Endoscopic management for congenital esophageal stenosis: A systematic review

Abstract

Congenital esophageal stenosis (CES) is an extremely rare malformation, and standard treatment have not been completely established. By years of clinical research, evidence has been accumulated. We conducted systematic review to assess outcomes of the treatment for CES, especially the role of endoscopic modalities. A total of 144 literatures were screened and reviewed. CES was categorized in fibromuscular thickening, tracheobronchial remnants (TBR) and membranous web, and the frequency was 54%, 30% and 16%, respectively. Therapeutic option includes surgery and dilatation, and surgery tends to be reserved for ineffective dilatation. An essential point is that dilatation for TBR type of CES has low success rate and high rate of perforation. TBR can be distinguished by using endoscopic ultrasonography (EUS). Overall success rate of dilatation for CES with or without case selection by using EUS was 90% and 29%, respectively. Overall rate of perforation with or without case selection was 7% and 24%, respectively. By case selection using EUS, high success rate with low rate of perforation could be achieved. In conclusion, endoscopic dilatation has been established as a primary therapy for CES except TBR type. Repetitive dilatation with gradual step-up might be one of safe ways to minimize the risk of perforation.

Key Words: Esophageal stenosis, Esophageal atresia, Tracheoesophageal fistula, Esophageal perforation, Dilatation, Endosonography, Deglutition disorders, Esophagoscopes, Esophageal ring, Plummer-Vinson syndrome

Core tip: Congenital esophageal stenosis (CES) is a rare malformation consisting of 3 types; fibromuscular thickening, tracheobronchial remnants (TBR) and membranous web. Endoscopic dilatation has been established as a primary therapy for CES except TBR type. Endoscopic ultrasonography is useful to distinguish TBR from other types of CES. Repetitive dilatation with gradual step-up is recommended to minimize the risk of perforation.

INTRODUCTION

Congenital esophageal stenosis (CES) is an extremely rare malformation, and diagnostic criteria and standard treatment have not been completely established. By years of clinical research, evidence for the management of CES has been accumulated. In the management of CES, surgery and endoscopic modalities play a key role. Endoscopic management could be an effective and less-invasive, however, the risk of therapies and therapeutic margin should be considered. The aims of this systematic review were to identify all published studies of endoscopic management of CES and to assess outcomes in terms of relief of the stricture and complication rates. Frequency and characters of 3 categories of CES, and relationship with associated anomalies were also reviewed.

RESEARCH

A Definition of CES was based on the description by Nihoul-Fékété[1]; “an intrinsic stenosis of the esophagus, present although not necessarily symptomatic at birth, which is caused by congenital malformation of esophageal wall architecture”.

Systematic review of English-language articles reporting CES was conducted by searching the PubMed database, in July 2014. Search terms “congenital” AND “esophageal stenosis” AND “endosc*”, and MeSH term“Esophageal Stenosis” AND the term “congenital” were used. The references of each of the included studies were then screened for any additionally relevant articles. Studies were selected according to the following inclusion/exclusion criteria: the only inclusion criteria was diagnosis of CES, defined as intrinsic stenosis of the esophagus. Esophageal stricture due to compression by cardiac/vascular malformations or intrathoracic tumor was excluded, if it is “congenital”. Secondary esophageal stenosis due to gastro-esophageal reflux, postoperative anastomotic stricture of esophageal atresia (EA) with/without tracheal fistula, leiomyoma and dermatological diseases including epidermolysis bullosa, dyskeratosis congenita, Rothmund Thomson syndrome and Goltz syndrome were also excluded. Review articles and mere letters were excluded. There were no exclusions based on patient numbers or length of follow-up. Accordingly, a total of 570 studies were identified by the initial searches, of which 144 studies satisfied the selection criteria (Figure 1). All the studies included were case reports or retrospective observational studies.

Figure 1 Flow chart of systematic search. CES: Congenital esophageal stenosis.

INCIDENCE

Investigators have commented on the rarity of CES, but the true incidence is still unknown. Bluestone et al[2] treated 24 cases of CES and approximately 200 cases of trachea-esophageal fistula in the single institution during the same 15 years, and estimated that the incidence of CES was one per 25000 births using that the incidence of tracheoesophageal fistula (TEF) was one per 2500 births[2]. Nihoul-Fékété et al[1] found 20 cases of CES and 484 cases of EA in the single institution during the same 25 years (1960-1984). According to this data, incidence of CES was lower than 1/20 of that of EA. Therefore, 1/25000-50000 live births is thought to be the incident rate of CES. These data are reliable and basically correct, but the frequency data should be revised based on the data at least in the 2000s.

