Noninvasive Evaluation of Intracranial Pressure by Transcranial Doppler Ultrasound in Patients With Traumatic Brain Injury


 Background

As a noninvasive monitoring measure, transcranial Doppler ultrasound (TCD) has been widely used to monitor the secondary brain injury in patients with traumatic brain injury (TBI). There are different physiological theories on the noninvasive assessment of intracranial pressure by TCD parameters, including ONSD and PI, which may cause that the change of ONSD and PI is not always synchronous with that of ICP. Therefore, the objective of this study was to investigate the relationship between PI or ONSD and ICP at different levels or in different periods after the operation, and the ability of prediction intracranial hypertension with these parameters in patients with TBI.
Methods

The clinical data of 68 patients with TBI were retrospectively analyzed. The statistical correlation analysis was performed to investigate the relationship between the PI or ONSD and ICP one week after the operation. Besides, the area under the curve (AUC) of ONSD or PI alone or a combination of them was calculated to determine the ability of intracranial hypertension.
Results

1. There was a correlation between ONSD and ICP ≥ 20 mmHg (r = 0.665, p < 0.001), ICP < 20 mmHg (r = 0.358, p = 0.006). The correlation still remained at ONSD ≥ 5 mm (r = 0.644, p < 0.001), but no correlation at ONSD < 5 mm (p = 0.137). 2. There was a strong correlation between PI and ICP at ICP of 15–20 mmHg (r = 0.705, p < 0.001), and ICP ≥ 20 mmHg (r = 0.716, p < 0.001). Nevertheless, it revealed a weak correlation at PI < 1.2 (r = 0.271, p = 0.021), PI ≥ 1.2 (r = 0.350, p = 0.020). In different period after the operation, there was a moderate correlation between ICP and PI on days 3, 4, and 5 (r = 0.508, p < 0.001), a strong correlation on days 6 and 7 after the operation (r = 0.645, p < 0.001). 3. For prediction intracranial hypertension with PI ≥ 1.2 or ONSD ≥ 5 mm alone or a combination of ONSD ≥ 5 mm and PI ≥ 1.2, the AUC value was 0.729 (p < 0.001), 0.900 (p < 0.001), and 0.943 (p < 0.001), respectively.
Conclusion

The correlation between the parameters of TCD, including ONSD and PI, and invasive ICP vary at different levels of ICP and in different periods in patients with TBI post-operation. It could also allow for a more accurate prediction of elevated intracranial pressure with a combination of ONSD ≥ 5 mm and PI ≥ 1.2.


Conclusion
The correlation between the parameters of TCD, including ONSD and PI, and invasive ICP vary at different levels of ICP and in different periods in patients with TBI post-operation. It could also allow for a more accurate prediction of elevated intracranial pressure with a combination of ONSD ≥ 5 mm and PI ≥ 1.2.

Background
Invasive intracranial pressure (ICP) monitoring is the gold standard approach for evaluating intracranial pressure in patients with traumatic brain injury (TBI). ICP -directed therapy, which is recommended by the guidelines for TBI treatment, can reduce the mortality of patients with severe TBI [1,2]. Due to the number of complications of invasive ICP monitoring, including bleeding, iatrogenic infection, bacteria-free operating environment, and zero drift, its application was limited [3]. As a noninvasive monitoring measure, transcranial Doppler ultrasound (TCD) monitoring may help identify cerebral hypoperfusion in patients with TBI, who lack invasive ICP monitoring in community hospitals, emergency departments, or intensive care units. Targetdirected therapy of cerebral blood ow measured by TCD can restore normal cerebral tissue perfusion within a short time, which is conducive to controlling secondary brain insults [4,5].
The optic nerve sheath is the continuation of the cranial dura mater in the optic canal, which is about 2.2-5.0 mm in healthy adults. Increased ICP and enlarged ONSD are independent risk factors of the mortality in patients with severe TBI [6]. Therefore, ONSD can theoretically re ect the levels of ICP. Currently, there is still a lack of uni ed ONSD diagnostic criteria for intracranial hypertension, and the correlation between ONSD and ICP at different ICP levels have not yet been studied.
PI is an essential index for evaluating the compliance and elasticity of distal cerebral arterioles resistance, calculated by PI = (peak systolic velocity -end diastolic velocity) / mean ow velocity. PI is considered to keep pace with invasive ICP measurements when cerebrovascular autoregulation is lost [7]. Therefore, the increase of ICP may lead to the rise of cerebrovascular resistance, the progressive increase of PI, and the decrease of cerebral blood ow. However, different clinical research programs and objectives have different conclusions about the relationship between PI and invasive ICP, and the correlation between them at different ICP levels has not yet been studied.
Additionally, there are different physiological theories on the noninvasive assessment of intracranial pressure by TCD parameters, including ONSD and PI, which may cause that the change of ONSD and PI is not always synchronous with that of ICP. Besides, the majority of previous studies were qualitative rather than quantitative in evaluating intracranial pressure by ONSD or PI, which was of limited value in guiding treatment.
Thus, this research aimed to explore the relationship between the TCD parameters and ICP at different levels or in different periods after the operation, and the ability of prediction intracranial hypertension for ONSD or PI alone or a combination of them in patients with TBI during the rst post-operative week.

