| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
|
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Attenuating intraoperative stress is a key factor in improving outcome. We compared neuroendocrine changes and heart rate variability (HRV) during balanced anesthesia (BAL) versus total IV anesthesia (TIVA). Forty-three patients randomly received either BAL (sevoflurane/remifentanil) or TIVA (propofol/remifentanil). Depth of anesthesia was monitored by bispectral index. Stress hormones were measured at 7 time points (P1 = baseline; P2 = tracheal intubation; P3 = skin incision; P4 = maximum operative trauma; P5 = end of surgery; P6 = tracheal extubation; P7 = 15 min after tracheal extubation). HRV was analyzed by power spectrum analysis: very low frequency (VLF), low frequency (LF), high frequency (HF), LF/HF ratio, and total power (TP). LF/HF was higher in TIVA at P6 and TP was higher in TIVA at P37 (P3: 412.6 versus 94.2; P4: 266.7 versus 114.6; P5: 290.3 versus 111.9; P6: 1523.7 versus 658.1; P7: 1225.6 versus 342.6 ms2). BAL showed higher levels of epinephrine (P7: 100.5 versus 54 pg/mL), norepinephrine (P3: 221 versus 119.5; P4: 194 versus 130.5 pg/mL), adrenocorticotropic hormone (P2 10.5 versus 7.7; P5: 5.3 versus 3.6; P6: 10.9 versus 5.3; P7: 20.5 versus 7.1 pg/mL) and cortisol (P7: 6.9 versus 3.9 µg/dL). This indicates a higher sympathetic outflow using BAL versus TIVA during ear-nose-throat surgery.
The stress response to surgery is characterized by increased release of catabolic and immunosuppressive pituitary hormones and activation of the sympathetic nervous system (1). Excessive intraoperative stress may influence outcome, length of hospital stay, and overall costs of hospital care (2,3). Because there is evidence that the choice of anesthetic may influence the intraoperative stress response (4), we compared sevoflurane-based balanced anesthesia (BAL) with total IV anesthesia (TIVA). Although both regimens are commonly used, there are no data comparing the stress response during administration of TIVA versus BAL technique in a consistent surgical population. Measurement of circulating stress hormones requires blood samples and laboratory analyses and is therefore not suitable for bedside monitoring of intraoperative stress. Thus, noninvasive measurement of stress-related physiological changes is highly desirable. Heart rate variability (HRV) may reflect autonomic nervous system (ANS) activity. Different frequency bands of the HRV power spectrum are related to the sympathetic and parasympathetic activity of the ANS (5). Thus, stress-related sympathetic activation may induce a similar change in the HRV pattern observed. Because HRV is influenced both by volatile and IV anesthetics (5), the time course and extent of these changes may differ depending on the anesthetic used (6,7). Thus, we assessed HRV intraoperatively to determine these changes as well as to investigate potential correlations with stress hormones. We hypothesized that 1) catecholamine concentrations differ between groups reflecting a variable degree of hypothalamic and sympathetic activation and 2) that sympathetic activation is reflected by differing HRV patterns, thus allowing for a noninvasive monitoring of the ANS.
After approval by the local ethics committee and written informed consent, 43 patients scheduled for minor elective ear-nose-throat (ENT) surgery were randomly assigned to one of two groups: BAL (anesthesia maintained with sevoflurane) or TIVA (anesthesia maintained with propofol). Exclusion criteria were as follows: age <18 yr or >65 yr, ASA physical status >2, any medication or organ failure known to interact with either stress hormones or HRV. Patients in both groups were premedicated with 10 mg dipotassiumchlorazepate (Tranxilium®), a benzodiazepine drug. Surgery always commenced in the morning to avoid bias caused by the circadian rhythm of circulating stress hormones. Patients in both groups received propofol (2 mg/kg) and remifentanil (0.5 µg · kg1 · min1 within 2 min) before tracheal intubation. After tracheal intubation, patients lungs were ventilated with an oxygen-air mix (Fio2 = 0.3) and ETco2 was stabilized at 3540 mm Hg. Propofol (starting at 6 mg · kg1 · h1 and adjusted in steps of 0.5 mg · kg1 · h1) was used for maintenance of anesthesia in group BAL and sevoflurane (starting with 2 Vol% end-tidal and adjusted in steps of 0.2 Vol% end-tidal) in group TIVA. The dosage of propofol and sevoflurane was adjusted every 3 min to depth of anesthesia assessed by bispectral index (BIS; BIS A-2000 monitor®, Aspect Medical Systems, Leiden, The Netherlands; intraoperatively aimed at 4555). Remifentanil was used for intraoperative analgesia in both