CLASSIFICATION

The classification of CES has been confusing mainly because of its infrequency. Histological classification has been difficult because surgical specimens cannot be obtained if the only bougie can improves the symptom. Furthermore, it has also been difficult to differentiate CES from other non-congenital esophageal stricture such as achalasia, peptic esophageal stenosis due to gastroesophageal reflux and herpetic esophageal stenosis[3,4].

Various classification of CES had been proposed to date. Ohkawa et al[5] (1975) reported 5 entities of CES including tracheobronchial remnants, fibromuscular thickening, esophageal epithelioma, short esophagus and achalasia. Sneed et al[6] (1979) considered that there are congenital fibromuscular thickening (FMT), tracheobronchial remnants (TBR) and membranous web (MW) in the category of CES. Nihoul-Fékété (1989) clearly define CES and categorized the cases based on these 3 entities[1]. This categorization based on this sophisticated study has been broadly accepted to date. Ramesh et al[7] (2001) categorized CES into 3 groups; isolated segmental type, isolated diaphragm type and combined type. Isolated segmental type corresponds FMT and TBR, isolated diaphragm type corresponds MW and combined type corresponds segmental stenosis distal EA/TEF or MW. Although this classification involves the etiological consideration of CES, it is too complicated to use in clinical practice.

Frequency of 3 categories of CES were assessed by using the 3 observational studies including pediatric CES cases with detailed categorization (Table 1)[1,8,9]. Accordingly, overall frequency of FMT, TBR and MW were 53.8%, 29.9% and 16.2%, respectively. Locations of stenosis in each categories were assessed by using 52 case reports including 64 patients (Figure 2). Trends were as follows; MW mainly in the upper or middle third of the esophagus[10-27], FMT mainly in the middle or lower third[28-39], and TBR mostly in the lower third[6,40-60].

Table 1 Frequency of 3 categories of congenital esophageal stenosis.

Ref.	FMT	TBR	MW	Total
Nihoul-Fékétéet asl[1] (1987)	10 (50.0%)	4 (20.0%)	6 (30.0%)	20
Takamizawa et al[8] (2002)	13 (36.1%)	15 (41.7%)	8 (22.2%)1	36
Michaud et al[9] (2013)	40 (65.6%)	16 (26.2%)	5 (8.2%)	61
Total	63 (53.8%)	35 (29.9%)	19 (16.2%)	117

¹Including cases of multiple web. FMT: Fibromuscular thickening; TBR: Tracheobronchial remnants; MW: Membranous web.

Figure 2 The locations of stenosis in each categories of congenital esophageal stenosis.

Additionally, multiple web type of CES has been reported mainly in adults[61]. Only 1 pediatric case with multiple web has been reported[62].

ASSOCIATION WITH ESOPHAGEAL MALFORMATION

CES associated with esophageal atresia (EA) and/or tracheoesophageal fistula (TEF) is not so rare, and 44 cases have been reported as case(s) report to date[12,22,26,28,31,33,37,44,47,50,55,63-75]. To assess relationship and EA and/or TEF, 14 observational studies of pediatric cases were reviewed[1,2,8,9,76-85]. According to the 4 observational studies[76,80,81,84], overall incidence rate of CES among patients with EA and/or TEF was 9.6% (Table 2). All the CES located in the middle to lower third of the esophagus; 13.5% in middle third of the esophagus, and 86.5% in lower third of the esophagus. Pathological findings of CES associated with TEF were not clear, because not all the cases had surgical specimens. In 15 cases (27% of CES cases), pathological assessment was performed; 10 cases (67%) had tracheobronchial remnant and 5 cases (33%) had fibromuscular stenosis. CES in TEF/EA is not a rare association, therefore, careful attention is required during the management of TEF/EA, especially in postoperative esophagogram.

Table 2 Incidence rate of congenital esophageal stenosis among patients with esophageal atresia and/or tracheoesophageal fistula.

Ref.	Cases	Incidence rate	Location of CES
			Middle	Lower
Holinger et al[76] (1963)	4/79	5.1%	0 (0%)	4 (100%)
Newman et al[80] (1997)	18/225	8.0%	NA	NA
Kawahara et al[81] (2001)	11/80	13.8%	2 (18%)	9 (82%)
Yoo et al[84] (2010)	22/187	11.8%	3 (14%)	19 (86%)
Total	55/571	9.6%	5 (13.5%)	32 (86.5%)

CES: Congenital esophageal stenosis.

According to the 10 observational studies[1,2,8,9,77-79,82,83,85], overall incidence rate of EA and/or TEF among patients with CES was 24.8% (Table 3). Variation of the incident rate in each study may depend on study period, the role of institution and study design. Type of EA were not so different from original proportion; EA in 2.4%, EA+TEF in 92.7% and TEF in 4.9% of the cases. CES cases with complicated form of EA/TEF which cannot be classified were also reported[6,64].

