Article Text

Original research
Intravascular optical coherence tomography for the evaluation of arterial bifurcations covered by flow diverters
  1. Christina Iosif1,2,
  2. Suzana Saleme1,2,
  3. Sebastien Ponsonnard3,
  4. Pierre Carles4,
  5. Eduardo Pedrolo Silveira1,
  6. Eduardo Waihrich1,
  7. Gilles Trolliard4,
  8. Catherine Yardin2,5,
  9. Charbel Mounayer1,2
  1. 1Interventional Neuroradiology Department, Dupuytren University Hospital (CHU Limoges), Limoges, France
  2. 2Applied Medical Research Team (ERMA), University of Limoges, Limoges, France
  3. 3Anesthesiology Department, Dupuytren University Hospital (CHU Limoges), Limoges, France
  4. 4Departments of Science of Ceramic Processes and Surface Treatments, CNRS, UMR 7315, European Ceramic Center, University of Limoges, France
  5. 5Department of Histology, Cytology, Cellular Biology and Cytogenetics, Mother and Child (HME) University Hospital, Limoges, France
  1. Correspondence to Dr Christina Iosif, Interventional Neuroradiology Department, Dupuytren University Hospital (CHU Limoges), 2 Avenue Martin Luther King, Limoges 87042, France; christina.iosif{at}, christinaiosif{at}


Background and objective Due to its high spatial resolution, intravascular optical coherence tomography (OCT) has been used as a valid method for in vivo evaluation of several types of coronary stents at straight lumen and bifurcation sites. We sought to evaluate its effectiveness for flow diverting stents deployed in arterial bifurcation sites involving jailing of a side branch.

Methods Four large white swine were stented with flow diverting stents covering the right common carotid artery–ascending pharyngeal artery bifurcation. After 12 weeks of follow-up the animals were evaluated by digital subtraction angiography and intravascular OCT and subsequently sacrificed. Neointimal thickness on the parent arteries and the free segments of the stent were measured. The stented arteries were harvested and underwent scanning electron microscopy (SEM) imaging. Ostia surface values were measured with OCT three-dimensional (3D) reconstructions and SEM images.

Results All endovascular procedures and OCT pullback runs were feasible. Stent apposition was satisfactory on the immediate post-stent OCT reconstructions. At 12-week controls, all stents and jailed branches were patent. Mean neointimal thickness was 0.11±0.04 mm on the free segments of the stent. The mean ostia surface at 12 weeks was 319 750±345 533 μm2 with 3D-OCT reconstructions and 351 198±396 355 μm2 with SEM image-derived calculations. Good correlation was found for ostia surface values between the two techniques; the values did not differ significantly in this preliminary study.

Conclusions Intravascular OCT appears to be a promising technique for immediate and follow-up assessment of the orifice of arterial branches covered by flow diverting stents.

  • Catheter
  • Flow Diverter
  • Material
  • Stent
  • Vessel Wall
View Full Text

Statistics from


Intravascular optical coherence tomography (OCT) has shown interesting results regarding in vivo evaluation of cardiac stent apposition and neointimal growth owing to its high spatial resolution.1 From drug eluting stents2 ,3 to bioabsorbable scaffolds,4 this technique is able to evaluate stent apposition in vivo, allowing for balloon correction of the result if necessary and for monitoring of the neointimal coverage of the device. The latter has proved useful in anti-aggregation regimen management, especially in patients with comorbidities.5

Flow diverting (FD) stents are devices with very dense braiding of metallic struts with much thinner diameters6 compared with cardiac and conventional neurovascular stents. Their particular mechanical properties are responsible for their mechanism of action, provoking a significant reduction in blood inflow in the treated aneurysm sac.7 However, they also present some technical and clinical challenges. The fine struts are not easily evaluated in vivo by three-dimensional rotational angiography (3DRA) or digital subtraction angiography (DSA), and jailing arterial branches may result in their narrowing or occlusion.8 Moreover, it seems from clinical observations that, in some cases, emboli may be caused by interference of the dense metallic struts at the level of jailed ostia.9

Intravascular OCT has already been successfully used in some clinical cases with very interesting results. Moreover, a recent animal study proposed a valid grading method for the evaluation of stent apposition,10 providing an important advancement in the field. Nevertheless, the effectiveness of intravascular OCT specifically on quantitative evaluation of jailed bifurcations by FD stents is still the subject of research. In an attempt to verify the effectiveness of the technique and provide useful data regarding neointimal coverage at the level of the jailed ostia in FD stenting, an animal study was conducted.


