Comparative study lumbar plexus block and lumbar erector spinae plane block for postoperative pain relief after proximal femoral nail for proximal femoral fractures

Abstract

Background: 

The clinical outcomes (time to ambulation, length of stay, and home discharge) after proximal femoral nail (PFN) for proximal femoral fractures (PFF) is dependent on successful pain management. Currently, the lumbar erector spinae plane block (LESPB) is in vogue and is associated with favorable outcomes in the postoperative period. Our study aimed to evaluate whether a LESPB provided equivalent analgesia and clinical outcomes as compared to LPB in PFN for PFF.

Material and Methods: 

We compared LPBs [L] with LESPBs [E], with 30 patients in each group, performed from June 2020 to June 2021 for PFN in PFF's. The primary outcome of this study was the average NRS pain scores over 24 hours postoperatively. Secondary outcomes included pain scores at different time points over 24 hours, opioid consumption between the groups at 24 hours postoperatively, time for request of first parenteral analgesia, quadriceps weakness and adverse events.

Results: 

The average pain scores over 24 hours were better in the LESPB group as compared to the LPB group (p = 0.02). Further, only n = 5 (30%) of patients in the LESPB group required opioids, while n = 13 (43.333%) of patients in the LPB group required opioids. Moreover, the median time for request of first parenteral analgesia was 615 (480–975) minutes, weakness of quadriceps function occurred in 2 patients in the L group, which recovered at 3^rd and 5^th month, respectively, with no incidences of hemodynamic instability and respiratory complications.

Conclusions: 

This trial demonstrated that single bolus LESPB is superior to LPB in terms of analgesic outcomes, has low adverse events, and is an agreeable substitute for patients with PFF undergoing a PFN.

Introduction

Elderly patients above the age of 65 years with proximal femoral fractures (PFF) and associated comorbid conditions present distinct challenges for perioperative pain management. Debate on timing of surgery continues but is agreed upon that early surgery between 48 hours and 4 days in patients is beneficial.[¹²] The clinical outcomes can be reduced with single shot nerve blocks when non-steroidal anti-inflammatory drugs (NSAID's) and opioids are ineffectual. Moreover, pain is reduced in 30 minute after implementation of block.[³]

Opioid sparing is observed with single shot and continuous femoral nerve blocks in the perioperative period after hip surgeries.[⁴] However, the previous claims of decreased incidence of delirium after nerve blocks is refuted.[⁵] Postoperative reduction in opioid consumption, decreased adverse effects of opioids, increased patient satisfaction and a stable hemodynamic are the hallmark of an appropriately performed LPB.[⁶⁷⁸] However, the neural target is deep and closer to neurovascular structures. This not only makes the block challenging bit at the same time is contraindicated in patients on anticoagulants.[⁹] Moreover, the ultrasound lumbar paravertebral in elderly has been recently demonstrated to be distorted and disorganized, and unlike in adults the needle – nerve contact to establish an optimal block is poor.[¹⁰]

After introduction of a LESPB, case reports, and observational studies illustrated effective pain control in the postoperative period after total hip replacements.[^1112131415] The diffusion of solution from the lumbar erector spinae plane to lumbar plexus area suggests it to be a lumbar plexus by proxy.[¹⁶]

Commensurate with clinical benefits and proposed diffusion, we hypothesized that the LESPB would be equally effective as LPB in postoperative pain management and prospectively assessed the opioid consumption and pain scores.

Material and Methods

This single-center, randomized, patient- and assessor-blinded, parallel-group, clinical trial was approved by the Research Ethics. The study was registered on Clinical Trials Registry of India (CTRI) and patients are randomized between June 2020 and June 2021. We adhered to the guidelines of the Consolidated Standards of Reporting Trials (CONSORT).[¹⁷]

Study participants

We enrolled elderly patients (age greater than 65 years) with American Society of Anesthesiologists’– physical status (ASA-PS) classification I to III and body mass index not exceeding 25 kg/m² or less scheduled to undergo unilateral proximal femoral nail (PFN) surgery under spinal anesthesia. All study subjects provided written informed consent before participating in this trial. Patients for a revision surgery, affected with lumbar spinal canal stenosis, contraindications for plexus blocks (infection, bleeding diathesis, allergy to local anesthetics), Parkinsonism, contraindication to any component of multimodal analgesia and history of significant psychiatric conditions that may affect patient assessment qualified for exclusion. All patients were identified, informed, and interviewed bed-side upon admission in ward or room. A baseline assessment and informed consent was obtained by study coordinator.

