Ultrasound-guided single erector spinae plane block versus thoracic paravertebral block for patients undergoing video-assisted thoracoscopic lobectomy: a single center randomized controlled trial

Background: Whether ultrasound-guided erector spinae plane block (ESPB) can replace thoracic paravertebral block (TPVB) remains unknown. This study aimed to determine the efficacy of ESPB compared with TPVB for postoperative analgesia after video-assisted thoracoscopic lobectomy under general anesthesia. Methods: This prospective randomized controlled trial divided patients into a control group, a TPVB group (0.3 mL/kg, 0.5% ropivacaine), and an ESPB group (0.5 mL/kg, 0.5% ropivacaine). Dermatomes with loss of pinprick sensation, were recorded during 30 min after block administration. Visual analog scale (VAS) scores, total analgesic dose, and complications after surgery were recorded. Results: Whether at rest or during coughing, the VAS scores were lower in ESPB group at 1, 6, 18, 24, and 48 h after surgery compared with the Control group. VAS scores were similar in the ESPB and TPVB groups at 1 h, but were lower in the ESPB group at 6, 18, 24, and 48 h postoperatively. Conclusions: Single ESPB provided superior postoperative analgesia than TPVB, without causing any adverse effect.

depression [5], while TPVB is associated with a risk of spinal cord trauma, and pneumothorax [6]. There has been a recent effort to increase the use of ultrasound guidance when conducting TEA or TPVB, it reduced the failure rate and complications, but it is still technical challenges for a less skilled anesthesiologist.
Ultrasound-guided single erector spinae plane block (ESPB) was first reported by Forero et al. in 2016 [7]. ESPB may be technically simpler and safer than TPVB, because the transverse process act as a clear target for the injection, and acts as an anatomical barrier to avoids critical structures injury, such as pleura and spinal cord [8,9]. ESPB has been reported by many case reports that it can produce cutaneous sensory block, indicating that both the ventral and dorsal spinal rami were anesthetized [10,11], even can produce visceral analgesia by blocking the sympathetic chain [12,13]. This raises the question of whether the ESPB can serve as a technically simpler alternative to TPVB with a similar analgesic effect We hypothesized that ESPB is as effective as TPVB for relieving acute pain in patients undergoing video-assisted thoracoscopic lobectomy. The primary outcomes of the study were the postoperative visual analog scale(VAS) for pain under the status of rest and cough at 1, 6, 18, 24, and 48 h after surgery. The secondary outcomes were comparisons of the number of anesthetized dermatomes, the intraoperative remifentanil dosage, the doses of postoperative patient-controlled analgesia and rescue analgesia, the incidence of adverse effects, such as pneumothorax, local anesthetic intoxation, nausea and vomiting.

Patients and study design
This prospective randomized controlled study was approved by the ethics committee of Patients suffering from lung cancer were eligible for enrollment in this study if they were between 20 and 70 years of age, were scheduled to undergo video-assisted thoracoscopic lobectomy for a single lobe and systematic lymphadenectomy, and had an American Society of Anesthesiologists (ASA) class of I or II. Patients with local infection, coagulopathy, allergy to local anesthetics, obesity (body mass index > 30 kg/m 2 ), decreased pulmonary reserve, major cardiac disorders, liver and renal dysfunction, preexisting neurological deficits, or psychiatric illness were excluded.

Randomization and blinding
The patients were randomized using sequential sealed envelopes prepared by an independent statistician using a computer-generated random number table. Patients were randomly allocated to an ESPB group, a TPVB group, and a control group using computergenerated random numbers. Patients in the control group did not receive any regional block, and patients in ESPB and TPVB groups received the corresponding blocks. The postoperative assessors were blinded to grouping.

