The Optimal Administration Time of Butorphanol in Suppressing Sufentanil-induced Cough: A Single-Center Randomized Double-Blind Trial

Backgroud: To evaluate the optimal administration time of butorphanol in suppressing sufentanil-induced cough(SIC) during the induction of general anesthesia. Methods:180 patients were randomly divided into 4 equally sized groups: at 5min, 2min and 0min before anesthesia induction, all patients were sequentially injected the drug labeled A/B/C (butorphanol 1mg or normal saline), GI received intravenously drug A/B/C (All were normal saline),GII received intravenously drug A/B/C (A was butorphanol, B and C were saline), GIII received intravenously drug A/B/C (B is butorphanol, A and C were saline), GIV received intravenously drug A/B/C (C is butorphanol (cid:0) A and B were saline). ALL Patients were then administrated with sufentanil 0.4 µg/kg in 5s after drug C. The incidence and severity of SIC was recorded within 2 minutes after sufentanil injection. MAP, HR, and SpO 2 were recorded at T0 (before the administration of any drug),Ta, Tb, Tc (before the injection of drug A/B/C), T1( 2minutes after sufentanil injection) and T2 (1 minutes after endotracheal intubation). Results: The incidences of cough in GII, GIII, and IV were lower than that in GI (0.09, 0.01, and 0 vs 42.2%, P<0.01), while there were no signicant differences between GII, GIII, and GIV. The HR of all 4 groups at T2 were signicantly higher than their levels at any other time (P<0.05,T2 vs other time), but there’s no signicant difference among 4 groups at T2. The MBP of all 4 groups at Ta and Tb were signicantly lower than their levels at any other time (P<0.05, Ta and Tb vs other time), but there’s no signicant difference among 4 groups at Ta and Tb. At Ta,


Background
During the induction of anesthesia, sufentanil induced-cough(SIC) is less likely to attract the attention of clinical anesthesiologists than fentanyl, yet this phenomenon may be undesirable in any patients, the cough rate caused by sufentanil can be as high as 47% [1] .
Butorphanol is considered to be a mixed agonist-antagonist opioid analgesic that has a nity for µ-, δ-, and κ-opioid receptor subtype(a nity in vitro of 1:4:25), with extensively used in clinical practice due to a potent analgesic effect and minimum side effects [6,7] . But the time of administration on preventing SIC is unintelligible,which was ranged from 5 s-15 minutes.When it should be given before sufentanil was still unclear [8][9][10] , and there is no investigation regarding the administration time of butorphanol on preventing unwanted SIC. So this trial was conducted to evaluate the optimal administration time of butophanol on preventing SIC during the induction of general anesthesia. Patients were unpremedicated and fasted. After arriving the operation room, a venous pathway was established with an intravenous (IV) cannula (18G) on the dorsum of the forearm, or a 18F catheter on the right internal jugular vein, Ringer lactate was infused at a rate of 4-6 ml/min. Then each patient was received routine electrocardiogram (ECG), pulse oxygen saturation (SpO 2 ), blood pressure (BP), heart rate (HR). Before anesthetic induction, the pretreatment drug, according to the random numbers, butorphanol (batch 16062232, Hengrui Medicine, Jiangsu ,China) or saline was prepared in labeled A, B and C 5 mL syringes outside the operating room by an anesthesiologist who was not involved in the induction of anesthesia.

