Predicting the Prognosis of Diffuse Axonal Injury Using Automated Pupillometry

doi:10.21203/rs.3.rs-2866927/v1

Background

Although previous studies have reported various predictive indicators of diffuse axonal injury (DAI), a consensus regarding the gold-standard predictive indicator has not yet been reached. The usefulness of automated pupillometry in patients with consciousness disorders has been widely reported; however, there are few reports of its use in patients with DAI. We thus investigated the usefulness of pupillary findings for predicting prognosis in DAI.

Methods

We included patients with a diagnosis of DAI admitted to our center from 1 June 2021 to 30 June 2022. Pupillary findings in both eyes were quantitatively measured by automated pupillometry every 2 hours after admission. We then examined the correlations between automated pupillometry values, patient characteristics, and outcomes such as the Glasgow Outcome Scale Extended (GOSE) score 6 months after injury and the time to follow commands.

Results

Of the 22 included patients, 5 presented oculomotor nerve palsy. Oculomotor nerve palsy was correlated with all outcomes. In contrast, Marshall computed tomography classification, injury severity score, and DAI grade were correlated with few outcomes. Some automated pupillometry values were significantly correlated with GOSE at 6 months after injury, and many values from the first 24 hours of measurement were correlated with the time to follow commands. In general, these results were not affected by adjusting for the sedation period, injury severity score, or Marshall computed tomography classification. A subgroup analysis of patients without oculomotor nerve palsy revealed that many automated pupillometry values from the first 24 hours of measurement were significantly correlated with most outcomes. The cutoff values that differentiated good prognosis (GOSE 5–8) from poor prognosis (GOSE 1–4) were a constriction velocity of 1.43 (area under the curve [AUC] = 0.81 [0.62–1], p = 0.037) and a maximum constriction velocity of 2.345 (AUC = 0.78 [0.58–0.98], p = 0.04). The cutoff values that differentiated the time to follow commands into within 7 days and 8 days or over were a percentage of constriction of 8 (AUC = 0.89 [0.68–1], p = 0.011), a constriction velocity of 0.63 (AUC = 0.92 [0.78–1], p = 0.013), a maximum constriction velocity of 0.855 (AUC = 0.9 [0.74–1], p = 0.017), and an average dilation velocity of 0.175 (AUC = 0.95 [0.86–1], p = 0.018).

Conclusions

Pupillary findings in DAI were a strong predictive indicator of prognosis. Quantitative measurements using automated pupillometry may facilitate the prediction of DAI prognosis.

Diffuse axonal injury

prognosis

reflex

pupillary

Diffuse axonal injury (DAI) is caused by high-energy trauma with acceleration–deceleration forces—especially rotational acceleration—that produce mechanical stress, leading to axonal severance and consequent axonal degeneration [1–4]. Patients with DAI generally present with severe consciousness disorder in the acute phase and higher brain dysfunction in the chronic phase. However, sequelae can vary from consciousness disorder to mild memory disorder and are difficult to predict. Some reports have suggested that clinical features such as magnetic resonance imaging (MRI) findings or the Glasgow Coma Scale (GCS) can predict the prognosis of patients with DAI, or that biomarkers can be predictive factors for DAI prognosis in terms of molecular pathology [1, 5–20].

Automated pupillometry is a non-invasive method for monitoring the neurological condition of patients; it has recently become more frequently used in neurocritical care. Previous studies have demonstrated the utility of automated pupillometry in patients with traumatic brain injury with elevated intracranial pressure or patients after cardiac arrest, as well as in patients with consciousness disorders such as non-convulsive status epilepticus or delirium [21–30]. Nevertheless, although automated pupillometry has been used in a DAI subgroup of patients with traumatic brain injury [25], it has not yet been investigated in patients with DAI only.

In the present research, we quantitatively evaluated pupillary light reflexes in patients with DAI and investigated the feasibility of these values as predictive indicators of DAI prognosis, along with other possible predictive indicators from previous research.

