Assessing differences in growth and shape between symptomatic and asymptomatic abdominal aortic aneurysms

Introduction

While a majority of abdominal aortic aneurysms (AAA) exhibit asymptomatic growth, some will become symptomatic prior to rupture, exhibiting surgical outcomes that are worse than asymptomatic AAA (aAAA) but better than ruptured AAA (rAAA) (1). The presence of a symptomatic AAA (sAAA) is a well-defined risk factor and marker of impending AAA rupture and is thus treated urgently/emergently (2,3). AAA rupture carries up to 50–90% mortality and accounts for approximately 175,000 deaths worldwide (4,5). Although the risk of AAA rupture generally increases with increasing maximum aortic diameter (Dmax), Dmax may be inaccurate in predicting overall rupture risk (5-8). Current guidelines recommend surgical repair of AAA in males with a Dmax ≥5.5 cm, females with a Dmax of ≥5.0 cm, symptomatic, and/or rapidly growing AAAs (3,5). However, up to 20% of AAAs ≤5 cm rupture, highlighting the need for improved patient-specific prediction methods for aneurysm rupture rather than using Dmax alone (9).

AAA rupture is a complex phenomenon that occurs when the wall stress exceeds the tensile strength of the aortic wall. Complex features of AAA morphology such as tortuosity, curvature, neck/aortic angle, AAA flow lumen (AFL), intraluminal thrombus (ILT), wall calcification, and the complex shape of the aneurysm influence the wall stress and thus potentially AAA growth/rupture risk (10-18). Dmax does not capture the complex morphological features of AAA and may miss important features that contribute to overall AAA rupture risk (10,11,19). AAA volume has been suggested to be a better marker of AAA growth and rupture risk than Dmax, given that volume changes reflect AAA growth in all directions (20-27). Moreover, assessment of aneurysm shape, curvature, and volumetric growth may offer a more complete and patient-specific assessment of important metrics which impact rupture risk.

In this study we aimed to quantify differences in shape and volumetric growth patterns between sAAA and aAAA. We hypothesized that sAAA will demonstrate more eccentric shapes and larger volumetric growth when compared to aAAA. We present this article in accordance with the STROBE reporting checklist (available at https://qims.amegroups.com/article/view/10.21037/qims-2024-2985/rc).

Methods

Population and study design

We conducted a single-center retrospective cohort study, identifying patients with infrarenal AAA from 2010 to 2023 using institutional records and International Classification of Diseases (ICD) (ICD-10 and ICD-9) codes from an internal dataset at the University of Michigan (28). We included patients with native infra-renal AAAs and ≥2 good quality abdominal aortic computed tomography angiograms (CTAs) that were completed prior to surgical repair (in cases where repair was indicated) at least one month apart and with AAA interval growth (which we defined as an increase of Dmax >0.1 cm between two CTA). Patients were excluded if they had poor quality CTAs (defined as suboptimal aortic enhancement (luminal contrast <200 Hounsfield Units), non-contrast computed tomography scans, large amounts of noise/artifact, large slice thickness (i.e., ≥5 mm), thoracoabdominal aortic aneurysms, and/or <2 CTAs that met inclusion. The earliest CTA was denoted as CTA1 and the CTA with the longest surveillance interval which met inclusion was defined as CTA2. Patient demographics, symptomatic status, past medical history, smoking history, and surgical history at time of CTA2 were collected by chart review. Specifically, symptomatic status was determined by chart review and was defined as documented acute onset back, flank, and or abdominal pain which was thought to be attributable to patient’s AAA and without other identifiable cause at time of CTA2. Repair and repair type were collected by chart review. At our institution indications for AAA repair are in-line with Society for Vascular Surgery guidelines (i.e., males with a Dmax ≥5.5 cm, females with a Dmax of ≥5.0 cm, symptomatic, and/or rapidly growing AAAs) (7). This study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments. This study was reviewed and approved by the University of Michigan Medical School Institutional Review Board (No. HUM00133798), which granted a waiver of informed consent due to the retrospective nature of the study.

This study consisted of comparisons amongst two groups. First, patient demographics, aneurysm and neck anatomy, and their change over time, were compared between all sAAA and aAAA in the cohort. Second, a subset of aAAA were matched with sAAA (12 each) based on sex and baseline Dmax. Shape, curvature, and 3D-growth were compared between matched sAAA and aAAA.

Image segmentation

DICOM data for included CTA scans were exported and de-identified. PRAEVAorta (Nurea, Bordeaux, France) was used to obtain segmentations of the aorta and its branches. PRAEVAorta is a fully automated and commercially approved volume segmentation tool which provides an overview of diagnostic information regarding the aortic segments including the diameter, length, tortuosity, and volumetric information of the infra-renal aorta (Figure 1A) and the diameter, length, tortuosity, volume, and angles (alpha/supra-renal and beta/infra-renal) of the aortic neck (Figure 1B). Automated segmentations included a range of aortic segments depending on the extent of CTA that was uploaded (i.e., CTA chest/abdomen/pelvis vs. CTA abdomen/pelvis). Automated segmentations were then manually edited to exclude any extra-vascular tissue and confirmed to have accurate segmentation of the AAA lumen, ILT, and calcium using 3D Slicer (29). Final, manually-confirmed, segmentations started approximately 1 cm proximal to the renal arteries and ended at the iliac bifurcation approximately 1 cm distal to the bifurcation. The inferior mesenteric artery was excluded. All automated segmentations were verified by one blinded investigator (D.J.B.) with >5 years of image segmentation experience. PRAEVAorta defines the aortic neck as starting just distal to the lowest renal artery and ending at the point when the diameter increases >10% of the diameter at the lowest renal artery. The aneurysm is defined as starting distal to the neck and ending just proximal to the iliac bifurcation. These definitions were used when/if manual segmentation was needed.

