Controlled,non-inferiority trial of bupivacaine liposome injectable suspension

Abstract

Objectives

Recently, a bupivacaine liposome injectable suspension (BLIS) became commercially available in the veterinary market that may provide local analgesia for 72 h. The purpose of this study was to compare a BLIS incisional block with a control protocol in cats after ovariohysterectomy (OHE). The hypothesis was that a BLIS block would provide equivalent pain relief.

Methods

This study was designed as a randomized, double-blind, non-inferiority trial. Students performed an OHE followed by a two-layer incisional and body wall block with either standard bupivacaine (control) or BLIS. Postoperatively, cats in the control group received robenacoxib, whereas the BLIS cats received saline. All cats were evaluated using the feline Glasgow Composite Measure Pain Scale (GCMPS) at multiple time points postoperatively.

Results

There were 24 control cats and 23 BLIS cats. One cat from each group required rescue medication. The mean GCMPS scores were low and the groups were equivalent at all time points (P <0.05). This study showed that BLIS was equivalent to the control group up to 42 h and pain scores remained low up to 68 h after surgery.

Conclusions and relevance

BLIS incisional block is equivalent to a control pain protocol and reduces the need for continued postoperative drug administration.

Introduction

High-volume, low-cost spay–neuter programs are prevalent throughout the USA. These programs decrease shelter euthanasia rates, decrease shelter intake rates and provide training for veterinary students. ¹ Similarly, many areas offer trap–neuter–return (TNR) programs for feral cats. These programs are often provided free of charge and are funded through donations, making budgetary concerns paramount. In addition, staff time is at a premium and efficiency is important.

One of the drains on staff time in shelter environments is dosing medications. Postoperative pain is of paramount concern for animal welfare, but methods of pain relief with frequent or difficult dosing can be logistically difficult and time consuming. Additionally, for TNR programs, 24 h pain relief may be provided, but the cat is unlikely to be re-dosed. The minimum recommended duration of pain relief after ovariohysterectomy (OHE) for cats is 3 days. ² The 2015 American Animal Hospital Association guidelines for postoperative pain suggest at a minimum that a non-steroidal anti-inflammatory drug (NSAID) be used, though multimodal analgesia is preferred.^3,4

Ideally, an analgesic protocol would provide good sedation prior to anesthesia and at least 72 h of postoperative analgesia. Current protocols typically include drugs with varied mechanisms, including N-methyl-D-aspartate receptor antagonism, opioid agonists, NSAIDs, alpha-2 adrenergic receptor agonists and sodium channel blockers (local anesthetics).^4–12 The majority of the studies investigating analgesia for OHE are truncated to the first 24 h, with multiple protocols providing adequate analgesia.^{4,5,7,8,11,13} However, to extend treatment to 72 h, medications are more limited.

Of the NSAIDs, there are two that are approved by the US Food and Drug Administration (FDA) for cats. Meloxicam is approved as a one-time single-use injection, and robenacoxib is approved for three consecutive days of dosing with both injectable and oral formulations. Meloxicam has been used more frequently, but the FDA now requires a boxed label warning of renal failure and death with repeat administration. ¹⁴ Even though robenacoxib does not have this stipulation, prescribing the medication without preoperative bloodwork may be worrisome in some cats. Both of these medications would require daily dosing.

In general, opioids are good pain relievers, but they are controlled substances, which increases barriers in high-volume, low-cost practices. Fentanyl patches have been used effectively in cats, ¹⁵ but they are expensive, difficult to dose ¹⁶ and variable in absorption. ¹⁷ Additionally, there is the potential for abuse or the possibility of accidental exposure once out of the hospital. Recently, a sustained-release formulation of compounded bu-prenorphine was reported in the literature, which shows promise for up to 72 h, otherwise there is a formulation FDA-labeled for 24 h.^6,18 Butorphanol is an oral opioid agonist–antagonist that has been used for cats, but it requires frequent dosing and provides inadequate analgesia alone. ¹⁹

