Trendelenburg positioning in combination with pneumoperitoneum during robotic-assisted prostatic surgery possibly impairs cerebrovascular autoregulation. If cerebrovascular autoregulation is disturbed, arterial hypertension might induce cerebral hyperaemia and brain oedema, while low arterial blood pressure can induce cerebral ischaemia. The time course of cerebrovascular autoregulation was investigated during use of the Trendelenburg position and a pneumoperitoneum for robotic-assisted prostatic surgery using transcranial Doppler ultrasound. Cerebral blood flow velocity was correlated with arterial blood pressure and the autoregulation index (Mx) was calculated. In 23 male patients, Mx was assessed at baseline, after induction of general anaesthesia, during the Trendelenburg position (40–45°), and after repositioning. During the Trendelenburg position, Mx increased over time, indicating an impairment of cerebrovascular autoregulation. After repositioning, Mx recovered to baseline levels. It can be concluded that with longer durations of Trendelenburg position and pneumoperitoneum, cerebrovascular autoregulation deteriorates, and, therefore, blood pressure management should be adapted to avoid cerebral oedema and the duration of Trendelenburg position should be as short as possible.
Robotic-assisted radical prostatectomy reduces short-term complications and provides better functional results in patients with prostate cancer compared with the conventional technique [1, 2]. For the duration of robotic-assisted prostatectomy, patients have to be positioned in the extreme Trendelenburg position and the peritoneum has to be inflated with carbon dioxide. This increases airway pressure, reduces pulmonary compliance and induces atelectasis [3, 4]. The haemodynamic effects are increases in mean arterial blood pressure and systemic vascular resistance, with controversial effects on cardiac output [3, 5, 6]. After longer surgical procedures in this extreme position, such as robotic-assisted cystectomy with conduit formation, neurological deterioration and brain oedema have been observed . However, the mechanism of this neurological impairment is unclear.
To maintain adequate brain function and neuronal integrity, cerebral blood flow is, under normal physiological circumstances, autoregulated within a wide range of cerebral perfusion pressures. When cerebrovascular autoregulation is disturbed, low cerebral perfusion pressures can lead to global cerebral ischaemia, while high cerebral perfusion pressures can cause cerebral hyperaemia and brain oedema. An increase in mean arterial blood pressure in combination with impaired cerebrovascular autoregulation can lead to neurological deterioration. The hypothesis of the present prospective cohort study was that long-term steep Trendelenburg positioning in combination with pneumoperitoneum impairs cerebrovascular autoregulation. Some of the investigated patients in the present study were also part of a study comparing two different near-infrared spectroscopy monitors for measurement of cerebral oxygenation before, during and after Trendelenburg positioning, and the data presented here are a subset of that trial .
Following Ethical Care Committee approval of the Federal Republic of Rhineland-Palatinate, Germany (approval number: 837.009.10(7018)) and written informed consent from the patients, 23 adult men were included. All patients were recruited and anaesthetised at the Department of Anaesthesiology at the University Medical Centre Mainz, Germany. Patients were included in the cohort study if they were listed for robotic-assisted prostatic surgery. Exclusion criteria were pre-existing neurological disease, acute intracranial lesions (intracranial infection, brain trauma, stroke, intracerebral haemorrhage), cerebrovascular diseases, symptomatic ischaemic heart disease, and if it proved impossible to use transcranial Doppler.
Thirty minutes before induction of anaesthesia, all patients received oral loracepam 1 mg or midazolam 7.5 mg. Electrocardiography, non-invasive arterial blood pressure and peripheral arterial oxygen saturation (SpO2) were monitored. General anaesthesia was induced with sufentanil 0.5 μg.kg−1 followed by propofol 2 mg.kg−1 and then atracurium 0.5 mg.kg−1 for neuromuscular blockade. Anaesthesia was maintained with 0.8–1.5 age-adjusted minimum alveolar anaesthetic concentration of sevoflurane calculated by the monitor software (Primus; Draeger, Luebeck, Germany) and repeated injections of sufentanil and atracurium if necessary. Pressure-controlled mechanical ventilation was performed with constant positive end-expiratory pressure of 6.8 cmH2O and an inspired fractional oxygen concentration of 0.4. Upper airway pressure and respiratory rate were adapted to maintain an end-expiratory carbon dioxide partial pressure (PETCO2) between 4.7 and 6.0 kPa throughout the period of measurement. After induction of anaesthesia, invasive arterial blood pressure measurement was established. The transduer for invasive blood pressure monitoring was fixed at heart level and calibrated to atmospheric pressure at the beginning and after each change in the patient's position. During steady-state anaesthesia, the rate of intravenous crystalloid infusion was kept constant at approximately 400 ml.h−1. Postoperative analgesia was initiated with piritramid 0.1 mg.kg−1 and metamizole 1 g, 20 min before the end of anaesthesia.
