The natural history of celiac artery aneurysms

Introduction

Celiac artery aneurysms (CAAs) are rare, accounting for only 3–4% of all visceral artery aneurysms, fourth in incidence behind splenic, hepatic, and superior mesenteric artery aneurysms, respectively.^1–3 With the wider availability and utilization of computed tomography (CT) scans ⁴ ; however, CAAs are being more frequently detected in asymptomatic individuals. ² Despite this, our understanding of the natural history of these aneurysms, in particular the risk of growth and rupture, remains poor.

Historically, rupture rates were thought to be as high as 72% as the majority of patients presented symptomatically and CAAs were thought to be infectious in etiology. ¹ Although the modern rate of rupture is lower, between 6-13%, data on mortality after rupture is scarce and ranges widely depending on the source, particularly since a significant number of patients will not make it to the hospital.^1,2 Given the concern of this dreaded complication, elective repair is advocated by the Society for Vascular Surgery in patients with asymptomatic aneurysms >2 cm, symptomatic aneurysms regardless of size and pseudoaneurysms, while surveillance is recommended for aneurysms <2 cm in asymptomatic patients. ⁵ Although these guidelines have provided an important framework for optimal management of CAAs, there is still limited data to support them. Considering the increasing incidence of CAAs in an aging population, a contemporary analysis of the behavior of CAAs is paramount. As such, the present study aims to document the natural history of CAAs managed at a single institution with a particular interest in recording the mean growth rate of CAA and identifying factors associated with aneurysm growth.

Methods

Results

Patient demographics

Between 2015 and 2019, an average of 18,300 abdominal CT scans were performed yearly. We identified 76 patients with 76 CAAs over this time frame. Our yearly incidence of CAA was 0.10%. The mean age at presentation was 69.8 +/− 12.1 years (range 29–93 years) and 66 patients (86.8%) were male. Hypertension and hyperlipidemia were the most frequently observed comorbidities (64.5% and 46.1%, respectively). Twenty-six patients were former smokers (34.2%), but only 3 (3.9%) were active tobacco users. Only two patients were receiving immunosuppressive therapy, including prednisone, and only one patient (1.3%) was pregnant. No patients had a history of pancreatitis. A full list of medical comorbidities can be found in Table 1.

OPEN IN VIEWER

Table 1.

Demographics of patients with celiac artery aneurysm (CAA).

Characteristic	Number of patients (%)
Demographics
Total patients	76 (100%)
Age at initial presentation, years	69.8 +/− 12.1 years
Male	66 (86.8%)
Comorbidities
Active tobacco use	3 (3.9%)
Prior tobacco use	26 (34.2%)
Hypertension	49 (64.5%)
Hyperlipidemia	35 (46.1%)
Diabetes	7 (9.2%)
Coronary artery disease	14 (18.4%)
Prior stroke	8 (10.5%)
Peripheral artery disease	5 (6.6%)
Prednisone or immunosuppressive medication use	2 (2.6%)
Obesity	14 (18.4%)
Pregnancy	1 (1.3%)
Pancreatitis	0 (0%)

Imaging and CAA characteristics

The most common indication for imaging was follow-up of a pathology unrelated to CAA (48.7%). Other indications included abdominal pain (22.4%), follow-up of a known CAA (21.1%), and other/unknown (6.8%). Of note, no patients underwent imaging for hemodynamic instability, defined as hypotension or tachycardia, for presumed rupture.

The mean diameter at diagnosis was 15.4 +/− 3.8 mm. The majority of aneurysms measured less than 15 mm (48.7%) or between 15 and 20 mm (40.8%) with the remainder (10.5%) found to be 20 mm or larger on initial imaging. All aneurysms were radiologically identified as true aneurysms. Of these, 76.3% were fusiform and 23.7% were saccular. Additional aneurysm features included calcification (26.3%), thrombus (13.2%), and dissection (11.8%). There were no radiologically identified ruptures. Associated imaging findings included the presence of abdominal aortic aneurysm (13.2%), atherosclerosis of the abdominal aorta (38.2%), and additional visceral aneurysms (10.5%). Of the total cohort, 47.4% were referred to a vascular surgeon. Of patients with an initial aneurysm size less than 15 mm, 51.4% were referred to a vascular surgeon while of those with an initial size between 15 and 20 mm, only 20% were referred. Finally, in patients with a CAA 20 mm or greater, 50% were referred to a vascular surgeon. Celiac artery aneurysm characteristics and associated imaging findings are listed in Table 2.

OPEN IN VIEWER

Table 2.

Celiac artery aneurysm characteristics and associated imaging findings.

