Ultrasound Measurement of Pupillary Dilation Reflex During Induction of General Anesthesia Using Sufentanil: A Randomized Controlled Trial

Introduction

Insufficient analgesia can result in severe stress responses, significant hemodynamic variations, and the risk of awareness during surgery. Conversely, an excessive dose of opioids might increase the incidences of general opioids-related adverse effects. Anesthesiologists often titrate opioid dosage by observing somatic reaction as well as signs of sympathetic activation such as elevated heart rate (HR), blood pressure, tearing, and sweating [1,2]. In recent years, new monitoring devices have been developed to evaluate nociception during general anesthesia. Few monitoring devices for assessing the analgesic component of general anesthesia, like electroencephalography-based sedation monitoring for hypnosis, have been developed widely for clinical use because of their unverified precision and specificity, and it remains unclear which of the indices is the most useful. Among them, pupillometry is a promising method. The physiological mechanism of monitoring nociception is based on pupillary dilation reflex (PDR), which is the increase in pupil diameter (PD) in response to noxious stimulation. In patients anesthetized with propofol or inhalational anesthetics, PDR correlates with the intensity of stimulation, and may be a more sensitive index for assessing nociception compared with vital signs [3]. In addition, many confounding factors, such as hypovolemia, sepsis, and anemia, can affect blood and heart rate but do not directly affect PDR [4]. Pupillometry-guided opioid administration showed a significant reduction in intra-operative opioid consumption without increasing postoperative complications compared with standard analgesic care at the discretion of the anesthesiologist in charge [5,6].

Clinical examination of pupil size includes estimation of PDs by a cursory visual assessment followed by accurate measurement using sophisticated systems like infrared video pupillometry devices [7]. However, these instruments are not widely available for clinical use in departments of anesthesiology. Ultrasound, a standard instrument of anesthesiologists, can be useful as a simple, noninvasive, and widely available method for pupillary measurement. Even the pupillary light reflex (PLR) of patients with severe ocular soft-tissue injury or anterior chamber hemorrhage can be observed by ultrasound [8]. Ultrasonic pupillometry allows real-time and measurable imaging of pupillary structures and can be a good alternative to infrared pupillometry [9]. Few clinical studies of ultrasonic pupillary dilation reflex during the induction period of general anesthesia have been reported, and further research is needed to assess the possible usefulness of ultrasonic pupillometry in clinical practice. We intend to apply this real-time and visual method to guide the appropriate dose of sufentanil and even other opioids in the future.

The objectives of this study were to determine whether PDR induced by tetanic stimulation can be observed by ultrasound, and to assess the possible correlation between pupillary parameters and sufentanil dose during the induction period of general anesthesia.

Material and Methods

GROUP AND PATIENT RANDOMIZATION:

The patients were randomized into 3 groups (P, S1, and S2) based on a computer-generated randomization list, by blocks of 6 patients, in a 1: 1: 1 ratio. A sealed envelope containing the allocation group was opened by anesthesia nursing staff after entry of the patients into the operating room. Patients in groups P, S1, and S2 received a first sufentanil dose of 0 (saline), a bolus of 0.2 μg/kg and a bolus of 0.4 μg/kg, respectively. The anesthesia nursing staff calculated and prepared the correct dose of sufentanil based on the ideal body weight (IBW), which was determined as height (cm)-100 for man and height (cm)-105 for woman. The anesthesiologists and investigators who recorded the vital signs and measured PDs was blinded to group allocation. After taking measurements, groups P, S1, and S2 received a second bolus sufentanil dose of 0.6 μg/kg, 0.4 μg/kg, and 0.2 μg/kg, respectively, until the total dose of sufentanil reached 0.6 μg/kg before tracheal intubation.

INDUCTION:

Upon the subject’s arrival in the operating room, an intravenous access of upper limb and standard monitoring of electrocardiogram, blood pressure, oxygen saturation, electric nerve stimulator (TOF watch; Spacelabs Healthcare, USA), and bispectral index (BIS; Covidien, Ireland) were performed. Ringer’s lactate was infused intravenously at a constant rate. Baseline measurements, including SBP, diastolic blood pressure (DBP), HR, BIS, and PD, were taken 10 min after the patients’ arrival before general anesthesia. General anesthesia was induced by effect-site target-controlled infusion of propofol and the initial effect-site target concentration was set at 5 μg/ml until loss of consciousness. Afterward, sedation was maintained by continuous BIS-guided propofol infusion with target values of 40 to 60. Once BIS dropped below 60, the patient was infused with saline or different boluses of sufentanil (0.2 μg/kg or 0.4 μg/kg) according to the group. After 1 min of the bolus infusion, a bolus of 0.6 mg/kg of rocuronium was injected. Five min after the completion of first sufentanil injection, patients were immediately administered a standardized tetanic stimulation (50 mA, 60 Hz, 5 s duration) of electric nerve stimulator at the ulnar side of the wrist. The process of the pupillary dilation reflex induced by tetanic stimulation was continuously recorded using B mode ultrasound technique as a 30 s video. During the induction of general anesthesia, the ventilation mask was tightly applied on the patient’s face, and hand-controlled assisted breathing was used when the patient had no spontaneous breathing. Tracheal intubation was performed after a second dose of sufentanil (0.6 μg/kg, 0.4 μg/kg, or 0.2 μg/kg) was injected intravenously until the total dose of sufentanil reached 0.6 μg/kg (Figure 1).