groups (starting dosage 0.25 µg · kg1 · min1, adjustment in steps of 0.025 µg · kg1 · min1). The dosage of remifentanil was adjusted every 3 min to maintain mean arterial blood pressure (MAP, ± 20% from baseline) and heart rate (HR, ± 20% from baseline). Body temperature was maintained within normal limits (36°C37°C). Thirty minutes before the anticipated end of surgery all patients received 1 g metamizole for postoperative analgesia. After skin closure, propofol/sevoflurane and remifentanil were stopped. After tracheal extubation, emergence was classified as "good" (calm, cooperative), "medium" (anxious, restless) or "poor" (uncooperative, agitated). At seven event-related time points (P1= baseline before anesthesia; P2 = after tracheal intubation; P3 = after skin incision; P4 = at maximum operative trauma, defined intraoperatively by the surgeon [this time point was used for stress assessment during the surgical procedure, although it is obvious that the time of maximum trauma can be only an estimation]; P5 = end of surgery, i.e., time of skin closure; P6 = after tracheal extubation; P7 = 15 min after tracheal extubation) blood samples were taken for measurement of epinephrine, norepinephrine, adrenocorticotropic hormone (ACTH) and cortisol. Samples were immediately placed into iced water, cool-centrifuged within 15 min, and stored at 25 C° until further analysis. Regarding analysis of stress hormones, a high-performance liquid chromatography technique was used for catecholamines (autosampler: AS 2000®, Merck-Hitachi, Darmstadt, Germany; column: cation-exchange column PCAT Analytical Column®, Bio-Rad Diagnostics, München, Germany; software: HPLC-Manager D-6000 A interface®, Merck, Darmstadt, Germany); detection limit epinephrine: 4 pg (plasma level 10 pg/mL), norepinephrine: 10 pg (plasma level 25 pg/mL); linearity: 102000 pg/mL; normal values for unpremedicated patients: epinephrine 10196 pg/mL, norepinephrine 78521 pg/mL. For detection of ACTH, immunoluminometric assay (LUMItest®; BRAHMS inc., Hennigsdorf, Germany) was used; normal values for measurement in the morning: 1060 pg/mL. Competitive immunoassay (IMMULITE®; Diagnostic Products Corp., Los Angeles, CA) was used for detection of cortisol; normal values for measurement in the morning: 623 µg/dL. HRV was recorded continuously (sampling rate, 1024 Hz; VariaCardio TF4®, Sima Media, Olomouc, Czech Republic) for periods of 256 beat-to-beat intervals and analyzed by power spectrum analysis: very low frequency (VLF: 0.020.04 Hz), low frequency (LF: 0.040.15 Hz), high frequency (HF: 0.150.4 Hz), LF/HF ratio, and total power (TP: 0.020.4 Hz). During HRV assessment, respiratory rate was adjusted to 10 breaths/minute. For correlation analysis with stress hormones, we used the period of 256 beat-to-beat intervals matching the time points P1P7.
For sample size calculation we used data from a previous study by Crozier et al. (4). A sample size of 21 subjects in each group was estimated for 80% power and
Demographics were compared with Students t-test and
A total of 43 subjects (31 male, 12 female) were analyzed. There were no significant changes with regard to sex, age, height, body weight, duration of surgery, dosage of remifentanil, blood loss, and time from skin closure to tracheal extubation (Table 1). Pattern of emergence from anesthesia was similar in both groups (BAL: 18 good, 1 medium, 2 poor; TIVA: 21 good, 1 poor).
MAP showed no significant difference between groups. Patients in the BAL group showed a significantly more rapid HR at P4P7 and lower BIS values at P3 and P4 (Fig. 1). For MAP, HR, and BIS responses to tracheal intubation and skin incision was not significantly different between BAL and TIVA.
In both groups, plasma concentrations of epinephrine tended to decrease after induction of anesthesia (not significant) and did not change at skin incision and maximum trauma. After the end of surgery, plasma concentrations increased significantly (P5 to P6, P < 0.05) in both groups with significantly (P < 0.05) higher levels in BAL at P7 (15 min after tracheal extubation) (Table 2). Plasma concentrations of norepinephrine decreased in both groups after induction of anesthesia (P1 to P2, P < 0.01) but showed a peak (BAL: significant increase P2 to P3, P < 0.01) at skin incision (P3) and maximum trauma (P4) in BAL, which was significantly (P3 P < 0.05; P4 P < 0.01) higher than in TIVA. In TIVA, patients showed no reaction to skin incision and at maximum trauma. Directly after and 15 min after tracheal extubation plasma concentrations of norepinephrine showed a small increase above baseline values (P5 to P6 significant for both groups, P < 0.05) that were not significantly different (Table 2).
