Is it possible that particulars of perioperative anesthetic management (eg, using epidural local anesthetics during general anesthesia) might influence long-term recurrence risk after cancer surgery? Not too long ago, this would have been considered an impossibility, and at first glance, it does seem an unlikely proposition. How could a change in management during a procedure that lasts at most several hours lead to effects on cancer outcome years later? Yet, suggestive (though far from conclusive) retrospective clinical data indicate that there might be an effect, and even a significant one. Use of paravertebral blocks¹ or the addition of ketorolac² in breast cancer, or the addition of epidural anesthesia in prostate³ or colon cancer⁴ might reduce risk of recurrence. The story is far from clear, however, as other studies^5,6 do not indicate benefit. Prospective randomized trials are in progress, but it will be many years before the results of these will be known.

Much more data are available at the basic science level, and in this chapter we therefore will focus on some of this evidence, discussing, where possible, animal rather than cellular data. Before reviewing the effects of specific perioperative interventions, however, it is necessary to discuss briefly the mechanisms by which metastasis occurs in the perioperative period. Then we will describe the main effects of anesthetic drugs on these mechanisms. Although outside the domain of pharmacology, we will also briefly touch on potential effects on cancer recurrence of some other perioperative events, such as hypovolemia, anemia, transfusion, and hypothermia.

Why does cancer recurrence happen at all, after what was supposed to be curative surgery? This is a highly complex issue, and here we can provide only a brief and admittedly simplified description of the process.

Two main mechanisms explain why recurrence happens. The first mechanism is that cancer cells are released into the circulation during surgery. This is well documented,⁷ unavoidable, and takes place even if a “no touch” technique is used for surgical resection. Pressure on the tumor will result in malignant cells being disseminated through the body. Tumor cell release into the circulation happens also in the absence of surgery, but the body has effective defense mechanisms that prevent (almost) all of these cells from surviving. Critical here are natural killer (NK) cells, which have the ability to recognize tumor cells and eliminate them.⁸ It follows that decreases in the number or activity of NK cells induced by perioperative interventions might increase the number of cancer cells surviving in the body. Indeed, it has been shown for a variety of tumor types that low levels of NK cell activity predict worse outcome after cancer surgery. For example, NK cell activity levels less than 20% correlate with poor survival after “curative” colon cancer surgery.⁹

The second mechanism is the immune suppression induced by surgery, some drugs, and some perioperative events. The role of the immune system in cancer is exceedingly complex and only partially understood. Only the briefest of outlines can be given here. Even when seemingly a solitary tumor is identified by the best diagnostic imaging techniques, micrometastases are frequently present. These are small collections of tumor cells, and they are unable to grow into full metastases for 2 reasons. First, they are kept under control by the immune system, in a process known as “immune surveillance.”¹⁰ Second, their growth is limited by the absence of a supply of oxygen and nutrients. If a tumor is to grow beyond the millimeter stage, it can no longer depend on diffusion for its nutrient supply and has to “recruit” blood vessels from the surrounding tissue, inducing them to grow into the tumor stroma. In order for this to happen, angiogenic factors need to be released by the tumor. This would suggest that perioperative events that suppress immune functioning, or that are angiogenic, would lead to the “escape” of micrometastases and their development into larger tumors.

As we will see in subsequent sections, the hypotheses stated here are essentially correct. Decreases in NK cell number and activity, decreases in cellular immune functioning, and angiogenic stimulation appear to be the main pathways by which perioperative management influences cancer recurrence. NK cell number and activity assays are easily performed and can be done in humans. Animal studies allow further direct investigation. Animals can be subjected to surgery, and tumor cells can then be injected into the bloodstream and their fate assessed. If cells are labeled, retention in the lungs can be measured, as an indicator of the ability of tumor cells to survive in the blood stream and to migrate into tissues. Alternatively, investigators can let the animals survive for several weeks after tumor cell injection and then count the number of metastases as well as the total tumor load.

These are the main techniques used in the studies that will be described in the subsequent sections. They allow us to answer the question posed in the beginning of this chapter: “can perioperative management affect cancer recurrence?” The answer is an unequivocal “yes” – at least in animals. Figure 24–1 provides an example.¹¹ In this study, rats were subjected to laparotomy under halothane anesthesia, and MADB106 tumor cells were injected. As shown, the stress of surgery/anesthesia approximately tripled the number of metastases. So, the surgical setting itself is a major stimulus for cancer progression—and largely unavoidable. But other perioperative management also has significant impact. As shown, the combined administration of indomethacin and nadolol decreased the number of metastases almost by half.

In this section, we will review several classes of drugs used in the perioperative period for which evidence suggests a potential influence on tumor recurrence. These will include the typical anesthetic drugs, as well as some other compounds used commonly in the perioperative setting.

Induction Drugs

Thiopental, ketamine, and propofol are the main induction drugs for which data are available. Etomidate has not been studied in this setting, although its effects on steroid synthesis and the expected subsequent influence on the immune system suggest that it might be a worthwhile topic of investigation.

Of these drugs, propofol is by far the one used most commonly, and as it turns out, it is also likely to be the most beneficial drug in the setting of cancer surgery. Thiopental has immune suppressant effects and greatly reduces NK cell number and activity. Ketamine significantly increases lung retention of injected tumor cells.¹² Propofol showed fewer effects on NK cell number and activity than ketamine or thiopental and had no effect on lung tumor retention. In fact, when administered for longer periods of time, propofol may have protective actions in the setting of cancer. When mice received daily intraperitoneal injections with propofol for 3 days and were inoculated with tumor cells, their T cells showed increased tumor cell killing activity, and tumor load 28 days later was decreased by more than half as compared with animals who received saline injections.¹³ The invasion ability of HeLa, HT1080, HOS, and RPMI-7951 cancer cell lines was reduced to negligible amounts by propofol 5 mcg/mL (which is approximately the upper concentration limit obtained in blood during propofol-induced general anesthesia), and continuous infusion of propofol (40 mcg/kg/d for 4 weeks) in mice reduced by about 50% the metastatic load after inoculation with LM8 tumor cells.¹⁴ Taken together, these data suggest that our usual choice of propofol as induction drug is appropriate in the setting of cancer surgery and that there may be potential additional benefits to longer-term propofol administration.

Volatile Anesthetics

For volatile drugs, unfortunately, the verdict is not as good. Although some variability exists between studies, the inhaled anesthetics generally have been found to increase tumor load in animal models. Halothane and isoflurane, for example, administered for 1.3 minimum alveolar concentration (MAC)-hour in mice, each more than doubled the number of metastases counted 21 days after intravenous administration of melanoma cells.¹⁵ There are little data available on sevoflurane and desflurane. Nonetheless, the available basic science data suggest that of the compounds used routinely in anesthetic practice, the inhaled drugs may be some of the least appropriate in the setting of cancer surgery.

Muscle Relaxants

There are no data available on the ability of muscle relaxants to affect cancer recurrence. There also is no physiologic rationale as to why they would affect any of the parameters relevant to perioperative metastasis. It seems unlikely that muscle relaxants would significantly affect cancer recurrence.