Hepatic Hilar Nerve Block for Adjunctive Analgesia in Thermal Ablation of Liver Tumors: A Prospective Randomized Controlled Trial

Keywords

Nerve block - Pain - Adult liver cancer - ablation procedures - interventional procedures

Introduction

As the complexity of interventional radiology (IR) procedures increases, providing adequate procedural sedation and analgesia (PSA) is essential. The most commonly used drugs for PSA are combinations of local anesthetics, hypnotics, opioids, and non-opioid analgesics [1]. However, complications such as agitation, hypoxemia, and aspiration may arise, particularly in older adults or those with liver disease [2] [3].

In many surgical and palliative settings, local analgesic nerve blocks have been used with varying outcomes over the years [4]. Interventional radiologists occasionally perform a hepatic hilar nerve block (HHNB) at the level of the hepatoduodenal ligament in an attempt to control pain of a hepatic origin [5] [6]. He et al. showed in a non-randomized pilot study of 12 patients undergoing liver tumor ablation that an HHNB could be safely performed to provide periinterventional anesthesia [5]. The study demonstrated a lower mean visual analog scale (VAS) score for pain and a decreased need for intraprocedural fentanyl. Parhar et al. in 2023 conducted a retrospective cohort analysis of 177 patients undergoing HHNB with a significant decrease in PSA during thermal ablation of hepatic tumors [6].

This prospective randomized controlled study aimed to evaluate whether HHNB can be safely performed before liver tumor ablation to reduce the intraprocedural need for i.v. PSA and to minimize the subjective perception of pain.

Materials & Methods

Study design and sample population

This was a prospective randomized controlled clinical trial, which was conducted on a monocentric basis at a tertiary hospital between March 2020 and January 2022. Fifty patients with primary or secondary hepatic malignancies were included. As no prior data was available at the time of study commencement, we assumed a large effect size of d=0.8 as proposed by Cohen, a two-sided level of significance of 0.05, and an actual power of 0.8 yielding an estimated sample size of 50 patients [7]. A study flowchart is provided in [Fig. 1]. Inclusion and exclusion criteria are listed in [Table 1]. Patients, upon written informed consent, were randomly assigned before the procedure (1:1 randomization, Mersenne twister) to receive either HHNB in addition to intravenous PSA (test group) or PSA alone (control group). Patients were not informed of their group assignment. The primary outcome of this study was PSA medication requirement during and after the intervention. Secondary outcomes included the patients general health status assessed on the day of admission using the EQ-5D health assessment (EuroQol Research Foundation), as well as pain assessment using the numeric rating scale (NRS) before, one hour after, and one day after ablation. Ancillary analyses were performed to explore the impact of pain on various life aspects (e.g., general activity, mood, sleep, enjoyment of life).

Fig. 1 CONSORT flowchart of the clinical trial.

Table 1 Inclusion and exclusion criteria.

Inclusion criteria	Exclusion criteria
1. 18+ years	1. Allergic reactions to contrast agents or other substances to be administered
2. Patients with malignant hepatic tumors, including primary hepatic cancer and metastatic hepatic tumors	2. Pathologies in the region around the porta hepatis
3. Patients eligible for percutaneous liver ablation

The methods and reporting of this study adhere to the CONSORT guidelines [8].

Ethics committee

The study protocol received approval from the Institutional Review Board (0006/20) and adhered to the revised Helsinki Declaration of 2008. Our study was registered in a primary registry within the WHO International Clinical Trials Registry Platform (ICTRP).

Procedure technique

A preprocedural standardized imaging protocol was completed to locate intrahepatic lesions and determine the ablation path and regional block compartment. Prior to the procedure, an intravenous dose of 8 mg ondansetron and 8 mg dexamethasone was administered by slow injection.

