Introduction

Rationale

Cholecystectomy is among the most commonly performed procedures worldwide [1]. The implementation of minimal invasive principles during the previous decades resulted in improved perioperative outcomes, such as postoperative pain, morbidity, quality of recovery, and cosmesis [2, 3]. These benefits, combined with a relatively short learning curve [4], enabled an early and widespread adoption of laparoscopic cholecystectomy.

Even though laparoscopic cholecystectomy is currently the gold standard for the management of most elective and emergency gallstone disease, perioperative analgesia still remains an important clinical issue [5,6,7]. More specifically, a considerable proportion of patients, in some cases up to 50% [8], complain about postoperative pain. Among the reported symptoms are superficial incision pain, abdominal pain, and post-laparoscopy shoulder pain [8, 9]. Suboptimal postoperative pain control after laparoscopic cholecystectomy results to prolonged hospitalization and worsening quality of recovery, thus inhibiting the effective implementation of a day-surgery policy [10, 11].

Therefore, to enhance postoperative pain management and reduce opioid consumption, multimodal analgesia protocols were introduced. These protocols consist of multiple synergetic interventions and medications, such as patient-controlled analgesia, local anesthetics, and non-steroidal anti-inflammatory drugs [10, 12]. Administration of intravenous lidocaine is another strategy whose role has been recently studied in abdominal surgery, with promising results [13,14,15]. Besides the potent analgesic and anti-inflammatory effect of intravenous lidocaine, there are also several other properties of this drug that accelerate postoperative recovery [14].

Current evidence regarding the role of intravenous lidocaine infusion in laparoscopic cholecystectomy is conflicting. Initial reports proposed a significant analgetic and opioid-sparing effect that resulted to reduced morbidity and shortened length of hospital stay [7, 10, 12, 16]. These findings were further validated by two successive meta-analyses [10, 12]. However, these pooled results were based on small size studies and heterogeneous lidocaine dosages. Moreover, recent clinical data questioned the analgesic role of intravenous lidocaine, thus adding to the overall discrepancy [11, 16].

Objectives

We designed and conducted the present study to elucidate the impact of intravenous lidocaine infusion on postoperative opioid consumption and pain intensity after laparoscopic cholecystectomy.

Methods

Design

This study was designed as a prospective randomized controlled trial (RCT) and conducted at the University Hospital of Larissa. A local ethics committee approval was obtained prior to the onset of the trial (UHL2782 /06/28/2018). The protocol record was, also, enlisted in an electronic registry (ClinicalTrials.gov Identifier: NCT03620591). All eligible patients provided a signed informed consent. The present study report adheres to the CONSORT statement guidelines [17].

Participants

All consecutive adult patients (age ≥ 18 years) scheduled for an elective laparoscopic cholecystectomy during the recruitment period, with an American Society of Anesthesiologists (ASA) score I or II, and a body mass index (BMI) ≤ 35, were considered eligible. The following exclusion criteria were included: (1) age < 18 years, (2) ASA physical status ≥ III, (3) patient’s refusal to provide a signed informed consent, (4) allergy in local anesthetics, (5) medical history of liver, renal, or heart failure, (6) chronic pain or (7) systematic analgesics consumption, (8) psychiatric disorders, and (9) inability to perceive and evaluate postoperative pain levels.

Interventions

After the outpatient pre-anesthetic evaluation, all eligible patients were submitted to an elective laparoscopic cholecystectomy. The numeric rate scale (NRS) for pain at rest (NRS rest) and at stress (NRS stress) was explained to patients (0 = no pain, 10 = the worst imaginable pain) during their evaluation. On patients’ arrival in the operating room, non-invasive monitoring (ECG, NIBP, SPO2) was established. Patients were monitoring continuously intraoperatively, and their vital signs (systolic, diastolic, and mean arterial pressure, heart rate) were recorded at a 5-min interval. Prior to anesthesia onset, 1 mg midazolam, 40 mg ondansetron, and 50 mg ranitidine were administered. In the experimental group, patients received a bolus (1.5 mg/kg) lidocaine dose, followed by a continuous lidocaine infusion (2 mg/kg/h), until the procedure end. Control patients received a corresponding placebo infusion (normal saline 0.9%). To avoid intraoperative adverse events, standard safety protocols were utilized. To be more specific, a decrease in the mean arterial pressure > 20% of the base value was managed with the bolus administration of 50 mcg i.v. phenylephrine.

