A Novel Parameter Derived From Photoplethysmographic Pulse Wave to Predict Risk of Postreperfusion Syndrome

Postreperfusion syndrome (PRS), observed after reperfusion of the grafted liver, was associated with poor outcome. The end-stage liver disease (ESLD) with autonomic dysfunction in the cardiovascular system has greater risk of developing of PRS, due to the poor ability in sympathetic vasoconstriction. Surgical Stress Index (SSI) is a novel parameter derived from photoplethysmographic pulse wave to assess central sympathetic modulation in awake volunteers. In this study, we determined the relationship between values and the risk of developing of PRS during orthotopic liver transplantation.


Abstract Background
Postreperfusion syndrome (PRS), observed after reperfusion of the grafted liver, was associated with poor outcome. The end-stage liver disease (ESLD) with autonomic dysfunction in the cardiovascular system has greater risk of developing of PRS, due to the poor ability in sympathetic vasoconstriction. Surgical Stress Index (SSI) is a novel parameter derived from photoplethysmographic pulse wave to assess central sympathetic modulation in awake volunteers. In this study, we determined the relationship between SSI values and the risk of developing of PRS during orthotopic liver transplantation.

Methods
We retrospectively studied 163 patients who had undergone OLT, and divided the patients into PRS group and non-PRS group. SSI and related parameters were determined 5min before and after clamping of the inferior vena cava, the occurrence of PRS were recorded during reperfusion.

Results
The clamping of the inferior vena cava modi ed the SSI signi cantly, accompanied with signi cant hemodynamic response. The SSI increased signi cantly after clamping (47.0 (43.0-49.0 ) vs.81.0(69. .0), p<0.001). The SSI increased by 45.3% at 5min after clamping of the inferior vena cava in the PRS group, as opposed to 81.7% in the non-PRS group (P = 0.037). PRS occurred in only 19.4% of patients in whom the SSI increased by more than 50%. Based on a multivariate analysis, percentage of the variation in the SSI was associated with a signi cant increased risk in developing the PRS (OR 2.49, 95% CI 1.15-5.02; P=0.021).
Conclusions SSI can sensitively indicate the central sympathetic modulation function during liver transplantation procedure. SSI might be a sensitive marker of risk of developing PRS.

Background
Severe hemodynamic instability de ned as postreperfusion syndrome (PRS), observed after reperfusion of the grafted liver, is associated with poor outcome [1]. The etiology of this syndrome is not certain with the incidence rate about 8%-30%. In addition to the metabolic acidosis, the release of vasoactive substances by the grafted liver [2], the autonomic dysfunction is responsible for the developing of PRS during reperfusion [3]. The autonomic dysfunction can amplify the haemodynamics instability, characterized as blunt cardiac and pressure compensatory under stressful conditions. Non-invasive tests including lying to standing, valid in diabetics, are introduced to diagnosis the cardiovascular autonomic function before operation [4]. ] However, most of the cardiovascular autonomic dysfunction in liver cirrhosis is undiagnosed before surgery for asymptomatic. It is urgent to nd a sensitive tests of detecting cardiovascular autonomic dysfunction and ability of coping with severe reperfusion related hypotension during liver transplantation procedure [4].
The surgical stress index (SSI), calculated from normalized analysis of the photoplethysmographic waveform and heart rate, has been regarded as an index of assessing surgical stress reactions [5][6][7][8] In addition, the SSI is con rmed to re ect the sympathetic modulation directed to the vessels through gravitational sympathetic stimulation in awake volunteers [9]. Changes in SSI were correlated with the autonomic nervous system modulation in the context of a balanced general anaesthesia undergoing laparoscopic abdominal surgery [10]. To our knowledge, there was no report related the application of SSI monitor during liver transplantation.
Orthotopic liver transplantation (OLT) is known as the key treatment for ESLD. Major haemodynamics changes were frequently observed during clamping and unclamping of the portal vein, hepatic artery, and inferior vena cava (IVC) [11]. IVC clamping caused the decrease of the venous return and subsequent a compensatory barore ex-mediated increase of peripheral vascular resistance and heart rate to keep proper arterial pressure. Thus, SSI might change signi cantly after the IVC clamping, and the variation in SSI can partially re ect the central sympathetic modulation function.
The purpose of this study was to evaluate whether variation in SSI values response to the IVC clamping is a sensitive marker of risk of PRS during liver transplantation procedure.

