cells are especially inhibited, whose function is "to recognize and lyse cells lacking self human leukocyte antigen class I molecules," and are therefore our number one immunological defense against cancer.2 They also produce regulatory cytokines which elicit immune responses.4 Kutza et al4 found that natural killer cell cytotoxicity was significantly (p<0.001) reduced twenty four hours postoperative from their preoperative levels in twenty-six ASA 1-3 individuals without cancer undergoing various surgical procedures. The effect was seen even in surgeries lasting just forty-five minutes. A drawback of the study was that it did not show the duration of the natural killer cell suppression, as postoperative levels were only drawn once.
The role-players in the mechanism of immunosuppression postoperatively is thought to be threefold: the neuroendocrine system, the inflammatory system, and the hypothalamic-pituitary-adrenal axis.3 As for the neuroendocrine role, according to Gottschalk3, stress biomarkers epinephrine and norepinephrine are elevated in the perioperative period, which interact with Î²1 and Î²2 receptors expressed by tumor cells, increasing invasive potential.3 The inflammatory role is no less important. Inflammatory mediators such as cytokines, chemokines, prostaglandins, and cyclooxygenase are thought to "promote cancer progression through immunosuppression, resistance to apoptosis, and promotion of angiogenesis." These inflammatory mediators may be what accounts for the suppression of natural killer cells, as well.3 Finally, the hypothalamic-pituitary adrenal (HPA) axis, stimulated by pain, plays a role in causing immunosuppression and can lead to tumor proliferation.3
With an understanding of the physiology, and using what the current research shows, we can attempt to tailor anesthetic plans for our cancer patients which optimize their ability to achieve a lasting cure. The following discuss various components of anesthesia, and what the literature shows.
Prevention of pain is not only an paramount component of an anesthetic, but as mentioned above, also is important in preventing the HPA axis from being stimulated, which can perhaps negatively affect individuals' immune function. However, some research shows us that opioids themselves may cause, with unknown etiology, immunosuppression which could lead to cancer recurrence and/or metastasis, complicating our care.
The landmark study in 1995 by Yeager et al5 built upon previous rodent studies, and showed a clinically significantly (p<0.05) depressed natural killer cell cytotoxicity (NKCC) in two groups of healthy human subjects who received either low dose (0.025 mg/kg load followed by 0.015 mg/kg/h infusion for 24 hours) or high dose morphine (0.05 mg/kg, then 0.03 mg/kg/h). The decrease in cytotoxicity was seen at 2 hours post initiation of infusion in both groups, and remained decreased at 24 hours (at termination of infusion) in both group. At 48 hours (24 hours post infusion termination), the high-dose group remained significantly depressed, while the low-dose group was no longer. This study's weaknesses lie in the fact that they had a small study group (n=23), although it was large enough to show statistical importance. Also, it is hard to say if the results in healthy volunteers not having surgery or anesthesia can be extrapolated to patients with malignancies who are having surgery and receiving anesthesia. Also, results may be specific to morphine and not of opioids in general. One strength of the study is that they controlled for the possibility that the hospitalization itself could cause the decrease in NKCC, by giving two participants placebo saline infusions in a blinded fashion. Those participants did not have any decrease in NKCC.
Beilin et al6 studied natural killer cell cytotoxicity with two different anesthetic techniques (low vs. high dose fentanyl) among patients undergoing abdominal surgery. They had two subsets of participants, those with benign processes, and those with malignant. This made four groups (low dose malignant, low dose benign, high dose malignant, high dose benign). Strengths of the study are that it was prospective, randomized, and highly controlled (all were abdominal surgeries, all were ASA 1-2, none received blood products, and other medications were kept the same). Also, it studied the real group of interest (patients actually having surgery, and with malignancy). Statistical significance was determined by p<0.05. They found no differences between the benign versus malignant groups when the type of anesthesia (low vs. high dose fentanyl) was the same. NKCC was decreased from baseline in both low dose (regardless of disease process presence) and high dose groups at 24 hours. There was no difference in cytotoxicity between the two groups. However, at 48 hours, the low dose group had returned to baseline NKCC, while the high dose group's cytotoxicity was still significantly reduced. This suggests the prolonged decrease in NKCC is attributable to the high-dose fentanyl anesthetic technique, and that high-dose opioid based anesthetics may be detrimental to our cancer patients. Other studies have shown however that fentanyl does not cause suppression of natural killer cells.
