Endoscopic features for early detection of superior mesenteric artery syndrome in children

Superior mesenteric artery syndrome (SMA) syndrome is a rare cause of duodenal obstruction. Diagnostic delay of SMA syndrome is common due to its rarity and lack of index of clinical suspicion. Present study aims to explore the endoscopic features for early decision to evaluate SMA syndrome in children. In case controlled observation study, we identied three endoscopic ndings in enrolled cases like as, nding I: a pulsating band like luminal narrowing of the third part of the duodenum without no expansion over one third during air insuation for at least 15 seconds, nding II: marked dilatation of the duodenal second part during air insuation at the duodenal third part, and nding III: bile lake in the stomach. SMA syndrome was conrmed with upper GI series in patients with endoscopic nding I or I plus more. We analyzed positive endoscopic ndings related with SMA syndrome.


Abstract
Background Superior mesenteric artery syndrome (SMA) syndrome is a rare cause of duodenal obstruction. Diagnostic delay of SMA syndrome is common due to its rarity and lack of index of clinical suspicion.
Present study aims to explore the endoscopic features for early decision to evaluate SMA syndrome in children.

Methods
In case controlled observation study, we identi ed three endoscopic ndings in enrolled cases like as, nding I: a pulsating band like luminal narrowing of the third part of the duodenum without no expansion over one third during air insu ation for at least 15 seconds, nding II: marked dilatation of the duodenal second part during air insu ation at the duodenal third part, and nding III: bile lake in the stomach. SMA syndrome was con rmed with upper GI series in patients with endoscopic nding I or I plus more. We analyzed positive endoscopic ndings related with SMA syndrome.

Conclusions
Endoscopic examination down to the third part of duodenum can provide a clue making a decision to evaluate SMA syndrome, which includes a feature composed of three ndings like as a pulsating band like luminal narrowing of duodenal third part, a marked expansion of duodenal second part, and a bile lake in the stomach.

Background
Superior mesenteric artery (SMA) syndrome is a symptom complex condition caused by the external compression of the third part of the duodenum between the aorta and SMA [1]. It commonly manifests nonspeci c gastrointestinal (GI) symptoms such as upper abdominal pain, postprandial fullness, and chronic food intolerance with easy satiety, anorexia, nausea and vomiting.
Making the early diagnosis of SMA syndrome is challenging. Diagnostic delay commonly occurs due to its rarity, nonspeci c symptoms, and lack of high indices of suspicion [2,3]. Early decision to perform the optimal diagnostic tests is essential to avoid diagnostic delay of SMA syndrome. However, it is often diagnosed by incidentally identifying external compression of the duodenal third part during investigative process of the exclusion approach. The external duodenal compression can be demonstrated with upper GI series, contrast enhanced computed tomography (CECT) scan or magnetic resonance angiography (MRA). These radiologic studies can show in ation of proximal duodenum with or without gastric expansion and evidence of abrupt vertical or oblique compression of the duodenal third part by SMA [4,5]. Furthermore, CECT scan and MRA can make measurements of the angel and distance between superior mesenteric artery and aorta.
Most patients with SMA syndrome present with upper GI symptoms. So most may commonly undergo esophagogastroduodenoscopy (EGD) for initial investigative process. Till now, it has been known that EGD hardly affords an information to make a suspicion or diagnosis of SMA syndrome. We hypothesized that upper GI endoscopy can provide the evidences re ecting the external compression of the duodenal third part in patient with SMA syndrome. We aimed to search for endoscopic features related with external compression of the third part of the duodenum by SMA through the analysis of endoscopic ndings of patients who con rmed SMA syndrome.

Methods
Patient selection, study design, and data analysis We retrospectively or prospectively had collected data on patients who underwent EGD and upper GI series with or without CECT scan because of upper GI symptoms at the Department of Pediatrics, Chungnam National University Hospital since 2007. EGD performed by the single same pediatric gastroenterologist.
The recruitment in our study was limited to patients who had endoscopic ndings showing the following features: a pulsating vertical or oblique band like luminal narrowing with partial luminal opening less than one-third during air insu ation more than 15 seconds at the duodenal third part ( nding I), with or without an over expansion of duodenal second part during air insu ation at the duodenal third part ( nding II) with or without a large amount of bile mixed uid in the stomach ( nding III) (Fig. 1). We classi ed endoscopic ndings into 4 features as followings: feature A by nding I, feature B by nding I and II, feature C by nding I and III, and feature D by the constellation of nding I, II and III.
All enrolled cases underwent upper GI series with or without CECT scan to con rm SMA syndrome. A pediatric radiologist performed and interpreted the upper GI series. Positive ndings of upper GI series consistent with SMA syndrome included ndings of (i) dilatation of the rst and second parts of the duodenum, (ii) abrupt vertical or oblique compression of the duodenal third part, (iii) antiperistaltic to and pro waves of contrast from the proximal to the obstruction, and (iv) signi cant delays in gastroduodenal transit [4,5] (Fig. 2).
All enrolled cases were divided into 2 groups according to the results of upper GI series: group I included patients with SMA syndrome and group II included control patients without SMA syndrome. The medical records, including endoscopic ndings, demographics, clinical presentations, growth status with weight, weight-for-age Z score, body mass index (BMI), BMI-for-age-Z score, and weight-for-height Z score were analyzed. The Z scores were assessed according to the stature percentiles of the Korean National Growth Charts drafted by the Korea Centers for Disease Control and Prevention.