Introduction to anesthesia included the administration of propofol (2 mg/kg), fentanyl (3 μg/kg), and rocuronium bromide (0.6 mg/kg). Anesthesia maintenance was based on a desflurane air/O2 mixture with a 40 to 60 bispectral index (BIS). Neuromuscular blockade was reversed with sugammadex. Extubation was performed when patients could obey simple demands. Intraoperatively, all patients received paracetamol 1 g and dexamethasone 8 mg. Post-anesthesia care unit discharge was based on the Aldrete score [18].

Surgical approach was based on the standard 4-port laparoscopic technique. Following the open Hasson technique, pneumoperitoneum was set at 12 mmHg. Intraoperatively, the patients were placed in a reverse Trendelenburg, combined with a left-sided tilt, position. Prior to trocar placement, local wound infiltration (ropivacaine solution, 0.375%, 20 mL) was performed.

Postoperative care was standardized for all patients and included the administration of crystalloids, thromboprophylaxis (20 mg enoxaparin sodium), and the early mobilization and onset of diet. Standard analgesia consisted of paracetamol every 8 h. In case of increased pain levels (NRS > 3), supplementary opioid analgesia was administered (tramadol 100 mg i.v.). All patients were discharged at 24 h, provided the fulfillment of the Clinical Discharge Criteria [19].

Outcomes

The primary outcome of this study was the difference between the two groups in terms of opioid consumption (total tramadol dose) at 24 h postoperatively. Secondary outcomes included tramadol requirements at various time endpoints 0 (discharge of post care anesthesia unit), 6, 12, and 24 postoperative h. Pain evaluation included NRS measurements at rest and during movement at the same time intervals. Similarly, the incidence of headache and shoulder pain was assessed. Sedation level was appraised by the Ramsay scale [20], while postoperative nausea and vomiting were graded by a 4-point Likert scale (0 = nil, 1 = mild nausea, 2 = moderate nausea, 3 = severe nausea and vomiting).

Randomization and blinding

A specialized electronic software was used for the 1:1 parallel randomization process. The allocation group masking included the use of sealed opaque envelopes. Blinding existed at the level of the patient and the independent third-party investigator, responsible for data recording.

Statistical analysis

Sample size estimation was based on the primary endpoint. In a pilot sample of 10 control laparoscopic cholecystectomies, the mean tramadol dose at 24 h was 150 (57.7) mg. Therefore, assuming a 25% decrease in the opioids consumption in the experimental group, with a 5% alpha, 90% power, and a 5% dropout rate, the total estimated sample size was 98 patients.

A per-protocol analysis was implemented. Prior to any analyses, all data underwent a Shapiro–Wilk normality test. Categorical variables were compared by a Pearson chi square test. An independent Student t-test was used for the comparison of normal continuous outcomes. In non-normal variables, a Mann–Whitney test was introduced. Correlation between continuous variables was assessed by Spearman’s rank-order test. Multiple group comparisons were performed by the Kruskal–Wallis test, while ad hoc analyses were based on the Mann–Whitney test. The Friedman test, followed by the Wilcoxon signed-rank test, was introduced for the evaluation of time-dependent pairwise comparisons of continuous data. Correspondingly, repeated measures of categorical variables were evaluated by the Cochran Q test.

The acceptable percentage of missing data was < 10% [21]. Missing values were managed through the implementation of the multiple imputation method. Continuous and categorical variables were presented as mean (standard deviation (SD)) and N (percentage), respectively. In non-normal distributions, continuous outcomes were provided as median (interquartile range (IQR)). Statistical significance was considered at the level of p < 0.05. All analyses were performed in the IBM SPSS Statistics v23 software.

Results

Due to the COVID-19 pandemic, all elective operations and outpatient departments were halted during the 2020–2021 period. Following a steering committee convention, the study recruitment period was extended until the completion of the estimated sample size. Overall, 126 patients (Fig. 1) were screened for eligibility during the study accrual period (July 2018–January 2022). From these, 19 cases failed to fulfill the inclusion criteria, while 9 patients refused to participate in the study. Therefore, 98 patients were eventually included and randomized to the study subgroups. There was no conversion to open, and no patient was lost to follow-up. Therefore, all cases were submitted to statistical analysis.

Fig. 1
figure 1

Consort study flow diagram

Full size image

Table 1 summarizes the characteristics of the included patients. There was no significant difference in demographics between the two groups. In total, 37 males and 61 females (p = 0.835) were submitted to laparoscopic cholecystectomy. Most patients had mild comorbidities (ASA I: 35.7%, ASA II: 64.3%, p = 0.527). The two groups received an equivalent intraoperative dose of fentanyl (p = 0.107). Similarly, operation duration was comparable between the experimental and the control arm (50.6 vs 50 min, p = 0.319). Interestingly, the lidocaine subgroup was associated with a significantly prolonged extubation period (7 vs 6 min, p = 0.018). Further explanatory analyses (Supplementary Material Tables) confirmed that extubation duration was associated only with the allocation group.

