Prognostic Value of the Neurological Pupil Index in Patients With Acute Subarachnoid Hemorrhage

Background: The Neurological Pupil index (NPi) provides a quantitative assessment of pupil reactivity and may have prognostic value in patients with subarachnoid hemorrhage (SAH). We aimed to explore associations between the NPi and clinical outcomes in patients with SAH. Methods: A retrospective analysis of 79 consecutive patients with acute SAH. Age, sex, Acute Physiology and Chronic Health Evaluation-II score, and respiratory failure and NPi in each eye were recorded at admission. The primary outcomes included death and poor clinical outcome (defined as inpatient death, care withdrawal, or discharge Glasgow Outcome Score <4). Groups were compared using the Fisher exact test, and predictive models developed with fast-and-frugal trees (FFTs). Results: A total of 53 patients were included: 21 (40%) had poor clinical outcomes and 2 (4%) died. Univariate analysis found that only APACHE-II score (P < 0.001) and respiratory failure (P = 0.04) were significantly associated with poor clinical outcomes. NPi was lower among patients with poor clinical outcomes (mean 4.3 in the right eye and 4.2 in the left eye) vs those without (mean 4.5 in the right eye and 4.5 in the left eye), but neither was significant. However, the most accurate FFTs for death and poor clinical outcome included NPi after accounting for age in the death FFT and APACHE-II score in the poor outcome FFT (sensitivity [sn] = 100%, specificity [sp] = 94%, and accuracy (ac) = 94% in a model for death; sn = 100%, sp = 50%, and ac = 70%) in a model for poor clinical outcome. Conclusions: Our study supports the NPi as a useful prognostic marker for poor outcomes in acute SAH after accounting for age and APACHE-II score.

Original Contribution Section Editors: Clare Fraser, MD Susan Mollan, MD Prognostic Value of the Neurological Pupil Index in Patients With Acute Subarachnoid Hemorrhage Rahul A. Sharma, MD, MPH, Philip S. Garza, MD, MSc, Valérie Biousse, MD, Owen B. Samuels, MD, Nancy J. Newman, MD, Beau B. Bruce, MD, PhD Background: The Neurological Pupil index (NPi) provides a quantitative assessment of pupil reactivity and may have prognostic value in patients with subarachnoid hemorrhage (SAH). We aimed to explore associations between the NPi and clinical outcomes in patients with SAH. Methods: A retrospective analysis of 79 consecutive patients with acute SAH. Age, sex, Acute Physiology and Chronic Health Evaluation-II score, and respiratory failure and NPi in each eye were recorded at admission. The primary outcomes included death and poor clinical outcome (defined as inpatient death, care withdrawal, or discharge Glasgow Outcome Score ,4). Groups were compared using the Fisher exact test, and predictive models developed with fast-and-frugal trees (FFTs). Results: A total of 53 patients were included: 21 (40%) had poor clinical outcomes and 2 (4%) died. Univariate analysis found that only APACHE-II score (P , 0.001) and respiratory failure (P = 0.04) were significantly associated with poor clinical outcomes. NPi was lower among patients with poor clinical outcomes (mean 4.3 in the right eye and 4.2 in the left eye) vs those without (mean 4.5 in the right eye and 4.5 in the left eye), but neither was significant. However, the Department of Ophthalmology (RAS, PSG, VB, NJN, BBB), Emory University School of Medicine, Atlanta, Georgia; Department of Neurology (VB, NJN), Emory University School of Medicine, Atlanta, Georgia; Department of Neurological Surgery (OBS, NJN), Emory University School of Medicine, Atlanta, Georgia and Department of Epidemiology (BBB), Emory University School of Medicine, Atlanta, Georgia. Supported by the Emory Eye Center Pilot Grant ($2000). B. B. Bruce is a medicolegal consultant for Bayer and individual litigants on the topic of idiopathic intracranial hypertension. V. Biousse and N. J. Newman are consultants for GenSight Biologics. They are supported in part by an unrestricted departmental grant (Department of Ophthalmology) from Research to Prevent Blindness, Inc, New York, by NIH/NEI core grant P30-EY06360 (Department of Ophthalmology, Emory University School of Medicine), and by NIH/NINDS (RO1NSO89694). N. J. Newman is a consultant for Santhera Pharmaceuticals and Stealth BioTherapeutics. Presented at the 2020 North American Neuro-Ophthalmology Society Meeting; Winner