Procedures were performed under CT fluoroscopy alongside additional ultrasonography or under MRI fluoroscopy by an interventional radiologist with approximately 10 years of experience in image-guided procedures. Unlike the control group, patients in the test group additionally received an HHNB with 25 mg of bupivacaine hydrochloride (10 mL of 0.25 % or 5 ml of 0.5% bupivacaine and 10 mL of sterile natrium chloride), immediately before the introduction of the thermal ablation applicators. The block was administered with a 23G Chiba needle using an intercostal right transhepatic approach in the area of the liver hilum anterior to the main portal vein. A low concentration of contrast agent (1 ml Imeron/Accupaque 300 diluted in 10 ml sterile NaCl) was added in CT procedures to visualize the block distribution around the portal vein ([Fig. 2]) while distribution was assessed on MRI using T2 sequences [9] [10]. Correct block placement was determined based on the overall distribution of bupivacaine/contrast agent admixture at the hepatic hilum area on CT/MRI, taking into account the target tumor location and the extent of the affected hepatic capsular treatment area after ablation.

Fig. 2 Documentation of the block in the control images after application of the bupivacaine/contrast agent at the hepatic hilum area.
After application of the bupivacaine/contrast agent at the hepatic hilum area using a 23G Chiba needle (a), using an intercostal, right transhepatic approach, documentation of the block in the control images (b).

Percutaneous thermal ablation was then performed using one or more applicator(s) (CT: NeuWave Microwave Ablation System, Ethicon or RF 3000 percutaneous Radiofrequency Ablation System, Boston Scientific; MRI: Microwave Therapeutic System, MML-Medical). In lesions close to critical structures, thermoprotective techniques were used.

All participants received, as per our protocol, a starting dose of 0.50–1.00 mg midazolam (0.010–0.035 mg/kg) and 50–100 μg fentanyl (1 μg/kg) intravenously based on body weight. Persistent pain was addressed with an additional 25–50 μg of fentanyl. Supplemental oxygen (2–4 l/min) was provided and vital signs were monitored throughout the procedure to manage potential complications. In cases of severe pain reported by the patient during the procedure (NRS 10/10 or a desire to discontinue therapy) despite adequate medication, or in the event of unmanageable autonomic reactions (e.g., hypertensive crisis), crossover and an additional HHNB were performed in the control group.

Repositioning of the probe and sequential ablations were executed if needed for satisfactory lesion coverage.

Patient monitoring, treatment evaluation, and follow-up

To quickly identify the first signs of deterioration of any vital parameters and avoid procedure- and sedation-related complications, dedicated personnel (IR nurse) continued appropriate monitoring in the recovery area for at least 60 minutes. Regular ward rounds while patients were on bedrest ensured consistent monitoring and continued as per the local institution’s practice for two days after the intervention until patient discharge.

Non-enhanced MRI with T1- and T2-weighted imaging, performed one day post-ablation, documented the procedure’s technical success. To assess the ablation’s extent considering that visceral receptors are mainly located in the liver capsule, we quantified the affected hepatic capsular area using manual surface segmentation [11] [12]. After eight weeks to three months post-procedure and subsequently every three months, we assessed for any delayed adverse events and reevaluated the efficacy of the thermal ablation using the (Modified) Response Evaluation Criteria in Solid Tumors (RECIST/mRECIST) evaluation system. Complications were documented using the Cirse Classification System for Complications [13].

Statistical analysis

Descriptive statistics for demographic characteristics of the patients and features related to the procedure were presented as mean and standard deviation (SD) for continuous variables and as frequency and percentage (%) for categorical variables.

Continuous variables were compared with the Student t-test and categorical variables using the chi-squared test or Fisher’s exact test. Analysis of Covariance (ANCOVA) was used to adjust endpoint variables for individual procedural characteristics and analyze the independent effect of PSA. The Mann-Whitney U-test was used to assess pain scores. A p-value of less than 0.05 was considered statistically significant. IBM SPSS version 26.0 was used (IBM Corp., Armonk, NY, USA).

