24.1 Goals

1. 
   

   Understand the indications for balloon test occlusion (BTO).

   
   
2. 
   

   Analyze standard techniques for BTO.

   
   
3. 
   

   Evaluate the nuances and complications associated with this procedure.

   
   
4. 
   

   Review adjunct modalities aimed at improving prediction of ischemic complications with BTO.

24.2 Case Example 1

24.3 Case Example 2

24.4 Case Example 3

24.4.1 History of Present Illness

A 68-year-old, right-handed man presented with progressive left-sided hearing loss and facial weakness. Imaging revealed a mass involving the left temporal bone and ear canals. Biopsy confirmed a squamous cell carcinoma, and the patient was planned to undergo a temporal bone resection with parotidectomy and possible left ICA sacrifice.

Past medical history: Hypertension, hyperlipidemia, hypothyroidism.

Past surgical history: None.

Family history: Noncontributory.

Social history: Noncontributory.

Review of systems: As per the above.

Neurological examination: Left-sided sensorineural hearing loss and subtle left nasolabial flattening (House-Brackmann scale 2). The remainder of the neurological examination was unremarkable.

Imaging studies: Mass associated with the left external auditory canal, middle ear cavity, and mastoid temporal bone with superior extension into the middle cranial fossa, as well as anterior-inferior extension into the temporomandibular joint (Fig. 24.3a).

24.5 Level of Evidence

BTO is an important test for prediction of ischemic complications following a permanent carotid occlusion.

The sensitivity of BTO is increased with addition of adjunctive modalities such as hypotensive challenge, HMPAO SPECT, positron emission tomography (PET), magnetic resonance imaging (MRI), and transcranial Doppler (TCD), among others. Majority of evidence for these procedures are based on observational case series and expert opinion only (Level of Evidence C).

24.6 Case Summaries

1. 
   

   What is the purpose of BTO?

   
   

   Elective carotid sacrifice is necessary in certain clinical settings involving resection of head and neck tumors and in complex vascular lesions, such as fistulas or large aneurysms.

   
   

   ICA ligations are associated with ischemic complications in 49% of patients. Similarly, in common carotid artery occlusions, ischemic complications were noted in 28% of patients ^{1 , 2 , 3} with a mortality of about 12%. ⁴ These complications may arise due to poor collateral circulation across the circle of Willis, inadequate reserve despite good collateral flow, and thromboembolism. However, with favorable anatomy, vessel sacrifice can be asymptomatic. BTO is an endo-vascular angiographic procedure which simulates an arterial occlusion in a controlled environment to predict ischemic complications.

   
   
2. 
   

   What are the procedural complications and associated prevention strategies associated with BTO?

   
   

   The major complications associated with BTOs may be asymptomatic or symptomatic vessel injury and/or transient or persistent ischemic events. The largest BTO case series to date by the University of Pittsburgh reported asymptomatic events in 1.6% of patients and symptomatic events in 1.6% (1.2% transient and 0.4% permanent). ⁵ Similar results have been reported by others which are within the range for diagnostic angiography alone. ^{6 , 7} The major preventative strategies include correct sizing of the balloon and preinflation heparinization.

   
   
3. 
   

   What adjunctive modalities are used with BTO to increase sensitivity and specificity?

   
   

   The common adjunctive modalities to improve predictions of ischemic complications include clinical testing with a simulated hypotensive challenge, TCD ultrasonography, Xenon-enhanced computed tomography, 0-15 labeled PET, or 99mTc-HMPAO SPECT. The literature on these techniques is discussed below.

   
   
4. 
   

   What are the common causes of ischemic complications and preventative strategies following permanent occlusions in patients with a negative BTO?

   
   

   Following a negative BTO, there is still a 4.7 to 25% risk of ischemic complications, depending upon the adjunct modal-ity_4,5,8,9,io The major causes of these ischemic complications include both hemodynamic insufficiency and thromboembolism. A rescue bypass has been utilized to correct hemodynamic complications with success. ^{11 , 12} Short-term anticoagulation or antiplatelet therapy can be considered following permanent occlusions to reduce risk of stump embolism. ¹³ There are, however, no evidence-based guidelines for these strategies.

