Graefes Arch Clin Exp Ophthalmol DOI 10.1007/s00417-014-2804-2

RETINAL DISORDERS

Distribution of internal carotid artery plaque locations among patients with central retinal artery occlusion in the Eagle study population C. Leisser & T. A. Kaufmann & N. Feltgen & M. Schumacher & C. Schmoor & S. Meckel

Received: 13 July 2014 / Revised: 3 September 2014 / Accepted: 8 September 2014 # Springer-Verlag Berlin Heidelberg 2014

Abstract Purpose Arterial emboli in the internal carotid artery (ICA) mainly cause cerebral ischemia; only 10 % of emboli reach the retinal arteries. Computational blood flow studies suggest that plaques situated in the ICA siphon may be a source of embolism to the ophthalmic artery (OA). To validate these calculated probabilities in patients with central retinal artery occlusion (CRAO), we reanalyzed digital subtraction angiography (DSA) images from the Multicenter Study of the European Assessment Group for Lysis in the Eye (EAGLE) study, a multicenter randomized study in patients with nonarteritic CRAO. Methods A reevaluation of 34 DSA studies was done from the interventional arm of the EAGLE study with regards to distribution of arterial plaques at specific ICA siphon locations and ICA stenosis. A comparison was made of plaque distribution to calculated probabilities for emboli reaching the OA

from a computational fluid dynamics (CFD) model of a patient-specific ICA siphon. Results Most of the ICA plaques near the OA’s origin were located in the cavernous ICA portion (31.3%). Of these, 12.5 % had plaques in the curvature opposite the OA origin, a location carrying the highest risk for embolization into the OA (according to the CFD model 12.6–13.2 % probability of embolisation into the OA). Also, 15.6 % had plaques in the paraclinoid ICA portion distal to the OA origin. Conclusions There were 40.6% of the patients that had plaques in the cavernous and clinoid ICA portions presenting possible sources for embolic material generating RAO. Keywords Retinal artery occlusion . EAGLE study . Plaques of the internal carotid artery . Computational blood flow studies

Introduction C. Leisser (*) Department of Ophthalmology, Heinrich-Collinstrasse 30, A-1140 Hanuschkrankenhaus, Vienna e-mail: [email protected] T. A. Kaufmann Department of Cardiovascular Engineering, Institute of Applied Medical Engineering, Helmholtz Institute, RWTH Aachen University, Aachen, Germany N. Feltgen Department of Ophthalmology, University Goettingen, Goettingen, Germany M. Schumacher : S. Meckel Department of Neuroradiology, University Hospital Freiburg, Freiburg, Germany C. Schmoor Clinical Trials Unit, University Hospital Freiburg, Freiburg, Germany

A recently presented computational fluid dynamics (CFD) model showed that most emboli in the internal carotid artery (ICA) originating from cardiac valves or carotid plaques are washed into the brain [1, 2]. Only 10 % of emboli in the ICA reach the retinal arteries. This would imply that most patients with retinal artery occlusion (RAO) might have had a history of clinical or silent embolic cerebral ischemia. On the contrary, the current literature reports stroke rates up to 20 % among RAO patients [3–5] and the 3-year risk of stroke after central retinal artery occlusion (CRAO) is only doubled [6]. As Hayreh et al. [7] reported plaques of the ICA being the main source for emboli causing RAO, this reevaluation of DSA images of the EAGLE study should give insight on the distributions of plaques in the ICA in that specific group of patients.

