Assessment of celiac artery compression using color-coded duplex sonography

Abstract

BACKGROUND:

Median arcuate ligament syndrome (MALS) is a rare condition due to compression of the celiac artery (CA) by an anatomically abnormal median arcuate ligament. With ultrasonography (US) as first-line diagnostic modality in patients with unclear abdominal pain, there is limited data on its diagnostic performance in MALS.

OBJECTIVE:

To investigate the value of CA peak systolic velocity (PSV) in the workup of patients with suspected MALS.

METHODS:

Patients with diagnosis of MALS between 2009 and 2019 were referred by Department of Visceral Surgery after clinical and gastroenterological workup. Diagnosis was confirmed by surgery or further cross-sectional imaging. B-mode US findings and PSV in the CA during various respiratory states were compared between patients with a final MALS diagnosis and patients not meeting the diagnostic criteria.

RESULTS:

Patients with proven MALS (n = 10) had higher median CA PSV during normal inspiratory breath-hold (239 [IQR, 159–327] vs. 138 [IQR, 116–152] cm/s; p < #x003C;< #x200A;0.001), and expiratory breath-hold (287 [IQR, 191–412] vs. 133 [IQR, 115–194] cm/s; p < #x003C;< #x200A;0.001) compared to patients without MALS (n = 26). CA PSV in both inspiratory breath-hold (AUC 0.88, 95% CI 0.77–1.00) and expiratory breath-hold (AUC 0.89, 95% CI 0.78–1.00) was of diagnostic value for confirming MALS. The best diagnostic performance (100% sensitivity, 80% specificity) was found for the combination of CA PSV_expiration + 2.4 · PSV_inspiration > 550 cm/s .

CONCLUSIONS:

Since results on optimal cutoff values are inconsistent, a combination of CA PSVs during breathing maneuvers may help to diagnose or rule out MALS.

Keywords

Median arcuate ligament syndrome MALS Dunbar syndrome color-coded duplex sonography coeliac artery compression syndrome

1 Introduction

Median arcuate ligament syndrome (MALS) is a rare condition caused by compression of the celiac artery (CA) and celiac ganglion, secondary to an anatomical abnormality of the median arcuate ligament [1]. Arterial compression by the crus tends to be intermittent and has been described to be more severe during inspiration than expiration [2]. Typically a diagnosis of exclusion, MALS can cause symptoms such as epigastric pain, postprandial pain, nausea, vomiting, and weight loss, which have been attributed to extrinsic compression of the vasculature and surrounding neural ganglia [3]. After the first description of this rare disorder by Harjola in 1963 [4], the first surgical repair was reported by Dunbar in 1965 [5].

Computed tomography angiography (CTA) can play a role in the diagnosis of MALS by demonstrating the characteristic focal narrowing of the celiac artery in patients presenting with the appropriate clinical symptoms [6]. However, it must be borne in mind that Petnys and colleagues observed celiac artery compression by the MAL on clinical CTA performed for other indications in ∼ 3.5% of patients without MALS-related symptoms [7]. In clinical routine, ultrasonography (US) is often used as the first-line modality in patients with unclear abdominal pain as it is widely available, involves no ionizing radiation exposure, and requires no contrast medium. Most publications on MALS are case reports or small case series with a focus on the therapeutic approach and surgical treatment, while studies investigating imaging techniques are rare. Conventional US and color-coded duplex sonography (CCDS) allow evaluation of flow and function under real-time conditions, even in upright subjects, and provide information on possible functional or collateral mechanisms that is not available with other (nonfunctional) imaging modalities [8, 9]. Gruber et al. demonstrated high sensitivity (83%) and specificity (100%) for the combination of celiac artery end diastolic velocity (EDV) ≥350 cm/s, a 210% change in pulse volume amplitude with inspiration and expiration, and a celiac artery deflection angle of 50° compared to an asymptomatic control group; however, this study included only six patients with confirmed MALS [8].

Thus, the aim of our study was to investigate the value of CA peak systolic velocity (PSV) measured by color-coded duplex sonography (CCDS) in the diagnostic workup of patients with suspected MALS.