Biki and others7 explored how anesthetic technique for open radical prostatectomy affected prostate cancer recurrence (as evidenced by increased PSA after the postoperative PSA nadir). In this retrospective analysis, the two groups were general anesthesia with epidural analgesia or general anesthesia with postoperative opioids (PCA morphine), and was comprised of patients who had surgery over a course of ten years. They were then followed for an additional 2.8 years, (for follow up range 2.8-12.8 years). After correcting for confounding factors, they found a staggering 57% lower rate of recurrence in the general plus epidural group compared to the general with postoperative opioids group. Of course, the results seen here do not solely reflect opioid usage, but also may reflect other advantages to regional anesthesia, as discussed later. In fact, by their own admission, total opioid usage was not measured, but only presumed to be lower in the group with epidurals. Weaknesses of the study are characteristic of retrospective studies, and include lack of randomization, non standardization of clinical care, and effects of unmeasured variables. There also were found to be differences between the groups, with the epidural group having slightly higher ASA scores, more complications, slightly shorter surgeries, and smaller fraction of patients with clear margins. All of the mentioned differences besides shorter surgical time however, would favor better outcomes for the non-epidural group.
Hand in hand with the discussion of opioids is a look at cyclooxygenase (COX) inhibitors, and their usefulness in the anesthesia care of a cancer patient. While affording one to decrease the amount of opioids used, studies show they also may help prevent cancer growth in themselves. Some of the studies shed additional light on possible mechanisms of opioid induced tumor growth, and how COX inhibitors may attenuate these effects.
Cyclooxygenase-2 serves as an enzyme for the production of prostaglandins (including PGE2), prostacycline, and thrombaxane via metabolism of arachadonic acid.8-10 PGE2 and other COX pathway products promote angiogenesis and tumor progession,10 and cells with high levels of COX-2 are resistant to apoptosis-inducing stimuli.8 Cyclooxygenase-2 is known to be over expressed by many cancerous and pre-cancerous cells, including pancreatic and colon cancers.8,10
Non steroidal anti-inflammatory drugs (NSAIDs) have been used for their anti-tumor properties, and in fact celecoxib is FDA approved for treatment of familial adenomatous polyposis, but some data has suggested that the efficacy of NSAIDs in treating cancer is not due solely to their COX inhibitory properties.8 Grosch et al8 tested the effects of both a COX-2 selective NSAID, celecoxib, and a COX-1 selective NSAID, SC560, on three different human colon cancer cell lines with varying levels of COX-2 expression. The tests were done both in vitro, and in vivo with mice. They found that both in vivo and in vitro, celecoxib, and to a lesser extenxt, SC560, had antiproliferative effects on tumor cells, regardless of their COX-2 expression. This suggests a mechanism (at least partly) independent of COX inhibition. They found through immunoassay that celecoxib, but not SC560, induced apoptosis in the tumor cells, irrespective of degree of COX-2 expression. Both celecoxib and (less so) SC560 also caused a block in the cell cycle of the tumor cells, evidenced by enumerating cells in each phase of the cell cycle by flow cytometry, again without regard to COX expression. Drawbacks to this study are that it was done in vitro and with mice, and no human subjects were used. Therefore translation to human models are only suppositional. It also does not explore the compounding effects of anesthesia or surgery, and so has limited direct application to our particular focus. Its relevance may lie in the fact that we as anesthetists are not likely to know whether our patients' tumor heavily expresses COX or not, but that COX inhibitors may still be a valuable part of our anesthetic, not only to augment analgesia, but to help fight the cancer. Of note, the amounts of NSAIDS used correlate to blood levels achieved after administration of 800mg of celecoxib,8 quite a high dose. The study was well designed, prospective, and highly controlled.
Wei et al9 studied the effects of COX-2 selective inhibitor celecoxib on vascular endothelial growth factor (VEGF), a potent mediator of tumor angiogenesis,9 with in vivo (mouse) and in vitro models. They found celecoxib decreased VEGF expression in a dose dependent fashion on pancreatic cancer cells. It achieved this through blocking the Sp1 protein binding site on the VEGF promoter, thereby reducing its transcription and expression. This study has many of the same strengths and weaknesses as the previously mentioned study by Grosch, and gives further support to the theory that COX inhibitors can be a very important part of our anesthetic plan to help decrease the metastasis and recurrence of cancer that may be stimulated by the surgical and anesthetic processes.