of Best Abstract by Fellow Award ($1,000). The authors report no conflicts of interest. Address correspondence to Beau Bruce, MD, PhD, NeuroOphthalmology Unit, Emory Eye Center, The Emory Clinic, 1365-B Clifton Road NE, Atlanta, GA 30322; E-mail: [e-mail] 256 most accurate FFTs for death and poor clinical outcome included NPi after accounting for age in the death FFT and APACHE-II score in the poor outcome FFT (sensitivity [sn] = 100%, specificity [sp] = 94%, and accuracy (ac) = 94% in a model for death; sn = 100%, sp = 50%, and ac = 70%) in a model for poor clinical outcome. Conclusions: Our study supports the NPi as a useful prognostic marker for poor outcomes in acute SAH after accounting for age and APACHE-II score. Journal of Neuro-Ophthalmology 2022;42:256–259 doi: 10.1097/WNO.0000000000001474 © 2022 by North American Neuro-Ophthalmology Society T he manual pupillary examination is an integral component of the neurological examination in critically ill patients but is limited by subjective categorization of pupil reactivity (e.g., nonreactive, sluggish, or brisk) and interrater variability. Handheld, noninvasive automated pupillometers use infrared light and a digital camera to measure pupil responses, likely with greater accuracy and reproducibility than manual examination (1). Pupillometry yields numeric parameters including pupil latency, maximum constriction amplitude, pupillary light reflex amplitude, constriction velocity (CV), and dilation velocity (DV). An additional output, the Neurological Pupil index (NPi), is a composite measure calculated by a proprietary algorithm from latency, CV, and DV; it provides a value from 0 to 4.9 that is intended to reflect the overall robustness of the pupil light reflex relative to a normative database of controls (2). Automated pupillometry has generated a significant interest in the field of critical care. Several studies suggest the prognostic value of the NPi in patients with subarachnoid hemorrhage (SAH) (1): Changes in NPi have been associated with current or impending intracranial hypertension (3), cerebral vasospasm (4), and overall clinical severity in aneurysmal SAH (5). The association between poorer clinical outcomes and the lower NPi in patients with SAH is believed to relate to direct oculomotor Sharma et al: J Neuro-Ophthalmol 2022; 42: 256-259 Copyright © North American Neuro-Ophthalmology Society. Unauthorized reproduction of this article is prohibited. Original Contribution nerve compression, midbrain distortion, or uncal herniation due to either the direct effect of blood products or a secondary rise in intracranial pressure. METHODS Ethics Approval and Consent to Participate Ethics approval was granted by our institutional review board. Eligible patients who lacked decision-making capacity as per the treating physician were enrolled in the study under a waiver of informed consent. Study Design, Setting, and Selection of Participants A retrospective subanalysis of the prospective FOTO-ICU study (6), which enrolled patients with acute SAH admitted to the neurological intensive care unit (ICU) between September 22, 2014, and April 10, 2015. Patients were excluded if younger than 18 years or had a known history of chronic ocular fundus abnormalities or a medical condition associated with chronically elevated intracranial pressure. Exposures and Outcome Measures Exposure of interest was the NPi, as measured using the NPi-100 pupillometer (NeurOptics, Irvine, CA) as soon as the treating ICU physician deemed the patient clinically stable. All outcome variables and covariates were obtained from an unmasked review of the electronic medical record. Outcomes of interest were poor clinical outcome and inpatient death; poor clinical outcome was defined as death, withdrawn care, or a discharge from the hospital with a Glasgow Outcome Score , 4. Covariates of interest included age, sex, APACHE-II score (7) at ICU admission (ordinal, ranging from 0–71), and respiratory failure at ICU admission (binary; 1 if patient required invasive mechanical ventilation at admission and 0 otherwise). Analytic Plan (18.9%) had respiratory failure at ICU admission. Two (4%) had an abnormal NPi. Twenty-one (40%) had poor clinical outcomes, and 2 of the 53 patients (4%) died. FFTs were developed for death and poor clinical outcome (Fig. 2). For patient death, age and the NPi in each eye produced the best