Correlation of transcranial Doppler based parameters with computed tomography assessed cerebral oedema score in patients with traumatic brain injury: A prospective observational study

ABSTRACT

Background and Aims:

Cerebral oedema and increased intracranial pressure are associated with poor neurological outcomes in traumatic brain injury (TBI). This study aimed to examine the correlation between transcranial doppler (TCD) derived indices and computed tomography assessed cerebral oedema score in patients with TBI.

Methods:

This prospective observational study was conducted between April 2021 and December 2021 after approval from the University Research Ethics Committee (R75/2021) and registration with the clinicaltrials.gov (NCT04834453). Cerebral oedema on computed tomography (CT) imaging of the brain was scored as (0 = no cerebral oedema, 1 = mild cerebral oedema, 2 = moderate cerebral oedema, and 3 = severe cerebral oedema). The daily neurological assessment was performed using Glasgow coma scale score. TCD-based parameters, mean flow velocity (MFV), and pulsatility index (PI) in middle cerebral arteries were simultaneously obtained.

Results:

There was a significant negative correlation between MFV and cerebral oedema score (r = - 0.840, P <.001) and a significant positive correlation between PI and cerebral oedema score (r = 0.825, P <.001) on the seventh day of assessment. Significant differences were noted in MFV [53.17 ± 7.52 cm/s vs. 34.55 ± 3.35 cm/s] and PI [1.02 ± 0.16 vs. 1.46 ± 0.07] in patients with improvement and no improvement in cerebral oedema after seven days of TBI management.

Conclusion:

Bedside assessments of TCD-based parameters of MFV and PI correlate well with CT-based assessment of cerebral oedema in patients with TBI.

INTRODUCTION

Traumatic brain injury (TBI) results in disruption of normal brain function.[1] The degree of cerebral oedema and raised intracranial pressure (ICP) are associated with poor neurological outcomes after TBI. Cerebral oedema is a nonspecific pathological brain swelling occurring after neurological damage and manifests in localized or diffuse patterns.[2] Cerebral oedema is assessed indirectly with surrogate measures such as tissue shifts and structural abnormalities on brain computed tomography (CT) imaging or by an increase in ICP. It is one of the common causes of increased ICP and has been linked to poor prognosis in individuals suffering from TBI, stroke, and other cerebral diseases.[3,4] Assessment of cerebral oedema on brain CT imaging requires patient transportation and carries the risk of hemodynamic instability and radiation exposure. Transcranial Doppler (TCD) sonography is a noninvasive bedside technique for rapid assessment of cerebral blood flow velocities and ICP.[5] In patients with high ICP, mean and diastolic blood flow velocities (MFV and DFV) decrease but pulsatility index (PI) increases.[6] The purpose of this study was to examine if bedside TCD study can be used to assess severity of cerebral oedema in patients with TBI. The primary objective was to determine the relationship of TCD-based parameters of MFV and PI with CT-based cerebral oedema score during the management of TBI.

METHODS

This prospective observational study was performed after approval from the Ain Shams University Research Ethics Committee (REC), Cairo, Egypt (R75/2021) and registration in clinicaltrials.gov (NCT04834453). The study was conducted between April 2021 and December 2021 at the Intensive Care Unit of the hospital, after obtaining a written informed consent from the legal guardians of the patients in accordance with the principles of Declaration of Helsinki.

Patients belonging to both genders, aged 18–60 years, with blunt TBI associated with diffuse brain oedema were enrolled in the study and followed prospectively. Exclusion criteria were pregnancy, focal brain injury, intracerebral hemorrhage, and patients whose blood flow velocities could not be obtained from trans-temporal acoustic bone windows.