Figure 1 Representative examples of automated segmentation outputs using PRAEVAorta (Nurea, Bordeaux, France). (A) Diagnostic information of the infrarenal aortic segment including maximum diameter, tortuosity, and volume (lumen, thrombus, calcium, and total). (B) Diagnostic information of the aortic neck including maximum diameter, length, and neck angles.

Evaluation of aneurysm and neck anatomy

Clinical Dmax measurements for CTA1 and CTA2 were obtained from clinical diagnostic imaging reports, performed by 3D post-processing technicians using standardized protocols. Anatomical features of the aneurysm and aneurysm neck were collected from automated PRAEVAorta reports of CTA1 and CTA2 for cases that passed manual verification of accuracy. For cases which required manual editing, anatomical features of the aneurysm and aneurysm neck were collected manually in 3D Slicer. Specific anatomical features of the aneurysm included, total AAA volume, ILT volume, AFL volume, calcium volume, and AAA tortuosity. Tortuosity was calculated as the ratio of the aortic centerline length to the straight-line distance between the corresponding endpoints. Specific anatomical features of the aneurysm neck included neck Dmax, neck length, and neck alpha and beta angles. The alpha-angle (also known as: supra-renal) was defined as the angle between the longitudinal axis of the supra-renal aorta and longitudinal axis of the AAA neck. The beta-angle (also known as: infra-renal) was defined as the angle between the longitudinal axis of the AAA neck and AAA sac (30). Height and weight were obtained from the electronic medical record and were used to calculate body mass index (BMI) and body surface area (BSA). BMI was calculated from the following formula: BMI = weight (kg)/height (m²). BSA was calculated using the Dubois and Dubois formula BSA = [0.007184×weight (kg)^0.425×height (m)^0.725] (31). Aortic size index (ASI) was defined as aortic diameter (cm)/BSA (32). BMI, BSA, and ASI were calculated from height and weight at time of CTA2. AFL/ILT Ratio, which has been suggested to be an independent predictor of ruptured AAA (rAAA), was calculated at CTA1 and CTA2 using the following formula: (AFL volume/ILT volume) (14). The interval of time between the CTA1 and CTA2, defined as the CTA time interval, was calculated and reported in months.

Change in AAA Dmax, AAA volume, ILT volume, AFL volume, and calcium volume was calculated via the difference in the aforementioned variables between CTA1and CTA2 and is presented over time [Δvariable/time interval between CTAs (months)].

AAA centerlines, curvature, shape, segments, and 3D-growth

Aortic centerlines were calculated from baseline segmentation meshes using the Vascular Modeling Toolkit (VMTK) and baseline shape was calculated by measuring the Euclidean distance (mm) between each point on the aneurysm surface and the aortic centerline (33). Curvature was defined as the inverse of the radius of a best-fit circle, quantifying the extent to which a curve deviates from a straight line (34). We defined eight discrete infrarenal aortic segments for each aorta that was included, via methods previously described by our group (28). Longitudinal zones were assigned by designating a value of 1 to the iliac bifurcation and then increasing by 1 for every 5% along the centerline toward the renal arteries (range: 1–19). The AAA was defined as spanning longitudinal zones 1–14, while the aneurysm neck was defined as spanning longitudinal zones 15–19. Lateral zones were assigned by designating a value of 1 for points aligned perpendicular to the plane of the iliac bifurcation, incrementing by 1 for every 30° in the circumferential direction (range: 1–12). These were grouped into four lateral regions (anterior, right lateral, posterior, and left lateral), each consisting of three lateral zones. Shape was compared across 8 segments [2 longitudinal and 4 lateral regions (Figure 2)]. Shape is calculated as the distance (mm) of the aneurysm surface from the aortic centerline. Thus, 3D-growth can be obtained and was defined as the difference in surface shape between CTA2 and CTA1 (mm/month). Shape, curvature, and 3D-growth were calculated across each of the eight aortic segments.

Figure 2 Aortic regions for quantitative analysis were defined using both longitudinal and lateral zoning. Each aorta was divided into 19 longitudinal zones by assigning a value of 1 at the iliac bifurcation and incrementing by 1 for every 5% of centerline length toward the renal arteries. Two longitudinal regions were established: (I) AAA (zones 1–14) and (II) aneurysm neck (zones 15–19). Lateral zones were assigned by designating zone 1 to points normal to the iliac bifurcation plane and increasing by 1 every 30° around the circumference (range: 1–12). These were grouped into four lateral regions: (I) anterior (zones 1, 2, 12), (II) right lateral (zones 3, 4, 5), (III) posterior (zones 6, 7, 8), and (IV) left lateral (zones 9, 10, 11). This zoning scheme divides the aorta into 8 total regions (2 longitudinal × 4 lateral), enabling localized comparison of aortic growth. AAA, abdominal aortic aneurysms.

Statistical shape modeling (SSM)

SSM was utilized to qualitatively compare mean shape between groups. SSM analysis began by standardizing each individual mesh generated from segmentation. Each mesh was aligned to a common coordinate system. Specifically, the iliac bifurcation served as the origin, the positive x-axis was directed perpendicular to the plane of the iliac bifurcation, and the positive z-axis was directed toward the centerline at the level of the lowest renal artery. Meshes were scaled to fit within a unit sphere and registered to a template of a cylinder to establish point-wise correspondence across all models (35). A mean mesh was then generated by averaging the coordinates of the corresponding points across all meshes.

