Abstract
Background and Aims:
In traumatic brain injury (TBI) patients, elevated intracranial pressure leading to brain–lung interactions is well described. This study aimed to find the association between the optic nerve sheath diameter (ONSD) variations and respiratory mechanics, that is, static lung compliance (Cstat), and driving pressure (DP) in TBI patients. The secondary objectives were to observe the association between ONSD and lung ultrasound score (LUS), arterial partial pressure of oxygen and fraction of inspired oxygen ratio (PaO2/FiO2), and clinical pulmonary infection score (CPIS).
Methods:
The prospective observational study, conducted at a tertiary care centre, recruited patients with moderate to severe TBI admitted to the neuro-intensive care unit within 24 h of injury. From day 1 to 5, once a day, ONSD, Cstat, DP, minute ventilation (MV), and LUS were recorded. The association between the ONSD variations and respiratory mechanics, LUS, PaO2/FiO2 ratio, and CPIS was studied using linear mixed-effect models.
Results:
A total of 27 patients were included. No significant association was found between ONSD and Cstat [ipsilateral ONSD: model coefficient (β) = 1.13, P = 0.70; contralateral ONSD β = 2.0, P = 0.46] and driving pressure (ipsilateral ONSD: β = 0.18, P = 0.82; contralateral ONSD: β = 0.29, P = 0.7). Higher MV was observed with higher ONSD (ipsilateral ONSD: β = 1.44, P = 0.033; contralateral ONSD: β = 2.23, P = 0.001). ONSD variations were not associated with LUS, PaO2/FiO2, or CPIS.
Conclusion:
Variations in ONSD did not impact the respiratory mechanics of ventilated acute TBI patients, though patients with higher ONSD were found to have higher MV.
Keywords: Acute lung injury, clinical pulmonary infection score, Glasgow coma scale, intracranial pressure, lung ultrasound score, mechanical ventilation, neuro-intensive care unit, optic nerve sheath diameter, traumatic brain injury
INTRODUCTION
Brain–lung interaction is well described in the literature. However, a very few clinical studies have actually investigated it in acute brain injury (ABI).[1,2] The mechanisms described are inflammation, neurotransmitter release, autonomic system dysfunction, neurogenic pulmonary oedema, and adverse effects of mechanical ventilation.[1] The most common recognised theory is the “blast injury” theory, where a sudden increase in intracranial pressure (ICP) leads to increased sympathetic activity along with a transient increase in intravascular pressure, which in turn leads to disruption of the alveolar-capillary membrane.[1,3] In addition, as per a widely accepted dual hit theory in ABI, first hit corresponds to the adrenergic surge and increased production of systemic inflammatory mediators making the lungs vulnerable to “second hit”, that is, the stress induced by mechanical ventilation.[4] Brain–lung interactions in acutely raised ICP can be studied by correlating the measured ICP with the lung mechanics. Although there are recent studies in traumatic brain injury (TBI) patients assessing the effect of application of positive end expiratory pressure (PEEP) on ICP changes (ΔICP),[5] as well as correlating ΔICP with lung ultrasound score changes (ΔLUS),[6] the evidence for ΔICP, per se, due to the ABI, affecting the lung mechanics, is lacking.
In the present study, we hypothesised that in mechanically ventilated acute moderate to severe TBI patients, changes in optic nerve sheath diameter (ONSD), which is found to be a reliable guide of ICP,[7] would correlate with the changes in respiratory mechanics.
The primary objective of this study was to find the association between the ONSD changes and the static lung compliance (Cstat) and driving pressure (DP) in moderate to severe TBI patients. The secondary objectives were to find the association of ONSD with lung aeration score using lung ultrasound (LUS), subsequent changes in need for oxygenation by seeing a ratio of arterial partial pressure of oxygenation (PaO2) to fractional inspiratory oxygen (FiO2) [PaO2/FiO2], and the occurrence of ventilator associated pneumonia (VAP) as determined by the clinical pulmonary infection score (CPIS).