Results

Baseline characteristics

Fifty patients (mean age: 65 ± 11 years; 28 men) with primary liver cancer (n = 14) or liver metastasis (n = 36) treated by microwave ablation (n = 49) or radiofrequency ablation (n = 1) who met the inclusion and exclusion criteria were included in this prospective study ([Fig. 1]). Most of the treated patients had liver metastases, with colorectal carcinoma as the main primary tumor (22 patients). Other primary tumors included hepatocellular carcinoma (12 patients), breast cancer (3 patients), cholangiocarcinoma (2 patients), and other tumors (11 patients). There was no statistically significant difference in both groups regarding patient demographic, clinical, and tumor characteristics ([Table 2]).

Table 2 Baseline patient characteristics.

Baseline patient characteristics	Test group (n=25)	Control group (n=25)	P* value
SD: standard deviation; BMI: body mass index * T-test was used for metric variables, Fisher’s exact test was used for frequencies
Sex, n (%)			1.000
Female	11 (44%)	11 (44%)
Male	14 (56%)	14 (56%)
Age, yrs.
mean ± SD	66.16 ± 11.79	63.92 ± 10.17	0.475
BMI, kg/m2
mean ± SD	25.56 ± 4.60	27.36 ± 3.64	0.132
Previous operation, n (%)			0.609
at hepatic hilum	1 (4%)	3 (12%)
at the ipsilateral liver lobe	2 (8%)	5(20%)
Liver cirrhosis, n (%)			1.000
	5 (20%)	5 (20%)
Tumor entity, n (%)			0.623
Primary liver tumor	7 (28%)	7 (28%)
Liver metastasis	18 (72%)	18 (72%)
Shortest distance between tumor and liver capsule, mm			0.230
mean ± SD	6.4 ± 8.7	9.8 ± 10.5

Intervention

The characteristics of the thermal ablation procedures are listed in [Table 3].

Table 3 Characteristics of thermal ablation procedures.

Characteristics of TA procedures	Test group (n=25)	Control group (n=25)	P* value
TA: thermal ablation; CT: computed tomography; US: ultrasonography; MRI: magnetic resonance imaging; SD: standard deviation; * T-test was used for metric variables, Fisher’s exact test was used for frequencies aAdjunctive techniques, like hydrodissection, were employed when the distance between the lesion and critical structures was less than 5 mm.
Imaging guidance during TA, n (%)			1.000
CT without US	2 (8%)	2 (8%)
CT with US	10 (40%)	9 (36%)
MRI	13 (52%)	14 (56%)
Adjunctive techniquesa in TA, n (%)			1.000
	3 (12%)	4 (16%)
Initial quantity of TA probes deployed, n			0.835
mean ± SD	1.40 ± 0.71	1.36 ± 0.64
TA duration, minutes			0.488
mean ± SD	13.16 ± 5.14	14.32 ± 6.50
CI 95%	[11.04, 15.28]	[11.64, 17.00]
Applied energy during TA, Joules			0.394
mean ± SD	84 840 ± 44 676	96 132 ± 48 096
CI 95%	[66 399, 103 281]	[76 287, 115 985]
Affected hepatic capsular area, mm2			0.034
mean ± SD	1139.12 ± 916.38	674.80 ± 542.16
CI 95%	[760.86, 1517.38]	[451.01, 898.59]

We considered the HHNB technically successful in nineteen patients (76%). In twelve patients, the block was distributed around the right portal vein; in four patients around the left portal vein; and in nine patients around both portal veins. Bupivacaine application involved the liver capsule in 10% of the patients (n = 5, all in the test group). Crossover and HHNB were successfully performed in three control group patients. Due to differences in the shortest distance of the tumor to the liver capsule, statistically significant differences in the affected hepatic capsular treatment area were recorded between groups (P = 0.034).

Patient medication

During the intervention, the fentanyl requirement in patients not receiving HHNB was 190 ± 57 μg, while in patients receiving HHNB it was 189 ± 89 μg. The midazolam requirement in the control group was 1.34 ± 0.57 mg compared to 1.60 ± 0.79 mg in the test group. T-test revealed nonsignificant differences (P=0.963 for fentanyl and P=0.187 for midazolam).

