Lumbar spine surgery is a common procedure associated with severe postoperative pain1 that, if poorly controlled, can increase complications and delay recovery. Opioids are the mainstay of therapy but are associated with adverse effects and a risk of long-term habituation and dependence.2

Table Case summaries
Full size table

Regional analgesia techniques can play a significant role in multimodal analgesia,3,4 but descriptions of their use in spine surgery are sparse.5,6,7 The erector spinae plane (ESP) block technique was first described for thoracic and abdominal analgesia via its action on the ventral rami of spinal nerves.8,9 Nevertheless, it also anesthetizes the dorsal rami, which innervate the paraspinal muscles and vertebrae (Fig. 1). In this report, we describe our adaptation of the ESP block to provide perioperative analgesia as part of a multimodal opioid-sparing regimen in a series of six patients undergoing lumbosacral spine surgery.

Fig. 1
figure 1

Graphic illustration of the erector spinae plane block. Local anesthetic is injected between the erector spinae muscle and the tip of the transverse processes. This anesthetizes the dorsal rami of the spinal nerves and their branches that innervate the paraspinal muscles and bony vertebrae. (Image adapted and used with permission from Maria Fernanda Rojas Gomez)

Full size image

Description of the ESP block and intraoperative anesthetic care

All blocks were performed with the patient in the prone position after induction of general anesthesia. The skin was disinfected with 2% chlorhexidine in 70% alcohol. Surface anatomy or ultrasound (counting up from the 12th rib) was used to identify the appropriate thoracic level and a high-frequency linear-array ultrasound transducer (SonoSite Edge, Bothell, WA, USA) covered in a sterile sleeve was placed in a longitudinal parasagittal orientation 3 cm lateral to the midline to identify the tip of the transverse process (Fig. 2A). A 21G 100-mm block needle (Pajunk, Geisingen, Germany) was used for single-injection blocks and an 18G catheter-over-needle set for continuous blocks (E-cath Plus; Pajunk, Geisingen, Germany; this catheter extends a fixed distance of 15 mm beyond final needle tip position). The needle was inserted in plane with the ultrasound beam in a cranial-to-caudad direction to gently contact the transverse process (Fig. 2B). Correct needle tip position was signaled by linear spread of the injectate solution (20-30 mL in total) separating the erector spinae muscle from the transverse processes (Fig. 2C and 2D). This process was repeated on the other side.

Fig. 2
figure 2

A) The ultrasound transducer is placed in a longitudinal parasagittal orientation approximately 3 cm lateral to the midline to visualize the tips of the transverse processes (TP) deep to the erector spinae muscle (ESM). B) The block needle is inserted in plane in a cranial-to-caudal direction to contact the TP. C) Injection of local anesthetic (LA) lifts the ESM off the TPs. D) Cranial-caudal linear spread of LA is clearly seen, separating the ESM from the TPs

Full size image

General anesthesia was maintained with propofol infusion 55-100 µg·kg−1·min−1 iv titrated using a Sedline® brain function monitor (Masimo, Irvine, CA, USA) to achieve a patient state index of 25-50 and bilateral spectral edge frequencies of 6-12 Hz. Rocuronium provided muscle relaxation for intubation in all cases. All patients were extubated prior to transport to the post-anesthesia care unit (PACU).

The bilateral ESP block catheters (Fig. 3) were connected in the PACU to two electronic infusion pumps (Sapphire™, Hospira, ICU Medical, San Clemente, CA, USA), which were each programmed to deliver patient-controlled boluses of 10 mL 0.2% ropivacaine at a lockout interval of 90 min with no background infusion. Patients were instructed to initiate boluses every 90 min when awake and at least every three hours during periods of sleep. Compliance was assisted by the use of the timer on patients’ smartphones and reminders from nursing staff.

Fig. 3
figure 3

A) Bilateral erector spinae plane block catheters inserted at T12 vertebral level prior to surgical incision. B) Subsequent incision for sacral laminoplasty and Tarlov cyst repair

Full size image

Case descriptions

Written informed consent was obtained from all patients for this report. Clinical details are summarized in the Table.

