Effect of the Location of Tetanic Stimulation on Autonomic Responses: A Randomized Cross-Over Pilot Study

Introduction

Surgery and painful intensive care unit (ICU) procedures elicit noxious stimuli that can cause hormonal, metabolic, and autonomic changes.^1,2 To attenuate the resulting physiological stress responses, clinicians use various pharmacological and non-pharmacological interventions to optimize nociception-antinociception balance.^3–10

Noxious input is needed to assess nociception-antinociception balance. Unfortunately, naturally evoked nociception, such as surgical stimulation and painful ICU care procedures, can be unpredictable in intensity, duration, and timing of occurrence, complicating its use to quantify nociception-antinociception balance. To overcome this limitation, a tetanic stimulus has been used as a “standardized” noxious stimulus to study autonomic nervous system activation, antinociceptive therapies, and to evaluate performance of nociception monitors.^3–13 Previous studies have applied the tetanic stimulus to various anatomical locations (ulnar-, median- and sural nerve, thoracic, lumbar and cervical dermatomes, finger and the forearm) with no prior validation that the resulting physiological responses are independent of the tetanic stimulus location.^{3–11,13–17} Thus, the aim of this pilot study was to investigate the effects of three anatomical tetanic stimulus application sites on physiological variables that reflect the autonomic nervous system responses.

Materials and Methods

In a study, that confirmed to the Declaration of Helsinki, with approval of the University of California San Francisco Institutional Review Board (IRB# 18–25325, ClinicalTrials.gov NCT03648853) and written informed consent we studied 12 patients between October 3, 2018 and November 30, 2018, who were scheduled for elective neurological surgery under general anesthesia and were over 18 years of age. We excluded patients who were unable to provide informed consent, were taking opioids, or whose anticipated duration of general anesthesia was less than one hour. This study was conducted at the University of California, San Francisco Medical Center.

All the patients fasted overnight. Prior to the induction of anesthesia, a 20 g or 18 g intravenous catheter was inserted into the vein of the right hand to administer intravenous fluids and medications. Standard anesthesia monitors were used (5-lead electrocardiography, noninvasive blood pressure monitoring in the right upper arm, and a pulse oximeter probe on the finger of the right hand). A second pulse oximeter probe was applied to the distal phalanx of the left index finger to collect data. During the study, patients were placed in the operating room in a supine position covered with blankets and/or surgical drapes.

Patients received fentanyl (up to 100 µg) and midazolam (up to 2 mg) premedication. Baseline blood pressure and heart rate were recorded prior to anesthesia induction. After preoxygenation with 100% oxygen, anesthesia was induced with intravenous propofol as per the attending anesthesiologist. Rocuronium (minimum, 50 mg) was administered after loss of consciousness to facilitate endotracheal intubation and prevent movement during data collection. Anesthesia was maintained with inhaled oxygen (2 L/min) and sevoflurane, targeted at an ET of 2 vol%. An intravenous phenylephrine infusion was used if necessary to maintain the systolic blood pressure within 20% of the baseline values. Positive pressure ventilation was mechanically controlled using an 8 mL/kg tidal volume, and the respiratory rate was adjusted to maintain the end-tidal carbon dioxide between 30 and 35 mmHg.

In this single-blinded (patient) study, a five second, 100 hertz (Hz), 70 milliamp (mA) tetanic stimulus (Stimpod NMS450, Xavant Technologies, Pretoria, South Africa) was delivered sequentially to three anatomical locations. The stimuli were delivered using standard electrocardiogram electrodes applied to the area of the left ulnar nerve, tibia and thorax. Electrodes for the ulnar nerve site were placed so that the distal electrode was on the skin approximately one centimeter above the flexor crease of the wrist, and the other electrode was immediately proximal to the first electrode parallel to the flexor carpi ulnaris tendon. Electrodes for the tibial site were placed on the skin on the medial surface (bony area) of tibia. Electrodes were placed next to each other parallel to the axis of the tibia at the mid-tibia level. Electrodes for the thoracic site were placed next to each other parallel to the axis of the body at the T5 dermatome of midaxillary line. The positive electrodes were placed distally. Three sets of two electrodes and connecting wires were left in place throughout the study so that tetanic stimuli could be delivered using the same Stimpod device.