Four large white swine aged 12 weeks (two male and two female) were used; their mean weight on the day of the procedure was 20.3±1.2 kg. The animals were followed up for 12 weeks before euthanasia.

The institutional ethical committee for animal care and experiments on animals approved the study and validated adherence to the national guidelines and regulations for the care and use of laboratory animals.

Anesthesia and animal care

Premedication with aspirin (10 mg/kg orally) and clopidogrel (10 mg/kg orally) for 3 days was applied to all animals before endovascular stenting and this regimen was maintained throughout the follow-up period. Stenting procedures were performed under general anesthesia (premedication with intramuscular 20 mg/kg ketamine and 2 mg/kg xylazine, intubation, propofol and sevoflurane11 to maintain anesthesia). Euthanasia was performed after 12 weeks of follow-up with a barbiturate overdose (sodium pentobarbital 390 mg plus sodium phenytoin 50 mg/mL, 0.22 mL/kg intravenously) in animals anesthetized with propofol.

Endovascular stenting technique and DSA control

The animals were randomly allocated to two groups: group A had anastomotic type of circulation for the right ascending pharyngeal artery (APhA) and group B had terminal type of circulation for the same branch. All animals were stented at the level of the right common carotid artery (CCA)–APhA bifurcation under general anesthesia. Procedures and follow-up were performed on a biplane flat-panel DSA unit (Allura Xper FD20; Philips, Eindhoven, The Netherlands). Femoral puncture was performed using the Seldinger technique after local percutaneous anesthetic injection (lidocaine) and was followed by the introduction of a 6 F percutaneous right femoral introducer sheath (Radifocus Introducer II; Terumo Medical Corporation, Somerset, New Jersey, USA).

For animals allocated to group B, endovascular coiling and N-butyl-cyanoacrylate glue (Glubran; Gem Srl, Viareggio, Italy) of the left APhA and anastomotic ramus preceded stenting by FD stents in order to simulate the terminal type of circulation for the right APhA. In this way, two groups of animals were created, one with competitive flow in the right APhA (group A) and one with terminal type of arterial flow (group B).

Endovascular stenting with an FD stent (Pipeline Embolization Device; Covidien-Medtronic, California, USA) was performed for all animals through a coaxial microcatheter system (6 F Navien guiding catheter, 2.8 F/3.2 F Marksman microcatheter; Covidien-Medtronic); the stent was deployed in the right CCA in order to cover the ostium of the right APhA.

FD stent dimensions and choice

During endovascular stenting, 3DRA was performed in each animal by selective injection through the guiding catheter at the level of the right CCA. The diameters of the CCA and external carotid artery (ECA) were measured using this technique and stents of the appropriate diameter were chosen for the corresponding arterial segment in order to obtain similar pore opening for every specimen. In this way, all ostia were covered by stent pores featuring similar angles of the braided struts and similar metal coverage.

Immediate OCT pullback controls

Intravascular OCT (Saint Jude Medical, St Paul, Minnesota, USA) was used before and immediately after each FD stent deployment. An optical imaging monorail catheter (2.7 F, Dragonfly Duo; Saint Jude Medical) was navigated with a 0.014 inch guidewire through the 6 F guiding catheter. Cross-sectional images were generated with a motorized pullback technique during contrast medium injection with acquisition of 180 images/s. Pullback length was 5 cm, with the bifurcation site in the middle of the pullback region, 7 mL/s contrast medium injection rate; each pullback lasted for 4 s. Cross-sectional and 3D images were post-treated in a dedicated console (Ilumien Optis; Saint Jude Medical).

Twelve-week intravascular OCT controls

All animals underwent percutaneous (by right femoral access) selective DSA controls under general anesthesia, including 3DRA and OCT pullback run, with the same material and acquisition parameters as during endovascular stenting. Primary endpoints were:

  1. Detection of neointimal formation on the free segments of the stent (FSS) and neointimal coverage of the FD stent covering the parent artery (PA); neointimal thickness measurements were performed in six concentric locations for the FSS and six for the PA for each animal on cross-sectional images at the level of the right CCA–APhA bifurcation and mean values were calculated for each animal.