Patient 1

A 73-yr-old female underwent an L2-L3 lumbar spine decompression with Coflex® interlaminar stabilization (Paradigm Spine, New York, NY, USA).10 She was taking hydrocodone/acetaminophen 10/325 mg several times per day for lower back pain. She received bilateral single-injection ESP blocks at T12 with 30 mL 0.375% bupivacaine and 2 mg dexamethasone per side. Additional intraoperative analgesics included hydromorphone 1 mg iv at induction, acetaminophen 1 g iv, and ketamine 20 mg iv pre-incision. Wound infiltration using 10 mL 0.5% bupivacaine with 5 µg·mL−1 epinephrine was performed at surgical closure. The patient reported 0/10 pain on an 11-point numerical rating scale11 (NRS; 0 = no pain, 10 = worst pain imaginable) in the PACU. Neurologic examination revealed full motor strength and normal sensation to pinprick in both lower extremities. The patient was continued on acetaminophen 1 g iv six hourly for the next 48 hr and opioids as needed. During the first 24 postoperative hours, her NRS pain scores ranged from 2-4/10 and she received one dose of morphine 4 mg iv, 13 hr after surgery. During the next 24 postoperative hours, her NRS pain scores ranged from 0-6/10 and she received three doses of morphine 4 mg iv. During postoperative hours 48-72, her NRS pain scores ranged from 0-6/10 and she received six doses of oral hydrocodone/acetaminophen 10/325 mg. She was discharged home on the third postoperative day.

Patient 2

An 81-yr-old female underwent sacral laminoplasty and microsurgical repair of two Tarlov cysts. She reported sensitivity to opioids resulting in significant nausea and vomiting. She received bilateral single-injection ESP blocks at T12 using 23 mL 0.375% bupivacaine and 2 mg dexamethasone per side. Additional intraoperative analgesics included hydromorphone 1 mg iv at induction and wound infiltration using 10 mL 0.25% bupivacaine with 5 µg·mL−1 epinephrine at surgical closure. Her NRS pain score in the PACU was 0/10. She had normal motor strength and sensation in the lower extremities on neurologic testing. The patient was continued on acetaminophen 1 g iv six hourly for the next 48 hr and did not require any opioids during her hospital stay. Her NRS pain scores ranged from 0-2/10 and she was discharged home on the third postoperative day.

Patient 3

A healthy 46-yr-old male presented for coccygectomy. He received bilateral single-injection ESP blocks at T12 using 27 mL 0.375% bupivacaine and 2 mg dexamethasone per side. Additional intraoperative analgesics consisted of fentanyl 100 µg iv at induction and wound infiltration using 10 mL 0.25% bupivacaine with 5 µg·mL−1 epinephrine at surgical closure. His NRS pain score in the PACU was 0/10. The patient was continued on acetaminophen 1 g iv six hourly for the next 24 hr. Two hours after completion of surgery, he received hydromorphone 0.5 mg iv for a pain score of 5/10. The patient’s pain scores subsequently ranged from 2-5/10 during his overnight admission, and he received a total of three doses of hydromorphone 0.5 mg iv, two doses of morphine 2 mg iv, and two doses of hydrocodone/acetaminophen 10/325 mg. He was discharged from the hospital 20 hr after his arrival in the PACU.

Patient 4

A 67-yr-old female presented for sacral laminoplasty and microsurgical repair of a Tarlov cyst. She had multiple reported drug allergies, including morphine, oxycodone, duloxetine, gabapentin, and topiramate. Bilateral ESP catheters were placed at T12 and a loading injection of 25 mL 0.375% ropivacaine with 0.25 µg·kg−1 dexmedetomidine and 2 mg dexamethasone was administered per side. Additional intraoperative analgesics included hydromorphone 2 mg iv at induction, acetaminophen 1 g iv, magnesium sulfate 2 g iv pre-incision, and wound infiltration using 10 mL 0.25% bupivacaine with 5 µg·mL−1 epinephrine at surgical closure. Somatosensory evoked potentials were monitored throughout the case, with no changes noted from the baseline measurements obtained prior to the ESP block (Fig. 4).