Patients with a diagnosis of DAI who were admitted to the Emergency Medicine and Critical Care Department of Saitama Medical University from 1 June 2021 to 30 June 2022 were retrospectively screened for this study. We did not exclude cases with coexisting traumatic intracranial hemorrhage other than DAI or cases with coexisting multiple traumas in the torso and extremities; however, we did exclude cases with previously known ophthalmic disease. The diagnosis of DAI was conducted using a 3T MRI with fluid-attenuated inversion recovery, diffusion-weighed imaging, and susceptibility-weighed imaging sequences. The classification of DAI grade was also performed using MRI based on the Adams classification and previous studies [6, 10, 31]. We used an NPi-200 Pupillometer (NeurOptics, Laguna Hills, CA, USA) for the quantitative measurement of pupillary light reflex variables. This device measures the following pupillary light reflex variables: pupil size before and after light irradiation, percentage of constriction (CH), constriction velocity (CV), maximum constriction velocity (MCV), average dilation velocity (DV), and latency of constriction. The neurological pupil index is then computed by the device by integrating these seven values using a proprietary algorithm. In each patient, these values were recorded from both eyes every 2 hours after admission for a maximum of 14 days. If patients presented anisocoria on admission despite a lack of previously known ophthalmic disease, we considered that they had oculomotor nerve palsy and continued measuring both eyes. In these cases, an ophthalmologist confirmed the lack of an optic nerve disorder as soon as the patient was able to undergo ophthalmic examination. Except in cases with oculomotor nerve palsy, if the automated pupillometer was unable to track the pupil (because of eye blinks, closures, or eye movements) and was unable to display many of the eight values, the measurements were discarded and repeated.

All patient data on admission were collected from their medical records: age, sex, GCS, Marshall computed tomography (CT) classification [32], existence of abnormal posture, injury severity score (ISS), with or without blood transfusion, hemoglobin (Hb) level, requirement of vasopressor, and existence of oculomotor nerve palsy. If patients needed to be sedated for treatment, we recorded the duration of the sedation period (in which sedative agents and/or opioid were used). The primary outcome was the Glasgow Outcome Scale Extended (GOSE) score at 6 months after injury, which was assessed through face-to-face or telephone interviews with each patient or their family. Secondary outcomes were the time to follow commands, the duration of mechanical ventilation, intensive care unit (ICU) stays, hospital stays, and GCS at the time of hospital discharge. Time to follow commands was defined as the period until the patient was able to obey a simple order with their eyes open. If patients underwent permanent tracheostomy and did not give any signs of speaking at the time of hospital discharge, we assessed their verbal GCS score as 1 and calculated the total score.

For the statistical analysis of correlations between explanatory variables and GOSE or GCS at the time of hospital discharge, ordinal logistic regression and the Mann–Whitney U test were used. Simple regression analysis and the Mann–Whitney U test were used to analyze correlations between explanatory variables and the time to follow commands, ventilator days, ICU stays, and hospital stays. Degrees of correlation were measured using Spearman’s rank correlation coefficient; degrees of difference were measured using the effect size r. Multiple regression analysis and ordinal logistic regression were used together to reduce the influence of confounding factors. We defined GOSE 1–4 as poor prognosis and GOSE 5–8 as good prognosis, and created receiver operating characteristic curves to find the cutoff median values of automated pupillometry that were able to differentiate good from poor prognosis. Because the median values of automated pupillometry during the first 24 hours of measurement were strongly correlated with the time to follow commands, we also used these values to assess the cutoff median values for differentiating the time to follow commands (within 7 days versus 8 days and over) using receiver operating characteristic curves. A value of p < 0.05 was considered significant. Correlations between values of automated pupillometry and outcomes were considered significant when they were observed in both eyes. All statistical analyses were conducted using RStudio, version 2022.12.0-353 (RStudio Inc., Boston, MA, USA; available at: https://rstudio.com/).

Informed consent for this study was obtained in the form of an opt-out choice on the institution’s website prior to the study. The study was approved by the institutional ethics committee board and was performed in accordance with the ethical standards of the 1964 Declaration of Helsinki and its later amendments.

Of the 22 patients included in this study, 11 were men and 11 were women. Craniotomy to remove intracranial hemorrhage was performed in two cases. Age at the time of injury ranged from 12 to 85 years (median 46 years). On admission, the median GCS was 7 (interquartile range [IQR] 6–9.75), median Marshall CT classification was 2 (IQR 2), median ISS was 26 (IQR 21–39), median sedation period was 0 (IQR 0–2.275), and median DAI grade was 2 (IQR 1.25–2). Two patients presented abnormal posture and five presented oculomotor nerve palsy. Eleven patients received a transfusion and six were administered a vasopressor on admission. The median GOSE at 6 months after injury was 6 (IQR 4–8). Medians and IQRs are provided in Table 1.

