Measuring artery diameter in rabbit angiograms for neurovascular disease modeling applications and device testing

Abstract

Background and purpose

New Zealand White Rabbits are a useful preclinical model for neurovascular research. Inducing right common carotid artery aneurysms in New Zealand White Rabbits gives researchers the ability to test neurovascular devices in vivo. This study takes advantage of archived angiograms to obtain vascular measurements of interest for future research in rabbit models.

Methods

Rabbit angiograms from endovascular studies conducted in-house between 2005 and 2024 were analyzed using ImageJ at points of interest along the thoracic and abdominal aorta. Using scale references, the diameters of various New Zealand White Rabbit (2.5–3.5 kg) arteries were measured for aneurysm models and controls. Anatomic vascular variations were categorized and recorded.

Results

Measurable artery diameters at 39 points of interest were recorded for 170 female New Zealand White Rabbits. Type 1 vascular anatomy was the most common rabbit aortic arch, characterizing 85% of identifiable rabbit anatomies in this study.

Conclusion

By making these measurements available, this study simplifies future neurovascular studies, allowing for vasculature to be selected by size depending on the application. Additionally, the study offers insight into the relationship between elastase-induced aneurysms and surrounding vasculature size and draws size comparisons between some key arteries of interest.

Introduction

Neurovascular disease presents a unique challenge for researchers in the field. Anatomically, neurovasculature varies between individuals, adding a layer of complexity to navigation and device sizing.^1–4 Neurovascular procedures can be invasive and costly to correct when incidents arise^5,6 and preclinical research is limited by size differences between human and animal vasculature. ⁷

Endovascular procedures utilizing coils, stents, and flow diverters are particularly effective in treating cerebral aneurysms, with a comparable success rate to vascular surgery and a lower complication rate.^8,9 As a result, endovascular procedures are now more common than traditional invasive neurovascular procedures. ¹⁰ It is thus imperative that researchers have a reliable and accurate neurovascular model to test endovascular devices and procedures prior to clinical use.

Because of their accessibility and size, New Zealand White Rabbits are an ideal preclinical animal model for human neurovascular disease treatments, including aneurysm embolization and thrombectomy.^7,11,12 Kallmes et al. previously introduced a novel elastase-induced aneurysm model that allows for the controlled study of endovascular treatments in vivo using the right common carotid artery (RCCA) of the New Zealand White Rabbit. ¹¹ Since then, the use of New Zealand White Rabbits for creating elastase-induced aneurysms and testing promising treatments for vascular disease has increased.^12,13 In particular, researchers have focused their work on the main branches and surrounding vasculature of the aorta, which mimic the size of human neurovasculature. ⁷

Due to their size and accessibility, aortic arteries in New Zealand White Rabbits can be matched to human cerebral arteries of interest to optimize the model for specific applications. By recording the diameter of thoracic and abdominal aortic arteries, we aim to create a database for future studies to reference when deciding which arteries to use for neurovascular device and treatment testing. This study records vascular anatomy to offer further insight into anatomic variation within the New Zealand White Rabbit subclavian aortic anatomy.

Methods

Thoracic and abdominal aortic angiograms taken between 2005 and 2024 were obtained from 170 female New Zealand White Rabbits using archives from our preclinical research laboratory. The imaged rabbits weighed 2.5–3.5 kg at the time of purchase and use. All studies were approved by the institutional animal care and use committee.

Elastase-induced RCCA aneurysms were created following the procedure as previously described ¹¹ and recently updated in a video article. ¹⁴ Briefly after anesthetic induction with an intramuscular injection of ketamine and xylazine, the rabbits were intubated and anesthesia was maintained with 2%–3% isoflurane. A cutdown was made to the RCCA and vascular access was achieved with a 5 Fr introducer sheath. Then, using fluoroscopic guidance, a 3 Fr balloon catheter was advanced through the sheath to the RCCA origin and inflated. Following confirmation of occlusion of the RCCA origin using iodinated contrast, 100 U of elastase was incubated within the lumen of the proximal RCCA for 20 min. After this, the balloon was deflated and the catheter was removed. The vessel was then ligated at its midportion and the skin was sutured. The saccular aneurysm typically forms 2–3 weeks after the procedure.

