Extrapleural infusion of levobupivacaine versus levobupivacaine‐sufentanil‐adrenaline after video‐assisted thoracoscopic surgery (VATS): A randomised controlled trial

Peripheral blocks are increasingly used for analgesia after video‐assisted thoracic surgery (VATS). We hypothesised that addition of sufentanil and adrenaline to levobupivacaine would improve the analgesic effect of a continuous extrapleural block.


Editorial Comment
This trial compared 72 h postoperative extrapleural infusion of levobupivacain to a mixture of lower concentration levobupivacaine + adrenalin + sufentanil.All patients received standard multimodal non-opioid analgesia, intraoperative subpleural bolus of mepivacaine and opioid as needed.No difference was found between groups, but the trial had low sensitivity as both groups had low opioid usage and pain levels.Furthermore, no placebo group was used to show the true effect of postoperative infusions.Future trials could consider investigating the addition of perioperative glucorticoids to reduce surgical stress and prolong the effect of perioperative blockades.

| INTRODUCTION
Minimally invasive techniques have largely replaced thoracotomy for lung surgical procedures.[3][4][5][6][7] Whereas continuous epidural analgesia (CEA) is the preferred pain treatment for open lung surgery, this is less obvious after VATS. 8,9e goal of any thoracic block, whether as CEA or a peripheral nerve block, is to provide opioid-sparing postoperative analgesia that facilitates recovery. 10Our earlier findings 11 showed that despite a continuous extrapleural block [12][13][14][15][16] with levobupivacaine at 13.5 mg h À1 , additional multimodal analgesia consisting of oral paracetamol, naproxen, and oxycodone was needed for adequate pain control.
Various adjuncts are added to local anaesthetics (LAs) with the goal to improve neuraxial and peripheral block efficacy through an additive or even synergistic effect.[23][24] This trial aimed to investigate whether a 72-h continuous extrapleural infusion of a lower concentration of levobupivacaine with the adjuncts sufentanil and adrenaline, a mixture that is effective in CEA, would result in lower postoperative opioid requirement than our standard infusion of plain levobupivacaine at the maximum recommended dose after lung VATS.

| Trial design
In this prospective, parallel, double-blinded trial, we randomised 60 study participants to receive either levobupivacaine only or levobupivacaine/sufentanil/adrenaline in a continuous extrapleural block.
The intervention covered the first 72 h after surgery with a follow-up at 3 weeks postoperatively.

| Participants
All patients scheduled for lung resection by VATS at the Karolinska University Hospital were considered for enrolment.Patients aged >20 years, with an American Society of Anesthesiologists (ASA) physical status classification of II or III, and the ability to give informed consent were eligible for participation.Reasons for ineligibility were allergy to LAs, pronounced hepatic disease, psychiatric disease or use of psychoactive medication, cognitive impairment and/or inability to understand written and/or oral instructions, and a history of chronic pain or chronic use of analgesics.Exclusion criteria were inability to place an extrapleural catheter.
Written informed consent from each patient who agreed to participate in the study after having received oral and written information about the study was obtained by the investigator ML.

| Intervention
Except for the allocated treatment and the use of a patient-controlled analgesia (PCA) pump, all study participants received care according to standard department practice.Prior to surgery, participants were informed how to use an intravenous PCA pump.Consistent to department practice, participants received intravenous anaesthesia with remifentanil and propofol, were intubated with a double-lumen endotracheal tube, and underwent the procedure in the lateral decubitus position with one-lung ventilation.

| Procedure for regional anaesthesia
Before lung resection was initiated, and under thoracoscopic control, a multi-hole 19-cm silver-coated catheter (ON-Q Soaker; Avanos Medical, Alpharetta, GA) was placed extrapleurally by the surgeon.A tunnelling stylet covered by a sheath was inserted percutaneously approximately two costal levels below the caudal port and 5-8 cm lateral to the spine.Subpleurally and parallel to the spine, the stylet was carefully advanced in a cephalad direction to a position superior to the most cranial port incision.After removal of the stylet and insertion of the catheter, the sheath was peeled back and removed.Then, a bolus containing 10 mL of levobupivacaine at 7.5 mg mL À1 was administered.

