Lower dose of sufentanil was associated with higher plasmatic level of cortisol in patients undergoing elective cardiac surgery at the same depth of anaesthesia: a prospective randomised pilot study. Do we overlook "stress under the surface"?

1. Introduction

Maintaining adequate depth of anaesthesia and analgesia during surgery is one of the most important goals of per-operative anaesthetic care.

While monitoring depth of anaesthesia has become a widely available and accepted technique of per-operative monitoring, objective monitoring of actual level of nociceptive stimulation remains a mostly unanswered challenge [1].

Anaesthetists still dose opioids relying mostly on signs of activation of a sympathetic nervous system such as an increase in heart rate and/or blood pressure.

Cardiac surgery can represent a special challenge for detecting signs of sympathetic stimulation. These signs can be often mimicked by pre-operative medication involving beta-blockers, calcium blockers, per-operative administration of beta-mimetics, vagolytics, vasoconstrictors or vasodilators. Profound changes in hemodynamics during cardiac surgery can also occur: large and rapid shifts of intravascular volume, vasoplegia, low cardiac output, arrhythmias. These situations can also make assessment of nociception based on sympathetic stimulation impossible. Effective management of nociception should be one of the objectives of anaesthetic care [2].

Where techniques of regional anaesthesia cannot be used, per-operative analgesia is provided mainly by administration of opioids.

Opioids are very potent drugs but have a relatively narrow therapeutic window and can have serious side effects. Opioids therefore should be titrated to the actual per-operative patient’s need, e.g., actual intensity of nociceptive stimulation [3, 4].

There were big differences in the dose of opioids used during cardiac surgery historically-different approaches shifted from the concept of “high dose opioid” or “stress free” [5, 6] anaesthesia to the recently described concept of “opioid free” anaesthesia” [7, 8, 9].

Both overdosing and underdosing of opioids can have a specific impact on patients’ safety, morbidity and mortality. When given in an insufficient dose, the nociceptive stimulation is not suppressed effectively and can provoke a stress response with many consequences [10, 11, 12, 13, 14].

Overdosing of opioids can cause prolonged intubation, ventilation and other complications [15, 16, 17, 18].

There are many reasons why per/operative opioid administration should be based on reliable feedback using objective monitoring of nociception in real time. Although there are already several commercially produced devices for the monitoring of nociception, none have been shown to have had a satisfactory level of specificity and sensitivity [19, 20, 21, 22].

Recent meta analyses did not bring an unequivocal answer on how useful existing nociceptive monitors are in per-operative management and dosing of opioids [3, 23, 24].

The vast majority of existing technologies are based on the evaluation of changes in the tone of the vegetative nervous system, which represents a fundamental limitation for reliable use during cardiac anaesthesia.

From all the reasons mentioned above, cardiac surgery has a certain possible potential of for creating conditions under which reaction to nociception can be blurred or overlooked. Subsequent stress reaction can be developing unrecognised as “stress under the surface”.

The main objective of our study was to verify the hypothesis as to whether patients anaesthetised at the same level of unconsciousness would express different signs of stress response caused by nociceptive stimulation when given a different dose of analgetic.

Patients undergoing cardiac surgery would be randomised in two groups each receiving a different dose of opioid sufentanil. At the same time all patients would be anaesthetised by a standardised anaesthetic protocol and maintained in the same level of monitored depth of anaesthesia. All patients would be managed with the same hemodynamic protocol to maintain mean arterial pressure within the same range. We would also record the total dose of vasoconstrictor and vasodilator used to achieve this goal.

Activation of hypothalamic–pituitary–adrenal axis would be assessed by analysis of plasmatic levels of cortisol.

There were several reasons why we focused our observation on the phase prior to connection to the cardiopulmonary bypass. Firstly, we wanted to avoid the diluting effect of primary filling of the extracorporeal circulation circuit on plasmatic concentration of cortisol. Secondly, we wanted to avoid changes in pharmacokinetics and pharmacodynamics of sufentanil during extracorporeal circulation, caused by the dilution effect of priming as well as by temperature fluctuations. Thirdly, we wanted to avoid the interference of observed hemodynamic parameters with possible hemodynamic fluctuations during the cannulation of large vessels and heart. Last but not least, the initial phase of cardiac surgery is suitable for evaluating the response to nociception due to significant surgical stimulation during the sternotomy and the opening of the chest.

2. Materials and methods

Our prospective randomised observational study was performed from July to November 2020 at the Department of Cardiac Surgery, University Hospital Kralovske Vinohrady and 3rd Faculty of Medicine, Charles University, Prague, Czech Republic.

The exclusion criteria for participation in the study were as follows: previous sternotomy or thoracotomy, signs of hemodynamic or other systemic instability, non-elective procedure, sepsis, signs of renal or liver impairment, expected difficult airway, LVEF (left ventricle ejection fraction) below 40%, diabetes mellitus, corticoid treatment, suprarenal gland disease, use of etomidate, hypoalbuminemia below 25 g/L, BMI (body mass index) above 40.