24.7 Landmark Papers

Serbinenko FA. Balloon catheterization and occlusion of major cerebral vessels. J Neurosurg 1974;41(2):125-145.

Prior to the introduction of formal BTO, ¹⁴ methods to determine the functional adequacy of the circle of Willis involved indirect cerebral blood flow (CBF) measurements such as angiography with manual compression, ICA stump pressure measurements, and oculoplethysmography (OPG). ¹⁵ However, these techniques had several limitations, including inconsistencies in applied manual pressure and variations in the pressure of the injected dye during an angiogram. Stump pressure measurements were proven unreliable in carotid endarterectomy case series models. ¹⁶ Similarly, OPG measured flow indirectly by measuring the flow in the ophthalmic artery.

In his groundbreaking paper, Serbinenko used self-designed, silicone balloon catheters for temporary occlusion of more than 300 vessels. One hundred and eighty-seven of these cases involved occlusion of the ICA. His procedure involved direct puncture of the carotid artery for introduction of the balloon catheter which was then flow navigated to various portions of the ICA. He reported only two complications out of his over 300 temporary balloon occlusions due to unexplained thrombosis of the middle cerebral artery. Unfortunately, both of these cases resulted in the patients’ death.

While this is the original study demonstrating the feasibility of BTOs, it served as an introduction to the technique rather than a rigorous, controlled validation of it. While Serbinenko successfully provided a very detailed manual on the use of BTO in various pathologies and locations within the extra- and intracranial cerebrovasculature and successfully introduced the practice of BTO to the neurosurgical community, he did not provide any information concerning the predictive value of the study following permanent occlusion.

De Vries EJ, Sekhar LN, Horton JA, et al. A new method to predict safe resection of the internal carotid artery. Laryngoscope 1990;W0(l):85-88.

With the incorporation of clinical BTO in the 1980s, prediction of ischemic complications from vessel occlusions were markedly improved, with nearly 100% prediction of ischemic complication for patients failing the test. ^{5 , 15 , 17} However, a significant percentage of patients, up to 10%, had strokes despite passing the clinical BTO. ^{4 , 8 , 9 , 18 , 19 , 20}

In light of an insufficient negative predictive value of a clinical BTO, authors reported benefits of adding CBF analysis to improve risk prediction. ^{4 , 21 , 22} The initial case series on BTO with Xenon-enhanced CT (Xe-CT) for CBF analysis was published by Vries et al in 1990. ¹⁵ The authors defined compromised flow as less than 20mL/100mg/min based on prior data. ²³ Their case series demonstrated less than 3% incidence of ischemic complications following permanent carotid occlusion in patients that passed BTO with Xe-CT. Similarly, Linskey et al demonstrated an approximately 3% risk of ischemia when BTO was paired with Xe-CT scan. ⁴ In Sen et al’s series of ICA bypass patients, ischemic complications were seen in 7% of patients that passed both clinical and Xe-CT components of BTO, 56% of patients who passed the clinical component of BTO but failed Xe-CT (with a CBF less than 30 mL/100 mg/min), and 100% of patients who failed both. ¹⁷

One of the technical limitations associated with using Xe-CT is the need to deflate the balloon, move the patient to a Xenon CT scanner, and re-inflate the balloon for Xe-CT. This adds significantly to the overall procedural complexity and risk. Administration of 133 radioactive Xenon followed by gamma activity measurements using cranial probes has also been described to assess CBF. ^{24 , 25} While this method can be completed in the angiography suite, 133 radioactive Xenon does not provide regional blood flow information and may be less sensitive than newer methods.

Matsuda H, Higashi S, Asli IN, et al. Evaluation of cerebral collateral circulation by technetium-99 m HM-PAO brain SPECT during Matas test: report of three cases. J Nucl Med 1988,29 (W):1724-1729.

