Prognostic value of intracranial pressure monitoring for the management of hypertensive intracerebral hemorrhage following minimally invasive surgery

Xiao-ru Che, Yong-jie Wang, Hai-yan Zheng Department of Cardiology, Zhejiang Province People's Hospital, Hangzhou, China

2 Department of Cardiology, People's Hospital of Hangzhou Medical College, Hangzhou, China

3 Department of Neurosurgery, the Second Aff liated Hospital of Zhejiang University School of Medicine, Hangzhou, China

4 Department of Neurosurgery, the Fouth Aff liated Hospital of Zhejiang University School of Medicine, Yiwu, China

KEY WORDS: Hypertensive intracerebral hemorrhage; Intracranial pressure; Minimally invasive surgery

INTRODUCTION

Intracerebral hemorrhage (ICH), accounting for 20% of strokes, leads to higher mortality and morbidity rates than its ischemic counterpart. Around 50%-70% of ICH is caused by hypertension, with the older population most commonly involved and peak incidence in winter and spring seasons. It has been estimated that almost 80% of the hypertensive intracerebral hemorrhage (HICH) patients had a bad prognosis.[1]Conservative medical treatment and surgical evacuation are the two options for HICH treatment. The surgical approaches can reduce mass effect, mainly including the conventional decompressive craniectomy, stereotactic puncture, and minimally invasive craniotomy.[2]Careful design of small craniotomy in accordance with the hematoma location allows for its efficient evacuation without significant damage to the intact brain. Therefore, the minimally invasive craniotomy is now increasingly gaining popularity.[3]

Postoperative management is of great importance for the recovery of patients. Re-hemorrhage and uncontrolled brain edema are the key issues causing increased intracranial pressure (ICP) and devastating consequences.[4]One study showed that a good recovery was associated with no ICP monitoring, and it included patients with surgical intervention or medical therapy.[5]But the patients with lower Glasgow Coma Scale (GCS) scores in this study tended to be divided into the ICP monitoring group. It was unusual that the ICP monitoring led to a worse prognosis. Thus, the aim of our retrospective cohort study is to evaluate the effect of ICP monitoring on the prognosis of HICH patients whose GCS score ranging from 5 to 9 after minimally invasive surgery.

METHODS

Inclusion of patients

We retrospectively reviewed a prospectively acquired database involving patients who presented with HICH to the Second Affi liated Hospital of Zhejiang University School of Medicine between 2014 and 2016. All patients were treated with surgical procedures.

Design and setting

The research protocol has been approved by the Hospital Ethics Board. HICH was defined as a spontaneous, nontraumatic, abrupt onset of severe headache, altered level of consciousness, or focal neurological deficit that was associated with a focal collection of blood within the basal ganglion region, with a medical history of hypertension. The study subjects were recruited if they met the following criteria: adult patients (≥18 years old); past medical history of hypertension; basal ganglion hematoma (amount ＞ 30 mL) confirmed by CT scan; disease onset within 6 hours and GCS score within 5-9. The exclusion criteria included hemorrhage caused by trauma, aneurysm or arteriovenous malformation, venous sinus thrombosis, involvement of ventricular system, severe systematic comorbidities, preoperative brain herniation, death within 24 hours postoperatively and preoperative administration of antiplatelet or anticoagulant drugs. The implantation of the ICP monitor was determined by the neurosurgeon's experience and the choice of the patients' legal surrogates. Patients were thus divided into ICP monitoring group and no ICP monitoring group.

Patient management protocol

Compared with direct visualization, minimally invasive approach has the advantages of reducing the surgical trauma injury and maximizing the removal of hematoma.[6]In our study, hematoma evacuation was conducted via either transcortical or transsylvian transinsular approach by microscope. A 5-cm straight incision was designed according to the location of the hematoma and the surgical approach, and the area of bone window was approximately 4-5 cm2. For transcortical approach, a small incision was made at the non-eloquent cortex closest to the hematoma. For transsylvian transinsular approach, after splitting the distal sylvian fissure and locating the insular cortex, a small incision was made parallel to the sylvian fissure. After entering the hematoma cavity, evacuation was performed using mild suction and continuous irrigation. The persistent sources of hemorrhage were cauterized, and a small amount of hematoma might not be completely removed around the cavity when the blood clot adheres tightly to surrounding neurovascular structures. In the ICP monitoring group, the ICP probe (Camino, Integra Lifescience) was implanted in the hematoma cavity during the surgery.

