|Year : 2019 | Volume
| Issue : 4 | Page : 446-454
A randomized trial comparing three different anesthetic techniques on immune response for patients undergoing abdominal hysterectomy operation
Amira Abdelfattah Shabaan
Department of Anesthesia and Intensive Care, Al-Azhar University
|Date of Submission||10-Jul-2018|
|Date of Acceptance||26-Nov-2018|
|Date of Web Publication||06-Jan-2020|
Amira Abdelfattah Shabaan
Assistant Professor of Anesthesia and Intensive Care, Department of Anesthesia and Intensive Care, Al-Azhar University
Source of Support: None, Conflict of Interest: None
Background In the past decade, the published studies indicated that different anesthetic techniques may have different effects on the immune response of patients undergoing the same type of surgery. The aim of the work was to evaluate the effects of three anesthetic techniques [sevoflurane anesthesia, total intravenous anesthesia (TIVA) with propofol, and epidural anesthesia with bupivacaine] on interleukin-6 (IL-6), tumor necrosis factor-α (TNF-α), and IL-10 in patients undergoing abdominal hysterectomy operation.
Patients and methods A total of 60 female patients with an age range of 39–60 years old and American Society of Anesthesiologists status I and II were scheduled for an elective abdominal hysterectomy. They were allocated randomly into three equal groups (n=20): general anesthesia with sevoflurane (group I), TIVA anesthesia (group II), and epidural anesthesia (group III). Vital signs (heart rate and mean arterial blood pressure) were recorded at the baseline and every 10 min. IL-6, IL-10, and TNF-α were also recorded just before the induction of anesthesia and then at 2 and 24 h after the end of the surgery.
Results Mean arterial pressure and heart rate showed no statistically significant differences between the three groups. The IL-6 was highly significantly increased at 2 and 24 h after surgery in all groups, but the increase was less in the epidural group in comparison with sevoflurane and TIVA groups. The TNF-α was highly significantly increased at 2 h in all groups in comparison with the preoperative value. This increase was more increased in sevoflurane group in comparison with TIVA and epidural groups. IL-10 increased significantly in epidural group 2 and 24 h after surgery in comparison with sevoflurane group and TIVA group. IL-10 was significantly higher in TIVA group than inhalation group 2 and 24 h after surgery.
Conclusion Epidural anesthesia had a better profile in relation to cytokine levels, whereas propofol had little effects on the immune response compared with sevoflurane anesthesia in patients undergoing abdominal hysterectomy operation.
Keywords: anesthetic technique, hysterectomy operation, immune response
|How to cite this article:|
Shabaan AA. A randomized trial comparing three different anesthetic techniques on immune response for patients undergoing abdominal hysterectomy operation. Res Opin Anesth Intensive Care 2019;6:446-54
|How to cite this URL:|
Shabaan AA. A randomized trial comparing three different anesthetic techniques on immune response for patients undergoing abdominal hysterectomy operation. Res Opin Anesth Intensive Care [serial online] 2019 [cited 2020 Apr 4];6:446-54. Available from: http://www.roaic.eg.net/text.asp?2019/6/4/446/275142
| Introduction|| |
Different anesthetic techniques may have different effects on the immune response of patients undergoing the same type of surgery. As both anesthesia and surgery affect the immune system in many ways, a large number of studies have been paying concern to the perioperative immune responses. Moreover, the effect of anesthesia per se on the immune system has taken a lot of attention .
Anesthesia and surgical stress can lead to activation of immunosuppressive mechanisms both during and after surgery. Several in-vitro studies have shown that intravenous anesthesia and inhalational anesthesia affect cells involved in nonspecific and specific immune responses likely by disturbing the balance between proinflammatory and anti-inflammatory cytokines and stimulating leukocyte activity .
Cytokines are low-molecular-weight heterogeneous proteins, which include interleukins (ILs) and interferon. The cytokines released are IL-1, tumor necrosis factor-α (TNF-α), and IL-6. They have a major role in the inflammatory response to, trauma and surgery, inducing local and systemic effects which attempt to limit injury and seem to be directly involved in the symptomatology of sepsis, multiorgan failure, and septic shock .
The anesthesia method for surgery has been a major concern in plenty of cancer research studies. General anesthesia, local anesthesia, and other anesthesia methods had significant effects on postoperative immune cells activity, which may lead to tumor recurrence and metastasis so as to affect prognosis through suppressing immune cell activity .