TIME POINT:

We measured clinical data, including SBP, DBP, HR, BIS, and PD, 4 times at T0 (baseline awake state), T1 (after loss of consciousness), T2 (5 min after intravenous injection of sufentanil), and T3 (the moment of the maximum PD after stimulation).

THE PUPILLARY ULTRASOUND TECHNIQUE:

All patients were placed in supine position in the operating room under standard light condition (room lighting 90~130 lux) and silence. Ambient brightness was measured using a digital light meter (TA632A; TASI, China). B mode ultrasound (EDGE; SONOSITE, USA) was performed with the subject’s right eye closed using the linear probe (6–13 Hz). The ultrasound probe was positioned flatly on the lower eyelid of the right eye (Figure 2A) at suitable pressure and adjusted slightly toward the head side (Figure 2B) until a satisfactory pupil and iris plane was obtained for measurement (Figure 2C). At this level, the pupil presented as an anechoic round structure surrounded by an hyperechogenic ring, the iris. All insonations were performed by the same experienced investigator. Power settings were reduced to minimum, according to the ALARA (as low as reasonably achievable) insonation approach, and we adhered to current guidelines for orbital insonation [10]. Ultrasound was well tolerated during the observation of PDR. Saline was applied in place of contact gel to avoid eye discomfort. Ultrasonic pupil images of T0~T2 and a 30-s video of pupillary changes response to tetanic stimulus were obtained and stored in the same ultrasound machine. PD at T3 was measured after stimulus at the timepoint of maximum mydriasis. Each measurement of PD at T0~T3 was manually performed after induction of general anesthesia by the same investigator. The primary endpoints of our study were PDR amplitude (PDRA) and PDR rate (PDRR). PDRA was defined as the difference between PDT3 and PDT2 (PDT3-PDT2), whereas PDRR was defined as (PDT3-PDT2)/PDT2, which represented PD reaction to the tetanic stimulation.

Additionally, whether PDR disappeared, as well as PDR latency and PDR duration, were obtained from the 30-s ultrasound video. PDR latency was the time interval from the completion of the noxious stimulation to the beginning of pupillary dilation. PDR duration was the time interval from the pupil beginning to dilate to recovery of stable state.

STATISTICAL ANALYSIS:

The sample size was estimated, and PDRR was determined as the primary endpoint in this study. There were 5 cases in each of the 3 pre-experiment sample groups: 44.8%, 36.2%, 19.0%, 2.4%, and 5.4% in group P; 0%, 15.2%, 5.6%, 19.8%, and 21.0% in group S1; and 0%, 5.1%, 0%, 11.0%, and 0% in group S2, with a level of significance of 0.05 and a power of 0.9. Using PASS 15 software (NCSS, Utah, USA), we estimated a 15% dropout rate, resulting in the final enrollment of 31 patients in each group.

All statistical analyses were performed using SPSS statistical software 29.0.0 (IBM SPSS Statistics, New York, USA) and GraphPad Prism 10.2.0 (GraphPad Software, Inc, California, USA). Continuous variables were tested for normality using the Shapiro-Wilk test, and are expressed as mean (SD) if the distribution was normal. Otherwise, they are expressed as medians (IQR). Categorical variables are expressed as category counts and percentages. Demographic date and baseline data were compared among the 3 groups using one-way analysis of variance (ANOVA). The results of the PDR assessment were correlated to sufentanil doses using Spearman correlation. Statistical analysis was performed using ANOVA for repeated measures or the Kruskall-Wallis H test across all measurements of PD, SBP, DBP, HR, BIS, and χ² for category counts and percentages. Pairwise comparisons among the 3 groups were performed using Bonferroni correction. Mean SBP, DBP, HR, BIS, and PD before and after tetanic stimulation were compared using the paired t test. P values less than 0.05 were considered statistically significant.