Graefes Arch Clin Exp Ophthalmol

Methods The EAGLE study [8, 9] was a prospective randomized controlled multicenter study, which recruited 84 patients with nonarteritic CRAO; 40 patients were randomized in the conservative arm and 44 in the interventional arm with intraarterial fibrinolysis. The study design of the EAGLE study followed the tenets of the Declaration of Helsinki; informed consent was obtained from all subjects and the study was approved by the ethic committees of the participating centers. Patients in the conservative arm did not undergo angiographic examination of the ICA. From the interventional arm, seven patients did not receive intraarterial fibrinolysis (four patients due to technical failure, three patients due to failure of the center to follow the randomization procedure) as reported earlier [9]. From the remaining subjects, digital subtraction angiography (DSA) studies were available for reevaluation in 34 patients. Analysis of the intracranial portions was not possible in two of these patients due to inadequate visualization (aortic arch angiogram only) and ipsilateral ICA occlusion, respectively, leaving 32 patients for reevaluation of the intracranial ICA. In another three patients the carotid artery bifurcation was not visualized, which ruled out evaluation of any proximal ICA stenoses or plaques; these patients were included in the reevaluation of intracranial ICA plaques. The images were evaluated by a neuroradiologist with longstanding expertise in DSA studies, including the presence and extension of vascular plaques in the cavernous, clinoid, and paraclinoid ICA portions as potential sources of embolism into the ophthalmic artery (OA). We focussed especially on plaques located within the inner curvature of the ICA siphon opposite the OA. Furthermore, the presence, location, and degree of ICA stenosis were noted. To assess clinical relevance of our findings, CFD calculations of blood flow within the carotid siphon in the vicinity of the OA origin were performed. These were based on a realistic anatomy derived from TOF-MRA data that was performed in a CRAO patient having undergone MRI and MRA because of cerebral ischemia. Probabilities of particle behaviour at the origin of the ophthalmic artery were calculated by 11 CFD calculations with 1500 particles in each calculation. Further technical details of this CFD application were described previously by Kaufmann et al. [1, 2]. These calculations provided insight on the distribution of wall shear stress (WSS) along the carotid siphon as well as probabilities of embolism into the OA from plaques at specific locations within the ICA anatomy.

patients (31.3 %) displayed plaques in the cavernous ICA segment proximal to the OA origin (Fig. 1), and five patients (15.6 %) had plaques in the paraclinoid ICA segment distal to the OA origin, one of which was mildly stenosing the ICA (3.1 %). Two patients (6.5 %) had high-grade stenosis of ICA origin. The ICA was occluded in one patient (3.1 %), two patients (6.5 %) showed low-moderate grade stenosis of the proximal ICA, and nine patients (29 %) had nonstenosing plaques at the ICA origin. Overall, 40.6 % of patients had nonstenosing distal intracranial ICA plaques within the carotid siphon close to the OA origin, excluding those located distal the OA origin; 45.2 % of patients had proximal ICA plaques at or near the ICA origin, including those with or without stenoocclusive disease of the ICA origin. According to the CFD calculations, the probability of embolism into the OA originating from plaques situated within the ICA siphon opposite the OA origin was 13.2+/−6.9 %. In case of a further proximal stenosis of the ICA, the probability of embolization into the OA from plaques opposite the OA origin was 12.6+/−7.7 %. In contrast, emboli originating from the cardiac valves, ascending aorta, and proximal carotid plaques displayed a risk of embolism into the OA of 7.3+/−1.1 % and 10.7+/−3.6 % in cases of an ICA stenosis, respectively. The higher rates of OA embolism from the location of plaques opposite the OA origin was attributed to helical blood flow within the carotid siphon. The probability of emboli within the laminar flow that originate from ICA plaques in the vicinity of the OA origin, i.e., within the cavernous and clinoid ICA portions was 11.1 +/−2.2 % (Table 1).

