Thoracic Epidural Analgesia for Lumbosacral Spine Surgery: A Randomized, Case-Control Study.

BACKGROUND: Traditional analgesics such as diclofenac and celecoxib have long been used in lumbosacral spine surgeries. Recently, preemptive single-shot caudal analgesia has been investigated by some workers with favorable results. We hypothesized that the thoracic route would not only allow preemptive but also postoperative analgesia through catheter insertion. AIM: We aimed at studying the feasibility and efficacy of thoracic epidural analgesia (TEA) in lumbosacral spine surgeries. SETTINGS AND
DESIGN: This was a prospective, randomized, controlled study that comprised 60 American Society of Anesthesiologist (ASA) Physical Status I and II patients posted for lumbosacral spine surgeries.
MATERIALS AND METHODS: Sixty ASA I and II patients were randomly divided into two groups: Group T - TEA was given using 0.2% ropivacaine 10 mL preemptive and postoperatively. Group C patients were given analgesia with intramuscular diclofenac 75 mg. Hemodynamic parameters, postoperative Visual Analog Scale scores, and neurological complications were noted. STATISTICAL ANALYSIS: Student's independent t-test for comparing the continuous variables and Chi-square test for the categorical variables. Kruskal-Wallis test was used for postoperative pain data.
RESULTS: Duration and quality of analgesia were superior in Group T. There were more hemodynamic alterations in Group C but no neurological complication in any patient.
CONCLUSION: TEA proves to be an effective analgesic technique for lumbosacral spine surgeries. Copyright:

INTRODUCTION

Lumbosacral spine surgeries, done for a variety of purposes, are one of the most commonly performed orthopedic surgeries today. Providing adequate analgesia, not just during surgery, but also postoperatively becomes a very important concern in these patients who have been moribund for weeks together and inadequate pain relief only adds to their morbidity and incapacitation.

There are, but, a limited number of analgesic techniques that can be used for these surgeries with success. Apart from the traditional analgesics used such as diclofenac and paracetamol, some workers such as Sekar et al.[1] and Samagh et al.[2] have quite recently pointed out preemptive use of single-shot caudal epidural analgesia as a very effective technique for use in lumbosacral spine surgeries.

Preemptive analgesia is the administration of analgesic medication before the onset of painful stimulus. The mechanism of action of preemptive analgesia has not been fully understood. Pain signals are transmitted by the peripheral nervous system causing plasticity of higher brain centers. This results in pain perception even in the absence of stimulus.[3] Thus, the administration of analgesia preemptively prevents the development of neuronal plasticity and provides effective pain relief.

One of the biggest strengths of the epidural route is that it provides the ability of extending analgesia into the postoperative period by insertion of a catheter in the epidural space. The challenge in case of lumbosacral spine surgeries is that catheter insertion through the traditional lumbar route might interfere with surgical field, thus leaving behind the use of preemptive single shot bolus dose as the more commonly utilized option.

We hypothesized that thoracic route, if used, would make catheter insertion feasible. This would allow not only preemptive analgesia but would allow us to extend the analgesic advantage into the postoperative period. Furthermore, thoracic technique would allow the use of lesser dose of the drug to produce the desired effect, thus reducing the hemodynamic alterations and chances of local anesthetic toxicity.

However, thoracic epidural analgesia (TEA) still remains an underrated technique which finds use in a limited number of scenarios. Jonnesco,[4] in 1909 first proposed puncture of the cord at the thoracic level for surgeries of neck and thorax. However, insertion of the needle above the termination of the cord incites fear in the mind of anesthesiologist about possible cord injury. Recently, the anatomy of the spinal cord was investigated with magnetic resonance imaging by Lee et al.[5] which showed that the thoracic cord is seated anteriorly within the vertebral column thus reducing the chances of needle contact with the neural tissue. Furthermore, the thoracic vertebral spines are directed obliquely downward, hence the direction of needle insertion, which has to be oblique, further elongates the distance between the needle tip and thoracic cord.