The diagnosis of cerebral oedema was based on the CT imaging. The cerebral oedema was scored as follows: 0 = no brain oedema, 1 = mild brain oedema, 2 = moderate brain oedema, and 3 = severe brain oedema. Oedema was defined as mild when the ventricles and cisterns were narrow in the CT in the acute stage; moderate, when in addition to the mild form, the sylvian fissure and cortical gyri were not visualized; and severe, when in addition to mild and moderate, the perimesencephalic cisterns could not be visualized.[7] The neurological examination was performed daily using the Glasgow coma scale (GCS). A trained intensivist performed bilateral TCD for obtaining middle cerebral artery (MCA) MFV calculated as (peak systolic velocities (PSV) + (end diastolic velocities (EDV) × 2))/3 and PI calculated as (PSV- EDV)/MFV, through the temporal cranial window with the patient in the supine position, TCD was performed using (Siemens Healthineers, Erlangen, Germany) phased array probe of 1.1–4.8 MegaHertz frequency [Figure 1]. The intensivist was blinded to the CT findings to reduce bias. Improvement in oedema was defined as a reduction in the CT score. The primary outcome was the correlation between TCD indices (MCA MFV, and PI) and severity of cerebral oedema (measured by CT score).

Figure 1.

Transcranial Doppler of the left middle cerebral artery in a patient with traumatic brain edema

The study sample size was calculated using PASS 11 - Power Analysis and Sample Size software (2011), Number Cruncher Statistical System, LLC. Kaysville, Utah, United States of America. Using the Spearman rho to calculate the correlation between quantitative and ordinal variables, a sample size of at least 32 will achieve a power of 0.80 to detect a correlation coefficient of at least 0.5 with a level of significance of 0.05 when compared to the null hypothesis of zero correlation coefficient.

The information was gathered, updated, coded, and entered into the Statistical Package for Social Science (SPSS) (Released 2015, Version 25.0. Armonk, New York: International Business Machines Corporation, United States of America). The qualitative data were presented as numbers and percentages and the Chi-squared test was used for comparision. Quantitative data with parametric distribution were reported as mean, standard deviation, and range and were compared between two independent groups using the independent t-test and between paired groups using paired t-test. Also, quantitative data with nonparametric distribution were reported as median with interquartile range (IQR) and compared between two independent groups using the Mann-Whitney test and between paired groups using the Wilcoxon Rank test. Spearman correlation coefficient was used to assess the correlation between two quantitative parameters in the same group and receiver operating characteristic curve was used to assess the best cut-off point with its sensitivity, specificity, positive predictive value (PPV), negative predictive value (NPV), and area under the curve of improvement predictors. The confidence interval was set to 95% and the margin of error accepted was set to 5%. So, the P value was considered significant at the level of <0.05.

RESULTS

From April 2021 to December 2021, 48 patients with TBI were eligible for the study, of which 16 patients were excluded: one patient was pregnant, three patients died within 24 hours, six patients had an intracerebral haemorrhage, and temporal acoustic window was not found in six patients. Finally, we enrolled 32 consecutive patients with cerebral oedema after blunt head trauma. The mean age of the study population was 42.47 ± 11.74 years and there were 12 (37.5%) females and 20 (62.5%) males. On day 1, the median (IQR) Acute Physiology and Chronic Health Evaluation (APACHE) II score was 11 (10-12), the median GCS was 12.0 (IQR, 11.0-13.0), the median CT oedema score was 3 (IQR, 2-3), mean MFV was 36.99 ± 5.12 cm/s, and mean PI was 1.44 ± 0.15. On day 7, the median GCS was 13.0 (IQR, 10.0-15.0), the median CT oedema score was 1.0 (IQR, 0.0-1.5), mean MFV was 48.52 ± 10.57 cm/s, and mean PI was 1.13 ± 0.24. When comparing the results between days 1 and 7, GCS score showed no statistically significant difference (P =0.402), while CT oedema score, MFV, and PI were statistically different (P < 0.001) [Table 1]. After treatment, cerebral oedema improved on CT assessment in 24 patients and no improvement was seen in 8 patients. There was a significant difference in MFV (53.17 ± 7.52 cm/s vs. 34.55 ± 3.35 cm/s), PI (1.02 ± 0.16 vs. 1.46 ± 0.07), and GCS score (10 [9.5-11] vs. 14 [12-15]) between patients who showed improvement on CT oedema scores and who did not, P <.001 for all comparisons [Table 2].