"Opioids at physiologically relevant concentrations promote angiogenesis in vitro, and breast cancer concentrations"10 in rodents, as seen in several studies. According to Farooqui et al,10 a key mechanism of this is the stimulation of nitric oxide (NO) and nitric oxide synthase (NOS) by the endothelium. NO stimulates COX activity, increasing the end products, including PGE2, an angiogenic factor.10 Farooqui explored the effects of celecoxib on morphine-induced tumor progression in a controlled experimental study with mice. Four groups of mice injected with mammary carcinoma were formed in a random process and were subjected to controlled conditions of light, temperature, feed etc. The groups were normal saline (control), morphine, celecoxib, and morphine + celecoxib. The effects on tumor PGE2 levels, angiogenesis (by qualitative and quantitative measures), tumor growth (weight) and metastasis, and mice survival at 14 days were measured. Pain was also measured in the mice using latency of paw withdrawal from a heat stimulus as a measure. One result they found was a fivefold increase in COX-2, accompanied by a five-fold increase in PGE2 in the morphine group vs. control (p<0.05). The morphine + celecoxib group lacked this morphine induced increase. The morphine group of mice tumors had increased angiogenesis (higher density, dilated, more branches) than control. The morphine + celecoxib group had significantly less angiogenesis than the morphine alone group. The celecoxib alone group was not significantly different than the control. Morphine treated mice also had significantly larger tumor weights (p<0.001) than control, while morphine + celecoxib weights were significantly lower than control mice (p<0.001). Celecoxib-only treated mice had tumors with no difference in weight from control. 100% of the morphine-only treated mice had lung metastasis upon completion of the study, while 70% of control mice did, while only 30% of celecoxib alone and 25% of morphine + celecoxib. Survival to day 14 was lowest with the morphine group (70% dead at day 14), compared to 30% of control, and 20% in combination therapy group. Surprisingly, the celecoxib group had worse survivability (50% dead at day 14) than control. This was not thought to be related to the tumor progression however, as many of these mice died early on, when tumors were barely palpable and non-metastatic, suggesting a toxic etiology. Further experimentation revealed "when administered together even at higher doses, morphine and celecoxib blocked the adverse effect of the other, resulting in increased survival."10 With regard to analgesic effects, the morphine treated group had increased paw withdrawal latency (decreased pain) at 5 days of treatment, but no longer at days 10 or 14. The combination therapy group however retained analgesic effects throughout the duration of the 14 day experiment (p<0.0001). Limits to this study are again, the use of mice, and not using the context of surgery and anesthesia. However, it was a very highly controlled study, with many important outcomes measured, and had highly significant p values. Overall, it gives convincing evidence of the use of COX-2 inhibitors in combination with opioids in the treatment of cancer patients, which can be carried over to our treatment of pain intraoperatively.
Forget et al,11 in their 2010 study, examined the effects of different perioperative analgesia techniques, including sufentanil, NSAIDs, ketamine, and clonidine effects on postoperative cancer recurrence. Theirs was a retrospective study of 327 breast cancer patients who underwent unilateral mastectomy with axillary node dissection at one institution over 5.5 years. In addition to surgery, other treatments were given based on standards of practice. All surgeries were performed by the same surgeon, and followed henceforth by said surgeon and the same oncologist. All patients were induced with sufentanil 0-0.2 mcg/kg, and thiopental (4 mg/kg) or propofol (2-3 mg/kg), and maintained with propofol infusion plus sevoflurane or desflurane in oxygen/air mixture. Intraoperative analgesics were up to the anesthesiologist, and consisted of sufentanil 0-0.5 mg/kg, preincisional clonidine 0-6 mcg/kg, preincisional ketamine 0-0.5 mg/kg, and ketorolac (a non selective COX inhibitor), either not administered, or given in a dose of 20 mg for patients <60 kg or 30 mg for patients >60 kg. Postoperatively, all patients received acetaminophen 3-4 g/day in the first 48 hours, and diclofenac 50 mg TID prn for the first 3 days. The end point used was length of recurrence free survival through February 2009. A multivariate analysis was used to adjust for confounding baseline factors and a p<0.05 was considered significant. Patient characteristics were the same between groups with different treatments, except age was lower in ketorolac treated groups. Cancer recurrence was found to be significantly more frequent (17% vs. 6%) in patients not treated with ketorolac compared to those who were. When adjusted for age, tumor histology, and lymph node involvement, the recurrence remained significant, p=0.019. There were no differences seen in patients with regards to treatment with sufentanil, clonidine, ketamine, or others drugs. This supports the use of intraoperative NSAIDS for our cancer patients. The retrospective nature of this study gives it inherent limitations. There are likely unrecognized biases and uncontrolled factors. It did examine actual humans undergoing surgery and anesthesia. These things do not intrinsically make it a good study, but they do help us to understand more closely the effects we may have on our patients as anesthetists.
