Hypothermia in Barbiturate-anesthetized Rats Suppresses Natural Killer Cell Activity and Compromises Resistance to Tumor Metastasis : A Role for Adrenergic Mechanisms

Fischer-344 male rats (Harlan Laboratories, Jerusalem, Israel) were housed four in a cage with free access to food and water. Animals were acclimatized to the vivarium for at least 4 weeks before the beginning of experiments and were 13–19 weeks old at that time. In any given experiment, all animals were of the same age. All experiments were approved by the Institutional Animal Care and Use Committee of Tel Aviv University.

Anesthesia was induced by intraperitoneal injection of 70 mg/kg sodium thiopental (Abbott S.p.A, Milan, Italy). Hypothermia was induced by positioning anesthetized rats on cotton-covered ice packs. Normothermia was maintained at 38°C (the normal rat body temperature) by positioning anesthetized rats on electric heating pads. Core body temperature was monitored using a 4-cm rectal probe (400 series; YSI Inc., Yellow Springs, OH). Anesthesia lasted for approximately 6 h in the normothermia group and 7 h in the hypothermia group.

Fluorescence-activated cell sorter analysis was used to determine the number of NK cells in the blood. These cells were identified as NKR-P1^brightlymphocytes. 35To determine the total number of NK cells per milliliter of blood, a fixed number of polystyrene microbeads (20 μm; Duke Scientific, Palo Alto, CA) were added to the blood samples before they were prepared for cytometric analysis. For further details, see Shakhar et al.  29 

The activity of NK cells was assessed using the whole-blood cytotoxicity assay described in detail elsewhere. 29Briefly, this procedure analyzes antitumor cytotoxicity of NK cells per milliliter of blood without previous purification of peripheral-blood mononuclear cells. One milliliter of blood is drawn by cardiac puncture, plasma is replaced with culture medium, and aliquots of the washed blood are placed in a microtiter plate. To assess NK cytotoxicity at six different effector-to-target ratios, six concentrations of radiolabeled YAC-1 tumor cells were added to the blood. Spontaneous and maximal releases of radioactivity from target cells are determined by substituting blood with the culture medium or hydrochloric acid, respectively. After a 4-h incubation period, samples of supernatant are recovered from each well so radioactivity could be assessed and specific target cell lysis could be calculated. Earlier studies indicated that cytotoxicity measured using this procedure is attributable to NK cells rather than to other cell types or soluble factors. 36,37 

MADB106 is a selected variant cell line obtained from a pulmonary metastasis of a mammary adenocarcinoma (MADB100) chemically induced in the inbred Fischer-344 rat. 14MADB106 cells were maintained in complete medium at 100% humidity, 5% carbon dioxide at 37°C, and detached from the flask using 0.25% trypsin.

To assess lung tumor retention, DNA radiolabeling of tumor cells was accomplished by the addition of 0.4 mCi/ml of ¹²⁵iododeoxyuridine (ICN Radiochemicals, Irvine, CA) to the growing cell culture 1 day before the cells were harvested for injection.

For tumor cell injection, rats were lightly anesthetized with halothane (Trofield Surgicals, London, UK), and 4 × 10⁵/kg ¹²⁵iododeoxyuridine-labeled MADB106 tumor cells in approximately 0.5 ml phosphate-buffered saline (supplemented with 0.1% bovine serum albumin) were injected into their tail veins. Nine hours later, rats were killed with halothane, and their lungs were removed and placed in a gamma counter for assessment of radioactive content. The percentage of tumor cells retained was calculated as the ratio of radioactivity measured in the lungs to total radioactivity in the injected tumor cells. Our previous studies have indicated that the level of lung radioactivity reflects the number of viable tumor cells in the lungs (for more information on this, see Ben-Eliyahu et al.  13).