METHODS
After obtaining the approval from the institute ethics committee, we conducted this prospective observational study from December 2022 to June 2023 (Approval No. NIMH/DO/IEC (BS & NS DIV)/2022), letter dated 18/11/2022. The study was conducted in accordance with the principles of the Declaration of Helsinki, 2013 and Good Clinical Practice guidelines. The study was enlisted with the Clinical Trials Registry-India (vide registration number CTRI/2022/12/048220, available at www.ctri.nic.in/) and abides by the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) guidelines. Written informed consent was obtained from the next of kin of the patients for participation in the study and for the use of patient data for research and educational purposes.
All adult patients, 18–70 years of age, admitted to the neuro-intensive care unit (NICU) within 24 h of TBI with Glasgow coma score (GCS) ≤12, operated or not operated, were eligible. Patients with a history of pre-existing cardiac or lung disease, who sustained lung injuries with TBI or had abnormal initial blood gases done within 6 hours of NICU admission or required initial higher PEEP or PaO2/FiO2 < 200, and who refused to provide informed consent were excluded.
All eligible consecutive patients were recruited after NICU admission and managed as per the standard NICU management protocol. As a standard practice, all patients received a PEEP of 5 cmH2O and intermittent sedation with intravenous morphine boluses. Bilateral ONSD was measured once a day, starting from day 1 to a maximum of day 5. Apart from routine daily measurements, ONSD was measured if any neurological deterioration, that is, a 2-point drop in patient’s GCS, was noted. Ipsilateral ONSD was measured on the side of the injury, and in bilateral injuries, the value from the more severe side of the injury was considered. However, contralateral ONSD denoted the non-injured or relatively lesser injured side. Ventilatory parameters such as Cstat and DP were noted when the patients were on mechanical ventilation. Pressure assist-control mode was the most common primary mode of mechanical ventilation used in all the patients. As decided by the treating intensivist treating intensivists, patients were gradually weaned off from the mechanical ventilation and the settings were changed from assist-control to pressure support ventilation. In patients who were weaned off within 5 days of mechanical ventilation, lung mechanics data were recorded till the time patients received mechanical ventilation.
The ONSD was measured using a linear transducer probe (8–13 MHz) in B-mode (M-Turbo, FUJIFILM Sonosite, Bothell, USA). For ocular insonation, settings were used to achieve the lowest possible thermal and mechanical index according to the “as low as reasonably achievable” (ALARA) principle of ultrasonic evaluation. The ultrasound probe was placed on the closed eye (transverse or sagittal plane) after applying a generous amount of ultrasound gel. The optic nerve was recognised as a linear hypoechoic area arising posteriorly from the hypoechoic eyeball. The ONSD measurements were performed 3 mm posterior to the eye globe, where maximal distension occurs consequent to raised ICP. Space among the two hypoechoic areas (representing optic nerve sheath) was measured as described by Ohle et al.[8] Three measurements were performed for each eye, and the average value was recorded. ONSD ≥6 mm was considered as high ICP.
A bedside LUS examination was performed in the supine position by using M-Turbo, FUJIFILM Sonosite, Bothell, USA, with either a linear or curvilinear probe. For each hemithorax, six regions were examined: from the front to the back, three areas were demarcated by the sternum, anterior, and posterior axillary lines; each of these areas was further divided into two regions (superior and inferior).
Each of the 12 standard regions was scored to calculate LUS score. Score 0: A-lines or two or less B-lines, Score 1: ≥3 well-spaced B-lines, Score 2: coalescent B-lines, and Score 3: tissue-like pattern. All the individual regional scores were added to calculate the final LUS score, which ranges from 0 to 36. LUS >15 is considered as a significant lung change.[9]
Data included patient demographics, clinical parameters like GCS, ONSD, mechanical ventilation variables such as peak airway pressure, plateau pressure, DP, Cstat (measured by inspiratory hold on the ventilator), minute ventilation (MV), tidal volume (TV), PEEP, LUS score, PaO2/FiO2, and end-tidal carbon dioxide (EtCO2). The CPIS score was recorded from day 3 to 7 of NICU admission.