| Start of the intervention
After conclusion of the procedure, a 200-mg bolus of mepivacaine was administered subpleurally.Depending on allocation, a single-use elastomeric pump containing either levobupivacaine at 2.7 mg mL À1 (LB group) or levobupivacaine at 1.25 mg mL À1 , sufentanil at 0.5 μg mL À1 , and adrenaline at 2 μg mL À1 (LBSA group) was then attached to the extrapleural catheter for infusion at a rate of 5 mL h À1 .

| Postoperative care and assessment
In the recovery ward, intravenous morphine could be given if required to achieve mild pain.When pain was rated as mild at the most, a PCA pump with morphine was connected for the duration of the intervention (morphine bolus dose of 1 mg, 6-min lock-out time, 4-h maximum dose of 30 mg).Standard basic analgesic treatment was provided with oral paracetamol at 1 g four times daily and naproxen at 500 mg twice daily.In the morning and afternoon of each day during the intervention, participants scored their pain from 0 to 10 on a numerical rating scale (NRS), at rest and after two deep breaths.No pain was scored as 0, mild pain as 1-3, moderate pain as 4-6, and severe pain as 7-10.
Preoperatively and in the afternoon of all postoperative days (PODs) participants performed three consecutive peak expiratory flow (PEF) measurements.The quality of recovery (QoR)-15 questionnaire 25 was completed preoperatively (POD0), 1 day postoperatively (POD1) and 3 weeks postoperatively (POD21).Blood samples were taken before surgery, 1 h after the start and at the end of the intervention.After centrifugation at 2000g for 10 min, serum was transferred to micro tubes and stored at À70 C until analysis.
Participants were monitored for any adverse events.Investigator ML performed the daily follow-up.

| Analysis of levobupivacaine
The serum levobupivacaine concentrations were analysed at the Department of Clinical Pharmacology, St. Olav University Hospital, Trondheim, Norway using ultra-high-performance liquid chromatography-tandem mass spectrometry.Sample preparation was performed using a Hamilton Microlab STAR pipetting robot (Hamilton, Bonaduz, Switzerland).Fifty microlitre of serum samples and 25 μL of the internal standard bupivacaine-d 9 were pipetted onto an Ostro 96-well plate (Waters Corp., Milford, MA, USA).Thereafter, 900 μL ice cold acetonitrile with 1% formic acid was added for protein precipitation, and the content in each well was filtrated using a positive pressure processor.Standards and quality controls were prepared as the serum samples.The limit of quantification was 0.5 μmol/L (0.14 μg/mL) and the method was linear at least up to 25 μmol/L (7.20 μg/mL).Samples with detectable concentrations under 0.5 μmol/L were reanalysed using a more sensitive method with standards reaching down to 0.005 μmol/L (0.0014 μg/mL).In this preparation, the volumes were 100 μL for serum samples/standards/quality controls, 25 μL for the internal standard bupivacaine-d 9 and 300 μL for ice cold acetonitrile with 1% formic acid.Principles for separation and detection were similar to what has been described above.

| Primary outcome
The intended primary outcome was the cumulative intravenous PCAmorphine administration at 72 h.

| Secondary outcomes
The secondary outcomes were the opioid requirement on POD1-3; NRS scores at rest and after two deep breaths at 6, 24, 30, 48, 54 and 72 h; the average of three consecutive PEF measurements on POD0-3; the QoR-15 score on POD0, POD1 and POD21; the serum levobupivacaine concentrations at 1 h after the start and at the end of the intervention; and the incidence of sedation, nausea, vomiting and pruritus.

