Article Text

Download PDFPDF

Analgesic efficacy of liposomal bupivacaine for surgical site infiltration: a single-outcome meta-analysis (the best we could do)
  1. Faraj W Abdallah1,2,
  2. Nasir Hussain3 and
  3. Richard Brull2,4
  1. 1 Anesthesiology and Pain Medicine, University of Ottawa, Ottawa, Ontario, Canada
  2. 2 Anesthesiology and Pain Medicine, University of Toronto, Toronto, Ontario, Canada
  3. 3 Department of Anesthesiology, Ohio State University Wexner Medical Center, Columbus, Ohio, USA
  4. 4 Women's College Research Institute, Univeristy of Toronto, Toronto, Ontario, Canada
  1. Correspondence to Dr Faraj W Abdallah, Anesthesiology and Pain Medicine, University of Ottawa, Ottawa, Ontario, Canada; mank_abda{at}

Statistics from

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

We recently presented the findings of our meta-analysis investigating the analgesic effects of peri-articular local anesthetic infiltration with liposomal bupivacaine during the first couple of days after total knee arthroplasty,1 which is one of three quantitative investigations we had undertaken to evaluate some of the most interesting and immediate clinical applications of liposomal bupivacaine, namely perineural, peri-articular and direct surgical site infiltration. Having completed the first two meta-analyses,1 2 we began our data acquisition for the third and final application—direct surgical site infiltration—which we fully expected would be the most challenging of the three, largely because surgical site infiltration was the original and first application to receive Food and Drug Administration (FDA) approval, and the relevant evidence involved a number of early industry-sponsored trials. It was not long before our expectations were realized.

While we were able to identify the titles of potentially relevant trials from the FDA records with relative ease, we were unable to retrieve important clinical outcome data for most.3–8 Many of the documents made available by the FDA contained data and results that were either blacked out, summarized in textual format precluding data extraction, or limited to comparison with placebo (negative control). On contacting the FDA in order to obtain the missing outcome data, we faced additional challenges relating to requisite fees, undue lengthiness of the process (1–2 years), and the requirement for permission to release these data from the industry sponsors. Fortunately, we were able to locate some related unpublished pain severity data on the ClinicalTrials registry; yet, these datasets were largely incomplete for other relevant outcomes, such as opioid consumption and functional recovery. With such limited data available, our intended comprehensive quantitative review was brought to an abrupt halt. The latter notwithstanding, we were able to synthesize data for a single but arguably most important clinical outcome—the area under the curve (AUC) of pain over 0–72 hours postoperatively—for which we had the most complete data from both industry and non-industry sponsored trials. We believe our sole finding to be worthy of dissemination.

Briefly, for each relevant trial, we calculated the mean difference (95% CI) in AUC of pain over the 0–72-hour time interval if data were extracted for at least four time points between 0 hour and 72 hours, including at least one point assessing short term pain (≤12 hours) and another assessing pain at 72 hours. We then conducted a meta-analysis over time that sequentially pooled data by year of publication, using the inverse variance method and random-effects modeling. This approach allows evaluating the change in treatment effect as additional data become available. Of the eligible 15 trials3–6 8–18 examining surgical field infiltration, three12 14 15 were excluded because they did not allow constructing an AUC and one8 was excluded because AUC over 0–96 hours was evaluated. The remaining 11 trials3–6 9–11 13 16–18 included 43–6 that were industry-sponsored and 79–11 13 16–18 that were not. Each of the four industry-sponsored trials had unpublished3–6 data that were retrieved from the ClinicalTrials registry for three3–5 and from a prior review19 for one.6 The types of surgeries examined included orthopedic,9 10 17 urologic,3 13 18 breast,4 16 general,5 6 and gynecologic.11 The trials involved a total of 712 patients (liposomal bupivacaine: 421, plain bupivacaine: 291). Chronologically, only the first two trials3 6 (industry sponsored) depicted a treatment effect (strong) favoring liposomal bupivacaine, with a mean difference (95% CI) of 109.0 cm.hour (67.8 cm.hour, 150.2 cm.hour) (figure 1). In contrast, all subsequent nine trials,4 5 9–11 13 16–18 including two that were industry sponsored,4 5 were negative and found no difference between liposomal and plain bupivacaine. Meta-analysis over time by pooling all 11 trials brought the treatment effect to null, with a mean difference (95% CI) of 28.00 cm.hour (−1.40 cm.hour, 57.40 cm.hour) (figure 1), suggesting no difference in postoperative pain over the 0–72-hour period between liposomal and plain bupivacaine for surgical field infiltration.

Figure 1

Meta-analysis over time forest plot of the mean difference in area under the curve (AUC) pain scores over 0–72 hours for plain local anesthetic vs liposomal bupivacaine. Pooled estimates of the mean difference are shown with SE and 95% CIs. Each new study listed depicts a pooled estimate, which includes AUC data from the studies published prior to it. Pooled estimates are represented by vertical red lines and the horizontal black lines represent the 95% CI. CI, confidence interval; IV, inverse variance; SE, standard error.

Despite the challenges we faced, we are able to definitively demonstrate that pain relief, as measured by AUC over the 0–72-hour postoperative period, is not superior for liposomal bupivacaine compared with plain bupivacaine when used for surgical site infiltration. Our results were not altered when the trial reporting AUC over 0–96 hours8 was included post-hoc. These results are consistent with our findings for the peri-articular1 and perineural2 administration routes. Taken together and based on the available evidence, our findings do not support the routine use of liposomal bupivacaine over traditional long-acting local anesthetics for these three important clinical applications, particularly when the cost implications are considered.

Additional lessons were also learned in the course of acquiring and analyzing relevant data. The first relates to the importance of reporting clinical trial data and making them available and transparent to researchers without undue impediments. The second emphasizes the importance of confirmatory trials to overcome the strong treatment effect reported by initial studies. While only the first two industry-sponsored studies indicated benefit, it took data from nine subsequent trials to bring the treatment effect back to null. This peculiar discrepancy may be random, or could even represent optimism bias, but we cannot ignore that it may signal industry influence, which has been shown to be associated with positive study outcomes.20 The third underscores the limitations of the current FDA requirement to demonstrate superiority compared with placebo, which does little to help practitioners interested in understanding any real clinical advantages over plain bupivacaine. Scientific comparisons against placebo are limited to informing safety and efficacy under optimal experimental conditions. In contrast, scientific comparisons to “usual care” inform effectiveness and can determine whether or not the intervention “works under real life conditions”.21 Consistency in both efficacy and effectiveness is essential to drive change in clinical practice. While liposomal bupivacaine is demonstrably efficacious, it, unfortunately, cannot be considered effective.



  • Twitter @Faraj_RegAnesth

  • Correction notice This article has been corrected since it published Online First. The author affiliations have been updated.

  • Contributors All authors equally contributed to the manuscript.

  • Funding The authors have not declared a specific grant for this research from any funding agency in the public, commercial or not-for-profit sectors.

  • Competing interests None declared.

  • Patient consent for publication Not required.

  • Provenance and peer review Not commissioned; externally peer reviewed.

Linked Articles