- Email: [email protected]
The radiological assessment of colonic replacement of the esophagus in children: A review of 43 cases
European Journal of Radiology 84 (2015) 2625–2632
Contents lists available at ScienceDirect
European Journal of Radiology journal homepage: www.elsevier.com/locate/ejrad
The radiological assessment of colonic replacement of the esophagus in children: A review of 43 cases Amr Abdelhamid AbouZeid a,∗ , Shaimaa Abdelsattar Mohammad b , Leila Mohamed Rawash b , Ahmed Bassiouny Radwan a , Khaled M. El-Asmar a , Ehab El-Shafei a a b
Faculty of medicine, Department of Pediatric Surgery, Ain Shams University, Abbasia, Cairo 11657, Egypt Faculty of medicine, Department of Radiodiagnosis, Ain-Shams University, Abbasia, Cairo 11657, Egypt
a r t i c l e
i n f o
Article history: Received 5 May 2015 Received in revised form 10 September 2015 Accepted 17 September 2015 Keywords: Esophagus Colon bypass Pediatrics Corrosive injuries Thoracic surgery Barium swallow
a b s t r a c t Purpose: To define the characteristic radiological features following colonic replacement of the esophagus in children. Materials and methods: The upper gastro-intestinal contrast studies of 43 patients who underwent colonic replacement of the esophagus at our pediatric surgery unit were available for analysis. UGI contrast studies were performed routinely in the post-surgical period in 17 cases (first asymptomatic group), while the rest of contrast studies (26) belonged to a second group of out-patients complaining of dysphagia (18) or dyspepsia (8) following colonic replacement of the esophagus. Based on our observations, we proposed a grading system to describe the degree of colonic redundancy in the thorax. Results: Redundancy of the colonic conduit in the thoracic cavity was a common radiological finding (62.8%). The redundancy was mild (grade 1) in 18 patients, moderate (grade 2) in eight, and severe (grade 3) in only one patient. In 88.9%, the redundancy was in the right hemi-thorax. Patients presenting with postoperative dysphagia had a stricture at the site of the esophago-colic anastomosis in the neck, which should be differentiated from other sites of anatomical narrowing at the inlet and outlet of the thoracic cavity. Gastro-colic reflux was common among patients who underwent colonic replacement of the esophagus without an anti-reflux procedure. Conclusion: Colonic replacement of the esophagus in children results in considerable anatomical alterations. Knowledge about the normal post-surgical changes and imaging features of the commonly encountered complications can increase the diagnostic confidence among radiologists and clinicians when dealing with these cases. © 2015 Elsevier Ireland Ltd. All rights reserved.
1. Introduction The esophagus is the best conduit for food and fluids to the stomach, and every effort should be exhausted to keep children with their native esophagus. However, there are certain circumstances when esophageal replacement in children becomes unavoidable: long gap esophageal atresia, and resistant corrosive strictures [1]. In 1911, the colon was first introduced as a substitute for the esoph-
∗ Corresponding author. E-mail addresses: [email protected] (A.A. AbouZeid), [email protected] (S.A. Mohammad), [email protected] (L.M. Rawash), ahmed b [email protected] (A.B. Radwan), [email protected] (K.M. El-Asmar), ia [email protected] (E. El-Shafei). http://dx.doi.org/10.1016/j.ejrad.2015.09.014 0720-048X/© 2015 Elsevier Ireland Ltd. All rights reserved.
agus [2,3]; and in our pediatric surgery unit, the colon has been used to replace the esophagus since 1972 [4,5]. The procedure has been subjected to a continuous process of technical evolution starting by the historical subcutaneous placement of the colonic conduit [6], trans-pleural retro-hilar route [7], posterior mediastinal ‘transhiatal’ route [8], and lastly the anterior retro-sternal colon bypass (which is considered our preferred technique at the present time) [9]. There are two common complaints (complications) following colonic replacement of the esophagus: the first is usually related to mechanical obstruction ‘stricture’ at the site of esophago-colic anastomosis in the neck causing dysphagia; and the second is a functional problem related to the mechanism of gastric emptying and regurgitation (reflux) of gastric contents into the colonic conduit [10,11]. An upper gastrointestinal (UGI) contrast study is the
2626
A.A. AbouZeid et al. / European Journal of Radiology 84 (2015) 2625–2632
Fig. 1. Barium swallow of children with post-corrosive esophageal strictures for assessment before esophageal replacement. (a) Stricture in the mid-esophagus with a patent proximal esophagus suitable for the usual esophago-colic anastomosis in the neck. (b) A higher stricture in the proximal esophagus that would necessitate a higher anastomosis (pharyngeo-colic).
