Staged treatment for pancreaticoduodenal artery aneurysm with coeliac artery revascularisation: Case report and systematic review

Abstract

Background

Despite being rare, pancreaticoduodenal artery aneurysms (PDAAs) carry a risk of rupture of up to 50% and are frequently associated with coeliac artery occlusion.

Methods

PubMed and Embase databases were searched using appropriate terms. The systematic review was conducted according to PRISMA guidelines.

Results

We present the case of a 2 cm pancreaticoduodenal artery aneurysm pre-operative angiography demonstrated that the coeliac artery was occluded and the pancreaticoduodenal artery was providing collateral blood supply to the liver. Treatment was a staged hybrid intervention inclusive of an aorto-hepatic bypass using a 6 mm graft, followed by coil embolisation of the aneurysm. We also present a systematic review of the management of PDAAs. Two hundred and ninety-two publications were identified initially with 81 publications included in the final review. Of the 258 peripancreatic aneurysms included, 175 (61%) were associated with coeliac artery disease either occlusion or stenosis. Abdominal pain was the main presentation in 158 cases. Rupture occurred in 111 (40%) of patients with only ten (3.8%) cases being unstable on presentation. Fifty (18%) cases were detected incidentally while investigating another pathology. Over half the cases (n=141/54.6%) were treated by trans arterial embolisation (TAE) alone, while 37 cases had open surgery only. Twenty-one cases needed TAE and a coeliac stent. Seventeen cases underwent hybrid treatment (open and endovascular). Sixteen cases were treated conservatively and in 26 cases, treatment was not specified.

Conclusion

PDAAs are commonly associated with coeliac artery disease. The most common presentation is pain followed by rupture. The scarcity of literature about true peripancreatic artery aneurysms associated with CA occlusive disease makes it difficult to assess the natural history or the appropriate treatment. Revascularisation of hepatic artery is better done with bypass in setting of median arcuate ligament compression and occluded celiac trunk.

Introduction

Visceral artery aneurysms (VAAs) are rare with an incidence of between 0.01% and 0.2%. ¹ The clinical presentation is non-specific ranging from incidental finding to rupture and shock. Being rare, the actual incidence of rupture is unknown.^1,2 Pancreaticoduodenal artery aneurysms (PDAAs) comprise about 2% of visceral artery aneurysms. Multiple pathologies are reported to be involved in its aetiology, including pancreatitis, surgical trauma, collagen vascular disease and occlusion or stenosis of the coeliac artery due to arteriosclerosis or median arcuate ligament compression. The risk of rupture in these aneurysms is independent of the size of aneurysm ³ treatment of PDA aneurysms appropriate even if asymptomatic. Endovascular treatment is considered to be the first line treatment due to its lower morbidity and mortality; however, flow to the hepatic artery should not be compromised.^4,5

Our aim is to present a case report of a patient with a PDAA and coeliac artery occlusion in addition to a systematic review of this rare condition.

Case report (Figure 1)

A 39-year-old lady presented with an incidental finding of a PDA aneurysm on CT following four episodes of intermittent abdominal pain, one of which resulted in ED attendance. The epigastric pain was associated with vomiting, initially of undigested foods and then became bilious in nature. The patient also complained of intermittent dyspepsia, with no identifiable triggers. She denied haematemesis, melaena, diarrhoea, constipation or weight loss. Her abdominal exam was unremarkable. Her past medical history was significant for a papillary thyroid carcinoma (thyroidectomy 2010), non-smoker, no history of portal hypertension or liver disease. CTAP revealed a 2.7 cm aneurysm involving the proximal inferior pancreaticoduodenal artery with no other aneurysms and occluded proximal celiac artery.

The case was discussed in our Multi-disciplinary meeting (MDT) and the decision was to treat the aneurysm giving the young age of the patient and size of aneurysm. Initially coil embolisation was considered; however, first angiography demonstrated complete occlusion of the coeliac artery from the origin to bifurcation and the PDA supplied the entire hepatic circulation. This led to concern that embolisation of the aneurysm would risk foregut ischaemia and so embolisation was aborted and a decision was made to perform first a bypass from the aorta to the hepatic artery and to embolise the aneurysm at a later stage. (Figure 1(a)–(d))

Figure 1.