Local and regional blocks have received more attention in the literature in recent years. ² Advantages include that they are effective, safe (when used correctly) and not controlled. Lidocaine and bupivacaine are inexpensive but relatively short acting. Recently, a bupivacaine liposome injectable suspension (BLIS [Nocita; Aratana Therapeutics]) has become commercially available for use in the veterinary market. It has been FDA approved for 72 h of pain relief as an incisional block for tibial plateau leveling osteotomy in dogs and as a peripheral nerve block for onychectomy in cats. ²⁰

The purpose of this study was to determine if a BLIS incisional block would be non-inferior and cost-effective compared with a control protocol (standard bupivacaine and robenacoxib) for postoperative pain relief in cats after OHE. Our hypothesis was that this method (BLIS) would not be inferior to a standard bupivacaine block and repeated robenacoxib in pain control, and that it would be cost-effective in a high-volume situation.

Materials and methods

Study design

This study was designed as a randomized, double-blind, intent-to-treat (ITT), non-inferiority trial. The number of cats needed for this study was determined using a non-inferiority sample size calculator for continuous variables. The primary variable was the feline Glasgow Composite Measure Pain Scale (GCMPS). ²¹ The estimated number of cats for 90% power and 5% error with a SD of 1 was 18 per group. The non-inferiority limit was set at 1 point on the scale.

Shelter cats were used for this study with client consent and Institutional Animal Care and Use Committee approval (1506-32706A). This study was performed within the live animal veterinary surgery course and was therefore subject to the laboratory timetable. The laboratory consists of junior veterinary students at various levels of surgical skill. The animals for the laboratory are delivered on a Monday. The students perform physical examinations and measure packed cell volume and total protein. Then, half the animals receive surgery on Tuesday and half on Wednesday. The animals are picked up by the shelter on a Friday. This schedule was repeated for a second week in the same order because the student groups switch days and the authors felt that balancing student groups across the BLIS and control cats was important.

Patients were assigned to the student group in the order that they were housed. Student groups were randomized to Tuesday or Wednesday by the college administration (effectively randomizing the cats to the day). The first surgery day (Tuesday) was randomized by coin toss to the BLIS group, which meant that Wednesday’s cats were assigned to the control protocol. This same schedule was maintained on the second week to ensure that students were balanced across the BLIS or control protocol groups. However, this also meant that treatment cats were assessed for pain for 24 h longer than the control cats. Thus, comparisons were only made for 2 days postoperatively.

All cats were assessed for pain by a single blinded investigator using the GCMPS the morning after surgery and every 12 h thereafter until the cats were transferred back to the humane society. This investigator also interacted with the cats the day before surgery for baseline behavior.

Anesthesia and surgery

All cats were premedicated with buprenorphine (0.24 mg/kg SC), dexmedetomidine (0.005 mg/kg IM) and ketamine (5 mg/kg IM). Anesthesia was induced with propofol (4–6 mg/kg IV to effect) and maintained with isoflurane in oxygen. Cefovecin (8 mg/kg SC) was given after induction of anesthesia. Postoperatively, but prior to anesthetic recovery, all cats were given robenacoxib (2 mg/kg SC). Cats randomized to the BLIS group were given saline injections on subsequent days, whereas the control group received repeated doses of robenacoxib daily.

An incisional block was performed by the students in all cats with bupivacaine hydrochloride at 4 mg/kg diluted to 4 ml (control protocol) or BLIS at 5.3 mg/kg diluted to 5 ml with saline (per the manufacturer’s recommendation). A single investigator aware of group allocation and not participating in outcome assessments calculated and diluted the solution. The prepared local anesthetic solution was transferred sterilely to the student surgeon (blinded to treatment allocation) who performed the blocks under direct supervision and guidance of veterinary instructors. After linea alba closure, half of the content of a syringe was injected using a 12 ml syringe with a 1 inch 22 G needle along the incision under the rectus sheath in a continuous line on both sides of the incision. After the subcutaneous layer closure, the remaining content was slowly injected at least 5 mm away from the incision such that a continuous line of anesthetic extended around the entire incision.