Results

ULTRASONIC ASSESSMENT OF PDR:

Measurements of pupillary dimeters are detailed in Table 2. With the gradual infusion of propofol and sufentanil, there was a decrease in pupil diameter in groups S1 and S2 (3.21 vs 1.65 mm and 3.24 vs 1.63 mm, respectively). There was a significant increase in pupil diameter in group P (3.03 vs 4.22 mm), group S1 (1.65 vs 2.14 mm), and group S2 (1.63 vs 1.71 mm) after tetanic stimulation was applied. Use of the Kruskall-Wallis H test showed that PDRA and PDRR in group P was significantly higher than in group S1, and group S1 was significantly higher than group S2. Both PDRA and PDRR were significantly and negatively correlated with sufentanil dose [Spearman r=−0.84 (95% CI, −0.88 to −0.77), P<0.01, Spearman r=−0.74, (95% CI, −0.82 to −0.64), P<0.01] (Figures 4, 5). Strong negative correlation coefficients were observed. PDRA and PDRR decreased significantly with increasing sufentanil doses. The number of cases in which PDR was inhibited was statistically different among 3 groups (P<0.01). There were more cases in group S1 (9 cases) than in group P (0 cases) and there were more cases in group S2 (17 cases) than in group S1. Of the 89 patients included in this study, we compared the PDR latency and PDR duration for 63 patients because the PDR of 26 patients disappeared and the time parameters could not be obtained. PDR latency was statistically different among 3 groups, and after pairwise comparisons, PDR latency of group P (2.4±0.6 s) and S1 (2.5±0.5 s) was shorter than that of group S2 (3.0±0.6 s) (P<0.05). PDR duration was significantly different among the 3 groups, with group P (24.8±7.5 s) having longer PDR duration than groups S1 (7.7±1.9 s) and S2 (6.3±1.8 s) (P<0.05) (Table 2). A higher sufentanil dose during the induction period was associated with longer latency and shorter duration of PDR response to standardized stimulation.

CHANGES BEFORE AND AFTER STIMULATION:

The SBP, DBP, HR, BIS, and PD values of patients before and after tetanic stimulation were compared among the 3 sufentanil doses (Table 3). The differences in SBP, DB, P and BIS before and after tetanic stimulation were not statistically significant (P>0.05). Groups P and S1 showed increased HR and PD after stimulation (P<0.05), whereas group S2 showed only 1 significant increase in PD after stimulation (P<0.05), demonstrating that PD can respond promptly to tetanic stimulation even though high-dose sufentanil was administered.

Discussion

Previous studies demonstrated the criterion standard for quantitative pupillometry was infrared pupillometry, which was convenient and accurate [11]. Nevertheless, ultrasound pupillometry was also a good alternative to infrared pupillometry because there was a significant correlation between ultrasound pupillometry and infrared pupillometry [9]. Additionally, measurement of infrared pupillometry commonly requires measuring time and interval time, while ultrasonic pupillometry can be continuously monitored in real time without eyes open, which may provide better continuous monitoring during general anesthesia in the future.

The purpose of our study was to determine the potential clinical use of ultrasonic PDR to assess nociception. In our study, the ultrasonic PDR decreased with increasing doses of sufentanil. We found that ultrasonic PDR could be used to differentiate analgesic levels. Compared with group P, which was sedated by propofol alone without analgesia, sufentanil showed higher probability of inhibiting PDR completely in groups S1 and S2. The dose of sufentanil was strongly correlated with PDRA and PDRR, and sufentanil inhibited PDR in a dose-dependent manner. These results demonstrated that the ultrasound method of observing PDR revealed the noxious/anti-noxious balance in patients during induction of general anesthesia. Larson et al found that the correlation between PDR and alfentanil concentration was higher than that between blood pressure and heart rate, and that PDR effectively differentiated the analgesic intensity of alfentanil [12], which is consistent with our results. After opioid use, higher stimulus intensities are required to achieve the same pupil dilation reflex [13]. It has been demonstrated that PRD-guided antinociception reduced intra-operative opioid administration and provided a better understanding of the origins of hemodynamic alterations [6].

However, there are also controversial aspects in clinical use of PDR. Funcke et al [14] found that PD did not detect intracutaneous stimulation reliably with high doses of infused opioid, and PD was the index with the highest sensitivity and specificity in detecting stimuli under propofol sedation only. At moderate doses of infused opioid, the probability of PDR complete suppression began to increase compared with at the low doses of opioid. Therefore, PDR had limited predictive value of evaluating nociception. Several autonomic nervous system-derived indexes of the nociception-antinociception (NAN) balance are currently commercially available, including surgical pleth index (SPI), analgesia nociception index (ANI), and pupil dilation in response to noxious stimulation [15–17]. Defresne et al [18] found that the responses of the 3 NAN balance indexes, including pupil dilation to calibrated stimulation, were not able to predict the hemodynamic response to tracheal intubation and surgical incision. The concordance and correlation between the gradients of SPI, ANI, and PD was poor, possibly due to noxious stimulus of different intensities, population heterogeneity, and drug interactions.