For randomization, two blocks of the six potential stimulus location sequences were placed in sealed envelopes according to a randomly generated sequence. One of the sequentially numbered envelopes was opened by the principal investigator at the beginning of each trial. Randomization of the stimulus location was designed to reduce potential biases due to the stimulus sequence or time effects. The stimulation sequence was repeated twice for each participant. Each participant received six tetanic stimuli. The tetanic stimuli were delivered at a minimum of 2 min apart.

To achieve quasi-steady-state anesthesia and to minimize nociceptive inputs other than the tetanic stimuli used for study purposes, the study data collection was started at a minimum of 15 min after the beginning of anesthesia and endotracheal intubation and was completed before surgical incision.

Tetanic stimuli-induced peripheral vasoconstriction and changes in the pulse rate (PR) were measured using photoplethysmography (PPG). Infrared light transmitted through the fingertip was measured using a Masimo Radical-7 pulse oximeter (Masimo Corp., Irvine, CA, USA; Masimo SET software version 7.0.3.3), for which an adhesive LNCS Adtx sensor (Masimo Corp., Irvine, CA, USA) was placed on the left index finger. The sensor was optically isolated from ambient light by covering a pulse oximeter probe with black 6 mil plastic. PPG data were recorded continuously throughout the study period using an automated data-acquisition system.

The pulse oximeter sensor contains a low-voltage, low-intensity light-emitting diode that emits infrared light (approximately 910 nm). A portion of the light is transmitted through the finger. The detector photodiode in the sensor generates an electrical current proportional to the amount of light received.¹⁸ This electrical current is low-pass filtered (10 Hz) and then converted into analog-to-digital converter counts. The generated electrical current data were transmitted to a computer, sampled at 62.5 Hz, and saved with no further signal processing. These data were recorded using Pulse Ox Automated Data Collection software (Masimo Corp., Irvine, CA, USA; ADC v3.1.1.0).

Sample size calculations were not performed for this pilot study because preliminary data were not available, and this study was meant to provide preliminary data for potential future studies. From the 62.5 Hz PPG data, we identified the minimum and maximum light transmittances for each cardiac pulse using programs written in MATLAB (MathWorks, Inc., Natick, MA). These data were used to determine the AC and DC values for each pulse (Figure 1).¹⁸ The AC values signify the pulsatile portion of the light transmittance, and the DC values signify the nonpulsatile portion of the light transmittance. The AC and DC values were used to calculate the derived variables: ACDC = AC/DC × 100. The pulse rate was calculated using the beat-to-beat time interval data from the PPG recordings.

Figure 1 Illustration of photoplethysmogram DC and AC components. The DC value corresponds to the smallest blood volume in the finger (end diastole), which corresponds to the highest light transmission value of each pulse through the finger. The AC values correspond to the difference between the highest (end diastole) and lowest (end systole) light transmission values of each pulse representing the pulse added volume of blood in the finger.

The baseline values for AC, DC, ACDC, and PR were defined as the median values of 12 heartbeats immediately before the beginning of the tetanic stimulus. To analyze the stress response, we included data from 60 heartbeats following the beginning of the tetanic stimulus. We averaged the data from the two PPG recordings at each site. AC, DC, and ACDC data are expressed as percent changes relative to the baseline values. The maximum DC and PR, and minimum AC and ACDC values were defined as the highest and lowest values during the 60-heartbeat recording period, respectively.

For the primary endpoint, the maximum percent changes in AC, DC, ACDC, and PR values from baseline were compared among the three stimulus locations using ANOVA. The effect of tetanic stimuli on PPG measurements at each stimulus location was determined using repeated-measures ANOVA followed by Dunnett’s post hoc test. The AC, DC, ACDC, and PR baseline values and times to reach the post-stimulus maximum or minimum values were compared among the three stimulus locations using ANOVA. Data are reported as the mean ± standard deviation (SD), with p < 0.05 signifying statistical significance.

