Effect of esketamine on inflammatory factors in opioid-free anesthesia based on quadratus lumborum block: A randomized trial

Abstract

Background:

As strong analgesics, opioids provide the analgesic component of general anesthesia, but they have bidirectional effects on the immune system, promoting the production of pro-inflammatory factors. The idea of opioid-free anesthesia is to replace the analgesic effect of opioids in the treatment of acute pain with comparably effective drugs that do not affect the immune system and thereby decrease the production of inflammatory factors. Therefore, this study aims to observe the effect of opioid-free esketamine anesthesia based on quadratus lumborum block on inflammatory factors in patients undergoing lower abdominal or pelvic surgery.

Methods:

A total of 122 patients who underwent lower abdominal or pelvic surgery in our hospital from March 2021 to June 2022 were selected and divided into the esketamine (E) group (n = 62) and control (C) group (n = 60) according to the random number table method. According to the surgical field, the 2 groups underwent unilateral/bilateral quadratus lumborum block in the supine position under ultrasound guidance. In addition, both groups received a target controlled infusion of propofol 3 to 3.5 μg/mL and intravenous rocuronium 0.8 mg/kg. Group E was given opioid-free anesthesia, group C was given opioid-based anesthesia. A 3 to 5 laryngeal mask was inserted according to body weight, and rocuronium 0.5 mg/kg was added intermittently. The levels of interleukin-6 (IL-6), interleukin-8 (IL-8), C-reactive protein (CRP), procalcitonin, tumor necrosis factor-α (TNF-α), numeric rating scales, dosage of propofol, dexmedetomidine and rocuronium, as well as the numeric rating scales score and analgesic complications were monitored in the 2 groups.

Results:

There was no significant differences in general outcomes between the 2 groups (P > .05). The blood pressure in group E was higher than in group C at T1 (P < .05). The levels of IL-6, TNF-α, CRP and IL-8 in group E were significantly lower than in group C at T1, T2, T3, and T4 (P < .05). The levels of IL-6, TNF-α, procalcitonin, CRP and IL-8 in the 2 groups at T1, T2, T3, and T4 were significantly higher than at T0 (P < .05).

Conclusion:

Opioid-free esketamine anesthesia based on quadratus lumborum block achieved perfect postoperative analgesia with little effect on inflammatory factors in patients undergoing lower abdominal or pelvic surgery.

1. Introduction

The activation of the inflammatory response is directly related to the degree of surgical trauma, and the levels of inflammatory mediators are related to clinical outcomes, such as postoperative pain, activity, and length of hospital stay. Opioid-free anesthesia (OFA) has 2 pillars: regional anesthesia and multimodal analgesia (paracetamol, steroidal and nonsteroidal anti-inflammatory drugs, alpha-2 agonists, N-methyl-D-aspartate receptor antagonists, local anesthetics, and drugs such as gabapentin).^[1,2] With the popularization of multimodal analgesia and the continuous development of anesthetic methods and drugs, regimens with low or zero-dose opioids have been applied for general anesthesia in the clinic.^[1,3] Quadratus lumborum block (QLB) is a new method of trunk nerve blockage that has a good analgesic effect, reduces the use of opioids, and is relatively safe for the peritoneal cavity, abdominal organs, and large vessels. The anterior QLB block has a wider range of visceral pain, with human and animal studies showing that it can be a safe and effective alternative to opioids, effectively inhibiting somatic and visceral pain.^[4–6] Esketamine (ESK) is the S-enantiomer of ketamine, which produces sedative and analgesic effects through noncompetitive inhibition of NMDA receptors, and also has a certain affinity for μ receptors.^[7] However, it is not clear what effect ESK has on inflammatory factors under OFA. This trial aimed to examine the effect of ESK on inflammatory factors in patients undergoing lower abdominal or pelvic surgery under OFA.

2. Materials and methods

2.1. Ethics and registration

This research scheme was approved by the Clinical Research Ethics Committee of Hainan Wanning People’s Hospital (SL-2021-002) and was registered in the Medical research registration and filing information system (registration number: No. MR-46-23-010135). Before randomization, all patients signed an informed consent form, and were free to choose whether to continue the study at any time.

The inclusion criteria were as follows: Age of 18 to 65 years old; American Society of Anesthesiologists I to III, no infection at the puncture site; Normal liver and kidney function and planned hospital stay ≥ 48 hours after surgery; undergoing open or laparoscopic lower abdominal or pelvic surgery; and no history of allergy to drugs used in this study.