Statistical analysis

Patient characteristics are reported as mean ± standard deviation for continuous variables with normal distribution or median (interquartile range) for continuous variables with a non-normal distribution and number (%) for categorical variables. Differences in shape, curvature, and 3D-growth between aAAA and sAAA across different aortic segments were compared using paired student’s t-test and Kruskal-Wallis. Statistical significance was defined as a P value less than 0.05 for all tests. All statistical analyses were performed using Stata 17.0 (StataCorp LP, College Station, TX, USA).

Results

Demographics

Fifty-five patients were included [median age: 70.4 years, 26 (47.3%) female, 50 (90.1%) Caucasian]. Fifty (90.9%) patients were either former or current smokers. Forty-one (74.5%) patients had a history of hypertension and 8 (14.5%) had a history of diabetes. Although all fifty-five patients were asymptomatic at time of CTA1, 12 (21.8%) patients had sAAA at time of CTA2.

Patients with sAAA were younger than those with aAAA (66.0 vs. 71.0 years, P=0.026), less likely to be Caucasian (75.0% vs. 95.3%, P=0.030), and less likely to have hypertension (50.0% vs. 81.4%, P=0.027). In total 26 (47.3%) patients were female. There was no difference in the percentage of female patients between aAAA and sAAA (48.8% vs. 41.7%, P=0.660). There were no differences in smoking status, diabetes, or BSA between sAAA and aAAA (Table 1).

Repair and repair type

Overall, 40 (72.7%) of patients eventually underwent AAA repair. All 12 patients with sAAA (100%) underwent AAA repair, while only 28 (65.1%) of patients with aAAA underwent repair (P=0.016). There were no differences in repair type (open vs. endovascular repair) between sAAA and aAAA (Table 2).

Baseline anatomic measurements from CTA1

Average AAA Dmax was 5.1±0.7 cm and median AAA volume was 110.6 (87.4, 137.9) cm³. Median ILT volume, AFL volume, and calcium volume were 38.9 (21.2, 72.7), 68.8 (50.0, 83.6), and 0.3 (0.1, 0.6) cm³, respectively. Median AFL/ILT ratio was 1.79 (0.92, 2.82) and median ASI was 2.7 (2.4, 3.0). Median AAA tortuosity was 1.09 (1.06, 1.16). In the aneurysm neck median Dmax was 2.8 (2.6, 3.0) cm and average length was 2.9±1.8 cm. Average neck alpha/supra-renal angle was 21.5°±10.8° and average neck beta/infra-renal angle was 31.5°±16.1°.

There was no difference in AAA Dmax, AAA volume, ILT volume, AFL volume, calcium volume, AFL/ILT ratio, or AAA tortuosity between the groups, although sAAA demonstrated a marginally-significant trend towards larger ASI compared to aAAA (3.0 vs. 2.6, P=0.050). There was no difference neck Dmax, length, alpha-angle, or beta-angle between aAAA and sAAA (Table 1).

Change in anatomic measurements over time

Average CTA time interval was 14.1±11.8 months. Median increase in AAA Dmax was 0.03 (0.02, 0.06) cm/month and median increase in AAA volume was 1.2 (0.6, 2.3) cm³/month. Median increase in ILT volume was 0.5 (0.1, 1.5) cm³/month and median increase in AFL volume was 0.6 (0.2, 1.3) cm³/month. Median change in calcium volume was 0.001 (−0.005, 0.011) cm³/month. Median change in neck Dmax was −0.001 (−0.025, 0.012) cm/month and median change in neck length was 0.001 (−0.022, 0.022) cm/month. On average the alpha-angle increased by 0.78±5.55° and the beta-angle increased by 1.84±4.02°.

There was no difference in CTA time interval between aAAA and sAAA (13.9±8.8 vs. 15.0±19.5 months, P=0.780). While there was no difference in change in Dmax [0.03 (0.01, 0.08) vs. 0.03 (0.02, 0.05) cm/month, P=0.752] between groups, sAAA demonstrated larger increases in AAA volume [1.6 (1.1, 7.8) vs. 1.1 (0.4, 2.1) cm³/month, P=0.019] and AFL volume [1.1 (0.5, 5.7) vs. 0.4 (0.2, 1.2) cm³/month, P=0.017] than aAAA. Interestingly, there was no difference in change of ILT volume between sAAA and aAAA [0.9 (0.3, 3.8) vs. 0.4 (0.1, 1.1) cm³/month, P=0.154]. There was no difference in change in calcium volume, neck Dmax, neck length, alpha-angle, or beta-angle in patients with aAAA and sAAA (Table 3).

Comparison of aneurysm shape and curvature

Figure 3 depicts the mean shapes of each patient group computed by SSM analysis of individual segmentations. One outlier sAAA was excluded from SSM analysis because of severe deviation in aneurysm shape. The SSM-derived mean shapes show that both aAAA and sAAA have eccentric shapes with larger deformation anteriorly than posteriorly. Qualitatively, the mean shape of the sAAA appears to have more focal dilation compared to the mean shape of the aAAA which appears to have more diffuse dilation.

Figure 3 Statistical shape modeling derived average three-dimensional shape of asymptomatic (A) and symptomatic (B) abdominal aortic aneurysms (AAA). Shape is displayed as distance to centerline in mm (blue =5 mm, red =20 mm). We excluded one symptomatic AAA which had severe deviation in shape.