Table 1.

GCS score, CT based cerebral oedema score, TCD based MFV, and PI on day 1 and day 7

Parameter		Day 1 N=32	Day 7 N=32	P
		GCS score	Median (IQR)	12 (11-13)	13 (10-15)	0.438≠
MFV cm/s	Mean±SD	36.99±5.12	48.52±10.57	<0.001*
PI	Mean±SD	1.44±0.15	1.13±0.24	<0.001*
CT cerebral oedema score	Median (IQR)	3 (2-3)	1 (0-1.5)	<0.001†

*: Paired t-test; ^†: Wilcoxon Rank test; GCS: Glasgow Coma Scale, CT: Computed tomography, MFV: mean flow velocity, PI: pulsatility index, SD: standard deviation, n: number

Table 2.

Comparison of GCS score, MFV, and PI in patients who showed improvement in CT cerebral oedema score on day 7

Day 7		No improvement n=8	Improvement n24	P
		GCS score	Median (IQR)	10 (9.5-11)	14 (12-15)	<0.001†
MFV cm/s	Mean±SD	34.55±3.35	53.17±7.52	<0.001*
PI	Mean±SD	1.46±0.07	1.02±0.16	<0.001*

*: Paired t-test; ^†: Wilcoxon Rank test; GCS: Glasgow Coma Scale, MFV: mean flow velocity, PI: pulsatility index, SD: standard deviation, IQR: interquartile range, n: number

Agreement of TCD-based MFV and PI with CT-based cerebral oedema score improvement was as follows: For MFV, when the cutoff was >40.6 cm/s; sensitivity was 95.83%, specificity was 100%, PPV was 100.0%, and NPV was 80%. For PI, when the cutoff was ≤1.34; sensitivity was 91.67%, specificity was 100%, PPV was 100.0%, and NPV was 80.9% [Figure 2 and Table 3]. On day 7, there was a significant negative correlation between MFV and cerebral oedema score (r = -0.840, P <. 0.001) and significant positive correlation between PI and cerebral oedema score (r = 0.825, P <.001) [Figure 3].

Figure 2.

Agreement for GCS, MFV, and PI with CT score as a predictor for cerebral oedema improvement on day 7. GCS, Glasgow Coma Scale; MFV, Mean Flow Velocity; PI, Pulsatility Index; CT, Computed tomography

Table 3.

Agreement for GCS score, MFV, and PI with CT cerebral oedema score improvement on day 7

Day 7	AUC	Cut off value	Sensitivity	Specificity	PPV	NPV
MFV cm/s	0.971	>40.6	91.67	100.0	100.0	80.0
PI	0.997	≤1.34	95.83	100.0	100.0	88.9

AUC: Area under the curve, PPV: positive predictive value, NPV: negative predictive value, GCS: Glasgow Coma Scale, MFV: Mean Flow Velocity, PI: Pulsatility Index, CT: Computed tomography

Figure 3.

(a) Correlation between MFV and CT cerebral oedema score on day 7. (b) Correlation between PI and CT cerebral oedema score on day 7. MFV, Mean Flow Velocity; PI, Pulsatility Index; CT, Computed Tomography

DISCUSSION

The TCD sonography is a low-cost, noninvasive, bedside technique for evaluating cerebral hemodynamics.[8] The two common TCD assessments performed by insonating the MCA through the temporal window are MFV and PI.[9]

TCD sonography has been used to evaluate cerebral blood flow velocity in clinical practice since 1982.[10] Apart from assessment of normal cerebral physiology, it is often used to assess pathological changes after subarachnoid hemorrhage, TBI, acute ischemic stroke, and brain death both in adults and children.[11-13]