There were significant correlations between GOSE and sex, GCS on admission, Hb on admission, and oculomotor nerve palsy. The automated pupillometry values of CV, MCV, DV, and latency of constriction were also significantly correlated with the primary outcome (Table 2). The patient characteristics of GCS on admission, abnormal posture, blood transfusion, Hb levels, use of vasopressor on admission, and oculomotor nerve palsy were correlated with many outcomes; oculomotor nerve palsy was the strongest predictive indicator, with significant correlations with all outcomes. No outcomes were correlated with Marshall CT classification, and three and one outcomes were significantly correlated with ISS and DAI grade, respectively. The median automated pupillometry values of CH, CV, MCV, and DV were correlated with many outcomes; of these, the median DV was significantly correlated with all outcomes. This trend was also observed in the median values during the first 24 hours of measurement; these values were correlated with almost all outcomes except GOSE. Notably, the medians of CH, CV, MCV, and DV during the first 24 hours of measurement were strongly correlated with the time to follow commands.

Multiple regression and ordinal logistic analyses were conducted to adjust for the influence of confounding factors on automated pupillometry values that were significantly correlated with outcomes. Marshall CT classification, ISS, and sedation period were used as the explanatory variables (Table 3). When adjusted for ISS, the significant correlations from the simple regression analysis were not observed in two of the eight median values for the primary outcome, and in approximately 40% (32/82) of the median values for the secondary outcomes. All median values that were correlated with the primary outcome in the simple regression analysis remained significantly correlated when adjusted for the sedation period; however, in 7.3% (6/82) of the median values, the significant correlations with secondary outcomes from the simple regression analysis were not observed. The results of the simple regression analysis were not affected when adjusted for Marshall CT classification.

Simple regression analyses were also performed in the same way in the subgroup without oculomotor nerve palsy, because this condition can affect automated pupillometry values (Table 4). Although a significant correlation with the primary outcome was not observed, the median CH, CV, MCV, and DV during the first 24 hours of measurement were significantly correlated with the time to follow commands, ventilator days, ICU stays, and hospital stays.

The cutoff values that differentiated good from poor prognosis, and those that differentiated the time to follow commands into within 7 days and 8 days or over, are described in Table 5. Although just two cutoff values were identified that differentiated good from poor prognosis, five cutoff values were able to differentiate the time to follow commands. Notably, four of these five cutoff values were median values from the first 24 hours of measurement, reflecting their strong correlation with the time to follow commands. The median values of neurological pupil index were not a useful predictive indicator of DAI prognosis.

Previous investigations into the prognosis of DAI have reported various predictive indicators, including grade as defined by MRI findings; lesions in the corpus callosum; or GCS, Hb, blood transfusion, or low blood pressure on admission [5–9, 11, 15, 16, 33]. Many studies have investigated the feasibility of DAI grade as a predictive indicator; however, findings have been inconsistent [5, 6, 9–11, 15, 34]. Although some correlations between DAI grade and outcomes were observed in the present study, its significance as a predictive indicator was not sufficiently confirmed. Importantly, we investigated the usefulness of automated pupillometry values as a novel predictive indicator of DAI for the first time.

In the current study, GCS, abnormal posture, blood transfusion, Hb levels, and the use of vasopressor on admission were significantly correlated with many outcomes; these findings are consistent with those of previous reports [6–10, 15, 33]. Notably, oculomotor nerve palsy was a strong predictive indicator that was correlated with all outcomes; this result is consistent with a previous report suggesting that pupillary abnormality is a predictive indicator of DAI prognosis [11, 15]. Additionally, there were significant correlations between automated pupillometry values—especially CH, CV, MCV, and DV—and many outcomes in the present study. These findings suggest the importance of pupil observation in DAI patients.