In previous studies, the parameters for standardized tetanic stimulation were not established as fixed values. The 50 mA, 50 Hz, 5 s stimulation in our study was based on previously published work by Sandra et al [14] (80 mA, 50 Hz, 30 s), Rantanen et al [19] (50 mA, 50 Hz, 30 s stimulus), and Guglielminotti et al [15] (60 mA, 100 Hz, 5 s stimulus). We sought to use a stimulus of lowest intensity, frequency, and duration, with no risk of harm to the patients. The application of ultrasound pupillometry to explore the relationship between the intensities as well as types of stimulation and PDR need further study.

In children receiving sevoflurane anesthesia, PD measured with a pupillometer increased significantly (200±40%) after skin incision, whereas mean heart rate and arterial blood pressure increased only 11±7% and 10±8%, respectively, and BIS remained essentially unchanged [20]. The change of PD was a more sensitive measure of noxious stimulation than other commonly used variables. Exposure of patients to a noxious stimulus results in almost complete dilation of the pupil during the latency period of the cardiovascular response. The peak cardiovascular response occurs rapidly after application of the stimulus and remains in effect 1 min after removal of the stimulus, while the pupil diameter is already in the recovery phase at the peak of the cardiovascular response [21]. Noxious reflexes like PDR are produced within seconds, which is significantly faster than the cardiovascular response [22]. In our study, only HR and PD were statistically different after stimulation in the P and S1 groups, and only PD was statistically different after stimulation in group S2. However, no significant changes in SBP, DBP, and BIS were observed within 30 s after stimulation. We also found that HR was sensitive to tetanic stimulation only at low doses of opioids, whereas PD was sensitive to stimulation at all 3 doses. Therefore, ultrasonic PDR showed the highest sensitivity in detecting noxious stimuli at all sufentanil levels compared with SBP, DBP, HR, and BIS.

The pupillary latency was shorter in the absence of sufentanil and at low doses of sufentanil. There was a statistically significant difference in the duration of dilation with and without sufentanil. Sufentanil significantly shortened the duration of pupillary dilation because opioids can attenuate C-fiber activation of spinal cord neurons [23]. After sufentanil administration, observable PDR was delayed, PDRA was reduced, and the time of recovery was shortened. It might be clinically helpful to confirm the sufficiency of opioids through observing the time parameters of PDR to noxious stimulation, skin incision, or intubation.

The current study has some limitations, mostly due to its methodological design. First, the study is limited to the induction of general anesthesia, and intra-operative and postoperative ultrasonic PDR still needs further observation. The stimulation in this study was quantitative and standardized, whereas there is different stimulation in surgical procedures such as skin incision, pneumoperitoneum, and burning with an electrosurgical knife. The relationship between these surgical procedures and ultrasonic PDR changes cannot be determined without further study. Second, another bias may be due to the use of target-controlled infusion of propofol during the induction period. It has been demonstrated that target-controlled infusion of propofol is not associated with PDRA and PDRR [24,25], but it affects PD before tetanic stimulation [26]. As in a previous study by our team, PD was affected by both propofol and sufentanil [27], so PD without the process of PDR may not be suitable for monitoring analgesia. A last concern is that we need to solve many problems related to the measurement of ultrasonic PDR parameters. To the best of our knowledge, ultrasound images of pupil can be observed in real time, but the parameters need to be measured manually, which is a general limitation of ultrasonic measurement. Although all ultrasound image acquisitions and measurements in our study were performed by the same experienced investigator to ensure procedural consistency, it should be acknowledged that manual measurements may lead to measurement errors. Furthermore, the transient nature of pupillary changes precluded the possibility of repeated measurements. Therefore, ultrasonic automated imaging and measuring function, similar to those incorporated into infrared pupillometry, is needed to reduce the technical barriers such as manual measurement and inter-observer variability.

Conclusions

Ultrasound can observe PDR induced by tetanic stimulation, and sufentanil inhibited PDR in a dose-dependent manner. Compared with SBP, DBP, HR, and BIS, ultrasonic PDR had a shorter latency and was more sensitive to noxious stimulation. Our results suggest that ultrasound is a simple and noninvasive method for quantitative assessment of PDR, which shows a significant correlation with sufentanil dose. Nonetheless, further studies are required to extend the technique to perioperative use in monitoring nociception and titrating opioids individually for anesthetized patients.