Subsequently, thoracic spinal anesthesia was studied for its feasibility in laparoscopic cholecystectomy by van Zundert et al.;[6] who then used this technique for cholecystectomy in a patient with severe lung disease.[7] Thoracic spinal anesthesia was even reported to be superior to general anesthesia in laparoscopic cholecystectomies by Yousef and Lasheen[8] Ellakany[9] used thoracic spinal anesthesia in breast cancer surgeries with favorable results. On the other hand, TEA has been used for postthoracotomy pain relief by Logas et al.[10] In addition, TEA has also been used in major colorectal surgeries with notable benefits.[11]

In this study, we aimed at evaluating the feasibility and efficacy of TEA in lumbosacral spine surgeries.

MATERIALS AND METHODS

This was a prospective, randomized, casecontrol study conducted in the Department of Anesthesia between November 2018 and August 2019. The study was performed after obtaining approval of Institutional Ethical Committee (L number - 25/2018-2019) and informed consent from all patients. Sixty American Society of Anesthesiologist (ASA) Physical Status I and II patients aged 20–65 years scheduled for first time lumbar discectomy/laminectomy were selected for the purpose of this study. The following inclusion criteria were followed for enrolment of patients in the study.

Inclusion criteria

Written informed consent from the patients

ASA physical status I and II

Patients aged 20–65 years of either sex

Patients scheduled for first time elective laminectomy and discectomy.

Exclusion criteria

Patient refusal

Body mass index (BMI) >30 kg.m−2

ASA physical status >II

Patients for emergency lumbar disc operations

Patients with major illnesses (e.g., cardiac, respiratory, renal, and hepatic)

Patients with coagulation abnormalities

A history of increased intracranial pressure, convulsions, and spinal stenosis

Infection at the needle insertion site

An allergy or contraindications to the drugs used in the study.

Using GPOWER software version 3.0.10 (Heinrich Heine University, Dusseldorf, Germany), it was estimated that the least number of patients required in each group with effect size of 0.25, 80% power, and 5% significance level is 25. We decided to include 30 patients to compensate for any drop outs or failures. Since we have to compare two groups in our study, we have included 60 patients in our study.

The patients were randomly divided into two groups:

Group T (thoracic group) - received general anesthesia with TEA with 10 mL 0.2% ropivacaine

Group C (control group) - received general anesthesia and systemic analgesia with diclofenac 75 mg.

Randomization was achieved by pulling envelops out of a partially sealed box. The opaque-sealed envelopes were opened by one anesthesiologist just before shifting the patient inside the operation theater (OT). This anesthesiologist, who was now aware of group allocation, administered the anesthesia according to group allocated; which included-performing the epidural block in Group T patients, induction of general anesthesia in both groups and administration of either epidural or systemic analgesia according to the respective group. The anesthesiologist who prepared and administered the drug was not involved in data collection. A blinded investigator, who was unaware of group allocation and drug administration, collected the data during the intra and postoperative period. This investigator did not witness the administration of anesthesia and entered the OT just before surgical incision.

All patients underwent a preanesthetic checkup, explained about the procedure in detail and were kept fasting for 6 h before surgery after a light evening meal. On the morning of surgery, an 18 G cannula was secured in all patients and ringers lactate 10 mL.kg−1 was started. Patients were then shifted to the operating room and all the monitors namely – pulse oximetry (SpO2), electrocardiogram (ECG), and noninvasive blood pressure were attached.

Epidural block

In Group T, patients were placed in the sitting position and under all aseptic precautions 18 G Tuohys epidural needle was inserted at T9–T10/T10–T11 interspace. The epidural space was confirmed using loss of resistance technique. The epidural catheter was then inserted 3 cm into the space with its tip-directed up. The catheter was aspirated for blood and cerebrospinal fluid (CSF). If blood was aspirated, the catheter was removed, and another trial was performed in a different space. If CSF was aspirated, the patient was excluded from the study. Then, a 3 mL test dose of 2% lidocaine with 1:200,000 adrenaline to exclude the subarachnoid/intravascular placement of the catheter was given. For the next 5 min, patients were asked for the signs of either intravascular (an abnormal metallic taste, tinnitus, dizziness, and tachycardia) or subarachnoid injection (by the ability of the patients to move their legs and the appearance of hypotension). If there were no signs, the catheter was fixed, and the patients were placed in the supine position.