Quantitatively, there was no difference in the shape of the anterior AAA, right lateral AAA, posterior AAA, or left lateral AAA between aAAA and sAAA. Additionally, there was no difference in the shape of the anterior aneurysm neck, right lateral aneurysm neck, posterior aneurysm neck, or left lateral aneurysm neck between aAAA and sAAA (Table 4). There was no difference in the curvature of the anterior AAA, right lateral AAA, posterior AAA, left lateral AAA, anterior aneurysm neck, right lateral aneurysm neck, posterior aneurysm neck, or left lateral aneurysm neck between aAAA and sAAA (Table 5).

Comparison of aneurysm 3D-growth

There was no difference in the 3D-growth rate of anterior AAA, right lateral AAA, posterior AAA, or left lateral AAA between aAAA and sAAA. At the left lateral aneurysm neck, sAAA had higher 3D-growth than aAAA [0.17 (0.05, 0.55) vs. 0.01 (−0.03, 0.14) mm/month, P=0.027]. However, there was no difference in 3D-growth of the anterior aneurysm neck, right lateral aneurysm neck, or posterior aneurysm neck (Table 6).

Discussion

In this study we aimed to comprehensively assess morphological and growth differences between sAAA and aAAA, to gain insight into the potential mechanisms underlying sAAAs that are at risk of rupture. We identified higher volumetric aneurysm growth in sAAA without corresponding differences in Dmax between groups. Second, we found that sAAA had larger increases in AFL volume compared to aAAA despite no significant difference in ILT volume change. When matched for risk factors for AAA growth, we noted that sAAA have a visual trend towards a more focal distribution of 3D-growth on statistical shape models, although there were no statistically significant differences in AAA 3D-growth on quantitative analysis. Lastly, our findings suggest that sAAA may demonstrate higher 3D-growth in the left lateral aneurysm neck compared to aAAA, although the mechanism underlying this observation requires further investigation. Taken together, these findings demonstrate that a comprehensive assessment of aneurysm shape, volume, and 3D-growth may capture subtle anatomic differences not currently appreciable by standard Dmax or volumetric measurements.

AAA rupture is a complex phenomenon that occurs when the wall stress exceeds the tensile strength of the aortic wall. Volumetric measurements, including that of the AAA, AFL, and ILT, have been proposed as promising adjuncts to explain AAA growth and rupture (13,14,20-23,26,27). Both baseline volume and Dmax have been shown to be predictive of increased volumetric growth (26). Additionally, AFL volume has been shown to be greater in rAAA compared with aAAA, in agreement with our observation that AFL volume was higher in the sAAA cohort (13,14). Although there is debate regarding the role of ILT in AAA growth, a meta-analysis demonstrated greater ILT volume in rAAA compared to intact AAA (12,15,36,37). Niklas et al. compared AAA, AFL, and ILT volume across rAAA, sAAA, and aAAA (50 AAA in each group) and found that rAAA had larger AFL and ILT volumes, larger AFL/ILT ratios, and smaller calcium volumes compared to aAAA (14). However, when comparing aAAA and sAAA they found no difference in AFL, ILT, or calcium volumes but did find that sAAA had larger AFL/ILT ratios than aAAA (14). While we did not observe any differences in ATL/ILT ratio, our observation that the longitudinal change in AFL volume was higher that ILT volume is in agreement with this report and suggests that disproportionate growth of AFL and AAA volumes may indicate a higher-risk phase of disease. The location of AAA rupture is variable but most commonly involves the posterior aneurysm or lateral aneurysm neck (38,39). Moreover, juxta-renal AAAs have been associated with higher rates of symptoms compared to infra-renal AAAs (40). Our results are consistent with these findings, as we found higher 3D-growth on the left lateral aneurysm neck in sAAA compared to aAAA. It is plausible that aneurysm growth in the neck leads to activation of visceral nerve fibers such as the splanchnic nerves and/or celiac plexus. Our work builds upon previous computational and volumetric studies, adding an assessment of changes in AAA, AFL, and ILT growth over time, which to our knowledge is novel in the context of comparing aAAA and sAAA.

Prior studies have demonstrated that aneurysm curvature, shape, and symmetry effect aortic wall stress and thus may impact rupture risk (16-18,41-43). Greater aneurysm curvature is associated with increased AAA growth and AAA rupture and has been thought to be a surrogate marker of aortic wall stress (16-18). Although our analysis did not compare differences between rAAA and intact AAA, we did not find a difference in the curvature of aAAA and sAAA across eight clinically relevant segments. It is possible that given the small sample size of sAAA this study was under powered to detect curvature differences. Nonetheless, in the context of prior studies that assessed differences in curvature, our result support the concept that differences in AAA shape may impact rupture risk. Future studies are warranted to better define these differences and provide mechanistic insight.

In this work we utilize SSM to qualitatively compare the mean shapes of aAAA and sAAA. SSM has been utilized in a wide variety of non-vascular and vascular diseases, including AAA, to identify conserved anatomic features that may otherwise be obscured by high degrees of variability between individual patients (28,44-47). Specifically, SSM uses geometric models to describe an assortment of similar objects in a succinct way, thus characterizing the average shape of multiple 3D structures and their variations in shape (44). SSM works by using a principal component analysis to provide information about each independent component of the object and describe the variations in shape across the assortment of objects or in our case AAA (45). SSM has been used to assess aneurysm neck anatomy prior to endovascular AAA repair and was deemed to provide a quantitative description of morphological variations in neck anatomy (47). Our group has recently employed SSM techniques to compare sex differences in AAA shape and growth, finding that women display more eccentric AAA growth patterns compared to a matched cohort of men (28). The current results further highlight the potential utility of SSM as a promising tool for assessing differences in AAA shape and growth, which may in turn lead to mechanistic insights. In future work with larger cohorts of sAAA patients, we plan to explore principal component analyses to assess whether particular modes of variation in shape and growth are significantly associated with sAAA.