Randomization and blinding

The randomization list was computer generated and independent of any other study procedures. Undisclosed to the patient and coordinator, a single investigator logged on the day of procedure to retrieve treatment allocation and performed the designated blocks. A manual case report form was used to collect and disperse information. Block performance and assessment were not evaluated by the coordinator and the single investigator, but by second anesthesiologist.

Intraoperative management

Spinal anesthesia was administered under a standard monitoring (non-invasive blood pressure, electrocardiogram, and pulse oximetry) at the level of L2/3 or L3/4 interspace with a 27 G Whitacre spinal needle with introducer (Becton Dickson, India Pvt Ltd) in sitting position. After egression of clear CSF 2–2.5 ml of 0.5% bupivacaine was injected over 15 seconds and the patient was made to lie supine for 10 minutes. Ringer lactate 500 ml was infused prior to positioning the patient in lateral decubitus with operated side non-dependent. PFN in all patients were performed in lateral position on non-fracture table. At surgical wound closure, all patients received 30 mg/kg intravenous paracetamol and every 8 hours, thereafter. Immediately after wound dressing, the blocks were accomplished in the same position corresponding with surgery.

Block technique

All blocks were performed in the lateral decubitus, with operative side non-dependent, under aseptic precautions. A 3 ml 1% lidocaine with adrenaline was injected as a test dose prior block injectate consisting of 0.2% ropivacaine and 30 mcg clonidine. The LPB was implemented using the Shamrock technique,[¹⁸] while the method described by Aksu was administered for LESPB.[¹⁹] A 100 mm stimulating needle (Pajunk, Germany) was used in both groups, while in the former neurostimulation unit (HNS B-Braun, Germany) was employed, in the later it was not.

Lumbar plexus group (Group L)

In the lateral decubitus position, after recognizing the L5-S1 interspinous space in the longitudinal view, the curvilinear probe [5-2 mHz M-Turbo, Fuji Sonosite] was shifted cephalad until the L3 spinous process, the probe was rotated at the transverse process of L3 [axial plane], and shifted on the abdominal trunk in the mid-axillary line. The insonation depicted the paravertebral sonoanatomy [Figure 1a]. With paravertebral sonoanatomy in view, under aseptic precaution, needle tip insertion and the pathway were recognized perpendicular to the US beam targeting the lumbar plexus in the posterior and medial quadrant.[²⁰] A 100 mm insulated, stimulating [Pajunk, Germany] was introduced at the above-mentioned point, and the tip was identified adjacent to the lumbar nerve root, L3 [Figure 1b]. Upon neurostimulation initiating at 0.4 mA quadriceps contractions was obtained as an evoked motor response (EMR). The current ranged from 0.4 mA to 0.8 mA and not less than 0.3 mA. A volume of 0.2 ml/kg body weight of LA was injected (test dose 3 ml + remaining volume).

Figure 1: (a) Sonoanatomy of lumbar paravertebral area (b) Needle tip in approximation with lumbar nerve root (c) LA injection [light blue zone] after confirmation with neurostimulation (d) Entire volume injection of LA, with Lnr [lumbar nerve root] engulfed by the LA

Lumbar erector spinae group (Group E)

The ultrasound scan for the LESPB was performed with the Aksu approach with a slight dorsal tilt to gain maximum visualization of the erector spinae dorsal to the L3 transverse process in the axial plane [Figure 2a]. After aseptic precaution, needle tip (100 mm Pajunk, Germany) penetrates in the path towards the tip of the L3 transverse process. On encountering the L3 transverse process [Figure 2b], aspiration followed by injection of 0.4 ml/kg body weight of LA was injected (test dose 3 ml + remaining volume). The spread of LA was observed in real time with axial views.