Rats were lightly anesthetized with halothane, and 10⁵MADB106 tumor cells were injected into their tail veins in 0.5 ml phosphate-buffered saline. Four weeks after tumor inoculation, rats were anesthetized, and lungs were removed and placed for 24 h in Bouin solution (72% saturated picric acid solution, 23% formaldehyde [37% solution], and 5% glacial acetic acid). After being washed in ethanol, visible surface metastases were counted by a researcher blinded to the donor of each lung. When the rats were killed, no signs of respiratory distress, body weight loss, or any noticeable illness-related behavior were observed. With the anticipated number of lung metastases, rats are not expected to show such signs before 8 weeks. 13 

In a preliminary study, we developed the procedure for inducing hypothermia. Core temperature was monitored continuously until rats emerged from anesthesia. No immune measures were studied in this initial phase.

All further experiments were devoted to the study of immune measures related to NK activity. Rat temperatures were taken only three times: immediately after anesthesia and 1 and 2.5 h into the procedure. These measurements were conducted to verify adherence to the expected course of changes in core temperature.

In all experiments, rats were allocated to one of three groups: control, normothermia, and hypothermia. Rats from the control group remained awake and undisturbed in their home cages. Rats from the normothermia group were anesthetized and kept for 2.5 h at normal body temperature, as described before. Rats from the hypothermia group were anesthetized and their temperatures were reduced for 2.5 h, as described before. After this 2.5-h period, rats from the latter two groups were returned to their home cages. All immune- and tumor-related measures were taken from this time point (T⁰) and onward, and their timing is recorded in relation to T⁰.

In the first experiment, blood was drawn by cardiac puncture at 2 or 6 h to assess the activity and numbers of circulating NK cells (n = 54; 9–14 per group).

In the second experiment, rats were injected with radiolabeled MADB106 tumor cells at T⁰, or at 3 h. Lungs were removed 9 h after tumor injection so tumor retention could be assessed (n = 48; 9–15 per group).

In the third experiment, rats were injected intravenously with MADB106 cells at 2 h; 4 weeks later, lungs were removed and the number of metastases was counted (n = 57; 19 per group).

In the fourth experiment, each of the three experimental groups was subdivided and treated with saline or nadolol, a β-adrenergic antagonist. 38A first subcutaneous injection of the drug (0.4 mg/kg) or saline was given at the induction of anesthesia (−2.5 h), and a second identical injection was given together with radiolabeled MADB106 cells at 2 h. Lungs were removed so that tumor retention could be assessed 9 h after tumor injection (n = 73; 8–18 per group).

Natural killer cell activity was analyzed using repeated-measures analysis of variance for the different effector-to-target ratios and an α error of 0.05. In all other cases, analysis of variance was not used because markedly different variances characterized the different groups, thereby violating an analysis-of-variance assumption. Instead, we used unpaired Student t  tests to compare groups and the Bonferroni α-correction procedure to compensate for multiple comparisons. Thus, α^correctedwas set at 0.015, allowing for as many as three independent comparisons ([1 −α^corrected]³> 95%). P  values indicate the significance levels of each t  test. Lytic units were calculated using the formula 100/ET³⁰, where ET³⁰is the effector-to-target ratio needed to lyse 30% of target cells. The regression exponential-fit method 39was used to infer ET³⁰from the data.

In the hypothermia group, the average target core temperature of 30°C was achieved within 50 min and began to increase 40 min later, to 32°C by T⁰. Normal body temperature was regained gradually within 5 h of T⁰(fig. 1). The body temperature of normothermic animals remained at approximately 38°C before T⁰and typically decreased thereafter by as much as 1°C until rats had emerged from anesthesia completely. Hypothermic treatment was uniform in all subsequent experiments and resulted in very similar changes in body temperature.

At 2 h and 6 h, hypothermia significantly suppressed NKCA compared with levels in the control (P < 0.0001) and the time-matched normothermic (P < 0.05) groups (fig. 2A). Normothermia significantly suppressed NKCA at 6 h (P < 0.05), but not at 2 h (P = 0.068). Because NKCA at the two time points was similar in each group (fig. 2A), they were combined for further analysis using lytic units. Normothermia suppressed NKCA (expressed as the mean lytic units at 30% specific killing [LU³⁰]± SEM) to 39% of control levels (from 49.30 ± 10.30 in the control group to 19.08 ± 3.03;P < 0.003). Hypothermia further suppressed NKCA to 15% of control levels (to 7.41 ± 3.24;P < 0.015)(fig. 2B).