2 Methods

2.1 Study cohort

Our department’s radiological information system (RIS) was queried for all US examinations performed for evaluation of suspected MALS between 2009 and 2019. MALS was the suspected diagnosis after the clinical workup, all examined patients have had typical clinically symptoms for MALS (abdominal pain, postprandial pain, weight loss, diarrhea, vomiting and nausea). Demographic data, clinical presentation, and imaging findings were collected from radiology reports (ultrasound parameters) and clinical reports and histories (clinical data). For analysis, any missing data were treated as normal values. While gastroenterological workup (including gastric endoscopy and coloscopy) prior to US examination was generally normal, other conditions explaining symptoms such as peptic ulcers, gall stones, pancreatitis, food intolerance, or atherosclerotic disease had been ruled out during diagnostic workup. Since maximum peak systolic velocity is changed postprandial or by medication, just data of examinations performed in sober state were used.

We excluded patients from the present analysis if: I) US imaging data were missing, II) image quality of CCDS (e.g., inaccurate blood flow velocity measurement, incomplete coverage of vessels) was poor, or III) postprandial examinations.

2.2 Ultrasonography imaging

The standardized ultrasound protocol used in the study patients included B-mode US and CCDS during various respiratory states (normal breathing, inspiration, expiration). First, gray-scale B-mode US of the CA and superior mesenteric artery (SMA) was performed in modified longitudinal and transverse planes with a 1–6 MHz convex array transducer to assess vessel structure (hook shape) and look for stenosis or occlusion. Afterwards, CCDS was performed in longitudinal plane during normal breathing and deep inspiration and expiration using high-end ultrasound systems (Aplio500/Aplio i900; Canon Medical Systems Corporation, Tochigi, Japan). Peak systolic velocity (PSV) of the CA was measured during normal breathing and during breath-hold in both expiration and inspiration using a 90° angle of insonation to the abdominal aorta and measured 2 cm away from the aortic junction with a sample volume of 2–3 mm. Patients with coeliac artery stenosis and artery calcifications were not included in analysis. All US examinations were performed by a specialized radiologist with more than 20 years of experience in the field of US (EFSUMB level 3), and documented B-mode findings were re-analized by two experienced radiologists in consensus.

2.3 Reference standard and surgical treatment

For confirmation of CCDS findings, additional contrast-enhanced computed tomography (ceCT) was performed in 13/36 (36.1%) and contrast-enhanced magnetic resonance imaging (ceMRI) in 19/36 (52.8%) of the patients. Of the 10 patients with a final diagnosis of MALS, seven underwent surgery and three were treated conservatively. One female patient has been diagnosed with recurrent MALS during follow-up.

Surgery was carried out under general anesthesia. All patients were operated on by visceral surgeons with long experience in vascular surgery. Robotic repair using the da Vinci system was performed in three patients. Surgery started with severing of the gastrohepatic omentum for dissection of the median arcuate ligament, followed by displacement of the stomach to the left. Next, the right crus of the diaphragm was identified, the anterior circumference of the aorta above the CA was exposed and dissected down to the origin of the SMA, completely exposing the CA and - if necessary - the upper SMA segment. Finally, the surgeon carefully checked the site to ensure that no musculotendinous tissue compromising vascular structures was left behind. Besides visible confirmation of relaxation of the target vessel(s), intraoperative CCDS was carried out to measure blood flow velocities when residual stenosis was suspected (without documentation). If there were no abnormal macroscopic findings, a control ultrasound was performed the following day.

2.4 Statistical analysis

Continuous variables were tested for normal distribution using the Kolmogorov-Smirnov test and descriptive analysis was performed; nonnormally distributed variables are presented as median and interquartile range (IQR) and were compared using the Mann-Whitney U-test. Categorical variables are presented as numbers and percentages and were compared using Fisher’s exact test or chi² test, as appropriate. Receiver operating characteristics (ROC) analysis was used to investigate the prognostic performance of continuous measures (radiological parameters) and to identify optimal cut-offs for the patient population by calculating Youden’s index; results are reported as area under the curve (AUC) with 95% confidence intervals (CIs). Prognostic accuracy of categorical radiological parameters with regard to study outcomes was determined using univariate logistic regression analysis, and results are reported as odds ratios (OR) with corresponding 95% CIs. For each patient, US findings and matched diagnostic reports were analyzed retrospectively for diagnostic accuracy by determining sensitivity, specificity, positive predictive value (PPV), and negative predictive value (NPV).