Methods
This was a retrospective observation study performed at the First A liated Hospital, Zhejiang University School of Medicine with the approval of the hospitals' Institutional Ethics Committee(2017 − 1095) and registered with the University of Zhejiang School of Medicine's Human Research Ethical Committee. This manuscript adheres to the applicable EQUATOR guidelines.

Study population
All patients undergoing OLT with modi ed piggy-back technique for the liver cirrhosis from January 2017 to January 2019 were included. Exclusion criteria were: age < 18 or > 65 years, a history of allergy, coronary artery disease and abdomen surgery, preexisting multiorgan failure, and receiving marginal donor liver.

Anaesthesia and monitoring
In addition to the regular monitors including ECG, pulse oximetry, continuous invasive blood pressure (IBP), and central venous pressure (CVP),BIS (BIS-Vista, Aspect Medical Systems, Newton, MA),the SSI was also monitored. In all patients, the SSI sensor was attached to the index nger of the arm. The plethysmographic waveform was continuously recorded. The SSI values were displayed on the S/5 Advance monitor and they were manually recorded. Pulse contour cardiac output monitoring (PiCCO) was introduced to monitor cardiac output (CO) and guide uid transfusion. Electrolytes and arterial blood gas analysis were checked periodically and corrected throughout the operation.
Anesthesia was induced with 2 mg.kg − 1 propofol, 5 µg.kg − 1 fentanyl and 0.6 mg.kg − 1 rocuronium ,and was maintained with propofol,remifentanyl, cisatracurium. The administration of propofol was guided by BIS (40 to 60).The remifentanyl administration was guided by SSI (20 to 50) during the dissection phase, and neohepatic phase. There was no additional fentanil bolus given except continuous infusion of same rate of remifentanil as that before clamping the IVC during the anhepatic stage. Warm blanket and warm uid or blood were introduced to keep body temperature above 35.5℃.
Fluids (crystalloid, albumin) were preferred during surgery and the transfusion rate was guided by stroke volume variation (SVV). Intraoperative blood and blood products were required according to the measured hemoglobin and coagulation parameters monitored by thromboelastogram(TEG). Hemoglobin value was kept above 7 g.dl − 1 while PaCO 2 was kept at 35-45 mmHg.

Measures to prevent PRS
We used Wisconsin solution as preservative solution in all the grafted livers. The grafted liver was refused before starting the anastomosis of hepatic artery. During anastomosis of the suprahepatic caval vein, the graft was ushed with 500 ml of 5% albumin (0℃) through the portal vein. Appropriate bicarbonate and calcium chloride were administered to prevent hyperkalemia and acidosis.

Record of SSI values and related parameters
SSI and related hemodynamic parameters were determined 5 min before clamping of the IVC(C-5)and 5 min after clamping of the IVC(C + 5).The data included demographic pro les, indications for transplantation, details of comorbidities, cold ischemia times, duration of surgery, monitoring, uid balance, vasopressin use, PRS, urine output, duration of mechanical ventilation, intensive care unit (ICU) stay and hospital stay, any adverse events noted during the hospital stay were also recorded.
PRS was characterized as a drop of MAP more than 30% for over 1 min during the initial stage after reperfusion. 12 The primary outcome was the relationship between the variations in SSI values response to the IVC clamping and the occurrence of PRS. The second outcome was the occurrence of PRS, and the variation in SSI values and related parameters response to the IVC clamping.

Statistical analysis
The variables considered are expressed as percentages when measured qualitatively and as means with the standard deviation or median with IQR when measured quantitatively. Chi-square test was used to compare the categorical variables. The Mann-Whitney U-test was used to analyze the proper continuous variables.
The relevant clinical parameters associated with the occurrence of PRS were chose in the univariate logistic regression analysis. Parameters with P values < 0.1 in the univariate analysis were chose in a multivariate logistic regression analysis to determine independent parameters predicting the PRS occurrence. All analyses were performed using SAS release 6.12 (SAS Institute, Cary, NC).

