Pupil size and eye position were once important indicators of anesthetic action. It has been 78 years since Arthur Guedel1 published his classic monograph describing pupil changes accompanying the different stages of ether anesthesia. The pupil is not a consistent or reliable way to estimate the effect of modern anesthetics, and most physicians (other than ophthalmologists or neurologists) have only a fleeting interest in the information derived from the pupil.a The article by Larson and Behrends2 in this issue of Anesthesia & Analgesia provides us with a broad review of the physiology and pharmacology of the pupil and its reflex responses. Much of the information they present has been made possible by the development of modern infrared pupillometers that accurately measure pupil size but do not elicit any reflex change. Prof. Larson has been a champion of pupillometry for many years, publishing articles on both the scientific basis for pupil responses and the clinical applications of pupillometry in the operating room and intensive care unit (ICU). He suggests that simply observing the pupil is unlikely to give us more than a crude estimate of drug effect or a change in physiologic state. This paper starts to build the case that the purchase of a pupillometer may be justified by the measurement of pupillary reflexes that may improve care in the operating room or ICU.
Much recent literature on pupillometry concerns its scientific applications as a measure of pharmacodynamic effect, especially for opioids. Analgesia and respiratory depression are the opioid effects of most clinical interest, but both are difficult to measure rapidly and repeatedly with great accuracy. Pain and pain relief, similar to all subjective behavioral measures, can be affected by numerous confounders that are difficult to control, and there is really no way to measure analgesia continuously. Respiratory depression is another important measure of opioid effect, but it too has a behavioral component and numerous confounders. Experimental studies of respiratory depression must use cumbersome methods such as the response of minute ventilation to hypercarbia or hypoxia. Measuring the opioid miotic effect has several advantages: it is a central nervous system (CNS) effect that is mediated by μ-opioid receptors, it occurs at relatively low opioid doses, and the measurement can be extremely accurate (<0.1 mm with a modern pupillometer). Because the measurement is painless and noninvasive, measurements can be made frequently, or even continuously, if desired. Evan Kharasch et al.3 have used repeated measurements of pupil diameter (actually, the area under the curve of pupil diameter versus time) to characterize the pharmacokinetics and pharmacodynamics of probe drugs for CYP3A. They have also used it to measure drug transport and drug interactions such as those that inhibit CYP3A4 metabolism.4 We and others have used pupil response to perform simultaneous pharmacokinetic/pharmacodynamic modeling of opioids.5 There are significant limitations to the use of this measurement. As Prof. Larson points out, most clinical situations do not permit adequate control of ambient light or retinal adaptation. The baseline size of the pupil decreases with age, and the musculature and stroma of the iris may also limit the amount of constriction that is possible. Luckily, measurements of drug effect usually consider change from baseline rather than absolute effect, so many of these confounders are taken into account. One important limitation that has not been addressed is the effect of chronic opioid administration. In opioid-naive subjects, miosis can be used as a surrogate for analgesia or respiratory depression because there is a good correlation between the relative potencies to produce these 3 effects. During chronic opioid administration, tolerance occurs to opioid analgesia and respiratory effects, but very little tolerance occurs to miosis. The pupil response should therefore be a progressively poorer surrogate for the other effects over time.
All of these research applications discussed thus far use measurements of the resting (i.e., unstimulated) pupil, but Prof. Larson makes a case for the increased information obtained from 2 important reflexes: the pupillary light reflex (PLR), which is the miotic response to visible light, and pupillary reflex dilation (PRD), the response to a noxious or alerting stimulus. Prof. Larson proposes that absence of the PLR can be used as a criterion for CNS dysfunction from injury, mass lesion, or hypoxia, and the return of this response can be a measure of the effectiveness of therapy (e.g., during cardiopulmonary resuscitation). The PRD is an objective response to pain that has been used successfully to measure the extent of regional or epidural blocks in patients who are under general anesthesia or otherwise unable to describe what they feel.
For both the PLR and the PRD, there appear to be important limitations that must be overcome before either can be adequately predictable and reliable for clinical use. PLR is strongly affected by the baseline size of the pupil, which, in turn, is a function of ambient light, drugs, and a wide variety of things that are not practical to control, particularly during emergency situations such as cardiopulmonary resuscitation. It is difficult to know how to interpret the significance of a small reflex response to light without some baseline measurement for comparison. In theory, any detectable response indicates the presence of some CNS function, but how much response is adequate? For example, during CPR, would a small PLR be sufficient evidence that CNS hypoxia is irreversible, and resuscitation efforts are futile?
The PRD also has numerous possible reasons for failure. Similar to PLR, this reflex is dependent upon ambient light and the baseline size of the pupil. It can be blocked by dopamine D2 antagonists, so one might anticipate problems in patients treated with typical neuroleptic agents such as haloperidol. One also wonders if the response would be intact in patients with Parkinson’s Disease and other types of neurodegenerative disease. Opioids can block PRD, and Dr. Larson presents data (Fig. 9 in his paper) that elimination of PRD is complete at remifentanil concentrations that are achieved or exceeded in many general anesthetics. Data in the same figure suggest that even in the absence of opioids, PRD has a high variability and sometimes a small amplitude. The resulting very small bandwidth could make it difficult to decide if the PRD is truly absent.
My intent is certainly not to condemn pupillometry as a clinical tool but simply to point out some of the important questions about this measurement that remain unanswered. Prof. Larson and his colleagues have created a fine body of baseline work on both the pupillometer and the physiology it measures. Pupillometry has already established itself as a valuable tool for pharmacologic research. With the right additional studies, this painless, noninvasive measurement may prove equally valuable for assessing CNS and autonomic function in the operating room and ICU.
Name: Carl Rosow, MD, PhD.
Contribution: This author wrote the manuscript.
Attestation: Carl Rosow approved the final manuscript.
This manuscript was handled by: Tony Gin, MD, FRCA, FANZCA.
a Evidence of this is our acceptance of the acronym, PERRLA, “pupils equal, round, and reactive to light and accommodation,” which often appears in the report of a physical examination. How many have taken the time to ask, “did that physician really assess the pupillary response to accommodation—and if so, what did it mean?”
1. Guedel AE Inhalation Anesthesia. 1937 New York, NY The Macmillan Company
2. Larson MD, Behrends M. Portable infrared pupillometry: a review. Anesth Analg. 2015;120:1242–53
3. Kharasch ED, Walker A, Isoherranen N, Hoffer C, Sheffels P, Thummel K, Whittington D, Ensign D. Influence of CYP3A5 genotype on the pharmacokinetics and pharmacodynamics of the cytochrome P4503A probes alfentanil and midazolam. Clin Pharmacol Ther. 2007;82:410–26
4. Kharasch ED, Bedynek PS, Walker A, Whittington D, Hoffer C. Mechanism of ritonavir changes in methadone pharmacokinetics and pharmacodynamics: II. Ritonavir effects on CYP3A and P-glycoprotein activities. Clin Pharmacol Ther. 2008;84:506–12
5. Dershwitz M, Walsh JL, Morishige RJ, Connors PM, Rubsamen RM, Shafer SL, Rosow CE. Pharmacokinetics and pharmacodynamics of inhaled versus intravenous morphine in healthy volunteers. Anesthesiology. 2000;93:619–28