A two-sided significance level of α= 0.05 was defined appropriate to indicate statistical significance. All statistical analyses were performed using the SPSS software package (IBM Corp. Released 2016. IBM SPSS Statistics for Windows, Version 26.0. Armonk, NY: IBM Corp.).

3 Results

Our retrospective search retrieved 52 patients who underwent US imaging for routine diagnostic workup of suspected MALS. The final retrospective study cohort included 36 patients with completely documented CCDS-based PSV measurement during normal breathing and during inspiratory and expiratory breath-hold (median age 34, IQR 24–49 years; 83.3% female). Sixteen patients (30.7%) were excluded for the following reasons: 11 patients (21.2%) because of missing PSV data (ultrasound data and clinical history), three because of poor image quality (5.8%), and two patients due to repeat postprandial US examinations (3.8%).

Gastrointestinal (GI) symptoms ranged from upper abdominal pain (32/36 patients; 88.9%), postprandial pain (20/36 patients; 55.6%), weight loss (18/36 patients; 50.0%), diarrhea (14/36 patients; 38.9%), vomiting (8/36 patients; 22.2%) and nausea (19/36 patients; 52.8%). Six patients had symptoms occurring predominantly during running or other sport activity. Patient characteristics are summarized in Table 1.

Table 1
Patient characteristics

Variable Overall No MALS diagnosis MALS diagnosis p-value

n = 36 n = 26 n = 10

Age (years) 34 (24–49) 31 (19–51) 40 (34–49) 0.227

Female sex 30/36 (83.3) 22/26 (84.6) 8/10 (80.0) 0.849

GI symptoms

Upper abdominal pain 32/36 (88.9) 22/26 (84.6) 10/10 (100.0) 0.497

Postprandial pain 20/36 (55.6) 15/26 (57.7) 5/10 (50.0) 0.741

Pain during physical exertion 6/36 (16.7) 4/26 (15.4) 8/10 (80.0) 0.849

Vomiting 8/36 (22.2) 4/26 (15.4) 4/10 (40.0) 0.876

Nausea 19/36 (52.8) 14/26 (53.8) 5/10 (50.0) 0.271

Diarrhea 14/36 (38.9) 8/26 (30.8) 6/10 (60.0) 0.189

Weight loss 18/36 (50.0) 15/26 (57.7) 3/10 (30.0) 0.214

US findings of celiac artery

Hook shape 8/36 (22.2) 3/26 (11.5) 5/10 (50.0) 0.080

Stenosis in B-mode 15/36 (41.7) 6/26 (23.1) 9/10 (90.0) 0.001

Occlusion 2/36 (5.6) 2/26 (7.7) 0/10 (0) 0.741

PSV, regular breathing 140 (104–199) 132 (103–174) 191 (110–301) 0.074

PSV, inspiratory breath-hold 147 (122–210) 139 (116–153) 239 (159–327) <0.001

PSV, expiratory breath-hold 175 (126–216) 135 (119–192) 287 (191–412) <0.001

Variable	Overall	No MALS diagnosis	MALS diagnosis	p-value
	n = 36	n = 26	n = 10
Age (years)	34 (24–49)	31 (19–51)	40 (34–49)	0.227
Female sex	30/36 (83.3)	22/26 (84.6)	8/10 (80.0)	0.849
GI symptoms
Upper abdominal pain	32/36 (88.9)	22/26 (84.6)	10/10 (100.0)	0.497
Postprandial pain	20/36 (55.6)	15/26 (57.7)	5/10 (50.0)	0.741
Pain during physical exertion	6/36 (16.7)	4/26 (15.4)	8/10 (80.0)	0.849
Vomiting	8/36 (22.2)	4/26 (15.4)	4/10 (40.0)	0.876
Nausea	19/36 (52.8)	14/26 (53.8)	5/10 (50.0)	0.271
Diarrhea	14/36 (38.9)	8/26 (30.8)	6/10 (60.0)	0.189
Weight loss	18/36 (50.0)	15/26 (57.7)	3/10 (30.0)	0.214
US findings of celiac artery
Hook shape	8/36 (22.2)	3/26 (11.5)	5/10 (50.0)	0.080
Stenosis in B-mode	15/36 (41.7)	6/26 (23.1)	9/10 (90.0)	0.001
Occlusion	2/36 (5.6)	2/26 (7.7)	0/10 (0)	0.741
PSV, regular breathing	140 (104–199)	132 (103–174)	191 (110–301)	0.074
PSV, inspiratory breath-hold	147 (122–210)	139 (116–153)	239 (159–327)	<0.001
PSV, expiratory breath-hold	175 (126–216)	135 (119–192)	287 (191–412)	<0.001

Continuous variables are given as median (IQR), categorical variables are given as absolute/total numbers (n/N) and percentages in brackets. Abbreviations: MALS, median arcuate ligament syndrome; PSV, peak systolic velocity.

