Cervical Carotid Pseudo-Occlusions and False Dissections
Intracranial Occlusions Masquerading as Extracranial Occlusions
Background and Purpose—Pseudo-occlusion (PO) of the cervical internal carotid artery (ICA) refers to an isolated occlusion of the intracranial ICA that appears as an extracranial ICA occlusion on computed tomography angiography (CTA) or digital subtraction angiography because of blockage of distal contrast penetration by a stagnant column of unopacified blood. We aim to better characterize this poorly recognized entity.
Methods—Retrospective review of an endovascular database (2010–2015; n=898). Only patients with isolated intracranial ICA occlusions as confirmed by angiographic exploration were included. CTA and digital subtraction angiography images were categorized according to their apparent site of occlusion as (1) extracranial ICA PO or (2) discernible intracranial ICA occlusion.
Results—Cervical ICA PO occurred in 21/46 (46%) patients on CTA (17 proximal cervical; 4 midcervical). Fifteen (71%) of these patients also had PO on digital subtraction angiography. A flame-shaped PO mimicking a carotid dissection was seen in 7 (33%) patients on CTA and in 6 (29%) patients on digital subtraction angiography. Patients with and without CTA PO had similar age (64.8±17.1 versus 60.2±15.7 years; P=0.35), sex (male, 47% versus 52%; P=1.00), and intravenous tissue-type plasminogen activator use (38% versus 40%; P=1.00). The rates of modified Treatment In Cerebral Ischemia 2b-3 reperfusion were 71.4% in the PO versus 100% in the non-PO cohorts (P<0.01). The rates of parenchymal hematoma, 90-day modified Rankin Scale score 0–2, and 90-day mortality were 4.8% versus 8% (P=0.66), 40% versus 66.7% (P=0.12), and 25% versus 21% (P=0.77) in PO versus non-PO patients, respectively. Multivariate analysis indicated that PO patients had lower chances of modified Treatment In Cerebral Ischemia 3 reperfusion (odds ratio 0.14; 95% confidence interval 0.02–0.70; P=0.01).
Conclusions—Cervical ICA PO is a relatively common entity and may be associated with decreased reperfusion rates.
Endovascular therapy has become the gold standard for large vessel occlusion strokes, with time and reperfusion being important predictors of outcome after acute ischemic stroke.1 Accordingly, any condition that potentially delays or prevents reperfusion has the potential to negatively impact patient care.
Cervical internal carotid artery (ICA) pseudo-occlusion (PO) is a described entity where flow-related artifact leads to an appearance of complete cervical ICA occlusion on computed tomographic angiography (CTA) or digital subtraction angiography (DSA), where in actuality the extracranial ICA is patent.2 This finding is likely because of rapid arterial phase imaging missing delayed images or because of blockage distal contrast penetration by a stagnant column of unopacified blood. The presence of PO can be confirmed by performing superselective microcatheter angiography distal to the PO site, which will have backfilling of contrast in the extracranial ICA, proving its patency. The finding of cervical ICA PO has the potential to impact care because of its misdiagnosis as a tandem extracranial–intracranial ICA occlusion or dissection and lead to the erroneous exclusion of isolated intracranial ICA patients from clinical trials.3 We aim to describe this phenomenon and evaluate its impact on patient treatments and outcomes.
Retrospective analysis of a prospectively collected acute ischemic stroke database for consecutive cases of endovascular therapy between September 2010 and December 2015. We identified all patients with intracranial ICA occlusions with preprocedural CTAs. Patients with evidence of steno-occlusive atherosclerotic disease or dissection leading to cervical ICA occlusion (or near-occlusion) during any point of the DSA were excluded. After crossing the cervical segment, microcatheter angiograms were performed in cases where there was a question of extracranial ICA patency, despite long-acquisition DSA, and used to confirm isolated intracranial occlusion. CTA/DSA images were then reviewed by 2 neurointerventionalists and categorized according to their apparent site of occlusion as (1) extracranial ICA PO or (2) discernible intracranial ICA occlusion. We also evaluated the morphology of the PO for the classic flame-shaped occlusion, characteristic of carotid dissection.4
Reperfusion was graded according to the modified Treatment In Cerebral Ischemia scale.5 Parenchymal hematoma was defined per European Cooperative Acute Stroke Study criteria (parenchymal hematoma-1/parenchymal hematoma-2).6 The Institutional Review Board approved the study.
Continuous variables are reported as mean±SD or median (intraquartile range). Categorical variables are reported as proportions. Between groups, comparisons for continuous/ordinal variables were made with Student t test, Mann–Whitney U test, or analysis of variance. Categorical variables were compared by Chi-square or Fisher exact test. Significance was set at P<0.05. Multivariate logistic regression analyses for factors associated with CTA PO was performed for variables at the 0.2 level of significance on univariate analysis using IBM SPSS Statistics 21.
Of 898 patients who underwent endovascular therapy, 173 patients were noted to have intracranial ICA occlusion, and 46 of these patients met the inclusion criteria of not having cervical dissection or atherosclerotic disease. Cervical ICA PO was found on CTA in 21 patients (46%): 17 had proximal and 4 had midcervical ICA PO. Fifteen (33%) of the total patient cohort were also found to have PO on pre-thrombectomy DSA. All patients with PO on DSA also had PO on CTA. Flame-shaped imaging mimicking carotid dissection was seen on CTA in 7 (33%) and on DSA in 6 (29%) of the PO patients.
