The Efficacy of 20 mg Parecoxib as an Adjunct to 0.75% Ropivacaine in Supraclavicular Brachial Plexus Block for Upper Limb Surgery

Brachial plexus block (BPB) has enjoyed ubiquitous popularity for upper limb surgeries.There are a few common techniques for BPB and one of the most frequently-utilised one is supraclavicular BPB since it provides a quick-onset, efficient, safe and dense anaesthesia for surgical procedures that involve the proximal mid-humerus down to the distal hand. At present, most BPB techniques have been widely performed under ultrasound guidance to enhance the BPB's success rate and lessen its complications.

There are several adjunct drugs that can be mixed with local anaesthetics agents (e.g. ropivacaine) to speed up the onset and prolong the duration of BPB and minimise the risk of local-anaesthestic(LA)-associated overdose and toxicity. Parecoxib, a prodrug of valdecoxib, is an ideal choice for such combination due to its excellent toxicity profile.It acts by inhibiting the function of the constitutive COX-2, an isoform of cyclooxygenase (COX), resulting in the diminution of prostaglandin H2 and E2 synthesis. Consequently, this diminishes nociception by reducing sensory neuron's excitability and bradykinin-associated hyperalgesia.

So far, there is only one study which investigates the efficacy of parecoxib as an adjunct to ropivacaine in patients requiring BPB performed via axillary approach for their upper limb surgeries. However, adjunct parecoxib's efficacy in BPB performed via supraclavicular approach is still not fully scrutinised. Therefore, this clinical trial endeavours to evaluate the efficacy of adjunct parecoxib in supraclavicular BPB.

This is a randomised two parallel groups active-controlled single centre double-blind (participant and outcome assessors) clinical trial comparing the effects of parecoxib as an adjunct to ropivacaine. This trial protocol received the approval from University of Science Malaysia's Human Ethics Research Committee (HERC) on the 20th June 2016 and all participants had provided written informed consent for study participation.

The study participants were then block-randomized using a block size of 4 with a balanced 1:1 allotment ratio without any covariate stratification. The block randomization was performed using permuted block design, with the random numbers generated by a random number generator package in Stata 9.0. This was performed by an independent third party statistician. To ensure selection bias was prevented, the allocation sequence was kept inside a password-protected STATA 9.0 file which was only accessible to the independent third-party statistician. The sequence of treatment allotment was only revealed after a study participant was properly recruited and subsequently randomised to treatment allotment. To prevent ascertainment and performance bias, the study participants and the independent assessors (2nd medical officers) were shielded from the knowledge of the type of intervention received. However, the intervention was administered by the primary investigator who was not blinded to the kind of intervention received by a study participant.

To mask the participants and outcome assessors to the types of intervention received by the participants, each visually indistinguishable syringe containing one of the 2 types of intervention were pre-filled to 20 ml and labelled with code numbers by an independent third-party pharmacist before they were sent to the operating theatre (OT) for intervention administration by the principal investigator (Vivekananda Gunasekaran).

DETAILS OF INTERVENTIONS ADMINISTERED

The premedication was first prescribed in the morning of the surgery. Upon arrival at the OT, all study participants underwent standard anaesthesia monitoring for relevant clinical parameters (baseline blood pressure (BP), saturation pressure of oxygen (spO2), electrocardiography (ECG) and heart rate (HR)) which were obtained using the electric B30 monitor (Stimuplex D® plus 50mm, B. Braun, Melsungen, Germany) and documented before the BPB commenced. Subsequently, IV access was introduced using at least 20G IV cannula which was inserted on the selected hand of study participants.

Intravenous (IV) loading of Ringer's Lactate solution (B. Braun, Melsungen, Germany) at a dose of 10 ml/kg was administered before performing the block. Brachial plexus block (BPB) was performed in the block corner at the recovery bay. Drugs regime and other standard equipments for BPB were prepared and acquired and these include:

• 5 mls of Lignocaine 2% for skin infiltration
• 20 mls of Ropivacaine (Naropin®, Astrazeneca) 0.75% + 20mg Parecoxib (Dynastat®, Pfizer)(1 mls) ---Group A
• 20mls Ropivacaine 0.75% + 1 ml Normal Saline----Group B.
• Ultra Sonographic machine (Mindray® Version M5, Mindray, Shenzen, China) with high frequency (10-15MHz) linear probe
• 50 to 80mm 22 G insulated peripheral nerve block needle. (Vygon, France)
• 2% chlorhexidine in 70% isopropyl alcohol solution for skin cleaning

The BPB was implemented by a single operator (principal investigator, Vivekananda Gunasekaran) and assessed by an independent 2nd medical officer in-charge who was blinded to the treatment administered. No peripheral nerve stimulator was employed during the procedure. The detailed descriptions of the techniques utilized in the BPB are as follows:

• The block site will be cleaned and draped. The US probe also was draped for the procedure as well.
• SUPRCLAVICULAR BPB TECHNIQUE
• Subjects were positioned semi-recumbent with the head turned to the contralateral side with the ipsilateral shoulder slightly elevated with the pillow.
• An exploratory scan was performed in all patients before the block, by positioning the probe on a coronal oblique plane above the clavicle.
• Hypoechoic and pulsating supraclavicular artery was then identified, which was lying above the hyper echoic first rib. While maintaining the view of the artery, the probe was then angled until both the first rib and the pleura were also seen simultaneously.
• After skin preparation and draping, the probe was next placed in the supraclavicular fossa and subcutaneous infiltration will be given on the targeted needle side
• The needle was then inserted from lateral to medial direction in the long axis of the transducer (in-plane technique).
• Nineteen (19) ml of 0.75% ropivacaine and 1 ml of 20 mg of parecoxib (Group I: parecoxib + ropivacaine) or 19 mls of 0.75% alone plus 1 ml of normal saline (Group II: ropivacaine alone) was then injected at the "corner pocket". Adrenaline was not added into any solution.
• The remaining 5 ml was later injected to a point approximately level with the superior/ cephalad aspect of the subclavian artery, but no further than 1 cm lateral to the artery

Block performance-related pain was then evaluated immediately after removing the needle by asking the patient to verbally quantify the level of pain using a score between 0 to 10 (0 meaning no pain, 10 meaning excruciating pain). The study participants were withdrawn from the trial and rescue medications were given if one of the following withdrawal criteria occurred:

• Patient developed local anaesthetic toxicity (seizure)
• Patient developed hemodynamically instability bradycardia/hypotension)
• Patient developed anaphylaxis

The procedures for the assessment of sensory and motor block are as follows:

• ASSESSMENT OF SENSORY BLOCK
• Sensory blockade was assessed every 5 minutes within the first 30 minutes following the completion of drug administration. The magnitude of sensory blockade is graded as follows:
• Grade 0 = normal sensory response
• Grade 1 = reduced sensory perception (partial sensory blockade)
• Grade 2 = no sensation (complete sensory blockade).
• Sensory loss was confirmed by the loss to cold sensation using 10 mls cold saline bottle and pinprick sensation using 23G needle in all dermatomes of the brachial plexus (Grade 2).
• Time zero was defined as the time at which LA was completely injected.
• Sensory block success was defined as complete pin-prick sensory blockade in all dermatomes of the brachial plexus (C5-T1).
• The block was considered incomplete if any supplemental local anesthetic is needed for complete anesthesia.
• The block was considered failed if the desired volume did not provide complete anesthesia or conversion to general anaesthesia was required prior to surgery.
• General anaesthesia was routinely performed with intravenous induction (sedation) agent, short acting opioids and muscle relaxant
• ASSESSMENT OF MOTOR BLOCK
• Motor block was assessed by subject's capability of flexing his / her elbow and hand against gravity. This was then graded according to the following scale:
• Grade 1: ability to flex or extend the forearm
• Grade 2: ability to flex or extend only the wrist and fingers
• Grade 3: ability to flex or extend only the fingers
• Grade 4: inability to move the forearm, wrist, and fingers
• INTRA-SURGICAL ASSESSMENTS
• Requirement of block supplementation, surgical wound infiltration and patient requested sedation or general anaesthesia
• Surgical anaesthesia success was defined as surgery without the requirements of block supplementation, general anesthesia (administered for incomplete block) or surgical site infiltration
• Haemodynamic monitoring was performed at baseline, after LA injection, after 15 min, 30 min, 1 hour of block procedure \& after completing surgery
• POSTOPERATIVE ASSESSMENT
• Sensory and motor blockade durations were assessed on half-hourly basis (up to 12 hours postoperatively), during the post surgical period.
• Patient-controlled analgesia (PCA) morphine or IV tramadol was given as rescue analgesia.
• Pain scores were assessed using the visual rating scale (VRS) (0-10) where pain was evaluated when the patients were resting at 1, 2, 4, 12, 24 hours postoperatively.
• The duration of analgesia (time interval from the completion of local anesthetic administration until the first need of rescue analgesia in the form of PCA morphine or IV tramadol) and the amount of morphine or tramadol consumed during the postoperative 24 hours were also documented.
• Any evidence of complications (e.g., bruises/swelling at the block site, chest pain/ breathing difficulty, dysaesthesia/ muscle weakness in the operated extremity not related to the site of operation) were also recorded.
• Surgeons were alerted to report any neurological problems not related to surgery during the clinical rounds prior to the patients were discharged from the hospital.
• Anaesthetic preferences (one of the following: 1) the BPB; 2) block under deep sedation; 3) block under GA).
• Preferred block for future hand operations was then recorded.

SAMPLE SIZE DETERMINATION

The sample size was calculated using power analysis method. The level of significance was fixed at 0.05, power (1 - type 2 error) at 0.80 and a dropout rate of 20%. Based on the standard deviation of 140 minutes and a hypothesised mean difference in total duration of sensory block of 104 minutes, the total sample size is 39 subjects per intervention arm.

STATISTICAL ANALYSES

The data was analysed using Statistical Package for Social Science for Windows version 23 (SPSS 23) and Stata version 11. Missing data were treated as Missing at Random (MAR) under Rubin's missing data mechanism. Multiple imputation was then used to address the missing data problem. The data were descriptively summarised using mean and standard deviations (or interquartile range) for continuous outcomes or frequency and percentage for categorical variables. Multiple imputation was then used to address the missing data. The mean differences in continuous outcome variables (onset and duration of sensory and motor blockades) between the intervention arms were analysed using independent t-test or Mann-Whitney test. For categorical outcome variables (complete motor and sensory blockades at 30 minutes per intervention), Chi-square or Fisher exact tests were utilised. The assumptions of normality and homogeneity of variances were checked using histogram with overlying normal plots, box plots, Mann-Whitney (normality) and Levene's tests (homoscedasticity of residuals).