The primary outcome of this study was to observe how the variations in the ONSD over the first 5 days of mechanical ventilation, in moderate to severe TBI patients, affected the respiratory mechanics, specifically, Cstat and DP. The secondary outcomes were to see if the changes in ONSD affected lung aeration, measured by LUS, PaO2/FiO2, and CPIS.
As there is a lack of precedence of similar studies, this pilot study was conducted based on the sample achieved over a fixed duration of 6 months, where all patients meeting the inclusion criteria were included. Data were analysed using R software (R Statistical language, version 4.1.2; R Core Team, Vienna, Austria, 2021). Interval/ordinal scale data were described as medians and interquartile ranges (IQRs), while nominal data were expressed as frequencies and percentages. For observing the effect of ONSD on study variables, linear mixed-effect models were constructed with random intercept by subject and assuming an unstructured covariance structure (using package “lmerTest” for R). The models were constructed for the main effect of ONSD and time predicting study variables. For modelling, ONSD was set as a continuous variable. For the visual representation, we have coloured the points on the graphs and panelled based on dichotomised ONSD, using a cut-off of 6 mm (poor ONSD ≥6 mm and good ONSD <6 mm). P < 0.05 was considered as the level of statistical significance.
RESULTS
A total of 27 patients with TBI were included, as shown in the STROBE diagram [Figure 1]. Among these, 23 (80%) were males with a median age of 35 years (26–52.2) [Table 1]. The most common lesion was subdural haemorrhage, followed by contusion, and all the included patients underwent surgical evacuation. Fifteen patients sustained bilateral injuries. The median [IQR] GCS on day 1 was 6 (4–7) and subsequently during the next 4 days was 6 (4.5–7). At NICU admission, the median [IQR] ONSD of the ipsilateral and contralateral eyes with respect to the site of injury was 6 mm [5.8–6.5] and 6.3 mm [5.9–6.5], respectively. The median [IQR] ipsilateral and contralateral ONSD on day 5 were 6.2 mm (5.9–6.5) and 6.2 mm (5.7–6.2), respectively. Out of 27 patients, 17 had poor ONSD, while 10 patients had a good ONSD with a median [IQR] of 5.75 mm [5.53–5.8] and 6.4 mm [6.3–6.5]. A drop in GCS by more than 2 points was observed in three patients on days 2, 3, and 4, respectively, with the corresponding increase in bilateral ONSD observed in two of three cases (in case 1, ipsilateral ONSD changed from 6.2 to 6.4 mm and contralateral from 6.3 to 6.5, and in case 2, ipsilateral ONSD changed from 6.1 to 7 mm and contralateral 5.9 to 6.2 mm). The median duration of mechanical ventilation in the assist-control mode was about 3 days, and the total duration was 6 days. The clinical information of the patients is described in Table 1.
Figure 1.
Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) flow chart of the study
Table 1.
