Clinical outcome of endovascular therapeutic occlusion of the celiac artery

Patients and methods 

Patients 

Between 1998 and 2006, 10 patients (2 women, 8 men) underwent endovascular occlusion of the celiac artery at our institution. Their average age was of 62.5 ± 9.8 years (range, 49 to 79 years).

Table I summarizes the type, location, and extent of the aneurysms of the patients. All aneurysms met the standards set by the Society for Vascular Surgery and the International Society for Cardiovascular Surgery. There was one thoracoabdominal mycotic pseudoaneurysm of the aorta with a broad proximal but narrow distal landing zone of only 14 mm. The thoracic stent graft implantation and occlusion of the celiac trunk was this patient’s last valid interdisciplinary treatment option. The technical result was good.

Table I. Demographics of patients, procedures, and concomitant disease

	Patient	Age	Aneurysm data			Liver blood supply⁎	Subsequent procedures	Concomitant disease
			Type	Location, extent	Size, mm (CA ratio)
	1	49	Dissecting	TAA	48×43	B	Stent implant SMA	Thyroid cancer
	2	57	Dissecting	CAtoCHA	13×12(2.2)	D	—	CHD, COPD
	3	51	Fusifom	CAtoRHA	21×29(3.5)	D	—	Pancreatic cystadenoma
	4	74	Mycotic pseudoaneurysm	TAA	56×48	C	—	CHD, diabetes
	5	62	Dissecting	TAA	50×42	B	—	CRI
	6	71	Nondissecting	TAA	78×72	B	—	Pancreatic carcinoma
	7	64	True	CA	30×21(2.6)	A	—	Arterial hypertension
	8	57	True	CA	31×19(2.6)	D	—	—
	9	61	True	CA	25×15(3,75)	A	—	Arterial hypertension
	10	79	Nondissecting	TAA	57×31	A	—	CHD, hypertension

CA, Celiac artery; TAA, thoracoabdominal aneurysm; CHA, common hepatic artery; CHD, coronary heart disease; COPD, chronic obstructive pulmonary disease; SMA, superior mesenteric artery; RHA, right hepatic artery; CRI, chronic renal insufficiency.

⁎ Letters refer to panels A-D in Fig 1.

The diagnosis in most patients was made by computed tomography (CT) for clarification of epigastric pain. Two of the celiac artery aneurysms were diagnosed during ultrasound scans performed for other diseases. Neither of the latter two patients showed any symptoms from the aneurysm. The patient with the fusiform aneurysm of the CHA was in a follow-up program after partial pancreas resection for a cystadenoma.

All patients were treated exclusively by endovascular means. Indication for therapy was an aneurysm with a diameter of more than twice the celiac artery, a pseudoaneurysm, a dissecting thoracoaortic aneurysm, or a thoracoabdominal aortic aneurysm with a diameter of >5 cm in the vicinity of the celiac artery. Before the occlusion of the celiac artery by thoracic stent grafting, the celiac artery was coil embolized to prevent a type II endoleak. In one case of a Stanford type A dissection, the celiac artery showed a subtotal stenosis of 98% and could therefore not be embolized before implantation of the stent graft. It was occluded solely by over-stenting (Table II).

Table II. Endovascular intervention, coils, stent grafts, coil-embolized branches of the CA, and treatment related complications