Oculomotor nerve palsy and sedative agents/opioids may have affected the automated pupillometry values in our study. For example, previous studies have reported that sedative agents, such as propofol or dexmedetomidine, and opioids can affect pupillary constriction and dilation, resulting in changes in automated pupillometry values [35–40]. Although the neurological pupil index is not considered to be particularly affected by these agents, some studies have reported that this index can be affected by anesthesia, especially in critically ill patients [35–38, 41]. Multiple regression and ordinal logistic analyses using the sedation period as an explanatory variable were therefore conducted in the present study to adjust for its effect on automated pupillometry values; these analyses revealed that many of the median values remained significantly correlated with outcomes, independently of the sedation period. In addition, a subgroup analysis was performed to eliminate the effects of oculomotor nerve palsy on automated pupillometry values. In this subgroup analysis, many median values were significantly correlated with the majority of the secondary outcomes. Together, these findings indicate that automated pupillometry values can directly predict the prognosis of DAI patients; these values do not just reflect the effects of oculomotor nerve palsy or sedative agents/opioids.

The coexistence of traumatic intracranial hemorrhage other than DAI or multiple traumas in the torso and extremities may also have affected the results of the present study [33, 42]. We therefore performed multiple regression and ordinal logistic analyses using the Marshall CT classification and ISS as explanatory variables. The median automated pupillometry values remained significantly correlated with outcomes independently of Marshall CT classification. In contrast, approximately 40% of the results of the simple regression analysis were affected by adjustment for ISS. One possible explanation for this finding is that the analysis may have been affected by the tendency of patients with DAI to also have high ISS, because the ISS is calculated using the Abbreviated Injury Scale score in the whole body, and a high Abbreviated Injury Scale score is set for DAI.

Together, the findings of our study suggest that automated pupillometry values are a useful predictive indicator of DAI prognosis, independent of coexisting oculomotor nerve palsy or other trauma, including intracranial hemorrhage. In addition, the strong correlations between the median automated pupillometry values during the first 24 hours of measurement and the time to follow commands, ventilator days, and ICU stays suggests that automated pupillometry may also be useful for predicting the prognosis of DAI patients in early stages of treatment, and to facilitate treatment determination for coexisting trauma in other body parts.

The main limitation of the present study was the low number of cases, which led to reduced statistical power. Moreover, the inclusion of patients with multiple traumas (including to the head), and not just patients with “pure” DAI, may have produced various confounding factors. Additionally, although many of the median automated pupillometry values were correlated with outcomes even after adjusting for confounding factors using multiple regression and ordinal logistic analyses, the possibility remains that unexpected confounding factors still exist. These limitations should be considered carefully when interpreting the results of our study.

We investigated many known predictive factors of DAI prognosis, along with a novel factor—automated pupillometry value—whose significance in DAI had not previously been examined. Although oculomotor nerve palsy strongly influenced the prognosis of DAI, the present study revealed that pupillary findings may also directly reflect the prognosis of DAI. The quantitative measurement of pupillary findings in DAI using automated pupillometry may facilitate the prediction of DAI prognosis and the subsequent development of appropriate treatment strategies.

Contributors:

Conception and design: MM and SY. Acquisition of data: MM and SY. Statistical analysis and interpretation of data: MM, SY, and MS. Drafting the article: MM and SY. Critically revising the article: MS. Reviewing submitted version of manuscript: MS. Study supervision: MS.

Patient consent for publication

For this type of study, formal consent is not required.

Data availability statement

The datasets used and/or analyzed during the current study are available from the corresponding author on reasonable request.

Acknowledgements

We thank Bronwen Gardner, PhD, from Edanz (https://jp.edanz.com/ac) for editing a draft of this manuscript.

Details

We confirm that the manuscript complies with all instructions to authors.
We confirm that authorship requirements have been met and the final manuscript was approved by all authors.
We confirm that this manuscript has not been published elsewhere and is not under consideration by another journal.
We confirm that the present study was performed in accordance with the ethical standards of the 1964 Declaration of Helsinki and its later amendments, and was approved by the institutional ethics committee board. Informed consent was obtained in the form of an opt-out choice on the institute’s website prior to the study.
The authors declare that they have no conflicts of interest.
We confirm completion of STROBE checklists.

This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.