Procedures
Both ESPB and TPVB were performed at level T5 with patients in lateral position, with affected side upward, under ultrasound guidance with a Navi Series ultrasound device In the ESPB group, the transducer was placed 2-3 cm from the midline in a sagittal orientation. After the T5 transverse processes and the erector spinae muscle had been identified, the needle was inserted laterally to medically with out-plane technique until close contact was made with the transverse processe, without air or blood aspiration, 0.5 mL/kg (but no more than 40 ml) of 0.5% ropivacaine was administered. And we could see the liquid spreading between the transverse process and the erector spinae muscle.
In the TPVB group, the transducer was placed 2-3 cm lateral to the midline in a vertical orientation. Once the T5 transverse process, internal intercostal membrane and parietal pleura had been identified, the needle was inserted laterally to medically with in-plane technique until the tip lay in the thoracic paravertebral space beyond the internal intercostal membrane. 2ml normal saline was injected to confirm the pressing of the parietal pleura, 0.3 mL/kg of 0.5% ropivacaine was administered for the block.
Starting from 30 min after block administration, pinprick sensation in the T1-T12 dermatomal distribution was monitored every 5 min by an independent observer who was blinded to the study group. After confirming sensory loss, general anesthesia was induced with midazolam 0.02 mg/kg, sufentanyl 0.3-0.4 μg/kg, and propofol 1.5-2 mg/kg. Rocuronium bromide 0.8 mg/kg was administered to facilitate tracheal intubation. The trochal ports were made at the 7th and 8th intercostal levels along the midaxillary line.
Anesthesia was maintained with propofol and remifentanil, and the dosage was adjusted according to the hemodynamic parameters and bispectral index. Additional boluses of intravenous rocuronium bromide were given as required. A chest drain to be placed for at least 48 hours after surgery was inserted before the skin closure at the 7th and 8th intercostal levels. During the 10 min before the end of the operation, only tramadol 100 mg was used for analgesia. The intraoperative remifentanil dosages were recorded. Nonsteroidal anti-inflammatory drugs were not used during the operation.
Postoperative analgesia via a patient-controlled intravenous analgesia (PCIA) device was initiated immediately after the surgery and continued until 48 h after the surgery. The PCIA device contained 100 μg sufentanyl and 10 mg tropisetron diluted with normal saline to a volume of 100 mL. The device was set up to administer a background dose of 1 mL/h and a bolus dose of 2 mL with a 15-min lockout interval. Before the operation, patients were taught how to use the PCIA device to manage their pain and the visual analog scale (VAS) to score their pain. If the PCIA device could not meet a patient's analgesic demand, and the VAS score at rest was 4 or more, 50 mg flurbiprofen axetil was intravenously administered as rescue analgesia.
An investigator blinded to the group allocation recorded the VAS scores at rest and during coughing at 1, 6, 18, 24, and 48 h after the operation. The number of doses administered via the PCIA device, the total volume of analgesic drugs administered via the PCIA device, the doses of flurbiprofen axetil administered, and the incidence of pneumothorax, local anesthetic intoxation, postoperative nausea and vomiting after the surgery were noted.

Sample size calculation
As the primary outcome of this study was the the VAS scores postoperatively, we took the VAS scores at rest 18h post-surgery for example. According to the preliminary study (n=10, unpublished data), the mean VAS scores after ESPB and TPVB was 2.1 and 3.6, respectively, and the SD of the VAS scores was 1.6. We defined an acceptable superiority margin as 1, according to the previous study [14]. A sample size of 7 in each group was required to provide a power of 0.8 and a one-sided α of 0.05, given the possibility of missed follow-up rate 10%.

Statistical analysis
Statistical analysis was performed using GraphPad Prism 7 software. Numerical data were expressed as mean and standard deviation or median and range, as appropriate.
Qualitative data were expressed as frequency and percentage. Qualitative parameters were analyzed using the chi-square test or Fisher exact test. One-way analysis of variance (ANOVA), followed by Holm-Sidak's multiple comparisons, was used to compare the quantitative parameters between groups. P<0.05 was considered statistically significant.

General information
A total of 81 patients were possibly eligible for enrollment into the study. Of these, 3 patients did not meet the selection criteria, and 11 patients did not consent to participate in the study. The remaining 67 patients were enrolled in the present study, and were randomly allocated to the three study groups. The block could not be performed in a patient because of a big lipoma at the puncture site, 1 patient refused to participate in the study before anesthesia..Additionally, 2 patients required conversion to open thoracic surgery, 1 patient suffered severe allergic reaction during the operation, 2 patients developed hemorrhage. Therefore, 60 patients were included in our final analysis. Of these, 19 patients were randomized to the control group, 20 patients were randomized to the TPVB group, and 21 patients were randomized to the ESPB group. A flowchart of patient selection and the study procedures has been presented in Figure 1. Patients in each of the three groups were comparable with regards to age, sex, body mass index, ASA status, and duration of surgery (Table 1).
The number of PCIA doses delivered, the volume of analgesic drugs administered via the PCIA device were significantly lower in the TPVB and ESPB groups than in the control group (P<0.001), and were lower in the ESPB group than in the TPVB group (P<0.001) (   Fig. 5and 6). The consumption of flurbiprofen axetil were significantly lower in the TPVB and ESPB groups than in the control group (0-48h: TPVB vs Control, P = 0.019; ESPB vs Control, P <0.001), and were lower in the ESPB group than in the TPVB group (0-48h: P=0.019) (Fig. 7). Nausea and vomiting were the only postoperative complications in our patient cohort (