	Patient	Treatment	Specification of coils	Stent graft	CA branches	Treatment related complications
	1	Aortic endoprosthesis (ET)	—	TAG Thoracic Endoprosthesis⁎ 37/20, 37/15	Not involved	Abdominal pain, followed by stent PTA of SMA
	2	CA coil embolization (ET)	IDC§ 8/20, 10/10, 10/20(2×), 12/10, 12/20(4×)	—	Not involved	—
	3	CA coil embolization (ET)	DCS‡ 8/20(3×), 8/30, 6/30, 6/15(2×), 8/15, 8/50, 6/50, 10/20	Talent† TB2222C116T	Occlusion of the LGA	—
			IDC§ 5/15, 6/20, 7/10, 8/10(2×), 9/10, 10/10, 10/20(2×)
	4	CA coil embolization, aortic endoprosthesis (SUT)	IMWCE‡ 4/3(3×), 5/5(4×)	Talent† TF4040C114T	Not involved	—
	5	CA coil embolization, aortic endoprosthesis (ET)	IDC§ 8/10, 10/20, 12/20(2×), 14/10, 14/20(4×)	Excluder⁎ 34-20-00, 37-10-00	Not involved	—
			MWCE‡ 8/12(2×), 15/15
	6	CA coil embolization, aortic endoprosthesis (ET)	IDC§ 5/15, /20(2×), 8/20, 10/20(5×), 12/20, 14/20(4×)	—	LGA occlusion	—
	7	CA coil embolization (ET)	DCS 6/20, 6/25(3×), 6/30, 6/50(3×), 8/20(3×), 8/25(3×), 8/30(3×), 8/50 2×), 10/30(4×), 20/50(3×)	—	CHA, LGA occlusion	—
	8	CA coil embolization (ET)	IDC§ 14/20, 10/20, 8/10(3×), 6/20(2×), 6/10(2×), 5/15(2×)	—	CHA, SA occlusion	—
			GDC§ 18/30 3D(2×), 14/30	—
			MWCE‡ 10/15(2×), 8/15(3×), 8/12(2×), 5/8(4×)	—
	9	CA coil embolization (ET)	IDC§ 8/20(4×), 12/20, 14/20(2×), vein of Galen coils 20/10(3×)	—	Not involved	—
			Macrospher‡ 12 mm (5×)	—
	10	CA coil embolization, aortic endoprosthesis (ET)	Microcoils 8/60, 6/60(7×)(MIT)	Talent† TF2828C100×	Not involved	One coil moved into the splenic and hepatic artery, respectively
			IDC§ 8/10	—

CA, Celiac artery; CHA, common hepatic artery; DCS, detachable coil system (mechanically detachable coil); ET, elective therapy; GDC, Guglielmi detachable coil (electrolytically detachable coil); IDC, interlocking detachable coil (mechanically detachable coil); IMWCE/MWCE, Micro/Mini Wallace Coiled Embulous (detachable coil); LGA, left gastric artery; SA, splenic artery; SMA, superior mesenteric artery; SUT, semi-urgent therapy.

⁎ W.L. Gore & Associates, Flagstaff, Ariz. † Medtronic, Minneapolis, Minn. ‡ Cook, Bloomington, Ind. § Boston Scientific, Natick, Mass.

Two patients had an underlying malignancy, one of which, a case of pancreatic carcinoma, was diagnosed during a follow-up examination shortly after the procedure. All patients had multiple comorbidities (Table I).

A thorough discussion of the procedure, in which all treatment alternatives were presented, was conducted with all patients, and all patients signed a written consent form before the procedure. Institutional ethical approval was waived because the study had no influence on treatment, and our institutional Ethical Review Board did not require its approval for this study.

Preinterventional vascular situation 

Nine patients had a complete celiac artery in the form of a gastrohepatosplenic celiac trunk, and one patient had an incomplete hepatosplenic celiac trunk. In this patient, the left gastric artery (LGA) arose as a robust vessel from the aorta immediately to the left of the celiac trunk. An accessory left hepatic artery (LHA) arose from this LGA that, in addition to the LHA, also supplied liver segments II and III. Four of the nine complete celiac trunks showed the classic trifurcation, and four others had the most common variation of an early branching of the LGA. One patient had a quartering instead of a trifurcation of the celiac trunk under the direct branching of a short dorsal pancreatic artery with the same caliber as the LGA from the celiac trunk. The four branching types of the trunks represented in our patient group are shown in Fig 1, A-D.

View Large Image Download to PowerPoint

Fig 1. A-D, The four branching types of the celiac trunks represented in the patient group. ALHA, Accessory left hepatic artery; CA, celiac artery; CHA, common hepatic artery; GDA, gastroduodenal artery; LGA, left gastric artery; LHA, left hepatic artery; SA, splenic artery; RHA, right hepatic artery; SMA, superior mesenteric artery.

Results 

Collateralization 

The pancreaticoduodenal arteries and the gastroduodenal artery via the SMA were the most common collateral pathways, followed by the gastric arteries.

There were five special situations:

1.The gastroepiploic artery and the gastroduodenal artery had been clipped and the spleen removed during surgery for a cystadenoma of the pancreas. Collateralization took place through the LGA, accessory LHA to the LHA, and intrahepatically to the right hepatic artery (RHA). There was a second collateral pathway by the middle colic artery to the RHA (Fig 2, A).