The fraction of inspired and expired oxygen, PETCO2, end-tidal sevoflurane concentration, SpO2, invasive arterial blood pressure, body temperature, urine output and neuromuscular blocking using train-of-four were measured throughout the anaesthetic (Primus and Infinity Delta; Draeger). The arterial carbon dioxide partial pressure (PaCO2) was measured at baseline and two to four times during use of the Trendelenburg position.
Measurement of cerebrovascular autoregulation was performed using transcranial Doppler ultrasound. After induction of anaesthesia, the cerebral blood flow velocity (CBFV) in one middle cerebral artery was measured using 2-MHz Doppler probes (Doppler Box®; DWL, Sipplingen, Germany). The probe was positioned over the temporal bone window above the zygomatic arch and fixed with a special frame (DiaMon; DWL). This procedure ensured that the angle and the individual depth of insonation were constant throughout the investigation. As the patients had no known intracranial lesions and no known stenosis of the internal carotid arteries, there was no expected difference between the two hemispheres and the measurement of CBFV was performed on one side only. The index of cerebrovascular autoregulation (Mx) was calculated using the software ICM+ (University Cambridge, UK; see www.neurosurg.cam.ac.uk/icmplus). Spontaneous variations in CBFV and arterial blood pressure were measured, digitised and then processed calculating the time-averaged values using waveform time integration for 6-s intervals. The Mx was calculated as a Pearson's correlation coefficient between 30 samples of mean CBFV and arterial blood pressure . Positive association between variations in blood pressure and CBFV (positive values of Mx) indicates passive dependence of CBFV, and, therefore, impaired cerebrovascular autoregulation. Negative or zero values of Mx imply active cerebrovascular responses to changes in blood pressure, and, therefore, preserved cerebrovascular autoregulation. The accepted cut-off level to discriminate between intact and impaired cerebrovascular autoregulation is 0.3 and values above 0.3 indicate impaired cerebrovascular autoregulation . The index Mx is a continuous variable and statistical analysis was performed with continuous statistical methods. This assessment of cerebrovascular autoregulation was selected because it needs no further manipulation unlike vasopressor administration or thigh cuff deflating, which potentially influence the results or leads to further complications.
After induction of general anaesthesia, invasive blood pressure monitoring, transcranial Doppler ultrasound and cerebrovascular autoregulation measurement were started. The data from commencing transcranial Doppler ultrasound up to positioning in the Trendelenburg position (~10 min) were collected at the beginning of Mx calculation (0 min) and after 3, 5 and 10 min, then averaged and used as the baseline for further calculations. The steep Trendelenburg position with 40–45° head-down tilt and the pneumoperitoneum with an estimated intra-abdominal pressure of 20 cmH2O were then established. Data were collected initially (0 min) and 3, 5 and 10 min after positioning and after this at 15-min intervals during the remaining period of Trendelenburg positioning. After repositioning in the supine position, data were collected at 0, 3, 5 and 10 min and then averaged for statistical analysis.
A power analysis was performed with software R (www.r-project.org) for pairwise comparison of Mx. Using a clinically relevant difference of 0.2, an estimated standard deviation of 0.25, a significance level of 0.05 and a power of 0.9 using a two-sided paired t-test, we required a minimum sample size of 20 patients. The primary endpoint of the study was the time course of Mx with adjustment for potentially influencing factors. Linear mixed models were applied to account for correlation within patients. To compare baseline data with data after repositioning, two-sided paired t-tests were used. A p value < 0.05 is reported as significant. SPSS (version 20; SPSS Inc., Chicago, IL, USA) and Prism (version 6; GraphPad Software Inc, La Jolla, CA, USA) software tools were used for statistical analysis and graph building, respectively.
Between March 2012 and February 2013, 23 of 123 potentially eligible patients undergoing robotic-assisted prostatic surgery were included (21 undergoing radical prostatectomy and two undergoing enucleation of a prostatic adenoma). One hundred patients did not fulfil the inclusion criteria because they had one of more of the exclusion criteria or did not give informed consent. The mean (SD) age was 65 (7) years; 20 patients were of ASA physical status 2, one was ASA 1 and two were ASA 3. The body mass index was 25.5 (4.5) kg.m−2 and the mean duration of surgery was 255 (69) min. Physiological values are shown in Table 1.
Table 1. Changes in measured variables in 23 men undergoing robotic-assisted prostatic surgery in a steep Trendelenburg position. Values are mean (SD).
Baseline (n = 23)
25 min (n = 23)
85 min (n = 23)
145 min (n = 20)
205 min (n = 11)
265 min (n = 3)
After repositioning (n = 23)
Mx, index of cerebrovascular autoregulation; MAP, mean arterial blood pressure; HR, heart rate; PETCO2, end-tidal carbon dioxide pressure; CBFV, cerebral blood flow velocity.
*p = 0.021 vs baseline; †p = 0.001 vs baseline.