In recent years, many studies have proven that cytokines and their receptors are suitable parameters for representing and predicting incidence rate and death rate of stress diseases caused by abdominal surgery, with TNF-α, IL-6, and IL-8 as the most crucial cytokines .
The aim of our study is to prove that anesthetic technique plays a key role in modulating the immune response. As an anesthesiologist, it is necessary for us to select appropriate anesthesia methods and drugs to maximize the long-term immune status of the patients during perioperative period.
| Patients and methods|| |
This randomized controlled trial was conducted in Al-Zahra Hospital, after approval of Research Ethics Committee of the Faculty of Medicine for Girls, Al-Azhar University and after consent was taken from the patients before operation. A total of 60 female patients with age range from 39 to 60 years old and having American Society of Anesthesiologists physical status I and II, were scheduled for an elective abdominal hysterectomy. They were allocated randomly into three equal groups according to the type of anesthetic techniques (20 patients each): general anesthesia with sevoflurane (group I), total intravenous anesthesia (TIVA) anesthesia (group II), and epidural anesthesia (group III).
Patients with any type of infection; white blood cells count greater than 10,000/µl; patients on regular antibiotics; steroid or immunosuppressive therapy within the past 6 months; endocrine disorders; chronic inflammatory disease; cardiovascular, respiratory, hepatic, renal, or metabolic diseases; alcohol or drug abuse; and any contraindication to epidural anesthesia were excluded from this study.
All patients were checked 24 h before surgery to fulfill the inclusion criteria and exclude any contraindication to the study through full history regarding any disease or any medication received. Routine investigations were done, for example, complete blood count, fasting blood sugar, kidney functions, liver function tests, coagulation studies, and ECG.
All patients fasted 8–10 h before surgery, in the preoperative holding area, a wide-bore cannula was inserted in a peripheral vein, and the first sample of venous blood (2 ml) was taken to measure IL-6, IL-10, and TNF-α. As soon as the patient arrived in the operating room, the patients were hydrated by infusion of lactated Ringer’s solution. Standard monitors were applied (ECG, noninvasive blood pressure, SpO2, and EtCO2).
The patients were randomly allocated to three equal groups of 20 patients each (N=20) according to the drug administered and the type of anesthetic technique given.
Group I: general anesthesia with sevoflurane (N=20 patients)
After preoxygenation for at least 3 min, anesthesia was induced with fentanyl (1–2 μg/kg, intravenous) and 3–6 mg/kg thiopental sodium. Tracheal intubation was facilitated with rocuronium 0.5 mg/kg, and maintenance with sevoflurane 1.5–2% in oxygen; inspired sevoflurane concentration was adjusted as necessary to maintain pulse and noninvasive blood pressure within 20% of the preinduction values.
Group II: total intravenous anesthesia anesthesia (N=20 patients)
After preoxygenation for at least 3 min, anesthesia was induced with fentanyl (1–2 μg/kg, intravenous) and propofol 1–2 mg/kg until loss of response to oral commands. Tracheal intubation was facilitated with rocuronium 0.5 mg/kg. Maintenance of anesthesia was done with continuous infusion of propofol 4–6 mg/kg/h and fentanyl 0.5 μg/kg/h using syringe pump.
The ventilation in both groups was achieved with a volume-controlled mode with a tidal volume of 4–6 ml/kg; ventilation was adjusted to maintain normocarbia using a closed circle system with a total fresh gas flow rate of 3 l/min.
Just before the end of surgery, the anesthetics were reduced, the lungs were ventilated with 100% O2 at a fresh gas flow rate of 8 l/min, and the residual neuromuscular blockade was reversed by atropine (0.01 mg/kg) and neostigmine (0.05 mg/kg). The duration of surgery was recorded.
Group III: epidural anesthesia (N=20 patients)
Overall, 500 ml of lactated Ringer’s solution was given before epidural anesthesia, in the sitting position under complete aseptic condition using a midline or par median approach; the lumbar epidural space was identified at the L3-L4 or L4-L5 interspace, and the procedure was done using an 18-G needle and with loss of resistance technique. After negative aspiration for cerebrospinal fluid, patients received 12–16 ml of 0.5% bupivacaine. The patients were laid supine until the block had established. In this group, an oxygen-enriched mixture was delivered through a facemask or a nasal cannula during surgery until discharge from the postanesthesia care unit.