The exclusion criteria were as follows: Refusal to participate in the study, reluctance to use patient-controlled intravenous analgesia (PCIA); Severe circulatory insufficiency, arrhythmias (especially bradyarrhythmia), hypovolemia, shock, coronary instability, autonomic neuropathy with orthostatic hypotension;^[8] Mental illness; Failure of QLB (puncture without block plane); Patients with serious risk of hypertension or intracranial pressure; Poorly controlled or untreated hypertension > 180/100 mm Hg; Untreated or under-treated hyperthyroidism; Patients requiring uterine muscle relaxation for surgery.

2.2. Randomization, blinding, and data collection

This was a prospective study, and the sample size was calculated using the following formula: $\text{n}=\frac{{(Z_{\alpha }+Z_{\beta })}^2(1+1/k)p(1-p)}{{(P_e-P_c)}^2},p=\frac{p_e+p_c}{1+k}$

where α = 0.05, β = 0.20, k is the ratio of the control group to the experimental group (k = 1 for this experiment), p_e = 0.87 is the measured probability value of successful block in the preexperiment, p_c = 1.00 is the measured probability value of the control group. The values of Z_α and Z_β can be found in the Z-score table, and n = 60 was obtained. The sample size was appropriately increased by 10% to 20% to account for factors such as exclusion criteria and loss to follow-up. A total of 140 patients who underwent lower abdominal or pelvic surgery in our hospital from March 2021 to July 2022 were selected. All patients were divided into 2 groups by the same anesthesia nurse according to the random number table method. The remainder was obtained by dividing the random number in the random number table by the number of groups. The aliquot was the control (C) group, and the experimental (E) group was the remainder (Fig. 1).

Figure 1.

Consort e-flowchart.

Diluted colorless esketamine (50 mg) 10 mL or sufentanil (50 µg) 10 mL (both drugs were injected at 0.2 mL/kg during induction of anesthesia, total ≤ 10 mL) and group labels were delivered to the anesthesiologist on duty in sealed and light-tight envelopes labeled 1 and 2, respectively. The anesthesia nurse who performed the grouping prepared a milky white analgesic pump containing flurbiprofen (100 mL) according to the grouping and handed it to the anesthesiologist on duty. The envelope 2 was opened by the anesthesiologist on duty for follow-up registration at the end of the surgery. Among them, 15 cases were excluded according to the exclusion criteria, 3 cases were lost to follow-up, and 122 cases were finally included in the experimental (E) group (62 cases) and the control (C) group (60 cases).

2.3. Anesthesia protocol

All subjects underwent routine ECG, BP, pulse oxygen saturation and end-tidal carbon dioxide monitoring under general anesthesia. The entropy index was monitored using a Datex-Ohmeda monitor. The surgical plethysmographic index (SPI) was monitored using a GE Healthcare monitor (Helsinki, Finland). The entropy index is used to evaluate the depth of anesthesia during surgery, and SPI is used to evaluate the “nociceptive” response of the central nervous system to pain.^[9]

Tropisetron 5 mg was given intravenously, penehyclidine hydrochloride 0.01 mg/kg was given intravenously, and dexmedetomidine (Dex) 0.8μg/kg/10 minutes was given by target controlled infusion. After sedation, ultrasound-guided single/bilateral QLB was performed through the anterior approach under local anesthesia according to the surgical field in the supine position using the “eye sign” (Fig. 2A) and “baby sign” (Fig. 2B). After the target was determined, 0.20% ropivacaine + 0.25% lidocaine, 20 mL/ side was injected. After 15 minutes, the block plane was evaluated by acupuncture.^[10] Patients without a blocked plane were excluded from this experiment.

Figure 2.

(A) “Eye sign”, (B) “baby sign.” N: needle indication;“Eye sign”: SPI-eyebrow; QL - eyeball; The three layers of abdominal wall muscles (transversus abdominis, internal oblique, external oblique) - crow’s feet”; Baby sign”: QL - infant head; PM - infant body; TP and VB -pillow and cradle. SF = subcutaneous fat, LD = latissimus dorsi, ES = vertical ridge, SPI = serratus posteriori, EO = external oblique muscle, IO = internal oblique muscle, TA = transversus abdominis, QL = quadratus lumborum, PM = psoas major, TP = transverse process, VB = vertebral body, AC = abdominal cavity.