The patients were given 0.2% ropivacaine slowly through the epidural catheter in 2 mL increments up to a total of 10 mL. This was done while keeping the hemodynamics parameters under close watch. The practice of administration of the drug slowly in increments was followed to further decrease the incidence of hypotension and reflex tachycardia which could be misinterpreted as inadequate analgesia and interfere with the observations of our study. After 10 min, the patients were given general anesthesia.

The dose of the drug to be given through the epidural route was calculated on the basis of a pilot study where three groups comprising 10 patients each were given TEA as follows: Group 1 – 7 mL of 0.2% ropivacaine, Group 2 – 10 mL 0.2% ropivacaine, and Group 3 – 12 mL of 0.2% ropivacaine. Whereas Group 1 showed inadequate analgesia, Group 3 patients had residual motor block (Bromage 1–2) at the end of surgery. Hence, it was decided to choose dose of 10 mL of 0.2% ropivacaine in our study. The pilot study patients were not included in the study.

General anesthesia

In both Group C and Group T, preoxygenation was done. Injection tramadol 2 mg.kg−1 was given, followed by injection propofol 1.5 mg.kg−1 slowly till loss of verbal contact occurred. Then, injection atracurium 0.5 mg.kg−1 was given, patients were ventilated through bag and mask for 4 min and endotracheal intubation was performed. End-tidal CO2 monitoring was started, and anesthesia was maintained with O2:N2O ratio of 33:66 and isoflurane 1%–1.5% minimum alveolar concentration. In Group C, injection diclofenac 75 mg was given intramuscular (IM) before the start of surgery. The patients were then turned prone and positioned with adequate padding at all pressure points.

In Group T, the time between the epidural block and the surgical incision was kept more than 20 min to provide adequate time for drug action. No other systemic analgesic was given.

At the end of surgery, patients were reversed with injection neostigmine 0.05 mg.kg−1 and injection glycopyrrolate 0.01 mg.kg−1. They were shifted to recovery for further monitoring.

Rescue analgesia was given when VAS score >3.

Group T – patients received postoperative analgesia with epidural top-up of 10 mL 0.2% ropivacaine.

Group C – patients received rescue analgesia with injection diclofenac 75 mg IM. The first postoperative dose was given on patient demand and time was noted. Subsequent doses of diclofenac were given every 8 h.

Rescue analgesia was administered by the first anesthesiologist who administered the anesthesia. He also made it a point to hide the epidural catheter so that it would not influence the data collector.

The following parameters were recorded by an independent observer:

All demographic parameters (age, sex, BMI, and ASA physical status class) and duration of surgery were recorded for all patients

All hemodynamic parameters such as heart rate (HR), systolic and diastolic blood pressure, ECG, and SpO2 were monitored just before the surgical incision; throughout the duration of surgery and immediately postextubation. Hypotension (defined as mean blood pressure falling more than 20% mmHg from baseline) was treated with bolus of 250 mL ringer lactate and injection mephentermine 6 mg, if need be. Bradycardia (fall in HR more than 20% from baseline) was treated with injection atropine 0.6 mg

Duration of analgesia (taken as the time from the administration of analgesia to demand of first rescue analgesic) was noted in both the groups

Examination of the motor function was done in the immediate postextubation period using the Modified Bromage score indicating the below:

Full movement – 0

Inability to raise extended legs/can bend knee - 1

Inability to bend knee/can flex ankle – 2

Complete paralysis/no movement - 3

Time to ambulation in both groups was noted

VAS scores were recorded on immediate extubation, then every hour for the first 3 h, then every 2 hourly for next 6 h and then every 3 h thereafter for the first 24 h

Patients were monitored for postoperative nausea and vomiting and any neurological complications for 24 h

At the end of 24 h, the patient and surgeon satisfaction scores (SSS) were recorded on a scale of 1–5 as follows:

Very satisfied – 5

Somewhat satisfied – 4

Neither satisfied nor dissatisfied – 3

Somewhat dissatisfied – 2

Very dissatisfied – 1.