Angiographic images for the aneurysm pre-embolisation and after A.B.C.D first preoperative angiogram showing the aneurysm and collateral vessels E.F second angiogram images before and after embolisation.

The operation was performed using a transabdominal approach with control of the supraceliac aorta at the hiatus. The hepatic artery was controlled at the lesser sac. The median arcuate ligament was divided as was very tight and we performed aorto-hepatic bypass was performed using 6 mm Dacron end to side anastomosis for both aorta and hepatic artery. The patient recovered well after the bypass and was discharged home for scheduled embolisation of the aneurysm 2 weeks later. The second stage involved Right femoral access and through the Superior mesenteric artery, cannulation of inferior PDA using microcatherters and 0.014 inch wire with help of our interventional radiology, coils were inserted in the aneurysm to occlude the aneurysm itself and completion angiogram showed complete occlusion of the aneurysm (Figure 1(e) and (f)). Follow-up CTA after 6 weeks and 6 months revealed complete occlusion of the aneurysm and a patent bypass with planned yearly CTA to assess the bypass.

Systematic review

Aim

To review the treatment of PDAAs in association with coeliac artery disease.

Eligibility criteria

Studies were included if they were prospective or retrospective studies which evaluate the relation between PDAAs and coeliac artery occlusive disease. Studies were excluded if they had insufficient data, duplicated studies, non-English, consisted of a comment or letter to a study or included aneurysms in other visceral arteries.

Literature search strategy

The literature review was conducted in PubMed and Embase databases. Two groups of key words were used and combined with Boolean operator ‘AND’. The first group comprised: ‘pancreaticoduodenal artery aneurysm’ or ‘gastroduodenal artery aneurysm’. The second group was: celiac artery OR celiac bypass OR aortohepatic bypass OR celiac artery stent. Additional studies were also examined from references of original studies on the topic. All studies were carefully examined, and selected as per the eligibility criteria outlined above. Each publication was assessed as regards patient characteristics, aneurysm site, presentation, treatment and outcome if mentioned. PRISMA was followed.

Data abstraction and quality assessment

Two reviewers (BG and LA) did the search independently and any disagreement was resolved by discussion or arbitration by the senior researcher (AOC). The main findings of included studies are presented in Table 1. The Down’s and Black Tool ⁶ was used for the assessment of quality of included studies. This tool has 27 questions that assess internal validity, external validity and reporting quality. For 26 of the 27 questions, a given study is awarded one point or no point and then for the final question on the sample size calculation, a score of 0–5 is given. Higher total scores signify higher quality. Case reports were not included in quality assessment.

Table 1.

Characteristics of included paper.