Both groups received target controlled infusion propofol 3 to 3.5 μg/mL and rocuronium 0.8 mg/kg, and group E received ESK 1 mg/Kg (manufacturer: Jiangsu Hengrui Pharmaceuticals Co., Ltd.; Approval number: H20193336, specification: 2 mL: 50 mg), total dose ≤ 50 mg for induction; Group C received sufentanil 1 μg/Kg (manufacturer: Yichang Humanwell Pharmaceutical Co., Ltd.; Approval number: H42022076; Specification: 1 mL: 50 μg) for induction with a total dose of ≤ 50μg, and a laryngeal mask 3 to 5 was inserted according to body weight. During the operation, QLB was used to maintain analgesia, and 0.5 mg/kg rocuronium was added intermittently according to the needs of surgery.

A low tidal volume lung protection strategy was used using the following parameters: tidal volume 5 to 8 mL/kg, positive end-expiratory pressure 5 to 6 cm H₂O, respiratory rate 12 to 15 bpm, end-tidal carbon dioxide 35 to 45 mm Hg, maintenance entropy index 40 to 65, SPI 30 to 50. Ephedrine and atropine were used for symptomatic treatment when the blood pressure was low and the heart rate was slow. The levels of interleukin-6 (IL-6), interleukin-8 (IL-8), C-reactive protein (CRP), procalcitonin (PCT), and tumor necrosis factor-α (TNF-α), the dosage of Dex, propofol, and rocuronium, as well as the numeric rating scales (NRS) and analgesic complications were monitored in both groups. At T0, T1, T2, T3, T4, 3 to 5 mL venous blood was collected from each group to detect IL-6, IL-8, CRP, PCT, and TNF-α.

Propofol was stopped at the skin suture, and PCIA was performed after intravenous injection of flurbiprofen 50mg. Group E: ESK 0.015 mg/(kg·hour)^[11] (total dose ≤ 50 mg) + flurbiprofen 200 mg + tropisetron 5 mg + 0.9% NS to 100 mL; Group C: sufentanil 0.040μg/(kg·hour) (total dose ≤ 100 ug) + flurbiprofen 200 mg + tropisetron 5 mg + 0.9% NS to 100 mL, 2 mL/hours, bolus 0.5 mL/15 minutes.

2.4. Observational indicators

The primary outcomes were the levels of inflammatory factors before anesthesia (T0), 0.5 hour after operation (T1), 12 hours after operation (T2), 24 hours after operation (T3), and 48 hours after operation (T4). The levels of IL-6 (Beijing Hotgen Biotech Co., Ltd.), IL-8 (Shandong Zhonghong Detection Biotechnology Co., Ltd.) and PCT (Kit: Shenzhen Mindray Bio-Medical Electronics Co., Ltd.) were measured using the chemiluminescence method. CRP (Shenzhen Mindray Bio-Medical Electronics Co., Ltd.) was measured by immunoturbidimetry. TNF-α (Wuhan Saipei Biotechnology Co., LTD.) was measured by ELISA.

Secondary outcomes included the NRS^[12] (0 to 10 points, 0 points: no pain, 10 points: unbearable severe pain) score, and postoperative complications observed 1, 12, 24, and 48 hours after the operation. The dosages of propofol, rocuronium and Dex were recorded in both groups.

2.5. Statistical analysis

SPSS 25.0 statistical software (IBM Corp., USA) was used for data analysis, and the measured data were expressed as means ± standard deviations ($\overline{x}\pm s$). Student t test was used to asses the significance of differences in continuous variables between the 2 groups, repeated analysis of variance was used for repeated measurement data, and statistical data were expressed as frequencies and compared between groups using Pearson’s χ2 test. Differences with P values < .05 were considered statistically significant.

3. Results

3.1. Comparison of general data between the 2 groups

There were no significant differences in gender, age, BMI, surgical method, department, and preoperative complications between the 2 groups (P > .05, Table 1).

Table 1.

Comparison of general data between the two groups ($\overline{x}\pm s$).