  2. Evaluation of the 3D reconstructions of the jailed arterial ostia and correlation with the corresponding scanning electron microscopy (SEM) image results.

  3. Quantification of ostia surfaces at 12 weeks with images reconstructed by the OCT pullback raw data.

  4. Comparison of the ostia surface values obtained in vivo at 12 weeks by OCT with those obtained by SEM images after harvesting. The ostium surface for each artery was defined as the total patent area or surface through which blood was able to flow at 3 months, as visualized by OCT reconstructions or by SEM images.

Sample harvesting and SEM

After 12 weeks of follow-up, arteries and FD stents were harvested, chemically fixed with glutaraldehyde in Sorensen's buffering and osmium tetroxide followed by critical point drying and longitudinal cutting. A plasma metallizator (AGAR sputter coater B7340) was used to cover the opened samples with 12 nm platinum.12 The inner surfaces of the opened arteries were observed and photographed with SEM (JEOL JSM-7400F). These images of the stented ostia were used to calculate the ostia surfaces at 12 weeks. Measurement was performed by hand selection of each SEM image with open source software (Image-J1, NIH, USA).13

Statistical analysis

Descriptive statistics were applied using the Student t test after normality verification with the Kolmogorov–Smirnov test. Independent variables were tested for equal variances with the F test and subsequently compared statistically using the Student t test or the Welch test, accordingly. Paired variables were compared either by the paired-samples t test or the Wilcoxon test (in cases of non-normal distribution). Each cross-sectional image was treated as an independent data point for neointimal thickness analysis; 24 measurements (12 each for groups A and B) were used for the neointimal thickness on the FSS of the ostia and 24 measurements for the neointimal thickness of the PAs on the same images. The level of statistical significance was ≤0.05. The electronic statistical package used was Statistica (StatSoft, Tulsa, Oklahoma, USA).


Endovascular procedures and intravascular OCT pullback runs were successfully carried out in all four animals. Absence of device migration, device fracture or deformation was confirmed by DSA during the follow-up period. Mean CCA diameter was 4.84±0.3 mm (95% CI 4.37 to 5.31), mean ECA diameter was 4.47±0.18 mm (95% CI 4.19 to 4.76), and mean APhA diameter was 2.29±0.22 mm (95% CI 1.95 to 2.65) on the day of stenting. Two FD stents of 5×20 mm, one stent of 4.5×20 mm, and one of 4.75×20 mm were used. All animals maintained normal weight gain curves during the follow-up period and yielded satisfactory results on veterinary examinations.

Immediate pullback controls

All immediate pullback runs were successfully performed. Within 2–4 min of post-processing, the region of interest was available in cross-sectional and 3D images in each case. Stent apposition on bifurcation sites was feasible, with the stent struts being easily detectable on the cross-section images. Visualization of the initial part of the jailed artery, 2–3 cm distal to the FD stent, allowed for the evaluation of acute thrombus formation on the stent struts and inside the jailed artery. Minimum thrombus on the stent struts in one case was resolved with in situ administration of abciximab. Lysis of the thrombus was verified with an additional OCT pullback run.

Twelve-week OCT and DSA controls

At 12 weeks, 3DRA runs were performed followed by selective DSA runs and eventually intravascular OCT pullback runs. All four cases were successfully evaluated and jailed arteries were patent in all cases. Neointimal formations on the FSS were clearly visible on cross-sectional images; the stent was visible on the PA and the neointimal lining was easily discerned (figure 1).

Figure 1

Case No 3. Digital subtraction angiography, selective contrast medium injection from the right common carotid artery (CCA) at working projection before stenting (A) (notice the undeployed stent still sheathed inside the microcatheter in the carotid lumen), (B) immediately after stent deployment (arrows show the borders of the deployed flow diverter (FD)) and (C) at 12 weeks control (notice the position of the three markers of the Dragonfly optical coherence tomography (OCT) microcatheter (thick arrows) before the initiation of the pullback with the ascending pharyngeal artery (APhA) ostium in the middle of the two proximal markers. (D) Cross-sectional image of the pullback run at the level of the APhA ostium (the lumen of the APhA is shown with an asterisk) showing the neointimal formation on the free segments of the stent (arrow); notice the OCT microcatheter (arrowhead) inside the right CCA lumen. (E) Three-dimensional OCT reconstruction showing the patent APhA ostium at 12 weeks and (F) scanning electron microscopic image of the same ostium after euthanasia and harvesting (notice the similarity in appearance of the two images). The quantification of this ostium with OCT resulted in a surface value of 753 500 μm2; with the scanning electron microscopy image the corresponding value was 873 506 μm2.

