Hypotension Prediction With HPI Algorithm During Major Gynecologic Oncologic Surgery

Overview

Intraoperative hypotension (IOH) is a rather common event during general surgery, with variable incidence that ranges between 5 and 99% based on the definition used. It is associated to significant complications including acute renal failure, myocardial damage, stroke and overall increased mortality, reason why the prevention and the reduction of hypotensive events through an appropriate proactive approach can potentially improve the patient's outcome. The Hypotension Prediction Index (HPI) is an algorithm derived from the analysis of the arterial waveform and it is expressed as an absolute value from 0 to 100. It has been demonstrated that the HPI is able to predict the occurrence of hypotensive events of patients undergoing major surgery under general anesthesia, providing also a guide for the appropriate treatment based on further calculated secondary hemodynamic variables that estimate patient's preload, cardiac contractility and afterload. Aim of this prospective randomized study is to compare the incidence of IOH during major gynecologic oncologic surgery among two groups of patients receiving standard hemodynamic monitoring versus HPI monitoring. The primary hypothesis is that hemodynamic management HPI-guided reduces the incidence, entity and duration of intraoperative hypotensive events, defined as mean arterial pressure (MAP) lower than 65 mmHg lasting more than one minute.

Intraoperative hypotension (IOH) represents a common event during general anesthesia (GA), with an estimated incidence between 5% and 99%, according to definition adopted \[1\]. Actually, a mean arterial pressure (MAP) below 65 mmHg is considered an appropriate definition of IOH \[2\]. Hypotension mainly occurs during anaesthesia due to three pathophysiological dysregulations: hypovolemia and consecutively decreased cardiac output, myocardial depression and low systemic vascular resistance \[3\]. IOH has been associated with postoperative acute acute kidney injury and myocardial injury; it seems that the cumulative time spent in hypotension increases the risk, and this has implications as even relatively short episodes of hypotension that are treated promptly can over time reach accumulated hypotension time associated with increased injury rates \[4-7\]. Therefore, anaesthetic management that aims to prevent IOH using a "pro-active" treatment protocol might potentially reduce the amount and severity of IOH, perioperative complications and mortality. In order to detect and treat these haemodynamic alterations, advanced haemodynamic monitoring combined with a treatment algorithm can be used \[8\]. The Hypotension Prediction Index (HPI) algorithm was recently established by Edwards Lifesciences (Irvine, USA). Based on the Edward´s monitoring platform (HemoSphere), HPI is a monitoring tool which aims to predict IOH up to 15 min before its onset \[9, 10\]. HPI is a unitless number that ranges from 1 to 100, and as the number increases, the risk of an event occurring in the future increases. The HPI was developed using machine learning methods and is a data-driven model developed from over 200,000 hypotensive patient events and it predicts upcoming hypotensive events based on features of the arterial pressure waveform \[9\]. When HPI rises over 85, the monitor Hemosphere provides a secondary screen showing the following hemodynamic parameters: stroke volume variation (SVV) as indicator of fluid responsiveness - preload, radial dP/dtmax as indicator of cardiac contractility, and dynamic elastance (Eadyn) as dynamic indicator of resistance - afterload. Hemodynamic variables, hemodynamic diagnostic guidance and the definition of a treatment protocol, should allow for determination and treatment of the underlying cause of the impending hypotension. Using an early warning system to predict hypotension does not necessarily lead to reduce hypotension. There are few randomized clinical trials to investigate the effectiveness of the use of HPI in reducing the amount of hypotension as measured by time-weighted average \[11\] during major noncardiac surgery \[3, 12\]. Eligible patients will be allocated in one of the two groups of the study according to a computerized randomization form (ww.randomization.com). Once the patient arrives in the premedication hall of the operating room, a peripheral intravenous access will be placed. Based on the patient's allocation group, the following hemodynamic monitoring will be started: * Group C - Control: through the EV1000 monitor (with Flotrac sensor) * Group HPI - Proactive: through the HemoSphere platform (with Acumen Flotrac sensor) The following timepoints will be recorded: T0: Machine calibration on the bed in neutral position, recordings of the parameters with annotation of the baseline MAP; T1: Infusion of RLS at 3ml/kg/h as in standard clinical practice; T2: cannulation of the radial artery under local anesthesia before induction of general anesthesia, as in normal standard clinical practice; T3: performance of neuraxial anesthesia if needed; T4: standard induction of general anesthesia GA) with