Results
Patient characteristics 185 patients were reviewed in this retrospective study, but 22 patients were excluded for the history of abdomen surgery and cardiac artery disease(n = 5),receiving marginal donor liver(n = 2), performed with partial clamping of the IVC technique(n = 2),the massive blood loss(> 2000 mL,n = 6),receiving vasopressor administration during the dissection phase(n = 7), data from 163 patients were included. PRS was observed in 48 of the 163 patients (29.4%) in our series (Fig. 1). The patient characteristics and relevant intraoperative data in the PRS and non-PRS groups were shown in Table 1. The median duration of surgery, anhepatic phase, the administered of crystalloids and colloids, and etiology for the liver disease were also shown, there was no difference between the two groups.  SVV and systemic vascular resistance index (SVRI) increased signi cantly from C-5 to C + 5 while the MAP decreased signi cantly (p < 0.05).  Association with variation in the SSI values and the PRS occurrence Table 3.showed that percentage of change of the SSI values from C − 5 to C + 5was greater in the patients with non-PRS than those with PRS(P < 0.0001), as well as the percentage of change of the SVRI values(P < 0.0001). The HR increased signi cantly after the IVC clamping, however, there was no difference in the percent of change of the HR from C − 5 to C + 5 between the two groups. The SSI increased by 45.3% at C + 5 in the PRS group, as opposed to 81.7% in the non-PRS group (P < 0.05). PRS occurred in only 19.4% of patients in whom the SSI increased at C + 5 by more than 50%.  Table 4.showed univariate and multivariate regression analysis of clinical risk parameters associated with the PRS occurrence. The factors including age, MELD scores, heart disease, the requirement of red blood cells, percent of the variation in the SSI and the SVRI were the signi cant determinants of the occurrence of PRS in the univariate logistic regression analysis. Then, multivariate analysis was also introduced to con rm if the variation in the SSI response to the IVC clamping could be an independent risk factor for the PRS occurrence. In addition to the age, MELD scores, the variation in the SSI (OR 2.49, 95% CI 1.147-5.017; P = 0.021) and the variation in the SVRI (OR 2.12,95%CI 1.09-4.95;P = 0.034)resulted as an independent risk factor for PRS (Table 4). The duration of PRS averaged 5.35 ± 7.02 min. The patients who developed PRS were treated effectively. The length of mechanical ventilation was 1.2 days (day 0 (0-2)). One patient in the PRS group died from multiple organ failure on POD 11 .The duration of ICU stay was longer in the patients with PRS than those with no PRS (5.9 (2.4), p = 0.031 vs. 2.5(1.8)), while the duration of hospital stay was no statistical signi cance.