Postoperative care and assessment

Following surgery, all cats were assessed for pain and assigned a GCMPS score by the blinded investigator. The veterinary students performed physical examinations on the cats, including a subjective pain assessment integral to the laboratory twice daily. Cats in the control group were given robenacoxib and those in the treatment group were given saline subcutaneously in the morning. This timing places the morning pain scoring at about 18 h after surgery followed by assessments at 30 and 42 h. Those cats in the treatment group had two more pain assessments at 54 and 68 h postoperatively. The same non-blinded investigator distributed medication or saline, but the students performed the injections.

Rescue analgesia was provided if the GCMPS scored ⩾5 by the blinded observer or if the students and a board-certified anesthesiologist (also blinded) agreed that rescue medication was indicated. Cats were not excluded from analysis if rescued or if analgesic administration was delayed. Rescue analgesia was performed using robenacoxib in the BLIS group and buprenorphine in the control group. The non-blinded investigator calculated and prepared medication to prevent overdoses.

Cost analysis

Cost was calculated on a per cat basis for the NSAID and bupivacaine in the control protocol group and the BLIS in the BLIS group. The BLIS formulation must be discarded after 4 h and thus the cost of wasted medication was distributed across the cats. Labor and supplies were not accounted for in the calculation.

Statistical analysis

Means, medians, SDs and confidence intervals (CIs) were calculated for all GCMPS time points. The age of the cats was estimated and therefore age is reported based on a determination of kitten vs adult. A cross-sectional equivalence test was performed for each of the first three time points using JMP Pro 13 (SAS Institute) with statistical significance set at 0.05 and clinical difference set at 0.5 GCMPS points. This is a more conser-vative approach than that used in the sample size calculation where a clinical difference of 1 was used. The results of non-inferiority testing were reported both with the P value and the absolute value of the difference and CIs. An ITT study design preserves the randomization and reduces bias by including all subjects in the analysis, regardless of adverse events or non-compliance. However, an additional analysis excluding the rescued cats was performed because analysis both with and without rescued subjects is recommended for non-inferiority trials. ²² Data are presented as mean ± SD unless otherwise indicated.

Results

There were 47 cats available for OHE in this trial (23 treatment and 24 control). There was no difference in age or weight between the groups. The treatment group had 12 adults and 11 kittens, and the control group had nine adults and 15 kittens. Body weights were 3.0 ± 0.8 kg and 2.8 kg ± 0.8 kg in the treatment and control groups, respectively. Delayed (1–6 h) postoperative administration of medication or saline occurred in three cats in the treatment and four cats in control group because some students misunderstood that they were supposed to administer the treatment regardless of whether they thought the cat was painful. One cat in the control group did not have a 42 h score because of adoption prior to this evaluation time point. The blinded observer recording the GCMPS did not rescue any cats. Each group had one cat where rescue analgesia was initiated by the student and anesthesiologist team. The cat in the control group that was rescued scored 3 on the GCMPS by the blinded observer prior to rescue at the 30 h time point. This cat also had iatrogenic intraoperative complications resulting in a splenectomy and partial pancreatectomy. The cat in the treatment group that was rescued scored 2 on the GCMPS at the 54 h time point. Neither cat required additional rescue.

There were no major complications aside from the cat requiring a splenectomy at the time of the OHE. Anesthesia (treatment: 169 ± 40 mins; control: 165 ± 46 mins) and surgery (treatment: 127 ± 37 mins; control: 129 ± 42 mins) times were not significantly different between the two groups. All cats had some swelling and redness or bruising associated with the incision. No cats had overt incisional infections or dehiscence prior to returning to the shelters. The majority of the cats exhibited signs of upper respiratory infections (sneezing, nasal or ocular discharge) at some point prior to returning to the shelter. They were not treated for these signs.