Overall, our analysis included 10 patients with proven MALS who underwent US workup in our department before treatment. During further diagnostic follow-up, patients without MALS (n = 26) were diagnosed with an anatomical variant of the celiac trunk (n = 1), somatoform disorder (n = 9), type Marsh IIIb enteropathy (n = 1), eating disorder (n = 1), intercostal neuralgia (n = 2), gastritis (n = 3), Crohn’s disease (n = 1), GERD (n = 2), gastric cancer (n = 1) and five patients had no final diagnosis.

Representative US images obtained during normal breathing and expiratory breath-hold are shown in Fig. 1. In the total study population (n = 36), median celiac trunk PSV was 140 cm/s (IQR 104–199) during normal breathing, 147 cm/s (IQR 122–210) during inspiratory breath-hold, and 175 cm/s (IQR 126–216) during expiratory breath-hold.

Fig. 1

Representative images of US, CCDS and cross-sectional imaging. a) B-Mode US in transversal plane depicts proximal coeliac artery stenosis (arrow). b–d) CCDS shows typical hook shape sign with relevant stenosis and decreased blood flow of the proximal CA in regular breath, in inspiratoy breath-hold (c, arrow) and increased stenosis in expiratory breath-hold (d, arrow). e, f) CT angiography visualizes the proximal CA stenosis due to fibrous compression of the median arcuate ligament in the same patients as presented in a–d, while ceMRI also show moderte stenosis of proximal CA in normal breathing in another patient. Abbreviations: US denotes ultrasound; CCDS, color-coded duplex sonography; ceCT, contrast-enhanced computed tomography; ceMRI, contrast-enhanced magnetic resonance imagin; MALS, median arcuate ligament syndrome.

As shown in Table 1, GI symptoms of patients with proven MALS differed from those in patients with other final diagnoses. Stenosis was depicted in 90% of the US examinations in patients with MALS.

Patients with proven MALS had higher median pretherapeutic CA PSV during normal breathing (191 [IQR, 110–301] vs. 133 [IQR, 104–184] cm/s; p = 0.074), inspiratory breath-hold (239 [IQR, 159–327] vs. 138 [IQR, 116–152] cm/s; p < #x003C;< #x200A;0.001), and expiratory breath-hold (287 [IQR, 191–412] vs. 133 [IQR, 115–194] cm/s; p < #x003C;< #x200A;0.001) compared to patients with other conditions (Fig. 2). While celiac trunk peak velocity in normal breathing was not associated with MALS (AUC 0.70, 95% CI 0.48–0.91), CA PSV during both inspiratory breath-hold (AUC 0.88, 95% CI 0.77–1.00) and expiratory breath-hold (AUC 0.89, 95% CI 0.78–1.00) was of diagnostic value in confirming MALS (Fig. 3) Calculated patient cohort-optimized PSVS cutoffs of 165 cm/s (normal breathing) and 160 cm/s (inspiratory breath-hold) were associated with a 6.3-fold higher likelihood (95% CI 1.2–31.6) and 16.8-fold higher likelihood (95% CI 2.7–104.8) of MALS, respectively. A cohort-optimized cutoff of 225 cm/s in expiratory breath-hold predicted MALS with a 58.3 times higher likelihood. An expiratory breath-hold PSV cutoff of ≥225 cm/s had 70% sensitivity (95% CI, 40–89), 96% specificity (95% CI, 40–89), 88% PPV (95% CI, 53–98), and 89% NPV (89%; 95% CI, 73–96) compared to CA PSV≥165 cm/s during normal breathing (70% sensitivity [95% CI, 40–89]; 73% specificity [95% CI, 54–86]; 50% PPV [95% CI, 27–73]; 86% NPV [95% CI, 67–95]) and ≥160 cm/s during inspiratory breath-hold (80% sensitivity [95% CI, 49–94]; 81% specificity [95% CI, 62–91]; 62% PPV [95% CI, 36–82]; and 91% NPV [95% CI, 73–98]).