Our study used a semi-automated segmentation protocol, where a combination of automated segmentation and manual quality assurance checks were utilized. Various automated imaging software have been designed and validated for fast and detailed analysis of the anatomic characteristics of infrarenal AAAs (48-50). Automated segmentations can provide comprehensive information about AAA volume, traditional size metrics (Dmax), and basic shape characteristics (tortuosity and neck angles) in rapid fashion, thus providing the surgeon/clinician with a comprehensive assessment of the morphological features that may impact rupture risk. Although further studies are needed before more nuanced, non-diameter measurements are applicable for routine clinical care, the rapid availability of such measurements from automated analysis tools open the door to translating these metrics to diverse AAA patient cohorts, either in the research setting or as adjunctive metrics to supplement diameter and volumetric measurements in the clinic.

Our group has previously used an image registration technique (vascular deformation mapping, VDM) in AAA to provide a more comprehensive analysis of AAA growth (28,51). In this study we elected to use differences in shape over time as a measure of 3D-growth given its improved sensitivity to low contrast at the boundary of the AAA and surrounding tissues which are important features in sAAA. The VDM technique depends on clear delineation of the aortic wall from adjacent structures (such as the bowel and inferior vena cava) and was originally developed for thoracic aortic aneurysms, which are considerably less prone to ILT formation. Thus, in cases with potential inflammatory soft tissue changes (as in sAAA) and in cohorts with extensive ILT (as was in this cohort), there can be misregistration errors that currently limit the reliability of VDM in this setting making our current approach more robust.

This study does have several limitations. We involved only a small cohort of largely Caucasian patients from a single center and thus our results may be subject to selection bias and/or not be generalizable to larger and more diverse AAA populations. Given that our study population was almost 50% female—a higher proportion than in typical AAA populations—this could affect the generalizability of our findings. Although patients were not included or excluded because of sex, it is possible that female patients, who have more challenging anatomy and are less likely to meet indications for use of endovascular aneurysm repair, are more likely to be referred to our tertiary center for care (52,53). A potential advantage of our relatively sex-balanced study population is that we may be less biased by sex-related differences in shape/growth, which our group has previously described (28). Moreover, we were only able to include 12 sAAA. Given that we are a tertiary referral center many patients with sAAA (and rAAA) are directly transferred to our center from other facilities to receive surgical treatment, and thus often do not have a prior surveillance CTAs for evaluation of volumetric and growth changes before development of symptoms and/or rupture. Due to a relatively small sample size we are likely not adequately powered to detect the effect of small differences in baseline AAA shape on clinical outcomes and/or rupture potential. Nonetheless, this work does provide important preliminary information suggesting the potential importance of increasing AFL and 3D-growth in sAAA, which can inform future studies in larger AAA populations. Additionally, we did not include rAAA in this analysis given that aneurysm rupture leads to disruption of the aneurysm shape and accurate segmentation and 3-dimensional analyses reply on clear delineation of the aortic wall, which is often obscured in rAAA by tissue edema or retroperitoneal hematoma. Although there are no required specifications for CT scans imported into PRAEVAorta, in our experience CTAs with finer slices (≤2.5 mm) provide the most accurate segmentations. Moreover, our study only included CTs with IV contrast to allow for delineation of AFL vs. ILT. Additionally, we elected to analyze shape, curvature, and growth across eight discreet segments of the aorta, an approach which could be less sensitive to quantitative differences on smaller scales, however, we believe that these eight segments help with the interpretability of our results (i.e., splitting the AAA and aneurysm neck into clinically meaningful segments). Although the PRAEVAorta platform is easy to use and only takes about 2–5 minutes of processing time per segmentation, the shape modeling workflow used in this study is a research-only technique, is not automated, and does require training and expertise. Moreover, PRAEVAorta is currently only commercially available in Europe (although it is anticipated to be commercially available in the United States in 2025). Given the associated cost and requirement for expert post-processing, the analyses described in this paper may not be feasible amongst all clinical practices. Lastly, while careful attention was paid to include only patients where the AAA was believed to the cause of the patients’ acute symptoms based on CT imaging and a vascular surgeon’s clinical assessment, we cannot exclude the possibility of misdiagnosis in patients with sAAA.

Conclusions

Volumetric changes, shape, and 3D-growth may be important adjuncts when considering the risk of AAA rupture. Despite no difference in Dmax, AAA volume, AFL volume, or change in Dmax over time, sAAA had larger increases in overall AAA volume and AFL volume over time when compared to aAAA. In a cohort of aAAA and sAAA matched on risk factors for growth, there were no quantifiable differences in shape or curvature. However, sAAA were found to have a significantly larger degree of 3D-growth at the left lateral aneurysm neck compared to aAAA. This study provides preliminary data that the use of three-dimensional analysis techniques for detection of regional growth during surveillance and as opposed to maximum diameter alone. It is possible that methods that allow for depiction of regional disease progression could allow for better identification of AAAs that are at risk of becoming symptomatic and/or rupturing. However, larger scale studies are warranted to confirm these preliminary findings, explore the mechanisms underlying these differences, and identify optimal treatment thresholds for AAA based on volumetric growth.

