Efficacy of Stroke Volume Variation, Cardiac Output and Cardiac Index as Predictors of Fluid Responsiveness using Minimally Invasive Vigileo Device in Intracranial Surgeries.

INTRODUCTION: Functional hemodynamic monitoring using dynamic parameters such as stroke volume variations (SVVs) based on pulse contour analysis is considered more accurate than central venous pressure and mean arterial pressure (MAP) in predicting fluid responsiveness. New device, i.e., Vigileo system, allows automatic and continuous monitoring of cardiac output (CO) based on pulse contour analysis and respiratory stroke volume. AIM: The study aims to test the above hypothesis using graded volume loading step (VLS) to assess the accuracy of SVV as a predictor of fluid responsiveness in patients undergoing intracranial surgery.
MATERIALS AND METHODS: After taking ethical committee approval and informed consent, 60 patients aged between 18 and 55 years belonging to the American Society of Anesthesiologists physical status Class I and II, of either sex, scheduled for brain surgery were included in the study. In this study, 5 min after intubation, with stable hemodynamics, patients received volume loading in successive steps (VLS) of 200 ml of lactated Ringer's solution until the stroke volume increased to <10%. Blood pressure (BP), heart rate (HR), stroke volume (SV), and SVV were measured before and after each VLS. Optimal preload augmentation required by each patient was measured by the number of VLS after which an increase in SV was <10%.
RESULTS: There was a significant decrease in the baseline BP and SV in responsive and nonresponsive groups for the first VLS, but there is no change in HR statistically. There was a significant change in SV after first VLS. Receiver operating characteristic analysis showed a larger area under the curve of 0.758 for SVV compared to other measured variables. The median number of VLS administered were 2 per patient equating to a mean ± SD requirement of 368 ± 176 ml of crystalloid per patient as the optimal preoperative infusion volume.
CONCLUSION: SVV is a better predictor of preload responsiveness measured with third-generation Vigileo device when compared to BP and HR.

INTRODUCTION

Hemodynamic stability and adequate cerebral perfusion are crucial in the management of patients with intracranial pathology.[1] Neurosurgical patients often experience rapid changes in intravascular volume by the administration of volatile anesthetics and potent vasodilators during anesthesia. Intracranial hypertension secondary to cerebral edema is now known to be one of the most common causes of morbidity and mortality in the operating room and postoperative periods.

Static indicators, such as central venous pressure (CVP), pulmonary capillary wedge pressure, and measuring left ventricular end-diastolic area, have been shown to be poor predictors of fluid responsiveness.[234567] The most commonly used device for CO determination is the pulmonary artery catheter with thermodilution capabilities.

Several new techniques have been developed in this regard: aortic transpulmonary thermodilution,[8910] esophageal Doppler, thoracic bioimpedance, partial CO2 rebreathing (NICO), and echocardiography. Nevertheless, these new tools are invasive (aortic transpulmonary thermodilution that requires a specific femoral arterial catheter), operator dependent (esophageal Doppler), poorly accurate (NICO), or require a long training (echocardiography).

Functional hemodynamic monitoring using dynamic parameters such as stroke volume variation (SVV) is considered more accurate than CVP and mean arterial pressure (MAP) in predicting fluid responsiveness. A new device, the Vigileo/FloTrac system (Edwards Lifesciences, Irvine, CA, USA), allows automatic and continuous monitoring of cardiac output (CO) based on pulse contour analysis and respiratory SVV. The main advantage of this new device is that it is minimally invasive and can be used with any arterial catheter. It requires a specific bedside monitor (Vigileo™) connected to a pressure transducer (FloTrac™) attached to the arterial catheter. It does not require any external calibration. The accuracy of this device to assess CO has been tested in numerous settings with various results.[11121314] SVV calculated by this system can predict fluid responsiveness in mechanically ventilated patients with acceptable levels of sensitivity and specificity.[15]

A number of minimally invasive and noninvasive diagnostic tools are currently available that allow clinicians to assess the volume responsiveness using dynamic procedures that challenge the patients’ Frank–Starling curve. These technologies complement one another; each has a useful place in the continuum of the resuscitation process.

Vigileo monitor

The FloTrac sensor and Vigileo™ monitor system is a recently introduced system for monitoring CO continuously. It does not require thermodilution or dye dilution, but rather bases its calculations on arterial waveform characteristics in conjunction with patient demographic data. It is unique among arterial waveform cardiac output systems, in that it does not require calibration with another method. Studies so far done indicate that even though the device is robust, it is accurate over a wide range of cardiac output conditions. This system is based on the analysis of the systemic arterial pressure waveform that does not require pulmonary artery catheterization or calibration with another method.

The possibility of determining the CO using the arterial pulse wave has intrigued both scientists and clinicians for decades.[16171819] Preliminary successes have been achieved using techniques involving determination of the area under the arterial pressure curve, as well as other methods involving analyses of various subtleties of the wave.[2021] The issue has been quantifying the relationship between the amount of blood flow and the pressure wave associated with it. This relationship can vary widely from one individual to another and in a single individual as clinical conditions change. Knowing this relationship for an individual patient and circumstance allows the calculation of a constant (K), which can be used for subsequent CO assessments.

The Vigileo system, using the FloTrac sensor attached to arterial pressure tubing, needs no such calibration and provides continuous CO measurements from the arterial pressure wave [Figure 1]. The system consists of a sensor (FloTrac, Edwards LLC) and a processing/display unit (Vigileo, Edwards LLC). The sensor is a transducer that preprocesses and sends a signal to both a cardiovascular monitor (for real-time waveform display) and the Vigileo monitor.

Figure 1

Vigileo monitor

The Vigileo is relatively small, weighing only 2.1 kg, and can be mounted on an intravenous pole. The processing unit applies a proprietary algorithm to the digitized wave and reports CO, cardiac index (CI; CO divided by body surface area), stroke volume (SV), SV index, and SVV. If a CVP catheter has been placed, its signal can be interfaced with the Vigileo, allowing the calculation or systemic vascular resistance (SVR) and SVR index. When used with a central venous oximetry catheter, the Vigileo also provides continuous central venous oxygen saturation (ScvO2). The rear panel of the Vigileo allows interfacing with CVP and oximetry, external video, and a printer (USB). The basic principle of the system is the linear relation between the pulse pressure and the SV.

The system also reports SVV. This is the change in SV (amount of blood pumped per heart beat) in one respiratory cycle. Patients suffering hypovolemia exhibit an exaggerated SVV. A large SVV (>10%), thus, indicates that the patient is likely to respond favorably to fluid administration.[22]

In conclusion, the Edwards FloTrac/Vigileo system represents a major advance in bedside hemodynamic assessment. It will probably be valuable in the care of a variety of critically ill patients, as well as healthier ones undergoing invasive procedures involving major fluctuations in hemodynamic status.

Aim of the study

We hypothesized that SVV displayed by Vigileo/FloTrac device can effectively predict the variation in SV with fluid load and, in turn, act as a surrogate for volume status of the patients on mechanical ventilation undergoing intracranial surgery.

The study aims,

To test the above hypothesis using graded volume loading step (VLS) to assess the accuracy of SVV as a predictor of fluid responsiveness

To determine the optimal preoperative infusion volume for these patients.

MATERIALS AND METHODS

After obtaining the institutional ethics committee approval and written informed consent, 53 American Society of Anesthesiologists (ASA) Grade I and II patients were included in the study.

All the patients in the study were scheduled for undergoing elective intracranial surgeries, i.e. either tumor resections or aneurysm clipping.

Type of study

This was a cross-sectional study.

Inclusion criteria

Patients belonging to ASA physical Class I and II

Age group 18–75 years

Tumor resection/aneurysm clipping.

Exclusion criteria

Patients with ASA physical status III and IV

Patients with documented coronary or peripheral artery disease

Pulmonary disease

Cardiac arrhythmias

Spontaneous ventilation.

Anesthesia

All patients’ height and weight were noted the previous day of surgery. Patients fasted for 6 h before surgery. Following placement of standard monitors, intravenous and radial arterial (20G) access were established and patients were started on an infusion of lactated Ringer (RL) at a rate of 10 ml/h. All the patients received injection glycopyrrolate 0.05 mg/kg, and anesthesia was induced with 0.04 mg/kg midazolam, 2 μg/kg fentanyl, and 2 mg/kg propofol. Endotracheal intubation was facilitated with 0.1 mg/kg of vecuronium bromide, and mechanical ventilation was set with a tidal volume of 8 ml/kg and a frequency of 12 breaths/min and was initiated to achieve an end-tidal carbon dioxide level of 30–35 mmHg. Sevoflurane 1%–2% in a mixture of oxygen and nitrous oxide was used for anesthetic maintenance.

Hemodynamic monitoring

FloTrac sensor kit was used to connect radial access to the Vigileo device for the invasive blood pressure (BP) monitor. The Vigileo device uses a novel algorithm based on the relationship of arterial pulse pressure with SV. SV = K × pulsatility, where K is a constant quantifying arterial compliance and vascular resistance and is derived from patient characteristics (gender, age, height, and weight) according to the method described by Langewouters et al.[22] and from pressure waveform characteristics (e.g., skewness and kurtosis of individual waves).[232425] This calibration constant was recalibrated every minute in newer versions of the Vigileo device. Pulsatility depended on the standard deviation of arterial pressure wave over a 20-s interval. The algorithm of third-generation Vigileo device allowed patient monitoring through expanded patient algorithm database. This database informed the algorithm to recognize and adjust for hyperdynamic and vasodilated patient conditions. Vigileo version 3.06 used in this study was also adjusted for certain arrhythmias. SVV, CO, and CI were calculated from percentage changes in SV during mechanical ventilation and was continuously displayed by the Vigileo device. Beat-to-beat variation of SV in the preceding 20 s was used to calculate SVV as, SVV% = (SVmax SVmin)/SVmean. Systolic BP (SBP), diastolic BP (DBP), MAP, heart rate (HR), SV, SVV, CO, and CI were measured continuously.

Experiment protocol

Baseline SBP, DBP, MAP, HR, SV, SVV, CO, and CI were documented 5 min after intubation to eliminate the bias of intubation response. Volume load was in steps after anesthesia and before brain surgery. A VLS of 200 ml of LR was performed over 3 min, and hemodynamic variables were recorded 1 min after each VLS. Patients were documented as having a if increase in SV ≥10% or a nonresponsive volume loading steps if increase in SV <10% from the baseline. All the responsive patients received successive VLS till a nonresponsive VLS was observed followed by a confirmatory VLS to calculate the optimum preload value. However, the final confirmatory VLS is not included in the calculation of optimum preload. Placement of cranial pins and surgical incision was deferred until the completion of the experimental protocol.

RESULTS AND ANALYSIS

Statistical analysis was done using MedCalc statistical package version 12.7.5(Acacialaan 22, Belgium).

The demographic characteristics of the 60 patients enrolled in the study are summarized in Table 1; no patient had cardiac rhythm disturbances during the study period. Seven patients who were administered vasopressors or additional fluids other than VLS, due to drop in MAP below 60 mmHg during the study period, were excluded from the study. Of 53 patients included in the study, 6 patients were scheduled for intracerebral aneurysms and 47 were scheduled for cerebral tumor resection. Patients responsive and nonresponsive to first VLS differed significantly in pre-VLS value of SBP (P < 0.05), DBP (P < 0.05), MAP (P < 0.05), SVV (P < 0.00001), CO (P < 0.0002), and CI (P < 0.003), but not the HR. Patients responsive and nonresponsive to first VLS did not differ significantly in pre-VLS value with respect to heart rate. P values for SVV, CO, and CI were statistically highly significant when compared to other variables (unpaired Student's t-test) [Table 2].

Table 1

Demographic data of 53 patients included in the study

Variables	Mean±SD
Age (years)	43.96±13.96
Height (cm)	155.81±6.88
Weight (kg)	59.51±12.53

Data were presented as mean±SD. SD=Standard deviation

Table 2

Hemodynamic variables in responsive and nonresponsive patients before volume loading step

Variables	Responders (a)	Nonresponders (b)	Statistical significance
SBP (mmHg)	110±19	119±15	P<0.05
DBP (mmHg)	68±9	73±11	P<0.05
MAP (mmHg)	84±12	91±12	P<0.05
HR (beats/min)	85±16	82±17	NS
SVV (%)	17±6	10±5	P<0.00001
CO (L/min)	4.3±1	5.7±1	P<0.0002
CI (L/min/M2)	2.8±0.7	3.5±0.9	P<0.003

Data were presented as mean±SD. (a). Responders defined as an increase in stroke volume of ≥10% (b). Nonresponders defined as increase of stroke volume <10% after first volume loading step. SBP=Systolic blood pressure, DBP=Diastolic blood pressure, MAP=Mean arterial pressure, HR=Heart rate, SVV=Stroke volume variations, CO=Cardiac output, CI=Cardiac index, SD=Standard deviation, NS=Not significant

The relationship between changes in SV and baseline hemodynamic variables was assessed using Karl Pearson's correlation. A significant correlation was found between the change in SV after first VLS and the pre-VLS values of SVV, CI, SBP, DBP, and MAP [Table 3].

Table 3

Correlation of hemodynamic variables before volume loading with the change in stroke volume after volume loading

Variable	Pearson’s correlation coefficient	Statistical significance
SBP	−0.3282	P<0.0164
DBP	−0.4	P<0.0049
MAP	−0.3742	P<0.0058
SVV	0.4339	P<0.0012
CO	−0.251	NS
CI	−0.211	P<0.025

A significant positive correlation was found with change in SV after first volume loading step with SVV and a significant negative correlation was found with blood pressures and cardiac index. SBP=Systolic blood pressure, DBP=Diastolic blood pressure, MAP=Mean arterial pressure, HR=Heart rate, SVV=Stroke volume variations, CO=Cardiac output, CI=Cardiac index, SD=Standard deviation, NS=Not significant

To assess the ability of different variables to discriminate between positive and negative response to fluid challenge, receiver operating characteristic (ROC) curves were generated. The area (± standard error) under the ROC curve was 0.582 ± 0.079 (95% CI: 0.439–0.716) for SBP, 0.610 ± 0.0794 (95% CI: 0.466–0.741) for DBP, 0.592 ± 0.0789 (95% CI: 0.449–0.752) for MAP, 0.758 ± 0.0660 (95% CI: 0.621–0.865) for SVV, 0.820 ± 0.0579 (95% CI: 0.691–0.912) for CO, and 0.704 ± 0.0747 (95% CI: 0.563–0.821) for CI. Only area under the curve for SVV, CO, and CI was statistically significant when compared to other variables [Table 4 and Figure 2].

Table 4

Area under receiver operating characteristic curve of hemodynamic variables of patients scheduled for brain surgery before volume loading as predictors of increase in stroke volume by >10% after volume loading

Variable	AUC±SE	Statistical significance
SBP	0.582±0.079	NS
DBP	0.610±0.0794	NS
MAP	0.592±0.0789	NS
SVV	0.758±0.0660	P<0.0001
CO	0.820±0.0579	P<0.0073
CI	0.704±0.0747	P<0.0074

Area under the cure was statistically significant for SVV, CO, and CI with higher statistical significance for SVV compared to CO and CI. Heart rate was not included as it would be a confounding variable. SBP=Systolic blood pressure, DBP=Diastolic blood pressure, MAP=Mean arterial pressure, HR=Heart rate, SVV=Stroke volume variations, CO=Cardiac output, CI=Cardiac index, SE=Standard error, NS=Not significant, ACU=Area under the curve