Methods
180 patients were assigned at a 1:1:1:1 ratio to one of 4 groups by using computer-generated table of random numbers: 5 min, 2 min and 0 min before induction, all patients were sequentially injected the drug labeled A, B and C (butorphanol 1 mg or normal saline), group I received intravenously drug A, B and C (All were normal saline, served as a negative control for sufentanil), group II received intravenously drug A, B and C (A was butorphanol,B and C were saline), group III received intravenously drug A, B and C (B is butorphanol A and C were saline), group IV received intravenously drug A, B and C (C is butorphanol A and B were saline). ALL Patients were then administrated with sufentanil 0.4 µg/kg in 5 s after drug C.The incidence and severity of cough was recorded for 2 minutes after sufentanil injection, the severity of cough was graded, based on the numbers of cough, as none (0), mild (1-2), moderate (3)(4), and severe (5 or more). If SpO 2 dropped below 95%, manually assisted mask ventilation with oxygen was applied immediately. Then propofol 2 mg/kg and rocuronium 0.6 mg/kg were given to facilitate tracheal intubation.
During experimental period, the mean arterial pressure(MAP), HR, and SpO 2 were recorded at T0 (before the administration of drugs), Ta, Tb and Tc (just before drug A, B and C injection, T1 (2minutes after sufentanil injection and T2 (1 minute after endotracheal intubation). Intravenous route, surgery type, duration and BMI were also recorded, and after operation, the total amount of sufentanil, postoperative VAS pain score were recorded.
In a pilot study, the incidence of cough induced by 0.4 µg/kg of sufentanil was 40% (20/50) when it was administered within 5 seconds. Therefore, we assumed that if the incidence of cough was reduced by 50%, the effect of butorphanol could be considered statistically signi cant. To achieve 80% statistical power with α = 0.05, We factored in a 10% dropout rate and enrolled 45 patients in each group.
Statistical Analysis: Data are expressed as the mean ± standard deviation, number, proportion, or percentage. Comparison of age, BMI, surgery duration, the total amount of sufentanil, the VAS pain score and the additional amount of sufentanil were analyzed by one-way Analysis of Variance. The Intravenous route, gender, ASA physical status, incidence of cough, and cough severity were compared using Pearson chi-square test or Fisher's exact test. Comparison of MAP, HR and SpO 2 were analyzed by repeated measures two-way analysis of variance. SPSS20.0 (SPSS Inc, Chicago, IL) and GraphPad Prism5.0 software (GraphPad Software Inc, San Diego, CA) were used for statistical analysis. A P-value of < 0.05 was considered to be signi cant.

Demographic data
In the present study, 200 patients were recruited and 20 patients were excluded, as they met the exclusion criteria (3 subjects declined participation, 7 had a habit of smoking, 3 had a history of asthma, 2 had impaired liver function and 5 had Aneurysm). Therefore, a total of 180 patients were randomized into 4 groups of 45 each and included in the nal analyses. There were no signi cant differences in age, gender, BMI, ASA status, intravenous route, surgery type, duration the total amount of sufentanil (and) postoperative VAS among the four groups(see Table 1 and Table 2).  The incidences of cough in group II, group III, and group IV were lower than that in group I (0.09, 0.01, and 0 vs 42.2%, P < 0.001), while there were no signi cant differences between group II, group III, and group IV. The moderate and severe FIC was not observed in group II, IIIand IV, but was recoded from 13 patients in group I. (see Table 3) Table 3  Incidence  The HR of all 4 groups at T2 were signi cantly higher than their levels at any other time (P < 0.05, T2 vs other time in all 4 groups), but there's no signi cant difference among 4 groups at T2 (Fig. 2). The MBP of all 4 groups at Ta and Tb were signi cantly lower than their levels at any other time (P < 0.05, Ta and Tb vs other time in all 4 groups), but there's no signi cant difference among 4 groups (Fig. 3). The SpO 2 of group II at Ta were signi cantly lower than group I, group III and group IV(P < 0.001), group II at Tb were signi cantly lower than group I(P < 0.05), and group IV(P < 0.01), group III at Tc were signi cantly lower than group IV(P < 0.01) (Fig. 4).