Study Design and Patient Enrollment
This study retrospectively analyzed the clinical data of 68 patients with TBI treated at the intensive care unit of the Second A liated Hospital of Air Force Medical University between January 2018 and April 2019.
Patients who presented with closed brain injury, age ≥ 16 years, time from onset to admission ≤ 4 h, craniectomy with invasive ICP monitoring, high-quality image of blood ow spectrum, and optic nerve sheath can be obtained by TCD, and the duration of ICP monitoring ≥ seven days were considered for inclusion. With a history of craniotomy, cerebral ischemic or hemorrhagic stroke, eyeball or optic nerve injury, endovascular stent implantation for a cephalic and cervical vessel, and open head injury were excluded. The baseline characteristics of the patients are shown in Table 1.

Monitoring Protocol
All patients included in the study received neurocritical care management. The operation was performed by an associate chief surgeon with ten years of experience. For the surgery procedure [1], the surgeon removed the extradural hematoma, subdural hematoma, brain contusion during the operation, and the bone ap was removed for the external decompression. was usually placed on the affected side or more severe side of brain injury and monitored continuously for seven days post-operation. When ICP ≥ 20 mmHg, the increase of intracranial pressure should be considered, and the practical strategy should be taken to maintain the ICP < 20 mmHg, CPP 60-70 mmHg.
Three quali ed sonographers (all with > 5 years of experience) conducted TCD at least once a day or whenever necessary. The bilateral middle cerebral artery (MCA), through the temporal ultrasound window, was monitored using a portable 2-MHz pulsed TCD device (LOGIQ E9, General Electric Healthcare, Wauwatosa, WI, USA) [8]. Peak systolic velocity, end diastolic velocity, mean ow velocity, and PI were recorded simultaneously. According to these parameters, abnormal cerebral hemodynamics, including cerebral ischemia, hyperemia, and vasospasm, was diagnosed and corrected.
The ONSD measurement method was done as follows: the width of the optic nerve sheath was measured 3 mm behind the optic disc with a 7.5-10 MHz ultrasound probe. Each eye was measured twice, and the average value was taken for further analysis.
The correlation between PI and ICP one week after the operation Generally, PI was moderately correlated with ICP during this period (r = 0.458, p < 0.001) (Fig. 2). Moreover, when ICP was strati ed, it revealed no correlation between PI at ICP < 15 mmHg (p = 0.366), but a strong correlation at ICP 15-20 mmHg (r = 0.705, p < 0.001) and ICP ≥ 20 mmHg (r = 0.716, p < 0.001); the difference between the two correlation coe cients was not statistically signi cant (Z = -0.078, p = 0.938). Besides, when PI was strati ed, there was a weak correlation between them at PI < 1.2 (r = 0.271, p = 0.021), and PI ≥ 1.2 (r = 0.350, p = 0.020) respectively; we found no signi cant differences between these correlation coe cients (Z = -0.440, p = 0.660). There was no correlation between ICP and PI on days 1 and 2 after the operation (p = 0.705), while a moderate relationship between them was found on days 3, 4, and 5 (r = 0.508, p = 0.001), and a strong relationship on days 6 and 7 post-operation (r = 0.645, p < 0.001); the difference between the two correlation coe cients was not statistically signi cant (Z = -0.784, p = 0.433).
The ability of ONSD or PI alone or a combination of them to predict intracranial hypertension (ICP ≥ 20 mmHg) one week after the operation Bland-Altman analysis of agreement between the different evaluation methods of intracranial pressure, there was no speci c trend to cause the difference between the two observers ( Fig. 3) ( Table 2.).  (Fig. 4), and 0.900 (95%CI: 0.831-0.969, p < 0.001) ( (Fig. 5) respectively; the difference between the two AUC values was statistically signi cant (Z = 2.647, p = 0.008). Further, a combination of ONSD ≥ 5 mm and PI ≥ 1.2 for predicting intracranial hypertension, the AUC value was 0.943 (95% CI: 0.866-1.000, P < 0.001) (Fig. 6). There was not statistically signi cant difference between the AUC value of a combination of ONSD ≥ 5 mm and PI ≥ 1.2 and ONSD ≥ 5 mm alone for predicting intracranial hypertension (Z = -0.819, p = 0.413).