Table 1 Patient characteristics
Full size table

The administration of lidocaine resulted in a statistically significant intraoperative reduction of systolic, diastolic, and mean arterial pressure, after the induction of anesthesia (Fig. 2; Supplementary Material Tables). The mean heart rate, though, was not affected. Similar results were estimated when the set time point was the introduction of pneumoperitoneum (Fig. 3; Supplementary Material Tables).

Fig. 2
figure 2

Mean a systolic pressure, b diastolic pressure, c arterial pressure, and d heart rate after induction of anesthesia

Full size image
Fig. 3
figure 3

Mean a systolic pressure, b diastolic pressure, c arterial pressure, and d heart rate after induction of pneumoperitoneum

Full size image

Regarding the primary outcome (Table 2), the administration of lidocaine did not affect the opioid requirements at 24 h, postoperatively (p = 0.542). Similar results were recorded at any time endpoint evaluated (tramadol dose at 0 h p = 0.391, at 6 h p = 0.767, at 12 h p = 0.296).

Table 2 Postoperative outcomes
Full size table

Postoperative pain at rest measurements (Table 2; Supplementary Material Figures) displayed a declining trend throughout the postoperative period in both groups (p < 0.001). More specifically, median NRS scores decreased from 3 at 0 h to 2 at 24 h. Pairwise comparisons could not identify any significant impact of lidocaine administration on pain scores, at any time endpoint evaluated. In addition to these, pain measurements at movement were, consistently, not affected by the allocation group.

A small number of the entire cohort of patients (14.3%) reported shoulder pain symptoms at 12 h postoperatively. The incidence of these symptoms was further decreased at discharge (24 h: 5.1%). Nonetheless, group allocation was not associated with shoulder pain rates.

Postoperative headache was recorded only in two cases in the lidocaine group (at 12 and at 24 h). No patient developed nausea or vomiting. Sedation was consistent throughout the postoperative period, with no variation, and all patients were co-operative, oriented, and tranquil.

Discussion

Summary of evidence

Lidocaine was first introduced in the early 50 s [22]. It is an amino-amide type local anesthetic, whose main mechanism of action is the blocking of voltage-gated sodium channels (VGSCs) [23]. More specifically, the free-base molecule passes through the lipophilic cellular membrane and binds to the segment 4 of domain 6 of the a-subunit of the ion channel, thus inhibiting the sodium influx and the action potential propagation [23]. The intravenous administration of lidocaine, though, may involve more complex biochemical interactions [24]. Animal studies suggested that in neuropathic pain, lidocaine binds to specific ion channels in the dorsal root ganglia and inhibits the activation of the respective nerves [24]. However, in patients with normal pain thresholds, the exact pathway of lidocaine is vaguer [22,23,24,25,26]. Among the cellular targets that have been examined are membrane receptors, modulation of K + and Ca2 + channels, N-methyl-d-aspartate (NMDA) receptors, α-amino-3-hydroxy-5-methyl-4-izoxazolepropionicacid (AMPA) receptors, and GTP-binding protein coupling receptors [22,23,24,25,26]. In a recent in vivo study, Kurabe et al. [24] proposed a glutamate-inhibition effect of intravenous lidocaine at the presynaptic terminals that results to a membrane potential shift and a post-synaptic hyperpolarization [24].

Initially, the role of lidocaine as part of a multimodal approach was examined in major operations, where analgesia requirements are increased [27]. In the study by Koppert et al. [27], patients who underwent major abdominal surgery received IV lidocaine until the 1st postoperative hour. The experimental group displayed significantly lower pain scores and morphine consumption during the first 72 h [27]. Subsequent pooled analyses further validated these findings, thus suggesting a potent analgesic effect of lidocaine [28, 29]. However, a recent Cochrane meta-analysis pointed out to a limited and early-phase benefit that does not extend to the subsequent postoperative days [15]. This discrepancy may be explained by the use different pain assessment tools, a wide range of lidocaine dosing regimens, and the extent of the operation [15]. Regarding the latter, the implementation of a minimal invasive approach decreases tissue injury, thus reducing the clinical impact of lidocaine [30, 31]. As shown by our research group, although lidocaine abates pain scores and opioid consumption in open colorectal resections [30], this is not the case in the laparoscopic setting [31].

