Introduction

Aneurysmal subarachnoid haemorrhage (aSAH) is associated with high morbidity and mortality. Delayed cerebral ischaemia (DCI) is considered one of the most common causes of complications [9, 14].

Whilst vasospasm refers to angiographic evidence of arterial narrowing, DCI refers to the related neurological deterioration which is not attributable to hydrocephalus, seizures or medical causes [22]. Typical features of clinical deterioration due to DCI are the development of neurological deficits as a result of focal ischaemia, or a decrease in the level of consciousness as a result of global ischaemia [22]. DCI remains a poorly understood phenomenon and is a major predictor of poor outcomes in aSAH [2, 5]. When DCI is found in combination with evidence of arterial narrowing on angiographic imaging, this is referred to as ‘symptomatic vasospasm’.

Current management strategies for symptomatic vasospasm focus on intensive care medical management with augmented haemodynamic therapies.

Symptomatic vasospasm refractory to such measures can be treated with angioplasty as a rescue measure, using intra-arterial vasodilators (chemical angioplasty (CA)), transluminal balloon angioplasty (TBA) or a combination of both (TBA plus CA) [9, 19]. Traditionally, angioplasty of symptomatic cerebral vasospasm uses a mix of TBA for proximal vasospasm, followed by CA for distal vasospasm or for vessels deemed at risk of rupture if a balloon was to be used [10]. Angioplasty has been shown to achieve significant improvement when employed early after onset of neurological deficits; therefore, timely diagnosis is essential [1].

For patients who are sedated in intensive care, prompt diagnosis of DCI can be challenging. Parenchymal measurement of partial pressure of brain tissue oxygen (PbtO2) allows continuous objective real-time assessment of brain oxygen tension and therefore early identification of patients at risk of DCI [4]. Lower mean PbtO2 and longer periods of compromised cerebral oxygenation are associated with poorer outcomes and death in aSAH [15].

Multi-modal monitoring including PbtO2 is routinely used at this centre to assess and guide management in DCI. PbtO2 is also monitored during angioplasty to objectively determine when successful reperfusion has been established. At this centre, combined approach (TBA plus CA) is preferred unless TBA is too high risk.

In this pilot study, we aimed to compare the effects of CA alone versus TBA and CA, on PbtO2. We hypothesised that the PbtO2 responses would be more sustained in patients treated with a combined approach.

Methods and materials

This study was performed as part of a retrospective service evaluation.

Population

We included all patients with aSAH admitted to this single centre between 2018 and 2019, with suspected DCI and an intra-parenchymal Raumedic NEUROVENT-PTO® in situ as part of their multimodal monitoring throughout the angioplasty process. Any patients who had already undergone a previous angioplasty procedure were excluded in order to capture the effectiveness of TBA plus CA versus CA alone as an initial treatment. All patients had secured aneurysms; were ventilated, sedated and monitored in the neurocritical care unit (NCCU); and received daily nimodipine and transcranial Doppler ultrasound (TCD) from the day of admission.

Placement of PbtO2 Bolts

The PbtO2 probe was inserted at the site that most likely would be affected by the vascular spasm as seen on either CTA or angiography (e.g. for anterior circulation vasospasm, the bolt could be placed slightly more medial, and for posterior circulation vasospasm the bolts were placed occipitally). This procedure was done at the bedside in the NCCU.

Criteria for interventional angiography

Patients with persistently low PbtO2 (which was defined as less than 15 mmHg at the time of this study) underwent a local protocol of escalated medical management for DCI, including (a) temporary hyperoxia (for example, increase FiO2 to 100% for 5 min to check there is a response (and no wire issues); (b) optimising blood gas, ensuring airway does not require suctioning and keep PaCO2 5.0–5.5 kPa, PaO2 13–16 kPa); (c) ensuring ICP was managed optimally; (d) optimisation of haemoglobin (keeping levels above 100 g/L with blood transfusion); (e) treatment of pyrexia (commencing cooling with target temperature of 36 °C); (f) increasing the level of induced hypertension (in patients with secured aneurysms can be increased incrementally to 200–220 mmHg with maximum noradrenaline rare 0.8 mcg/kg/min); and (g) and finally a trial of hyperoxia with FiO2 increased to achieve a PaO2 greater than 18 Kpa and maximum FiO2 60%. If PbtO2 did not improve (after 30 min) despite the above maximal medical management, a multidisciplinary discussion was undertaken regarding the value of interventional procedure for vasospasm. This was influenced by patient factors as well as findings on TCD and CT angiogram and in patients deemed appropriate, angiography followed by intervention (either CA or TBA plus CA) were performed directly.

