Brain-based measures of nociception during general anesthesia with remifentanil: A randomized controlled trial
Keerthana Deepti Karunakaran,Barry D Kussman,Ke Peng,Lino Becerra,Robert Labadie,Rachel Bernier,Delany Berry,Stephen Green,David Zurakowski,Mark E Alexander,David Borsook
DOI: https://doi.org/10.1371/journal.pmed.1003965
2022-04-22
Abstract:Background: Catheter radiofrequency (RF) ablation for cardiac arrhythmias is a painful procedure. Prior work using functional near-infrared spectroscopy (fNIRS) in patients under general anesthesia has indicated that ablation results in activity in pain-related cortical regions, presumably due to inadequate blockade of afferent nociceptors originating within the cardiac system. Having an objective brain-based measure for nociception and analgesia may in the future allow for enhanced analgesic control during surgical procedures. Hence, the primary aim of this study is to demonstrate that the administration of remifentanil, an opioid widely used during surgery, can attenuate the fNIRS cortical responses to cardiac ablation. Methods and findings: We investigated the effects of continuous remifentanil on cortical hemodynamics during cardiac ablation under anesthesia. In a randomized, double-blinded, placebo (PL)-controlled trial, we examined 32 pediatric patients (mean age of 15.8 years,16 females) undergoing catheter ablation for cardiac arrhythmias at the Cardiology Department of Boston Children's Hospital from October 2016 to March 2020; 9 received 0.9% NaCl, 12 received low-dose (LD) remifentanil (0.25 mcg/kg/min), and 11 received high-dose (HD) remifentanil (0.5 mcg/kg/min). The hemodynamic changes of primary somatosensory and prefrontal cortices were recorded during surgery using a continuous wave fNIRS system. The primary outcome measures were the changes in oxyhemoglobin concentration (NadirHbO, i.e., lowest oxyhemoglobin concentration and PeakHbO, i.e., peak change and area under the curve) of medial frontopolar cortex (mFPC), lateral prefrontal cortex (lPFC) and primary somatosensory cortex (S1) to ablation in PL versus remifentanil groups. Secondary measures included the fNIRS response to an auditory control condition. The data analysis was performed on an intention-to-treat (ITT) basis. Remifentanil group (dosage subgroups combined) was compared with PL, and a post hoc analysis was performed to identify dose effects. There were no adverse events. The groups were comparable in age, sex, and number of ablations. Results comparing remifentanil versus PL show that PL group exhibit greater NadirHbO in inferior mFPC (mean difference (MD) = 1.229, 95% confidence interval [CI] = 0.334, 2.124, p < 0.001) and superior mFPC (MD = 1.206, 95% CI = 0.303, 2.109, p = 0.001) and greater PeakHbO in inferior mFPC (MD = -1.138, 95% CI = -2.062, -0.214, p = 0.002) and superior mFPC (MD = -0.999, 95% CI = -1.961, -0.036, p = 0.008) in response to ablation. S1 activation from ablation was greatest in PL, then LD, and HD groups, but failed to reach significance, whereas lPFC activation to ablation was similar in all groups. Ablation versus auditory stimuli resulted in higher PeakHbO in inferior mFPC (MD = 0.053, 95% CI = 0.004, 0.101, p = 0.004) and superior mFPC (MD = 0.052, 95% CI = 0.013, 0.091, p < 0.001) and higher NadirHbO in posterior superior S1 (Pos. SS1; MD = -0.342, 95% CI = -0.680, -0.004, p = 0.007) during ablation of all patients. Remifentanil group had smaller NadirHbO in inferior mFPC (MD = 0.098, 95% CI = 0.009, 0.130, p = 0.003) and superior mFPC (MD = 0.096, 95% CI = 0.008, 0.116, p = 0.003) and smaller PeakHbO in superior mFPC (MD = -0.092, 95% CI = -0.680, -0.004, p = 0.007) during both the stimuli. Study limitations were small sample size, motion from surgery, indirect measure of nociception, and shallow penetration depth of fNIRS only allowing access to superficial cortical layers. Conclusions: We observed cortical activity related to nociception during cardiac ablation under general anesthesia with remifentanil. It highlights the potential of fNIRS to provide an objective pain measure in unconscious patients, where cortical-based measures may be more accurate than current evaluation methods. Future research may expand on this application to produce a real-time indication of pain that will aid clinicians in providing immediate and adequate pain treatment. Trial registration: ClinicalTrials.gov NCT02703090.