99 m Technetium HMPAO is a lipophilic radiotracer with a half-life of about 6 hours. It readily crosses the blood-brain barrier and is converted into a hydrophilic form, leading to contrast retention between normal and abnormal brain for up to 2 hours. The earliest report using HMPAO SPECT to assess CBF was published by Matsuda et al during manual compression. The findings in this initial study of three cases correlated well with the collateral status of the downstream occluded vessel, as observed on direct cerebral angiography. ²⁶ Used as an adjunct with BTO, HMPAO SPECT has been shown to add useful information to stratify risk after permanent carotid occlusion. ^{27 , 28} Monsein et al compared HMPAO SPECT scans at baseline followed by scans performed with BTO. The authors noted that patients with no change between baseline and test occlusion SPECT imaging had adequate collaterals to sustain flow after a permanent occlusion of the artery. ²⁹ These initial studies validated the negative predictive value of this adjunct test. Efforts to quantify CBF more objectively have since been pursued to correlate with clinical intolerance to vessel sacrifice. One study reported that a reduction in CBF to less than 85% of baseline in the MCA territory is associated with ischemic complications. ³⁰ Other authors have utilized similar approaches to calculate asymmetry indices for specific regions of interest on SPECT imaging. Similar cutoffs of 10 to 15% correlated well with ischemic complications after permanent occlusion. ^{31 , 32 , 33}

BrunbergJA, Frey KA, Horton J A, DeveikisJP, Ross DA, Koeppe RA. [150JH20 positron emission tomography determination of cerebral blood flow during balloon test occlusion of the internal carotid artery. AJNR Am J Neuroradiol 1994;15(4):725-732.

Despite the advantages of using Xe-CT or SPECT for CBF analysis in conjunction with BTO, these studies lacked specificity, as the semiquantitative asymmetry analysis was associated with high rates of false-positive results. ²¹ PET, which allows quantitative CBF analysis, improved both the sensitivity and specificity ofa clinical BTO ^{4 , 21 , 22 , 34 , 35}

PET technology in neuroimaging makes use of 0-15 labeled compounds to determine a variety of physiologic variables including CBF, cerebral blood volume (CBV), oxygen extraction fraction (OEF), and cerebral metabolic rate of oxygen (CMR0₂). These variables can then be utilized for assessing ischemia in both acute and chronic cerebrovascular disease states.

The first report employing PET during BTO for prediction of ischemia after permanent occlusion was published by Brunberg et al using [150]H20. This study showed that patients with CBF reduction to 25 to 35mL/100g/min during balloon occlusion are at risk for developing cerebral infarction after permanent ICA occlusion. ³⁶ Similar findings were reported by Murphy et al. ³⁷ Overall PET scans were more reliable than other available modalities; however, the prolonged inflation of the balloon during the entirety of the PET scan may pose significant procedural risks.

A more recent study utilized a novel dual tracer autoradiographic (DRAG) method in PET scans as an adjunct to BTO for shortening the PET examination period. The authors employed use of sequential administration of dual tracers of ¹⁵ 02 and C ¹⁵ 0₂ during a single PET scan, followed by computation of CBF and CMR0₂ autoradiographically. ³⁴ This protocol reduced the time of scan from 1 hour to 15 minutes. The study further consolidated the previously established CBF critical values and introduced the use of OEF/CMR0₂ ratio to further predict ischemic complications. Despite the promise of PET scans with BTO, these novel protocols are not available at most institutions. Furthermore, requirement of reinflation of the balloon outside of the angiography lab and prolonged balloon inflation times have limited the utilization of PET studies with BTO.

Schneweis S, Urbach H, Solymosi L, Ries F. Preoperative risk assessment for carotid occlusion by transcranial Doppler ultrasound. J Neurol Neurosurg Psychiatry 1997 May;62(5):485-489.

Another adjunctive test employed TCD ultrasonography to assess CBF. Early studies demonstrated that TCD velocities can be correlated with CBF as assessed simultaneously on Xe-CT following carotid occlusion. ³⁸ Schneweiss et al validated the use of TCD with BTO. They concluded that a drop in the mean velocities of more than 30% was associated with clinical manifestations during temporary occlusion. ³⁹ The addition of acetazolamide potentially not only made the test more reliable by exhausting cerebral autoregulation but also increased the risk of ischemic injury during the trial.