Statistical analysis
Statistical analysis on the data was performed by using SPSS version 19.0 (SPSS Inc. Chicago, IL, USA).
Demographics, duration of illness and growth status were assessed by the Mann-Whitney U test. Clinical manifestations and endoscopic ndings were assayed using the Fisher's exact test. A value of p < 0.05 was considered signi cant.

Results
The total number of patients available for this study was 29; group I included 18 patients with SMA syndrome and group II included 11 patients without SMA syndrome. The median ages of group I and II were 11.5 years (range, 7.8-16.2 years) and 13.2 years (range, 10.5-15.8 years) and male to female ratios were 0.8:1 in group I and 0.6:1 in group II, respectively. The median duration of symptoms before diagnosis was 68 days (range, 5-760 days) and 30 days (range, 5-1825 days) in each of the groups. The three most common presenting symptoms were postprandial discomfort (61.1% in group I and 54.5% in group II), abdominal pain (55.6% in group I and 90.9% in group II), and early satiety (50.0% in group I and 18.2% in group II). Less common symptoms included anorexia (33.3% in group I and 9.1% in group II), vomiting (27.8% in group I and 18.2% in group II), and nausea (16.7% in group I and 9.1% in group II). Weight loss was noted in 38.9% and 18.2% in group I and II, respectively. These presenting symptoms and signs were not statistically different between the groups (Table 1).  Table 2 shows growth status including the mean body weight, weight-for-age Z score, BMI, BMI-for-age Z score, and weight-for-height Z score. These parameters were not signi cantly different in both groups. The initial impression and nal diagnosis of 29 cases are shown in Table 3. The three most common initial impressions before EGD were functional dyspepsia (37.9%), gastritis or gastric ulcer (31%), and SMAS (24.1%). Other impressions were re ux esophagitis (10.3%), functional abdominal pain (6.9%), small bowel obstruction (3.4%), and Crohn's disease (3.4%). All cases underwent EGD and upper GI series with or without CECT scan. SMA syndrome was con rmed in 62.1% of the enrolled cases. Some patients had more than one disease at initial impression and diagnosis.
The results of EGD were as below: one case for feature A, three cases for feature B, nine cases for feature C, and sixteen cases for feature D. The cases of SMA syndrome con rmed by upper GI series according to each EGD feature were as follows: none (0%) of one case with feature A, one (33.3%) of three cases with feature B, four (44.4%) of nine cases with feature C, and thirteen (81.3%) of sixteen cases with feature D. Figure 3 shows the evaluation ow of the total enrolled cases and the results.
The most common endoscopic nding associated with SMA syndrome was feature D, which was documented in 72.2% in group I and 27.3% in group II, respectively (P = 0.027) ( Table 4). There were no statistically signi cant in feature A, B, and C between the groups.