Discussion
The major nd in this study was that the frequency of occurrence of PRS in our series was similar to the previous reports [3,13,14]. There was signi cant difference related to the variation in SSI values response to IVC clamping between the recipients with PRS and without PRS. Multivariate analysis showed that the variation in SSI response to IVC clamping was an independent risk factor for the PRS occurrence. These results suggested that the risk of PRS might be associated with the blunt changes in SSI values response to the IVC clamping.
As we can see from the equation for the SSI value, it is determined by 2 factors, heart beat interval (HBI) and photoplethysmographic amplitude (PPGA). PPGA contributes 66% of SSI values. 9,15 In other words, the SSI re ects sympathetic vasoconstriction re ex and the HR reaction to stimulus including nociception stimuli [10,15]. The increased SSI values along with the increased HR,CVP,SVV and SVRI were observed after the IVC clamping in our series. The increase of SSI values might be ascribed to the signi cantly increased HR and SVR, resulting from the compensatory barore ex for the sudden reduced preload and hypotension. Clamping of the IVC reduced 50% venous return and cardiac output, together with increased systemic vascular resistance and heart rate [11,16,17]. The similar change of the hemodynamics had also been reported in the healthy awake volunteers while changing to head-up tilt position [9,17]. Head-up tilt caused shift of blood toward lower body and reduction of the venous return. These effects induced a compensatory barore ex-mediated increase of heart rate and peripheral vascular resistance aiming to keep arterial pressure near to the baseline. Thus, variation in the SSI response to the IVC clamping seems to indicate the sympathetic out ow directed to peripheral vessels. The variation of the SSI response to the IVC clamping, might also be helpful to assess the intactness of central sympathetic function.
In recipients with liver cirrhosis, autonomic dysfunction is a common nding; usually it is asymptomatic but it may correlate with increased PRS occurrence [4,[18][19][20]. Reperfusion of the grafted liver through the portal vein might result in signi cant cardiovascular collapse, worsening the clinical outcome [12]. To decrease the risk of PRS during liver transplantation, it is advisable to diagnosis the autonomic dysfunction even in the asymptomatic recipients [4]. To date, the autonomic function evaluation of the recipients is another challenging issue because there was no consensus involving the diagnosis of autonomic dysfunction, especially during the procedure [4]. Only a few studies investigated the autonomic dysfunction and its associated with the risk of developing of PRS [3,4,18]. The recipients with autonomic dysfunction were at a great risk of hemodynamic instability during the reperfusion of the grafted liver. In one study, sympathetic withdrawal evidenced as lower LF/HF measured before reperfusion of grafted liver was found signi cantly associated with the risk of developing PRS after reperfusion [18]. The LF/HF of heart rate variability was used as an index of sympathovagal balance, depressed LF/HF was regarded as a relative decline in sympathetic to parasympathetic tone [21]. Another study showed that the variation of the SVRI response to the IVC clamping could be help to assess the intactness of the cardiocirculatory system [3]. Thus, sympathetic dysfunction might be associated with the risk of severe reperfusion-related hypotension. However, it remains not clear whether the PRS may be correlated with the peripheral or central sympathetic dysfunction. This observation indicated that those recipients who had the poor ability of compensatory increase of heart rate and peripheral vascular resistance to keep haemodynamics stability were at greater risk of developing PRS. The SSI values might be a sensitive marker of assess the central sympathetic dysfunction.
Some factors including uid balance, analgesic and hypnotics can affect the SSI values. All the recipients received similar intraoperative management strategies including uid, analgesic and hypnotics transfusion, especially during the dissection stage, thus the effect on the SSI response to clamping of the IVC was similar in all the recipients. The recipients received vasopressive drugs during the dissection stage were also due to the possible effect on the SSI values.
This study was valuable considering the application of a novel parameter during liver transplantation. The response of the SSI value to the IVC clamping might help to detect central sympathetic dysfunction during procedure. The less variation of the SSI showed the blunt barore ex in case the reduction of preload and hyoptension. The less variation of the SSI might help to remind anesthesiologist to pay more attention to the risk of developing PRS after reperfusion. In addition to correct hypocalcemia and acidosis just before reperfusion of graft liver, active pretreatment with vasoconstrictors should be required to prevent the risk of developing severe PRS.
There were some limitations in this retrospective study. First, it was associated with the usual inherent biases, multiple potential confounders, measured and unmeasured. Second, selecting the MELD scores as the only indicator for evaluation the severity of liver disease to predict the risk of developing PRS might be limited, it was reported that jaundice index, coagulation function, renal function was also associated with the risk of developing PRS [13,14,18]. MELD score, however, was a better indicator than individual parameter for evaluating the severity of liver disease, which integrated jaundice index, coagulation function, renal function [27]. Third, we investigated recipients mainly in the setting of LT with the IVC clamping. LT might be performed with partial clamping of the IVC or with VVB in some centers. However, clamping and unclamping of the IVC was associated with more severe hemodynamic instability, especially in the recipients with autonomic dysfunction.

Conclusion
The response of the SSI value to the IVC clamping might help to detect central sympathetic dysfunction. The variation in SSI values response to IVC clamping was associated with the PRS occurrence.