Induction Agents and Volatile Anesthetics
In their landmark study, Woods and Griffith12 studied the effects of inhaled anesthetics on natural killer cell cytotoxicity in vitro. Mononuclear cells extracted from healthy humans and suspended in radioactive-chromium labeled tissue culture medium were used. Culture trays were then placed in air-tight containers with exposure to halothane, enflurane, nitrous oxide, nitrous oxide+ enflurane, or control gas for four hours at a controlled temperature. Mass spectrometry was used to confirm gas concentrations. NK cell cytotoxicity was measured by the amount of chromium released (and therefore cell lysis). The experiments were done in triplicate. They also studied the reversibility of any effects of the gases by doing another experiment where cells incubated in the gases for several hours, were then allowed to equilibrate for an hour with air, and then checked for cytoxicity. Their results showed a dose dependent reduction in NK cell cytotoxicity, with halothane and enflurane being equally depressant on activity, and nitrous oxide being less so, at equipotent doses. Additive effects were seen with nitrous+enflurane. The reversibility studies showed that the reduction in cytotoxicity occurs only during the exposure to the gases, and not after. Strengths of the study are its experimental design and controlled nature, the fact that it was performed in triplicate, and its in-vitro design studied the ability of the gases themselves to cause immunosuppression via NK cell inhibition, without confounding factors. This last fact can also be seen as a drawback of the study. In vitro studies may not be transferrable to what happens in vivo, as there are many other variables involved. Halothane and enflurane are no longer widely used, and there could also be questions raised about whether currently used gases have the same qualities.
Further studies have corroborated the results of Wood and Griffith, including a study by Markovic and others,13 which evaluated both halothane and isoflurane, a gas that is currently in wide use. The study has not been critically evaluated in the current literature review.
Melamed et al14 did rat experiments, investigating metastesis (lung tumor retention) after IV injection of tumor cells and NK activity following anesthesia with ketamine, thiopental, halothane, or propofol. They found ketamine to be the most deleterious to the subject, causing over a 5.5 fold increases in lung tumor retention. The thiopental group also had an increase in tumor retention. (p<0.0001) When looking just at NK cell activity, ketamine, halothane, and thiopental (but not good controls. propofol) caused a significant suppression of NK cell activity. (p= 0.02). This study was highly controlled, and had an experimental design. The results were obtained from in vivo studies, which are one step closer to real-life scenarios than in vitro studies. However, the rodent subjects again leave the question of the ability to extrapolate to humans or not.
In the above study, Melamed showed that propofol did not reduce NK cell activity. There may be another reason why propofol could be a good anesthetic to use for cancer patients, too. In 2009, Inada15 and others examined the effects of propofol on prostaglandin E2 and thromboxane B2 production in a human monocyte cell line. As discussed above, PGE2 and TXB2 are products of the COX pathway, which uses arachadonic acid (AA) as a substrate, and is proinflammatory in nature. These products are believed to promote tumor progression via angiogenic mechanisms. Inada used the monocyte cell line THP-1, and stimulated the cells with lipopolysaccharide to induce the AA pathway, in either the presence or absence of propofol (at concentrations that equate clinically relevant plasma concentrations for moderate to deep sedation.) After 18 hours, they were analyzed for amount of COX expression and amount of PGE2 and TXB2. They found that propofol (in all concentrations tested) significantly (p<0.001) reduced amount of TXB2 and PGE2. They importantly did not achieve this through changing viability of the cells, as there was no increase in necrotic cells. Also, COX expression was not changed, leading the experimenters to conclude that the mechanism was instead via suppression of the COX enzyme activity. This study shows that propofol may be of benefit for use in our cancer patients requiring anesthesia for surgery. The in vitro design limits the study's relevance for what happens in vivo. Also, as cited by the authors, they used the human monocytic cell line rather than human monocytes. The two types of cells may have different physiology owing to mutations.