Discussion
The aim of the present study was to evaluate whether the EGD provides clues on early decision to perform a diagnostic test for SMA syndrome. To the best of our knowledge, no study investigating on the endoscopic features associated with SMA syndrome has yet been conducted. Our study suggests that endoscopic observation up to the third part of the duodenum can give a signi cant information in determine to perform diagnostic tests of SMA syndrome. The endoscopic features in our study were the three constellations of pulsating vertical or oblique band like luminal narrowing of the third part of the duodenum with luminal expansion no more than one third during air insu ation over 15 seconds, proximal duodenal over distension during air insu ation in the duodenal third part, and bile lakes in the stomach.
SMA syndrome is a disease caused by external compression of the duodenal third part between the aorta itself and the SMA branching from the aorta at the level of the rst lumbar vertebra. Although the exact pathogenesis of this syndrome is unknown, it is often associated with rapid linear growth without or small weight gain, rapid weight loss, external cast compression, an abnormally high position of the ligament of Treitz, and unusually a low origin of the SMA [4][5][6][7][8]. Rapid weight loss and certain metabolic states cause depletion of the mesenteric and retroperitoneal fat, subsequently decreasing the aortomesenteric distance (AMD) (1). However, weight loss and low BMI do not necessarily relate with SMA syndrome. Indeed, no substantial difference were observed in the weight and BMI between the SMA syndrome and non-SMA syndrome groups in our study.
SMA syndrome often presents with non-speci c upper GI symptoms such as nausea, vomiting, early satiety, anorexia, postprandial discomfort, and abdominal pain [9,10]. In our study, postprandial discomfort, abdominal pain, and early satiety were the major symptoms. Symptoms can be acute or chronic. Making a diagnosis of SMA syndrome is challenge and diagnosis may commonly be delayed for a period of time after symptoms appear. The median duration of symptoms before diagnosis varies from 5 to 30 days and up to 18 months according to the literature [9,11,12]. In our study, the median was 68 days (range: 5 to 760 days) until diagnosis. The reasons for varying duration of diagnostic delays may be due to rarity, nonspeci c symptoms, and a lack of high suspicion of clinical index of SMA syndrome. Many clinicians remain unaware of this syndrome. So this syndrome may commonly be diagnosed in the process of excluding other suspected conditions [2,3]. And many patients may suffer from upper GI symptoms and related comorbidity such as weight loss, malnutrition, and poor quality of life until a diagnosis was made.
A con rm diagnosis of SMA syndrome is made mainly through the conventional upper GI series. CECT scan is also known to assist a diagnosis of SMA syndrome with ndings of proximal duodenal dilatation and decreased AMD and aortomesenteric angle (AMA). An AMA of less than 22-25° and an AMD of less than 8 mm correlated well with the development of symptoms of SMA syndrome in adults [13][14][15][16]. Sinagra et al. [17] reported that AMD seems to be more accurate rather than AMA to diagnose SMA syndrome. And these are consequently used as cutoff values for diagnosis of SMA syndrome in adults. However, there is no cutoff value for diagnosis in children and the normal ranges of these are variable in this age. So, interpretation should be performed with caution when using CECT scan for diagnosis of SMA syndrome in children [18,19]. Ultrasound color Doppler imaging or MRA has recently been proposed as alternative modalities rather than CECT scan for the diagnosis in cases of suspected SMA syndrome [20].
In the practice, SMA syndrome can be linked to inadequate investigation and ineffective treatment due to its rarity and unawareness by clinicians, which may lead to electrolyte imbalance, dehydration, weight loss, malnutrition, and even death [3,11]. So, early investigations and detection of SMA syndrome is important to avoid these problems. Most patients with SMA syndrome present nonspeci c upper GI symptoms. Thus EGD is usually rst performed to differentiate the upper GI diseases such as peptic diseases, functional dyspepsia, and other disease but, it is not possible to make a diagnosis of SMA syndrome with EGD. Lippl et al. [5] suggested that endoscopic ndings such as duodenal dilatation, liquid stasis, and antiperistaltic waves may indicate SMA syndrome. Meanwhile, Sundaram et al. [20] stated that although the Lippl's suggested endoscopic ndings may indicate duodenal obstruction, a diagnosis of SMA syndrome cannot always be made with certainty with those ndings. They recommended using both endoscopic ultrasound and endoscopy for the diagnosis of SMA syndrome. Unlike this, Sinagra et al. [17] suggested that pulsatile extrinsic compression in the third part of the duodenum by EGD is the most reliable nding to suspect SMA syndrome. Cappell et al. [21] reported a case with SMA syndrome with an endoscopic photograph showing pulsatile band-like luminal narrowing of the third part of duodenum which was gradually, partially opened by moderate air insu ation. We observed that diagnostic yield of SMA syndrome was signi cantly higher in patients with endoscopic feature D compared to those without that. The sensitivity and the positive predictive value having SMA syndrome in cases with endoscopic type D were 72.2% and 81.3%, respectively. So we suggest that EGD can be a useful in deciding to investigate SMA syndrome when endoscopic feature D is documented through examination up to the duodenal third part.

Conclusion
In conclusion, clinical symptomatology, physical examination, and endoscopic information are important for early suspicion of SMA syndrome. Clinicians should consider as differential diagnosis of SMA syndrome in patients with symptoms such as postprandial distress, abdominal pain, early satiety, anorexia with or without weight loss. Even though our study has some limitations due to the small number of enrolled patients and lack of endoscopic follow-up after improvement of symptoms, we recommend to examine up to the duodenal third part when thick bile stained uid retention (bile lake) is noted in the stomach during EGD in patient with upper GI symptoms like as postprandial discomfort, abdominal pain, and early satiety. We suggest endoscopic feature D would be a substantial clue to reach an early suspicion and make a decision for evaluation of SMA syndrome. Further clari cation with future large scale case studies are warranted. Availability of data and materials The datasets generated and analyzed during the present study are available from the corresponding author on reasonable request.

Competing interests
The authors declare that they have no competing interests.

Funding
Not applicable.
Authors' contributions JYK designed and performed most of the investigation, data analysis and wrote the most of manuscript; MSS and SHL contributed to write some part of discussion. All the authors have read and approved the manuscript for publication