The majority of the thoracic angiograms acquired were taken from studies involving elastase-induced aneurysms in the RCCA. The angiograms, which visualized the vasculature at creation and time points of interest, were analyzed for clear vasculature visualization and full contrast dispersion. One to two representative angiograms were selected per rabbit for the thoracic and abdominal aortic regions. Angiograms lacking visible vasculature or a clear measuring reference were omitted from the study. To ensure consistent measurement, locations of interest were numbered according to Figure 1 with the main value (1–27) representing the anatomical structure and the decimal value (0.1–0.3) representing the location along that structure. Decimal measurements occur 15 mm apart, with each point along the artery more distal to the ostium. The only exception is in the descending aorta, where locations 12 and 13 were 30 mm apart. Locations of interest included the brachiocephalic artery (BCA), right subclavian artery (RSCA), left subclavian artery (LSCA), RCCA, left common carotid artery (LCCA), right vertebral artery (RVA), left vertebral artery (LVA), superior mesenteric artery (SMA), right renal artery (RRA), left renal artery (LRA), IleoCecoColic artery (ICCA), ascending aorta (AA), descending aorta (DA), iliac arteries (IA), lumbar arteries (LA), and abdominal aorta.

Figure 1.

Anterior and posterior view of type 1 New Zealand White Rabbit vasculature showing the 39 locations (1–27) measured. Average measurements for each location are shown as mean ± standard deviation in millimeters. Measurement locations are organized by corresponding artery: right subclavian artery (RSCA), right vertebral artery (RVA), right common carotid artery (RCCA), brachiocephalic artery (BCA), left common carotid artery (LCCA), ascending aorta (AA), left subclavian artery (LSCA), left vertebral artery (LVA), descending aorta (DA), abdominal aorta, superior mesenteric artery (SMA), IleoCecoColic artery (ICCA), right renal artery (RRA), left renal artery (LRA), iliac arteries (IA), and lumbar arteries (LA).

Measurements along the BCA and RSCA were compared to a small dataset (n = 36) of angiograms without RCCA aneurysms to determine how aneurysm creation impacts vessel size. Of these angiograms, 15 (41.7%) were from arteriovenous fistula studies, 11 (30.6%) from LCCA aneurysm studies, 5 (13.9%) from upper LCCA aneurysm studies, and 3 (8.3%) from upper RCCA aneurysm studies, respectively. In addition, one angiogram was from a sidewall aneurysm study and one was from a vascular graft creation.

Subclavian aortic anatomy was assessed using the five anatomic variations reported by Ding et al. ¹⁵ as follows: type 1: the LCCA stems from the bifurcation of the BCA and aortic arch; type 2: the LCCA branches off the aortic arch independently; type 3: the LCCA originates directly from the BCA; types 4 and 5: the RVA is located off the RSCA and BCA, respectively.

Measurements were obtained using basic ImageJ software functions. ¹⁶ For each image, the pixel-to-millimeter scale was set using the measuring balls imaged and the Set Scale command. Prior to setting the scale, measuring balls were measured multiple times to ensure scale accuracy. If measurements differed by diameter drawn across the measuring ball or were inconsistent with other measuring balls in the image, measurement data from the angiogram was omitted. Images were also omitted if the measuring ball set was not present or could not be identified. Measurements at anatomic points of interest were then collected by measuring the diameter of the artery. Measurements at points of bifurcation were taken as close to the ostium as possible. Measurements, along with vascular variation type and imaging date, were recorded for each rabbit.

Continuous variables are reported as mean ± standard deviation (SD) with n number of available subjects for each measuring point, while categorical variables are expressed as frequency (%). Welch's t-tests were performed using mean, sample variance, and sample size to assess quantitative variables as appropriate. P-values < 0.05 were considered statistically significant. All statistical analyses were performed using Microsoft Excel 2024 (Version 16.93.1, Build 25011917). Quantitative figures were created using GraphPad Prism 2024 (Version 10.4.1, Build 532).