The following data were recorded by an anesthetist blind to the study groups:
- Vital signs [heart rate (HR) and mean arterial blood pressure (MAP)] were recorded at the baseline and every 10 min.
- IL-6, IL-10, and TNF-α were also recorded just before the induction of anesthesia and then at 2 and 24 h after the end of the surgery.
- Time of surgery.
Our sampling size was based on referring to previous studies using a two-sided design by adjusting the confidence interval to 95% and the margin of error accepted to 5% and power of the study 80% using Epi info program, version 7.0. The sample size was found to be 60 patients divided into three groups, with 20 patients each .
Data were collected, revised, coded, and entered to the statistical package for the social science (IBM SPSS, IBSS Inc., Chicago, Illinois), version 20, and the following were done: qualitative data were presented as number and percentages, whereas quantitative data were presented as mean, SDs, and ranges. The comparison between two groups with qualitative data was done by using χ2 test. The comparison between two independent groups with quantitative data and parametric distribution was done by using independent t test. The comparison between two paired groups with quantitative data and parametric distribution was done by using a paired t test. The comparison between more than two groups with parametric distribution was done by using one-way analysis of variance.
The confidence interval was set to 95% and the margin of error accepted was set to 5%. So, the P value was considered significant as follows: P value more than 0.05: nonsignificant, P value less than 0.05: significant, and P value less than 0.01: highly significant.
| Results|| |
There was no significant difference between the three groups regarding patient’s demographic data ([Table 1]).
There were no significant differences between all groups through the operative time regarding HR ([Figure 1]) and MAP ([Figure 2]).
|Figure 1 HR (beat/min) changes between the three studied groups. HR, heart rate.|
Click here to view
|Figure 2 MAP changes (mmHg) in the three studied groups. MAP, mean arterial blood pressure.|
Click here to view
The preinduction values of IL-6 showed no statistically significant difference between the three groups (2.88±0.60 pg/ml in sevoflurane group, 3.06±0.98 pg/ml in TIVA group, and 3.37±1.15 pg/ml in the epidural group), with P value more than 0.05.
At 2 and 24 h postoperatively, the IL-6 was highly significantly increased in all groups (P<0.001), but the increase was less in the epidural group in comparison with sevoflurane and TIVA groups and more increased in sevoflurane group in comparison with TIVA group ([Figure 3]).
|Figure 3 IL-6 level (preoperative and postoperative) in the three studied groups. IL-6, interleukin-6.|
Click here to view
Tumor necrosis factor-α
Regarding the preinduction values, TNF-α showed no statistically significant difference between the three groups (19.47±5.40 pg/ml in sevoflurane group, 19.93±7.37 pg/ml in TIVA group, and 21.25±8.75 pg/ml in the epidural group), P value more than 0.05.
At 2 h postoperatively, the TNF-α was highly significantly increased in all groups in comparison with preoperative value (P<0.001), but it was less in epidural group than other groups (58.29±9.43 pg/ml in sevoflurane group, 41.75±7.46 pg/ml in TIVA group, and 30.81±8.38 pg/ml in epidural group).
At 24 h postoperatively, TNF-α returned to preoperative value in all groups (21.75±5.44 pg/ml in sevoflurane group, 22.48±6.96 pg/ml in TIVA group, and 20.55±7.58 pg/ml in the epidural group), with P value more than 0.05 ([Figure 4]).
|Figure 4 TNF-α level (preoperative and postoperative) in the three studied groups. TNF-α, tumor necrosis factor-α.|
Click here to view
The anti-inflammatory cytokine (interleukin-10)
Regarding the preinduction values, IL-10 showed no statistically significant difference between the three groups (1.42±0.48 pg/ml in sevoflurane group, 1.25±0.46 pg/ml in TIVA group, and 1.19±0.35 pg/ml in the epidural group), with P value more than 0.05.
At 2 h postoperatively, IL-10 increased significantly in the epidural group in comparison with TIVA group and increased highly significant in relation to sevoflurane group (6.03±2.81 pg/ml in sevoflurane group, 7.88±2.40 pg/ml in TIVA group, and 10.18±1.96 pg/ml in the epidural group), with P value less than 0.001.