Figure 2: (a) Needle tip medial to the tip of TP [transverse process] (b) LA spread dorsal to the TP. (c) Diffusion of LA along the posterior and anterior thoracolumbar fascia. (d) Spread of LA from dorsal of TP to ventral of TP in the lumbar plexus area

Postoperative management

In the post operative period, all patients were admitted in post anesthesia care unit (PACU) and monitored for hemodynamic stability and pain at rest using a Numerical Rating Scale (NRS) – 0 = no pain, 10 = worst pain imaginable. NRS scores at various time points, the time to first analgesic [TTFA] request and postoperative analgesics administered in PACU were documented. At any point of time, patients with a NRS 4 or more, received IV infusion of 50 mg tramadol. The site of nerve blocks for any block-related complications, adverse events like quadriceps weakness, sensory loss, and dysesthesias in the territory of lumbar plexus were assessed prior to discharge.

Primary outcome

In this single-center randomized controlled comparative trial, the primary outcome was to compare the average postoperative pain severity scores measured over 24 hours postoperatively.

Secondary outcomes

The secondary objectives investigated were the NRS pain scores at different time points at 0, 4, 8, 16, and 24 hours, the TTFA, opioid consumption between two groups at 24 hours, quadriceps function, and adverse events (20% fall in arterial blood pressure, oxygen saturation less than 95% in PACU and quadriceps weakness, sensory loss, and dysesthesias in the territory of lumbar plexus prior to hospital discharge). The block related outcomes monitored were needle tip to lumbar plexus or transverse process contact and spread of LA. Postoperatively, a routine institute protocol is hospital discharge on the 5^th postoperative day. However, hospital stay is extended in the event of cardiac, respiratory, neurological, nephrological, and electrolyte imbalances.

Sample size estimation

The sample size was calculated with the use of a two-tailed t-test. For the primary analysis, we determined that a sample size of 26 patients in each group (for a total of 52 patients) would provide the trial with a power of 80% to detect an effect size d = 0.8 at the 5% significance level in the average NRS over 24 hours between the LESPB and the LPB group. A sample size of 30 in each group was decided to tackle attrition rate and possible exclusions due to adverse perioperative events and drop-outs.

Statistical analysis

The data on categorical variables is shown as n (% of cases) and the data on continuous variables is presented as mean and standard deviation (SD). The inter-group statistical testing for continuous variables is done using the independent samples t-test. The inter-group statistical testing for categorical variables is done using the Chi-square test of independence. The underlying assumption for normality (Shapiro-Wilk) and equality of variances (Levene) is tested before subjecting the study variables to the statistical tests. All the results are shown in tabular as well as graphical format to visualize statistically significant differences more clearly.

The P values less than 0.05 are considered to be statistically significant for the primary outcome. All the hypotheses were formulated using two tailed alternatives against each null hypothesis (hypothesis of no difference). Analyses are performed according to the intention-to-treat method. Sample size calculation is done using G*Power ver. 3.1.9.7 and statistical analysis is done using JASP ver. 0.16.3.0 for MS Windows.

Results

There was a total of 60 patients who were randomized into two groups of 30 patients in each group with no drop-outs. Figure 3 depicts the CONSORT flow diagram of patient allotment in both groups.

Figure 3: CONSORT flowchart

Demographic data is depicted in Table 1. The mean NRS scores obtained postoperatively with standard deviations are provided in Table 2.

Table 1: Comparison of demographic data between both groups

Table 2: NRS scores at various time intervals in both groups

Primary outcomes

LESPB (Mean = 1.413, SD = 0.795) was statistically superior to LPB (Mean = 0.960, SD = 0.663) in reducing the average pain scores over 24 hours with a medium to large effect; t (58) = −2.399, P = 0.02, Cohen's d = 0.619 [Figure 4].