The primary outcome of our study was – duration of analgesia (time to first rescue analgesia) and quality of pain relief (reflected by VAS scores) provided by thoracic epidural technique. The secondary outcomes were hemodynamic variations, neurological complications, and surgeon and patient satisfaction scores.

Statistical analysis

The compiled data were entered in a spreadsheet (Micro Excel) and then exported to the data editor of SPSS software version 20.0 (SPAA Inc., Chicago, IL, USA). The continuous variables were expressed as mean ± standard deviation and categorical variables were summarized as frequencies and percentages. Student's independent t-test was employed for comparing the continuous variables. The Chi-square test was applied for comparing the categorical variables. Kruskal–Wallis test was used for postoperative pain score as data was expressed as median and range. P < 0.05 was considered statistically significant. All P values were two-tailed.

RESULTS

Sixty patients who were scheduled for elective lumbosacral spine surgeries were included in this study. Patients in all the groups were comparable with respect to demographic characteristics age, sex, weight, ASA physical status, and duration of surgery [Table 1].

Table 1

Comparison of baseline demographic variables between group thoracic and group control (n=30)

	Group T	Group C	P
Age	48.30±6.67	47.86±5.43	0.721
Weight	72.80±5.43	74.05±6.20	0.682
Sex (female/male)	16/14	17/13	0.456
Duration of surgery (min)	250.06±11.6	248.15±10.3	0.542
ASA status, n (%)
ASA I	14 (46)	17 (56)	0.321
ASA II	16 (53)	13 (43)

The data are expressed as mean±SD and analyzed using unpaired t-test or as, n (%) and analyzed using Chi-square test. ASA=American Society of Anesthesiologist, Group T=General anesthesia + thoracic epidural analgesia, Group C=General anesthesia alone, SD=Standard deviation

The duration of analgesia was significantly longer in Group T (10.15 ± 1.50 h) than in Group C (7.35 ± 1.02 h) [Table 2]. None of the patients showed any motor deficit postextubation. Time to ambulation was significantly lesser in Group T (6.02 ± 0.75 h) than Group C (12.20 ± 1.02 h).

Table 2

Comparison of duration of analgesia in Group T and Group C (n=30)

	Group T	Group C	P
Duration of analgesia (h)	10.15±1.50	7.45±1.02	<0.001*

*Statistically significant. The data are expressed as mean±SD and analyzed using unpaired t-test. Group T=General anesthesia + thoracic epidural analgesia, Group C=General anesthesia alone, SD=Standard deviation

The incidence of tachycardia and hypertension was higher in Group C (40%) than in Group T (10%). There were no significant hemodynamic variations (bradycardia/tachycardia and hypertension/hypotension) in Group T.

VAS score comparison showed that Group T had significantly low scores at all time intervals than Group C [Table 3]. There was no significant difference in mean SSS between the two groups. However, the patient satisfaction scores were significantly higher in Group T [Table 3].

Table 3

Comparison of Visual Analog Scale, patient and surgeon satisfaction score between Group T and Group C

VAS score	Group T	Group C	P
Immediate postoperative	0 (0-0)	0 (0-0)	0.762
2 h	0 (0-0)	3 (0-0)	<0.001*
6 h	0 (0-0)	5 (5-5)	<0.001*
10 h	0 (0-0)	6 (5-6)	<0.001*
12 h	3 (3-3)	6 (6-6)	<0.001*
18 h	3 (3-4)	7 (6-7)	<0.001*
24 h	2 (2-2)	6 (5-6)	<0.001*
Patient satisfaction score	4.8 (4-5)	3.2 (3-4)	<0.001*
Surgeon satisfaction score	4.2 (4-5)	4.4 (4-5)	0.362

*Statistically significant. Data expressed as median and IQR. Analyzed by Kruskal–Wallis test. VAS=Visual Analog Scale, Group T=General anesthesia + thoracic epidural analgesia, Group C=General anesthesia alone, IQR=Interquartile range

There were no significant neurological or any other postoperative complications in any patient. However, the incidence of paraesthesia in Group T was 6.6%.

DISCUSSION

This study evaluates the feasibility and efficacy of TEA in lumbosacral spine surgeries. Apart from analgesia, we evaluated this technique for any possible serious neurological complications and also its impact on patient satisfaction when used in combination with general anesthesia.