Demographic and clinical parameters of the studied patients (n=27)
| Variable | Descriptive |
|---|---|
| Age (years) | 35 (26-52.5) |
| Weight (kg) | 71 (68-73) |
| Height (cm) | 158 (155-161.5) |
| GCS Day 1 | 6 (4-7) |
| TBI | |
| Unilateral lesion Bilateral lesion |
12 (44.4%) 15 (55.6%) |
| Duration of assist-control mode of Mechanical Ventilation (days) | 3 (1.5-6.5) |
| Duration of Mechanical Ventilation (days) | 6 (4.5-9) |
| Duration of NICU stay (days) | 9 (6-10) |
| Duration of Hospital stay (days) | 17 (10-22.25) |
| Ipsilateral ONSD (mm) | |
| Day1 Day2 Day3 Day4 Day5 |
6 (5.9-6.5) 6.2 (5.8-6.5) 6.2 (5.9-6.5) 6.1 (5.8-6.5) 6.2 (5.9-6.5) |
| Contralateral ONSD (mm) | |
| Day1 Day2 Day3 Day4 Day5 |
6.3 (5.9-6.5) 6.1 (5.8-6.5) 6.2 (5.9-6.4) 6.1 (5.8-6.5) 6.2 (5.7-6.5) |
Data presented as median (IQR) for quantitative variables and frequency (%) for qualitative variables, IQR=interquartile range, GCS=Glasgow coma score, NICU=Neuro-intensive care unit, ONSD=Optic nerve sheath diameter
On observing the relation of the ipsilateral and contralateral ONSD with the actual Cstat [Figure 2a; Left panel] and predicted Cstat [Figure 2a; Right panel] over the five consecutive days, no significant association was observed between the Cstat and ipsilateral ONSD [model coefficient (β) = 1.13 and P = 0.701], and contralateral ONSD (β = 2.009 and P = 0.469). In the predicted Cstat graph, there was a steady decline in Cstat up to day 3, following which a steady increase till day 5 of mechanical ventilation was noted, though the changes were not statistically significant compared to baseline. We also did not find any association between ONSD and DP (β = 0.18 and P = 0.827) [Figure 2b]. However, an increase in ONSD was associated with an increase in MV and TV [Figure 3]. For each unit increase in ONSD, 1.4 units increase in MV for ipsilateral (P = 0.033) and 2.2 units increase in MV for contralateral ONSD (P = 0.001) were noted.
Figure 2.
Association of static lung compliance (compliance) (a; upper panel) and driving pressure (DP) (b; lower panel) with optic nerve sheath diameter (ONSD). The patients were grouped into good ONSD (<6 mm; green dots) and poor ONSD (≥6 mm; red dots) groups. Actual compliance and DP are depicted on the left graphs, while the predicted are shown in the right graphs over the 5 days for ipsilateral and contralateral ONSD. The table on the right-hand side shows the model coefficients of the correlation between ONSD with compliance and DP and P values
Figure 3.
Association of minute ventilation (MV) (a; upper panel) and tidal volume (TV) (b; lower panel) with optic nerve sheath diameter (ONSD). The patients were grouped into good ONSD (<6 mm; green dots) and poor ONSD (≥6 mm; red dots) groups. Actual MV and TV are depicted on the left graphs, while the predicted are shown in the right graphs over the 5 days for ipsilateral and contralateral ONSD. The table on the right-hand side shows the model coefficients of the correlation between ONSD with MV and TV, and P values
At the same time, an increase in ONSD was associated with a decrease in EtCO2, and the association was significant for the contralateral ONSD (P = 0.008) but not for the ipsilateral ONSD (P = 0.816) [Figure 4a].
Figure 4.
Association of end tidal carbon dioxide (EtCO2) (a; upper panel) and ratio of arterial partial pressure of oxygenation (PaO2) to fractional inspiratory oxygen (FiO2) (b; lower panel) with optic nerve sheath diameter (ONSD). The patients were grouped into good ONSD (<6 mm; green dots) and poor ONSD (≥6 mm; red dots) groups. Actual EtCO2 and PaO2/FiO2 are depicted on the left graphs, while the predicted values are shown in the right graphs over the 5 days for ipsilateral and contralateral ONSD. The table on the right-hand side shows the model coefficients of the correlation between ONSD with EtCO2 and PaO2/FiO2 and P values
No significant association was found between ONSD and LUS score. However, it is noteworthy to mention that an ascending trend of LUS scores, in both, patients with good and poor ONSD, was noted over 5 days [Figure 5], with a significant increase being noted from day 3 onwards. On day 2, PaO2/FiO2 was observed to decrease from baseline, though not statistically significant, and remained reduced for the studied period [Figure 5b]. CPIS was positively associated with ipsilateral ONSD with a trend towards statistical significance (β = 0.47 and P = 0.06). But, no significant association was observed with contralateral ONSD (β = 0.298 and P = 0.221). CPIS was found to increase compared to baseline till day 3 with a reduction thereafter.