FFT with a sensitivity of 100%, specificity of 94%, and accuracy of 94%. For poor clinical outcome, APACHE-II at admission and the NPi produced the best FFT with a sensitivity of 100%, specificity of 50%, and accuracy of 70%. CONCLUSIONS SAH is associated with high patient morbidity and mortality. The initial risk stratification of patients with SAH most in need of supportive care is typically performed through the utilization of clinical severity scores such as Hunt and Hess and APACHEII (8). Such scoring systems have limitations, including subjective elements and poor correlation with the actual degree of nervous system injury (9). More quantitative neuromonitoring techniques such as direct intracranial pressure measurement are available but subject patients to additional risks. Most importantly, no single approach to risk stratification has demonstrably improved outcomes of patients with SAH (9,10). For these reasons, noninvasive biomarkers for risk of death and secondary brain injury in SAH would be valuable. To evaluate NPi as a possible prognostic marker in patients with SAH, we included several clinical variables in the development of FFTs which sequentially order the most useful variables or “cues” to optimize the prediction of the outcome. In an FFT, every cue has 2 branches with at least a branch exiting to a prediction (11); the final cue in the sequence has 2 exits such that the tree always yields a decision. FFTs allow for optimized clinical decision-making based on limited clinical information, such as the initial assessment and risk stratification of patients with SAH. There are several limitations to this study including its retrospective design, variable timing of NPi assessment All analyses used R v. 3.5.1 (The R Foundation; rproject. org). Means and medians were calculated for NPi, age, and APACHE-II score. Frequencies and 95% confidence intervals were calculated for death, poor outcome, and all remaining covariates of interest. Groups were compared using the Fisher exact test. Predictive models were developed with default settings in the FFTrees package v. 1.4.0 (maximizing balanced accuracy). RESULTS Fig. 1 shows study recruitment and participant exposure status. The median age was 53 years (interquartile range [IQR] 40–59 years). The median APACHE-II score was 7 (IQR 5–11), and 9 patients (17%) had APACHE-II scores .12. Twenty patients (38%) were male, and 10 Sharma et al: J Neuro-Ophthalmol 2022; 42: 256-259 FIG. 1. Overview of study recruitment procedures and participant exposure status. neuro-ICU, neurological intensive care unit; SAH, acute subarachnoid hemorrhage; NPi, neurological pupil index. 257 Copyright © North American Neuro-Ophthalmology Society. Unauthorized reproduction of this article is prohibited. Original Contribution There is thus no consensus on the particular clinical scenario(s) in which the technology should be used and how NPi values should be incorporated into clinical decision-making. It should be emphasized that the accuracy of the NPi may be affected by a number of factors (e.g., concurrent medications [including opioids], patient movement, periorbital edema, and ocular/facial injuries (1)), and thus, pupillometry must be considered in the broader clinical picture. In our study, the most accurate FFTs for both patient death and poor clinical outcome included the NPi, indicating its potential prognostic value in the management of patients with acute SAH. Although our results do not establish a definitive association between initial NPi and adverse patient outcomes, they add to the emerging body of literature suggesting NPi is of prognostic value in patients with acute SAH. FIG. 2. A. A fast-and-frugal tree incorporating patient age, mean initial Neurological Pupil index in the right eye, and mean initial Neurological Pupil index in the left eye yielded a sensitivity of 100%, specificity of 94%, and accuracy of 94% for predicting the clinical outcome of patient death. The model included an age of 39 years, Neurological Pupil index in the left eye of 4.5, and Neurological Pupil index in the right eye of 4.5. B. A fast-and-frugal tree incorporating APACHE-II score, mean initial Neurological Pupil index in the right eye, and mean initial Neurological Pupil index in the left eye yielded a sensitivity of 100%, specificity of 50%, and accuracy of 70% for predicting our composite outcome of “poor clinical