| Amended trial outcomes
Twenty-three study participants were discharged before 72 h, resulting in premature termination of the intervention.We therefore report the primary outcome as cumulative PCA-morphine consumption at both 48 and 72 h.
We describe pain in ambulation as secondary outcome in the European Union Drug Regulating Authorities Clinical Trials Database (EudraCT).Our intention was to assess pain when moving.We concluded deep breaths is the easiest reproducible and most painprovoking movement of the thorax.We therefore report pain after two deep breaths.
The outcome measurement of serum levobupivacaine concentrations is not stated in EudraCT.However, we planned for this outcome when we planned the study and it was approved by the Swedish Ethical Review Authority and explained in the trial participant information.
Post-operative Quality Recovery Scale (PQRS) and QoR-15 are both scores for evaluation of post-operative recovery.During study design we chose to include PQRS as a secondary outcome, as stated in EudraCT.In preparation of our trial, QoR-15 became the predominantly reported score.As both scores measure postoperative recovery, we decided to use the most frequently used to facilitate future comparison.

| Sample size calculation
Based on clinical follow-up of patients receiving an extrapleural catheter during the year 2014, we assumed a mean accumulated 72-h intravenous morphine requirement of 28 mg (standard deviation 13 mg).
We hypothesised that the LBSA solution would decrease the morphine requirement by 10 mg.Assuming an alpha error of 0.05 and power of 80%, we calculated that 27 patients in each treatment group were needed to detect a statistically significant difference.Allowing for a 10% dropout and protocol violation rate, we planned to enrol 60 patients.

| Randomisation
Treatment was randomised in a 1:1 fashion in blocks of 10.Before the trial, sequentially numbered opaque envelopes were prepared by investigator ML.With the text faced down, treatment allocations were then randomly inserted.The sealed envelopes were kept at the locked local pharmacy at our department.At enrolment, each trial participant received an inclusion number.After surgery started, one of two nurse anaesthetists who were otherwise not involved in the trial opened the envelope with the number corresponding to the participant's inclusion number.Based on the allocation, an elastomeric pump prepared by the hospital pharmacy with the correct treatment was selected, blinded and delivered to the team caring for the trial participant.

| Blinding
The trial participants, investigators, surgeons, attending anaesthesiologist and nurse anaesthetist and nurses in the recovery room and surgical ward were blinded to the allocation.
After conclusion of surgery and until the PCA pump was started in the recovery room, the LB group received 5 mg (IQR, 3-6 mg) and LBSA group 8 mg (IQR, 4-10 mg) of morphine, respectively ( p = .046).

| Pain scores
Median NRS score at rest and after two deep breaths was 3 or lower at all time points for both treatment groups (Table 3).At rest, there was no difference between groups in NRS score at the individual time points.Following two deep breaths, the NRS score at 24 h was 2.5 (IQR, 1-3) in the LB group and 3 (IQR 2-5) in the LBSA group (p = .029).At 30 h, the NRS score was 1 (IQR, 0-3) respectively 3 (IQR, 2-4; p = .022).

| Other outcomes
PEF PEF values decreased significantly from baseline to POD1 (Figure 2).
There was no difference between groups at any time point.

QoR-15
The QoR-15 score decreased significantly on POD1 for all participants, without difference between groups (Figure 3).Three weeks after surgery, the QoR-15 score increased significantly from 118 to 131 in the LB group.In contrast, the LBSA group had a non-significant increase from 117 to 124.

| Serum levobupivacaine concentration
There was no difference in the serum levobupivacaine concentration between the two groups at 1 h after administration of the 75-mg levobupivacaine bolus (Figure 4).In contrast, the levobupivacaine concentration was significantly higher in the LB group at both 48 and 72 h.
An outlying levobupivacaine concentration of 2.54 μg mL À1 was found at 1 h after the levobupivacaine bolus in a patient who underwent intraoperative rib resection to allow removal of the tumour.