Discussion
This prospective randomized study showed that a single ESPB provided superior postoperative analgesia than the TPVB in patients undergoing video-assisted thoracoscopic lobectomy, which was different from recent studies. The other three studies all have proved that the analgesic effect of ESPB block is equal to or weaker than that of TPVB [14][15][16].
The most important reason was that high dose of local anesthetics were used in the ESPB block in our study. As with all interfascial plane blocks, ESP blocks are volume-dependent, and thus dermatomal coverage increases with increased volume [17,18]. We had performed a preliminary experiment prior to this study which showed that ESPB with 0.3 mL/kg ropivacaine resulted in a narrow distribution of sensory loss (tested using pinprick sensation) at 30 min after administration. When the block was performed with a dose of 0.5 mL/kg ropivacaine, obvious loss of pinprick sensation over a wide area could be measured within 30 min. But in TPVB, we injected local anesthetics into the paravertebral region to block the anterior and posterior branches of spinal nerves, even sympathetic nerve fibers directly. Thus, the optimal analgesic effect of TPVB could be achieved after a single ropivacaine dose of 0.3 mL/kg, according to previous studies [19,20] . The diffusion mechanism of local anesthetics in ESPB is different from that of TPVB, and it is appropriate for us to use different volumes of ropivacaine for the comparison of the two blocks.
Block failure or lack of efficiency of ESPB has also been reported by other literatures [21].
The focus of the dispute is whether ESPB can block the anterior branch of spinal nerve and sympathetic chain. And previous cadaveric studies radiological imaging reports differed on the spread of injectant of ESPB [17]. Some cadaveric and clinical evidences suggested that the local anesthetics could penetrate anteriorly through the intertransverse connective tissues into the paravertebral space where it blocked the ventral rami of the spinal nerves [7]. Additionally, the anesthetic might reach and block the rami communicantes and sympathetic chain to produce visceral analgesia [22][23][24]. The cadaver study by Adhikary and the magnetic resonance imaging study by Schwartzman concluded that the local anesthetics could spread to the the paravertebral space, neural foramina and epidural space [13,25]. Vidal et al found that the injectant could also spread to the intercostal space, except for the paravertebral space [26]. However, Ivanusic et al. found that dye couldn't spread to the anterior of the paravertebral space to involve the origins of the ventral and dorsal branches of the thoracic spinal nerves [27]. The multilayered fascia deep to the erector spinae muscle may be the cause of different results from anatomic studies [17]. In our study, during the ESPB block we used a needle with a sharp tip in close contact with the transverse process to ensured that the tip of the needle broke through the multilayered fascia beneath the erector spinal muscle. That may be the second reason for the superior analgesia effect of ESPB in our study Interestingly, the slower effect was observed despite the higher local anesthetic dose in the ESPB group than in the TPVB group. First, the dosage of intraoperative remifentanil in ESPB group was not less than that in TPVB group. Second, postoperative pain control was superior in the ESPB group as compared to the TPVB group during 6-48 h after the operation, but not at 1 h. This implies that the effects of the ESPB were more persistent than those of the TPVB. This may occur because an anesthetic drug injected into the space between the transverse process and the erector spinae muscle may only gradually penetrate anteriorly through the intertransverse connective tissues and into the paravertebral space, epidural space or intercostal space [22,23].
There were no block-related complications in our study, the only complications observed were nausea and vomiting. No case of local anesthetic intoxation and no case of pneumothorax occurred in both ESPB and TPVB groups. This suggests that the ESPB is safe for use in patients undergoing video-assisted thoracoscopic lobectomy. Additionally, the ESPB is less technically challenging than the TPVB. The erector spinae muscle and transverse process can be very easily imaged using a linear ultrasound transducer. Fang et al. has demonstrated that less block performing time, and higher success rate of puncture by once in ESPB group than in TPVB group [15].
There are several limitations to the current study. First, due to safety and time limitations, we only evaluated the block range 30 min after the block, but whether the block plane had been stabilized at this time needs further study. Second, we didn't measure the blood concentration of ropivacaine. There was no case of local anesthetic intoxation occurred in ESPB group in our study, after all, high volume and high concentration of ropivacaine increased the risk of local anesthesia toxicity [14,17]. Thirdly, our study was a randomized controlled trial with a small sample size from a single center. A multicenter study and more detailed experiments are needed to verify the effect and mechanism of ESPB.     The dosage of remifentanil used in the operation. *P < 0.05 compared with the control group.    The volume of analgesic drugs administered via the PCIA device. *P < 0.05 compared with the control group; #P < 0.05 compared with the TPVB group. Additional file1.xlsx