Discussion Hayreh [7] postulated ICA plaques being the major embolic source of RAO and found the presence of ICA plaques in 65– 70 % of patients, though not distinguishing between proximal

Results Four patients (12.5 %) presented plaques in the clinoid segment of the ICA siphon directly opposite the OA origin. Ten

Fig. 1 DSA image showing multiple plaques (arrows) located within the ICA siphon on the inner and outer curvature of the cavernous segment proximal to the OA origin

Graefes Arch Clin Exp Ophthalmol Table 1 Probabilities (in % +/- standard deviation) of emboli from plaques opposite the OA origin or ascending emboli for reaching the OA (cardiac emboli, emboli from plaques of the aortic arch, or proximal carotid artery), based on a realistic patient’s anatomy with CRAO and 11 CFD calculations with 1500 particles in each calculation

ICA siphon opposite the OA origin ascending emboli

No ICA stenosis

Proximal ICA stenosis

13.2 %+–6.9 %

12.6 %+–7.7 %

7.3 %+–1.1 %

10.7 %+–3.6 %

ICA plaques and distal intracranial plaques in the vicinity of the OA origin. In our reevaluation of the EAGLE study data, distal intracranial ICA plaques located within the ICA siphon were found in 45.2 % of nonarteritic CRAO patients. Among those, 40.6 % of patients had plaques in the cavernous and clinoid ICA portions proximal to or just at the OA origin portions, including the area of highest arterial WSS in the region near the OA origin (Fig. 2). CFD calculations [1, 2] showed that embolic material in the laminar flow (originating more likely from plaques in the vicinity of the OA origin, than the more proximal ICA plaques) had a higher probability to be washed into the OA than did the embolic material from cardiac valves, the aortic arch, or more proximal ICA plaques. Nevertheless, the highest probability for OA embolism was calculated from plaques in the clinoid ICA portions directly opposite the OA origin (Table 1).

Apart from ICA-plaques, other potential sources for ascending thrombotic material are vegetation of the cardiac valves and atrial fibrillation or plaques of the aortic arch. Vegetation of the cardiac valves, for example, could be found on transesophageal echocardiography in 4.3 to 4.6 % of patients with RAO, atrial fibrillation in 9.4 to 10 %, and plaques of the aortic arch in 18.8 to 19.1 % [10, 11]. According to the abovementioned CFD studies, those emboli reach the ophthalmic artery in 7.3 to 10.7 % of patients. As these potential embolic sources are present in a minority of patients with RAO, in accordance with Hayreh [7], plaques of the ICA can be regarded as the major source for embolic material causing RAO. However, a history of stroke among patients with RAO is only present in about 20 % [3–5], a fact that leaves open further questions about the etiology of RAO, potentially having different pathomechanisms than stroke for some patients. Whether retrograde embolism from plaques in the ICA siphon located distal of the OA origin (15.6 % of patients) could be a cause of RAO remains unclear to us. Like the flow hypothesis of retrograde embolism from aortic arch plaques in ischemic stroke [12, 13], ICA plaques from distal of the OA origin might be a potential cause of RAO due to retrograde helical blood flow patterns within the bent anatomy of the carotid siphon. This new hypothesis requires further research on the specific ICA blood flow patterns. There are several limitations to the CFD study:, a) the simulations were performed considering rigid walls, thus, the movement of the arterial walls was neglected; b) the inlet profile in the carotid artery was based on literature values that might not represent the specific patient case; c) the blood clots were assumed to be spherical: and d) simulation of the blood flow could not be performed until the central retinal artery. This was due to inadequate visualisation of the central retinal artery in our TOFMRI data.

Conclusion

Fig. 2 Computational reconstruction of the ICA siphon including the OA origin that indicates WSS in different colors. The area in red reveals the location with the highest WSS (232Pa versus 15Pa in other ICA regions)

In this retrospective analysis of nonarteritic CRAO patients derived from the EAGLE study, 45.2 % of this cohort had distal intracranial ICA plaques located within the ICA siphon close to the OA origin with or without steno-occlusive disease of the ICA origin. Among those, 40.6 % presented with plaques in the cavernous and clinoid ICA portions that demonstrated a higher probability of embolism into the OA than other embolic sources for RAO in a patient-specific CFD analysis.