ACTH plasma concentrations decreased after induction of anesthesia (P1 to P2 significant in both groups, P < 0.01) and during anesthesia, with a minimum level at the end of operation. Although this pattern was seen in both groups, it was more pronounced in TIVA (significantly lower values at P2 and P5; P < 0.05). After tracheal extubation plasma concentrations in both groups increased again and reached a peak 15 min after extubation. Similar to the intraoperative situation, we observed significantly (P < 0.05) higher values in BAL for both P6 and P7 (Table 2). Cortisol plasma levels decreased after induction and during the operation (P1 to P2, P2 to P3, P3 to P4, P4 to P5 significant in both groups, P < 0.01) in both groups without significant differences. After tracheal extubation, values increased again but did not reach baseline levels. This increase was significant in both groups (P5 to P6, P < 0.01) but was more pronounced in BAL; 15 min after extubation this group showed significantly higher plasma levels of cortisol (P < 0.05) (Table 2). After induction of anesthesia, TP decreased in both groups and increased after tracheal extubation (P1 to P2, P5 to P6, P6 to P7 in both groups, P < 0.01). Patients in the BAL group showed significantly lower TP intraoperatively and up to 15 min after tracheal extubation (P37) (Table 3). Although there was a significant reduction in TP, differences in VLF, LF, and HF did not reach the level of significance (except for significantly lower LF in BAL at P5 and P 7).
HF increased after induction of anesthesia and decreased after tracheal extubation in both groups without any significant differences (Table 3). In the TIVA group, LF/HF ratio did not change significantly throughout the operative procedure but increased above the baseline value after tracheal extubation (P5 to P6 significant in both groups, P < 0.01). In BAL, LF/HF ratio did not change after induction and reached a significantly (P < 0.05) lower level at the end of surgery (P6) compared with TIVA. After extubation, LF/HF ratio increased similarly to TIVA. No significant correlation was observed among VLF, LF, HF and TP and any of the stress hormone plasma levels or with MAP, HR, or BIS. There was a weak correlation between the LF/HF ratio and the plasma level of norepinephrine in both groups at P2 and P6 (P2: r = 0.334, P < 0.05; P6: r = 0.314, P < 0.05).
Choice of anesthetic appears to impact intraoperative stress (4). Previous reports suggest a correlation of intraoperative catecholamine concentration with a negative postoperative outcome, such as postoperative hypertension, vascular graft occlusion, and development of morbid cardiac events (3). We have compared, for the first time, the intraoperative stress response of two of the most commonly used fast-track anesthetic regimens: TIVA with propofol/remifentanil and BAL using sevoflurane/remifentanil. Both anesthesia regimens showed hemodynamic stability within the ranges of our study protocol, but HR was slightly slower with BAL. Previous reports comparing TIVA with BAL are contradictory. Comparing sevoflurane/fentanyl with propofol/fentanyl, Fredman et al. (8) demonstrated a similar MAP but a slower HR in the sevoflurane group. In contrast, Juckenhofel et al. (9) found the HR to be more rapid in a sevoflurane/fentanyl versus a propofol/remifentanil anesthetic regimen, the latter possibly caused by an effect of remifentanil We found no significant correlation between BIS and the release of stress hormones. A previous investigation of Kussman et al. (10) supports our suggestion that BIS is not well correlated with intraoperative stress: they could not find a correlation between BIS and stress hormones in infants during cardiac surgery. BIS is a well-validated measure of anesthetic depth, but it apparently does not reflect the level of intraoperative stress. Although it was part of our study protocol to achieve similar BIS values in both groups, we found significantly lower BIS values at P3 (skin incision: 37 versus 47) and P4 (maximum trauma: 38 versus 45) in the BAL group. We compared the hormonal stress response of two potentially favorable, short-acting anesthetic regimens: sevoflurane/remifentanil and propofol/remifentanil. Although there is a previous investigation comparing the same anesthetic regimens, it was done with a different emphasis. In their study, Eroglu et al. (11) analyzed the plasma concentration of stress hormones 10 minutes before, during, and 10 minutes after a wake-up test during scoliosis surgery and did not find significant differences between the anesthetic regimens. Analyzing the complete intraoperative period at 7 event-related time points, we demonstrated larger plasma concentrations of stress hormones in the BAL regimen. In BAL, norepinephrine plasma concentrations at skin incision and maximum operative trauma almost reached the preoperative baseline, whereas these values remained low in TIVA. In both groups, catecholamine plasma concentrations were still larger than preoperative values 15 minutes after tracheal extubation, with a significantly larger epinephrine plasma concentration in BAL. Our findings of less stress response in the TIVA group are supported by the results of Juckenhofel et al. (9). Comparing the hemodynamic reaction to