After positioning in the steep Trendelenburg position and instituting the pneumoperitoneum, the Mx increased from −0.23 (0.37) to 0.14 (0.39) and then further increased with a slope of 0.0011 (0.00028) per min (p < 0.001) (mixed model without adjustment; Fig. 1). With this ascending slope, the Mx increased after 170 min to more than the cut-off (0.3) indicating impaired cerebrovascular autoregulation. After repositioning supine, the Mx recovered to 0.06 (0.30), which was not different from the baseline value. Adjusting Mx for mean arterial blood pressure, PETCO2 and end-tidal sevoflurane concentration, the Mx increase was still significant with a slope of 0.001 (0.00035) per min (p = 0.024) and, therefore, these potential influencing factors did not explain the increase of Mx during the use of the Trendelenburg position.
The mean arterial blood pressure increased after Trendelenburg positioning from 85 (11) mmHg to 101 (10) mmHg and decreased during the positioning back to baseline levels (Fig. 2). Mean differences between PETCO2 and PaCO2 at baseline were 0.4 (0.4) kPa and during Trendelenburg positioning 0.5 (0.5) kPa. No changes in PETCO2 during the time of positioning were observed. Heart rate, PETCO2, end-tidal sevoflurane concentration and SpO2 did not show clinically relevant fluctuations during the measurement period.
The present observational study shows that Mx increased over time during robotic-assisted prostatic surgery in a steep Trendelenburg position combined with the use of a pneumoperitoneum. These data validate the hypothesis that extreme head-down positioning in combination with a pneumoperitoneum during robotic-assisted radical prostatectomy slowly impairs cerebrovascular autoregulation over a long time period.
Trendelenburg positioning leads to increased hydrostatic pressure with elevated MAP and central venous pressure, with a downward trend in the cerebral perfusion pressure over a period of 165 min . In the present study, the observed deterioration of cerebrovascular autoregulation over time and the impairment of cerebrovascular autoregulation after 170 min (Mx > 0.3) might be the pathophysiological correlate for the deterioration in neuronal function and formation of brain oedema, which have been observed in previous case reports after 8 h of surgery using an extreme Trendelenburg position combined with a pneumoperitoneum . However, whether the evolving cerebral oedema impairs the cerebrovascular autoregulation, or if cerebrovascular autoregulation is first impaired by another mechanism that fosters the development of cerebral oedema due to the high hydrostatic pressure, cannot yet be answered. The late onset of impairment of the cerebrovascular autoregulation might explain why neurological deterioration after robotic-assisted prostatic surgery is commonly not observed, because the duration of surgery is normally shorter. Neurological deteriorations after operations in the Trendelenburg position have, up to now, only been described in case reports when the duration of operation was longer than in this study . Besides these case reports, no study has investigated the effect of an extreme Trendelenburg position in robotic-assisted radical prostatectomy on neuronal outcome. No neurological deteriorations were observed after the termination of anaesthesia in our cohort, but explicit neurological or neuropsychiatric examinations were not performed. Therefore, it will be of further interest if prolonged positioning over a period of longer than 170 min leads to neuropsychiatric deterioration.
The initial increase in MAP after positioning in the Trendelenburg combined with the pneumoperitoneum has already been observed in previous studies, and was caused by an elevation of cardiac output, central venous pressure and systemic vascular resistance [5, 6]. Mean arterial pressure recovers over time to baseline levels. Intact cerebrovascular autoregulation at the beginning of Trendelenburg positioning leads to a constant CBFV, a surrogate for cerebral blood flow, despite the initially increased MAP. If cerebrovascular autoregulation is impaired (after, say, 3 h of Trendelenburg positioning), an increase in MAP may directly threaten the brain via potential complications such as oedema or intracranial bleeding. In the present study, the MAP was maintained in a normal range after 3 h and, therefore, the CBFV remained stable, although cerebrovascular autoregulation was impaired. This may suggest that during prolonged head-down positioning of a patient, MAP should be kept within the normal range.
A limitation of this study was that PETCO2, and not PaCO2, was used to calculate the influence of carbon dioxide on cerebrovascular autoregulation. The observed differences between PETCO2 and the spot tests of PaCO2 during the observation period were minimal. This finding was in accordance with a previous study, which reported no changes in the ratio of PETCO2 to PaCO2 during steep Trendelenburg positioning with carbon dioxide pneumoperitoneum .
Our data have shown that during extreme Trendelenburg positioning combined with pneumoperitoneum, cerebrovascular autoregulation is slightly impaired over time. In these patients, high arterial blood pressure might trigger or aggravate the formation of cerebral oedema. To avoid neurological deterioration in patients placed in an extreme Trendelenburg position for more than 3 h, it may be beneficial to maintain the MAP within the normal range and to minimise the duration of Trendelenburg positioning as much as possible.
Data presented in this manuscript are part of a doctoral thesis presented by Adrian-Hennig Treiber (Medical Student, Department of Anaesthesiology, Medical Centre of the Johannes Gutenberg-University, Mainz, Germany) to the Medical Faculty.
No external funding and no competing interests declared.