Is the Anatomy of the Inferior Mesenteric Artery Root Different in Patients with a Tortuous Abdominal Aorta?

Background: Catheterization of the inferior mesentery artery (IMA) during angiography can sometimes be dicult due to anatomic variations. Our study aimed to evaluate the relationship between abdominal aortic tortuosity (AAT) and the anatomy of the IMA root. Methods: In this retrospective study, a total of 45 patients were selected and 3 groups were divided using a qualitative method by visual estimation of vessel tortuosity. A quantitative method was performed by recording the patient age, tortuosity index, and features of the IMA root anatomy, including orice level, vertical diameter, clockwise angle, and angulation. Pearson correlation coecients (p < 0.05) were calculated to analyse the strength of the linear association between tortuosity and other variables. Results: The AAT index was signicantly associated with age. When the abdominal aorta was tortuous, the IMA root was more likely to be distorted toward the convex side, with a larger angulation between the IMA root and the abdominal aorta. Conclusions: Our study provides information which could help clinicians plan endovascular interventions of IMA. When cannulation of the IMA appears to be dicult, a rotation of tip of the catheter to the convex side of the abdominal aorta and a less sharp looping of the catheter may be attempted.


Background
Catheterisation of the inferior mesentery artery (IMA) during angiography can sometimes be di cult due to variations in the patients' anatomy. A previous study using three-dimensional CT to evaluate the anatomy of IMA showed that it arises from the aorta approximately 7 cm below the origin of the superior mesentery artery (SMA). It is usually at the level of L3, and is a relatively straight vessel arising from left side of aorta [1]. Another study evaluating the positional relationships of the inferior mesenteric artery using the coeliac trunk and SMA as landmarks in 32 cadavers revealed that the site of branching of the IMA can be inferred to some extent from the celiac axis and superior mesenteric artery [2].
In this study, we aimed to investigate whether the anatomy of the IMA root was different in patients with increased abdominal aortic tortuosity (AAT).

Methods
This retrospective clinical study was approved by the Institutional Review Board.

Patient population
A total of 45 patients who underwent abdominal CT angiography (CTA) between January 2017 and December 2019 were selected and divided into three groups. The groups were composed of 15 patients without AAT (group A), 15 patients with signi cant AAT and a leftward convexity (group B), and 15 patients with signi cant AAT and a rightward convexity (group C). Patients were grouped using a qualitative method involving visual estimation of vessel tortuosity. None of the patients had any history of vascular diseases (e.g. abdominal aortic aneurysm, aortic surgery, or severe atheromatous plaque).

Quantitative image analysis
Patient age, tortuosity index, and features of the IMA root anatomy, including ori ce level, vertical diameter, clockwise angle, and angulation were recorded. The imaging study was performed in Philips CT workstation (Intellispace Portal v7.0.1, Philips Healthcare) with automatic batch creation. The tortuosity index of the abdominal aorta was calculated as the ratio of the centreline to the shortest distance from the diaphragm level to the aortic bifurcation. The clockwise angle was measured in a clockwise fashion from 12:00 to the IMA ori ce in the tangential plane of the abdominal aorta. The angle of the IMA was measured at the intersection of the centreline of the abdominal aorta with the IMA root ( Fig. 1).

Statistical analysis
Data were analysed using SPSS Statistics 23 software (IBM Corp, Armonk, NY), and all numerical values were reported as mean ± standard deviation. Independent-sample t-test was used for comparison between groups. A p-value of < 0.05 was considered statistically signi cant. Pearson correlation coe cients (p < 0.05) were calculated to analyse the strength of the linear association between tortuosity and other variables.

Discussion
In this study, we demonstrated that the AAT index was signi cantly related to patient age. The IMA root was more likely to be distorted toward the convex side, with a larger angulation between the IMA root and the abdominal aorta when the abdominal aorta was tortuous.
The IMA is responsible for supplying blood to the distal third of the transverse colon, the descending colon, sigmoid colon, and upper rectum. It is a relatively straight vessel with several branches, all of which arise from the left side, typically between the L2 and L4 vertebral bodies [3,4]. In our study, the majority (80%) of IMA ori ces originated at L3, while 18% originated at L4. Only one patient (2%) had an IMA ori ce originating at L2.
Previous studies have demonstrated a strong association between vessel tortuosity and mechanical factors, including blood pressure, blood ow axial tension, and structural changes to vessel walls [5]. The in uence of age on aortic tortuosity, however, remains controversial [6]. In recent studies, the tortuosity and length of the thoracic aorta were reported to be moderately to signi cantly associated with age in the ascending and descending segments [7,8]. In our study, the AAT index was also shown to be signi cantly associated with age.
The tortuosity of an artery can be measured using both qualitative and quantitative methods; however, the visual estimation of a vessel's tortuosity is most often used in clinical settings [9]. In our experience, the AAT can be estimated using radiography. Abdominal computed tomography or a guidewire tract during angiography studies may be use if calci ed plaques are present in the abdominal aorta.
In our study, a right counter clockwise rotation of the IMA ori ce was shown to have a moderately positive correlation with AAT and a rightward convexity. This nding may be explained by recognising parallels in patients with scoliosis. Using a common computed tomography (CT) method for measuring vertebral rotation in patients with scoliosis, it has been shown that the vertebrae are distorted toward the convex side, with spinous processes deviated to the concave side during disease progression (Fig. 2) [10,11].
Similarly, a clockwise rotation of the IMA ori ce was observed in patients with AAT and with a leftward convexity (Fig. 3). However, there was no statistically signi cant association between these two variables.
We hypothesised two possible reasons for this nding. First, we observed less severe AAT with leftward versus rightward convexities (tortuosity index: 1.089 vs.1.119, respectively). Additionally, IMA ori ces were not always located at the curvature sites.
A moderately to highly positive correlation was also identi ed between angulation and AAT. A kyphotic curvature of the abdominal aorta was observed in both groups B and C except in cases with lateral deviation. During a xed peritoneal attachment, a wider intersection angle can be achieved with greater kyphotic curvatures (Fig. 4).
There were two limitations in our study. First, the tortuosity index can change depending on the two end points selected. In our study, we chose the diaphragm and the aortic bifurcation as our two end points because they were easily identi able on imaging studies. Second, as described above, intersecting points are not always the greatest curvature site of the selected vessel, especially when there are more than two consecutive curvatures.

Conclusion
In our study, the AAT index was shown to be signi cantly related to patient age. When the abdominal aorta was tortuous, the IMA root tends to be distorted toward the convex side, with a larger angulation between the IMA root and the abdominal aorta. Our study provides information which could help clinicians plan endovascular interventions of IMA. When cannulation of the IMA appears to be di cult, a rotation of tip of the catheter to the convex side of the abdominal aorta and a less sharp looping of the catheter may be attempted.

Declarations
Ethics approval and consent to participate: