Arterial Pulse Pressure Variation versus Central Venous Pressure as a Predictor for Fluid Responsiveness during Open Major Abdominal Operations

Introduction: Fluid resuscitation is the cornerstone in the management of hemodynamically unstable patients. Dynamic parameters of fluid responsiveness, like pulse pressure variation, have the advantage of being more reliable index for fluid management. Objective: The aim of our study was to compare between arterial pulse pressure variation (PPV) versus central venous pressure (CVP) as a predictor for fluid responsiveness during major open abdominal operations. Patients and Methods: 60 adult patients under general anesthesia with mechanical ventilation underwent open major abdominal surgical procedures were included in our prospective randomized controlled study. Intravenous fluid was infused and monitored by CVP in control group or by PPV in the other group. Hemodynamic variables (heart rate, invasive blood pressure, PPV and CVP) were measured at baseline after anesthesia induction and every 10 min, during first hour of operation, and then every 15 min, till end of surgery. Blood loss and total i.v. fluid & blood transfusion given to patients were recorded and compared between two groups intraoperatively. Results: Patients in the PPV group required more intraoperative fluid and blood transfusion than patients in CVP group to achieve more stable hemodynamic parameters. The fall in blood pressure (>20% of baseline) and increase in heart rate are more common in CVP group (p < 0.05). Conclusion: PPV is a better predictor and a good guide for fluid responsiveness. More stable hemodynamic variables are observed in PPV group.


Introduction
Over the last decades, fluid therapy was considered as the first line of therapy during resuscitation of hemodynamically unstable patients.Although, fluid therapy increases cardiac output (CO) and improves blood pressure, volume overload can lead to pulmonary edema and interstitial edema which will increase morbidity and mortality [1] [2].
It was found that only one half of the critically ill patients are responsive to fluid therapy.Thus, fluid intake can be harmful when hemodynamically unnecessary.Accurate fluid prediction is crucial as it has a great impact on patient outcome.It is important to be able to predict increase in CO before fluid therapy is given.Clinical signs of volume depletion (e.g.heart rate, blood pressure, skin turgor and urine output) are routinely used but of limited sensitivity and specificity [3].
Numerous hemodynamic variables are studied to predict fluid responsiveness.
Static variables (e.g.central venous pressure, pulmonary capillary wedge pressure, inferior vena cava diameter and left ventricle end diastolic volume) depend mainly on the preload estimation.These static variables are not reliable because Starling curves differ between patients.So, there may be a positive response at a given preload (preload dependency) or no response (preload independency) [4] [5] (Figure 1).Michard et al. [12] concluded that PPV more than 12% -13% predicted fluid responsiveness in patients with septic shock or acute respiratory distress syndrome (ARDS) (Figure 2).

Patients and Methods
The study was conducted in the theatre of the general surgery department, at faculty of medicine, A in Shams university during the period from September 2014 to February 2016.
After approval of local ethical committee and taking informed written consent from patients, 60 adult patients, of both sexes, aged between 20 to 60 years, ASA I/II physical status, undergoing open major abdominal surgery under general anesthesia were enrolled in our study.
Exclusion criteria: artery cannulation was done by 20 G catheter.Pressure transducer was leveled at mid-axillary line and zeroed to atmospheric pressure.A series of hemodynamic variables are measured from the indwelling radial artery catheter including heart rate (beats/minute), systolic & diastolic arterial pressure (mmHg), and pulse pressure variation (PPV).Induction of anesthesia was done using propofol (2 mg/kg), fentanyl (2 µg/kg), atracurium (0.5 mg/kg) with endotracheal intubation.Maintenance of anesthesia was done using sevoflurane 2% (1 minimal alveolar concentration) in oxygen/air gas mixture.Fentanyl 1 µg/kg bolus injection was given if required.Mechanical ventilation was maintained in volume control mode with tidal volume 8 ml/kg, respiratory frequency 12 breaths/min.And positive end expiratory pressure (PEEP) 0 cm H 2 O to maintain an ET CO 2 between 35 -40 mmHg.Following induction of anesthesia, atriple-lumen central venous catheter (Certofix® Trio, B. Braun, Melsungen, Germany) was inserted in right internal jugular vein for central venous pressure (CVP) monitoring.
Patients who met the inclusion criteria were randomly allocated into either two groups (30 patients each) by computer generated program: CVP group (Control group) and PPV group.
Lactated ringer and normal saline (crystalloids) were used alternatively to maintain normal CVP range (5 -10 cm H 2 O) and to maintain PPV < 13%.In case of bleeding exceeding allowable blood loss, colloid and blood transfusion were given to maintain hemoglobin level around 10 gm % depending on patients' preoperative hemoglobin.
Systolic blood pressure (SBP), diastolic blood pressure (DBP), heart rate (HR), CVP (normal range 5 -10 cm H 2 O) and PPV (normal range 10% -13%) were measured at baseline after intubation, then every 10 min.during first hour of surgery, then every 15 min.Till end of surgery.The total volume of fluid (crystalloid & colloid) and blood given to the patients during surgery were calculated.

B 3. Sample Size Calculation
Group sample sizes of 28 patients per group achieve 81% power to detect a difference of 500 ml in blood loss between both groups assuming the mean blood loss in the control group is 2 liters with estimated group standard deviations of 650 ml and with a significance level (alpha) of 0.05 using a two-sided two-sample t-test.Thirty patients were included to replace any dropout.
gorical variables were analyzed using the χ 2 test or fisher exact test and are presented as number (%).All P values are two-sided.p < 0.05 is considered statistically significant (Figure 3).

Results
Comparison of demographic data between the two groups showed no significant difference.The two groups were comparable in the general characteristics including age, sex, body weight, ASA, duration and type of surgery (p > 0.05) (Table 1).Figure 6.Diastolic blood pressure (DBP) comparison between two groups.Data are expressed as mean ± SD. p < 0.05 is considered statistically significant.

Discussion
In this study, patients in the PPV group received more intraoperative crystalloid but the difference between the two groups was insignificant.Colloid, blood and Pulse pressure variation measured by IntelliVue MP system possess the ability to predict fluid responsiveness in patients undergoing major operation with mechanical ventilation.This provides a useful guide for fluid management in these patients.Several clinical studies over the past years demonstrated the effectiveness of using PPV monitor for intraoperative and postoperative fluid guidance and how this can reduce the incidence of perioperative complications and hospital stay [14].
IntelliVue MP monitoring system can provide continuous automatic monitoring of PPV through analysis of the arterial pressure waveform without the need for special devices.Other methods for PPV monitoring can be used like PiCCO system and LiDCO system, but these methods require special device [15].Kim et al. investigated the use of PPV in patients undergoing cardiac surgery and found good clinical application [16].
Fischer et al. demonstrated that the accuracy of PPV in predicting fluid requirement might be compromised in open chest operations.This is due to inaccuracy in reflection of phasic changes in preload and stroke volume, decreased cyclic changes in intrathoracic pressure and increased aortic impedance [17].
Despite the strong predictive value of PPV, Cannesson and colleagues [18] demonstrated that PPV may be inaccurate and inconclusive in predicting fluid responsiveness in 25% of patients under general anesthesia.Also, Biais and colleagues [19] used the term "gray zone" while investigating the clinical use of PPV.The practical value of this term allows determination of three zones: A zone with positive response, a zone with negative response and a third zone with uncertainty (i.e.gray zone).This zone occurs in patient receiving small tidal volume or those with low heart rate/respiratory rate ratio.There are many limitations for the use of PPV as a reliable predictor for fluid responsiveness.Tidal volume must be at least 8 ml/kg with controlled ventilation at fixed rate.This is because the low tidal volume ventilation cannot produce significant cyclic change in the intrathoracic pressure to induce preload variations.The cardiac rhythm must be sinus rhythm without any bradycardia or arrhythmia.PPV measurement is also influenced by the presence of spontaneous breathing [22].PEEP must be avoided as it enhances cyclic changes in pleural pressures and hence increase PPV [23].Also, drugs like β-blocker, norepinephrine and vasodilators interfere with PPV accuracy.Norepinephrine decreases DOI: 10.4236/ojanes.2018.82005 PPV while vasodilators increase PPV [24].
Fluid resuscitation guided by CVP may leads to inaccurate volume replacement.CVP guided fluid therapy is effective only when the patients are on the ascending portion of the Frank-Starling curve, but when the left ventricle reach the flat portion of the curve, fluid intake will increase tissue edema and tissue dysoxia [3].
There were some limitations to our study: First, small sample size of the studied groups.Second, we did not measure the cardiac output which is the best method to differentiate between fluid responder from non-responder.Third, we did not continue monitoring the patients during the postoperative period.Colloid, like hydroxyethyl starch, was used during fluid resuscitation although evidence suggesting increased risk of nephropathy and coagulopathy associated with starch products [25].

Conclusion
PPV when combined with CVP can be a good guide to monitor fluid therapy in patients undergoing major abdominal operations.

B References
In contrast to static variables, dynamic variables (e.g.PPV and stroke volume variation, pleth variability index (PVI) and aortic blood flow measured by Doppler) are more reliable and more predictive for fluid responsiveness.These dynamic variables depend on the relation between cardiopulmonary interaction with mechanical ventilation.Mechanical ventilation produces cyclic changes in intrathoracic and trans-pulmonary pressures.These cyclic changes affect the left ventricle preload with resultant cyclic changes in left ventricle stroke volume and systemic arterial pressure in the preload dependent patients only [7] [8].

Figure 3 .
Figure 3. Flowchart showing the number of participants in each stage.

Figure 4 .Figure 5 .
Figure 4. Heart rate comparison between the two groups.Data are expressed as mean ± SD. p < 0.05 is considered statistically significant.

FFP transfusion were significantly
more in the PPV group.Intraoperative blood loss was comparable between the two groups.Monitoring of vital data (i.e.HR, SBP, DBP) were done and were more stable in the PPV group.No patients in either groups required any norepinephrine infusion.Assessment of volume requirement during major operations (e.g.Wipple operation) may be difficult due to use of irrigation fluids, blood loss beneath the drapes and inaccuracy of usual index for cardiac preload (i.e.CVP).So, fluid management should depend on the hemodynamics with volume adjustment guided by PPV.
Gouvea et al. investigated the use of PPV during orthotopic liver transplantation.They found inability of PPV to predict fluid responsiveness during operation of liver transplantation due to decreased systemic vascular resistance (SVR), decreased heart function and operation stimulation [20].A recent study done by Sundaram et al. investigated the use of PPV for intraoperative fluid management in adult patients undergoing craniotomy operation in supine and lateral position.They concluded that PPV monitoring resulted in better postoperative hemodynamic stability, avoided central line associated complications and reduced the additional cost [21].

Table 1 .
Demographic data comparison among the two groups.No difference between the two groups as regards the preoperative Hb.Level and the amount of intraoperative blood loss.The amount of crystalloid given intraoperatively was comparable in both groups (p > 0.05) (Table2).
Values are expressed as mean ± SD or number of patients.p > 0.05 was considered statistically non-significant.

Table 2 .
Intraoperative fluid management in the two groups.
Values are expressed as mean ± SD. *p < 0.05 was considered statistically significant.