first-line investigation; however, other imaging modalities (CT and MRI) may be of value [12]. The non-anatomical placement of the esophageal substitute can represent a source of confusion among radiologists and clinicians. The aim of this study is to define the characteristic radiological features of the commonly encountered complications as well as the normal appearance following colonic replacement of the esophagus in children. This may help to improve the diagnostic confidence when dealing with these cases. 2. Materials and methods Between January 2013 and April 2015, 23 children underwent colonic replacement of the esophagus at our pediatric surgery unit. The indication for operation was either corrosive strictures or long gap esophageal atresia. UGI contrast studies were performed in some of these patients routinely in the post-surgical period (first group). Also, UGI contrast studies were performed in another group referred from our outpatient clinic, who presented with dysphagia or dyspepsia following previous colonic replacement of the esophagus (second group). The UGI contrast and other radiological studies of both groups were collected, and retrospectively analyzed. Pediatric patients who undergo esophageal replacement require radiological assessment throughout the different stages of the procedure: pre-operative, post-operative, and in the follow up. 2.1. Pre-operative imaging For a case with long-gap esophageal atresia, an echo-cardiogram and renal ultrasound are performed to screen for possible associating anomalies. The patient undergoes esophagostomy and feeding
gastrostomy at birth, to be scheduled later for esophageal replacement (usually at 9–12 months of age). For patients with post-corrosive esophageal strictures, the preoperative esophagogram is important to examine for the patency and degree of scarring of the proximal esophagus (Fig. 1). Usually, the colonic conduit is anastomosed with the proximal esophagus in the neck (esophago-colic anastomosis) to bypass a mid or lower esophageal strictures. A scared proximal esophagus should be an indication to shift for a higher-level anastomosis (pharyngeo-colic); otherwise, dysphagia will persist after the esophageal replacement. Similarly, it is important to check for gastric emptying by a barium meal (with or without a follow-through study), especially with a history of acid ingestion that is known to cause antral scaring as well. 2.2. Early post-operative imaging Plain chest X-ray immediately after the colon bypass procedure is essential to check for complete lung inflation (Fig. 2), and for the position of the intravenous central line (if present). At the eighth to tenth post-operative day, a water-soluble esophagogram may be requested to check for any possible leakage at the esophago-colic anastomosis before starting oral intake. 2.3. Imaging in the follow-up An UGI contrast study is the investigation of choice for evaluating both the anatomical and functional results following an esophageal replacement procedure. This includes either a barium swallow, meal, or follow-through, according to the patient’s complaint. For uncooperative younger children, it may be necessary
A.A. AbouZeid et al. / European Journal of Radiology 84 (2015) 2625–2632
2627
Fig. 2. After esophageal replacement, early post-operative chest X-ray is mandatory to check for complete lung inflation. (a) 6-year-old boy showing post-operative complete lung inflation. (b) 9-month-old female showing post-operative bilateral lung collapse; arrows are pointing to the pleural line.
Fig. 3. Upper gastro-intestinal contrast studies of a group of children following colonic replacement of the esophagus, and showing the different degrees of colonic redundancy. (a) grade 0–straight colon; (b) grade 1–colonic conduit extending into the medial two thirds of one hemi-thorax; (c) grade 2–colonic conduit extending into the lateral third of one hemi-thorax; (d) grade 3–severe redundancy of the colonic conduit in both thoracic cavities; (e) grade 1 redundancy in the left hemi-thorax; (f) redundancy of the colonic conduit in both thoracic and abdominal cavities.