Table 3 Incidence rate of esophageal atresia and/or tracheoesophageal fistula among patients with congenital esophageal stenosis.

Ref.	Cases	Incidence rate	EA	EA + TEF	TEF
Bluestone et al[2] (1969)	0/24	0.0%	0	0	0
Nishina et al[77] (1981)	4/81	4.9%	0	3	1
Dominguez et al[78] (1985)	5/34	14.7%	0	5	0
Nihoul-Fékété et al[1] (1987)	2/20	10.0%	0	1	1
Yeung et al[79] (1992)	6/8	75.0%	1	4	1
Vasudevan et al[82] (2002)	4/6	66.7%	1	2	1
Takamizawa et al[8] (2002)	13/36	36.1%	0	13	0
Amae et al[83] (2003)	4/14	28.6%	0	4	0
Romeo et al[85] (2011)	15/47	31.9%	0	15	0
Michaud et al[9] (2013)	29/61	47.5%	0	29	0
Total	82/331	24.8%	2 (2.4%)	76 (92.7%)	4 (4.9%)

EA: Esophageal atresia; TEF: Tracheoesophageal fistula.

Additionally, another esophageal malformation with CES, including esophageal duplication[22,50,86], diverticulum[18] and achalasia[11] were also reported.

ASSOCIATED ANOMALIES OTHER THAN ESOPHAGEAL MALFORMATION

Seven observational studies with detailed description about associated anomalies were reviewed[1,8,77-79,82,83]. These studies included a total of 199 cases of CES. The cases without any anomalies accounted for 55.3% of CES cases. Associated anomalies other than esophageal malformation were miscellaneous. Relatively frequent anomalies were as follows; congenital heart disease (4.5%), 21trisomy (4.0%), anorectal anomaly (2.0%), duodenal atresia (1.5%), tracheal malacia (1.5%), esophageal hiatal hernia (1.0%).

ADULT CASES

It is difficult to prove whether the adult cases with esophageal stenosis are truly “congenital”. Actually, webs of the cervical esophagus have been commonly associated with Plummer-Vinson syndrome. In the largest series of adult CES cases, 62% of cases with upper esophageal webs had anemia, and all of them were female[87]. Khosla et al[88] also reported that among 117 patients with iron deficiency anemia, 6 cases (5.1%) had upper esophageal webs. Meanwhile, esophageal stenosis may also be found without the Plummer-Vinson syndrome. We found 24 case reports including 30 adult cases of CES with the categorization[10,11,13,15-18,20,21,40,41,59,89-99]. In these, 26 cases (86.7%) had MW type of CES[10,11,13,15-18,20,21,89-97,98,99]. In these, 16 cases had multiple webs[89-99], which was similar to ring of the trachea. Younes et al[61] treated 10 adult cases of multiple esophageal webs during 7 years, and stated that CES in adults is under-recognized cause for intermittent, long-standing dysphagia. Although extremely rare, TBR[40,41,59] and FMT[34] type of CES were also reported in adults.

FAMILY INCIDENCE

Occurrence of CES within a family was reported only in the 2 literatures; in father and son[94], and sisters[96]. They all were over middle age, suffered from dysphagia and/or food impaction for long duration, and had multiple esophageal webs (one of the sisters had no detail). In the former family, the son had male sibling who died 1 wk after birth because of an inability to swallow. In earlier reports, the nature of multiple esophageal webs has been speculated to be either congenital or acquired[89], and still remains unclear.

DIAGNOSIS

In diagnosis of CES, it is essential to exclude postnatally acquired stenoses (peptic, caustic, infectious, neoplastic), extrinsic compression, and achalasia[1]. Careful medical interview is of key importance. Both esophagogram and esophagoscopy is required to know location, range, form and degree of stenosis. To exclude peptic stenosis, pH monitoring may be useful. To exclude achalasia, measure of esophageal pressure is also informative.

Endoscopic ultrasonography (EUS) is brilliant way to classify the CES, especially distinguishing TBR from FMT[8,54,100,101]. By using this modality, the cartilage in the esophageal wall is visualized as low echoic area[54,100] or high echoic area[8,101]. Whether CES is classified as TBR or not is important information to determine the therapeutic strategy, because CES of TBR should be managed by surgery, not bougie due to high rate of perforation[55].

TREATMENT

Therapeutic option consists of dilatation and surgery. Although surgery tends to be reserved for ineffective dilatation, efficacy and risk of dilatation has been controversial. We, therefore, reviewed the literatures in which more than 5 cases of CES were treated by dilatation[1,8,9,79,81-83,85]. Studies were divided into two groups by whether EUS was used for case selection or not. EUS was to distinguish TBR type of CES. Accordingly, overall success rate of dilatation for CES with or without case selection was 89.7% and 28.9%, respectively (Table 4). Overall rate of perforation with or without case selection was 7.4% and 23.9%, respectively (Table 5). By using EUS, high success rate with low rate of perforation could be achieved. On the basis of this knowledge, flow chart of treatment is shown in Figure 3.