All patients underwent elective cardiac surgery. All patients came from home and were admitted on the day before surgery. All patient had the standard preoperative fluid and food restriction from midnight and were listed for morning procedure. During the investigation period (ending 15 min after the sternotomy) patients received 250 ml of physiological solution intravenously. There was no hemodilution in the period of our investigation.

Patients were randomised into two groups subsequently receiving different doses of sufentanil TCI (target controlled infusion). Group LS (lower sufentanil), n = 9, received TCI sufentanil of 0.25 ng/mL. Group HS (higher sufentanil), n = 10, received TCI sufentanil of 0.75 ng/mL. The pharmacokinetic model of Gepts was used in both groups [25].

All patients were anaesthetised using the same anaesthetic protocol. The depth of anaesthesia was monitored by index qCON (Quantium Consciousness Index) using the Conox monitor (a non-invasive depth of anaesthesia monitor manufactured by Quantum Medical, Barcelona, Spain). All patients were maintained within the same range of depth of anaesthesia (index qCON: 30–50).

Five minutes before induction to general anaesthesia we took first blood samples for analysis of baseline plasmatic cortisol levels.

After preoxygenation, we started a sufentanil TCI (dose according to randomisation) and subsequently performed an induction to general anaesthesia. Our standardised anaesthetic protocol consisted of i.v. administration of midazolam 0.075 mg/kg, propofol 0.75 mg/kg and inhalation of sevoflurane (end-tidal concentration 1–2%). When index qCON dropped below 60, the muscle relaxant rocuronium was administered in a dose of 1.0 mg/kg and endotracheal intubation was performed. General anaesthesia was then maintained with continuous propofol infusion (1 mg/kg/hr) and inhalation of sevoflurane (ET 1–2%) which was dosed according to index qCON. We maintained the depth of anaesthesia in all patients within the range of index qCON: 30–50. Muscle relaxant rocuronium was administered via infusion in a dose of 0.4 mg/kg/hr.

The same protocol of hemodynamic management was used for all patients. We maintained the mean arterial pressure within a range of change of not more than 10% up or down from the pre-induction baseline; in all patients MAP (mean arterial pressure) was maintained within the range 60–80 mmHg. To achieve this goal, boluses of vasopressor phenylephrine or vasodilator glyceryl trinitrate were used if needed, and administrated as a single dose of 50 or 100 mcg each. We recorded the dose of vasoactive medication given to each patient from induction of anaesthesia to the point of blood sampling.

Second blood samples from all patients were obtained 15 minutes after sternotomy. At this time, chest spreader had been applied and surgical manipulation inside the chest had already started. We assumed that this was the most stimulating period of surgery. At this point we also recorded values of systolic blood pressure, pulse rate and depth of anaesthesia index qCON.

3. Results

There were no significant differences found between LS group (patients receiving lower TCI sufentanil) and HS group (patients receiving higher TCI sufentanil) in the basic demographic data: age, weight, BMI (body mass index), LVEF (left ventricle ejection fraction).

These basic data are presented by median (25th–75th percentile): age (years): 72 (64–75) vs. 68 (61–70), p = 0.217, weight (kg): 84 (76–95) vs. 89.5 (85–101), p = 0.251, BMI: 27.6 (25.6–28.7) vs. 27.9 (26.9–33.5), p = 0.720, LVEF (%): 60 (50–60) vs. 63 (55–65), p = 0.113. All patients included in our study had normal levels of plasmatic albumin-presented by median (25th–75th percentile): HS group 41.5 (40.2–45.2) g/L, LS group 45.6 (43.8–46.9) g/L.

There were no significant differences found between the groups in the total dose of vasoconstrictor or vasodilator in the period from induction of general anaesthesia to the point of blood sampling. There were no significant differences in systolic blood pressure, pulse rate or index qCON (depth of anaesthesia) between the groups.

There was no significant difference found in the plasmatic level of cortisol at time before induction to general anaesthesia-data are presented by median (25th–75th percentile): LS group had plasmatic level of cortisol 464 nm/L (372–525), HS group had plasmatic level of cortisol 417 nm/L (298–517), p = 0.624.

There was a significant difference found in the plasmatic level of cortisol at time after sternotomy. Data are presented by median (25th–75th percentile): LS group had a significantly higher plasmatic level of cortisol 700 nm/L (599–765) compared to the level of HS group 328 nm/L (243–386), p = 0.006.

Results are presented in Table 1 and Fig. 1.

Fig. 1.Plasmatic levels of cortisol. LS group, patients receiving lower TCI sufentanil (0.25 ng/mL); HS group, patients receiving higher TCI sufentanil (0.75 ng/mL); nm/L, nanomoles per litre; T1, time of the first blood sample—5 min before induction to general anaesthesia; T2, time of the second blood sample—15 min after sternotomy. p value for T1 is p = 0.624. p value for T2 is p = 0.006.