Ma J, Zhang K, Wang Z, Chen G. Progress of research on diffuse axonal injury after traumatic brain injury. Neural Plast. 2016;2016. doi:10.1155/2016/9746313
Frati A, Cerretani D, Fiaschi AI, et al. Diffuse axonal injury and oxidative stress: A comprehensive review. Int J Mol Sci. 2017;18(12):1-20. doi:10.3390/ijms18122600
Hill CS, Coleman MP, Menon DK. Traumatic Axonal Injury: Mechanisms and Translational Opportunities. Trends Neurosci. 2016;39(5):311-324. doi:10.1016/j.tins.2016.03.002
Johnson VE, Stewart W, Smith DH. Axonal pathology in traumatic brain injury. Exp Neurol. 2013;246:35-43. doi:10.1016/j.expneurol.2012.01.013
van Eijck MM, Schoonman GG, van der Naalt J, de Vries J, Roks G. Diffuse axonal injury after traumatic brain injury is a prognostic factor for functional outcome: a systematic review and meta-analysis. Brain Inj. 2018;32(4):395-402. doi:10.1080/02699052.2018.1429018
Lee H-J, Sun H-W, Lee J-S, Choi N-J, Jung Y-J, Hong S-K. Clinical Outcomes of Diffuse Axonal Injury According to Radiological Grade. J Trauma Inj. 2018;31(2):51-57. doi:10.20408/jti.2018.31.2.51
Zheng WB, Liu GR, Kong KM, Wu RH. Coma duration prediction in diffuse axonal injury: Analyses of apparent diffusion coefficient and clinical prognostic factors. Annu Int Conf IEEE Eng Med Biol - Proc. 2006:1052-1055. doi:10.1109/IEMBS.2006.259627
Zheng WB, Liu GR, Li LP, Wu RH. Prediction of recovery from a post-traumatic coma state by diffusion-weighted imaging (DWI) in patients with diffuse axonal injury. Neuroradiology. 2007;49(3):271-279. doi:10.1007/s00234-006-0187-8
Kim M, Ahn JS, Park W, et al. Diffuse axonal injury (DAI) in moderate to severe head injured patients: Pure DAI vs. non-pure DAI. Clin Neurol Neurosurg. 2018;171(March):116-123. doi:10.1016/j.clineuro.2018.06.011
Lohani S, Bhandari S, Ranabhat K, Agrawal P. Does Diffuse Axonal Injury MRI Grade Really Correlate with Functional Outcome? World Neurosurg. 2020;135:e424-e426. doi:10.1016/j.wneu.2019.12.017
Matsukawa H, Shinoda M, Fujii M, et al. Genu of corpus callosum as a prognostic factor in diffuse axonal injury: Clinical article. J Neurosurg. 2011;115(5):1019-1024. doi:10.3171/2011.6.JNS11513
Sandhu S, Soule E, Fiester P, et al. Brainstem diffuse axonal injury and consciousness. J Clin Imaging Sci. 2019;9(32):1-10. doi:10.25259/JCIS-11-2019
Moen KG, Brezova V, Skandsen T, Håberg AK, Folvik M, Vik A. Traumatic axonal injury: the prognostic value of lesion load in corpus callosum, brain stem, and thalamus in different magnetic resonance imaging sequences. J Neurotrauma. 2014;31(17):1486-1496. doi:10.1089/neu.2013.3258
Tong KA, Ashwal S, Holshouser BA, et al. Diffuse axonal injury in children: Clinical correlation with hemorrhagic lesions. Ann Neurol. 2004;56(1):36-50. doi:10.1002/ana.20123
Van Eijck M, Van Der Naalt J, De Jongh M, et al. Patients with Diffuse Axonal Injury Can Recover to a Favorable Long-Term Functional and Quality of Life Outcome. J Neurotrauma. 2018;35(20):2357-2364. doi:10.1089/neu.2018.5650
Abu Hamdeh S, Marklund N, Lannsjö M, et al. Extended Anatomical Grading in Diffuse Axonal Injury Using MRI: Hemorrhagic Lesions in the Substantia Nigra and Mesencephalic Tegmentum Indicate Poor Long-Term Outcome. J Neurotrauma. 2017;34(2):341-352. doi:10.1089/neu.2016.4426
Izzy S, Mazwi NL, Martinez S, et al. Revisiting Grade 3 Diffuse Axonal Injury: Not All Brainstem Microbleeds are Prognostically Equal. Neurocrit Care. 2017;27:199-207. doi:10.1007/s12028-017-0399-2
Catharine VL, Helena V, Eva G, et al. Is diffuse axonal injury on susceptibility weighted imaging a biomarker for executive functioning in adolescents with traumatic brain injury? Eur J Paediatr Neurol. 2019;23(3):525-536. doi:10.1016/j.ejpn.2019.04.003
Li J, Li XY, Feng DF, Pan DC. Biomarkers associated with diffuse traumatic axonal injury: Exploring pathogenesis, early diagnosis, and prognosis. J Trauma. 2010;69(6):1610-1618. doi:10.1097/TA.0b013e3181f5a9ed
Tomita K, Nakada Taki, Oshima T, Motoshima T, Kawaguchi R, Oda S. Tau protein as a diagnostic marker for diffuse axonal injury. PLoS One. 2019;14(3):1-9. doi:10.1371/journal.pone.0214381
Pansell J, Hack R, Rudberg P, Bell M, Cooray C. Can Quantitative Pupillometry be used to Screen for Elevated Intracranial Pressure? A Retrospective Cohort Study. Neurocrit Care. 2022;37(2):531-537. doi:10.1007/s12028-022-01518-y