Another case series of 32 patients demonstrated that a reduction in both mean velocity and pulsatility index (PI) greater than 50% was associated with neurological deficits. A reduction of up to 30%, however, had no significant ischemic complications. For patients with an intermediate reduction between 30 and 50% in mean velocity and PI, motor vasoreactivity (representative of cerebrovascular auto regulation following voluntary motor activity) was used to further stratify risk. ⁴⁰ Similar results were noted in another study reviewing changes in MCA velocities in 22 patients undergoing percutaneous transluminal angioplasty of the ICA. A greater than 50% drop in velocities was associated with transient or persistent neurological deficits. ⁴¹ In a study of carotid balloon occlusion monitored with electroencephalogram (EEG), SPECT, and TCDs, a drop of up to 40% in mean TCD velocity was not associated with changes in EEG, SPECT, or clinical examination. However, a drop greater than 40% was correlated with changes in each of these. ⁴² Based upon the available data, TCD velocity reductions > 30% may predict higher risks of ischemic complications.

van Rooij WJ, Sluzewski M, Slob MJ, Rinkel GJ. Predictive value of angiographic testing for tolerance to therapeutic occlusion of the carotid artery. AJNR Am J Neuroradiol 2005;26( 1): 175-178.

Abud DG, Spelle L, Piotin M, Mounayer C, Vanzin JR, Moret J. Venous phase timing during balloon test occlusion as a criterion for permanent internal carotid artery sacrifice. AJNR Am J Neuroradiol 2005;26( 10)2602-2609.

A simple adjunct which can be performed during test occlusion is assessment of collateral status. Evaluation of cortical venous delay on the occluded side compared to the contralateral control was used to assess collateral circulation. ^{43 , 44 , 45} van Rooij et al followed 74 patients who underwent BTO prior to therapeutic carotid occlusion. Synchronous venous filling with< 0.5 seconds delay was the cutoff criteria for a negative angiographic BTO. Only 1 patient in 51 therapeutic occlusions of the carotid artery following a negative study developed a transient, hemiparesis with hypoperfusion infarcts on follow-up imaging. However, two patients with ruptured aneurysms died from diffuse vasospasm following permanent occlusion. ⁴⁴

Similarly, Abud et al showed that a venous delay of< 3 seconds on a BTO angiogram was associated with low risk of ischemic hemodynamic complications following permanent vessel occlusion. One of the three patients in this study with venous delay > 3 seconds had a border zone ischemic stroke following permanent occlusion. ⁴³ More recently, venous delay on angiogram was compared with HMPAO SPECT imaging in 56 patients. Twenty-six patients had no hypoperfusion on SPECT with an average 0.5 seconds of venous delay. Of these, eight went on to have carotid occlusions with no ischemic complications. Patients with mild hypoperfusion on SPECT had an average 0.65 seconds of venous delay and patients with moderate-to-severe hypoperfusion on SPECT had an average of 1.08 seconds of delay. ⁴⁵ These results suggested a cutoff closer to 0.5 seconds in contrast to reports of negative results with up to 3 seconds of venous delay. There is risk of significant inter-rater variability due to the lack of objective assessment of venous delay time on angiograms.

Tanaka F, Nishizawa S, Yonekura Y, et al. Changes in cerebral blood flow induced by balloon test occlusion of the internal carotid artery under hypotension. Eur J Nucl Med 1995;22(11): 1268-1273.

While the adjuncts above are useful in stratifying risk profiles, the role of blood pressure was not taken into account. Pa-terman et al noted an incidence of ischemic strokes in patients following BTO with a symmetric SPECT, related to intraoperative hypotension. ²⁷ Similarly, Eckard et al reported that patients with both symmetric and asymmetric SPECT perfusion patterns presented with delayed strokes. ⁴⁶

Tanaka et al reviewed changes in CBF using SPECT following induced hypotension during a BTO. The study was remarkable for 20 to 40% reduction in ipsilateral CBF during the hypotensive challenge. While the patients remained asymptomatic during the study, the authors identified significant alterations of CBF during hypotension in many patients who would otherwise have normal HMPAO SPECT BTOs. ⁴⁷ Linskey et al similarly noted three risk profiles, including low risk in patients passing both clinical and imaging studies and high risk in patients failing both studies. Moderate-risk patients passed the clinical study but demonstrated asymmetry with Xe perfusion and were noted to have ischemic complications following carotid occlusion. ⁴

Unfortunately, these adjunctive tests lacked specificity in predicting ischemic outcomes. ²¹ Furthermore, these quantitative tests reviewed hemodynamic profile in a controlled environment and may not accurately predict stroke risk during periods of physiologic stress states.