Results

Out of 12 subjects, data from 2 subjects were excluded from the analysis. One subject had no clearly defined responses and one study was aborted because the surgical procedure started before data collection was complete (protocol violation).

The patient characteristics and baseline hemodynamic and medication data of the remaining 10 subjects are shown in Table 1. A representative PPG recording of tetanic stimulus-induced vasoconstrictive response is shown in Figure 2.

Table 1 Patient Characteristics, Baseline Hemodynamic and Medication Data

Figure 2 A representative PPG recording of one tetanic stimulus induced vasoconstrictive response. Tetanic stimulus induced vasoconstriction results in an increase in light transmittance through finger. The data are in analog-to-digital converter units (ADC units). The slower (7–8 heartbeats) PPG oscillations are due to mechanical ventilation.

The AC, DC, ACDC or PR baseline values did not differ significantly between the stimulus locations (p=0.99, 0.99, 0.99 and 0.98 for AC, DC, ACDC and PR respectively). The AC and ACDC values decreased, and the DC values increased in response to tetanic stimuli-induced peripheral vasoconstriction at each location (p<0.001 for all) (Figure 3). The maximum changes (primary outcome) in AC, DC, and ACDC values did not differ significantly between locations (p=0.36, 0.39 and 0.06 for AC, DC and ACDC respectively; Figure 4). The PR values did not change significantly in response to tetanic stimuli at any location (p=0.37).

Figure 3 % change in PPG derived variable data (DC, AC, ACDC) from baseline values. Data are means (thick lines) ± SD over 45 heart beats. Time zero is the beginning of the vasoconstrictive response. The horizontal bars at the bottom of the figure illustrate values that are significantly (p <0.05) different from baseline values.

Figure 4 Maximum changes in AC, DC and ACDC values from baseline (in percent) after tetanic stimuli. Data reported as means ± SDs. Abbreviations: Ul, ulnar; Th, thorax; Ti, tibia.

Tetanic stimulus-induced AC, DC, and ACDC vasoconstrictive responses started during the 5 s tetanic stimulation. There were no significant differences in the times for AC, DC, or ACDC values to reach maximum changes between the stimulus locations (p=0.68, 0.07 and 0.07 for AC, DC and ACDC respectively; Figure 5).

Figure 5 Number of heartbeats to reach maximum AC, DC and ACDC values after beginning of tetanic stimulus induced vasoconstriction. Data reported as means ± SDs. Abbreviations: Ul, ulnar; Th, thorax; Ti, tibia; hb, heartbeat.

Discussion

Our results show that the anatomical location at which a tetanic stimulus is applied does not affect autonomic nervous system-mediated physiological responses, as measured using PPG. Our results further show that under these experimental conditions, changes in heart rate are not reliable indicators of the stress response. These results should be extrapolated with caution to anatomical locations other than those tested in this study.

There are several reasons for studying the effects of tetanic stimulation sites on autonomic response. First, the distribution and density of the peripheral endings of polymodal primary sensory neurons are inhomogeneous.² Thus, there are theoretical reasons why electrical stimuli may not induce similar levels of nociception when applied to different anatomical locations. Second, it is not feasible to place stimulating electrodes at a singular anatomical location in every subject for clinical or research purposes. Our results provide options for three sites at which the application of tetanic stimulus yields similar physiological responses. Third, tetanic stimulus-induced muscle contraction may be undesirable in certain situations and may cause additional sympathetic nervous system activation.² Therefore, we included stimulation sites that would elicit minimal muscular contraction (tibia, thorax). However, since our subjects underwent complete neuromuscular blockade to optimize PPG signal quality, future studies need to be conducted to determine whether our results apply when neuromuscular blockade is not used.

Nociception can cause hormonal, metabolic and autonomic changes. Autonomic nervous system activation induces physiological responses, such as increases in heart rate, blood pressure, peripheral vasoconstriction, changes in skin conductance, heart rate variability, and temperature.^1,2,19 These variables can be monitored continuously, in real time, and are being used in various devices. We chose to monitor changes in peripheral vasomotor tone and heart rate using PPG, which is currently used for surgical and postoperative pain assessments.^{3–5,7,13,20} PPG allowed us a unique possibility to use high-frequency data recordings to analyze the peak stress responses and beat-by-beat time course of the stress response.