After adjusting the medication dosage for ablation times, total ablation energy, and affected hepatic capsular area, ANCOVA revealed significantly higher fentanyl application in the control group compared to the HHNB group (206 µg vs. 184 µg; P = 0.020). Contrarily, there was no statistically significant difference in midazolam administration between the groups (1.56 vs. 1.28 mg; P = 0.240). [Table 4] presents the primary endpoint results concerning PSA medication before and after covariate adjustment.

Table 4 Primary endpoint results: PSA medication required intraprocedurally before (a) and after (b) adjusting the dosage based on ablation time, total energy, and affected hepatic capsular area post-ablation.

PSA medication	Test group (n=25)	Control group (n=25)	P* value
PSA: procedural sedation and analgesia; SD: standard deviation; CI 95%: confidence interval at 95% * T-test was used for metric variables
a. t-test
Fentanyl, µg
mean ± SD	189 ± 88.99	190 ± 57.28	0.963
CI 95%	[152, 225]	[166, 213]
Midazolam, mg
mean ± SD	1.60 ± 0.79	1.34 ± 0.57	0.187
CI 95%	[1.27, 1.92]	[−0.65, 0.13]
b. ANCOVA with adjusted means
Fentanyl, µg
mean ± SD	184 ± 61.55	206 ± 69.80	0.020
CI 95%	[159, 234]	[209, 234]
Midazolam, mg
mean ± SD	1.56 ± 0.65	1.28 ± 0.75	0.240
CI 95%	[1.29, 1.83]	[1.11, 1.57]

Analgesia was required in the recovery area for 16% of the patients (n = 8), of whom seven received 1g of metamizole (P = 0.247 between group). Two test group patients, one of whom received 350 μg of fentanyl during the procedure, required additional antiemetics (8 mg ondansetron) due to nausea.

Health status and pain assessment

Despite similar results in all five areas (mobility, self-care, performance of usual activities, pain & discomfort, anxiety & depression), indicating no statistical difference in quality of life between groups (P= 0.798), the test group exhibited a significantly higher pain intensity before ablation (P= 0.018), while no statistically significant differences in pain intensity were reported after ablation (P= 0.223) or before hospital discharge (P= 0.755). [Table 5] presents these secondary endpoint results.

Table 5 Results of the pain assessment as given by the patients initially, after the intervention, and before hospital dismissal.

Numerical rating scale	Test group (n=25)	Control group (n=25)	P* value
* T-test was used for metric variables
Initial pain assessment (PI 1)
median (min.-max.)	2 (0–9)	0 (0–6)	0.018
Post-interventional pain assessment (PI 2)
median (min.-max.)	4 (0–7)	2 (0–7)	0.223
Pain assessment before hospital dismissal (PI 3)
median (min.–max.)	0.50 (0–4)	1 (0–6)	0.755

Additionally, on a 10-point pain scale, the post-procedural pain score increased by a mean of 1.55 ± 2.48 points in the control group and 1.14 ± 3.86 points in the test group (P= 0.98), while the pain score decreased before hospital discharge by a mean of 1.10 ± 1.92 points in the control group and 1.82 ± 2.72 points in the test group (P= 0.31). Moreover, the test group showed statistically significant differences in the pain’s impact on mood (P = 0.004), social relations (P = 0.019), and enjoyment of life (P = 0.035) compared to the control group.

Complications and therapy success evaluation

None of the 28 patients who received a temporary HHNB experienced any related adverse events during a mean follow-up period of nearly six months (range: 0–17 months). Nevertheless, complications occurred in 8% of the patients (n=4) due to the ablation procedure, with the most severe being a grade 5 colon perforation directly attributable to ancillary hydrodissection. One patient developed a liver abscess peri-procedurally, resulting in a prolonged hospital stay (>48 h, grade 3 complication), while two patients experienced grade 2 complications (non-infected hepatic bilomas) without necessitating extended hospitalization.

Furthermore, one patient from each group exhibited residual unablated tumors in non-enhanced MRI performed one day after ablation. During the follow-up period, local tumor progression was observed in one of the 50 patients with a local time to tumor progression of 8 months.