Figure 5.
Association of lung ultrasound score (LUS) (a; upper panel) and clinical pulmonary infection score (CPIS) (b; lower panel) with optic nerve sheath diameter (ONSD). The patients were grouped into good ONSD (<6 mm; green dots) and poor ONSD (≥6 mm; red dots) groups. Actual LUS and CPIS are depicted on the left graphs, while the predicted are shown in the right graphs over the 5 days for ipsilateral and contralateral ONSD. The table on the right-hand side shows the model coefficients of the correlation between ONSD with LUS and CPIS, and P values
DISCUSSION
In the present study, we did not find any significant association between ONSD and Cstat or DP. This was probably because clinically significant change in sensorium happened in a very few patients and all patients underwent decompressive surgeries. Despite a thorough literature search, we did not find any similar study exploring this concept. There are studies in TBI patients observing the association between the ICP and respiratory mechanics, but through dynamic PEEP changes. In 15 ABI patients, Robba C et al.[5] observed the effect of PEEP applied at two levels (baseline 5 and when increased to 15 cmH2O) on the changes in computed tomography-based lung recruitment, ΔICP, respiratory system compliance, and so on and correlated ΔICP with these parameters. Following the application of 15 cmH2O PEEP, ΔICP had significant negative correlation with changes in compliance, plateau pressure, and amount of lung recruitment (rho = −0.77, P = 0.0008, rho = −0.54, P = 0.0002, rho = −0.78, P = 0.0006, respectively) [Figure 4b]. However, the baseline lung parameters at a PEEP of 5 cmH2O did not predict the response of ΔICP following higher PEEP application. We did not assess for acute changes in ONSD and in lung dynamics by doing any lung manoeuvres or dynamic PEEP change but rather observed the variables only at a PEEP of 5 cmH2O. As we intended to know how the higher ONSD, assuming it reflects higher ICP, affects the lungs during the acute phase of care.
Altered respiratory patterns are observed in severe TBI patients with raised ICP, where significant compression of upper brainstem and pons leads to a fast and deep neurogenic breathing pattern.[9] This probably explains our observation of higher MV and TV and correspondingly a decreased in EtCO2 in patients with higher ONSD, pointing towards neurogenic hyperventilation. Apart from the impact of ICP on lung mechanics, it is also equally important to determine the subsequent cascading effect on the lungs. We used LUS score and PaO2/FiO2 ratio to assess the impact on the lung aeration and the concomitant changes in the requirement of oxygenation. In severe TBI patients, higher initial ICP was found to have increased incidence of acute lung injury (ALI) (25.59%) and systemic inflammatory response syndrome (SIRS) (60.47%). Significantly higher initial ICP was observed in patients with ALI (P < 0.001), and SIRS in turn increased ALI risk by fourfold.[2] In the present study, irrespective of ONSD size, we observed an increasing trend of LUS with a significant difference from day 3, and a non-significant decrease in PaO2/FiO2 on day 2 onwards. This is also the first study to demonstrate the effect of increased ICP through the ONSD and its impact on the occurrence of VAP as determined by CPIS. However, no direct association was found between these two variables, probably because in the present cohort, none of the patients developed VAP (CPIS >6) during the first 7 days, and ONSD remains an estimate of ICP, and thus may not reflect enough information, especially when done once a day.
The strength of the study is that it is one of the first to explore the concept of brain–lung interactions in acute TBI, with respect to correlating the ICP changes, measured by a non-invasive technique (ONSD), with changes in lung mechanics over a period of 5 days of acute intensive care.
Limitations of this study include a small pilot cohort size; thus, a larger study is desirable to confirm the findings. Furthermore, there are many caveats associated with ONSD, such as its operator dependency, though in our study, only one person did all the measurements; it is not a real-time monitor of ICP; and there are variable cut-off values of ONSD for increased ICP, though we have chosen a higher cut-off of ≥6 mm to define poor ONSD and raised ICP.