endotracheal intubation, they found it to be more pronounced in patients with sevoflurane/fentanyl versus propofol/remifentanil. In contrast to the previous authors, Chung et al. (12) described significantly increased responses to tracheal intubation and skin incision in a TIVA (propofol/remifentanil) compared with a BAL group (isoflurane or enflurane/remifentanil)although this seem to contradict our results, the variables of stress were not comparable because plasma concentrations of stress hormones were not measured. Castillo et al. (13) investigated the hormonal stress response in 48 patients with either isoflurane/fentanyl or propofol/fentanyl. They found larger plasma concentrations of cortisol and dopamine in the propofol group. These results, however, might have been affected by the lack of a validated control of anesthetic depth. Considering the differences in study design, it is not surprising that some authors did not find any differences between anesthetic regimens (14). In our study, the moderate concentration of norepinephrine and epinephrine in both groups compared to other investigations (3) suggest that both regimens produced adequate and comparable results despite statistically significant differences. Perioperative measurement of HRV is a relatively new method of assessing the balance of the ANS. It has been shown that the ratio of LF to HF parts of the power spectrum reflects the balance between the sympathetic and the parasympathetic outflow (15). Surgical stress provokes hypothalamic activation of the sympathetic ANS. On the other hand, HRV was shown to reflect sympathetic activation during orthostatic as well as mental stress (16,17). HRV is influenced by anesthesia (5), and anesthesia generally produces a reduction of TP (18). There are few studies with a focus on differences of HRV in either anesthesia with sevoflurane or propofol. Sato et al. (19) described a decrease of LF/HF ratio attributable to a decrease of LF in patients with either sevoflurane or propofol anesthesia. They concluded that choice of the anesthetic did not seem to play an important role for HRV. In contrast, Kanaya et al. (20) demonstrated more distinct changes of HF in patients with propofol versus sevoflurane anesthesia, assuming that sevoflurane has no or little effect on the cardiac parasympathetic tone. Although the time course of TP and LF/HF ratio seems to follow the BIS and the plasma levels of stress hormones, there were only weak correlations between the LF/HF ratio and the level of norepinephrine. This may be explained, at least in part, by the problem of timing. The way of sampling HRV (256 beat-to-beat intervals) and stress hormone plasma concentrations (one time point at the beginning of HRV sampling) might explain the lack of a more distinct correlation. The interrelation of sympathetic activation and HRV warrants further investigations. Some limitations of our study should be noted. As we focused on intraoperative changes of catecholamines and HRV, we cannot comment on the outcome of our patients. However, as our patients reflected a rather young and healthy population, we did not expect any serious adverse events influencing morbidity or length of hospital stay. It may be objected that BIS values during skin incision and maximal surgical trauma were significantly different between groups and that it maybe therefore difficult to differentiate between the effects of a differing depth of anesthesia versus the impact of the anesthetic. Although these differences might be considered as a significant change of experimental conditions at T3 and T4, a systematic bias, as one might expect to find, was not observed: BIS values at T3 and T4 were lower in BAL compared with TIVA, but the plasma levels of catecholamines were higher in BAL at the same time. Given the recommended range reflecting an adequate depth of anesthesia (range, 4555), the differences found are small and may not be clinically relevant. As sample size calculation was based on catecholamine concentration, the lack of differences in other variables may be attributed to a lack of power. Important differences with regard to HRV have been reported in rather small groups of patients, however. We conclude that minor ENT surgical procedures under general anesthesia supplemented with remifentanil cause only a small stress response. The choice of anesthetic modifies this response. Compared with propofol anesthesia, sevoflurane-based anesthesia resulted in higher levels of epinephrine, norepinephrine, ACTH, and cortisol because of a more pronounced reaction to tracheal intubation, skin incision, and extubation. A marked reduction of HRV was observed in both groups. The decrease of TP was more distinct in the BAL group. In both groups, a slight increase of HF and a large decrease of LF indicate a shift towards a more parasympathetically dominated ANS. Both findings emphasize the fact that the choice of the anesthetic has an impact on the individuals response to surgical trauma and encourage discussion of an individually tailored anesthetic.
Presented, in part, at the ASA Annual Meeting, Orlando, Florida, October 1216, 2002. Accepted for publication May 23, 2005.
This article has been cited by other articles:
|
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|