Discussion
In our study, we found that the incidences of cough in group II, group III, and group IV were lower than that in group I (0.09, 0.01, and 0 vs 42.2%, P < 0.001), while there were no signi cant differences between group II, group IIIand group IV, it means that preemptive infusion of 1 mg butorphanol before sufentanil bolus administration reduced the incidence of SIC. However, SpO 2 of group II at Ta and Tb were signi cantly lower than other groups, group III at Tc were signi cantly lower than groupIV, the drop of SpO2 may be due to the butorphanol of respiratory depression, so we can suppose that without waiting time is an effective and clinically feasible method for suppressing sufentanil-induced cough during general anesthesia induction. As for MBP of all 4 groups at Ta and Tb were signi cantly lower than their levels at any other time, may be due to the psychological comfort caused by the injection of drugs to the patient.
Sufentanil is an opioid with less impact on hemodynamics and strong analgesic effect, which makes it as an ideal option in anesthesia induction [11] . However, opioid-induced cough(OIC) caused by injecting sufentanil was commonly observed during the induction of anesthesia. A meta-analysis reported that relational factors of OIC can be divided into 2 aspects: the patients' individual physical conditions (age, sex, smoking and disease history), and the usage of opioids (drug category, dosage, concentration, administration order, injection rate, and site) [12] . Lin et al [13] .found that only light smoking(< 10 cigarettes per day or < 10 smoking years or < 10 packyears), not heavy smoking, could decrease the frequency of fentanylinduced cough, but the association between age and incidence of cough was not observed in their study. Oshima et al [14] .completed a cohort study from 1311 adult patients, these authors reported that increasing age, cigarette smoking were associated with a decreased risk of fentanyl-induced cough. OIC is a risk factor for postoperative nausea and vomiting, which means young female nonsmoker may decrease the incidence of postoperative nausea and vomiting by reducing the happening of OIC [15] . However, in a randomized, double-blind study, they found no association of age or smoking status on the incidence of cough in any of the groups [15] . So the relationship between patients' individual physical conditions and OIC is remain controversial. However ,there is no doubt that with the increase of opiods dose and concentration can rise the happening of OIC [17][18][19] . While does selection of central or peripheral administration of sufentanil affect opioid induced cough? A prospective, randomized, controlled trial was inconsistent to previous literature,they suggested that sufentanil administered through central vein reduces the occurrence and severity of SIC [17][18][19] .
The mechanisms of OIC remains unclear, but some theories have been proposed to explain this phenomenon. First, the dualism of opioid receptor, substances like histamine and neuropeptides, released by an action on prejunctional µ-opioid receptors after fentanyl administration [17][18][19] , Kamei et al [23] .showed that fentanyl enhances the excitability of rapidly adapting receptors to cause cough via the enhancement of histamine release in rat's airways. In our study,intravenous butorphanol without waiting time before general anesthesia induction effectively suppressing sufentanil-induced cough may though this mechanisms.Second, inhibition of central sympathetic out ow causes vagal predominance inducing cough and re ex bronchoconstriction [24] . Third, pulmonary chemore ex resulting from the stimulation of C-ber receptors(Juxta-capillary receptors) or irritant receptors(rapidly adapting receptors) from deformation of the trachea-bronchial wall by tracheal smooth muscle constriction [25][26][27] . Finally, Sufentanil and fentanyl preparations are available as citrate salts; there is a possibility that, when sufentanil and fentanyl are injected intravenously, their citrate component may stimulate the pulmonary chemore ex [28] .
Butorphanol, an agonist-antagonist opioid, can antagonize the µ-receptor or preempted the lung related receptors, to prevent the cough re ex while acting on the C-ber receptor to inhibit the afferent pathway of the cough re ex [29][30][31] . The more possible reason that butorphanol suppressing cough just before injection of sufentanil could be explain by priming use of butorphanol may result in depletion of neurotransmitters in peripheral nerve bers [2] , as Hoffmann et al.indicated that the interval between the start of injection and the beginning of signal ampli cation in the ICA was noted as 'arm-to-head time' (AHT), and AHT was 9.0-22.0 s(14.3 ± 3.0 s; mean ± SD) [32] . Otherwise, it is hard to explain that butorphanol suppress cough just before or in the same time to inject sufentanil. Moreover, the administration of intravenous butorphanol prior to induction of anaesthesia helps on better attenuation of the haemodynamic response to laryngoscopy and endotracheal intubation [33] .
There are some limitations that are relevant to our study. First, Our butorphanol did not calculate the required dose based on the patient's weight. It has been reported that 0.1 mg butorphanol can also completely inhibit SIC,just 5 seconds right before sufentanil bolus. Second, we did not record the effects of butorphanol on postoperative nausea and vomiting. Third, our study provides evidence for the e cacy of butorphanol in suppressing cough just before the sufentanil injection in clinical practice; but we did not verify the exact mechanisms, more studies need to be performed to reveal the exact mechanism.

Conclusion
We suggest that intravenous butorphanol 1 mg just before the sufentanil bolus, without waiting time is an effective and clinically feasible method for suppressing sufentanil-induced cough during general anesthesia induction.

Declarations
Ethics approval and consent to participate Ethical approval was given by the institutional ethics committee (Zhongshan Hospital, Fudan University (B2019-129),Shanghai China) and registered at Chinese Clinical Trial Registry with registration number ChiCTR1900024394.Written informed consent of participation is obtained from all participants.

Consent for publication
Not applicable.

Competing interests
The authors declare that they have no competing interests.

Funding
No funding received for the work.
Authors'contributions XZ analysis, paper writing and most part of clinical studies. SY data mining and part ofclinical studies.LQ & CY experimental design,ideal conception,and paper writing. All authors read and approved the nal version of the manuscript.