A PPG has two components, commonly referred to as AC and DC. The AC component is caused by pulsatile changes in the tissue volume (mainly arterial pulsation). The DC component is due to transmission of light by nonpulsatile tissues, including venous blood and nonpulsatile portions of arterial blood.¹⁸ Noxious stimuli-induced stress responses mediate peripheral vasoconstriction.^19,21–24 This in turn, should decrease the AC component (reduced arterial pulsation) and increase the DC component (reduced venous and arterial blood volume) of the PPG, as observed in this study.

Peripheral vasoconstriction may have different effects on the AC and DC components of PPG, depending on the cause of vasoconstriction. For example, temperature-induced vasoconstriction has a larger effect on the DC than on the AC component of the PPG, whereas intubation-induced noxious stimuli have a larger effect on the AC than on the DC component.^20–24 Our study adds to the current literature by analyzing the effect of a tetanic stimulus on both the AC and DC components and a derived variable (ACDC) of the PPG. Our data demonstrated that tetanic stimulus-mediated peripheral vasoconstriction has a larger effect on the AC and ACDC components of the PPG than on the DC component.

In our study, peripheral vasoconstriction started within seconds after the beginning of the tetanic stimulus, peaked at an average of 10 heartbeats later, and lasted less than a minute. This fast time-course suggests a neural response mediated by norepinephrine release from presynaptic nerve terminals after the activation of the nociceptive medullary autonomic pathway. The time course of this vasoconstrictive response was identical to that observed intraoperatively after the surgical stimuli (personal observations). In addition to a relatively brief neural response, surgical noxious stimuli may activate catecholamine secretion from the adrenal medulla. The resulting hemodynamic changes can last several minutes beyond the end of the surgical stimulus that evokes the stress response, reflecting the pharmacokinetics of circulating catecholamines rather than the ongoing surgical stimuli.

Most available physiological measurements detect, but do not predict, inadequate nociception. This is a significant limitation of clinical anesthesia. This is partly due to the continuously changing surgical stimulation and variations in anesthetic concentrations. Instead of relying on unpredictable surgical noxious stimuli to assess the nociception-antinociception balance, we suggest that the use of periodic tetanic stimuli-evoked autonomic responses should be tested as a potential means to improve the predictability of antinociception monitoring.

Our study had several limitations. First, the sample size is small. No prior data were available for use in the sample size calculations. Power calculation using our ACDC data showed that this study had a 0.8 power (alpha=0.05) to detect a 30% difference between the measurement sites. The results of these pilot studies could be used to design future research. We measured peripheral vasomotor tone at the fingertip, the most commonly used site for pulse oximeter probes. As the magnitude of noxious stimulus-induced peripheral vasoconstriction is not the same in all vascular beds, our results are applicable only when PPG is measured from the fingertip.^19,22–30, We applied tetanic stimuli to three anatomical locations. We cannot infer weather our results would be applicable to the other anatomical locations that have been used in previous studies.

Our study had two subjects that were obvious outliers. One subject (not analyzed) had no clearly defined responses to any of the six tetanic stimuli. The patient was taking an alpha-1 adrenoceptor blocker. The second outlier exhibited a typical initial peripheral vasoconstrictive stress response. However, this subject’s recovery from vasoconstriction was significantly prolonged compared with that of the other subjects. This patient was taking a selective serotonin reuptake inhibitor (SSRI), suggesting the potential for reduced norepinephrine or neuropeptide Y reuptake from the synaptic cleft. These outliers contribute to concerns regarding concomitant medications that may affect nociceptive monitoring.

Conclusions

In Conclusion, this pilot study showed that the magnitude of peripheral vasoconstriction was independent of the location where tetanic stimulus was applied. These results should aid in the development and use of nociception-antinociception balance monitors and in conducting pain research using electrical noxious stimuli.