Biki's7 study is discussed under "opioids", and was an analysis of outcomes from radical prostatectomy patients who received either general anesthesia + postoperative parenteral morphine, or general anesthesia + epidural. They found a 57% lower rate of recurrence in the general+epidural group. Discussed above was the possibility that this was due to lower total amounts of opioids given to the epidural group, bypassing the potentially detrimental effects which opioids have on the immune system and cancer recurrence. Biki also proposes that regional anesthesia may be of benefit to the cancer patient perioperatively because it blocks afferent transmission and subsequent efferent SNS activation by the surgical stress response, which also depresses cell mediated immunity (including NK cells) and is proangiogenic. Thirdly, regional anesthesia may also allow lower levels of general anesthesia, which as discussed above may be immunosuppressive and therefore detrimental to the cancer patient.
Exadaktylos et al16 retrospectively examined the effects of anesthesia technique (general+ postoperative PCA opioid versus general+paravertebral block) on breast cancer recurrence in 129 patients undergoing mastectomy and axillary node dissection. They excluded patients with low risk for recurrence. Follow up time from surgery was 27-37 months. As it was retrospective, care was not standardized, and patients were not randomized, which are limitations to the study. There are likely other factors that play into the results that are not accounted for. The general anesthesia group had slightly larger tumors, smaller margins, and higher chemotherapy rates, although none of these variables reached statistical significance. They found a four-fold reduction of recurrence or metastasis in the paravertebral group over the 2.5 to 4 year follow up period. Results are specific to breast cancer, and should probably not be generalized to all cancer types.
A study by Wada17 and others used mouse subjects, subjected them to laparotomy with either general anesthesia (with sevoflurane) or general anesthesia (sevo) +spinal (bupivacaine and morphine). In one experiment, the subjects were injected with liver-metastatic cells immediately postoperatively, and hepatic metastases were then counted 12 days later. Laparotomy with general anesthesia increased the number of liver metastasis from a 15.5 average in control group to 33.7. This increase was attenuated by the addition of spinal anesthesia: the laparotomy group with general plus spinal had an average number of metastases of 19.8 (p<0.05). In another experiment, the subjects were not injected with tumor cells, but rather their livers were removed and examined 5 hours postoperatively to assess the degree of cytotoxicity of liver mononuclear cells (MNC, including NK cells). The cytotoxicity was reduced to 18.5 after surgery with general anesthesia, from the control of 32.7 (p<0.01). In the group who had surgery with general and spinal, liver MNC activity was maintained closer to control, with a level of 26.6. There was a statistical difference between the general alone (18.5) and general+spinal (26.6), p<0.01. Strengths of this study are its highly experimental, controlled design. It also shows in vivo effects. However, once again, it does not deal with human subjects and may have limited extrapolation. Also, it is specific to liver metastasis, and again, results cannot necessarily be generalized to other types of cancers.
In contrast to the previous studies, there are two recent studies,18,19 which do not show convincing benefit of epidural anesthesia on cancer recurrence. Wuethrich et al18 did a retrospective review of 261 patients who had either general anesthesia+thoracic epidural (TEA), or general anesthesia+postop IV analgesia (consisting of ketorolac 30 mg q8h, paracetamol 1000 mg q6h, and morphine IV prn) for radical prostatectomy. The follow up time was 8.5 to 11.9 years. They found a significant difference in clinical progression-free survival between the two groups, with the benefit going to the general/TEA group. However, they found no difference in biochemical recurrence-free survival (PSA), or in either cancer-specific or overall survival between the two groups. There are a lot of problems with the study, which may temper its validity. First, it was retrospective and non-randomized. This gives the possibility that selection bias occurred. In fact, there were baseline characteristic differences between the two groups. The general/IV analgesia group had a significantly lower ASA score, and received less blood products than the general/TEA group. There are likely also confounding variables. One that easily comes to mind is the standard use of ketorolac in the IV group, where none was used in the general/TEA group. As discussed above, perioperative NSAID use may improve outcomes for cancer patients.
