This retrospective observational study aimed to investigate the efficacy of lipid emulsion pretreatment in reducing the incidence of LAST after LIA for fast-track THA and TKA. Our study introduces a novel concept and tests its efficacy in an Asian population with a calculated sample size, providing valuable evidence for clinical practice. Our data revealed that lipid emulsion pretreatment reduced the incidence of arrhythmia in patients who underwent THA but increased the requirement for rescue opioid analgesia compared to those without pretreatment. Lipid pretreatment reduced the incidence of LAST, particularly cardiovascular events, possibly due to the lipid shuttle effect from heart tissue or the direct cardiotonic effect of the LCT/MCT emulsion [11]. Conversely, lipid pretreatment did not significantly reduce the occurrence of LAST in TKA patients and was associated with an increased postoperative rescue opioid requirement. As a result, lipid pretreatment might reduce the analgesic efficacy of LIA and hinder enhanced recovery by increasing opioid-related adverse effects. The efficacy of lipid pretreatment in alleviating LAST remains controversial, and its routine clinical use for LIA is not recommended unless its benefits are confirmed by further prospective studies.

Fast-track or ERAS pathways are multi-modal and multi-professional approaches designed to mitigate surgical and anesthesia risks and improve the quality of medical care, rather than focusing solely on expediting discharge [18, 19]. LIA provides effective opioid-free analgesia when combined with multi-modal analgesia after joint arthroplasty and is recommended by the ERAS Society, particularly for TKA [4, 20]. However, the clinical evidence supporting the use of LIA for THA is limited. A recent meta-analysis found that LIA improved pain relief and decreased opioid consumption in patients undergoing THA compared to controls [21]. Judging from our analysis of LIA in fast-track TKA and THA, motor-sparing LIA has significantly reduced postoperative pain and facilitated the swift return of patients to their daily lives after joint replacement surgery. Although most LAST events in our patients resolved within two hours with conservative treatment in the PACU, some patients experienced self-limited symptoms such as dizziness, nausea, or vomiting the day after surgery [5]. Therefore, it is important to acknowledge the clinical scenario and explore strategies to reduce the incidence and severity of LIA-induced LAST, such as adding dexmedetomidine, dexamethasone, or vasoconstrictor to the LIA mixture, reducing LA dose for LIA, prolonged lipid infusion following the initial bolus rescue of lipid infusion, or considering novel measure like lipid pretreatment [22].

Our result found that the incidence of severe LAST events requiring rescue lipid infusion after lipid pretreatment slightly decreased from 2.54 to 2.28 per 1000, which is close to the rate reported in a previous cohort (2.0 per 1000 after LIA for THA) [7]. However, previous small-scale population pharmacokinetic analyses of LIA for THA and TKA did not report LAST events, even with LIA doses up to 400 mg ropivacaine, and their reported free plasma ropivacaine concentrations were all below the pre-defined toxic threshold (0.600 mcg/ml) [23]. Affas et al. evaluated 15 patients undergoing LIA with 200 mg ropivacaine for primary THA. The 95% upper prediction value of maximal unbound plasma concentration of ropivacaine was 0.032 mcg/ml within 30 h after LIA [24]. Gromov et al. investigated 28 patients who received 400 mg ropivacaine for LIA in unilateral TKA, and reported peak free-form ropivacaine concentrations of 0.030 mcg/ml in unilateral TKA and 0.095 mcg/ml in bilateral TKA [25]. Other small-population pharmacokinetic studies of LIA using ropivacaine for TKA yielded similar results, with free-form ropivacaine levels all below the threshold for CNS toxicity [26,27,28]. However, the sample sizes in these studies were relatively small, and they did not report the incidence of clinical LAST events. Our analysis involved 1,621 Asian individuals with naturally lower α1-acid glycoprotein levels, yet there is limited literature focusing on this population [29]. This retrospective cohort reported various presentations of clinical LAST events and found that lipid pretreatment reduced the incidence of LAST events, particularly in cases of bradycardia and new-onset arrhythmia. Pharmacokinetic analysis of plasma bupivacaine concentrations after LIA and lipid pretreatment to correlate with our observed LAST events is required to explain our observation [30].