No	Publication	Publication year	Number of cases	Gender	Age	Presentation	Treatment	FU
1	Zhang et al 24	2019	1	Male	50	Rupture on CT epigastric pain	TAE then celiac stent 2 weeks later for pain development	24 m with CTA and good
2	Gabrielli et al 27	2019	1	Male	56	Pain and prerupture signs on CTA. Recurrent pancreatitis history and open AAA 10 years earlier	TAE	Uneventful
3	Degheili et al 28	2017	1	Male	77	Severe upper abd pain and drop of HB rupture on CT	TAE with preservation of post branch	Uneventful
4	Takeuchi et al 25	2017	4	1 male		One ruptured but stable upper abdominal pain in all the others with CT showed MAL	TAE for all MAL resection one patient celiac stents for 2 patients bypass from rt renal to CHA by GSV for 1	The 2 stents were thrombosed one underwent stent again with high radial force
5	Kamarajah et al 29	2019	1	Female	71	Incidental asym	Conservative	6 m fu no issues
6	Miyahara et al 30	2019	8	3 males		One ruptured	NS	NS
7	Ikeda et al 31	2013	11	2 males		9 asym detected on US one hypoglycaemia work up one back pain with chronic alcoholic pancreatitis	5 TAE 6 isolation of aneurysm by covered stent 1 celiac artery stent SES	One stent was occluded
8	Tien et al 32	2004	2	1 male	46 and 66	Rupture with abdominal pain one stable and the other is unstable	Stable: TAE and celiac stent same time unstable: Failed TAE with open surgery was done then 2 weeks later CA stent	NS
9	Yamamoto et al 33	2015	63 female PDA, splenic aneurysm and MAL and gall stone	Female	63	Pain	Cholecystectomy as gall stone. Splenectomy as aneurysmal also with ligation of splenic and PDA and MAL release	NS
10	Otsuka et al	2018	2 59 female 53 male	1 male	59–53	Ruptured	Exclusion of aneurysm by coils but no repair of celiac	NS
11	Boll et al 34	2017	20 15 had celiac stenosis 65 years old mean age	8 male	65 mean age	9 rupture all stable retroperitoneum 3 asym 8 abdominal pain	13 TAE (9 had celiac stenosis with only one had prophylactic bypass) 3 open exclusion with bypass 4 conservative	No hepatic ischaemia
12	Bonardelli et al 16	2020	25	7 male	64 mean age	3 asym 17 ruptured ( 2 surgery and 15 endo	16 TAE. 6 open repair 4cMAL release and 2 aortohepatic bypass. 3 conservative were asym	4 cases had visceral ischaemia after EV without revasc and three of them needed revascularised open 4 patients died in 3 EV group and 1 OR
13	Ohta et al 35	2020	1	Female	78	Pain	Aortohepatic fist then 2 weeks later TAE	Good
14	Suzuki et al 36	1998	1	Male	43	Asym us detected	Open resection of aneurysm with MAL release	NS
15	Ikoma et al 37	2012	2	Male	45	Pain	TAE with SMA stent	NS
16	Chivot et al 38	2016	10	4 male		Rupture and abdominal pain only 1 is unstable	2 surgical excision 8 TAE	No MAL ttt 1 death
17	Quandalle et al 39	1990	2	Male	42–62	One incidental with haematemesis for varices one rupture	One unfit and conservative treatment one surgical repair with MAL release	Unfit patient died after one month
18	De Santis et al 13	1	57 female	Female	57	Pain and rupture	TAE of 3 PDA aneurysm with no revascularisation	No comp
19	Shibata et al 21	2017	1	Male	59	Asym	TAE	Pressure monitoring of CHA pressure showed no drop
20	Kitaoka et al 40	2014	1	Male	54	Pain	TAE with SMA stent	NS
21	Dave et al 41	2010	7	5 males	55 mean age	Abdominal pain all 2 ruptured shocked 2 rectal bleeding	7 TAE one failed but had CT guided injection. One CA stent	No visceral ischaemia 11 years FU
22	Kobayashi et al 42	2004	1	Female	57	Rupture and pain	TAE	Ant branch was patent and 2 years later no comp
23	Teng et al 12	2006	1	Male	40	Abd pain rupture unstable	Aortohepatic SMA bypass by bifurcated graft then coiling	Good
24	Suzuki et al 43	2008	7	4 males	67 mean age	Pain sudden upper with rupture in all	TAE with no revascularisation	Uneventful
25	Miyamotto et al 44	2018	1	Male	65	Pain	Laparscopic MAL release and then TAE later	Uneventful
26	Yin et al 45	2015	1	Male	84	Obstructive jaundice	Coil embolisation with no need to drain the liver	Uneventful
27	Pavić et al 46	2020	2	Females	45–59	1 postprandial pain asym	NS	NS
28	Sugiyama et al 47	2007	5	1 male	54 mean age	3 rupture (2 GDA) 2 incidental	3 rupture TAE 2 asym conservative MAL resection one case	Uneventful
29	Yamana et al 48	2020	1	Female	59	Rupture stable with abdominal pain	TAE	Enlarged on FU and new aneurysm developed in arcade so MAL release and aortohepatic bypass done and regression of aneurysms
30	Heo et al 49	2018	1 GDA 31 other visceral arteries	5 males	64 mean age	One rupture 31 asym	TAE only for ruptured one but conservative treatment for the other	NS
31	Hildebrand et al 50	2007	3	Female	48–53-46	Incidental all	ALL had open surgery ligation with monitoring circulation intraop US and if compromise bypass was done 2 cases end to end and one just ligated	NS
32	Yamamoto et al 10	2011	1	Male	77	Ruptured and unstable	TAE post PDA coiled	NS
33	Ritter et al 51	2010	1	Female	36	Asym	Bypass aortohepatic then resection open	NS
34	Gandini et al 52	2017	1	Female	64	Asym	TAE for 2 aneurysm post PDA and stent to celiac	NS
35	Wattez et al 11	2013	1	Female	61	Jaundice compression	Hybrid coiled and bypass aortohepatic PTFE then US guided intraop coiling and aneurysm injection then intaop SMA angio confirmed	NS