Fig. 4
figure 4

Somatosensory evoked potential waveforms recorded from the right and left lower limbs of patient #4, showing no significant changes following erector spinae plane block

Full size image

Her NRS pain score in the PACU was 0/10 and she had full motor strength on neurologic testing of the lower extremities. Continuous ESP blockade was commenced in the PACU using the regimen described above. The patient received acetaminophen 1 g iv six hourly for the next 72 hr but did not require any opioids during her admission. Her NRS pain scores ranged from 1-4/10 during the first 24 hr, 1-3/10 during postoperative hours 24-48, and 1-2/10 during postoperative hours 48-72. The ESP catheters were removed just prior to her discharge home on the third postoperative day.

Patient 5

A 76-yr-old male presented for L1-L3 decompression with Coflex® interlaminar stabilization. He was taking 800 mg ibuprofen once or twice per day for low back pain. Bilateral ESP catheters were placed at T10 and a loading injection of 25 mL 0.375% ropivacaine with 0.25 µg·kg−1 dexmedetomidine and 2 mg dexamethasone was administered per side. General anesthesia was maintained with 55-75 µg·kg−1·min−1 propofol without a volatile agent. Additional intraoperative analgesics included fentanyl 250 µg iv at induction, ketamine 20 mg iv, magnesium sulfate 2 g iv pre-incision, and wound infiltration using 10 mL 0.5% bupivacaine with 5 µg·mL−1 epinephrine at surgical closure.

The patient’s NRS pain score in the PACU was 0/10. Continuous ESP blockade was maintained using the regimen described above and he received acetaminophen 1 g iv six hourly for the next 48 hr. Throughout his hospital admission, the patient reported NRS pain scores of 0/10 and required no opioids. He had normal motor power in both lower extremities. The ESP catheters were removed just prior to his discharge home on the second postoperative day.

Patient 6

A 55-yr-old male presented for L2-S1 decompression and excision of a L3-L4 intradural lesion. He was taking hydrocodone 30-40 mg daily and marijuana twice daily to manage chronic pain. Bilateral ESP catheters were placed at T10 and a loading injection of 2 0 mL 0.5% ropivacaine with 2 mg dexamethasone was administered per side. Additional intraoperative analgesics included hydromorphone 2 mg iv at induction, ketamine 0.5 mg·kg−1 iv pre-incision and 0.25 mg·kg−1 every hour, dexmedetomidine 0.4 µg·kg−1·hr−1, and wound infiltration using 10 mL 0.25% bupivacaine with 5 µg·mL−1 epinephrine at surgical closure. Somatosensory evoked potentials were monitored throughout the case, with no changes noted from the baseline measurements obtained prior to the ESP block.

His NRS pain score in the PACU was 0/10 and there was no change from his preoperative neurologic examination. Continuous ESP blockade was maintained using the regimen described above. He was started on oral gabapentin 300 mg every eight hours and acetaminophen 1 g six hourly. He did not receive any postoperative opioids until the morning of the first postoperative day, when he was started on oral extended-release oxycodone 10 mg twice daily to avoid symptoms of opioid withdrawal. No additional doses of opioid were required during his admission. He reported NRS pain scores of 0-3/10 during the first 48 hr and 0-2/10 during postoperative hours 48-72. The ESP catheters were removed just prior to his discharge home on the third postoperative day.