Results of the GCMPS for each time point are given in Table 1. Overall, the GCMPS scores were low for both groups and did not change if rescued animals were censored. The median pain score at all times was 0. The pain scores of the two groups were equivalent for all of the time points (P <0.05), regardless of whether the rescued animals were excluded or not. The absolute value mean difference in scores were 0.2 (95% CI –0.06 to 0.3) at 18 h, 0.2 (95% CI –0.8 to 0.4) at 30 h and 0.1 (95% CI –0.2 to 0.4) at 42 h. With the rescue cat excluded at 42 h, the absolute value mean difference was 0.1 (95% CI –0.3 to 0.4).

Table 1

Glasgow Composite Measure Pain Scale score (mean ± SD) and weight by group

	Control	Treatment
18 h	0.5 ± 1.0	0.4 ± 0.7
30 h	0.5 ± 1.2	0.3 ± 0.9
42 h	0.2 ± 0.4	0.3 ± 0.7
54 h		0.5 ± 1.1
68 h		0.1 ± 0.3
Weight (kg)	2.8 ± 0.8	3 ± 0.8

At our institution, the cost of the BLIS protocol was US$7 per cat, whereas the control protocol was US$3.3 per cat for injectable robenacoxib (excluding the cost of syringe and needle). If oral robenacoxib were used, the cost would be US$4.20 per cat in the control group. This is a US$2.80–US$3.70 difference in cost per cat between BLIS and bupivacaine hydrochloride and repeated robenacoxib dosing.

Discussion

The pain relief provided by a single BLIS block during closure after OHE was equivalent to block performed with bupivacaine hydrochloride followed by postoperative NSAID up to 42 h after surgery. Although the groups could not be compared at 68 h after surgery, the cats in the BLIS group scored low on the GCMPS up to 68 h after surgery. This may be because the surgical pain was waning or that there was still activity of the BLIS at that time.

One cat from each group required rescue analgesia initiated by the student–anesthesiologist team rather than the GCMPS. To ensure adequate treatment of pain, rescue could also be triggered by the single evaluator or the agreement of the student and a board-certified anesthesiologist who were also blinded to treatment allocation. The two cats rescued in this trial had a higher than average GCMPS score, but it was not high enough to trigger rescue by the single evaluator. There are several possible causes for the difference between the GCMPS trigger for rescue and the student–anesthesiologist team. First, a more intimate knowledge of the cat and extent of surgical trauma could bias the student–anesthesiologist team toward rescue where unnecessary. Secondly, a more intimate knowledge of the cat may have allowed detection and interpretation of more subtle behavioral cues and a more accurate assessment than the GCMPS. Thirdly, the order of evaluation may have altered behavior. The GCMPS score was given to the cats prior to student evaluation. In other words, the students may have contributed to the pain with handling, thus increasing the score.

The cat that required rescue analgesia in the control group had more extensive surgery and the increased invasiveness likely contributed to the breakthrough pain. The cat in the treatment group was rescued at the 54 h time point. There are several possibilities, including more aggressive surgical handling causing increased inflammation, incomplete or inadequate administration of the medication at the time of the block or faster metabolism of the liposomes. Most students completed the block with ease, but there were a few that struggled. The medication must completely encircle the incision at high enough concentrations for effective sensory block. If there is a gap or an under-dosed area, there is a possibility of breakthrough pain. Additionally, an incisional block is not expected to control pain originating from internal organs.

The manufacturer suggests the use of BLIS in three layers in the stifle; however, in this study only two layers were used. The majority of the cats were of lean body condition, leaving only two effective areas for infusion: under the rectus sheath and subcutaneously. This paper describes an off-label use of the BLIS. Other blocks (ie, transversus abdominus plane block) or different distributions of the BLIS may be more efficacious.

Pain studies acknowledge the limitations of the ability to accurately assess feline pain.^{4,5,7–9,11,13,15,16,18,19} We used a common, validated scoring system with a single experienced, blinded evaluator not associated with the laboratory. ²¹ The students and anesthesiologist were also blinded to the treatment allocation. Despite the challenges in accurately assessing pain in cats, the use of a validated pain assessment tool and blinding provide a good level of confidence in our findings. However, it is possible that comorbidities such as upper airway disease common to shelter animals may affect behavioral responses to pain. ²³

Blinding was easily maintained for the anesthesiologist and GCMPS rater. They were not present in the laboratory and there are no physical external clues on the cat between the standard bupivacaine and BLIS. However, the BLIS is cloudy and the bupivacaine is clear giving some indication of a difference in the medications. All of the students had no prior experience or knowledge of either drug and could not assign a label based on visualization. It would be possible with effort for the students to figure out their cat’s assigned group. This is a limitation of the study.