Figure 4: Average NRS over 24 hours between the two groups (points represent the mean and whiskers represent the 95% confidence intervals)

Secondary outcomes

LESPB was clinically superior to LPB in reducing NRS pain scale scores in most of the different time points over 24 hours as shown in Table 2 and Figure 5. LESPB was superior in reducing NRS at time to first analgesia than the LPB as shown in Table 2. There was no statistically significant difference between the T-times between the two groups with t (58) = −0.005, P = 0.996, Cohen's d = −0.001 (95% CI: −0.507 to 0.505). The graph illustrates [Figure 6] the opioid requirement in the LPB and LESPB groups.

Figure 5: NRS pain scale scores at various time points (points represent the mean and whiskers represent the 95% confidence intervals)

Figure 6: Opioid requirements between the two groups

In the LPB group, 13/30 (43.333%) patients required opioids, whereas in the LESPB group, 5/30 patients (16.667%) required opioids. A Chi-square test of independence is performed to examine the relation between requirements of opioids between the two groups. There is a statistically significant association between requirements of opioids in the two groups, X² (1, N = 60) = 5.079, P = 0.024. The LPB group required more opioids than the LESPB group.

Weakness of quadriceps function prevailed in two patients in the L group, which recovered at 3^rd and 5^th month, respectively. There are no incidences of hemodynamic instability and respiratory complications.

Results of ultrasound findings

The needle to lumbar plexus contact is visualized in all patients in their respective groups. A circumferential spread is visualized [Figure 1c and d] in real time at the site of injection in L group and a linear spread [Figure 2b] along the lamina and transverse process in the E group. A slight cephalad and caudal tilt of the probe revealed spread in the lumbar nerve root engulfed by the LA in the L group, while in the E group, with the same tilt local anesthetic spread was observed near the lumbar nerve root area in 13/30 [43.33%] patients. In the E group, the LA diffusion in the posterior and anterior thoracolumbar fascia were observed [Figure 2c and d].

Discussion

This single-center randomized clinical trial in patients undergoing PFN surgery for PFF's found that single-injection LESPB was comparable to a LPB in terms of postoperative pain relief for the first 24 hours after surgery.

Though case reports and case series have successfully implemented LESPB for postoperative pain relief after hip arthroplasty and PFN's, it is pertinent to test with a hypothesis, and, hence a trial comparing two approaches is advocated.[^2122232425] LPB is an attractive recommendation for unilateral hip surgical procedures with analgesic duration of single bolus of LA lasting for 4–8 hours and more than 8 hours with continuous infusion.[^2627282930] Moreover, a single injection LPB is superior to single injection femoral nerve block but inferior to subarachnoid morphine.[³¹] A meta-analysis did not reveal, plasma concentrations of LA reaching toxic levels after single bolus LPB, and in only 1/30, depicted LA toxicity.[⁸³²] Translocation of LA in the epidural space after a single bolus LPB has been reported at 3–27%.[^323334353637] Isolated case reports of total spinal, renal injury, and retroperitoneal hematoma are published in literature. Considering its high-risk profile, it is necessary to understand if ultrasound lumbar plexus blocks have reduced the incidence of complications.

Ultrasound of the lumbar paravertebral demonstrated the relation of the paraspinal muscles with the vertebral components, although, the lumbar plexus is seldomly visualized.[³⁸³⁹] The needle to lumbar plexus contact was discerned only after injection of local anesthetic agent.[⁴⁰] With the Shamrock approach the MEVLA_50/95 are 20.4 and 36.0 ml, respectively, of 0.5% ropivacaine and MRI studies illustrated contrast containment between the anterior and posterior psoas muscles highlighting the anterior rami L2, 3, and 4 nerves.[⁴¹⁴²] Based on a previous study, we considered injection of 0.2% ropivacaine at 0.2 ml/kg body weight.