At 24 h postoperatively, IL-10 increased significantly in the epidural group in comparison with TIVA group and increased highly significant in relation to sevoflurane group (11.11±6.17 pg/ml in sevoflurane group, 19.57±10.16 pg/ml in TIVA group, and 45.68±13.81 pg/ml in the epidural group) ([Figure 5]).
|Figure 5 IL-10 level (preoperative and postoperative) in the three studied groups. IL-10, interleukin-10.|
Click here to view
| Discussion|| |
Surgery and anesthesia lead to stress response which causes inflammatory changes and immune response during and after surgery that is responsible for postoperative complications such as infection and hypercatabolism. This stress response is initiated by two mechanisms, the stimulation of the afferent nerve and the release of humeral substances such as cytokines (IL-1 and TNF-α), prostaglandins, and others. The most important is the neural pathway which is responsible for the release of catabolic hormones whereas the humoral response is responsible for the hyperthermic reactions, change in capillary permeability, and coagulation changes .
The systemic inflammatory response likely depends on the invasiveness of the surgical procedure. Although the nature of cytokine response is determined by surgical trauma, the anesthetic technique is capable of modifying its extent . Cytokine mediators, which are released as a result of surgical trauma or anesthesia, have a role in infection and tissue repair. The increase in the proinflammatory cytokines leads to increase in postoperative complications. The increase in the anti-inflammatory cytokines has a protective effect against infection and tissue trauma. So, the balance between the two mediators is important to decrease the postoperative complications and help tissue repair .
The present study was done to evaluate the effects of three anesthetics techniques: TIVA with propofol, general anesthesia with sevoflurane and epidural anesthesia with bupivacaine on proinflammatory and anti-inflammatory cytokines (IL-6, TNF-α, and IL-10) in patients undergoing abdominal hysterectomy operations.
Regarding the hemodynamic changes (HR and MAP), the results of the current study showed no statistically significant differences between the three groups regarding HR and MAP whether in the preoperative, intraoperative, or postoperative values in comparison with the baseline values.
Hadade and colleagues supported the results of the current study by comparing the effect of TIVA with propofol and inhalation anesthesia with sevoflurane on the plasma concentrations of two potent proinflammatory cytokines (IL-32 and TNF-α) in patients scheduled for laparoscopic cholecystectomy. They reported that there were no significant changes between the two groups regarding intraoperative MAP or intraoperative HR .
Moreover, Kahveci and colleagues designed a study to compare the effects of general anesthesia versus epidural anesthesia in 60 patients undergoing major lower extremity surgery on stress hormone response. They reported that there were no statistically significant differences among both groups regarding MAP and HR .
Proinflammatory cytokine IL-6 is a molecule produced by mononuclear phagocytes, endothelial cells, and fibroblasts in response to several stimuli. During the innate response, it stimulates the acute-phase proteins, whereas it promotes the B-lymphocytes growth when specific immunity is initiated; it also plays a role in hematopoiesis.
In the present study, the IL-6 was highly significantly increased at 2 h and 24 h after surgery in all groups, but the increase was less in the epidural group in comparison with sevoflurane and TIVA groups. It was more increased in sevoflurane group in comparison with TIVA group.
Wiryana and colleagues found that patients who received regional anesthesia had a lower level of IL-6 compared with patients who received general anesthesia 24 h after the surgery. They concluded that regional has a better effect of reducing inflammatory response caused by surgery compared with general anesthesia . Moreover, Zhao and Mo  reported that the IL-6 in epidural anesthesia group at 30 min after anesthesia and 2 h after operation was significantly lower than that in general anesthesia group.
Gu and colleagues showed that epidural blockade suppresses the stress-induced increase in IL-6 levels. This observation may be explained by the ability of the sympathetic nerve block induced by epidural anesthesia to reduce the surgical stress response, including reductions in plasma catecholamine and cortisol levels, thus improving the overall immune response and limiting inflammation .
The results obtained by Kvarnström and colleagues were in contrary to the results of the present study. They found that activation of the complement cascade and the release of proinflammatory and anti-inflammatory cytokines in response to major colorectal surgery were similar regardless of whether TIVA or inhalation anesthesia was used .