4. Discussion

The main finding of our study shows that patients who were receiving a lower dose of sufentanil (TCI 0.25 ng/mL) had a higher level of plasmatic cortisol (700 nm/L; 599–765) compared with patients who were receiving a higher dose of sufentanil (TCI 0.75 ng/mL) and whose cortisol level was (328 nm/L; 243–386); figures are presented as median; 25th–75th percentile. The difference found between the groups was statistically significant; p = 0.006. There was no difference between the groups in level of cortisol before induction to general anaesthesia; p = 0.624.

Under physiological conditions the plasmatic cortisol levels follow circadian oscillations. In healthy individuals, normal plasmatic concentrations are the highest in the morning (range 250–850 nmol/L) and the lowest at night or early morning (range 110–390 nmol/L). The cortisol levels increases during the stress response to various stimuli including pain perception [26].

Nociceptive stimulation during surgery provokes stress response and is also associated with an increase in plasmatic cortisol levels. It has been documented that plasmatic cortisol during surgery can range from 270 to 1452 nmol/L and that cortisol levels positively correlate with the extensiveness of surgery-major surgery was associated with higher levels of cortisol than the minor surgery [27, 28].

It has also been suggested that a high plasmatic cortisol level correlated with worse outcomes in patients presenting cardiac failure [29].

The interpretation of our findings is that the higher level of cortisol signified a more potent activation of hypothalamic–pituitary–adrenal axis, and thus that patients with a lower dose of opioid experienced a more intense stress reaction to nociceptive stimulation. Our findings are consistent with the results of other authors [27, 28, 30, 31, 32].

In our study all patents from both groups were anaesthetised under the same protocol and maintained within the same range of depth of anaesthesia measured by monitor Conox. Index qCON was maintained in a range 30–50. There is convincing evidence that index qCON is reliable in the measurement of depth of anaesthesia when compared with other monitors [33, 34, 35, 36].

As one of our secondary findings we can present the fact that there was no significant difference found in index qCON in between the two groups of patients, which is an indication that we succeeded in maintaining the same depth of anaesthesia for patients of the both groups.

At the same time, we found no statistically significant differences between both groups in systolic blood pressure, pulse rate or dose of vasoactive drugs (phenylephrine or glycerol trinitrate).

This secondary finding can be seen as consistent with the specifics of cardiac anaesthesia. Hemodynamic disturbances (change in blood pressure and/or heart rate) are often present during cardiac surgery and can be caused by underlying medication, cardiovascular pathology and impact of the ongoing surgery itself. Cardiac anaesthetists, on the other hand, are skilled and proactive in management of hemodynamic determinants with powerful vasoactive and inotropic drugs when having the objective of maintaining perfusion pressure within the desired range [37, 38, 39].

These specific conditions might facilitate situations where classical signs of sympathetic activation (as in an increase of heart rate and/or blood pressure) are blurred and/or missing.

We acknowledge that size of our group of patients might represent one of the limitations of the study.

5. Conclusions

The main findings of our study proved our hypothesis that patients who were receiving a lower dose of the opioid sufentanil presented higher levels of the stress hormone cortisol. More potent activation of the hypothalamic–pituitary–adrenal axis in this group of patients can be explained by a higher intensity of ongoing nociceptive stimulation.

Our study demonstrates, that the signs of ongoing nociceptive stimulation may not be easily visible and/or easily recognised under certain conditions. Patients may then be exposed to an ongoing “under the surface” stress reaction provoked by nociception even when the same patients are being anaesthetised at the right depth of anaesthesia/unconsciousness and maintained within desired hemodynamic parameters.

Cardiac anaesthesia might represent a special case where there is extensive potential for classical clinical signs of nociceptive stimulation to be blurred and/or overlooked. Our results strongly support the need for developing a method of an objective monitoring of nociception.

Availability of Data and Materials

The data associated with the paper are archived and can be provided by the corresponding author upon reasonable request.

Author contributions

VR—was the author of the hypothesis, design, protocol and the main investigator; TV—was the main collaborator and advisor; VM—was the scientific advisor; MM—provided an independent statistical data analysis; PK—was the consultant. All authors contributed to the study. All authors read and approved the final version of the manuscript.

Ethics approval and consent to participate

The study obtained the approval of the Ethics Committee of the University Hospital Kralovske Vinohrady in Prague and the approval of University Hospital Kralovske Vinohrady and 3rd Faculty of Medicine, Charles University in Prague, ID: EK-VP/041012, ID: EK-VP/041012. The study was performed in line with the principles of the Declaration of Helsinki. Nineteen adult patients were enrolled after signing informed consent for participation in the trial.

Acknowledgment

We thank Julian Fisher, a Lecturer of English Language, University of East Anglia, Norwich, England, for the linguistic revision of the manuscript.

Funding

The study was supported by Charles University Research Program 260530/SVV/2020 and University Research Centres UNCE/MED/002.

Conflict of interest

The authors declare no conflict of interest.