Intervention

All patients in this study underwent either CA (verapamil) or TBA plus CA as part of the standard clinical management in this centre. The treatment decision was made by the treating interventional radiologist. The general principles underlying this decision were that patients with proximal vasospasm amenable to balloon angioplasty underwent this procedure whilst those with more distal, or anatomically riskier vasculature, underwent chemical angioplasty alone. Data from the NEUROVENT-PTO® were continuously collected pre-, during and post-angioplasty and downloaded as second-by-second data.

Comparison

Comparison between pre-angioplasty PbtO2 and post-angioplasty PbtO2 median values (4 h before angioplasty, 4 h after and 12 h after).

Outcome

The recorded outcome measure was PbtO2 levels and how long they were sustained for after treatment.

Angioplasty procedure

Angiographic procedures were performed under general anaesthesia via common femoral artery access. Bilateral internal carotid and, depending on the location of vasospasm, vertebral artery angiography was performed. Cerebral vasospasm was assessed by the interventional neuroradiologist. CA was performed using intra-arterial infusion of 10 mg of verapamil into each target artery over up to 30 min. Response to verapamil was determined angiographically. Where there was proximal vasospasm, systemic heparinisation was administered adapted to weight, and then TBA of the target vessel segment was performed. TBA was performed via 4- to 6-F catheters with Transform (Stryker) or Maverick balloons using Envoy guidewires. Response to TBA was again determined angiographically.

Statistical analysis

All comparative tests were performed in GraphPad Prism version 10.0.1. Continuous data are presented as medians and ranges and categorical data are presented as counts. Continuous demographic data were compared using an unpaired t-test. Categorical data were compared using chi-square test for trend and Fisher’s exact test. Comparisons between PbtO2 results (in mmHg) for (1) from baseline to 12 h post procedure within a group and (2) between chemical versus chemical plus TBA were performed using paired and unpaired t-tests respectively. A p-value of less than 0.05 was considered significant.

Results

Demographics

A total of ten patients with PbtO2 monitors underwent angioplasty, of which six had intra-arterial verapamil only, and four patients underwent a combination of chemical and mechanical balloon angioplasty. All patients underwent treatment within the first 14 days of ictus (Table 1).

Table 1 CA vs. TBA plus CA demographics
Full size table

PbtO2 response to angioplasty

For both interventions, there was an upward trend in PbtO2 of 13 mmHg and 15 mmHg for CA (Fig. 1A) and TBA plus CA (Fig. 1B) respectively, in the first 4 h post procedure. This improvement in PbtO2 was lost at 12 h in the CA group (Fig. 1A) whereas in the TBA plus CA group the effect was maintained at 12 h (Fig. 1B).

Fig. 1
figure 1

A Chemical angioplasty (n = 6) and B transluminal balloon angioplasty plus chemical angioplasty (n = 4). Brain tissue oxygenation before angioplasty (− 4 h) and after, at 4 h and 12 h (in mmHg)

Full size image

Within the CA group, PbtO2 levels increased by 6 mmHg from baseline to 12 h post angioplasty from 14.6 ± 14.4 to 20.7 ± 10.9 mmHg (mean ± SD) (Fig. 1A). This was not a significant increase (p = 0.18). Within the CA plus TBA group, PbtO2 levels increased by 13.5 mmHg from baseline to 12 h post angioplasty from 22.5 ± 8.5 to 36.0 ± 4.1 mmHg (mean ± SD) (Fig. 1B). This was a significant increase (p = 0.03).

The difference between the CA and TBA plus groups, at 12 h post procedure, was significantly different being 20 mmHg for those treated with CA, compared to 36 mmHg for those treated with TBA plus CA (p = 0.02) (Fig. 2).

Fig. 2
figure 2

The difference in brain tissue oxygen (in mmHg) at 12 h following CA vs. TBA plus CA (p = 0.02)

Full size image

One patient in the CA group developed an anterior circulation infarction post 14 days.

Discussion

Key findings

In this short pilot study, we compared the effect of chemical angioplasty versus a combined treatment of transluminal balloon plus chemical angioplasty on PbtO2 in patients with DCI post-aSAH. Both methods were effective within the first 4 h, but only the combined angioplasty caused a sustained improvement in PbtO2 at 12 h post-procedure. Our findings suggest that PbtO2 is useful as a neuromonitoring tool in this cohort and supports the routine use of PbtO2 monitoring in patients undergoing angioplasty.