outcomes.” The model included an APACHE-II score of 6, Neurological Pupil index in the left eye of 4.1, and Neurological Pupil index in the right eye of 4.2. APACHEII, Acute Physiology and Chronic Health Evaluation II score; NPi, Neurological Pupil index; OD, right eye; OS, left eye. (based on the requirement of the treating physician to deem patients clinically stable), few deaths, and overall relatively few enrolled subjects. Ultimately, our results may suggest an inverse association between the NPi and poorer clinical outcomes but do not aim to statistically establish one, as could be achieved using a regression model. Yet, our results show NPi may be a useful tool in the risk stratification of patients with SAH. This could be more definitively explored in a prospective study with broader patient enrollment, and such a study may be forthcoming (12). Noninvasive biomarkers for critically ill patients with neurological impairment are needed, and NPi may be a useful one. However, there remain no established clinical guidelines for the use of pupillometry in clinical practice. 258 STATEMENT OF AUTHORSHIP Category 1: a. Conception and design: P. S. Garza, V. Biousse, O. B. Samuels, N. J. Newman, and B. B. Bruce; b. Acquisition of data: R. A. Sharma, P. S. Garza, V. Biousse, O. B. Samuels, N. J. Newman, and B. B. Bruce; c. Analysis and interpretation of data: R. A. Sharma, P. S. Garza, V. Biousse, O. B. Samuels, N. J. Newman, and B. B. Bruce. Category 2: a. Drafting the manuscript: R. A. Sharma, P. S. Garza, and B. B. Bruce; b. Revising it for intellectual content: V. Biousse, O. B. Samuels, and N. J. Newman. Category 3: a. Final approval of the completed manuscript: R. A. Sharma, P. S. Garza, V. Biousse, O. B. Samuels, N. J. Newman, and B. B. Bruce. REFERENCES 1. Phillips SS, Mueller CM, Nogueira RG, Khalifa YM. A systematic review assessing the current state of automated pupillometry in the NeuroICU. Neurocrit Care. 2019;31:142–161. 2. NeurOptics—the Leader in the Science of Pupillometry [NeurOptics website]. Available at: https://neuroptics.com. Accessed June 1, 2021. 3. Jahns FP, Miroz JP, Messerer M, Daniel RT, Taccone FS, Eckert P, Oddo M. Quantitative pupillometry for the monitoring of intracranial hypertension in patients with severe traumatic brain injury. Crit Care. 2019;23:155. 4. Aoun SG, Stutzman SE, Vo PN, El Ahmadieh TY, Osman M, Neeley O, Plitt A, Caruso JP, Aiyagari V, Atem F, Welch BG, White JA, Batjer HH, Olson DM. Detection of delayed cerebral ischemia using objective pupillometry in patients with aneurysmal subarachnoid hemorrhage. J Neurosurg. 2019;132:27–32. 5. Natzeder S, Mack DJ, Maissen G, Strässle C, Keller E, Muroi C. Portable infrared pupillometer in patients with subarachnoid hemorrhage: prognostic value and circadian rhythm of the Neurological Pupil Index (NPi). J Neurosurg Anesthesiol. 2019;31:428–433. 6. Sharma RA, Garza PS, Biousse V, Samuels OB, Newman NJ, Bruce BB. Ocular fundus abnormalities in acute subarachnoid hemorrhage: the FOTO-ICU study. Neurosurgery. 2021;88:278–284. 7. Knaus WA, Draper EA, Wagner DP, Zimmerman JE. Apache II: a severity of disease classification system. Crit Care Med. 1985;13:818–829. 8. Hunt WE, Hess RM. Surgical risk as related to time of intervention in the repair of intracranial aneurysms. J Neurosurg. 1968;28:14–20. Sharma et al: J Neuro-Ophthalmol 2022; 42: 256-259 Copyright © North American Neuro-Ophthalmology Society. Unauthorized reproduction of this article is prohibited. Original Contribution 9. Tjahjadi M, Hernesniemi J. Aneurysmal subarachnoid hemorrhage grading scales: something old, something borrowed, something new, and silver sixpence in our shoes. World Neurosurg. 2015;83:1037–1038. 10. le Roux A, Wallace M. Outcome and cost of aneurysmal subarachnoid hemorrhage. Neurosurg Clin N Am. 2010;21:235–246. Sharma et al: J Neuro-Ophthalmol 2022; 42: 256-259 11. Gigerenzer G, Todd PM. Simple Heuristics that Make Us Smart. New York, NY: Oxford University Press, 1999. 12. Oddo M, Taccone F, Galimberti S, Rebora P, Citerio G; on behalf of the Orange Study Group. Outcome Prognostication of Acute Brain Injury using the Neurological Pupil Index (ORANGE) study: protocol for a prospective, observational, multicentre, international cohort study. BMJ Open. 2021;11:e046948. 259 Copyright © North American Neuro-Ophthalmology Society. Unauthorized reproduction of this article is prohibited.