There were no adverse events attributed to the block itself. The infusion of the medication caused a visually detectable presence under the skin, but there were no prolonged incisional oozing or infections. Swelling and bruising were common in both groups, but as all cats had incisional blocks, it was not possible to determine the relative contribution of the blocks themselves vs normal student suturing.

The cost of the liposomal encapsulated formulation was more expensive but not exorbitant on a per cat basis in this study. Cost-effectiveness would be based on the day’s caseload. A 20 ml bottle of BLIS would treat 18–20 cats on average. Therefore, the cost per cat would double if only nine or 10 cats were spayed as the manufacturer recommends discarding unused portions after 4 h. In this study, the discarded portion was accounted for in the cost calculation. Expenses not included in the calculation were missed or delayed doses, accidental loss of medication during administration, time, labor, supplies and injury to students. There are other intangible savings that a long-lasting (days rather than hours) incisional block would provide, including stable pain relief rather than fluctuating pain relief, time savings and flexibility as caregivers are not tied to a dosing schedule, feral or fractious cats where daily dosing is not practical, and decreased stress experienced by the cat in a shelter environment where interaction may be limited to pilling or injecting. These unaccounted costs may be worth the US$2.80–US$3.70 per cat difference in cost in different situations.

The cost difference may also be altered depending on the specific analgesic protocol being compared. Ideally, BLIS could be used as a standalone drug eliminating the cost and time associated with opioid administration. In this trial, all cats were given buprenorphine and robenacoxib perioperatively, which are labeled for pain relief for 24 h. It is possible that the buprenorphine lasts longer than the labeled expectation. ²⁴ If this were true, it is possible that the frequency of robenacoxib dosing was excessive and the BLIS was not needed, adding unnecessary costs. Future studies may address protocols without opioids entirely (ie, NSAIDs and BLIS).

It is also possible that pain decreases quickly postoperatively after OHE and continued pain medications are not needed after the first 24 h. In that case, even placebos would test as equivalent because pain was already at a low level. A study assessing this possibility would strengthen the evidence for use of BLIS. However, a placebo control could not be ethically used in this environment.

This study was performed in a surgery laboratory for third-year veterinary students, thus the methods were constrained to the laboratory timeline and design. Ideally, the cats would have been purely randomized to treatment or control. This would have potentially led to a significant student group bias because student skill varied and it is possible that surgical trauma affects pain.^13,18 Thus, the days were randomized to allow each student group a cat in both the control and treatment group in a random order, balancing the student group influence across the treatment and control groups. However, this limited the time of observation for the control group. Consequently, cats receiving the BLIS were hospitalized 24 h longer than the control group. Ideally, we would have kept the cats in both groups for >72 h, but this was not possible.

In addition, cefovecin was used for antibacterial prophylaxis. The use of a third-generation cephalosporin is not recommended because of contribution to increasing antimicrobial resistance. In this laboratory, cefovecin was chosen because of its actions against Staphylococcus pseudintermedius, and duration of action without the need to redose. ²⁵ Although postoperative use of antibiotics is not generally recommended for short procedures in healthy patients, postoperative surgical site infections are decreased significantly with postoperative antibiotics in student surgical environments. ²⁶ This is likely due to breaks in aseptic technique, poor tissue handling and prolonged procedures that occur with student learning. ²⁶ Other antibiotics may have been used; however, they would need 1–2 times daily dosing that could not be continued after returning to the shelter.

Conclusions

BLIS incisional blocks are equivalent to bupivacaine hydrochloride block followed by daily robenacoxib postoperatively for up to 2 days after surgery. This would reduce the need for continued postoperative drug administration, and may be cost-effective in high-volume, low-cost spay–neuter facilities.