Gottschalk et al19 also conducted a retrospective study, of 448 patients who underwent surgery for colorectal cancer at a single institution with either epidural or no epidural perioperatively. Average follow-up time was 1.8 years, and p<0.05 was considered significant. They found no difference in cancer recurrence between the two groups, with non-epidural recurrence being 16%, and epidural recurrence reaching 13%. When other variables were included in the analysis, they did find significantly better outcomes for the epidural group for patients >65 years. Per the authors, this may suggest a difference in tumor type which presents in each age. Weaknesses of the study include its retrospective nature with lack of randomization and lack of control and/or documentation of other variables. The author19 also notes that they "cannot determine whether analgesia in the epidural group was sufficient and similarly cannot determine whether the epidural infusion was stopped in the perioperative period." Also, the time period used for follow-up was relatively short. Perhaps a longer follow up time or a larger sample group would have brought the slight difference seen (16% vs 13% recurrence) to statistical significance.
Many human tumor cell lines express Î²1 and Î²2 receptors, and catecholamine agonism to these receptors effects migration of carcinoma cell types and secretion of proangiogenic factors, thereby promoting cancer growth and spread.20 Benish20 and others studied the effects of the nonselective beta blocker propanolol on lung tumor retention (LTR) in mice injected with a cell line that metastasizes to lung tissue. They used 88 mice, divided into 6 groups. Three groups underwent laparotomy under halothane anesthesia with either 0, 1.5, or 4.5 mg/kg propanolol injected one hour before surgery. The other three groups were nonsurgical controls, given the same three concentration of propanolol. They found surgery caused a significant increase in LTR over the nonoperative controls, but that both doses (1.5 and 4.5 mg/kg) of propanolol decreased the effects by 50% (p<0.05). In a different experiment, they tested the effects of the combination of the selective COX-2 inhibitor etodolac with propanolol, and each drug separately, against no treatment on LTR in mice undergoing laparotomy under halothane. There were also non-operated controls given the same treatments (no drug, propanolol only, etodolac only, or combination). They once again found that surgery increased lung tumor retention over non surgical groups. Both etodolac and propanolol alone attenuated this effect significantly, by about 50% (p<0.05). The combination treatment group also had significant decreases in LTR compared to propanolol alone (p<0.05), and nearly significant decrease compared to etodolac alone (p=0.058). This study was well designed experimentally, and very controlled. Its drawback is the mouse model.
Blood transfusions are known to have immunosuppressive effects, first recognized when renal transplant patients who received blood transfusions had fewer rejections.21 There are therefore potential detrimental effects to cancer patients who receive blood transfusions by similar immunosuppressive mechanisms. A Cochrane Database meta-analysis21 evaluated data from 36 studies involving 12,127 patients. Inclusion criteria were patients undergoing curative resection for colorectal cancer and had received blood products within 1 month of surgery. They found a pooled result of perioperatrively transfused patients to have an odds ration of recurrence of 1.46 as compared to non transfused patients. Although causality cannot be stated, the enormity of the data provides convincing evidence of detriment by perioperative blood transfusion in colorectal patients.
The current literature on anesthesia techniques and cancer recurrence is neither complete nor completely in agreement. Additionally, besides the cancer, other patient and surgical factors will play into the anesthetic plan. Continued research in needed in this area before drastic changes in anesthesia practice are warranted for cancer patients. Specifically, more human based randomized controlled prospective studies are needed. However, keeping in mind the current literature, the anesthetist can consider tailoring the plan to include techniques which may help decrease the recurrence risk for their patients. There is certainly not enough evidence to preclude opioid use wholly in the anesthesia of a cancer patient, but the current literature should cause one to take pause and consider their use in a prudent manner with this patient population. Additionally, use of COX inhibitors in appropriate populations seem an effective, easy way to both decrease opioid usage perioperatively and provide benefit of their own. Propofol is perhaps a good choice for induction agent, and TIVA may even be considered if appropriate. Our inhaled anesthetic gases, which may contribute to immunosuppression, seem to do so in a very time limited fashion and much more evidence against them would be needed before avoiding their use as a general rule. If regional anesthesia is an option based on patient and surgery planned, its use should be heavily considered as it may be of benefit in certain types of curative cancer resection. One must keep abreast of the new and changing base of literature, and tailor practice as appropriate to the benefit of one's patient.