Figure 3 Flow chart of diagnosis and treatment for congenital esophageal stenosis.

Table 4 Success rate of dilatation for congenital esophageal stenosis with/without case selection by endoscopic ultrasonography.

Ref.	Case selection by EUS		Modality
	+	-
	Success rate
Takamizawa et al[8] (2002)	16/21 (76.2%)	-	BD
Romeo et al[85] (2011)	45/47 (95.7%)	-	BD
Nihoul-Fékété et al[1] (1987)	-	7/14 (50.0%)	BD or TD
Yeung et al[79] (1992)	-	0/7 (0.0%)	BD or TD
Kawahara et al[81] (2001)	-	2/9 (22.2%)	BD
Vasudevan et al[82] (2002)	-	3/7 (42.9%)	TD
Amae et al[83] (2003)	-	3/11 (27.3%)	BD or TD
Michaud et al[9] (2013)	-	13/49 (26.5%)	BD or TD
Total	611/68 (89.7%)	28/97 (28.9%)

BD: Balloon dilatator; TD: Tapered dilator; EUS: Endoscopic ultrasonography.

Table 5 Rate of perforation during dilatation of congenital esophageal stenosis.

Ref.	Case selection by EUS		Modality
	+	-
	Rate of perforation
Takamizawa et al[8] (2002)	0/21 (0.0%)	-	BD
Romeo et al[85] (2011)	15/47 (10.6%)	-	BD
Nihoul-Fékété et al[1] (1987)	-	6/14 (42.9%)	BD or TD
Yeung et al[79] (1992)	-	1/7 (14.3%)	BD or TD
Newman et al[80] (1997)		3/18 (16.7%)	BD
Kawahara et al[81] (2001)	-	4/9 (44.4%)	BD
Amae et al[83] (2003)	-	1/11 (9.1%)	BD or TD
Fan et al[104] (2011)	-	1/8 (12.5%)	BD
Total	5/68 (7.4%)	16/67 (23.9%)

BD: Balloon dilatator; TD: Tapered dilator; EUS: Endoscopic ultrasonography.

As a technique of dilatation, there are tapered dilator and balloon dilatator, but there has been no comparison study of these. Some prefer balloon dilator because it enable expanding force to focus on the stenotic segment without shear stress, resulting in more effective and safer[8,102]. Appropriate diameter of dilatation for CES is still unknown. Kozarek et al[103] suggested that inflation of a single large-diameter dilator of less than 15 mm or an incremental dilation of more than 3 mm may be safe in simple esophageal strictures in adults. Fan et al[104] reported 9 procedures of balloon dilatation for CES including 1 esophageal perforation. Although there was no statistical significance, mean balloon diameter of the procedure with/without perforation was 12.1 mm and 15.0 mm, respectively. Mean dilation achieved with/without perforation was 5.4 mm and 8.4 mm, respectively. Not surprisingly, large dilatation with large increment might be a risk of perforation. Therefore, repetitive dilatation with gradual step-up might be one of safe ways to minimize the risk of perforation.

In cases of MW type of CES, efficacy of endoscopic dilatation with radial incision of the web has been reported. Instruments for incision include electrocoagulation[17,19,105], high-frequency-wave[27] and laser[23]. Nose et al[27] used balloon catheter for pulling up the web from the distal side during incision. Adverse events during dilatation with incision have not been reported.

LONG-TERM PROGNOSIS

It is well known that the association of Plummer-Vinson syndrome with carcinoma of the mouth, hypopharynx and upper esophagus. In the 58 adult cases of MW type of CES, 9 cases (15.5%) had carcinoma; buccal carcinoma in 6, esophageal carcinoma in 3[88]. Other than MW type, only one case has been reported, who had esophageal carcinoma associated with FMT type of CES; 65-year-old man who had suffered from dysphagia and vomiting since birth, but had not received any treatment because of mild symptom, underwent esophagectomy for worsening symptom. The resected specimen revealed squamous cell carcinoma in the region of fibromuscular stenosis[34]. The authors speculated that chronic mechanical stimulation by food trapped above the stenosis may have induced dysplasia of the mucosa. Special attention should be paid to status of the esophageal passage. Long-term functional prognosis after dilatation of pediatric CES has not been reported. Further studies are still needed.

CONCLUSION

Endoscopic dilatation has been established as a primary therapy for CES except TBR type. EUS is useful to distinguish TBR from other types of CES. Repetitive dilatation with gradual step-up is recommended to minimize the risk of perforation.