Discussion

Posterior spine surgery is amongst the most painful surgical procedures, with median pain scores (using the 0-10 NRS) on the first postoperative day ranging from 5 (spinal decompression) to 7 (spinal fusion).1 Opioids have traditionally been the mainstay of analgesia therapy, but they may not always adequately control pain and, at high doses, are associated with significant adverse effects (sedation, cognitive impairment, constipation) and the risk of long-term habituation and dependence.2 Regional anesthesia is an important component of multimodal analgesic regimens3,4; however, in spine surgery, this has been primarily confined to neuraxial techniques, namely epidural analgesia and intrathecal opioid.5,12 These have side effects and limitations and are not widely used. Local anesthetic wound infiltration is commonly performed but its benefit tends to be short-lived.13 Nevertheless, we employed it in all our patients as a matter of surgical routine as well as a means of delivering epinephrine to promote wound hemostasis.

The paraspinal muscles and posterior bony elements of the spine are innervated by the dorsal rami of the spinal nerves. These originate shortly after the spinal nerves exit the vertebral foramina and travel posteriorly through the intertransverse connective tissues and the paraspinal muscles to reach the superficial tissues.14 In the ESP block, local anesthetic spreads within the musculofascial plane deep to erector spinae muscles and acts on the dorsal rami of spinal nerves at multiple levels (Fig. 1). Evidence to date indicates that spread with 20 mL of injectate extends 3-4 vertebral levels or more from the site of injection in a caudal direction.8,9,15 Physical spread to the lumbar paraspinal area from a thoracic site of injection has also been documented, supporting the existence of a discrete anatomical pathway.9 We therefore aimed in all cases to target the T11 or T12 transverse process. This capacity for extensive cranial-caudal spread is a unique advantage of the ESP block, allowing it to be performed at a distance from the surgical field, thus minimizing the risk of microbial contamination and permitting the preoperative insertion of catheters to prolong postoperative analgesia. This is in contrast to another recently described regional analgesic technique for spine surgery, the thoracolumbar interfascial plane block, which requires injection at a vertebral level congruent with the surgical site.6,7

The observed lack of impact on intraoperative electrophysiologic monitoring and the absence of a motor block16 that might hinder postoperative neurologic testing and mobilization are additional potential advantages of the ESP block that should be confirmed in a larger patient population. The lack of correlation between the degree of analgesia and motor or sensory block achieved may be explained by the limited amount of local anesthetic that actually reaches the lumbar ventral rami or nerve roots. Low concentrations of local anesthetic applied to nerve targets have been shown to preferentially inhibit pain generation and transmission compared with motor and sensory function.17,18 At the same time, given the need for relatively large injectate volumes to achieve spread, we employed the maximum recommended dose of bupivacaine/ropivacaine in the initial bolus to avoid excessively low local anesthetic concentrations. While the ability of dexamethasone and dexmedetomidine to augment analgesia in ESP blocks is currently unsubstantiated, we chose to add them to the local anesthetic mixture based on data from peripheral nerve blockade19,20,21 and the principle that opioid sparing is best achieved by using as many multimodal analgesic strategies as possible2 rather than relying on a single “silver bullet”.

Regarding continuous ESP blockade, we chose a regimen of intermittent bolus dosing rather than continuous infusion to again ensure adequate local anesthetic spread from the catheter tip to the spinal nerves congruent with the surgical wound. At present, this choice is based on our understanding of the mechanics of the ESP block and anecdotal evidence.22 We note, however, that intermittent boluses appear superior to continuous infusion in epidural labor analgesia,23 in contrast to peripheral nerve blockade where the current evidence is equivocal.24 The former technique is more relevant to the ESP block given that both rely on local anesthetic spread within a relatively large anatomical space. A programmed intermittent bolus function was unavailable on our pumps and we instead improvised with a schedule of patient-initiated boluses. In practice, this worked well because of a high level of motivation amongst patients and nurses to maintain the degree of analgesia that was being provided.

In summary, pre-incision ESP blocks performed at the T10-T12 level contributed to effective perioperative opioid-sparing analgesia in this preliminary series of six patients undergoing lumbosacral spine surgery. Catheter insertion in more major surgeries and patients with complex pain issues allowed prolongation of this benefit and avoidance of opioid dose escalation.