36	Flood et al 53	2013	14	10 female	Mean 46	2 asym 1 rupture and shock 11 pain	13 TAE 4 celiac stents 1 PTA CA 1 laparscopic MAL resection and the aneurysm thrombosed after	NS
37	Ignashov et al 54	2005	2	Female	46–52	asym	TAE then MAL release one case the other MAL first then open surgery for aneurysm resection 3 months between in each	NS
38	Antoniak et al 55	2017	15	7 males	64 mean age	NS	NS	NS
39	Ikeda et al 56	2007	3	3 female	45	asym	TAE	NS
40	Sanal et al 57	2016	1	Female	41	Pain	TAE with help of retrievable stent	Good circulation at the end
41	Reddy et al 58	2019	1	Male	72	Ruptured abd pain and nand V with duodena obst with MAL	TAE	NS
42	Hayashi et al 59	2008	1	Female	56	Rupture and duodenal obst	TAE	NS
43	Armstrong et al 60	2014	1	Female	25	Incidental as donor for kidney work up	TAE	NS
44	Skalický et al 26	2021	1	Male	45	Ruptured	TAE with CA stent	Developed compartment and laparotomy was done
45	Antoniak et al 15	2018	21	3 males	55 mean age	All abdominal pain	14 TAE 6 open surgery 1 TAE and bypass	2 open surgery has major comp one infected aortohepatic one liver injury one graft occluded
46	Aryal et al 23	2017	1	Female	54	Asym	Ilio hepatic bypass and resection of one aneurysm then 3 weeks later coils for the other one	NS
47	Katsura et al 61	2010	8	3 males	54 mean age	5 ruptured 3 asym CA occluded 2 CA stenosis 4	7 TAE 1 surgery with TAE 1 open surgery	6 years FU no growth
48	Ducasse et al 62	2004	1	Male	54	Rupture stable	TAE then 4 weeks later MAL resection	NS
49	Sakatani et al 63	2016	1	Female	46	Pain	TAE	NS
50	Kimura et al 17	2020	1	Male	72	Ruptured post PDAA	TAE	Developed gall bladder necrosis 5 days after so CC and MAL release diaed 10 days after with another aneyrysm in ant br was not treated
51	Salomon du Mont et al 18	2017	1	Female	53	Rupture MAL	Revascularization of celiac artery open with no TAE	3 m asym
52	Bauder et al 64	2018	1	Female	46	Upper abd pain with prior rupture that caused haematoma compressed duodenum	TAE	Good FU
53	Kaszczewski et al 65	2020	1	Female	45	Abd pain	MAL release and vein aortohepatic bypass then TAE 2 weeks later	48 m FU
54	Matsumura et al 66	2013	3	1 male	55	Abd pain and shock ruptured 2 mod 1 severe shock	TAE in all	NS
55	Takeuchi et al 67	2017	1	Female	80	Vomiting and pain rupture haematoma compressed duodenum	TAE	NS
56	Sgroi et al 68	2015	3	2 males	57 mean	2 asym 1 rupture	TAE for the three	No CA TTT with no sym
57	Uher et al 69	1995	4	2 male	45	One rupture 3 asym	2 TAE 2 conservative	NS
58	Peyrottes et al 70	2018	4	1 male	46	2 ruptured 2 asym	2 TAE (ruture) 3 MAl release	One case ruptured developed abd pain after coies and MAL release was done another 2 cases MAL release same time of surgery 4th good collateral and no release done
59	Storm et al 71	2015	I	Male	65	Initially with stroke then developed pain and shock 3 h later No thrombolysis giveCoiled	TAE	He has CT brain showed infarction which likely ischaemic by acute hypotension but he left well
60	Rebelos et al 72	2019	1	Female	44	Ruptured and abd pain	TAE	MAL not treated stenosis
61	Hong et al 73	2015	37 female	Female	37	Pain	Surgery with end-to-end anastomosis and clips for aneurysm as intracranial aneurysm	NS
62	Bersani et al 74	1989	1	Male	45	Asym	NS	Celiac stenosis
63	Jensen et al 75	2020	1	Male	40	Rupture and haematoma compressed duodenum	TAE then 6 m later MAL resection	Uneventful
64	Hiramatsu et al 20	2014	1	Female	37	Incidental	Laparscopic release of MAL only	CT FU 1 month good flow in CA and no aneurysm
65	Klonaris et al 76	2013	53 female	Female	53	Abdominal pain pancreatitis history	TAE	NS
66	Bageacu et al 77	2006	9 PDAAs 4–30 mm	3 males	45	7 ruptured with pain 2 asym	9 TAE one surgery for bleeding	Celiac stenosis in 3 cases
67	Tarazov et al 78	2001	1	Female	56	Abd pain and weight loss	TAE then 6 m later pain persist so surgery MAL release as stent tried first but failed to pass	NS
68	Imamura et al 79	1998	1	Female	43	Incidental	TAE	NS
69	Hughes et al 80	2016	1	Female	75	Incidental	PTFE aortohepatic with resection of aneurysm failed coiling initially	NS
70	Itoh et al 2	2010	4	1 male	45	Vomiting and duodenal obst	3 conservative one surgery for aneurysm and MAL	All MAL celiac stenosis 3 > 50% and one <50%
71	Takase et al 81	2014	2	Females	46	Ruptured	TAE	No TTT for MAl
72	Iusco et al 82	2006	1	Female	57	Rupture and shock	TAE	NS
73	Schoder et al 83	2000	1	Female	39	Asym	TAE	NS
74	Ogino et al 84	2002	1	Female	54	Rupture unstable	TAE	NS
75	Schweizer et al 85	1999	1	Female	72	Shock and rupture after hip replacement	Life saving open surgery No TTT of CA	Splenic infarction after
76	Majerus et al 86	1998	3	Female	45	Rupture	TAE in true aneurysms 2 open surgery for mycotic	2 were mycotic
77	Habre et al 87	2005	1	Female	54	Rupture as perforated DU	Ligation surgery	NS
78	El Alaoui M 88	2005	1	Male	55	Rupture	TAE	NS
79	Rai et al 89	2010	1	Female	36	Rupture	TAE	NS
80	El Hassani et al 90	2021	3	1 male	54	Rupture	TAE	NS
81	Mora J et al 19	1976	1	Female	44	Gall stone with abd pain and calcification	Aortohepatic Dacron graft and the aneurysm disappeared on FU	NS