Traylor JI, El Ahmadieh TY, Bedros NM, et al. Quantitative pupillometry in patients with traumatic brain injury and loss of consciousness: A prospective pilot study. J Clin Neurosci. 2021;91:88-92. doi:10.1016/j.jocn.2021.06.044
Stevens AR, Su Z, Toman E, Belli A, Davies D. Optical pupillometry in traumatic brain injury: neurological pupil index and its relationship with intracranial pressure through significant event analysis. Brain Inj. 2019;33(8):1032-1038. doi:10.1080/02699052.2019.1605621
Jahns FP, Miroz JP, Messerer M, et al. Quantitative pupillometry for the monitoring of intracranial hypertension in patients with severe traumatic brain injury. Crit Care. 2019;23(1):1-9. doi:10.1186/s13054-019-2436-3
El Ahmadieh TY, Bedros N, Stutzman SE, et al. Automated Pupillometry as a Triage and Assessment Tool in Patients with Traumatic Brain Injury. World Neurosurg. 2021;145:e163-e169. doi:10.1016/j.wneu.2020.09.152
Oddo M, Sandroni C, Citerio G, et al. Quantitative versus standard pupillary light reflex for early prognostication in comatose cardiac arrest patients: an international prospective multicenter double-blinded study. Intensive Care Med. 2018;44(12):2102-2111. doi:10.1007/s00134-018-5448-6
Riker RR, Sawyer ME, Fischman VG, et al. Neurological Pupil Index and Pupillary Light Reflex by Pupillometry Predict Outcome Early After Cardiac Arrest. Neurocrit Care. 2020;32(1):152-161. doi:10.1007/s12028-019-00717-4
Favre E, Bernini A, Morelli P, Pasquier J, Miroz J, Abed-maillard S. Neuromonitoring of delirium with quantitative pupillometry in sedated mechanically ventilated critically ill patients. Crit Care. 2020;24(1):66-73. doi:10.1186/s13054-020-2796-8.
Care N, Godau J, Bharad K, et al. Automated Pupillometry for Assessment of Treatment Success in Nonconvulsive Status Epilepticus. Neurocrit Care. 2022;36(1):148-156. doi:10.1007/s12028-021-01273-6
Godau J, Bierwirth C, Rösche J, Bösel J. Quantitative Infrared Pupillometry in Nonconvulsive Status Epilepticus. Neurocrit Care. 2021;35(1):113-120. doi:10.1007/s12028-020-01149-1
Adams JH, Doyle D, Ford I, Gennarelli TA, Graham DI, Mclellan DR. Diffuse axonal injury in head injury: Definition, diagnosis and grading. Histopathology. 1989;15(1):49-59. doi:10.1111/j.1365-2559.1989.tb03040.x
Marshall LF, Marshall SB, Klauber MR, et al. A new classification of head injury based on computerized tomography. J Neurosurg. 1991;75(SUPPL.): S14–S20. doi:10.3171/sup.1991.75.1s.0s14
Chelly H, Chaari A, Daoud E, et al. Diffuse Axonal Injury in Patients With Head Injuries: An Epidemiologic and Prognosis Study of 124 Cases. J Trauma. 2011;71(4) :838-46. doi:10.1097/TA.0b013e3182127baa
Hedaoo K, Rajpurohit V, Garg S, Thanvi S, Nagocha VB. GCS score and MRI grading does not predicts the outcome in dai patients - a prospective study. Rom Neurosurg. 2018;32(2): 314-321. doi:10.2478/romneu-2018-0039
Shirozu K, Setoguchi H, Tokuda K, Karashima Y. The effects of anesthetic agents on pupillary function during general anesthesia using the automated infrared quantitative pupillometer. J Clin Monit Comput. 2017;31(2):291-296. doi:10.1007/s10877-016-9839-3
Wang Pong Chan, Brenton R Prescott, Megan E Barra, et al. Dexmedetomidine and Other Analgosedatives Alter Pupil Characteristics in Critically Ill Patients. Crit Care Explor. 2022;4(5):e0691. doi: 10.1097/CCE.0000000000000691.
Mckay RE, Larson MD. Detection of opioid effect with pupillometry. Auton Neurosci. 2021;235:102869. doi:10.1016/j.autneu.2021.102869.
Tosi F, Gatto A, Capossela L, Ferretti S, Mancino A. Role of the pupillometer in the assessment of pain in the sedation of pediatric patients. Eur Rev Med Pharmacol Sci. 2021;25(20):6349-6355. doi:10.26355/eurrev_202110_27008.
Larson MD, Berry PD, May J, et al. Latency of pupillary reflex dilation during general anesthesia. J Appl Physiol (1985). 2004;97(2):725-30. doi:10.1152/japplphysiol.00098.2004.
Kostas N Fountas, Eftychia Z Kapsalaki, Theofilos G Machinis, et al. Clinical Implications of Quantitative Infrared Pupillometry in Neurosurgical Patients. Neurocrit Care. 2006;5(1):55-60. doi:10.1385/NCC:5:1:55.
Eliana L Jolkovsky, Felix E Fernandez-Penny, Maya Alexis, et al. Impact of acute intoxication on quantitative pupillometry assessment in the emergency department. J Am Coll Emerg Physicians Open. 2022;3(5):e12825. doi:10.1002/emp2.12825.
Stephen S Humble, Laura D Wilson, Li Wang, et al. Prognosis of Diffuse Axonal Injury with Traumatic Brain Injury. J Trauma Acute Care Surg. 2018;85(1):155-159. doi:10.1097/TA.0000000000001852.