In the ICP monitoring group, an ICP goal of ＜15 mmHg was maintained using a stepwise management strategy: for ICP between 15 and 25 mmHg, 125 mL of 20% mannitol compound was intravenously injected every 8 to 12 hours to maintain ICP lower than 15 mmHg; for ICP higher than 25 mmHg, 125 mL of 20% mannitol would be added every 6 hours and emergent CT scan would be performed. If ICP suddenly jumped 5-10 mmHg and lasted longer than 10 minutes, 125 mL of 20% mannitol with or without furosemide would be administered. For the control group, the dosage of mannitol was based on neuroimaging and clinical examination including CT scan, consciousness, pupil size, vital signs, GCS score. Head CT was obtained immediately after operation and at 1, 3, 5, and 7 days afterward: for patients with GCS＞9 and only mild midline shift, 125 mL of mannitol 2 or 3 times per day was prescribed; for unconscious patients with GCS＞6, and moderate midline shift, 125 mL of mannitol was given at 3 times per day, with furosemide when necessary; for patients with GCS ≤6 and obvious midline shift, mannitol and furosemide would be used 4 times per day, with or without albumin. For the recurrence of hemorrhage with severe mass effect, re-operation would be considered. External ventricular drainage was performed in cases of hydrocephalus. If refractory intracranial hypertension occurred, decompressive craniectomy was performed.

Data collection

The following data were recorded: age, gender, hypertension, diabetes, ischemic heart disease, coagulopathy, medication, and initial GCS. The hematoma volume was calculated by Coniglobus formula (v=1/2×a×b×c) based on preoperative CT. The midline shifts were determined using the distance from the midline to the septum pellucidum. The primary outcomes were the mortality rate and the Glasgow Outcome Score (GOS) at 6 months. GOS of 4 or 5 was classifi ed as a good recovery, while 1 to 3 as a poor recovery. The secondary outcomes were determined as re-operation rate, central nervous system (CNS) infection rate, and in-hospital duration.

Statistical analysis

Statistical analysis was performed with SPSS for Windows (version 21.0). Categorical variables were expressed as percentage and compared with Pearson chisquare or Fisher's exact test. Continuous variables were presented as mean±standard deviation format or median (25% quartile, 75% quartile) format and were compared using independent-samples t-test or Wilcoxon rank-sum test. Logistic regression was used for multiple factor analysis. Variables with significance of P＜0.10 in the univariate analyses were then entered into a multivariate analysis via binary logistic regression (forward: LR) to identify independent predictors of 6-month favorable outcomes. P＜0.05 (two sided) were considered statistically signifi cant.

RESULTS

Patient demographics

Preoperative and postoperative characteristics of the entire surgical cohort were summarized in Table 1. A total of 116 patients were included in the final analysis. Male patients accounted for 55.1%. The mean age was 61.7±9.3 years. The median GCS was 7. The basic characteristics including age, gender, comorbidities such as diabetes, ischemic heart disease, coagulopathy, hematoma volume, hematoma side and extend of midline shift were similar between the two groups. Four reoperations were performed in ICP monitoring group including two re-hemorrhage cases and two uncontrolled brain edema cases. Five re-operations were performed in no ICP monitoring group including four re-hemorrhage cases and one uncontrolled brain edema case.

ICP monitoring

The median duration of ICP monitoring was 6 days, and ICP was elevated in 70% of patients. Patients with ICP monitoring had a significantly better GOS score (3 [2-4] vs. 3 [2-3], P＜0.05). The average in-hospital duration for patients with ICP monitoring was 16.68 days compared with 20.47 days for the no ICP monitoring patients (P＜0.05). Mortality rates, CNS infection and re-operation rates between ICP monitoring and no ICP monitoring did not differ signifi cantly.

Predictors of GOS

The good recovery and bad recovery groups are compared in Table 2. Good recovery group had a younger average age (P＜0.001), a greater initial GCS score (P＜0.001), smaller hematoma volume (P＜0.001), milder midline shift (P＜0.001). Good recovery group was associated with more ICP monitoring implantation (61.8% vs. 15.9%, P=0.009). The characteristics including gender, comorbidities (such as diabetes, ischemic heart disease, coagulopathy), hematoma side, re-operation and in-hospital duration did not differ signifi cantly between the two groups.

Multivariate analysis showed that age (OR 0.916, 95% CI 0.851-0.985, P=0.018), GCS on admission (OR 8.827, 95% CI 3.414-22.821, P＜0.001) and presence of ICP monitor (OR 17.167, 95% CI 3.075-95.833, P=0.001) were independent predictors of 6-month favorable outcomes (Table 3). We stratifi ed the data based on preoperative GCS score (5, 6, 7, 8 and 9 groups, Figure 1), and showed that both GCS score on admission and the placement of ICP monitoring predicted a better 6-month outcome.