Aneurysm characteristic	Percent of total
Indications for imaging
Abdominal pain	22.4%
Hypotension or tachycardia	0
Follow-up known aneurysm	21.1%
Follow-up other pathology	48.7%
Unknown/other	6.8%
Characteristics
Mean diameter at diagnosis, mm	15.4 +/− 3.8
CAA <15 mm	48.7%
CAA 15–20 mm	40.8%
CAA ≥20	10.5%
Fusiform	76.3%
Saccular	23.7%
True aneurysm	100%
Calcification	26.3%
Thrombosis	13.2%
Dissection	11.8%
Rupture	0
Associated imaging findings
Presence of AAA	13.2%
Atherosclerosis of abdominal aorta/branches	38.2%
Additional visceral aneurysms	10.5%
Additional visceral artery anomalies other than aneurysm	11.8%
Referral to vascular surgery
Referral for any sized aneurysm	47.4%
Referral for CAA <15 mm	51.4%
Referral for CAA 15–20 mm	20%
Referral for CAA ≥20 mm	50%

AAA: abdominal aortic aneurysm.

Univariate comparison and multivariate regression

On univariate analysis (see Table 3), age over 70 years old was protective against aneurysmal growth (OR 0.087, 95% CI: 0.02–0.46, p = .004). T-testing demonstrated that the mean age of patients with aneurysmal growth was 62.3 +/− 11.7 years while mean age of patients with aneurysm stability was 74.3 +/− 11.6 years. This relationship persisted in multivariate analysis where older age was again found to be protective against growth (OR 0.027, 95% CI: 0.01–0.31, p = .004). Other factors, including gender, tobacco use, hypertension, hyperlipidemia, obesity, thrombosis, calcification, or initial diameter over 20 mm were not found to be predictive of or protective from aneurysm growth on either univariate or multivariate analysis. The full results of the multivariate analysis can be found in Table 4.

OPEN IN VIEWER

Table 3.

Univariate analysis of patient and aneurysm characteristics on CAA growth.

Variable	Odds ratio	95% confidence interval	p value
Age ≥70	0.087	0.02–0.46	.004
Gender	2.083	0.44–9.79	.35
Prior or current tobacco	1.636	0.48–5.62	.43
Hypertension	0.530	0.14–1.99	.35
Coronary artery disease	0.375	0.07–2.03	.26
Hypercholesterolemia	1.378	0.40–4.70	.61
Obesity	2.00	0.35–11.32	.43
Thrombosis	0.417	0.04–4.09	.45
Calcification	0.982	0.28–3.46	.98
Initial diameter >20 mm	1.442	0.28–7.44	.66

Bold face p values represent statistical significance.

OPEN IN VIEWER

Table 4.

Multivariate regression analysis of patient and aneurysm characteristics on CAA growth.

Factor	Odds ratio	95% confidence interval	p value
Age ≥70	0.027	0.01–0.31	.004
Gender	5.365	0.54–53.10	.151
Prior or current tobacco	0.673	0.09–5.03	.700
Hypertension	2.437	0.27–22.10	.428
Coronary artery disease	0.322	0.04–2.83	.307
Hypercholesterolemia	2.251	0.34–14.73	.397
Obesity	2.260	0.19–27.39	.522
Thrombosis	0.917	0.03–28.08	.960
Calcification	1.333	0.16–11.25	.791
Initial diameter >20 mm	0.205	0.02–2.48	.205

Bold face p values represent statistical significance.

Discussion

The Society for Vascular Surgery Guidelines for the management of visceral artery aneurysms published in 2020 recommends treatment of asymptomatic non-ruptured true celiac artery aneurysms measuring >2 cm. Despite being a grade 1 (strong) recommendation, the supporting evidence is deemed low quality. ⁵ As such, we performed a single institution, retrospective analysis of patients with untreated CAAs to further define the natural history of these entities. Our study found aneurysmal enlargement in slightly over one-third of patients with radiological follow-up with a mean growth rate of only 1.09 mm/year within this cohort. Moreover, there were no documented cases of rupture or symptomatic disease at mean clinical follow-up of 31.2 months. Our findings are consistent with prior reports suggesting that CAAs are insidious, slow-growing lesions with low risk of rupture and provide further evidence to support the current guidelines.

Given the rarity of CAAs, it is no surprise that the current literature consists mainly of case reports and small retrospective series. The largest of these contemporary studies focused solely on CAAs analyzed a total of 18 patients, of which there was a single initial presentation of a ruptured aneurysm (6%). Within this initial cohort, 8 patients were managed non-operatively. Mean aneurysm size within this group was 2.1 cm and no aneurysm was observed to significantly enlarge over follow-up. There was a single late rupture (6%) in a patient with a 2.5 cm aneurysm who had previously declined repair. ² It is difficult to draw comparisons between our study and this previously published report for many reasons, the most obvious being the small overall number of patients who were surveilled in the original report. It is unclear how many of these asymptomatic patients are referred to or followed by a vascular specialist, but of the 8 patients in this study with CAA >2.0 cm, half were not seen by a vascular surgeon and the remainder were managed expectantly for a variety of reasons, including high risk medical comorbidities.