The European Society of Intensive Care Medicine in 2021 recommended (weak recommendation) B-mode transcranial color-coded duplex (TCCD) assessment of MCA as a basic skill for the analysis of qualitative waveform and to measure PI for excluding intracranial hypertension impairing cerebral perfusion.[14] Consciousness deteriorates as a result of global cerebral hypoperfusion and ischemic encephalopathy caused by elevated ICP.[15] As per Frank, the most common cause of neurologic deterioration in individuals with massive supratentorial infarctions was decreased cerebral perfusion pressure due to elevated ICP.[15] TCD can replace invasive cerebral blood flow measurement techniques while still providing equivalent prognostic data. With an odds ratio of 3.9 (confidence interval [CI], 1.2–13), a low-flow velocity state defined as an MCA MFV of <35 cm/s within 72 hours following head injury has been demonstrated to predict an adverse outcome at 6 months.[16] Mayer et al.[17] found that when ICP was raised in individuals with intracerebral haematoma volumes >25 mL, MFV and DFV were low, and PI values were high.

In the present study, we observed that TCD-based parameters correlated well with CT-based scores with regards to improvement in cerebral oedema after TBI over a seven-day period.

There have been reports on sonographic monitoring of ICP in the last few years.[18,19] Asil et al.[20] tracked the increase in ICP caused by cerebral oedema with TCD in patients with MCA infarction. They observed that increases in PI were related to the midline shift observed on the third day. Patients in whom maximal PI was >1.5 during the first 10 days had a worse prognosis than those when maximal PI was <1.5. In a study by Bouzat et al.,[21] TCD was performed on 356 patients with a GCS score of 9 to 15 with mild lesions on cerebral CT scan. TCD thresholds had 80% sensitivity (95% CI, 56%-94%) and 79% specificity (95% CI, 74%-83%) for predicting neurologic worsening in patients with PI <1.25 and DFV >25 cm/s in the two MCAs. TCD had negative predictive values of 98% (95% CI, 96%-100%) and positive predictive values of 18% (95% CI, 11%-28%), indicating that it could be effective for in-hospital triage.

TCD being an operator-dependent technology requires detailed three-dimensional knowledge of cerebrovascular anatomy and its variations could be a limitation to the present study. Also, inadequate acoustic windows were prevalent in about 15% of the patients. This may be related to the thickness and porosity of the bone around the acoustic windows and attenuation of the ultrasound energy transmission. TCD measurements are also limited to the large basal arteries and can only provide an index of global rather than local cerebral blood flow velocity. Our findings may not be generalizable to patients with focal TBI lesions. Factors other than cerebral oedema such as raised ICP alone without oedema can also result in low MFV and high PI, and hence, TCD-based assessments may not be specific for cerebral oedema evaluation. Finally, while most centers that manage TBI are likely to have CT imaging facility, TCD equipment may not always be available.

CONCLUSION

TCD-based parameters of MFV and PI correlate well with CT-based scores of cerebral oedema in patients with TBI. The TCD parameters obtained at bedside can help assess improvement in cerebral oedema scores on brain CT imaging without the need for in-hospital transport. Further studies with larger sample sizes are needed to establish the utility of TCD sonography in serial assessments of cerebral oedema and prognostication after TBI.

Declaration of patient consent

The authors certify that they have obtained all appropriate patient consent forms. In the form, the patients’ legal guardians have given their consent for images and other clinical information to be reported in the journal. The patients’ legal guardians understand that their names and initials will not be published and their identities will be concealed.

Author contributions

We Dr M A A/W A/W A hereby declare that the article has not been published or submitted to or accepted for publication in any form in any other journal. I vouch that the authorship of this manuscript will not be contested by anyone whose names are not listed. On acceptance the article will become the copyright of the journal. The manuscript has been read and approved by all the authors. M A A and W A contributed to data collection, study design, and statistical analysis; W A contributed to the medical writing of the manuscript. M A A took the whole responsibility of the integrity of the work. All authors revised the manuscript and approved submission.

Financial support and sponsorship

Nil.

Conflicts of interest

There are no conflicts of interest.