| Post-hoc analyses
When three participants who underwent wedge resection were excluded from analysis, median cumulative PCA-morphine dose did not change for the treatment groups.
We categorised pain to mild (NRS = or <3) and moderateto-severe (NRS >3;  points. 26,27The LBSA group on the other hand showed a nonsignificant increase of 7 points during the same time.
8][19][20] In contrast, their efficacy in peripheral continuous blocks is not as obvious, [21][22][23][24] although animal models have shown the presence of opioid receptors in peripheral nerves and their upregulation in inflamed tissue. 28For a continuous paravertebral block after VATS, Bauer and colleagues 29 found the addition of 0.25 μg mL À1 of sufentanil to 2 mg mL À1 of ropivacaine at 0.15 mL kg À1 h À1 reduced neither pain nor morphine consumption.In contrast to our study, they observed a high 48-hour morphine requirement of >50 mg in both study groups.Considering that the ropivacaine dose they used has been effective in previous studies, 14,30 the lack of block improvement could be explained by an insufficient dose of sufentanil. 31 the other hand, increasing the adjunct opioid dose too much comes at a cost.Burlacu and colleagues 32 added 4 μg mL À1 of fentanyl to 0.5 mg mL À1 of levobupivacaine, comparing it to levobupivacaine at 1 mg mL À1 in a continuous paravertebral block after breast surgery.The 24-h rescue morphine doses were 7.9 mg and 27.7 mg for the levobupivacaine-fentanyl and the levobupivacaine-only groups, respectively.Despite the significantly lower morphine requirement, the levobupivacaine-fentanyl group experienced more nausea and vomiting.Importantly, the authors also concluded that levobupivacaine at 1 mg mL À1 did not deliver an optimal paravertebral block.
Using LA and an opioid at doses comparable to the doses used in our present trial, Mohta and colleagues 33 observed similar results.
T A B L E 3 Pain at rest and after two deep breaths according to the NRS score at different time points for the LB and LBSA groups.Treating pain after rib fractures, they added 2 μg mL À1 of fentanyl to ropivacaine at 2 mg mL À1 with adrenaline at 5 μg mL À 1 for a 0.1-0.2mL kg À1 h À1 continuous paravertebral block.The mixture performed as well as ropivacaine at 3.75 mg mL À1 with adrenaline at 5 μg mL À1 .
With the disadvantages of opioid adjuncts in mind and our trial showing a possibly superior effect of a block with the maximal dose of LA only, the question is whether a prolonged continuous high dose LA nerve block is a safe strategy?The serum levobupivacaine concentrations in our trial were well below the concentration limit of 2.62 μg mL À1 shown to cause toxic central nervous system symptoms. 34The initial bolus of 75 mg levobupivacaine was followed by a 24-hourly dose of 324 and 150 mg, resulting in 48-and 72-h serum concentrations of 0.97 and 0.94 μg mL À1 in the LB group and 0.63 and 0.31 μg mL À1 in the LBSA group, respectively.This is consistent with the mean plasma concentration of 0.68 μg mL À1 that Burlascu and colleagues 35 found in another trial after a 24-h paravertebral dose of 236.2 mg of levobupivacaine.
We found one outlying concentration of 2.54 μg mL À1 at 1 h after the levobupivacaine bolus.Probably it was the extended surgery that allowed rapid intravascular levobupivacaine uptake.The study participant experienced nausea and vomiting during the first postoperative night.These are not typical symptoms of mild LA systemic toxicity (LAST), 36,37 but we cannot rule out a possible association.At 72 h, the serum concentration was at 0.26 μg mL À1 .
Although our levobupivacaine bolus dose needs re-evaluation, the continuous infusion of levobupivacaine at 13.5 mg h À1 was reassuringly safe.

| Limitations
The power in our study was negatively affected by three factors.First, the morphine consumption was considerably lower than expected from our sample size calculation.Second, pain throughout the intervention was mild.Finally, 52% of our study participants stopped the intervention prematurely.We therefore lack data on pain scores and morphine consumption for the final day of the planned intervention.
Our current findings stand in contrast to our previous study. 11mpared to the cumulative 3-day oxycodone consumption of 35 mg, the current study's low 72-h cumulative morphine dose of 11 mg could be explained by the fact that the vast majority of procedures were performed with fewer surgical access ports (Table 1).