Figure 2

Receiver operating characteristic curves for CI, CO, SVV, MAP, SBP, and DBP before volume loading as predictors of increase in stroke volume by >10% after volume loading. Area under the curve for SVV was highly statistically significant (P < 0.0001) and statistically significant for CO and CI. CI=Cardiac index, CO=Cardiac output, SVV=Stroke volume variation, MAP=Mean arterial pressure, SBP=Systolic blood pressure, DBP=Diastolic blood pressure

The optimum threshold values by ROC analysis were SVV – 13%, CO – 4.37 L/min, and CI – 3.02 L/min/m2. Thus, if a patient had a SVV value of >13%, CO value <4.37 L/min, or CI value <3.02 L/min/m2, he/she was very likely to be responsive to a subsequent volume load by increasing the SV by >10% with a sensitivity of 58.1% and specificity of 86.4% for SVV, sensitivity of 90.1% and specificity of 71% for CO, and sensitivity of 70% and specificity of 80.6% for CI.

Patients responsive to first VLS received VLS till a nonresponsive VLS was reached to calculate the optimum preload volume. The median number of VLS administered was 2 per patient equating to a mean ± standard deviation requirement of 368 ± 176 ml of crystalloid per patient as the optimal preoperative infusion volume.

DISCUSSION

This study assessed the efficacy of SVV, CO, and CI using FloTrac/Vigileo third-generation device, which applies a minimally invasive technique (intra-arterial) to predict the responsiveness to fluid challenge for preoperative fluid optimization in intracranial surgeries. This study demonstrated that SVV, CO, and, to some extent, CI obtained via a simple arterial line are better indicators as compared to conventional noninvasive methods and can be used as a substitute for the gold standard invasive methods. Till date, there is no confirmed hemodynamic parameter that best predicts the volume status of a patient. In mechanically ventilated patients, volume responsiveness by dynamic tests which rely on heart–lung interactions is superior to static tests.[262728]

The concept of volume load depends on the fact that if SV does not increase with volume load, it does not serve any benefit. On the contrary, it could harm the patient. A volume load increases the SV until the left ventricle reaches the flat part of the Frank–Starling Curve, and once there is maximum overlap between the actin-myosin myofibrils, there is no increase in SV.[2627]

De Waal et al.[29] found that SVV obtained with the FloTracvigileosystem, using flrst-generation system software (version 1.01), which detected SVV every 10 min, failed to predict fluid responsiveness in coronary artery bypass graft patients. The shortcomings suggest that this version may not have been able to assess SVV accurately and further investigations with newer software versions were indicated. Our study used the later third-generation with improvised version 3.06 where calibration constant is recalculated every 1 min.

SVV >12%–13% has been reported to be highly predictive of volume responsiveness with a remarkable consistency. The present study also supports the above statement with a predictive value of 13% and above for SVV and also gives a predictive value of 4.37 L/min and less for CO in optimizing the volume status of patients perioperatively in intracranial surgeries with a minimally invasive technique using third-generation Vigileo device. Unlike the older generation devices, third-generation device does not require the calibration with other means of CO monitoring.

Limitations of the study

Findings were not compared with a standard monitoring device to keep the study minimally invasive

The study was not conducted in patients with large fluid shifts or in high-risk individuals with ≥3 ASA physical status. Studies in such subset of patient would help in confirming hemodynamic predictors to volume response.

CONCLUSION

The present study concludes by giving better predictors, SVV, and CO for fluid responsiveness via a minimally invasive technique in comparison with conventional noninvasive (BP and HR) methods by third-generation Vigileo device in intracranial surgeries. However, more studies in various subsets of patients and also in comparison with other preload indices will be required to confirm their utility.

Financial support and sponsorship

Nil.

Conflicts of interest

There are no conflicts of interest.