Retrospective single center studies analyzing the behavior of splanchnic artery aneurysms (SAA; i.e., arising from the celiac, superior mesenteric or inferior mesenteric arteries) provide complementary yet limited data on CAAs. A study published by Batagini et al. ⁶ evaluating 74 patients with 87 SAAs with at least two subsequent CT scans reported a median growth of 0.63 mm/year for all aneurysms. More specifically, 20 patients were noted to have a CAA with mean diameter at presentation of 15 mm and a reported growth rate of −0.1 mm/year (+/− 1.4 mm). Within a follow-up period of 57.3 months, there were no reported ruptures. Other studies have additionally confirmed the indolent nature of SAAs and CAAs amongst patients demonstrating aneurysmal enlargement. In 2017, Erben et al. ⁷ reported a mean growth rate of 0.64 mm/year amongst 122 patients with 138 SAAs. Of the 63 CAAs identified, mean initial size was found to be 14.6 mm. Growth rate stratified by aneurysm location was not provided, but of note, there were no ruptures within the CAA cohort.

The study population included in this report tended to be older men which is consistent with the existing literature,^2,8 but stands in contrast to patients with splenic and renal artery aneurysms, who are often be female.^8,9 The relationship between gender and CAAs has yet to be fully elucidated but the male predominance appears to date back to historic reports published prior to the 1950s. ¹ Interestingly, up to 18% of CAAs are associated with abdominal aortic aneurysms, ¹ a disease primarily affecting men. ¹⁰

The data is even more sparse in identifying risk factors for CAA aneurysm growth and rupture. In their presentation of 18 patients, Stone et al. ² were unable to identify any comorbid conditions or aneurysm characteristics associated with risk of rupture, which they attributed to being underpowered for such analysis. In their study of SAAs, Erben et al. ⁷ found that female sex was associated with faster annual aneurysm growth rate, while smoking was associated with a slower growth rate. However, a similar analysis of SAAs by Batagini et al. ⁶ identified only portal hypertension as a risk factor for faster SAA enlargement. Regardless, it is unclear how applicable these findings are to patients with CAAs given the diversity of SAAs.

Our study found on both univariate and multivariate analysis that age >70 was protective against aneurysm growth. The mean age of patients who were found to have aneurysmal growth at follow-up was 62.3 years, while the mean age of patients who demonstrated aneurysm stability was 74.3 years. This finding has important implications for future surveillance protocols. Currently, the SVS guidelines suggest annual surveillance scans of patients with CAAs who are being observed, which they list as grade 2 (weak) evidence. ⁵ Given our findings, future studies should be undertaken to determine whether the screening interval can be increased in patients over 70 years old in an effort to spare them from unnecessary radiation and preserve hospital resources. Identifying which patients warrant annual screening is particularly important as the use of CT scans continues to increase annually, ¹¹ leading to more patients being diagnosed incidentally.

As incidental diagnoses become increasingly common, guidelines will need to be updated to better reflect these trends. Currently, none of the major society guidelines,^5,12 put forth specific recommendations on when patients with asymptomatic splanchnic artery aneurysms, including CAA, should be referred to a vascular specialist. Given this lack of clarity, it is not surprising that within our own health system only 47.4% of patients with CAA were referred to a vascular surgeon. At our own institution, guidelines are being addressed to encourage referral at time of diagnosis.

There is a wide variety of treatment options available today for patients with larger aneurysms indicated for repair. Previously, celiac artery aneurysms were managed with open ligation, excision, or bypass—procedures associated with higher morbidity and mortality particularly in the face of rupture. However, management options have expanded with the development of coil embolization and endovascular stenting. A comparison of open and endovascular management of celiac and superior mesenteric artery aneurysms performed by Sachdev et al. ¹³ found that the technical success rate for endovascular treatment approached 90% and was safe even amongst hemodynamically unstable patients presenting with rupture.

There are several limitations to this study. Although it is larger than previously published reports on the subject, the overall numbers are still small and data collection was performed retrospectively. Additionally, clinical follow-up time was relatively short at 31.2 months and only 45 patients (59.2%) had radiological follow-up. To better quantify the risk of growth and rupture, longer and more consistent follow-up will be needed. Symptoms at follow-up were identified through chart review of encounters often unrelated to the pathology at question, which created the potential for incomplete assessment. For example, it was not always possible to identify the exact etiology of the patient’s abdominal pain after aneurysm related issues were excluded. It is additionally worth mentioning the limitations of CT scanning, which although currently is the gold standard for diagnosis, cannot fully characterize CAAs or reliably distinguish between true and false aneurysms. Furthermore, while we did not collect information on whether all patients were seen by a vascular surgeon, our intervention rate is potentially explained by the fact that many of our subjects were not assessed by a vascular specialist or other provider, such as an interventional radiologist, who manages this disease. Lastly, but perhaps most importantly, there were no documented ruptures so analysis of risk factors for this event could not be performed. Despite these limitations, we believe this is the most comprehensive contemporary review of the natural history of celiac artery aneurysms to date.

Conclusions

This single institution study of patients undergoing observation of CAAs found that the majority of CAAs remained stable in size over a mean follow-up 31.2 months with only one patient undergoing intervention and no documented ruptures. Given the benign nature and indolent growth of these lesions, particularly of those <2 cm, our findings support the current SVS recommendations.