Discussion

Providing adequate PSA is essential during IR procedures. At the same time, conserving general anesthesia resources allows for more flexible planning of IR procedures, early patient mobilization, and a possible reduction in hospital stays. However, despite appropriate PSA, thermal ablation can still result in varying degrees of intra- and post-procedural pain [14].

In our investigation, performance of an HHNB significantly reduced fentanyl administration (ANCOVA, 184 µg in the test group vs. 206 µg in the control group; P = 0.020). In contrast to the study by Parhar et al. [6], which demonstrated a correlation between HHNB and reduced midazolam dosage, our study found no statistically significant difference in midazolam administered between the test and control groups (1.56 vs. 1.28, respectively; P= 0.240). This difference is likely because persistent pain in our study was managed with additional doses of fentanyl. In the study by Parhar et al., ongoing or worsening pain during the procedure was addressed with additional boluses of both fentanyl and midazolam, as determined by the treating interventional radiologist and nursing team. Our findings may also reflect the effect of HHNB on pain intensity rather than on alteration of consciousness. Compared to the study by Chao et al. [15], which retrospectively assessed medical charts of 8 patients undergoing CT-guided MWA of hepatic tumors with additional HHNB with a mean fentanyl dose of 197.2 μg and mean midazolam dose of 3.3 mg, we observed reduced amounts of both fentanyl (189.0 μg) and midazolam (1.6 mg). Similarly, in comparison to the approach of Chandrashekhara et al. [16] for pain control during RFA of liver lesion, we performed HHNB with a lower dose of bupivacaine hydrochloride (25 mg vs. 75 mg). This reflects our efforts to minimize medication doses to mitigate adverse effects.

After successful HHNB in control patients who were unwilling to proceed with the intervention due to extreme pain (crossover), thermal ablation was successfully completed in all three cases without the use of general anesthesia, thereby reducing resource utilization, costs, time, and risks.

Although no evident improvement in pain intensity was observed after thermal ablation (P= 0.223) or upon hospital discharge (P= 0.755) between groups, we found a significant difference in pain intensity between the test group and the control group patients before thermal ablation (P= 0.018), with the test group reporting higher initial pain levels, irrespective of whether this initial pain was localized to the area of the liver tumor to be treated or elsewhere in the body. Additionally, significant differences were observed between the two groups in the impact of pain on mood (P= 0.004), social relations (P= 0.019), and enjoyment of life (P= 0.035), with pain having a greater negative impact on these aspects in the test group. While it could be implied that the test group patients may have had heightened pain sensitivity or lower pain tolerance, which could potentially explain the lack of a significant decrease in pain after the HHNB, it is important to note that this remains an assumption.

No adverse events were observed during follow-up. Additionally, factors such as anatomical variabilities (including changes induced by resection or inflammation) und individual responses to medications can limit HHNB’s effectiveness. Lastly, considering experimental publications suggesting that a celiac plexus block may increase splanchnic perfusion, it is important not to ignore potential alterations of the HHNB in liver perfusion [17].

Study limitations include its monocentric setting and a relatively small sample size. Interventional and medication-administering radiologists were aware of HHNB performance, potentially introducing bias. Furthermore, although patients were not informed of their group assignment, a sham procedure was not performed in the control group patients to mimic the HHNB. A future double-blinded study with a sham procedure could enhance control over psychological and placebo effects. Additionally, the use of bupivacaine hydrochloride aimed to eliminate neurotoxic or cardiotoxic effects. Alternative doses or drugs with different action and potency might yield different NRS outcomes.

Conclusion

This prospective randomized controlled trial demonstrated a modest yet statistically significant reduction in fentanyl medication during liver thermal ablation after HHNB, aligning with the retrospective findings of previous publications. Interventional radiologists might consider this technique as an adjunct to intravenous medication, particularly for patients at risk of oversedation or for those undergoing ablation of tumors located close to the liver capsule, thereby avoiding the need for general anesthesia. Nevertheless, no difference in post-procedural pain between groups was observed. Given the variability in analgesic practices across different countries, it’s important to acknowledge the practicality of administering a nerve block to achieve a relatively modest reduction in intraprocedural fentanyl. This should be weighed against the potential clinical benefits of opioid reduction, justifying the additional procedure time and associated risks in clinical practice. Further research is needed to address the limitations of this study and to confirm the efficacy and safety of HHNB in larger multicentric studies.