Table shows the comparison between CTA PO and discernible intracranial occlusion groups. Baseline characteristics were similar. Merci retriever use was nonstatistically higher in the discernible intracranial occlusion group (20% versus 4.8%; P=0.12). The modified Treatment In Cerebral Ischemia 2b-3 and modified Treatment In Cerebral Ischemia 3 reperfusion rates were higher in the discernible intracranial occlusion group (100% versus 71%; P<0.01, and 60% versus 23%; P=0.01). The rates of good functional outcomes (90-day modified Rankin Scale score 0–2), parenchymal hematoma, and 90-day mortality were similar between the groups.
The multivariate model indicated an independent association between dependent variable of CTA PO and the independent variables of lower rate of modified Treatment In Cerebral Ischemia 3 reperfusion (odds ratio 0.14;95% confidence interval [0.029–0.706];P=0.017) and lower use of Merci retriever (odds ratio 0.076; 95% confidence interval 0.007–0.835; P=0.035). The chance of achieving modified Rankin Scale score 0–2 at 90 days (odds ratio 0.60; 95% confidence interval 0.13–2.68; P=0.50) was not independently associated with CTA PO. There were no radiographic or clinical predictors of CTA PO found.
The concept of carotid PO on CTA because of discrepancy between the timing of image acquisition and flow was first described in a study on computed tomography image quality.2 The authors analyzed 100 consecutive CTAs for patients with acute ischemic stroke and found 5 cases of carotid PO.2 Half of these patients had only an intracranial occlusion without extracranial stenosis/occlusion. All patients were able to be deciphered as not having a cervical ICA occlusion on delayed contrast-enhanced computed tomography.2 Although delayed CTA, multiphase CTA, or carotid ultrasonography may improve diagnostic accuracy, we have demonstrated that PO remains a significant issue even after long-acquisition DSA, with angiographic microcatheter exploration being often required for the proper diagnosis.
Other authors have advocated the use of angiography to diagnose PO,7 but in our series, 71% of the patients continued to have a PO on angiogram. This is related to a delay or absence of contrast opacification because of a downstream occlusion, making microcatheter exploration the only way to accurately diagnose the occlusion location.
Other studies have looked at PO in comparison to true cervical ICA occlusion. One study identified 6 POs in a series of 54 acute ischemic stroke patients with occluded cervical ICAs.8 In 4 of these cases, the ultimate DSA finding was isolated intracranial ICA occlusion. Given the small series sizes, the authors only describe the imaging finding, but are unable to give any conclusions regarding incidence or outcomes.
We describe the largest published series of endovascularly treated carotid POs and their potential impact on the rate of recanalization. We found that nearly half of isolated intracranial occlusions presented as CTA cervical ICA POs. Furthermore, nearly a third of the intracranial occlusions also presented as POs on both DSA and CTA, which points to the slow speed or absence of anterograde flow.
In addition, 7 of our patients had preprocedure CTAs and 6 patients had initial DSAs where the PO was flame shaped, which has been described as characteristic of dissection.4 This further proves the importance of microcatheter exploration, given that the treatment of a dissection is different than that of an isolated thrombus (Figure). PO patients had a trend toward longer procedure times, possibly because of the need for additional angiographic evaluation.
The rate of full reperfusion was worse in PO patients in our study. We also saw a more frequent use of the Merci retriever in the group with discernible intracranial occlusions on the CTA. Although this is related to the products on the market at the time of the procedure, it could represent a potential bias. However, given that stent retrievers are associated with higher recanalization rates versus Merci retrievers,9 this strengthens the argument that POs may be more challenging to treat. This may be related to a larger thrombus burden occluding the outflow of intracranial ICA branches. This situation may also promote greater adherence and compaction of the clot by the water-hammering effect of arterial flow.
Our study has limitations, mainly related to its retrospective design and relative small sample size. The small sample size was particularly limiting in our statistical analysis and could have potentially masked correlative variables. Given the reliance on microcatheter angiogram, which may not have been performed or documented in all cases, there is a potential of ascertainment bias.
Cervical ICA PO is a common entity in the large vessel occlusion strokes population, and its presence seems to be associated with a decreased rate of reperfusion. Treating physicians should be familiar with its presence to optimize procedural planning.
- Received June 22, 2016.
- Revision received November 7, 2016.
- Accepted November 28, 2016.
- © 2017 American Heart Association, Inc.
- Goyal M,
- Menon BK,
- van Zwam WH,
- Dippel DW,
- Mitchell PJ,
- Demchuk AM,
- et al
- Kim JJ,
- Dillon WP,
- Glastonbury CM,
- Provenzale JM,
- Wintermark M
- Zaidat OO,
- Yoo AJ,
- Khatri P,
- Tomsick TA,
- von Kummer R,
- Saver JL,
- et al
- Berger C,
- Fiorelli M,
- Steiner T,
- Schäbitz WR,
- Bozzao L,
- Bluhmki E,
- et al
- Nogueira RG,
- Lutsep HL,
- Gupta R,
- Jovin TG,
- Albers GW,
- Walker GA,
- et al