Abbreviation: GSV:GREAT SAPHENOUS VEIN. NS: NON SIGNIFICANT. FU: FOLLOW UP. TTT: TREATMENT.

Results

Two hundred and ninety-two publications were identified and after duplicated publications were removed, 277 were included for initial assessment. 177 publications were excluded based on title and abstract as they did not fulfilling the search criteria. A further 19 reports were excluded as they included coeliac artery aneurysms leaving 81 publications for inclusion in the review. (Figure 2)

Figure 2.

Flowchart of search.

Table 1 summarises the main findings of the eligible studies. Seventy-one were case reports and 10 studies were case series. No randomised controlled studies were identified. Using the Down’s and Black Tool, quality assessment of included studies ranged between poor and moderate quality.

Presentation

Two hundred and fifty-eight pancreaticoduodenal aneurysms (PDAAs) were identified, 175 (61%) were associated with coeliac artery disease either occlusion or stenosis.

Abdominal pain was the main presentation in 158 cases (61%) (Figure 3) Rupture occurred in 111 (43%) patients with only ten patients (3.8%) being unstable on presentation. Fifty (19%) cases were detected incidentally while investigating another pathology. Nine cases (3.9%) presented with vomiting secondary to duodenal obstruction. Two cases (0.89%) presented with rectal bleeding and two cases presented with obstructive jaundice from compression on the common bile duct. (Figure 3)

Figure 3.

Presentation of PDAAs according to literature review.

Treatment

One hundred and forty-one (54.6%) cases were treated by coil embolisation alone. Of these five failed and subsequently three underwent open surgical repair, one, CT guided injection and one was treated conservatively.

Thirty-seven (14%) cases had open surgery alone. Sixteen of them had exclusion and bypass with nine cases have additional Median Arcuate Ligament release (MALR). Eleven cases had exclusion alone while three cases had MALR release alone and seven cases had exclusion with MALR.