View Large Image Download to PowerPoint

Fig 2. A-E, The five special collateral pathways represented in the patient group. ALHA, Accessory left hepatic artery; CA, celiac artery; CHA, common hepatic artery; DPA, dorsal pancreatic artery; GDA, gastroduodenal artery; HAP, hepatic artery proper; IPA, inferior pancreatic artery; LGA, left gastric artery; LGEA, left gastroepiploic artery; LHA, left hepatic artery; MCA, middle colic artery; RGA, right gastric artery; PDA, pancreaticoduodenal arteries; RGEA, right gastroepiploic artery; RHA, right hepatic artery; SA, splenic artery; SMA, superior mesenteric artery.

2.Collateralization occurred by way of the pancreaticoduodenal arteries to the gastroduodenal artery, hepatic artery proper, and LHA, and in this case continued to the accessory LHA (Fig 2, B).

3.The accessory LHA provided antegrade collateralization because the LGA had remained patent due to early branching off from the celiac trunk (Fig 2, C).

4.A robust dorsal pancreatic artery formed the anastomosis between the SMA and the celiac artery (Fig 2, D).

5.A very robust anastomosis existed between the right gastroepiploic artery and the left gastroepiploic artery, through which the splenic artery was well contrasted (Fig 2, E).

Discussion 

This study shows that therapeutic endovascular occlusion of the celiac artery can be performed without leading to clinically significant ischemia of the upper abdominal organs. The mortality rate in our study was formally 10%, and morbidity was also 10%. The patient who died of a myocardial infarction after the procedure had an infectious aortic aneurysm, had multiple morbidities, and had a poor prognosis. The stent graft implantation and occlusion of the celiac trunk, despite unstable angina pectoris, was her last valid interdisciplinary therapy option and the technical result was very good. One patient’s abdominal angina, which resulted from the partial blockage of the SMA by the aortic prosthesis, was successfully treated by PTA and stenting.

One limitation of our study is the relatively small group of 10 patients. The reason for this is that aneurysms of the celiac artery are rare, and paraceliac aortic aneurysms requiring endovascular over-stenting of the celiac trunk are not as frequent as more proximal aneurysms. Many patients with celiac trunk or aortic aneurysms have multiple comorbidities, as in our study, and for most of them, the risk of surgery is high in addition to the mortality rate of 5% for an operation of the celiac artery9 or even higher for aorta replacement surgery. Thus, endovascular therapy can be an appealing alternative to open surgery in selected cases.

Although treatment is generally recommended for aneurysms of the visceral arteries with a ratio of the aneurysm diameter to the parent vessel >2,8 in this study even relatively small aneurysms of the celiac trunk with a ratio of 2.2 to the parent vessel were treated. The rationale behind this is that acute therapy for a visceral artery aneurysm results in much higher mortality than elective endovascular treatment.10 The prognosis after the rupture of a celiac trunk aneurysm is generally poor.11 In view of these facts, we recommend early preventive treatment.

Before endovascular therapy for paravisceral thoracoabdominal aortic aneurysms, occlusion of the celiac artery had been generally accepted for prevention of a type II endoleak if the level of the celiac artery was needed as distal landing zone.2, 3 However, some evidence indicates that this procedure might be unnecessary in most cases.1 At the present time, no reliable data are available on which patients would or would not benefit from ligation or embolization of the celiac artery before the implantation of the thoracic stent graft.

Influencing factors may be the configuration of the aorta and the diameter and position of the celiac artery and the SMA, as well as the length of the celiac artery. Collateralization from the splenic to the hepatic artery, and vice versa, might be impaired by embolization material, especially when the celiac artery is short, although we did not observe such a situation. Although type II endoleaks do not seem to occur even without prior embolization,1 type I endoleaks may possibly evolve from incomplete over-stenting of the celiac artery as well. Further investigations of this point are needed, however.

In all 10 patients, the anatomic vascular situation was unproblematic or even favorable for occluding the celiac trunk. Only one patient presented with abdominal angina and functional underperfusion of the SMA. The immediate disappearance of these symptoms after PTA and stenting of the SMA branch confirmed that functional restriction of blood flow was a result of the blockage of the vascular branch by a stent strut of the aortic prosthesis. In this case, the replaced RHA from the SMA ensured the blood supply to the right liver segments.