Results

Overall results

Example thoracic and abdominal angiograms are shown in Figure 2. An elastase-induced aneurysm is visible at the RCCA in this figure. Supplemental Table S1 reports the mean diameter, standard deviation, and number of datapoints for each measurement location. At least 10 measurements were produced for each measurement site with 30 measurements or fewer retrieved for the RCCA before aneurysm creation (4, 4.1, 4.2), the LCCA (11), the SMA (16, 17, 18, 19, 19.1, 19.2), and location 15 on the descending aorta. All study types were used to calculate location diameter unless otherwise specified.

Figure 2.

Angiograms of (a) the thoracic and (b) abdominal aorta were utilized for obtaining measurements. An elastase-induced aneurysm is present at the right common carotid artery (RCCA) in (a).

This study also allows us to categorize the analyzed arteries by size range for reference in future neurovascular studies. Given the variations observed across rabbits, these categories are broad and provide only a rough expectation of average vessel diameter across all measured locations on a given artery. Falling within the 0–2 mm range are the RVA, LA, LVA, and upper LSCA. The RSCA and SMA fall in the 1–4 mm range, while the 2–3 mm range includes the RCCA, LCCA, RRA, and LRA. The 3–4 mm range is comprised of the BCA and IA, while the DA and AA fall in the 5–6 mm bracket. The abdominal aorta ranges from values between 3 and 5 mm. Notably, location 9, where the LSCA originates from the aortic arch, falls within a 6–7 mm range and is thus identified separately from the upper LSCA superior to the LVA. To simplify categorization, arteries were included in a given range provided the largest and smallest mean diameters on the artery were within 0.1 mm of the range.

Locations of note impacted by aneurysm presence

A decrease in average diameter in the presence of an induced RCCA aneurysm was observed at locations adjacent to the aneurysm site along the RSCA and BCA. Locations 5 and 6 on the BCA, directly inferior to the site of aneurysm induction, demonstrated a significant decrease (21.9% and 20.1%, respectively) in diameter compared to the control: 3.01 ± 0.76 mm versus 3.86 ± 0.42 mm for location 5 and 3.00 ± 0.73 mm versus 3.76 ± 0.53 mm for location 6, both P < 0.0001. Decreases in diameter were also observed at locations 1, 1.1, and 3, though the changes (2.6%, 5.8%, and 6.1%, respectively) were not statistically significant.

Arterial diameter comparisons: Left versus right and aortic evolution

Average arterial diameters were compared for the left and right arteries in the renal, common carotid, vertebral, and iliac arteries, as shown in Figure 3. LCCAs were found to be 16.97% smaller than RCCAs; LVAs were similarly 18.90% smaller than RVAs (P < 0.001). The left and right renal and iliac arteries, respectively, were not significantly different in diameter (defined as P < 0.05).

Figure 3.

Mean artery diameters for the right and left (a) renal arteries, (b) common carotid arteries, (c) vertebral arteries, and (d) iliac arteries. Diameters are averaged across all measurement locations of the specified artery. Error bars indicate standard deviation. Statistical significance was defined as P < 0.05.

Anatomic variation

Of the rabbits with visible thoracic aortae, 118/139 (84.89%) of the rabbits analyzed displayed type 1 vasculature. Type 2 and 3 vasculatures were less common, occurring in 12/139 (8.63%) and 9/139 (6.47%) of the rabbits, respectively. Figure 4 shows examples of vascular anatomy types 1–3 from the analyzed dataset. Type 4 and 5 vasculatures were not observed in this study.

Figure 4.

Angiograms showing anatomic variation types 1–3, respectively. Images were selected from angiograms analyzed in the study.

Discussion

New Zealand White Rabbits are a widely accepted animal model for safety testing ¹⁷ and are well established for use in neurovascular research.^11,12 While average cerebrovascular diameters are known for humans, ¹⁸ limited information exists on the vessel diameter of New Zealand White Rabbits. Existing knowledge about New Zealand White Rabbit vessel size is largely experiential among researchers, though some research exists on abdominal aortic vessel size. ¹⁹ Here, we provide a registry of artery diameters to simplify device and treatment testing in rabbits. This study builds upon the clinical relevance of work done by Ding et al., ²⁰ creating a registry for aortic vessel sizing. This registry allows neurovasculature to be accurately correlated with comparably sized New Zealand White Rabbit vasculature for neurovascular disease modeling applications.