• References

• 1 Romagnoli S, Fanelli F, Barbani F. et al. CIRSE Standards of Practice on Analgesia and Sedation for Interventional Radiology in Adults. Cardiovasc Intervent Radiol 2020; 43 (09) 1251-1260
• 2 Höhener D, Blumenthal S, Borgeat A. Sedation and regional anaesthesia in the adult patient. Br J Anaesth 2008; 100 (01) 8-16
• 3 Starczewska MH, Mon W, Shirley P. Anaesthesia in patients with liver disease. Curr Opin Anaesthesiol 2017; 30 (03) 392-398
• 4 Sag AA, Qadri YJ. Interventional Radiology Regional Anesthesia Approaches for Intra- and Postprocedural Pain Control. Semin Intervent Radiol 2022; 39 (02) 142-149
• 5 He KS, Fernando R, Cabrera T. et al. Hepatic Hilar Nerve Block for Hepatic Interventions: Anatomy, Technique, and Initial Clinical Experience in Thermal Ablation of Liver Tumors. Radiology 2021; 301 (01) 223-228
• 6 Parhar D, Baum RA, Spouge R. et al. Hepatic Hilar Nerve Block for Adjunctive Analgesia during Percutaneous Thermal Ablation of Hepatic Tumors: A Retrospective Analysis. J Vasc Interv Radiol 2023;
• 7 Cohen J. Statistical Power Analysis for the Behavioral Sciences. New York, NY: Routledge Academic; 1988
• 8 Schulz KF, Altman DC, Moher D. CONSORT 2010 Statement: Updated guidelines for reporting parallel group randomised trials. Ital J Public Health 2010; 7 (03)
• 9 Nitschke AM, Ray CE. Percutaneous Neurolytic Celiac Plexus Block. Semin Intervent Radiol 2013; 30 (03) 318
• 10 Stundner O, Meissnitzer M, Brummett CM. et al. Comparison of tissue distribution, phrenic nerve involvement, and epidural spread in standard- vs low-volume ultrasound-guided interscalene plexus block using contrast magnetic resonance imaging: a randomized, controlled trial. Br J Anaesth 2016; 116 (03) 405-412
• 11 Bhathal P, Grossman H. Nerve supply and nervous control of liver function. Bircher J, Benhamou JP, McIntyre N. et al. , ed. Nerve supply and nervous control of liver function. Oxford Textbook of Clinical Hepatology Oxford University Press; 1999
• 12 Alexander WF. The innervation of the biliary system. J Comp Neurol 1940; 72: 357-370
• 13 Filippiadis DK, Binkert C, Pellerin O. et al. Cirse Quality Assurance Document and Standards for Classification of Complications: The Cirse Classification System. Cardiovasc Intervent Radiol 2017; 40 (08) 1141-1146
• 14 Taylor AL, Gostin LO, Pagonis KA. Ensuring effective pain treatment: a national and global perspective. JAMA 2008; 299 (01) 89-91
• 15 Chao C, Park S. Evaluating our Initial Experience with Hepatic Hilar Nerve Block for Microwave Ablation of Liver Malignancies: Procedure Time, Efficacy and Duration [abstract taken from Journal of Vascular and Interventional Radiology]. Journal of Vascular and Interventional Radiology 2023; 34 (03) S54-S55
• 16 Chandrashekhara SH, Gupta N, Prakash A. A novel approach to pain control during radiofrequency ablation of liver lesion: Hepatic hilar nerve block. Cardiovasc Intervent Radiol 2023; 46 (07) 962-963
• 17 Sekiguchi Y. Effects of celiac plexus block on splanchnic circulation – II: Changes of the systemic hemodynamics and the blood flow of the liver and kidney in dogs. Jpn J Anesthesiol 1989; 38 (08) 1042-1047