In general however, occlusion of the celiac trunk should be avoided if the portal vein or the SMA are not patent6 or in case of a liver function disorder. Occlusion of the gastroduodenal artery also presents a relative contraindication. Incomplete celiac trunks with single branches of all vessels from the aorta, which are very rare—less than 1% of all cases12—as well as celiac trunk variations in the form of celiac trunks including the superior mesenteric artery, should be classified as very problematic. These include configurations such as a gastrohepatosplenomesenteric celiac trunk, which occurs in 2.4% of cases,13 a hepatosplenomesenteric celiac trunk, which occurs in 1.2%,13 and a hepatomesenteric celiac trunk, which occurs in up to 3%.12 For the latter, it must be taken into consideration that neither a common origin nor a direct connection exists between the splenic and the hepatic artery.

Numerous ways of collateralizing the arterial territories of the spleen, liver, and pancreas have been reported. The pancreatic arcades are the most important ones used when the celiac trunk is ligated.14 In addition, the transverse pancreatic artery, small intrapancreatic arteries, and the dorsal pancreatic artery14 are potential pathways. In case of an incomplete celiac trunk, the intraceliac collaterals by way of the gastroepiploic and the gastric arteries can be significant.14 The connections between the vascular territory of the RHA and LHA are important as regular intraceliac collaterals.15

Renal-celiac collaterals can become symptomatic as a renal-splanchnic steal and arterial hypertension if they are done at the expense of the renal supply, although to our knowledge this has never been described after artificial occlusion of the celiac artery. These are anastomoses involving the inferior phrenic or inferior adrenal arteries and are connected with branches of either the splenic or the hepatic artery.14

There are also connections from the celiac territory to the 12th intercostal artery, to the middle suprarenal artery from the aorta,16 and to the superior lumbar arteries.17 The superior suprarenal arteries from the inferior phrenic artery can also supply the celiac territory.18 The Arc of Bühler is a remnant of the embryonic vertical connection between the superior mesenteric artery and the celiac trunk and is regarded as an atypical collateral, as are potential connections to the middle colic artery.19

In nine of 10 of our patients, robust pancreatic arcades were observed before the procedure that ensured blood supply to the celiac artery territory. In the remaining case, the classic pathway by way of the pancreaticoduodenal arteries was occluded because the gastroduodenal artery had been clipped. This patient had a favorable constellation of an incomplete celiac trunk with collateralization occurring through the accessory LHA to the LHA, as well as atypical collaterals through the middle colic artery from the SMA. Aside from these, pathways from the gastroepiploic artery and the dorsal pancreatic artery were each seen once, but not, however, in any of the other paths described above.

The release of γ-GT, alanine aminotransferase, aspartate aminotransferase, and LDH indicates that the celiac trunk occlusion causes slight acute hepatic cell decay.20 It is only of short duration and, owing to its slight extent, appears to have no clinical relevance for patients with healthy livers. Although an effect on the liver synthesis parameters could not be proven from the slight cell decay found, it must be assumed that at times a slight reduction occurs in liver synthesis performance. The temporary impairment of the liver from occlusion of the celiac trunk was shown here for the first time, to our knowledge. This impairment must be taken into consideration for patients with existing liver function disorders.

Recent reports of individual cases of endovascular occlusion of the celiac trunk, either to treat a celiac trunk aneurysm7, 8, 21 or during stent graft treatment of a thoracoabdominal aortic aneurysm,22 prove the feasibility of endovascular therapy for these two indications. The results from this initial series of patients are encouraging. The treatment was technically successful for all patients, and the morbidity of the procedure was low. It should be noted, however, that occlusion of the celiac trunk could be impossible for some patients owing to complex vascular variations. Therefore, a treatment concept must be created for the individual anatomic situation in each patient before the intervention. Liver function disorders are a relative contraindication.

Conclusion 

Occlusion of the celiac artery by coil embolization for treatment of a celiac trunk aneurysm or for stent graft implantation to treat a thoracoabdominal aortic aneurysm is feasible, with a low risk of ischemia. Numerous collaterals, especially the gastric and pancreaticoduodenal arteries, ensure the supply of the celiac artery territory. Celiac trunks, including the SMA, incomplete celiac trunks, or occlusion of the gastroduodenal artery are relative contraindications. The existence of sufficient collateralization must be established by angiography before the procedure.

Author contributions