Project/group	Gender (male/female)	Age	BMI (kg/m·m)	Department (cases) gynecology general surgery urology			Type of operation (Laparoscopic/open)	Preoperative complications (case) Hypertension diabetes cardiac disease other
E group (n = 62)	30/32	51.91 ± 16.54	23.44 ± 3.67	16	20	26	40/22	10	6	5	16
C group (n = 60)	36/24	55.26 ± 13.15	22.88 ± 3.78	15	17	28	41/19	4	5	5	11
t/χ2	1.66	−1.24	0.83	0.01	0.22	0.28	0.20	2.69	0.07	0.00	0.99
P value	.20	.22	.41	.92	.64	.60	.66	.10	.80	.96	.32

3.2. Comparison of the dosages of propofol, rocuronium and dexmedetomidine, block point, and block level between the 2 groups

There was no significant difference in the dosages of propofol, rocuronium, dexmedetomidine, block point, or block plane between the 2 groups (P > .05, Table 2).

Table 2.

Comparison of propofol, rocuronium and Dex dosage, block point and block level between the two groups ($\overline{x}\pm s$).

Project/Group	Propofol (mg)	Rocuronium (mg)	DEX (ug)	Point of puncture	Sensory block level
				Unilateral/bilateral	T6/T7	L1/L2
E group (n = 62)	787.34 ± 472.75	73.44 ± 33.57	137.42 ± 18.22	19/43	13/79	48/14
C group (n = 60)	675.00 ± 377.44	67.42 ± 26.47	132.55 ± 11.03	15/45	15/45	43/17
t/χ2	1.59	−1.10	1.78	0.48	0.28	0.53
P value	.12	−1.10	.08	.49	.60	.47

Dex = dexmedetomidine.

3.3. Comparison of NRS scores and postoperative complications between the 2 groups

There was no significant difference in the preoperative NRS scores between the 2 groups (P > .05). The NRS scores of group E at 1 hour (P = .00), 12 hours (P = .03), 24 hours (P = .00) and 48 hours (P = .03) after the operation were lower than those of group C, and the difference was statistically significant. The NRS scores of the 2 groups at 12 hours, 24 hours and 48 hours after the operation were lower than those at 1 hour after the operation, and the difference was statistically significant (All P = .00). There were no significant differences in postoperative infections, respiratory depression, nausea, vomiting, or dizziness between the 2 groups (P > .05). Pruritus (P = .04) and hypotension (P = .02) were significantly lower in group E than in group C (Table 3).

Table 3.

Comparison of NRS scores and postoperative complications between the two groups ($\overline{x}\pm s$, cases).

Project/group	Before surgery	1 h after surgery	12 h after surgery	24 h after surgery	48 h after surgery	Postoperative sinfection	Nausea	Vomiting	Pruritus	Dizziness	Respiratory depression	Hypotension
E group (n = 62)	2.40 ± 0.83	3.23 ± 0.97	2.85 ± 0.94	2.48 ± 0.75	2.02 ± 0.85	1	8	3	0	9	0	0
C group (n = 60)	2.20 ± 0.94	3.97 ± 1.01	3.22 ± 0.85	3.27 ± 1.22	2.35 ± 0.86	1	5	2	4	7	0	5
t/χ2	1.26	−4.11	−2.24	−4.30	−2.15	0.48	0.67	0.18	4.27	0.22	/	5.39
P value	.21	.00	.03	.00	.03	.49	.41	.68	.04	.64	/	.02

NRS = numeric rating scales.

3.4. Comparison of IL-6, CRP, PCT, and TNF-α levels between the 2 groups

There were no significant differences in the levels of IL-6, IL-8, CRP, PCT, or TNF-α between the 2 groups at T0 (P > .05). However, the levels of IL-6, TNF-α, and CRP in group E were significantly lower than those in group C at T1, T2, T3 and T4 (All P = .00). The levels of IL-8 in group E were significantly lower than those in group C at T1 (P = .02), T2 (P = .00), T3 (P = .01), and T4 (P = .00). There was no significant difference in PCT levels between the 2 groups at T1, T2, T3, and T4 (P > .05). The levels of IL-6, TNF-α, CRP, and IL-8 in the 2 groups at T1, T2, T3, and T4 were significantly higher than at T0 (All P = .00).The levels of PCT in the 2 groups at T1(P = .00), T2(P = .00), T3(P = .00), and T4(P = .02) were significantly higher than at T0.(Table 4).

Table 4.

Comparison of IL-6, CRP, PCT, INF-α, and IL-8 levels between the two groups ($\overline{x}\pm s$).