Discussion
It was believed that the increase of ONSD could quickly and accurately re ect the rise of ICP. Maissan et al. [9] reported that when ICP increased to more than 20 mmHg during tracheotomy in 18 patients with TBI, ONSD rapidly expanded to more than 5 mm. If we consider that the longitudinal measurement of the ONSD width of 5.0 mm is the diagnostic threshold for intracranial hypertension [10,11], this study found that ICP and ONSD had a strong correlation (r = 0.679, p < 0.001) during seven days post-operation. The correlation was stronger at intracranial hypertension than that at normal ICP level (r = 0.665 vs. r = 0.358, p = 0.039). Rajajee et al. [12] found that ONSD rapidly increased following the increase of ICP. Nevertheless, when ICP returned to normal levels, the ONSD remained to widen. This study also found a strong correlation between ICP and ONSD ≥ 5 mm (r = 0.644, p < 0.001), and no correlation at ONSD < 5 mm (p = 0.137). Hence, the higher the intracranial pressure corresponds to a stronger correlation between ONSD and ICP. When the intracranial pressure is decreased, the tension of dura in the cranial cavity is released, but the nerve sheath may still be in the state of expansion. So when the intracranial pressure is reduced or less than 20 mmHg, ONSD may not allow for the accurate evaluation of the ICP for a weak correlation between them. This conclusion suggested that the therapeutic measures based on the decrease of ONSD width might prolong osmotic drug use or other programs for ICP management.
So far, there are different conclusions about the relationship between PI and invasive ICP. Bellner et al. [13] reported that PI was correlated with ICP, when PI > 2.13 or < 1.2, it was deduced ICP > 22 mmHg or < 12 mmHg respectively. Moreover, Prunet et al. [14] found that TCD-PI could accurately and effectively predict intracranial hypertension in patients with TBI: the area under the curve was 0.901, the optimal threshold was 1.35, the sensitivity was 80%, and the speci city was 90%. On the contrary, de Riva et al. [15]argued that TCD-PI could not accurately predict ICP. It was in uenced by cerebral perfusion pressure, heart rate, arterial pressure difference, cerebrovascular resistance, cerebral artery compliance, and cerebral vascular autoregulation function. The formula was put forward: (a1 is the pressure difference between systolic and diastolic pressure, CPPm mean arterial pressure, Ra vascular resistance, Ca vascular compliance, HR heart rate) [16].
In the present study, we found a moderate correlation between ICP and PI on the whole seven days postoperation (r = 0.458, p < 0.001). When the ICP was strati ed, there were no signi cant differences between these correlation coe cients (r = 0.705 vs. r = 0.716, p = 0.938). Furthermore, the intensity difference of correlation coe cient between invasive ICP and PI no matter at PI < 1.2 or PI ≥ 1.2 was also no signi cant differences too (r = 0.271 vs. 0.350, p = 0.660). Additionally, the intensity difference of correlation coe cient between ICP and PI at an early stage or a late-stage post-operation was not statistically signi cant (r = 0.508 vs. r = 0.645, p = 0.433). Therefore, all the ndings above mentioned, con rmed that PI should be regarded as a dynamic trend of ICP, rather than an absolute value of ICP. PI is not a pressure indicator, which may be affected by the severity of secondary brain injury, cerebrovascular autoregulation, intracranial pressure, and other factors [15]. So, we should carefully deduce the variation of ICP based on PI parameter in this case, and similarly, it does not mean that the higher intracranial pressure led to the stronger correlation between invasive ICP and PI.
The regression analysis of ONSD and PI evaluation intracranial hypertension was carried out in the present study. It showed that the AUC value of a combination of ONSD ≥ 5 mm and PI ≥ 1.2 for prediction intracranial hypertension was 0.943. Although there was not a statistically signi cant difference between the AUC value of a combination of ONSD ≥ 5 mm and PI ≥ 1.2 and ONSD ≥ 5 mm alone for predicting intracranial hypertension (p = 0.4119), it was a tendency to enhance the ability to predict intracranial hypertension and helpful for clinicians from qualitative to quantitative assessment of intracranial pressure [17]. Notwithstanding, considering the characteristics of patients and the level of intracranial pressure in this study, we should be comprehensively analysis of the clinical and imaging examination before intervention is taken based on the PI or ONSD.
There are several limitations in the present study that should be pointed out. First, we were not able to overcome the bias of observational research and the small number of patients with TBI included in this study. Second, TCD measurements, including ONSD, were intermittent; invasive parenchymal ICP monitoring was continuous, which may in uence the effectiveness of this study. Third, TCD was performed by different physicians, which may lead to variability in performance and differences in data acquisition. Finally, our results showed a different strength correlation between ONSD and TCD-PI with ICP, respectively, which does not suggest that these indicators would replace invasive ICP monitoring in patients with TBI.

Conclusions
The correlation between the parameters of TCD, including ONSD and PI, and invasive ICP vary at different levels of ICP and in different periods in patients with TBI post-operation. Additionally, it could allow for a more accurate prediction of elevated intracranial pressure with a combination of ONSD ≥ 5 mm and PI ≥ 1.  Scatter plots and linear regression between ICP and ONSD.

Figure 2
Scatter plots and linear regression between ICP and PI.