propofol, sufentanil and rocuronium bromide; maintenance of GA with sevoflurane and sufentanil; T5: differentiated hemodynamic management based on the patient's allocation group; T6: surgery under continuous monitoring. Group C: Standard Protocol In group C the hemodynamic management will be performed based on cardiac optimization. Fluid boluses will be given with the use of an individualized goal-directed therapy (GDT) protocol aiming to optimize stroke volume index (SVI) \[13\]. Patients will receive 250 ml fluid challenges, within duration of 5 minutes, with a RL solution. Fluid responsiveness will be defined as a SVI increase ≥10%. Maximal stroke volume will be defined as the absence of a sustained rise in SVI of at least 10% sustained for 20 minutes or more in response to a fluid challenge. No more than 500 ml of fluid will be administered for initial determination of the maximal value of SVI before start of the surgical procedure. Once the maximal value of SVI will be determined after induction of anesthesia, SVI must be maintained throughout the intervention period with subsequent boluses of fluids as required. For the treatment of eventual hypotensive episodes, etilephrine or efedrine will be used, depending on the clinical situation. If needed, continuous infusion of noradrenaline and/or other vasopressors or inotropes will be started. All of the hemodynamic parameters detected with the EV1000 monitor and the measures adopted will be recorded. Group HPI: Proactive Protocol In case of HPI\>85, hypotension will be prevented according to the following modalities. The first step will be the optimization of SVV (in case of SVV\>13%) with bolus of 250 ml RLS. In case of dP/dtmax \< 400 (or in decreasing trend) dobutamine infusion 2.5 - 5 mcg/Kg/min will be started. In case of Eadyn \< 0,9 norepinephrine infusion 0,1 mcg/kg/min will be started. All of the data relative to the hemodynamic monitoring and the measures adopted will be recorded and subsequently analyzed to compare the incidence and duration of the hypotensive events in the 2 groups. To calculate the entity of the intraoperative hypotension the following technique will be used: after removal of artifacts, the TWA MAP (time weighted average mean arterial pressure) \<65mmHg will be represented by the area between the threshold of 65 mmHg and the curve of the measured MAP, divided by the total recorded time in minutes \[12\]. Statistical plan The data will be described in their demographic and clinical characteristics through the application of descriptive statistics. The qualitative variables will be described using tables of absolute frequencies and percentages; the continuous quantitative variables will be presented as a median and interquartile range or as a mean and standard deviation when normally distributed; while the non-normal variables will be presented as minimum, maximum and median values. In the two groups of patients, the hypertensive events, defined as a MAP lower than 65mmHg for \> 1 minute (and the severe hypotensive events as having a MAP lower than 60 and 55 mmHg), will be analyzed in terms of frequency and absolute duration and, to estimate their severity, as the ratio between the area under the 65 mmHg threshold and the total length, expressed in minutes, of intraoperative monitoring using the TWA-MAP method. The Chi-squared test will be used to compare the two incidences. The t-student test for independent samples will be used to evaluate statistically significant differences among the patients' demographic and anthropometric characteristics. The Shapiro-Wilk test will be used to assess the normality of data distribution, while the sd test will be used to verify the equality of the variance. The Mann-Whitney test will be used for independent samples to compare the data that are not normally distributed. Regarding the assessment of the secondary outcomes, the investigators will report the relative risks with 95% confidence intervals and p-values based on either the chi square or Fisher's exact test, as more appropriate according to the frequencies expected and detected. A p-value \<0.05 will be considered statistically significant. All of the analysis will be performed through the STATA IC 15.1 (Stata Corp) for Windows, Microsoft Excel, Matlab (The MathWorks Inc., Natick, MA, USA) and the Acumen Analytics software (Edwards Lifesciences, Irvine, CA). Sample Size International literature agrees to consider clinically significant a reduction of 75% of the hypotensive events in terms of entity and duration. According to one of our internal databases (data not yet published), the mean TWA-MAP during gynecological oncology surgery is 0,5 mmHg. Thus, the estimated mean difference between the two groups according to the sample size calculations results to be 0.38 mmHg. Based on the previous clinical trials 11,12 the TWA-MAP standard deviation is estimated to be 0.51 mmHg. Dividing the mean difference by the standard deviation, the resulting dimension has a size effect of 0.74. The investigators calculated that a sample of 60 patients, 30 for each group, has 80% ability of discriminating such effect (t-test with α of 0.05).