The demographic characteristics were comparable in both the groups. TEA could be performed in the first attempt in 25 patients and on the second attempt in five patients with success. Dural puncture did not occur in any patient. However, two patients in Group T reported paraesthesia (6.6%) during catheter insertion. These paraesthesias were transient and too short lived to cause any inconvenience to the patient. This low incidence is similar to that seen by Imbelloni et al.[12] and van Zundert et al.[6] who reported an incidence of 6.6% and 5%, respectively. Yousef and Lasheen[8] pointed out that there is not any difference in the incidence of paraesthesias whether by thoracic or lumbar route.

We chose to use local anesthetic for epidural analgesia in our study because workers like Block et al.[13] demonstrated no better pain control using local anesthetic and opioid as compared with local anesthetics alone adding that using local anesthetics alone may decrease postoperative ileus in postlaparotomy patients. They further stated that TEA using opioids does not improve postoperative analgesia at rest compared with parenteral opioid therapy. Many workers such as Sekar et al.,[1] Samagh et al.,[2] and Kumar et al.[14] who have evaluated the efficacy of preemptive analgesia in lumbosacral spine surgeries also chose local anesthetics for providing analgesia in their studies.

Among local anesthetics, we chose ropivacaine as it produces sensory blockade leaving motor function largely unaffected.[15] Ropivacaine has also shown better central nervous system and cardiac safety profile than other local anesthetics. Hence, it would cause least interference in the assessment of motor function in the immediate postoperative period and also allow early ambulation.[14]

General anesthesia has long been the technique of choice for spine surgeries. This stems from the fact that it allows better airway control in the prone position and ability to extend anesthesia according to the length of surgery.[16] However, it has certain disadvantages also, such as increased chances of blood loss, delayed wound healing, and increased hemodynamic variability.

Regional anesthesia offers benefits such as decreased oropharyngolaryngeal morbidity, decreased blood loss, decreased chances of thromboembolism, and the ability to extend analgesia into the postoperative period. Many studies comparing general versus regional anesthesia for lumbar laminectomy and discectomies have shown some advantages for regional anesthesia with no significant differences in morbidity and mortality.[1718] However, still, regional anesthesia is an underutilized technique for spine surgeries, either because of fear of interference with the surgical site or because of relative inexperience with the technique.

Khajavi et al.[19] showed that general anesthesia combined with epidural analgesia (CGE) was a better alternative to general anesthesia alone because the combination carries the benefit of both. Subsequently, Alansary and Elbeialy[20] adopted CGE approach to compare the efficacy of dexmedetomidine and fentanyl as adjuvants to bupivacaine for elective lumbar disc surgeries.

Imbelloni and Gouveia[21] noted that as a consequence of the difference of the growth rhythm between the spinal column and the medulla, there is an increasing distance of the medullary segments from the corresponding vertebrae. To know which spinous process of the vertebra corresponds to which medullary segment we have the following rule: Between C2 and T10, we add two to the spinous process of the vertebra to find out the lumbar medullary segment while from T11 and T12 they correspond to the five lumbar segments. The process of L1 corresponds to the five sacral segments. This justifies our puncture at T10–T11/T9–T10 because nerves supplying lumbosacral segments derive from the medulla at this level. This is highlighted in the successful analgesia generated in all patients through the thoracic route in our study.

Furthermore, works by Imbelloni et al.[12] and Lee et al.[5] showed that thoracic puncture is safe as the posterior separation of spinal cord and dura is more at thoracic level than lumbar. Moreover, this separation is accentuated in the sitting position; hence, we administered the block in the sitting position in our study. The safety of this technique is reflected in the low incidence of paraesthesias and absence of neurological complications in our study.