to inject the contrast material through a Ryles tube inserted into the upper cervical esophagus (Fig. 3a). The following are important points to be looked for in the contrast study: (1) the condition of the esophago-colic anastomosis in the neck (the usual cause of dysphagia); (2) the position, and degree of redundancy of the colonic conduit; and (3) abnormalities in gastric emptying (delay, or regurgitation into the colonic conduit). Based on our observations, we proposed a grading system to describe the degree of colonic redundancy in the thorax (Fig. 3): grade 0–straight colon (the colonic conduit is contained within the medial one third of each hemi-thorax); grade 1–colonic conduit extending into the medial two thirds of one hemi-thorax; grade
2–colonic conduit extending into the lateral third of one hemithorax; and grade 3–severe redundancy of the colonic conduit in both thoracic cavities. Contrast stagnation in the redundant colonic conduit (Fig. 4) hinders proper assessment of the gastro-colic regurgitation (reflux). Therefore, in patients with a feeding gastrostomy (which is usually kept for the first 2–3 months following the esophageal replacement), barium is injected through the gastrostomy tube (Fig. 5), to be followed by the swallow. This offers higher sensitivity for detection of gastro-colic reflux.
2628
A.A. AbouZeid et al. / European Journal of Radiology 84 (2015) 2625–2632
Fig. 4. Six-year-old girl with post-corrosive esophageal stricture underwent colonic replacement of the esophagus. Upper gastro-intestinal contrast study showing stagnation of the contrast in both the stomach and the colonic conduit in delayed films.
3. Results The UGI contrast studies of 43 patients who underwent colonic replacement of the esophagus were included in the study. Their age ranged from 7 months to 17 years. Most cases (41) underwent retro-sternal colon bypass, while only two had a posterior trans-hiatal colonic replacement of the esophagus (Fig. 6). UGI contrast studies were performed routinely following the procedure in 17 cases (first asymptomatic group), while the rest of contrast studies (26) belonged to the second group of outpatients complaining of dysphagia (18) or dyspepsia (8) following previous colonic replacement of the esophagus. One patient underwent CT chest that was ordered by her pediatrician who misinterpreted the colonic conduit in the chest X-ray for a diaphragmatic hernia. A retention mucous cyst was an incidental finding in the study at the site of the retained scared esophagus (Fig. 7). Redundancy of the colonic conduit in the thoracic cavity was a common radiological finding (27 patients; 62.8%). The redundancy was mild (grade 1) in 18 patients, moderate (grade 2) in eight, and severe (grade 3) in only one patient. Higher grades of redundancy were more commonly seen in patients presenting with dyspeptic symptoms (vomiting; aspiration especially on bending and at night) (Table 1). In 88.9% (24 patients), the redundancy was in the right hemi-thorax. On the contrarily, redundancy of the colonic conduit in the abdomen was not a common finding (only 4 cases) (Fig. 3f). We examined the UGI contrast studies of 18 patients who presented with dysphagia following colonic replacement of the esophagus; all of them gave a history of post-corrosive esophageal
stricture as an indication for esophageal replacement. A persistent narrowing could always be depicted in their esophagrams at the site of the esophago-colic anastomosis in the neck (Fig. 8). Finding a stricture at the site of anastomosis provided a clue to the cause of dysphagia, which was managed either by endoscopic dilatation, or by surgical revision of the anastomosis. This should be differentiated from other sites of anatomical narrowing, which are usually seen along the course of the retro-sternal colonic conduit in the contrast studies of asymptomatic cases as well: the first at the thoracic inlet, behind the upper border of the manubrium, and the second at the site of exit of the colonic conduit through the diaphragm (Figs. 6 a,b and 8 c,f). The gastro-colic reflux was difficult to be assessed retrospectively in the contrast studies that were performed the conventional way (baruim swallow, then meal), due to contrast stagnation in the redundant colonic conduit (Fig. 4). To overcome this problem, we modified our technique by performing the barium meal first through the gastrostomy (14 patients), to be followed by the swallow. Reflux of the contrast from the stomach into the colonic conduit was evident in five cases, while it was absent in nine cases that had an anti-reflux mechanism added to the procedure (Fig. 5). In two patients who presented with severe vomiting and hematemesis following a previous retro-sternal colonic bypass procedure, their gastrograms showed gastro-colic reflux with marked distention and loss of haustrations of the colonic conduit (similar to the radiological picture of a toxic megacolon, Fig. 5b). These patients were successfully managed by a re-operation to add an anti-reflux mechanism to the procedure.
