Should CT Angiography be a Routine Component of Acute Stroke Imaging?

This editorial is not an argument against endovascular therapy for acute ischemic stroke. In recent months, 4 clinical trials, MR CLEAN (16 centers in the Netherlands),3 ESCAPE (22 centers worldwide),8 EXTEND-IA (17 centers across Australia and New Zealand),17 and Swift Prime (39 centers in United States and Europe), have made a compelling argument in favor of using stent retrievers to reduce disability in severe ischemic strokes. Several large-volume, endovascular-capable stroke centers have adopted multimodal computed tomography (CT) protocols (including CT angiography [CTA]) as part of their initial workup of the patient with stroke. This editorial should be perceived as a call to the stroke community to develop practical, efficient, and cost-effective protocols to triage patients for endovascular therapy that can be applied expeditiously and on a large scale in hospitals across the United States. Within the 4.5-hour window, IV tPA is the standard of care for ischemic stroke. Intravenous tPA protocols are based on a clinical assessment with a focused neurological examination combined with a noncontrast CT to exclude hemorrhage or a large area of completed infarct18 and do not require vascular evaluation with CTA. Numerous studies have demonstrated better functional outcomes with shorter time to treatment with IV tPA. In a large stroke registry, every 15 minutes of stroke onset to treatment time made a difference in functional outcome.19 Emergency departments across the country are modifying processes to administer intravenous tPA quickly. The addition of routine multimodal CT evaluations adds a few minutes (potentially several minutes in low volume centers) to acute stroke protocols. It is unacceptable for stroke centers to delay tPA administration in favor of obtaining a CTA during initial evaluation. The access to acute stroke therapies is inadequate. In 2011, 63% of hospitals across the United States did not give any IV tPA.20 A meager 7% of hospitals gave ≥1 IV tPA and performed ≥1 thrombectomy for stroke.20 Only 7% of patients with ischemic stroke in the Get with the Guidelines hospitals actually receive IV tPA.21 Although building systems to increase endovascular access is important (and the increased use of CTA to identify patients with large artery occlusion is part of this effort), we cannot afford to lose momentum in our efforts to improve rates and efficiency of intravenous tPA utilization. Of the large pool of patients with acute ischemic stroke, about 20% to 25% of patients with stroke have symptoms from occlusion of a large artery and have the potential to benefit from embolectomy.22 In clinical trials evaluating thrombectomy procedures, large number of patients need to be screened to arrive at the eligible patient. For instance, in the EXTEND-IA study, of the 1044 patients who received IV tPA, only 70 were ultimately deemed eligible for the trial, and the leading reason for exclusion listed by investigators was the absence of a large vessel occlusion.17 Although CTA may provide useful diagnostic information in the other 75% to 80% of patients with stroke without large vessel occlusion, it is rarely critical to acute management. On the other hand, all patients who are examined with CTA are exposed to its risks, which include an approximate doubling of the radiation dose, contrast exposure, and incidental findings such as unruptured aneurysms, thyroid nodules, and lung nodules.23 Emergent CTA protocols require injection of iodinated contrast without a laboratory assessment of renal function. Patients run a risk of developing contrast-induced nephropathy, especially if they have underlying chronic kidney disease (CKD). The incidence of developing this complication is about 2.5% in patients with underlying CKD, but it may be as high as 12% in patients with advanced CKD (glomerular filtration rate <30 mL/min).24 Injection of dye during CTA requires the insertion of a large bore intravenous line. This could be another potential source of delay for patients who qualify for IV tPA. Is there a way to quickly identify those patients who would benefit from thrombectomy without CTA? A high NIHSS bears a strong correlation with the presence of a vessel occlusion. In the 0- to 3-hour window, an NIHSS of 9 or more, and in the 3- to 6-hour window an NIHSS of 7 or more best predicts large vessel occlusion.25 In the 0- to 3-hour window, only 5% of patients with NIHSS <4 had large vessel occlusions.25 If we excluded patients with cortical symptoms such as language disturbance or visual disturbance, and patients with cerebellar signs, this percentage may even be lower. Thus, a rapid clinical examination could identify patients where an emergent CTA would be of low value. There is a concern that a small proportion of patients with a low NIHSS may have a large vessel occlusion and are particularly likely to have early neurological deterioration. In a large series of patients with NIHSS <4, 7.4% had serious early neurological deterioration, about three-quarters of whom had occlusion of a large artery.26 Would identifying these occlusions on emergent CTA, followed by thrombectomy, help these patients? While patients with low NIHSS were included in MR CLEAN (≥2) and EXTEND-IA (≥0), the median NIHSS scores (and interquartile range) of patients who were actually recruited into the endovascular arms of the studies was very high (18 [14-22] in MR CLEAN,3 17 [12-20] in ESCAPE,8 and 17 [13-20] for EXTEND IA17). Therefore, the ideal treatment of patients with low NIHSS is unclear, and whether thrombectomy would benefit them is uncertain from the existing data. Is a CTA truly needed in the patient who presents with a left middle cerebral artery (MCA) syndrome and has a hyperdense left MCA sign? If the ASPECTS score (Alberta Stroke Program Early CT Score) on the initial noncontrast CT is favorable, these patients can be taken to the catheterization laboratory directly. Only 56% of the US population has ground access and 85% of the US population has air access to endovascular capable hospitals within 60 minutes.20 There are vast expanses in the United States where the population with stroke lacks this luxury. In the states of Nebraska, South Dakota, New Mexico, Wyoming, and Alaska, patients with stroke cannot reach an endovascular capable hospital within 60 minutes.20 The majority of all patients with ischemic stroke will be initially evaluated at a hospital that does not have endovascular capability. How does one approach the patient who arrives at the small volume hospital that does not offer endovascular therapy and which may treat a patient with acute ischemic stroke only a few times a month? Over the last decade, telemedicine networks have improved the rate and efficiency of tPA administration in such hospitals. This infrastructure also enables potentially eligible candidates to access endovascular therapies. Protocols need to efficiently identify which patients should be transferred over to central sites for assessment for endovascular therapy. One argument in favor of universal use of CTA is that the information from this study could be used to make this triage decision. However, there are 3 practical drawbacks to implementing routine CTA use. (1) A large number of spoke sites do not possess CTA capabilities. (2) Many spoke sites may not have a large volume of strokes. It would be onerous for CT technicians to maintain levels of technical skill and retraining. Technically inadequate CTA is of no diagnostic value. (3) Even if the spoke institution was capable of multimodal CT imaging, would stroke physicians at the hub site base decisions to treat a dynamically evolving process such as a large vessel ischemic stroke based on imaging obtained an hour or so prior at the spoke site? This would lead to repeat imaging at the hub, as noted in ESCAPE trial, making the results of multimodal imaging at the spoke site redundant and potentially being a source of delay in onset to groin times. In practical experience, triaging is usually performed based on NIHSS and the ASPECTS score on a noncontrast head CT. Further efforts are ongoing to develop clinical bedside tools to identify large artery occlusions, and some have been validated for use even by prehospital personnel.22 In summary, stroke systems should be organized to offer endovascular therapies expeditiously to patients who are most likely to benefit from it. Routine CTA evaluations of all acute strokes may not be the optimal way to approach this. Clinical examination and careful evaluation of the head CT may allow us to target resources and protocols to those most likely to benefit from thrombectomy while minimizing risks such as radiation exposure to those who are least likely to require endovascular therapy.

Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.