Fig. 2

Boxplot of pretherapeutic median celiac artery PSV in patients with confirmed MALS compared to patients with other diagnoses. Median celiac artery PSV measured during a) normal breathing, b) inspiratory breath–hold, and c) expiratory breath-hold. Abbreviations: MALS, median arcuate ligament syndrome; PSV, peak systolic velocity.

Fig. 3

Diagnostic performance of celiac artery peak systolic velocity (PSV) during normal breathing, inspiratory breath-hold, and expiratory breath-hold with regard to diagnosis of MALS. AUC assessed by ROC analysis of celiac artery PSV during normal breathing (blue), inspiratory breath-hold (red), and expiratory breath-hold (green). Abbreviations: MALS, median arcuate ligament syndrome; PSV, peak systolic velocity; AUC, area under the curve; ROC, receiver operating characteristics.

An exploratory analysis of the dataset revealed two options to separate MALS patients from those with other diagnoses using CCDS-derived CA PSV during expiration and inspiration. The first is a decision tree, which was set to >250 cm/s for PSV during expiration and >158 cm/s during inspiration to differentiate MALS from other conditions (Fig. 4). For our dataset, this decision tree has 100% sensitivity and 76.9% specificity.

Fig. 4

Decision tree and algorithm for diagnosing MALS. Decision tree for diagnosis of MALS based on CCDS-derived CA PSV during expiratory and inspiratory breath-hold. Abbreviations: MALS, median arcuate ligament syndrome; PSV, peak systolic velocity.

A second option to analyze the dataset for relevant cutoffs is to use linear discriminant analysis. Figure 5 shows the separating straight for CA PSV during expiratory and inspiratory breath-hold for MALS patients versus those with other diagnoses. For this separating straight, we found the highest class margin distance.

In our data we could also find the following dependence. $PS V_{ex} + 2.4 \cdot PS V_{in} > 550 cm / s$ (1)

Fig. 5

Decision tree, discriminant analysis, and mental arithmetic method discriminant analysis for MALS (red) versus no MALS (blue). Decision tree (see Fig. 4) and discriminant analysis showing the optimal separating straight for CA PSV in expiratory and inspiratory breath-hold for patients with confirmed diagnosis of MALS (red) compared to patients without (blue). Abbreviations: MALS, median arcuate ligament syndrome; CA, celiac artery; PSV peak systolic velocity.

If equation 1 holds, we can identify MALS with 100% sensitivity and 80.8% specificity. We suggest a simplified version for routine use in the ultrasound lab. This version can be solved using mental arithmetic. $PS V_{ex} + 2 \cdot PS V_{in} > 480 cm / s$ (2)

With equation 2, we again achieve 100% sensitivity and 80.8% specificity in our dataset. However, once Eq. 2 is fulfilled, the more precise Eq. 1 should be used for confirmation, as it features a higher class margin distance.