Project/group	time	E group (n = 62)	C group (n = 60)	t	P value
IL-6 (pg/mL)	T0	4.68 ± 3.21	4.84 ± 3.78	−0.25	.80
	T1	21.24 ± 6.01	32.31 ± 16.86	−3.99	.00
	T2	16.32 ± 5.38	23.68 ± 9.86	−5.14	.00
	T3	16.13 ± 9.02	22.40 ± 13.66	−3.00	.00
	T4	8.46 ± 3.92	12.01 ± 7.07	−3.44	.00
TNF-α (pg/mL)	T0	18.37 ± 3.21	18.93 ± 3.29	−0.94	.35
	T1	49.21 ± 26.61	67.86 ± 30.32	−6.53	.00
	T2	44.41 ± 25.44	59.69 ± 28.46	−4.74	.00
	T3	37.64 ± 10.05	46.95 ± 11.64	−3.13	.00
	T4	21.91 ± 3.82	27.15 ± 4.99	−3.61	.00
CRP (mg/dL)	T0	2.18 ± 0.60	2.24 ± 0.83	−0.53	.60
	T1	3.42 ± 2.15	4.94 ± 2.54	−3.56	.00
	T2	6.28 ± 1.25	11.15 ± 5.84	−6.41	.00
	T3	9.81 ± 5.32	12.18 ± 2.28	−3.19	.00
	T4	15.09 ± 7.83	20.63 ± 9.70	−3.47	.00
PCT (ng/mL)	T0	0.33 ± 0.29	0.38 ± 0.29	−1.00	.32
	T1	1.26 ± 0.64	1.33 ± 0.86	−0.51	.61
	T2	2.05 ± 0.92	2.16 ± 1.40	−0.52	.60
	T3	2.47 ± 1.03	2.55 ± 1.66	−0.30	.76
	T4	0.47 ± 0.36	0.58 ± 0.44	−1.59	.11
IL-8 (pg/mL)	T0	5.62 ± 3.07	5.84 ± 1.88	−0.47	.64
	T1	7.43 ± 3.25	8.74 ± 2.55	−2.46	.02
	T2	8.12 ± 2.51	9.46 ± 2.60	−2.89	.00
	T3	10.75 ± 4.06	12.39 ± 2.25	−2.73	.01
	T4	7.68 ± 2.07	9.44 ± 3.84	−3.16	.00

CRP = C-reactive protein, IL-6 = interleukin-6, IL-8 = interleukin-8, PCT = procalcitonin, TNF-α = tumor necrosis factor-α.

Note: T0: before anesthesia; T1: 0.5 hour after the operation; T2: 12 hours after the operation; T3: 24 hours after the operation; T4: 48 hours after the operation.

4. Discussion

The side effects of perioperative opioid use include hyperalgesia, chronic postoperative pain, respiratory depression, postoperative nausea and vomiting, as well as postoperative delirium. Opioids regulate immune responses by stimulating immune cells to release cytokines via opioid receptors. Studies have shown that opioids can induce IL-4 release from T lymphocytes, and both buprenorphine and morphine can reduce the protein expression of IL-4, illustrating the 2-fold pro- and anti-inflammatory effects of opioids.^[13] OFA is a multimodal anesthesia strategy that combines a variety of non-opioid drugs and/or techniques without the use of opioids in the whole body, neuraxial or intracavitary administration during surgery to obtain high-quality anesthesia.^[14,15] QLB inhibits somatic and visceral pain, producing a wide range of local anesthetic effects, while the sensory inhibition level is mostly at T7-L1, which can be used for postoperative analgesia in the abdominal and pelvic regions.^[16]

Pro-inflammatory markers including IL-1β, IL-2, IL-6, IL-8, IL-10, TNF-α, and CRP were measured. There was no significant difference in the dosage of propofol, rocuronium and Dex, block point and block plane between the 2 groups, excluding their effects on inflammatory factors. PCT, with a specificity of 79%, is a biomarker for the early detection of (systemic) bacterial infections and can be used to determine the bacterial species and cause of systemic inflammatory responses.^[17]There was no difference in PCT between the 2 groups at any of the time points, excluding the effect of postoperative infection on inflammatory factors. In addition, the CRP level was used to assess the degree of surgical trauma and surgical stress response.^[18] IL-8, a multifunctional chemokine secreted by a variety of cell types, plays a key role in acute inflammation and promotes the chemotaxis and activation of neutrophils.^[19] Studies have shown that subjects did not have elevated serum IL-6 levels before surgery, but those who used opioid anesthesia had elevated serum IL-6 levels after surgery.^[14]