Laparoscopic cholecystectomy is performed through a standard four-port technique. Despite its benefits, the increased incidence of postoperative pain impedes the implementation of an efficient recovery protocol. Post-laparoscopic cholecystectomy pain includes three distinct subtypes [9]. First, tissue trauma in the port insertion sites results to somatic pain [9]. Moreover, intraoperative tissue handling combined with pneumoperitoneum-related diaphragmatic irritation is responsible for the development of visceral pain. Finally, abdominal distention and CO2-induced peritoneal inflammation that leads to arteriole tearing and nerve fiber injury promote phrenic nerve neuropraxia and, subsequently, the emergence of shoulder pain [32]. Visceral pain is the main form during the first postoperative day; on the contrary, shoulder-tip pain symptoms may exist up to 72 h, postoperatively [9].

Numerous studies attempted to quantify the clinical benefit deriving from the intravenous administration of lidocaine during laparoscopic cholecystectomy. In a RCT by Song et al. [7], intraoperative lidocaine infusion, when compared to placebo, decreased mean visual analog scores at 2 and 6 h after surgery. Two recent meta-analyses by Zhao et al. [10] and Li et al. [12] indicated a significant analgesic effect of lidocaine that extended up to the second postoperative day. In parallel with the subjective pain measurements, the respective opioid consumption was considerably reduced over the control subgroup [10, 12]. On the contrary, Ortiz et al. [16] reported that lidocaine infusion did not alter postoperative pain scores at rest and coughing and total morphine consumption. Our results are similar; NRS scores at rest and movement were comparable at all endpoints. Additionally, contrary to our initial study assumption, overall tramadol consumption did not decrease in the experimental group.

An important factor that should be commented on is the potential effect of local wound infiltrations on the overall pain outcomes. In our study, both subgroups received ropivacaine infiltrations in the port incisions, thus limiting any bias. Theoretically, local infiltrations may affect somatic pain, but not visceral or shoulder-tip pain. Although local anesthetic infiltration may act as a confounder in early pain measurements, it is a common practice in multiple large-scale RCTs in the field of minimal invasive surgery [33, 34]. Moreover, current evidence regarding the exact analgesic effect of local infiltrations is still inconclusive [35,36,37].

Besides analgesia, lidocaine displays multiple other properties, including anti-hyperalgesia, anti-allodynia, and anti-inflammatory effects [23]. Regarding the latter, it was shown that lidocaine inhibits the priming of human polymorphonuclear cells or neutrophils and inhibits the release of inflammatory cytokines and chemokines [23]. More specifically, Sridhar et al. [38] validated a significant reduction in postoperative total leukocyte count and circulating CRP and interleukin (IL)-6 levels after IV lidocaine infusion. Similarly, Ortiz et al. [16] showed that lidocaine reduces serum levels of IL-1, IL-6, interferon-γ, and tumor necrosis factor-a during laparoscopic cholecystectomy. These anti-inflammatory effects did not have an effect, though, on bowel function recovery or length of hospital stay [16]. On the contrary, Song et al. [7] reported a shorter time to first flatus and time to first bowel movement when lidocaine was administered. In our protocol, we did not look specifically at those parameters as all patients received per os diet and were discharged during the first 24 h on a day case basis.

Opioids are commonly used for perioperative pain management [39]. Intraoperatively, opioids minimize pain symptoms and inhibit the sympathetic response, thus facilitating anesthesia maintenance [39]. During the postoperative period, opioids modulate pain signal transmission and the response to painful stimuli, thus enabling efficient pain control [39]. However, opioids are associated with multiple side effects, such as nausea, vomiting, and respiratory depression [39]. Besides these, prolonged hospitalization and the development of addiction impact postoperative recovery, especially under the light of the recent opioid over-prescription crisis [39]. Thus, multimodal analgesia has been proposed as a mean to withhold opioid administration [39, 40]. Several studies confirmed the opioid-sparing effect of intravenous lidocaine in laparoscopic cholecystectomy [10] and the respective improvement on the quality of postoperative recovery [11]. As mentioned earlier, we estimated similar tramadol requirements in the lidocaine and the control group. Respectively, there was no difference in the nausea and vomiting rates at all estimated time endpoints.