CONCLUSION
In this pilot study, we did not notice any association between the ONSD changes and respiratory mechanics assessed by Cstat and DP when measured once a day for 5 consecutive days. A higher MV and TV and lower EtCO2 were observed with a higher ONSD. LUS scores were found to increase with each increasing day of mechanical ventilation, but it did not correlate with ONSD changes.
Presentation at conferences/CMEs and abstract publication
The abstract of the study was presented at World Congress of Anaesthesiologists (WCA), held from 3-7 March 2024 at Singapore and abstract published in Anesthesia & Analgesia 2024;139 (6S2):1173-1174.
Study data availability
De-identified data may be requested with reasonable justification from the authors (email to the corresponding author) and shall be shared upon request.
Disclosure of use of artificial intelligence (AI)-assistive or generative tools
The authors confirm that no AI tools or language models were used in the writing or editing of the manuscript, and no images were manipulated using AI.
Declaration of use of permitted tools
None.
Supplementary material
None.
Conflicts of interest
There are no conflicts of interest.
Acknowledgements
None.
Funding Statement
Nil.
REFERENCES
- 1.Ziaka M, Exadaktylos A. Brain–lung interactions and mechanical ventilation in patients with isolated brain injury. Crit Care. 2021;25:358. doi: 10.1186/s13054-021-03778-0. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Lou M, Chen X, Wang K, Xue Y, Cui D, Xue F. Increased intracranial pressure is associated with the development of acute lung injury following severe traumatic brain injury. Clin Neurol Neurosurg. 2013;115:904–8. doi: 10.1016/j.clineuro.2012.09.001. [DOI] [PubMed] [Google Scholar]
- 3.Theodore J, Robin ED. Pathogenesis of neurogenic pulmonary oedema. Lancet. 1975;306:749–51. doi: 10.1016/s0140-6736(75)90729-1. [DOI] [PubMed] [Google Scholar]
- 4.Mascia L. Acute lung injury in patients with severe brain injury: A double hit model. Neurocrit Care. 2009;11:417–26. doi: 10.1007/s12028-009-9242-8. [DOI] [PubMed] [Google Scholar]
- 5.Robba C, Ball L, Nogas S, Battaglini D, Messina A, Brunetti L, et al. Effects of positive end-expiratory pressure on lung recruitment, respiratory mechanics, and intracranial pressure in mechanically ventilated brain-injured patients. Front Physiol. 2021;12:711273. doi: 10.3389/fphys.2021.711273. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Robba C, Ball L, Battaglini D, Iannuzzi F, Brunetti L, Fiaschi P, et al. Effects of positive end-expiratory pressure on lung ultrasound patterns and their correlation with intracranial pressure in mechanically ventilated brain injured patients. Crit Care. 2022;26:31. doi: 10.1186/s13054-022-03903-7. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Chen L-M, Wang L-J, Hu Y, Jiang X-H, Wang Y-Z, Xing Y-Q. Ultrasonic measurement of optic nerve sheath diameter: A non-invasive surrogate approach for dynamic, real-time evaluation of intracranial pressure. Br J Ophthalmol. 2019;103:437–41. doi: 10.1136/bjophthalmol-2018-312934. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Ohle R, McIsaac SM, Woo MY, Perry JJ. Sonography of the optic nerve sheath diameter for detection of raised intracranial pressure compared to computed tomography: A systematic review and meta-analysis. J Ultrasound Med. 2015;34:1285–94. doi: 10.7863/ultra.34.7.1285. [DOI] [PubMed] [Google Scholar]
- 9.Wijdicks EFM. Fast and deep breathing in acute brain injury. Neurocrit Care. 2024 doi: 10.1007/s12028-024-01970-y. doi: 10.1007/s12028-024-01970-y. [DOI] [PubMed] [Google Scholar]