LPB could be implemented in elderly patients with comorbid conditions.[⁴³⁴⁴] However, a retrospective study demonstrated a distorted paravertebral sonoanatomy and a hyperechoic psoas muscle obscuring the needle to lumbar plexus contact.[¹⁰] Though in a porcine model the optimal current is 0.5Ma,[⁴⁵] in our study, the needle – lumbar plexus contact was divulged under ultrasound and confirmed with evoked motor response (quadriceps contractions) at 0.4 mA as the end-point.

LESPB is an acquiescent option in elderly patients with multi-comorbid conditions. Considering LPB in a high-risk profile category as a suitable alternative was unavailable till the advent of LESPB. Case report and case series demonstrated efficacy in terms of NRS and opioid consumptions.[¹¹¹³] These paved the path for further execution of clinical trials and cadaveric studies. Considering the deposition of local anesthetic dorsal to the lumbar transverse process and achieving the scores and clinical outcomes similar to LPB, the LESPB is ‘LPB by proxy’.[¹³] Local anesthetic diffusion into the lumbar plexus and paravertebral area could be the possible mechanism of action. Cadaveric injections at the level of L4 transverse process resulted in the diffusion to ventral rami in just 16% of specimens; while in another cadaveric investigation, no translocation of dye was observed at the dorsal root ganglion, ventral rami, or paravertebral space.[²¹²²] However, extensive cranial-caudal spread, dorsal to the transverse process is visible in both studies though the injection points differed. Nevertheless, in our study spread from LESPB to lumbar plexus area was divulged in real time at 43.33%.

Erratic diffusion after cadaveric LESPB injections can be explained based on morphology of the fascia, anatomic connection between spaces and location of the needle tip.[²⁰] Cadaveric LESPB results cannot be extrapolated to human trials since factors like, muscle tone, breathing, blood circulation, and accumulation in tissues can substantially affect the diffusion.

Clinical and radiological studies evaluating the spread of CT contrast after LESPB confirmed sensory-affection from T12-S1 (20–40 ml LA) and a contrast spread from T12 to L5 dorsal to the transverse process with a visible contrast in the neural foramina [L2-3], at L3 it surrounded the psoas muscle and at L5 the contrast engulfed the spinal nerves, [^14,23]. MRI performed during LESPB revealed contrast in lumbar paravertebral, lumbar plexus area, and partially in epidural spaces.

A median volume of 3.3 ml is suggested for thoracic ESPB; however, in the lumbar area, the volume required is twice than thoracic area (5 ml and 2.5 ml, respectively).[⁴⁶⁴⁷] We injected 0.4 ml/kg BW based on prior study which injected 20–30 ml of LA. Apart from obtunding knee and ankle reflexes, no serious complications have been reported with LESPB at the level of L3 transverse process.[⁴⁸] A retrospective comparative study between LPB and LESPB employing catheter techniques did not find a significant difference in postoperative opioid consumption and pain scores in patients undergoing revision total hip replacement.[⁴⁹] An LPB being deep and closer to vascular structures in paravertebral area is contraindicated in patients on anticoagulants. However, due to the superficial positioning of needle tip at the transverse process without major vascular structure in the vicinity, the LESPB is at an advantage over the LPB.

Strengths in our trial includes stringent methodology, single block performer with more than more than 15 years of experience and the blinding of patients. Our assessment of the success of patient blinding indicated that it was successfully maintained.

There were several limitations of this study. This trial was not applicable to hip arthroplasty, catheter-based techniques and is not germane to surgical anesthesia. Though this study is not large enough to detect complications, quadriceps weakness was observed in the L group.

In conclusion, an LESPB is better than LPB in terms of reducing the average pain postoperatively over 24 hours. The total opioid consumption in the LESPB group was also lower than the LPB group postoperatively. Moreover, the LESPB is easy to perform and devoid of significant adverse events that are associated with LPB and would be more preferable for postoperative management in patients undergoing PFN for PFF's.

Financial support and sponsorship

Nil.

Conflicts of interest

There are no conflicts of interest.