Chemical angioplasty

Few studies demonstrate a clear and sustained objective improvement of PbtO2 following chemical angioplasty treatment alone. In 1995, Leur and colleagues first demonstrated a brief improvement in brain oxygenation (using transcranial cerebral oximetry) following intra-arterial papaverine infusion for symptomatic cerebral vasospasm [11]. Other reports describe varying results, with either minimal benefit (no significant PbtO2 change 12 h post-CA), or clear sustained benefit (a minimum of 8 h or more significant increase in PbtO2) and good clinical outcomes[4, 21]. Near-infrared spectroscopy techniques also revealed inconsistent improvements in brain oxygenation after chemical angioplasty [12]. Interestingly, some studies found a transient (4–10 min) decrease in PbtO2 in patients receiving intra-arterial papaverine [20]. There was no mention of any beneficial effect after this but it was considered whether confounding factors such as a concomitant uncompensated decrease in intracranial pressure had contributed to the decrease in PbtO2 [20].

Balloon angioplasty

Initial large-scale studies showed favourable outcomes when transluminal balloon angioplasty (TBA) is performed at an early stage [1, 17]. Santillan and colleagues described their positive experience over a 9-year period of treating distal cerebral vasospasm with TBA, concluding that it is both safe and effective, with 82.1% of their 28 patients showing good outcome (mRS < 2) [18]. This is also supported with evidence that balloon angioplasty can result in up to 5 days of improvement in brain oxygen tension (measured via Licox PtiO2 probe) [8]. Labeyrie et al. demonstrated in a large series of patients with cerebral vasospasm, that distal balloon angioplasty decreased the risk of DCI and recurrence of vasospasm compared with conventional angioplasty. Nonetheless, these findings did not correlate with a clinical benefit [10]. Schacht et al. provide data on TBA for cerebral vasospasm in their centre which corroborate the findings which suggest that TBA is a safe and effective treatment for vasospasm [19]. However, they warn of recurrent cerebral vasospasm as a common finding post-TBA, especially in those with moderate to poor angiographic response [19]. Whilst the majority of evidence supports TBA, some studies show minimal benefit of TBA alone in terms of clinical outcome [13, 16].

Comparison of combined angioplasty and chemical angioplasty for DCI

Our findings are in agreement with existing evidence suggesting that combined angiography has a more sustained effect on brain oxygenation than chemical angiography alone.

The clinical outcome data supports a combined angioplasty approach. A literature review of 530 patients showed 62% improved clinically after TBA, compared to only 43% for CA alone (intra-arterial papaverine) [9]. Measures of brain oxygenation (transcranial Doppler velocities) were also significantly improved in the TBA group [9]. Furthermore, the clinical improvement in the CA group was only transient and there was a need for repeated interventions [9]. The same study found intra-arterial nicardipine to have more sustained improvement than papaverine.

Elliot et al. compared the effect of TBA with CA (a papaverine infusion) for the treatment of anterior circulation vasospasm post-SAH, using mean transcranial doppler (TCD) velocities pre- and post-treatment [6]. Both interventions demonstrated initial significant improvements; however, balloon angioplasty showed a more marked decrease (45% vs. 20%). Furthermore, in the papaverine group, velocities had deteriorated again by post-treatment day 2, necessitating re-treatment significantly more often than balloon angioplasty [6]. PbtO2 measurements have been shown to correlate with TCD findings, and so the fact that our data supports this study further validates the use of PbtO2 as a monitoring tool [3].

The comparison between CA (super selective intra-arterial papaverine infusion) and TBA has been analysed using continuous measurement of jugular bulb vein oxygen saturation (SvjO2) [7]. Nearly one third of patients who underwent TBA required additional papaverine in order to achieve normalisation of luminal diameter, suggesting that regardless of method used (TBA or CA), a combined approach is optimal [7].

Strengths and limitations

Owing to the COVID-19 pandemic, this study was interrupted and therefore only low numbers are included. Nonetheless, the data is presented as a pilot study, to ensure the information gleaned is available.

Despite the low numbers, there is a clear objective difference between the CA and TBA plus CA group, in terms of sustained increase to PbtO2. Due to the PbtO2-based protocol-led management at this centre, all patients were subjected to the same standardised decision-making protocol, thus minimising the variability of management. Future studies could use PbtO2 as a target for angioplasty procedures.

One limitation of this study relates to its pragmatic design. Patients with vasospasm affecting narrow or friable vessels (which were sometimes more distal) were treated with only CA. Certainly, spasm affecting more distal vessels can be considered a different pathology to spasm affecting the circle of Willis vasculature. Therefore, the difference in PbtO2 response between TBA + CA and CA alone, may, in part, relate to the difference in proximal and distal vasospasm pathophysiology.

Conclusion

In this study, balloon plus chemical angioplasty demonstrated greater and more sustained increases in brain tissue oxygen in patients with refractory vasospasm. Our findings suggest that PbtO2 is a useful tool for monitoring the response to angioplasty in vasospasm.