At the two time points when blood was drawn to assess NK activity, NK cell numbers were similar in all groups, ranging from 305 to 350 NK cells/μl (fig. 2C).

After tumor injection at T⁰, normothermia increased the percentage of MADB106 tumor cells retained in the lungs from 0.40 ± 0.05% in the control group to 0.64 ± 0.09%(P = 0.024). This effect completely dissipated within 3 h (fig. 3A). Hypothermia, on the other hand, caused more than a twofold increase in lung tumor retention at T⁰(as much as 1.01 ± 0.23;P < 0.004) and at 3 h (as much as 0.94 ± 0.16%;P < 0.001). The increase caused by hypothermia compared with normothermia was also significant at T⁰and 3 h (P < 0.01 and P < 0.002, respectively).

The average number of metastases identified in the hypothermia group was 4.68 ± 1.27, which is more than three times greater than in the control and normothermia groups, in which the metastases numbered 1.10 ± 0.4 and 1.47 ± 0.39, respectively (one-tailed test;P < 0.006 and P < 0.011, respectively)(fig. 3B).

Importantly, only 21% of the hypothermic animals were free of metastases, whereas in the normothermic and control groups, 36.8% and 52.6%, respectively, were free of metastases.

Normothermia and hypothermia significantly increased lung tumor retention, from 0.53 ± 0.05% in the control group to 1.30 ± 0.16% and 1.57 ± 0.22%, respectively (fig. 4). Although the effect of hypothermia was larger than that of normothermia, this difference was not statistically significant in this experiment. Nadolol attenuated the effects of hypothermia, keeping tumor retention at 0.93 ± 0.07%(hypothermia saline vs.  hypothermia nadolol;P = 0.0055). This attenuation was not statistically significant in the normothermia condition (normothermia saline vs.  normothermia nadolol;P = 0.0329).

To the best of our knowledge, this is the first study to systematically assess the effect of hypothermia under anesthesia on NK activity or tumor metastasis. Our findings show that 2.5 h of hypothermia during thiopental anesthesia markedly suppressed NK activity and promoted metastasis. Specifically, normothermic rats showed less suppression of NKCA than did hypothermic rats. This suppression occurred at the level of individual NK cell activity, because the numbers of NK cells per milliliter of blood were unaffected. Correspondingly, the retention of MADB106 tumor cells in the lungs and the number of lung metastases were significantly increased by hypothermia compared with rats in the control and the normothermia groups. The effects of normothermia (thiopental anesthesia alone) were smaller and reached significant levels only on one of the four occasions that the MADB106 was used. Finally, nadolol, a β-adrenergic antagonist, significantly attenuated the effect of hypothermia on lung tumor retention.

Overall, these findings suggest that the suppression of NKCA by hypothermia is robust and long-lasting and has consequences for the health of the animal. Specifically, NKCA was still markedly suppressed at 6 h, and lung tumor retention, reflecting cumulative clearance of tumor cells from 3 to 12 h, was greatly increased. This latter effect seems resistant to factors that influence tumor development at later stages, because lung metastases, counted 4 weeks after hypothermia, were four times as numerous in the rats in the hypothermia group compared with those in the control or normothermia groups. Considering the high sensitivity of the MADB106 to NKCA 12–14and our previous findings of parallel effects of various manipulations on NKCA and on resistance to MADB106 metastasis, 29,40the suppression of NKCA by hypothermia is likely to underlie the increase in MADB106 metastases.

The suppression of NKCA by hypothermia seems unrelated to the cooling of NK cells themselves or to the cooling of other leukocytes. The blood samples from all groups were kept at room temperature (23°C;i.e. , “hypothermia”) for approximately 3 h during the preparation for the assessment of NKCA. Natural killer cell cytotoxicity was then assessed in all samples at 37°C. Therefore, it appears that physiologic responses secondary to hypothermia are involved in the suppression of NKCA. Possible candidates are the various hormones and cytokines released after hypothermia, most prominently catecholamines, and to a lesser degree glucocorticoids and other hormones. 27,28 