4 Discussion

In this study we retrospectively analyzed 36 patients (83.3% female, 16.7% male) with suspected MALS who underwent conventional US and CCDS between 2009 and 2019. The focus was on analyzing the potential role of CCDS with PSV measurement during normal breathing and during inspiratory and expiratory breath-hold in the diagnosis of MALS. Overall, 10 patients were diagnosed with MALS and one patient had recurrent MALS two years after surgery. The median age was 34 years, which is comparable to Gruber et al. [8]. All patients with MALS suffered from epigastric abdominal pain, which is a very unspecific symptom that was also present in 23 of our patients without MALS. Other major symptoms of patients with proven MALS are postprandial abdominal pain, exercise-induced pain, nausea, and weight loss [1, 10]. CCDS identified the typical hook pattern of the celiac artery in half of our MALS patients; however, the hook pattern was also seen in 2 of 26 patients without MALS, making it another unspecific sign. Higher US-derived CA PSV is especially obvious in expiration, when pressure of the median arcuate ligament on the celiac artery is higher. PSV of 180 cm/s or above in expiration has been reported as a valid criterion for MALS [8, 11]. Gruber et al. performed duplex sonography in patients with suspected MALS and found 83% sensitivity with 10% specificity [8]. They proposed the combination of expiratory PSV > #x003E;> #x200A;350 cm/s and a celiac trunk deflection angle >50°; however, this is based on a study population including only six MALS patients. In our population we identified a cohort-optimized PSV cutoff of 225 cm/s in expiratory breath-hold, which was associated with a 58.3-fold higher likelihood (95% CI 5.2–651.7) of MALS and a PPV of 88% (95% CI, 53–98). While sensitivity and specificity are comparable, the wide range of CIs in our study using Youden’s index quantification is due to the small number of cases as in the study of Gruber and colleagues. Our dataset also shows that the inspiratory PSV threshold depends on a single datapoint (PSV_inspiration 166 cm/s; PSV_expiration 188 cm/s). This outlier is also the reason why our cutoff differs from that reprted by Gruber et al. [8]. Moreover, examiner dependence of US-derived measurement and the US device used need to be considered. Brody et al. introduced a guideline to predict symptom improvement following MAL decompression from preoperative data. Age and baseline celiac artery expiratory velocity can be used to guide postoperative expectations in patients with MALS [12]. Based on our findings, we propose a straightforward scoring algorithm that is based on PSV and has high sensitivity and specificity.

There are only a few case studies with a focus on surgical outcome in patients with MALS, which is rare and difficult to diagnose. Both diagnosis and treatment are challenging because the underlying pathophysiology is still debated. Hypotheses include celiac artery compression by the median arcuate ligament or by celiac ganglia and permanent overstimulation of celiac ganglia [8]. Compression of the celiac artery leads to high pressures in the poststenotic arterial segment with the risk of aneurysm formation and retroperitoneal bleeding, and early diagnosis is highly desirable. A prevalence of 62% has been reported for poststenotic dilatation, especially affecting the pancreaticoduodenal arteries [3, 14].

US is widely available and provides valuable imaging information next to clinical evaluation without any side effects and is the only functional primary diagnostic method to examine patients with suspected MALS. In addition, CCDS can be used to intraoperatively confirm the success of repair by showing a drop in PSV compared to preoperative velocities in the celiac artery. Used in postoperative follow-up, CCDS can guide further patient care [15 –18]. Other useful CCDS findings are turbulent flow and aliasing, as used in the diagnostic algorithm proposed by Kim et al. [3] US is readily available and provides functional imaging information without exposure to ionizing radiation or contrast agents that may damage the kidneys. This is an important advantage in the young and female patients typically presenting with MALS. For children and in cases of good scan conditions, linear probes can be helpful increasing US image quality. Furthermore, US and CCDS is used frequently, also in outpatient settings in indefinite abdominal pain, the diagnosis algorithm for MALS should be applied in these cases as a standard protocol and/or additional imaging method.

Newer microflow imaging tools such as B-Flow or SMI (superb microvascular imaging) may help in visualization of the stenotic area by increased spatial resolution and also allows flow measurements, potentially be superior in differentiation of CA compression due to fibrous ligament and stenosis due to atherosclerosis. Furthermore, contrast-enhanced ultrasound (CEUS) represents a real-time imaging method allowing the visualization of proximal stenosis in variable breathing maneuvers as well as depicting the perfusion of abdominal organs such as liver, spleen or mural perfusion of small bowel [19].