Graefes Arch Clin Exp Ophthalmol Acknowledgments We thank the collaborators of the EAGLE study [8, 9] for generously allowing us to reevaluate 34 of their patients. Conflict of interest The authors declare that they have no conflict of interest.

References 1. Kaufmann TA, Leisser C, Schmitz-Rode T, Steinseifer U (2012) Computational analysis of blood flow and emboli to understand retinal artery occlusion. ASAIO J 58(7):4 2. Kaufmann TA, Leisser C, Gemsa J, Steinseifer U (2014) Analysis of emboli and blood flow in the ophthalmic artery to understand retinal artery occlusion. Biomed Tech (Berl). doi:10.1515/bmt-2014-0002 3. Leisser C (2010) Hyperlipidaemia and increased systemic blood pressure–is there a connection to retinal artery occlusion? Spektrum Augenheilkd 24(3):154–156 4. Wang JJ, Cugati S, Knudtson MD, Rochtchina E, Klein R, Klein BE, Wong TY, Mitchell P (2006) Retinal arteriolar emboli and long-term mortality: pooled data analysis from two older populations. Stroke 37(7):1833–1836 5. Leisser C (2012) Risk factors of retinal artery occlusion. Spektrum Augenheilkd 26(4):202–204 6. Chang YS, Jan RL, Weng SF, Wang JJ, Chio CC, Wei FT, Chu CC (2012) Retinal artery occlusion and the 3-year risk of stroke in Taiwan: a nationwide population-based study. Am J Ophthalmol 154(4):645–652

7. Hayreh SS, Podhajsky PA, Zimmerman MB (2009) Retinal artery occlusion: associated systemic and ophthalmologic abnormalities. Ophthalmology 116(10):1928–1936 8. Feltgen N, Neubauer A, Jurklies B, Schmoor C, Wanke J, MaierLenz H, Schumacher M, EAGLE-Study Group (2006) Multicenter study of the European assesment group for lysis in the eye (EAGLE) for the treatment of central retinal artery occlusion: design issues and implications. EAGLE Study report no. 1. Graefe’s Arch Clin Exp Ophthalmol 244(8):950–956 9. Schumacher M, Schmidt D, Jurklies B, Gall C, Wanke I, Schmoor C, Maier-Lenz H, Solymosi L, Brueckmann H, Neubauer AS, Wolf A, Feltgen N, EAGLE-Study Group (2010) Central retinal artery occlusion: local intra-arterial fibrinolysis versus conservative treatment, a multicenter randomized trial. Ophthalmology 117(7):1367–1375 10. Leisser C (2014) Risk factor analysis in patients with retinal artery occlusion with respect to transesophageal echocardiography. Spectrum Augenheilkd. doi:10.1007/s00717-014-0218-5 11. Leisser C (2014) Are there differences between plaques of the internal carotid artery and aortic arch among patients with retinal artery occlusion and anterior ischemic optic neuropathy? Klin Monatsbl Augenheilkd. doi:10.1055/s-0034-1368574 12. Harloff A, Strecker C, Dudler P, Nussbaumer A, Frydrychowicz A, Olschewski M, Bock J, Stalder AF, Stroh AL, Weiller C, Hennig J, Markl M (2009) Retrograde embolism from the descending aorta: visualisation by multidirectional 3d velocity mapping in cryptogenic stroke. Stroke 40(4):1505–1508 13. Harloff A, Simon J, Brendecke S, Assefa D, Helbing T, Frydrychowicz A, Weber J, Olschewski M, Strecker C, Henning J, Weiller C, Markl M (2010) Complex plaques in the proximal descending aorta: an underestimated embolic source of stroke. Stroke 41(6):1145–1150

Distribution of internal carotid artery plaque locations among patients with central retinal artery occlusion in the Eagle study population.

Arterial emboli in the internal carotid artery (ICA) mainly cause cerebral ischemia; only 10 % of emboli reach the retinal arteries. Computational blo...
226KB Sizes 0 Downloads 7 Views