Proinflammatory cytokine TNF-α is responsible for the initial phase of the acute inflammatory response. It is mainly produced by mononuclear phagocytes after bacterial infection, but it can also be secreted by T-lymphocytes, natural killers cells, and mast cells. In general, TNF-α promotes the recruitment of neutrophils and monocytes thought the expression on the endothelial surface of specific receptors.
In the present study, the TNF-α was highly significantly increased at 2 h in all groups in comparison with preoperative value. This increase was less in epidural group than other groups. It was more increased in sevoflurane group in comparison with TIVA and epidural groups. At 24 h postoperatively, TNF-α returned to preoperative value in all groups.
Our findings are in the line with the study performed by Liu  who compared the influence of propofol and inhalation anesthetics on the levels of serum proinflammatory cytokines in 90 patients with cancer and proved that propofol is better in inhibiting proinflammatory cytokines secretion. Moselli et al.  concluded that the increase in IL-10 at 3 h after incision restrains the initial inflammatory burst and that the following stable high IL-10 levels are able to prevent the systemic inflammation.
Anti-inflammatory cytokine IL-10 is a potent anti-inflammatory cytokine and it is mainly produced by activated macrophages and T-lymphocytes. It has an inhibitory effect on macrophages activation and TNF production. Therefore, it has a pivotal role in regulating the immune response.
In the present study, IL-10 increased significantly in epidural group 2 and 24 h after surgery in comparison with sevoflurane group and TIVA group. IL-10 was significantly higher in TIVA group than inhalation group 2 and 24 h after surgery. These findings suggest that TIVA using propofol suppresses the inflammatory response caused by surgery to a greater extent than a balanced inhalation technique did.
In agreement with the results of the current study, the study done by Ahmed and Hamdi on 54 patients scheduled for lower abdominal surgery under either general with sevoflurane, TIVA, or epidural anesthesia showed that the IL-10 increased significantly in epidural group 2 and 24 h after surgery in comparison with sevoflurane group and in relation to TIVA group, but in contrary to our results, the increase was more in sevoflurane group in comparison with TIVA group .
Ionescu and colleagues also supported the results of our study by comparing the perioperative effects of TIVA and inhalation anesthesia on plasma concentrations of anti-inflammatory ILs and cell adhesion molecules in 88 patients scheduled for laparoscopic cholecystectomies. Their study showed that the largest cytokine responses to surgery were observed for IL-10 with significant peak concentrations 2 h after surgery. IL-10 was significantly greater in patients assigned to TIVA in comparison with those to inhalation anesthesia .
In contrast with our results, Margarit and colleagues compared the effects of TIVA and inhalation anesthesia on plasma concentrations of IL-6 and IL-10 in 60 patients undergoing surgery for colorectal cancer, where plasma cytokine levels were measured before induction of anesthesia, after intubation but before skin incision, and 2 and 24 h after extubation. They found no significant differences between the effects of TIVA and inhalation anesthesia on plasma concentrations of IL-10 after colorectal cancer surgery during the first 24 h postoperatively .
| Conclusion|| |
- Epidural anesthesia for abdominal surgery has a better profile in relation to cytokine levels. Epidural anesthesia can provide complete blockade of all afferent neurogenic stimuli from the surgical field and reduce the stress-related response.
- A balance between proinflammatory and anti-inflammatory processes is essential for the body to be able to react appropriately to injury.
- It is increasingly important to understand the effects of anesthetic drugs and techniques on immune function, not only to prevent harm but also to use them deliberately to protect patients from the risks of surgery.
Financial support and sponsorship
Conflicts of interest
There are no conflicts of interest.
| References|| |
Waly SH, Baqi HA, Nasr YM, Alazizi NM, Atfy M. Immunomodulation: TIVA versus VIMA. Egypt J Anaesthesia 2011; 27:241–247.
Tylman M, Sarbinowski R, Bengtson JP, Kvarnström A, Bengtsson A. Inflammatory response in patients undergoing colorectal cancer surgery: the effect of two different anesthetic techniques. Minerva Anestesiol 2011; 77:275–282.
Abdelhamid S, Talha A, Hamdy S, Abdelhalim AA, Elakany M. Serum IL-6 level in breast cancer surgery: evaluating the addition of hydrocortisone to two anesthetic regimens. Open Anesthesiol J 2015; 9:29–35.