A.A. AbouZeid et al. / European Journal of Radiology 84 (2015) 2625–2632
2629
Fig. 5. Barium-meal-first through the gastrostomy tube (gastrograms) to test for the presence of gastro-colic reflux in a group of children following colonic replacement of the esophagus. (a) and (b) showing reflux of contrast into the colonic conduit in the thorax. Note the marked distention of the colonic conduit in (b) with loss of colonic haustrations. (d) and (c) Gastrograms showing absence of gastro-colic reflux in those who had an anti-reflux mechanism added to the esophageal replacement procedure.
Table 1 Correlation between the post-operative symptoms and the degree of redundancy of the colonic conduit used to replace the esophagus in children. Degree of colonic redundancy
Asymptomatic group(17 cases)
Dysphagia group(18 cases)
Dyspepsia group(8 cases)
Grade 0 (no redundancy) Grade 1 (mild redundancy) Grade 2 (moderate redundancy) Grade 3 (severe redundancy)
10 cases (58.8%) 5 cases (29.4%) 2 cases (11.8%) –
5 cases (27.8%) 10 cases (55.6%) 3 cases (16.6%) –
1 case (12.5%) 3 cases (37.5%) 3 cases (37.5%) 1 case (12.5%)
4. Discussion Esophageal replacement is one of the technically challenging procedures in pediatric surgery. Different organs can be used to replace the esophagus: the stomach, the jejunum, or the colon [13]. For many years, the colon has been the organ of choice for replacing the esophagus in children, for its durability and good functional
outcome [14,15]. In the last two decades, the colon has lost some of its popularity against the gastric pull-up procedure that became more predominant [11,12,13,15][11–13,15]. However, the colon has remained the preferred organ for replacing the esophagus at our institution, with satisfactory outcome and acceptable morbidity. It has been shown that choosing between the different techniques tends to be based on personal preference and local experience
2630
A.A. AbouZeid et al. / European Journal of Radiology 84 (2015) 2625–2632
Fig. 6. Upper gastro-intestinal contrast study (lateral view) of three cases following colonic replacement of the esophagus. (a) and (b) The retro-sternal position of the colonic conduit showing ‘anatomical’ narrowing at two points (white arrows): the first behind the upper border of the manubrium at the thoracic inlet, and the second at the site of exit of the colonic conduit through the diaphragm. (c) Posterior mediastinal position of the colonic conduit.
Fig. 7. Three-year-old girl underwent colonic replacement of the esophagus for corrosive stricture. A CT scan was ordered by her pediatrician: (a) mediastinal window, (b) lung window showing a mucous cyst at the site of the retained scared esophagus (*). Five years later, a follow-up MRI (c) and (d) was performed showing reduction in the size of the cyst (*). (c) Axial T1-weighted spin echo showing the cyst exhibiting hypo-intense signal. (d) Axial T2-weighted fast spin-echo showing hyper-intense signal of the cyst. (Cd.: colonic conduit)
rather than on controlled studies and objective data [16]. There are many reports discussing the variable operative techniques and the different routes used for the colonic replacement of the esophagus [1,6,11,15,16]. However, to our knowledge, there is paucity of data describing the post-surgical imaging findings [12,17]. In this report, we presented the expected normal radiological findings following the retro-sternal colon bypass procedure (which is the commonly practiced technique at our institute), as well as the radiological features of the commonly encountered post-surgical complications.
Redundancy of the colonic graft has been a major concern for thoracic surgeons who perform esophageal replacement procedures [1,11,18,19]. In our series, redundancy of the colonic conduit in the thorax was present in 62% of the studied cases. However, redundancy of the colonic conduit is not always symptomatic. That is why we proposed a new grading scale to differentiate between different degrees of redundancy. Mild degrees (grade 1) were the most common (66%), and were usually asymptomatic. More severe degrees are more liable to produce symptoms (delay in transit of food; regurgitation; dysnea with exertion). Several theories about
A.A. AbouZeid et al. / European Journal of Radiology 84 (2015) 2625–2632
2631
Fig. 8. Barium swallow of a group of children who underwent colonic replacement of the esophagus. (a)–(c) belong to three patients complaining of dysphagia, demonstrating strictures at the site of the esophago-colic anastomosis (white arrows). Note in (b) for the diffuse scaring of the proximal esophagus and pharynx due to corrosive injury. (d)–(f) belong to another three patients without dysphagia. Note the site of anatomical narrowing (*) in the retro-sternal colonic conduit at the thoracic inlet behind the manubrium in (c) and (f).