As a first-line therapy, the surgical division of median arcuate ligament can be performed in traditional open surgery or with the use of minimally invasive techniques, while main advantage of minimally invasive techniques is a shorter hospital stay and reduce of postoperative wound complications. In selected patients, celiac artery bypass or balloon angioplasty with/without stenting are reported to be additional therapeutic approaches for refractory cases, especially when first-line surgical therapies may have proven unsuccessful [14, 20]. Of note, US and CCDS may serve as a first-line imaging tool, but cross-sectional imaging such as CTA or MRA are mandatory in pre-operative assessment of all vascular structures and parenchymatous organs for optimal treatment planning. US and CCDS may helpful both separation of occlusive and non-occlusive disease (by depicting sclerotic parts of the proximal CA) and differentiation of arteriosclerosis and inflammatory changes of the abdominal organs. Nevertheless, visualization of more distal arteries is limited in obese patients or with intense bowel gas. Furthermore, in case of non-occlusive mesenterial ischemia (NOMI), US may be limited by giving a less overview of abdominal structures compared to CT or MRI, especially in assessment both stenosis and organic perfusion. In good scanning conditions, CEUS may serve as a new imaging modality for assessment of organic perfusion, as recently demonstrated in Covid19 patients [19].

Our results confirm the importance of both conventional B-mode US and duplex sonography for both the diagnosis of MALS and posttherapeutic follow-up. Ultrasound is a straightforward functional imaging tool in the diagnosis and follow-up of MALS. While we here present a new diagnostic algorithm based on CCDS, further research in this field is needed to confirm our findings and establish high-quality criteria for the US diagnosis of MALS. Future studies might also investigate whether CEUS can further improve assessment of vascular and organic perfusion.

4.1 Limitations

Our study is limited by the retrospective, single-center design. US is user-dependent, and a high level of experience is especially relevant when performing CCDS to confirm or rule out MALS. In this study, all examinations were performed by a high-experienced radiologist (EFSUMB Level 3) with more than 20 years of experience who used ultra-wideband multi-frequency convex transducers (Canon Aplio i-series). Their superior quality compared to standard transducers may have improved visualization of flow and blood flow velocity measurement. Due to the deficit in documentation in patient records, especially during this period of time for ten years, not all diagnoses of patients without MALS could be identified retrospectively. We also assume a higher number of not documented MALS-cases in our patient records during the last ten years. Another limitation is that our patients were preselected in that they were referred by the Department of Visceral Surgery with suspected MALS after clinical evaluation.

5 Conclusion

Functional US is a fast first-line imaging tool in patients with suspected MALS. US has no adverse effects. The absence of radiation exposure in the diagnostic and therapeutic process is an essential asset in a mainly young and female patient clientele. We propose a straightforward scoring system for the likelihood of MALS (including CA PSV in expiratory and inspiratory breath-hold). Further studies are needed to confirm these findings.

Conflict of interest

None of the authors reports a relationship with industry and other relevant entities –financial or otherwise –that might pose a conflict of interest in connection with the submitted article.

Footnotes

Acknowledgments

The authors thanks Ms. Bettina Herwig for proofreading the manuscript.

References

Loukas

, Pinyard

, Vaid

, Kinsella

, Tariq

, Tubbs

. Clinical anatomy of celiac artery compression syndrome: A review. Clin Anat. 2007;20:612–7. DOI 10.1002/ca.20473.

Gloviczki

, Duncan

. Treatment of Celiac Artery Compression Syndrome: Does It Really Exist? Perspect Vasc Surg Endovasc Ther. 2007;19:259–63. DOI 10.1177/1531003507305263.

Kim

, Lamb

, Relles

, Moudgill

, DiMuzio

, Eisenberg

. Median Arcuate Ligament Syndrome-Review of This Rare Disease. JAMA Surg. 2016;151:471–7. DOI 10.1001/jamasurg.2016.0002.

Harjola

. A rare obstruction of the coeliac artery. Report of a case. Ann Chir Gynaecol Fenn. 1963;52:547–50.