Zhu J, Zhang XR, Yang H. Effects of combined epidural and general anesthesia on intraoperative hemodynamic responses, postoperative cellular immunity, and prognosis in patients with gallbladder cancer: a randomized controlled trial. Medicine (Baltimore) 2017; 96:e6137.
Visconti L, Nelissen K, Deckx L, Van Den Akker M, Adriaensen W, Daniels L, Buntinx F. Prognostic value of circulating cytokines on overall survival and disease-free survival in cancer patients. Biomarkers 2014; 8:297–306.
Inada T, Yamanouchi Y, Jomura S, Sakamato S, Takahashi M, Kambara T, Shingu K. Effect of propofol and isoflurane anaesthesia on the immune response to surgery. Anaesthesia 2004; 59:954–959.
Amin OAI, Salah HE. The effect of general or spinal anesthesia on pro- and anti-inflammatory intracellular cytokines in patients undergoing appendicectomy using the flow cytometric method. Egypt J Anesthesia 2011; 27:121–125.
Schilling T, Kozian A, Senturk M, Huth C, Reinhold A, Hedenstiema G, Hachenberg T. Effects of volatile and intravenous anesthesia on the alveolar and systemic inflammatory response in thoracic surgical patients. Anesthesiology 2011; 115:65–74.
Hildebrand E, Pape HC, Krette KC. The importance of cytokines in the posttraumatic inflammatory reaction. Unfallchirurg 2005; 108:793–803.
Hadade A, Ionescu D, Mocan T, Necula A, Cristea V. Total intravenous versus inhalation anesthesia in patients undergoing laparoscopic cholecystectomies. Effects on two proinflammatory cytokines, serum levels: IL-32 and TNF-alfa. J Crit Care Med 2016; 2:44–50.
Kahveci K, Ornek D, Doger C, Aydın GB, Aksoy M, Emre C et al.
The effect of anesthesia type on stress hormone response: comparison of general versus epidural anesthesia. Niger J Clin Pract 2014; 17:523–527.
Wiryana M, Sinardja K, Budiarta G, Tjokorda GA, Widnyana MG, Prakas D. The differences in the level of interleukin 6 in lower limb orthopedic surgery with general anesthesia technique compared with regional anesthesia combined spinal-epidural block technique. Int J Sci Res 2016; 5:716–721.
Zhao J, Mo H. The impact of different anesthesia methods on stress reaction and immune function of the patients with gastric cancer during the perioperative period. J Med Assoc Thai 2015; 98:568–573.
Gu CY, Zhang J, Qian Y, Tang Q. Effects of epidural anesthesia and postoperative epidural analgesia on immune function in esophageal carcinoma patients undergoing thoracic surgery. Mol Clin Oncol 2015; 3:190–196.
Kvarnström AL, Sarbinowski RT, Bengtson JP, Jacobsson LM, Bengtsson AL. Complement activation and interleukin response in major abdominal surgery. Scand J Immunol 2012; 75:510–516.
Liu TC. Influence of propofol, isoflurane, and enflurane on levels of serum interleukin-8 and interleukin-10 in cancer patients. Asian Pac J Cancer Prev 2014; 15:6703–6707.
Moselli NM, Baricocchi EB, Ribero D, Sottile A, Suita L, Debernardi F. Intraoperative epidural analgesia prevents the early proinflammatory response to surgical trauma. Results from a prospective randomized clinical trial of intraoperative epidural versus general analgesia. Ann Surg Oncol 2011; 18:2722–2731.
Ahmed WG, Hamdi N. Effects of different anesthetic techniques on inflammatory cytokines response and T-lymphocyte subsets in patients undergoing lower abdominal surgery. Ain Shams J Anesthesiol 2012; 5-1:43–56.
Ionescu DC, Margarit SC, Hadad AN, Mocan TN, Miron NA, Sessler DI. Choice of anesthetic technique on plasma concentrations of interleukins and cell adhesion molecules. Perioper Med (Lond) 2013; 2:8.
Margarit SC, Vasian HN, Balla E, Vesa S, Ionescu DC. The influence of total intravenous anaesthesia and isoflurane anaesthesia on plasma interleukin-6 and interleukin-10 concentrations after colorectal surgery for cancer. A randomised controlled trial. Eur J Anaesthesiol 2014; 31:678–684.
[Figure 1], [Figure 2], [Figure 3], [Figure 4], [Figure 5]