It is our hypothesis that adrenergic discharge plays an important role in the suppression of both NKCA and resistance to MADB106 metastasis observed in the current study. Anesthesia and surgery have been found to increase epinephrine levels. Similarly, hypothermia, under surgical conditions and in nonanesthetized persons, has been reported to increase norepinephrine levels compared with normothermia. 27,28,In vitro  suppression of NKCA by catecholamines has been demonstrated by various groups, 41–43and we recently reported that in vivo β-adrenergic stimulation markedly suppresses NKCA per NK cell and consequently promotes tumor metastasis of the same tumor used in the current study. 29Correspondingly, β-adrenergic antagonists have been reported to block the suppression of both NKCA and tumor resistance caused by various manipulations. 44,45Finally, in the current study the β-adrenergic antagonist, nadolol, almost completely blocked the effects of thiopental anesthesia and hypothermia on the retention of the MADB106 tumor cells. Although several studies in humans have suggested that adrenergic stimulation can enhance NKCA, 34,46there are indications that this increase is attributable to a transient elevation in the number of circulating NK cells, and that the long-term effect of catecholamines on NK activity is suppressive. 33,34 

Suppression of NKCA per NK cell, when induced in vivo , is likely to dissipate with time during preparations for the in vitro  assessment of NKCA. 13When blood is withdrawn, serum is replaced with an artificial medium, and factors such as cytokines and hormones are removed. In vitro  studies using human blood have shown that the suppressive effects of catecholamines dissipate after hormone removal. 43Therefore, a prolonged delay before the assessment of NKCA might conceal the effects of hypothermia. In the current study, only 3 h elapsed between blood withdrawal and assessment of NKCA. The whole-blood assay, or other timesaving procedures, may be considered to obviate these obstacles in future clinical studies.

Although several surgical procedures in humans are characterized by levels of hypothermia similar to or deeper than those reached here, 47most are not. To clarify the clinical applicability of the current findings, a wide range of hypothermic levels and durations should be tested to identify the conditions in which hypothermia begins to suppress NKCA. Simultaneous assessment of neuroendocrine responses could suggest potential mediators. These important elaborations are beyond the scope of the current study and should be addressed in experimental and clinical conditions. An additional unresolved issue in this study is the possible effect of hypothermia on the pharmacokinetics of thiopental metabolism. Because this anesthetic had effects of its own, some of the effects of hypothermia could be attributed to elevated plasma concentrations of thiopental. Nevertheless, we estimate this factor to be of minor significance: Hypothermia prolonged the duration of anesthesia by only 15% but typically had a much greater effect than normothermia on the immune measures, some of which (e.g. , the number of metastases) were not affected at all by normothermia.

Recent studies have suggested the molecular mechanisms that make NK cells important for preventing metastasis. By the time malignant cells evolve to a metastasizing stage, they have usually suppressed the expression of functional major histocompatibility complex class I molecules, making them invulnerable to cytotoxic T cells. The mechanism by which NK cells recognize their targets, however, involves the detection of aberrations in, or the absence of, these molecules. 48This, and the fact that a large proportion of NK cells reside in the blood and the tissues targeted by metastasizing cells (e.g. , the lung and liver), make NK cells most important at this stage. 49,50Apart from lysing single tumor cells dispersed from the primary tumor, NK cells have been shown to infiltrate and diminish established micrometastases. 16 

Our findings suggest that hypothermia may contribute to the clinically well-documented observation that surgery suppresses NKCA. 1–4In the current study in rats, hypothermia and suppression of NKCA were also associated with decreased resistance to metastasis. Relations of this kind would be difficult to assess clinically, because the delay between tumor dissemination and detection of possible related metastases is long and variable. In contrast, suppression of immunity by surgery or hypothermia may lead to an outbreak of latent or opportunistic infections within days, a phenomenon that is indeed well documented clinically. 5,21In sum, it is as yet unclear to what extent the morbidity and immunosuppression observed after surgery is caused by hypothermia. Nevertheless, given the current findings, exposing patients to intraoperative hypothermia should be considered carefully when there is a high risk of viral infection or metastatic spread. Of particular clinical relevance is the removal of metastasizing tumors in procedures that tend to cause hypothermia, such as open abdominal surgeries. Because hypothermia is imperative in several lifesaving procedures, prophylactic measures, such as preoperative stimulation of the immune system or the perioperative use of β-blockers, should be considered.