Twenty one (8%) cases underwent TAE and coeliac stenting. Three stents were occluded on follow up with one having subsequent restenting with a high radial force stent and the other two patients were treated conservatively. One patient of this group had additional open surgery for decompression after compartment syndrome from rupture.

Seventeen (6.5%) cases underwent hybrid treatment. Nine cases had TAE and bypass with six performed at the same sitting. Inflow was from the aorta in seven cases, right renal in one case and right common iliac artery in the remaining case. Seven cases needed TAE and median arcuate ligament release, four were performed at the same sitting and one before and one after TAE. One case had an aorto-hepatic bypass and intraoperative US guided coiling of the PPAA. Sixteen (6.2%) cases underwent conservative treatment. In 26 (10%) cases, treatment was not specified. (Figure 4)

Figure 4.

Treatment of PDAAs according to literature review.

Follow-up

Forty-two studies reported follow-up. Two stents were occluded on follow-up. Both were treated conservatively. One bypass was occluded on follow-up with no further treatment required. One study reported liver injury during the bypass surgery and another study reported graft infection which was treated conservatively. Seven cases died with three deaths in endovascular group and two in open surgery group and one died just before the surgery. Yamena and his colleagues reported that after ligation of PDA aneurysm, it continued to enlarge on follow-up and the patient developed a new PDAA besides the excluded one and these aneurysm regressed only after aortohepatic bypass was performed with normalisation of the pressure in peripancreatic arterial arcade.

Discussion

PPAAs are considered a rare form (2%) of all visceral artery aneurysms (VAAs) and are commonly associated with coeliac artery disease which has been reported in literature between 59% and 95%. We think that the common association between PDAAs and celiac artery occlusive disease is due to a haemodynamic change in the peripancreatic arterial arcade with higher pressure through collaterals with resultant increase in the incidence of PPAAs formation.^7,8

Yoon and his colleagues ⁹ had developed two theories to examine the relation between the CA stenosis and PDAAs development. They used electrical simulation circuit to test their theory. The stenosis first theory showed that when CA stenosis develops, it is associated with increase flow in peripancreatic arterial arcade and due to high pressure in the aorta it favours PDAAs development through higher flow through these weak collaterals. The other theory, the aneurysm first theory suggests that when a PDAA develops, it steals the blood from the CA with consequent collapse of the CA. They conclude that either condition could be the predisposing factor for the other. In our opinion, the stenosis first theory is more related to the formation of these PDAAs as shown by Yamamoto and his colleagues when they ligated the PDAA and the aneurysm continued to enlarge and the patient developed a new aneurysm in the peripancreatic arcade that did not regress except after normalisation of the pressure with aortohepatic bypass. ¹⁰

Our review showed that 40% of cases presented with rupture PDAAs but only 10% were unstable at the time of presentation. Despite being rare, treatment of PDAAs is advised as rupture is independent of size and the associated mortality is in excess of 50%. ¹¹ Most cases who presented with rupture were stable on presentation which could be explained by rupture in the retroperitoneal space rather than intraperitoneal with small, confined space in peripancreatic area limits haematoma expansion. Our case presented with abdominal pain and the aneurysm was picked up incidentally on CT. ^10,12,13

Treating these lesions should not be dependent on size of the aneurysm as many cases mentioned rupture even when the aneurysm was <1 cm. Surgery, endovascular, hybrid treatment and conservative treatment are all treatment options with endovascular favoured as it is less invasive and surgical accessibility can be difficult. Treatment recommendations are difficult to assess as most publications are case reports. ¹⁴ However, trans arterial embolisation (TAE) is getting popularity as first line treatment as it is less invasive and has a reduced mortality for these type of aneurysms with difficult surgical access. The main drawback and concern are preserving the foregut circulation in the presence of CA disease as TAE can compromise the hepatic blood flow with insufficient collateral circulation. ¹⁵

As shown in the review, trans arterial embolisation (TAE) with coils was the most common reported treatment (54%). Recommendations for revascularisation of associated coeliac artery disease cannot be clearly defined from the literature.