Acknowledgments

None.

Footnote

Reporting Checklist: The authors have completed the STROBE reporting checklist. Available at https://qims.amegroups.com/article/view/10.21037/qims-2024-2985/rc

Funding: This research was funded by the Baiardi Family Foundation (to J.L.E.) and the National Institute of Health(No. R44HI145953 to N.S.B.).

Conflicts of Interest: All authors have completed the ICMJE uniform disclosure form (available at https://qims.amegroups.com/article/view/10.21037/qims-2024-2985/coif). N.S.B. is entitled to related to licensure of intellectual property to Imbio/4D Medical. The other authors have no conflicts of interest to declare.

Ethical Statement: The authors are accountable for all aspects of this work and in ensuring the accuracy and integrity of all aspects of this work were appropriately investigated and resolved. This study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments. This study was approved by the Institutional Review Board of the University of Michigan Medical School (No. HUM00133798), which granted a waiver of informed consent due to the retrospective nature of the study.

Open Access Statement: This is an Open Access article distributed in accordance with the Creative Commons Attribution-NonCommercial-NoDerivs 4.0 International License (CC BY-NC-ND 4.0), which permits the non-commercial replication and distribution of the article with the strict proviso that no changes or edits are made and the original work is properly cited (including links to both the formal publication through the relevant DOI and the license). See: https://creativecommons.org/licenses/by-nc-nd/4.0/.