Table 1. Characteristics and the outcome of patients, and their correlations.

			Outcome
			GOSE	Time to follow commands (days)	Ventilator days	ICU stays (days)	Hospital stays (days)	GCS at the time of discharge
		Median (IQR)	6 (4-8)	3 (2-29.75)	2.5 (0-5.75)	6.5 (3-17)	34.5 (28-68.5)	15 (14-15)
Age	12-85	46 (24.5-62.5)
Sex	Male: 11 Female: 11		*
GCS on admission	3-13	7 (6-9.75)	*	*	*	*		*
Marshall CT classification	1-6	2 (2)
Abnormal posture	Yes: 2 No: 20			*	*	*		*
ISS	9-75	26 (21-39)		*	*		*
Transfusion on admission	Yes: 11 No: 11			*	*	*	*
Hemoglobin on admission (mg/dL)	8.3-16.6	13.35 (11.28-14.98)	*	*				*
Use of vasopressor	Yes: 6 No: 16			*	*	*	*
Sedation period(days)	0-16	0 (0-2.75)			*		*
DAI grade	1-3	2 (1.25-2)				*
Oculomotor nerve palsy	Yes: 5 No: 17		*	*	*	*	*	*

GCS; Glasgow Coma Scale, GOSE; Glasgow Outcome Scale-Extended, ICU; intensive care unit, IQR; interquartile range, ISS; injury severity score, DAI; diffuse axonal injury, *; p < 0.05

Table 2. The correlations between the outcomes and the median value of automated pupillometry.