As already discussed above, the anatomical data suggest that puncture at T10 allows deposition of the drug near the target dermatomes. This obviously decreases the dose and volume of the drug required to produce analgesia. This is highlighted in our study, where only 10 mL of 0.2% ropivacaine was adequate for producing analgesia in Group T. On the other hand, analgesia given through caudal or lumbar route requires almost double the volume of drug than used in thoracic route in our study. Samagh et al.[2] when evaluating preemptive analgesia with ropivacaine through the caudal route used 20 and 25 mL of the drug for surgeries above and below L5, respectively. Similarly, Kumar et al.[14] when administering preemptive ropivacaine epidural analgesia through caudal route had to use 20 mL of the local anesthetic. Khajavi et al.[19] when comparing CGE with general anesthesia administered 18 mL of 0.25% bupivacaine at the site or one segment below the site of surgery.

Because of the lesser volume and dose of the drug used in thoracic route with lesser subsequent sympathectomy, there are lesser hemodynamic variations in the thoracic group. Similar results were reported by Imbelloni et al.[22] who while comparing thoracic versus lumbar spinal anesthesia in laparoscopic cholecystectomies, reported lesser hypotension in thoracic group. On the other hand, general anesthesia group had a higher incidence of hypertension and tachycardia, especially when the patient entered lighter planes of anesthesia. The absence of such episodes with TEA could be because of the better pain control leading to attenuation of surgical stress. This has already been seen by workers like Li et al.[23] and Manion and Brennan[24] who found TEA to blunt physiological response to stress.

The duration of analgesia was longer and quality of pain relief (low VAS score) was superior in Group T than the control group. This resulted in higher patient satisfaction scores in thoracic epidural group. Our results are similar to those reported by Weber et al.[25] who compared TEA with patient controlled intravenous analgesia and found the former to be more effective. Halpern et al.[26] who compared the effect of epidural analgesia with parenteral opioid analgesia on progress of labor concluded that patient satisfaction and neonatal outcome were better with epidural analgesia.

Another important point of consideration is the motor block. Immediately after extubation, we evaluated all patients for any residual motor deficit by modified Bromage scale. Two patients in Group T had a score of Bromage 1 whereas none of the other patients had evidence of motor paralysis (Bromage 0). The patients with Bromage 1 also reverted to Bromage 0 within a span of 1 h. This is highly desirable as the analgesic technique to be used in such surgeries should not cause a sustained motor deficit as it will interfere with the postsurgical neurological evaluation of these patients. The absence of motor block can be partly attributed to the dose used by us which was decided on the basis of a pilot study, keeping in mind to exploit the analgesic potential of ropivacaine and avoiding unnecessary motor block. Another important factor is the use of ropivacaine. Studies have shown that ropivacaine demonstrates obvious sensory-motor separation, producing sensory nerve Aδ and C fiber blockage, while leaving the motor function of Aα fibers largely unaffected.[15] Mehta et al.[27] compared the epidural analgesic efficacy of bupivacaine and ropivacaine for lower limb surgeries and concluded that - 0.2% ropivacaine showed a distinct sensory-motor dissociation resulting in analgesia without motor blockade which improves ambulation and patient satisfaction compared to 0.2% bupivacaine.

Time to ambulation was significantly lesser in the thoracic group. This can be attributed to better pain relief in Group T. Similar results were reported by Kumar et al.[14] and Saoud et al.[28]

Another important point of discussion is the SSS. There was no significant difference in SSS between the two groups which means insertion of epidural catheter through thoracic route did not create any interference for the surgeon. Nor did it cause any serious neurological deficit in any patient making it a highly favorable technique for use in lumbar laminectomies and discectomies.

Limitations

In spite of blinding the observer, it was difficult to blind the patients, as those with epidural catheters could understand that they were being given analgesia using a different technique. This could have been avoided by inserting epidural catheters in all patients to prevent bias. However, in our study, unnecessary epidural puncture was avoided in the control group to prevent any infection. However, this remains a limitation in the study.

Second, patients were monitored for neurological complications for 24 h postoperatively. However, they should have been monitored for a longer period up to a week or more to rule out any delayed neuropathic complications.

CONCLUSION

TEA is a feasible and highly effective technique that not only helps in providing preemptive analgesia but also postoperative pain relief in lumbosacral spine surgeries. It produces a hemodynamically stable intraoperative period and is devoid of any serious neurological complications.

Financial support and sponsorship

Nil.

Conflicts of interest

There are no conflicts of interest.