Statistical analyses identified a consistent decrease in surrounding artery diameter in the presence of an elastase-induced RCCA aneurysm compared to control vasculature. A possible reason for this is that aneurysm formation may involve narrowing of the surrounding vasculature as the vessel tissue bulges outward at the aneurysm site.

For the common carotid and vertebral arteries, the left vessel was significantly smaller than the right. The observed differences between renal arteries and iliac arteries were not significant. Work by Ding et al. found the left iliac artery diameter to be 14.29% smaller than the right iliac artery, ⁷ but little else exists in the literature comparing left and right New Zealand White Rabbit aortic artery diameters. While some human studies report larger RCCA and RSCA diameters,^21,22 results can vary depending on the artery and other factors including gender.^23,24 Notably, only female rabbits were studied and researchers should be cautious when generalizing these results to male New Zealand White Rabbits. Still, it is informative to report a comprehensive comparison of female New Zealand White Rabbit artery diameters.

Future studies should consider some limitations when expanding on this work. The biggest limiting factor in this study is the sample size. Because only 36 angiograms without an RCCA aneurysm were obtained, it is difficult to comprehensively observe differences between aneurysm and no aneurysm conditions. Additionally, the relative scarcity of anatomical types 2–5 makes it more difficult to develop a comprehensive understanding of different anatomical types with a sample size of fewer than 200 New Zealand White Rabbits. With a large enough dataset, vascular diameter could be further compared across anatomical types, providing insight into the relationship between aortic anatomy type and size. Correlating vasculature size to the age, sex, and weight of each rabbit at the time of imaging would also be particularly beneficial to the research community, allowing for more robust vascular diameter estimations.

By virtue of the selected angiography modality, this study is also limited by image dimensionality. Though measures were taken to ensure the scale was consistent throughout the image, two-dimensional (2D) images can result in slight inaccuracies at varying distances and angles; three-dimensional (3D) imaging modalities could be utilized in future studies to address this limitation. This said, previous research comparing 3D rotational angiography with 2D digital subtraction angiography indicates that traditional angiograms may more accurately reflect neurovascular measurements, ²⁵ supporting the reliability of our two-dimensional analysis. Researchers may instead focus on determining how the spread of contrast impacts vessel visualization to optimize measurement accuracy. Thus, future studies may consider strategies for ensuring contrast dispersion is even at the time of imaging.

Finally, while we present preliminary data on the impact of elastase-induced aneurysm creation on surrounding vasculature size, future studies should explore the relationship more thoroughly. Understanding how aneurysm creation alters vessel size in New Zealand White Rabbits may provide further insight into aneurysm development and open doors for new aneurysm models.

Conclusion

In this study, we produced a registry for average artery diameter at 39 locations along New Zealand White Rabbit aortic vasculature, identifying type 1 vasculature as the most common anatomic variation. Left arteries tended to be smaller than their right-side counterparts, and vessel diameter decreased at all measured locations along the RSCA and BCA in RCCA aneurysm models. Future neurointerventional studies can reference this work when determining what arteries to use for modeling (cerebro)vasculature and selecting endovascular device sizes.

Abbreviations

AA

Ascending Aorta

BCA

Brachiocephalic Artery

DA

Descending Aorta

IA

Iliac Arteries

ICCA

Ileocecocolic Artery

LA

Lumbar Arteries

LCCA

Left Common Carotid Artery

LIA

Left Iliac Artery

LRA

Left Renal Artery

LSCA

Left Subclavian Artery

LVA

Left Vertebral Artery

RA

Renal Artery

RCCA

Right Common Carotid Artery

RIA

Right Iliac Artery

RRA

Right Renal Artery

RSCA

Right Subclavian Artery

RVA

Right Vertebral Artery

SMA

Superior Mesenteric Artery