Compared with opioid anesthesia, OFA based on ESK has no adverse hemodynamic effects on pancreatectomy, while offering better analgesia, lower comprehensive complication index, and shorter 4-day hospital stay, without increasing the recurrence or readmission rate.^[20] Clinical experiments have shown that intravenous infusion of Dex combined with ESK can reduce excessive excitation of brain neurons and alleviate the mental side effects of ESK.^[21] Ketamine can enhance descending inhibition and anti-inflammatory effects in the central nervous system.^[22] In this study, group E did not use opioids, with the aim to avoid the release of inflammatory factors from immune cells stimulated by opioids, while esketamine may have the same anti-inflammatory effect as ketamine,^[23,24] so the levels of inflammatory factors IL-6, IL-8, CRP, or TNF-α at T1, T2, T3, and T4 were lower than in group C. In an animal model of depression, ESK inhibited lipopolysaccharide-induced neuroinflammation, and the levels of pro-inflammatory cytokines (including TNF-α, IL-6, and IL-1β) increased after lipopolysaccharide administration, which was reversed after 7 days of ESK treatment.^[25]

ESK, the S-enantiomer of ketamine, has a higher affinity for N-methyl-D-aspartate receptor^[26] and is twice as effective as ketamine. PCIA with ESK (0.01 mg/kg/hour) as an adjuvant can significantly reduce the incidence of postpartum depression within 14 days and relieve pain within 48 hours after cesarean section without increasing the incidence of adverse reactions.^[27] Ultrasound-guided type II thoracic nerve block combined with ESK instead of sufentanil anesthesia in modified radical mastectomy can improve the quality of early postoperative recovery and reduce the level of IL-6 24 hours after surgery.^[28] In this study, there were no significant differences in postoperative nausea, vomiting, and dizziness between the 2 groups. There was less pruritus and hypotension in group E than in group C. There were no serious complications of such as myocardial infarction and respiratory failure in either of the groups. Because nausea and vomiting are common side effects of both esketamine and opioids, there were no significant between-group differences. The incidence of hypotension was higher in group C than in group E due to sympathetic inhibition by opioids. No opioids were used in group E, so no cases of pruritus occurred. The analgesic effect was better in group E at 1 hour, 12 hours, and 24 hours after surgery. ESK is used in OFA based on QLB to inhibit stress responses in the peripheral and central nervous system, prevent and treat postoperative pain during the whole process, cut off the vicious cycle of pain, reduce the secretion of inflammatory factors, and reduce pain, while also exerting an antidepressant effect to prevent the expansion of postoperative pain signals and central sensitization. A clinical study on elderly patients undergoing radical resection of esophageal cancer showed that the levels of TNF-α, IL-6 and IL-8 first increased and then decreased, whereby the changes in the ESK group were smaller than in the opioid group,^[29] which is consistent with our observation.

The block level was tested 15 minutes after completion of the QLB, which may be too short because it has been reported in the literature that the QLB has a slow onset and takes 20 to 30 minutes to fully take effect.^[30] There is no gold standard for monitoring nociception, which means that SPI and entropy index monitoring may be insufficient. These intrinsic shortcomings of the available methodology may affect the experimental results. The limitations of this study are listed as follows; This study was conducted only in our hospital and is not a multicenter study; The anti-inflammatory mechanism of ESK has not been fully elucidated; Further studies are needed to assess the long-term effects of ESK on inflammatory factors that were not monitored.

5. Conclusion

In this study, ESK was used for OFA in patients undergoing lower abdominal or pelvic surgery based on QLB, which led to less postoperative pain and complications, less postoperative increase in IL-6, IL-8, CRP, and TNF-α, as well as less stress and inflammatory reactions during anesthesia and surgery.

Acknowledgements

This work was supported by the Department of Anesthesiology,HainanWanningPeople’s Hospital, Wanning,Hainan China.

Author contributions

Investigation: Riyue Zheng.

Methodology: Juan Li.

Software: Shanliang Li.

Writing – original draft: Jingwei Dai.

Writing – review & editing: Jingwei Dai.