| CONCLUSION
For a continuous extrapleural block after VATS lung surgery, this trial compared plain levobupivacaine at 13.5 mg h À1 to a mixture of levobupivacaine at 6.25 mg h À1 , sufentanil and adrenaline and found no significant difference in postoperative PCA morphine consumption.
The higher dose of plain levobupivacaine possibly provided better pain control during movement on the first postoperative day and enhanced recovery after 3 weeks.Regardless, serum concentrations after 48 and 72 h of infusion were well below toxic, which supports the use of the maximally recommended dose of levobupivacaine for a 2-to 3-day continuous extrapleural block.
T A B L E 4 Percentage of participants experiencing moderate-to-severe pain (NRS >3) over time, at rest and after two deep breaths, for the LB and LBSA groups.
Acquity UPLC I-class coupled to a Waters Xevo TQ-S tandem-quadrupole mass spectrometer (Waters Corp., Milford, MA, USA).Chromatographic separation was achieved on an Acquity UPLC HSS T3 (2.1 Â 100 mm, 1.8 μm) column with an Acquity UPLC HSS T3 VanGuard pre-column (2.1 Â 5 mm, 1.8 μm), using gradient elution with mobile phase consisting of acetonitrile and 0.1% formic acid in water.Total run time was 2.00 min.The Xevo TQ-S was operated in positive electrospray ionisation (ESI) mode with multiple reaction monitoring.Bupivacaine was detected using the mass transitions m/z 289.2 > 84.1 and m/z 289.2 > 98.0, for quantification and qualification respectively.The mass transition m/z 298.3 > 140.0 was used for detecting bupivacaine-d 9 .

2. 8 |
Statistical methods Data were analysed on a per-protocol basis.All data are summarised as median [interquartile range (IQR)] for continuous variables and number and/or frequency (percentage) for categorical variables.Demographic and outcome data were analysed for continuity.Nonnormally distributed data were analysed with the Wilcoxon rank-sum test.Repeated measurements of non-normally distributed data were analysed with the Friedman test and Wilcoxon signed rank test for individual comparisons.Categorical data were analysed with Pearson's chi-squared test.A p-value of <.05 was considered statistically significant.Statistical analysis was performed using Stata 14.2 software (StataCorp, College Station, TX).

From 23
October 2017 to 27 April 2020, 76 patients scheduled for VATS were considered for participation.The CONSORT diagram is shown in Figure 1.Sixteen patients were ineligible, and the remaining 60 patients were enrolled.After enrolment, four participants were excluded.Data of the remaining 56 participants were analysed per protocol.The participants' characteristics, surgical procedure, number of ports used and what day the chest drain was removed are presented in Table1.Twenty-three trial participants (41%) were discharged before 72 h (14 in the LB group, 9 in the LBSA group; p = .423).For practical reasons, the intervention was stopped at 48 h for two additional participants in the LB group and four in the LBSA group.There was no significant difference in length of stay between the LB and LBSA groups [3 (IQR, 2-3) days and 3 (IQR, 2-5) days, respectively; p = .234].

T A B L E 2 15 Note:
Cumulative patient-controlled analgesia morphine use in mg over time in LB and LBSA groups.Data are presented as median (interquartile range) and n, number.Abbreviations: LB, levobupivacaine; LBSA, levobupivacaine/sufentanil/ adrenaline.This trial assessed whether a well-established epidural drug combination consisting of a low dose of long-acting LA with addition of sufentanil and adrenaline would provide superior postoperative analgesia, and thereby decrease postoperative opioid requirement, compared with the maximally recommended dose of the long-acting LA only, administered as a continuous extrapleural block during the first 72 h after VATS.We found no difference in our primary outcome of cumulative opioid requirement at 48 and 72 h.Overall, pain was well controlled and mild.The only difference in pain occurred after two deep breaths in favour of the LB group, but the clinical significance of this result is uncertain.Interestingly, the LB group recovered better 3 weeks after surgery.The LB group showed a significant increase in QoR-15 score of 13 from POD1 to POD21, a difference that is well over the suggested minimal clinically important difference of 6-8

Table 4
). Significantly fewer participants in the LB group experienced moderate-to-severe pain at 24 h, both at rest and after two deep breaths, as well as at 30 h after two deep breaths.Note: Data are presented as n (%) or median (interquartile range).Abbreviations: LB, levobupivacaine; LBSA, levobupivacaine/sufentanil/adrenaline; POD, postoperative day.