References

1. Soden PA, Zettervall SL, Ultee KH, Darling JD, Buck DB, Hile CN, Hamdan AD, Schermerhorn ML. Outcomes for symptomatic abdominal aortic aneurysms in the American College of Surgeons National Surgical Quality Improvement Program. J Vasc Surg 2016;64:297-305. [Crossref] [PubMed]
2. Aggarwal S, Qamar A, Sharma V, Sharma A. Abdominal aortic aneurysm: A comprehensive review. Exp Clin Cardiol 2011;16:11-5.
3. Chaikof EL, Dalman RL, Eskandari MK, Jackson BM, Lee WA, Mansour MA, Mastracci TM, Mell M, Murad MH, Nguyen LL, Oderich GS, Patel MS, Schermerhorn ML, Starnes BW. The Society for Vascular Surgery practice guidelines on the care of patients with an abdominal aortic aneurysm. J Vasc Surg 2018;67:2-77.e2. [Crossref] [PubMed]
4. van der Vliet JA, Boll AP. Abdominal aortic aneurysm. Lancet 1997;349:863-6. [Crossref] [PubMed]
5. Lederle FA, Wilson SE, Johnson GR, Reinke DB, Littooy FN, Acher CW, et al. Immediate repair compared with surveillance of small abdominal aortic aneurysms. N Engl J Med 2002;346:1437-44. [Crossref] [PubMed]
6. Lederle FA, Johnson GR, Wilson SE, Ballard DJ, Jordan WD Jr, Blebea J, Littooy FN, Freischlag JA, Bandyk D, Rapp JH, Salam AA. Veterans Affairs Cooperative Study #417 Investigators. Rupture rate of large abdominal aortic aneurysms in patients refusing or unfit for elective repair. JAMA 2002;287:2968-72.
7. Powell JT, Brown LC, Forbes JF, Fowkes FG, Greenhalgh RM, Ruckley CV, Thompson SG. Final 12-year follow-up of surgery versus surveillance in the UK Small Aneurysm Trial. Br J Surg 2007;94:702-8. [Crossref] [PubMed]
8. Parkinson F, Ferguson S, Lewis P, Williams IM, Twine CPSouth East Wales Vascular Network. Rupture rates of untreated large abdominal aortic aneurysms in patients unfit for elective repair. J Vasc Surg 2015;61:1606-12. [Crossref] [PubMed]
9. Nicholls SC, Gardner JB, Meissner MH, Johansen HK. Rupture in small abdominal aortic aneurysms. J Vasc Surg 1998;28:884-8. [Crossref] [PubMed]
10. Hua J, Mower WR. Simple geometric characteristics fail to reliably predict abdominal aortic aneurysm wall stresses. J Vasc Surg 2001;34:308-15. [Crossref] [PubMed]
11. Vorp DA, Raghavan ML, Webster MW. Mechanical wall stress in abdominal aortic aneurysm: influence of diameter and asymmetry. J Vasc Surg 1998;27:632-9. [Crossref] [PubMed]
12. Vorp DA, Lee PC, Wang DH, Makaroun MS, Nemoto EM, Ogawa S, Webster MW. Association of intraluminal thrombus in abdominal aortic aneurysm with local hypoxia and wall weakening. J Vasc Surg 2001;34:291-9. [Crossref] [PubMed]
13. Siika A, Lindquist Liljeqvist M, Hultgren R, Gasser TC, Roy J. Aortic Lumen Area Is Increased in Ruptured Abdominal Aortic Aneurysms and Correlates to Biomechanical Rupture Risk. J Endovasc Ther 2018;25:750-6. [Crossref] [PubMed]
14. Niklas N, Gutowski P, Kazimierczak A, Rynio P. Abdominal Aortic Aneurysm Morphology as an Essential Criterion for Stratifying the Risk of Aneurysm Rupture. J Clin Med 2022;11:933. [Crossref] [PubMed]
15. Singh TP, Wong SA, Moxon JV, Gasser TC, Golledge J. Systematic review and meta-analysis of the association between intraluminal thrombus volume and abdominal aortic aneurysm rupture. J Vasc Surg 2019;70:2065-2073.e10. [Crossref] [PubMed]
16. Teng B, Zhou Z, Zhao Y, Wang Z. Combined Curvature and Wall Shear Stress Analysis of Abdominal Aortic Aneurysm: An Analysis of Rupture Risk Factors. Cardiovasc Intervent Radiol 2022;45:752-60. [Crossref] [PubMed]
17. Chandrashekar A, Handa A, Lapolla P, Shivakumar N, Ngetich E, Grau V. Regent Lee. Prediction of Abdominal Aortic Aneurysm Growth Using Geometric Assessment of Computerized Tomography Images Acquired During the Aneurysm Surveillance Period. Ann Surg 2023;277:e175-83. [Crossref] [PubMed]
18. de Galarreta SR, Cazón A, Antón R, Finol EA. The Relationship Between Surface Curvature and Abdominal Aortic Aneurysm Wall Stress. J Biomech Eng 2017;
19. Finol EA, Amon CH. Blood flow in abdominal aortic aneurysms: pulsatile flow hemodynamics. J Biomech Eng 2001;123:474-84. [Crossref] [PubMed]
20. Kontopodis N, Lioudaki S, Pantidis D, Papadopoulos G, Georgakarakos E, Ioannou CV. Advances in determining abdominal aortic aneurysm size and growth. World J Radiol 2016;8:148-58. [Crossref] [PubMed]
21. Kauffmann C, Tang A, Therasse E, Giroux MF, Elkouri S, Melanson P, Melanson B, Oliva VL, Soulez G. Measurements and detection of abdominal aortic aneurysm growth: Accuracy and reproducibility of a segmentation software. Eur J Radiol 2012;81:1688-94. [Crossref] [PubMed]
22. Parr A, Jayaratne C, Buttner P, Golledge J. Comparison of volume and diameter measurement in assessing small abdominal aortic aneurysm expansion examined using computed tomographic angiography. Eur J Radiol 2011;79:42-7. [Crossref] [PubMed]
23. Kontopodis N, Metaxa E, Papaharilaou Y, Georgakarakos E, Tsetis D, Ioannou CV. Value of volume measurements in evaluating abdominal aortic aneurysms growth rate and need for surgical treatment. Eur J Radiol 2014;83:1051-6. [Crossref] [PubMed]
24. Czermak BV, Fraedrich G, Schocke MF, Steingruber IE, Waldenberger P, Perkmann R, Rieger M, Jaschke WR. Serial CT volume measurements after endovascular aortic aneurysm repair. J Endovasc Ther 2001;8:380-9. [Crossref] [PubMed]
25. Wever JJ, Blankensteijn JD, Th M, Mali WP, Eikelboom BC. Maximal aneurysm diameter follow-up is inadequate after endovascular abdominal aortic aneurysm repair. Eur J Vasc Endovasc Surg 2000;20:177-82. [Crossref] [PubMed]
26. Olson SL, Panthofer AM, Blackwelder W, Terrin ML, Curci JA, Baxter BT, Weaver FA, Matsumura JS. Role of volume in small abdominal aortic aneurysm surveillance. J Vasc Surg 2022;75:1260-1267.e3. [Crossref] [PubMed]
27. Siika A, Bogdanovic M, Liljeqvist ML, Gasser TC, Hultgren R, Roy J. Three-dimensional growth and biomechanical risk progression of abdominal aortic aneurysms under serial computed tomography assessment. Sci Rep 2023;13:9283. [Crossref] [PubMed]