DISCUSSION

HICH shares a notorious reputation of high mortality and morbidity rate. The primary brain injury is caused by the mass effect of hematoma, which is closely related to the rate, volume and the location of the hemorrhage. The peri-hematoma edema resulting from the dissociates of hemoglobin is the main source of secondary injury.[7]Decompressive craniectomy and expansive duraplasty with hematoma evacuation is the traditional standard treatment.[8]But it is time-consuming and exerts secondary trauma to the patients due to the blood loss, wide exposure and brain injury. Therefore, minimally invasive approaches, including stereotactic puncture, keyhole craniotomy and endoscope-guided hematoma evacuation, are gaining wider application.[9]However, attention should be paid to the postoperative ICP rebound, mainly caused by the rebuilding of intracerebral circulation and the consequent congestion, brain edema, and re-hemorrhage, which result in rapid clinical deterioration and even brain herniation. Thus a real-time supervision and early-warning system should be applied to guide the post-operative management.

In the field of head trauma and neuro-critical care, ICP monitoring has been considered as the standard of care, especially in severe TBI.[10]By comparing the outcomes of TBI patients in hospital with different frequencies of ICP monitoring, Aziz et al[11]found the adjusted OR of death was 0.52 (95% CI 0.35-0.78) in the quartile of hospitals with the highest use, compared to the lowest. Through a retrospective analysis of 123 severe TBI patients, decreased mortality rate was found in the ICP monitoring group (OR 0.32, 95% CI 0.35-0.78).[12]After propensity score matching, Dash[13]concluded that ICP monitoring was associated with a signifi cant decrease in 6-month mortality and had a significant impact on the 6-month favorable outcomes among Chinese patients. In all these series, continuous ICP monitoring helped to protect fragile brain from persistent high ICP, which was beneficial to save patients' lives and improve their functional outcomes.

Table 1. Basic characteristics of HICH patients with ICP and without ICP monitoring

Table 2. Comparison of good recovery and bad recovery

Table 3. Results of logistic regression model analysis

Figure 1. Average GOS at 6-month based on GCS scores on admission.

Minimally invasive craniotomy and hematoma evacuation has been proved effective and safe in the treatment of supratentorial HICH. Transcortical and transsylvian transinsular approaches are most commonly applied by the surgeons.[14-16]Unlike the decompressive craniectomy, the space of minimally invasive craniotomy is limited. The implantation of ICP monitor could help neurosurgeons to detect any abnormal increase of ICP and initiate intervention in time. However, the efficacy of accurate and continuous ICP monitoring after minimally invasive surgical evacuation of HICH about patients' outcomes has been poorly studied. In a prospective observational study recruiting 186 patients with severe ICH, ICP monitoring and early operation were predictors of longer survival and better functional outcomes.[17]Sykora et al[18]proved that average ICP, ICP variability and the frequency of ICP values ＞20 mmHg were independently associated with mortality and poor outcome after ICH, further implying the importance of ICP monitoring and control in ICH. In our study, we found that ICP monitoring was significantly associated with the 6-month functional outcomes. The core value of ICP monitoring was effective interventions before the irreversible cerebral injury, which could help clinicians to make rational treatment decisions. Therefore, we suggested that postoperative ICP monitoring should be applied to the patients with HICH.

There are both invasive and non-invasive methods to measure ICP. Ultrasonic measurement of optic nerve sheath diameter provides a non-invasive, rapid and convenient approach to evaluate elevated ICP, but continuous monitoring is impossible and the accuracy relies on the experience of the operator.[19]Although it runs the risks of infection, hemorrhage and brain injury, invasive implantation of ICP monitor is still the golden standard. The monitor could be intraventricular, intraparenchymal or subdural, with the highest accuracy achieved by the first two types.[20]In this study, the monitor was placed in the hematoma cavity, which could avoid further injury caused by insertion of the monitor into the ventricle or surrounding brain tissue, or the relatively unsteady readings when buried subdurally. The rate of CNS infection didn't differ signifi cantly between ICP monitoring and no ICP monitoring in our study (P=0.260).

This study is limited by its retrospective nature, and the number of patients enrolled was still not large enough. In addition, the value of ICP monitoring was only examined in patients with mini-bone window craniotomy, while for other minimally invasive modalities, including stereotactic puncture and endoscopic hematoma evacuation, the advantages of ICP monitoring remained to be assessed. In addition, our study included patients whose GCS scores ranging from 5 to 9, which may have some selection biases. Therefore, prospective, randomized controlled trial is needed to further evaluate the usage of ICP monitoring in HICH.

CONCLUSION

Here in our retrospective study, ICP implantation was associated with a better 6-month functional outcome. Besides, we found that age and initial GCS score were also associated with the prognosis. Future study is still needed to confirm our results and find the subgroup of HICH patients who will benefi t most from the minimally invasive surgical intervention and ICP monitoring.

Funding:None.

Ethical approval:This study was approved by the Hospital Ethics Board.

Conf ict of interest:There is no confl ict of interest.

Contributors:XC developed the concept of study; XC and YW collected the data; YW and HZ performed data analysis; XC wrote the draft; YW and HZ critically reviewed manuscript; HZ provide administrative support.