Neointimal thickness was measurable during the procedure on the dedicated reconstruction console. Mean neointimal thickness was 0.18±0.07 mm (95% CI 0.15 to 0.21) on the PAs and 0.11±0.04 mm (95% CI 0.09 to 0.13) on the ostia (FSS) (table 1). The ratio of neointimal thickness on the FSS to the neointimal thickness on the PA was 0.65. Neointimal thickness values on the FSSs were lower than on the corresponding PAs (paired sample t test: mean difference: 0.07083, SD 0.07506, 95% CI 0.03914 to 0.1025, test statistic t: 4.623, degrees of freedom (df): 23, two-tailed probability p=0.0001).

Table 1

Baseline characteristics (arterial diameters prior to FD stent deployment), FD dimensions, and mean neointimal thickness calculations at 12 months post stenting for the four animals performed by cross-sectional pullback OCT images (two columns on the right)

For group A, mean neointimal thickness was 0.10±0.03 mm (95% CI 0.07 to 0.11) on the FSS and 0.13±0.05 mm (95% CI 0.10 to 0.16) on the PA; the corresponding values for group B were 0.12±0.5 mm (95% CI 0.09 to 0.15), respectively. The two groups were similar in terms of statistical comparison of FSS endothelial thickness (independent sample t test after F test for equal variances, difference: 0.03083, test statistic t: 1.838, df: 22, two-tailed probability p=0.07).

Comparison of ostia surfaces obtained by 3D OCT reconstructions and SEM images

3D OCT reconstructions on the dedicated console were feasible during the DSA control session. On a case-to-case comparison with SEM images a posteriori, good qualitative correlation was found between the two types of images (figure 2). Quantification of ostia surface values was feasible with both techniques. The mean value obtained with OCT reconstructions was 319 750±345 533 μm2 and the mean value obtained with SEM image-derived calculations was 351 198±396 355 μm2 (table 2).

Table 2

Calculations of ostium surface at 12 months post stenting for the four animals performed by 3D-OCT images and by SEM images

Figure 2

Case No 1. Images derived from 12-week digital subtraction angiography (DSA) and optical coherence tomography (OCT) controls as well as optical loop photography and scanning electron microscopy images. The patent opening of the initial ostium is shown with thick and thin arrows, arrowheads and asterisks, which correspond to the same site on each type of image in order to evaluate the appearance with the different imaging techniques. (A, B) OCT cross-sectional images of the 12-week pullback control at the level of the right common carotid artery (CCA)–ascending pharyngeal artery (APhA) bifurcation with 0.9 mm difference; the patent ostia are depicted with the aforementioned symbols. (C) Optical loop photograph (with light from the top) of the chemically fixed (5% glutaraldehyde solution in Sorensen's buffer) and longitudinally opened artery; notice the same sites of opening as well as the shadow of the area corresponding to the jailed APhA (black parentheses). (D) Scanning electron microscopic image of the same area and enlarged views (E, F) showing the still patent parts of the jailed ostium exhibiting important neointimal coverage and endothelial cells which partially cover the remaining ostia (thick and thin arrows, arrowhead). (G) DSA at 12-week control, selective contrast medium injection from the right CCA at working projection showing important remodeling of the jailed right APhA with partially patent ostium (arrows and asterisk); the inset figure shows the DSA image with the same acquisition parameters at working projection before stenting, showing the initial aspect of the right APhA before stenting.

The ostia surface values derived by the two techniques did not differ significantly (Welch test, difference: −31448.25, test statistic t(d): −0.120, df: 5.9, two-tailed probability p=0.91). Good correlation was found between the two techniques (correlation coefficient r=0.9967, significance level p=0.003). Even though the OCT quantification technique had a tendency to overestimate the patent ostia at 12 weeks due to the lower spatial resolution compared with SEM, this was not statistically significant.