The stabilization of the intraoperative hemodynamic variables is a significant parameter that affects the perioperative mortality and morbidity [41]. More specifically, intubation and introduction of pneumoperitoneum result to notable cardiovascular fluctuations that derive from the activation of the sympathetic nervous system [41]. Therefore, the optimal drug should inhibit the excretion of catecholamines, without affecting the cerebral blood flow and the efficacy of the anesthetic regimen [41]. Lidocaine has exhibited these properties in multiple clinical scenarios, where a significant decrease in mean arterial pressure and cardiac index was found [41,42,43]. However, these were not confirmed in the study by Lu et al. [11]. Our RCT validated a significant reduction of systolic, diastolic, and mean arterial pressure measurements in the lidocaine group, regardless of the time set point. No effect in the mean heart rate was found, though.

Despite these, the development of intraoperative hypotension may result to organ ischemia and, therefore, suboptimal postoperative outcomes [44]. More specifically, in a large multicenter cohort of 368,222, non-cardiac procedures showed that the risk for a major cardiac or cerebrovascular adverse event in the 30-day postoperative follow-up period was 12% with mean arterial pressure ≤ 75 mmHg [44]. In cases of sever hypotension (≤ 55 mmHg), that risk increased to 26% [44]. In our study, a safety protocol was applied, where a reduction of over 20% of the base mean arterial pressure was managed with the administration of bolus phenylephrine. In addition to these, mean MAP values were systematically retained over the 75 mmHg threshold in both study subgroups.

In our cohort, a significant delay in the time to extubation (7 versus 6 min) was noted in the experimental subgroup. This finding is in parallel with the current literature. In a recent RCT by Dogan et al. [45], it was shown that compared to esmolol, lidocaine infusion was superior in terms of tracheal responses with a longer time to extubation, though. Similar results were provided by the meta-analysis by Yang et al. [46]; time to extubation was significantly longer when lidocaine was administered compared to placebo. The authors acknowledged the role of volatile agents and confirmed the effect of halothane, but not that of sevoflurane and isoflurane [46]. We applied a desflurane-based anesthesia maintenance which is a favorable agent, with minimal extubation periods [47, 48]. Moreover, to reduce the confounding of the administered agent to the clinical outcomes, we targeted the depth of anesthesia to pre-defined BIS measurements.

Patient safety is another major issue that should be addressed when evaluating a novel analgesic strategy. Intravenous lidocaine is associated with a very narrow therapeutic range, with early toxicity. Previous reports estimated that serious adverse events occur beyond the level of 9–10 μg/mL [26, 49]. Although initial toxicity may present with neuropsychiatric symptoms, such as light headedness, tinnitus, altered speech, and muscle twitching, higher plasma levels result to loss of consciousness, myocardial depression, and, ultimately, cardiac and respiratory arrest [26]. It must be noted, though, that factors that affect bioavailability and clearance (e.g., distribution volume, cardiac index, hepatic and renal impairment, acid–base status, hypoalbuminemia, and medications) alter the correlation between dosage and toxicity [26]. Large cohort analyses reported a minimal minor adverse event rate (6.8%), thus confirming the safety of lidocaine [26]. Moreover, a recent pooled analysis of 2802 cases [50] concluded that the administration of lidocaine did not result to an increased risk of agent-related complications. Although cardiac adverse events were reported, most of them were sporadic and unrelated to lidocaine [50]. Similarly, we did not record any mild or severe adverse event in the experimental arm.

Limitations

Prior to the appraisal of our results, several limitations should be acknowledged. First, despite the introduction of an adequate randomization and blinding algorithm, the inherent heterogeneity in patient characteristics may have affected the final outcomes of our trial. Moreover, a larger patient pool may have affected comparative results, although the estimated sample size was based on previous observations and adequate power. More specifically, failure to confirm data normality prohibited the performance of further explanatory and correlation analyses. Furthermore, despite the use of validated scales [20], the subjective nature of several, patient-assessed, secondary outcomes represented an additional source of bias. Additionally, based on the study protocol, evaluation of inflammatory markers, hospitalization duration, and bowel function recovery was not performed, thus narrowing the measured effect of lidocaine in the postoperative period. Finally, the intraoperative infiltration of local anesthetic may have had a confounding effect on the postoperative pain measurements.

Conclusions

Our study failed to confirm the presence of any benefit of lidocaine, in terms of opioid consumption, during the postoperative period after laparoscopic cholecystectomy. Additionally, lidocaine administration did not affect postoperative pain scores or the incidence of shoulder pain, at any time endpoint evaluated. Moreover, we did not identify any difference in terms of postoperative sedation levels and nausea rates. Due to the abovementioned study limitations, further RCTs with larger sample size are required to validate our results.