Bonardelli et al. ¹⁶ reported 5 open repair, 3 MAL release and 2 aorto-hepatic bypasses. They coiled 16 patients and 4 of them developed visceral ischaemia with three of them requiring further urgent open revascularisation. They had three mortalities in the endovascular group and one in the open group. Kimura et al. ¹⁷⁾ described a case of embolisation for posterior branch PDAA that complicated with gall bladder necrosis and they had to release MAL 5th day post op but the patient died on 10th day as another aneurysm in the anterior branch of PDAA which ruptured. TAE alone may jeopardise the foregut circulation with ischaemia of the liver or gall bladder and revascularisation seems to be a safe alternative way to revascularise the foregut before TAE. We thought initially that TAE will be sufficient treatment, however, on performing the angiogram, the entire hepatic circulation seems to get its supply from PDA and performing TAE could jeopardise the hepatic circulation. Consequently, the case was aborted and discussion in our MDT revealed that staged treatment with revascularization first then embolization later could be the safer treatment in this young female.

Additionally, some authors have reported continuous growth of the aneurysm despite coiling as the pressure inside the aneurysm was presumably still high through the collateral circulation. Yamana et al. ³³ used coils to occlude the PDAA first but the aneurysm continued to grow on follow-up with development of a new aneurysm in peripancreatic arterial arcade that regressed only after they did aortohepatic bypass to revascularise the celiac artery with resultant normalisation of the pressure inside the aneurysm. Several authors support normalisation of the pressure in the peripancreatic arterial arcade to avoid growth of these aneurysms and revascularisation of the foregut territory to avoid ischaemic complications.^20,21

Revascularisation alone was adopted by some authors in their case reports as they argue that the aneurysm will regress when pressure is normalised and several authors reported regression of aneurysm by revascularisation^18,19 or release of the Median arcuate ligament only (MAL). ²⁰ However, we do not feel that release of MAL only is completely safe as it was illustrated by case presentation by Kimura et al. ¹⁷ when they coiled ruptured posterior PDAA but the patient developed Gall bladder necrosis 5^th day postoperative and they have to revascularise the celiac artery by MAL release but the patient died 1 week later from ruptured anterior PDAA. This may denote that revascularisation alone with theory to normalise the pressure in the peripancreatic arterial arcade did not prevent rupture of PDAA and the aneurysm itself should be treated additionally to prevent progression and rupture.

Shibata and his colleagues used intraoperative haemodynamic pressure monitoring to guide revascularisation in their case report where the patient had anterior branch pancreaticoduodenal artery aneurysm and the posterior pancreaticoduodenal artery (PDAA) was small in calibre, and the patient was considered high risk for liver ischaemia. Surprisingly, no change in pressure on occlusion balloon test to the anterior branch PDA and posterior branch PDA enlarged with circulation maintained to liver. ²¹

Aorto-hepatic bypass as employed in our case was the most commonly used bypass in the literature. Takeuchi and his colleagues ²² reported bypass from the right renal to the common hepatic artery using great saphenous vein. Aryal and co-authors also used the common iliac artery as inflow. ²³

A staged hybrid approach appears to be the safer option combining both TAE to occlude the aneurysm and revascularisation to normalise the pressure and avoid foregut ischaemic complications. Wattez et al. ¹¹ used a hybrid approach differently; they performed an aorto-hepatic bypass as well as ultrasound-guided intraoperative coiling and aneurysm injection.

Endovascular treatment for both CA disease and PDAA could be an appealing useful option as used by some authors. However, the presence of associated MAL syndrome could jeopardise the stent and cause occlusion. ^24–26

Conclusion

PDAAs are commonly associated with coeliac artery disease. The most common presentation is pain followed by rupture. The scarcity of literature about true peripancreatic artery aneurysms associated with CA occlusive disease makes it difficult to assess the natural history or the appropriate treatment. Staged hybrid treatment seems to be safer approach for such disease entity. Revascularisation of hepatic artery is better done with bypass in setting of in presence of median arcuate ligament compression and occluded celiac trunk.

What does this study/review add to the existing literature and how will it influence future clinical practice

We hereby present one rare and difficult case with literature review. The association between pancreatic-duodenal arterial aneurysm and celiac artery stenosis is not well explored in literature and the management is not well understood. We present our case report with our approach explained to avoid foregut ischaemia and we did literature review to highlight this rare presentation.