Intravascular OCT has already been successfully used in the evaluation of coronary arterial bifurcations14 stented by conventional cardiac stents15 as well as in the quantification of neointimal coverage after drug-eluting stent deployment in the coronary vasculature.16 Although the use of OCT for assessment of neurovascular procedures has already been described, this is the first study to evaluate OCT specifically in arterial bifurcations covered by FDs.

As the struts of FD stents are thinner and denser than conventional or drug-eluting cardiac stents, they are more difficult to visualize. Nevertheless, in the present study the stents were visible both with immediate post-deployment pullback runs and at 12-week controls. In accordance with recent published data,10 initial evaluation of stent apposition on the PA wall close to the arterial bifurcations was efficient; poor apposition could be immediately corrected by balloon inflation in patients. At the same time, in vivo information on immediate thrombus formation could be useful in clinical practice in order to proceed to immediate lysis in cases of small thrombi not clearly visible on immediate DSA controls.

The 3D reconstructions derived from follow-up pullback runs produced jailed ostia images which correlated well with the SEM images at 3 months. Moreover, the ostia quantification results obtained with OCT showed good correlation with the SEM ostia quantification measurements. In this preliminary study, intravascular OCT seems to be an effective and rapid technique for in vivo evaluation of arterial bifurcations jailed by FD stents during intracranial aneurysm treatment. It could also prove useful for stent development research in animal studies with regard to the detailed evaluation of jailed branches such as with increased metal coverage FD stents.17

Although the statistical power was limited, the results suggest a significant difference between the neointimal thickness on the covered ostia or FSS compared with the neointimal thickness found on other parts of the stent for the same cross-sectional cut. Although these findings need further confirmation with a larger number of subjects, they are in accordance with clinical observations suggesting that neointimal coverage may be partially reversible in some cases of jailed branches in humans.18

The presence or absence of competitive flow has been correlated with the degree of vascular remodeling in jailed branches in the published literature.19 ,20 In the present study, we explored the effectiveness of the intravascular OCT technique in moderately or importantly remodeled jailed arteries. Although the sample used in this study was limited in number, OCT shows promising results with regard to in vivo evaluation of remodeled arterial ostia due to FD stent jailing, regardless of the degree of neointimal coverage. Further confirmation is necessary with a larger number of subjects in order to ensure the statistical power of the results.

Free struts covering an ostium may be the source of emboli. Knowing when full coverage is achieved and evaluating neointimal extent may be useful in anti-aggregation regimen management, especially in resistant patients or in cases where interruption of medication is necessary.5 The in vivo method described here may be effective in the evaluation of stented ostia during follow-up when anti-aggregation regimen interruption or retreatment is needed, especially the deployment of a second stent at the level of an already narrowed jailed artery.

Study limitations

Since this was a feasibility study, the number of animals was limited so the quantitative results lack statistical power. Further evaluation with an increased number of animals is needed in order to confirm the good correlation found between OCT and SEM surface measurements.

Both OCT and SEM surface measurements were assumed to be planar with regard to their surface configuration. Even though the small size of the covered ostia allows for this assumption, this fact represents a limitation of the study due to the cylindrical shape of the arteries.

Additionally, statistically significant differences could have been found between the neointimal thickness or ostia surface values for the two groups of cases (terminal or anastomotic circulation) with a larger number of subjects, but the limited number of data probably did not allow such a difference to be shown. Nevertheless, the results are valid for the feasibility purposes of the study. A study with a larger number of subjects may be interesting in terms of comparisons between the two groups.


Intravascular OCT seems to be a promising in vivo technique for assessment of the orifice of jailed branches after FD stenting for intracranial aneurysm treatment.


View Abstract

Supplementary materials

  • Supplementary Data

    This web only file has been produced by the BMJ Publishing Group from an electronic file supplied by the author(s) and has not been edited for content.


  • Contributors All authors made substantial contributions to the conception or design of the work or the acquisition, analysis, or interpretation of data for the work; drafting the work or revising it critically; and approval of the final version of the manuscript.

  • Competing interests None declared.

  • Ethics approval Ethics approval was obtained from the Comité Regional d'Ethique pour l'Expérimentation Animal (CREFAL).

  • Provenance and peer review Not commissioned; externally peer reviewed.

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.