Our study presents a novel concept of lipid pretreatment after LIA to reduce LAST occurrence, which has not been previously reported in the literature. The proposed multi-modal mechanisms of lipid emulsion as a resuscitation therapy for LAST include scavenging (lipid shuttle and organ redistribution) and non-scavenging effects (such as cardiotonic, postconditioning, and vasoconstrictive effects) [8, 9, 31, 32]. Existing evidence suggests that lipid emulsions do not significantly decrease LA concentrations in target organs [8]. A clinical pharmacokinetic study in 16 volunteers found that administering intralipids 2 min after intravenous infusion of ropivacaine or levobupivacaine reduced peak plasma concentration by only 26–30% and did not prevent the development of CNS toxicity [33]. However, the mechanism of preemptive administration of lipid emulsions before the onset of LAST is rarely discussed. In rat models, lipid pretreatment not only reduced the incidence of bupivacaine-induced cardiotoxicity [34], but also ameliorated convulsions by inhibiting the increase in blood-brain barrier permeability and enhancing GABA-mediated currents [35]. In a clinical pharmacokinetic study by Chen et al., intralipid pretreatment (1.5 ml/kg) before femoral and sciatic nerve blocks increased the volume of distribution and significantly reduced both free and total levobupivacaine concentrations [36]. Since the half-life of the lipid emulsion (21.9 ± 8.2 min) is substantially shorter than that of bupivacaine (4.6 ± 2.6 h) and ropivacaine (2.3 ± 0.8 h) [10, 16], our lipid pretreatment may exert only temporary cardiotonic effects, thus contributing to fewer CV events. In the lipid sink model, simulated lipid emulsion infusion requires much longer time to reduce bupivacaine concentration in brain tissues (15 min) than in cardiac tissues (3 min) [9]. This phenomenon might partially explain why we observed fewer CV events but no obvious change in CNS and respiratory events following lipid pretreatment in our results. Regarding the higher incidence of tremor, agitation, and respiratory events in the THA group receiving lipid pretreatment, a reduced effect of LA compared to those without lipid pretreatment might explain this shift from catastrophic LAST presentations like seizures and cardiac arrhythmias to relatively milder events. The actual mechanisms behind the different CV and CNS presentations of LAST after lipid pretreatment in our study required further investigation through pharmacokinetic studies.

In our results, we found no difference in LAST incidence between TKA patients with or without lipid pretreatment, but those receiving lipid pretreatment required more rescue opioid analgesia. Since lipid pretreatment may reduce peak plasma bupivacaine concentration by 30% [33], it could further decrease LIA’s analgesic efficacy in TKA. The observed differences in rescue opioid requirements in THA or TKA patients after lipid pretreatment might be partially explained by the variable analgesic efficacy of LIA for THA or TKA. However, due to the limited number of TKA patients, these findings should be interpreted with caution.

To summarize the pros and cons of lipid pretreatment, our results demonstrated a minimal reduction in severe LAST events, a lower incidence of cardiovascular events, but no clear benefit for neurologic or respiratory outcomes. Additionally, patients receiving lipid pretreatment required more rescue opioid analgesia and had higher pain scores in the first postoperative hour, after correction for multiplicity. Although no adverse events were reported in our study, the safety of lipid pretreatment in routine practice still requires further investigation [37]. Given the small sample size and retrospective nature of this study, the efficacy of lipid pretreatment for managing LIA-induced LAST cannot be justified. Overall, the efficacy of lipid pretreatment remains controversial, as it slightly reduced cardiovascular LAST events but might hinder postoperative recovery by increasing rescue opioid requirements. This decreased analgesic efficacy of LIA for THA and TKA by lipid pretreatment could exacerbate patient recovery, particularly in the context of modern multimodal opioid-minimizing analgesia. The slight reduction in cardiovascular LAST events does not offset the increased risk of opioid-related side effects, especially in patients where opioid-sparing techniques are essential for enhanced recovery. The increased postoperative pain after lipid pretreatment might ultimately undermine the initial benefits of fast-track protocols for THA and TKA. Therefore, lipid emulsion pretreatment should not be advocated until more robust prospective studies can demonstrate clear benefits without increasing other risks.

To the best of our knowledge, this study is the first to evaluate the efficacy of lipid pretreatment in reducing the incidences of LAST after LIA for THA and TKA. We utilized Benjamini–Hochberg correction to account for multiple testing and reduce the false discovery rate. Our retrospective analysis serves as a reminder that LAST is an ongoing concern that cannot be overlooked when using LA for LIA or nerve blocks, especially in Asian populations, who have lower levels of α1-acid glycoprotein compared with Western populations [28]. Adequate preparations for managing LAST should always be in place.

The present study has several limitations. First, as an observational cohort study, inherent demographic differences existed between the THA and TKA groups, as well as other comparisons. Despite using propensity score matching and correcting for multiple testing, these differences could not be completely eliminated. Second, the study participants underwent THA or TKA between 2020 and 2022, and were grouped based on lipid pretreatment status. When comparing the groups with or without lipid pretreatment, the distribution of patients was not randomly selected, which might have introduced temporal bias. However, all patients were treated by the same surgeons, anesthesiologists, and medical teams, following standardized protocols, which likely minimized this bias. Third, non-recorded LAST events, such as subclinical tremors or agitation, QT prolongation, or short-term arrhythmia, were not captured in this cohort, potentially leading to underestimation. Nevertheless, since our results were based on group comparisons, similar underestimations would likely have occurred in both groups. Fourth, plasma levels of albumin, α1-acid glycoprotein, bupivacaine, and lipid emulsion were not measured in this study. Patients with low plasma levels of albumin or α1-acid glycoprotein are more susceptible to LAST [13]. Without measuring plasma bupivacaine levels, it is difficult to conclusively attribute these events to LAST, leaving room for alternative explanations. Given these limitations, along with the retrospective design and small sample size of this study, the true efficacy and safety of lipid pretreatment for LIA-induced LAST remain unjustified. Further carefully designed prospective studies with larger patient populations and pharmacokinetic monitoring are required to verify the causal relationship between plasma bupivacaine concentrations following LIA and LAST events, and to establish the true efficacy and safety of lipid pretreatment.