28. Braet DJ, Baker TJ, Delbono L, Spahlinger G, Graham N, Arora A, Figueroa CA, Eliason JL, Burris NS. Three-dimensional characterization of sex differences in abdominal aortic aneurysm progression via vascular deformation mapping. Sci Rep 2024;14:24215. [Crossref] [PubMed]
29. Fedorov A, Beichel R, Kalpathy-Cramer J, Finet J, Fillion-Robin JC, Pujol S, Bauer C, Jennings D, Fennessy F, Sonka M, Buatti J, Aylward S, Miller JV, Pieper S, Kikinis R. 3D Slicer as an image computing platform for the Quantitative Imaging Network. Magn Reson Imaging 2012;30:1323-41. [Crossref] [PubMed]
30. van Keulen JW, Moll FL, Tolenaar JL, Verhagen HJ, van Herwaarden JA. Validation of a new standardized method to measure proximal aneurysm neck angulation. J Vasc Surg 2010;51:821-8. [Crossref] [PubMed]
31. Du Bois D, Du Bois EF. A formula to estimate the approximate surface area if height and weight be known. 1916. Nutrition 1989;5:303-11; discussion 312-3.
32. Davies RR, Gallo A, Coady MA, Tellides G, Botta DM, Burke B, Coe MP, Kopf GS, Elefteriades JA. Novel measurement of relative aortic size predicts rupture of thoracic aortic aneurysms. Ann Thorac Surg 2006;81:169-77. [Crossref] [PubMed]
33. Antiga L, Piccinelli M, Botti L, Ene-Iordache B, Remuzzi A, Steinman DA. An image-based modeling framework for patient-specific computational hemodynamics. Med Biol Eng Comput 2008;46:1097-112. [Crossref] [PubMed]
34. Piccinelli M, Veneziani A, Steinman DA, Remuzzi A, Antiga L. A framework for geometric analysis of vascular structures: application to cerebral aneurysms. IEEE Trans Med Imaging 2009;28:1141-55. [Crossref] [PubMed]
35. Heimann T, Meinzer HP. Statistical shape models for 3D medical image segmentation: a review. Med Image Anal 2009;13:543-63. [Crossref] [PubMed]
36. Speelman L, Schurink GW, Bosboom EM, Buth J, Breeuwer M, van de Vosse FN, Jacobs MH. The mechanical role of thrombus on the growth rate of an abdominal aortic aneurysm. J Vasc Surg 2010;51:19-26. [Crossref] [PubMed]
37. Di Martino E, Mantero S, Inzoli F, Melissano G, Astore D, Chiesa R, Fumero R. Biomechanics of abdominal aortic aneurysm in the presence of endoluminal thrombus: experimental characterisation and structural static computational analysis. Eur J Vasc Endovasc Surg 1998;15:290-9. [Crossref] [PubMed]
38. Schwartz SA, Taljanovic MS, Smyth S, O'Brien MJ, Rogers LF. CT findings of rupture, impending rupture, and contained rupture of abdominal aortic aneurysms. AJR Am J Roentgenol 2007;188:W57-62. [Crossref] [PubMed]
39. Sterpetti AV, Blair EA, Schultz RD, Feldhaus RJ, Cisternino S, Chasan P. Sealed rupture of abdominal aortic aneurysms. J Vasc Surg 1990;11:430-5. [Crossref] [PubMed]
40. Ayari R, Paraskevas N, Rosset E, Ede B, Branchereau A. Juxtarenal aneurysm. Comparative study with infrarenal abdominal aortic aneurysm and proposition of a new classification. Eur J Vasc Endovasc Surg 2001;22:169-74. [Crossref] [PubMed]
41. Fillinger MF, Racusin J, Baker RK, Cronenwett JL, Teutelink A, Schermerhorn ML, Zwolak RM, Powell RJ, Walsh DB, Rzucidlo EM. Anatomic characteristics of ruptured abdominal aortic aneurysm on conventional CT scans: Implications for rupture risk. J Vasc Surg 2004;39:1243-52. [Crossref] [PubMed]
42. Nathan DP, Xu C, Pouch AM, Chandran KB, Desjardins B, Gorman JH 3rd, Fairman RM, Gorman RC, Jackson BM. Increased wall stress of saccular versus fusiform aneurysms of the descending thoracic aorta. Ann Vasc Surg 2011;25:1129-37. [Crossref] [PubMed]
43. Stringfellow MM, Lawrence PF, Stringfellow RG. The influence of aorta-aneurysm geometry upon stress in the aneurysm wall. J Surg Res 1987;42:425-33. [Crossref] [PubMed]
44. Ambellan F, Lamecker H, von Tycowicz C, Zachow S. Statistical Shape Models: Understanding and Mastering Variation in Anatomy. Adv Exp Med Biol 2019;1156:67-84. [Crossref] [PubMed]
45. Jolliffe IT, Cadima J. Principal component analysis: a review and recent developments. Philos Trans A Math Phys Eng Sci 2016;374:20150202. [Crossref] [PubMed]
46. Khan N, Peterson AC, Aubert B, Morris A, Atkins PR, Lenz AL, Anderson AE, Elhabian SY. Statistical multi-level shape models for scalable modeling of multi-organ anatomies. Front Bioeng Biotechnol 2023;11:1089113. [Crossref] [PubMed]
47. van Veldhuizen WA, Schuurmann RCL, IJpma FFA, Kropman RHJ, Antoniou GA, Wolterink JM, de Vries JPM. A Statistical Shape Model of the Morphological Variation of the Infrarenal Abdominal Aortic Aneurysm Neck. J Clin Med 2022;11:1687. [Crossref] [PubMed]
48. Caradu C, Spampinato B, Vrancianu AM, Bérard X, Ducasse E. Fully automatic volume segmentation of infrarenal abdominal aortic aneurysm computed tomography images with deep learning approaches versus physician controlled manual segmentation. J Vasc Surg 2021;74:246-256.e6. [Crossref] [PubMed]
49. Abdolmanafi A, Forneris A, Moore RD, Di Martino ES. Deep-learning method for fully automatic segmentation of the abdominal aortic aneurysm from computed tomography imaging. Front Cardiovasc Med 2022;9:1040053. [Crossref] [PubMed]
50. Lareyre F, Adam C, Carrier M, Dommerc C, Mialhe C, Raffort J. A fully automated pipeline for mining abdominal aortic aneurysm using image segmentation. Sci Rep 2019;9:13750. [Crossref] [PubMed]
51. Braet DJ, Eliason J, Ahmed Y, van Bakel PAJ, Zhong J, Bian Z, Figueroa CA, Burris NS. Vascular Deformation Mapping of Abdominal Aortic Aneurysm. Tomography 2021;7:189-201. [Crossref] [PubMed]
52. Sweet MP, Fillinger MF, Morrison TM, Abel D. The influence of gender and aortic aneurysm size on eligibility for endovascular abdominal aortic aneurysm repair. J Vasc Surg 2011;54:931-7. [Crossref] [PubMed]
53. Shutze WP Sr, Shutze R, Dhot P, Forge M, Salazar A, Ogola GO. Sex as an independent risk factor for long-term survival after endovascular aneurysm repair. J Vasc Surg 2019;69:1080-1089.e1. [Crossref] [PubMed]