Dunbar

, Molnar

, Beman

, Marable

. Compression of the Celiac Trunk and Abdominal Angina. Am J Roentgenol. 1965;95:731–44. DOI 10.2214/ajr.95.3.731.

Horton

, Talamini

, Fishman

. Median Arcuate Ligament Syndrome: Evaluation with CT Angiography. RadioGraphics. 2005;25:1177–82. DOI 10.1148/rg.255055001.

Petnys

, Puech-Leão

, Zerati

, Ritti-Dias

, Nahas

, Neto

, et al. Prevalence of signs of celiac axis compression by the median arcuate ligament on computed tomography angiography in asymptomatic patients. J Vasc Surg. 2018;68:1782–7. DOI 10.1016/j.jvs.2018.04.044.

Gruber

, Loizides

, Peer

, Gruber

. Ultrasound of the median arcuate ligament syndrome: a new approach to diagnosis. Med Ultrason. 2012;14:5–9.

Wolfman

, Bluth

, Sossaman

. Median Arcuate Ligament Syndrome. J Ultrasound Med. 2003;22:1377–80. DOI 10.7863/jum.2003.22.12.1377.

10.

Köhler

, Schardey

, Bettels

, Berger

, Wohlgemuth

, Heindel

, et al. Median arcuate ligament syndrome - imaging presentation and interdisciplinary management. Rofo. 2018;190:907–14. DOI 10.1055/a-0591-5008.

11.

Sproat

, Pozniak

, Kennell

. US case of the day. Median arcuate ligament syndrome (celiac artery compression syndrome). RadioGraphics. 1993;13:1400–2. DOI 10.1148/radiographics.13.6.8290734.

12.

Brody

, Randall

, Amdur

, Sidawy

. A predictive model for patients with median arcuate ligament syndrome. Surg Endosc. 2018;32:4860–6. DOI 10.1007/s00464-018-6240-y.

13.

Chivot

, Rebibo

, Robert

, Regimbeau

J-M

, Yzet

. Ruptured Pancreaticoduodenal Artery Aneurysms Associated with Celiac Stenosis Caused by the Median Arcuate Ligament: A Poorly Known Etiology of Acute Abdominal Pain. Eur J Vasc Endovasc Surg. 2016;51:295–301. DOI 10.1016/j.ejvs.2015.10.025.

14.

Nasr

, Faraj

, Al-Kutoubi

, Hamady

, Khalifeh

, Hallal

, et al. Median Arcuate Ligament Syndrome: A Single-Center Experience with. 23 Patients. Cardiovasc Intervent Radiol. 2017;40:664–70. DOI 10.1007/s00270-016-1560-6.

15.

Reilly

, Ammar

, Stoney

, Ehrenfeld

. Late results following operative repair for celiac artery compression syndrome. J Vasc Surg. 1985;2:79–91.

16.

Kohn

, Bitar

, Farber

, Marston

, Overby

, Farrell

. Treatment options and outcomes for celiac artery compression syndrome. Surg Innov. 2011;18:338–43. DOI 10.1177/1553350610397383.

17.

Vaziri

, Hungness

, Pearson

, Soper

. Laparoscopic Treatment of Celiac Artery Compression Syndrome: Case Series and Review of Current Treatment Modalities. J Gastrointest Surg. 2009;13:293–8. DOI 10.1007/s11605-008-0702-9.

18.

Carbonell

, Kercher

, Heniford

, Matthews

. Multimedia article. Laparoscopic management of median arcuate ligament syndrome. Surg Endosc. 2005;19:729. DOI 10.1007/s00464-004-6010-x.

19.

Jung

, Stroszczynski

, Jung

. Contrast enhanced ultrasonography (CEUS) to detect abdominal microcirculatory disorders in severe cases of COVID-19 infection: First experience. Clin Hemorheol Microcirc. 2020;74. DOI 10.3233/CH-209003.

20.

Rubinkiewicz

, Ramakrishnan

, Henry

, Roy

, Budzyński

. Laparoscopic decompression as treatment for median arcuate ligament syndrome. Ann R Coll Surg Engl. 2015;97:e96–9. DOI 10.1308/rcsann.2015.0025.