		Outcome
	Value of automated pupillometry	GOSE	Time to follow commands	Ventilator days	ICU stays	Hospital stays	GCS at the time of discharge
Total period	NPi		*			*
	Size
	Min
	CH		*	*	*	*
	CV	*	*		*	*	*
	MCV	*		*	*	*	*
	DV	*	*	*	*	*	*
	Lat	*
24 hours after admission	NPi		*	*	*	*
	Size
	Min
	CH		*	*	*	*	*
	CV		*	*	*	*	*
	MCV		*	*	*	*
	DV		*	*	*	*
	Lat

CH; percentage of constriction, CV; constriction velocity, DV; average dilation velocity, GCS; Glasgow Coma Scale, GOSE; Glasgow Outcome Scale-Extended, ICU; intensive care unit, Lat; latency of constriction, MCV; maximum constriction velocity, Min; pupil size after light irradiation, NPi; neurological pupil, Size; pupil size before light irradiation, *; p < 0.05

Table 3. The number of automated pupillometry values significantly correlated with the outcomes and the number of those which lose significant correlation with the outcomes after adjusting using multiple regression analysis and ordinal logistic regression.

	Outcome
Explanatory variables	GOSE	Time to follow commands	Ventilator days	ICU stays	Hospital stays	GCS at the time of discharge
Marshall CT classification	0/8	0/18	0/16	0/18	0/20	0/10
ISS	2/8	6/18	6/16	8/18	6/20	6/10
Sedation periods	0/8	1/18	2/16	0/18	0/20	3/10

CT; computed tomography, GCS; Glasgow Coma Scale, GOSE; Glasgow Outcome Scale-Extended, ICU; intensive care unit, ISS; injury severity score.

Table 4. The correlations between the outcomes of patients without oculomotor nerve palsy and the median values of automated pupillometry.

		Outcome
	Value of automated pupillometry	GOSE	Time to follow commands	Ventilator days	ICU stays	Hospital stays	GCS at the time of discharge
Total period	NPi
	Size
	Min
	CH					*
	CV
	MCV
	DV
	Lat
24 hours after admission	NPi					*
	Size
	Min
	CH		*	*	*	*
	CV		*	*	*	*
	MCV		*	*	*	*
	DV		*	*	*	*
	Lat

CH; percentage of constriction, CV; constriction velocity, DV; average dilation velocity, GCS; Glasgow Coma Scale, GOSE; Glasgow Outcome Scale-Extended, ICU; intensive care unit, Lat; latency of constriction, MCV; maximum constriction velocity, Min; pupil size after light irradiation, NPi; neurological pupil, Size; pupil size before light irradiation, *; p < 0.05

Table 5. The cutoff values which differentiated poor prognosis (Glasgow Outcome Scale-Extended 1-4) and good prognosis (Glasgow Outcome Scale-Extended 5-8), and those which differentiated the time to follow commands within 7 days and 8 days or over.

	Value of automated pupillometry	Cutoff value of GOSE (AUC (95% confidence interval), p value)	Cutoff value of time to follow commands (AUC (95% confidence interval), p value)
Total period	CH	N.A.	N.A.
	CV	1.43 (0.81(0.62-1), p = 0.037)	N.A.
	MCV	2.345 (0.78 (0.58-0.98), p = 0.04)	N.A.
	DV	N.A.	0.848 (0.74 (0.68-1), p = 0.018)
24 hours after admission	NPi	N.A.	N.A.
	CH	N.A.	8 (0.89 (0.68-1), p = 0.011)
	CV	N.A.	0.63 (0.92 (0.78-1), p = 0.013)
	MCV	N.A.	0.855 (0.9 (0.74-1), p = 0.017)
	DV	N.A.	0.175 (0.95 (0.86-1), p = 0.018)

CH; percentage of constriction, CV; constriction velocity, DV; average dilation velocity, GOSE; Glasgow Outcome Scale-Extended, MCV; maximum constriction velocity, N.A.; not available, NPi; neurological pupil index.

STROBEchecklist.doc

Predicting the Prognosis of Diffuse Axonal Injury Using Automated Pupillometry

Status:

Version 1

Abstract

Introduction

Methods

Results

Discussion

Conclusions

Declarations

References

Tables

Supplementary Files

Status:

Version 1