the cause of redundancy have been proposed [20]. Surgeons are advised to avoid using too long grafts to bridge the gap, and to excise any excess length before completing the neck anastomosis [21]. In our series, we have noticed that the colonic redundancy was usually in the right hemi-thorax (89%), which may be explained by the position of the heart and pericardium normally toward the left. Also, we believe that major regurgitation of gastric contents into the colonic conduit would further aggravate existing redundancy through distension (Fig. 5b). Dysphagia was the most common complaint after colonic replacement of the esophagus. This was especially evident among patients with a history of post-corrosive esophageal stricture, probably due to the fibrosis and compromised blood supply of their proximal esophagus. The cause of dysphagia was usually a stricture
at the site of the esophago-colic anastomosis in the neck. Occasionally, the whole proximal esophagus was scared, which would emphasize on the importance of proper pre-operative assessment of the proximal esophagus (either radiologically or endoscopically). A scared proximal esophagus may indicate the need to perform the bypass at a higher level (with the pharynx). In colonic replacement of the esophagus for corrosive injuries, there is a debate about the necessity of excising the scared esophagus [1,4,5,22]. A colonic bypass procedure via a retro-sternal route saves a technically challenging step (esophageal resection) and considerably simplifies the procedure with comparable functional results. However, there is a concern about possible late complications in the retained scared esophagus (malignant transformation and mucous cyst formation) [4,5,22]. Usually the residual scared
2632
A.A. AbouZeid et al. / European Journal of Radiology 84 (2015) 2625–2632
esophagus is a harmless fibrotic tube [23], with only very few reports on patients developing carcinoma [22]. In our series, we had an incidental finding of a mucous cyst in the retained esophagus in one patient. This cyst was small in size (2.5 × 2.4 cm), and got even smaller on follow-up. The last complication is concerning the gastric emptying and regurgitation into the colonic conduit. Early reports highlighted the advantage of utilizing the colon in replacing the esophagus for having a ‘relatively’ resistant mucosal lining to peptic ulceration [4,13]; and there was some reluctance for adding a routine anti-reflux mechanism to the procedure [9]. However, reflux problems are not only restricted to peptic ulceration. We have seen fatal respiratory complications especially in patients with long gap esophageal atresia who underwent the colon bypass procedure early in infancy. Also, in some patients with major gastro-colic reflux, we have noticed severe interference with the normal gastric emptying. The stomach is emptying a majority of contents into the colonic conduit in the chest, which then goes back again down to the stomach (yo–yo movement). This abnormal ‘yo–yo movement’ of the gastric contents will be associated with little amounts passing each time through the normal path into the duodenum and small bowel. With our increased awareness of the possible adverse effects of the gastro-colic regurgitation, it is now our routine practice to add an anti-reflux mechanism to the retro-sternal colon bypass procedure. This study is limited by its retrospective nature. Also, it mainly included a single technique for esophageal replacement ‘the colon bypass procedure’. However, through collecting a relatively large number of contrast studies that belong to both symptomatic and asymptomatic cases, we could provide clear illustration of the typical radiological features of the commonly encountered complications as well as the normal appearance following colonic replacement of the esophagus in children. 5. Conclusion Colonic replacement of the esophagus in children results in considerable anatomical alterations. Knowledge about the normal post-surgical changes and imaging features of the commonly encountered complications can increase the diagnostic confidence among radiologists and clinicians when dealing with these cases. Conflict of interest None. References [1] S.A. Ahmad, K.G. Sylvester, A. Hebra, A.M. Davidoff, S. McClane, P.W. Stafford, et al., Esophageal replacement using the colon: is it a good choice? J. Pediat. Surg. 31 (1996) 1026–1031.
[2] G. Kelling, Oesophagoplastik mit hilfe des querkolon, Zbl. Chir. 38 (1911) 1209–1212. [3] H. Vulliet, De l’oesophagoplastie et des diverses modifications, Sem. Med. 31 (1911) 529–534. [4] A.F. Bahnassy, I.E. Bassiouny, Esophagocoloplasty for caustic stricture of the esophagus: changing concepts, Pediatr. Surg. Int. 3 (1993) 103–108. [5] A.F. Hamza, S. Abdelhay, H. Sherif, T. Hasan, H. Soliman, A. Kabesh, et al., Caustic esophageal strictures in children: 30 years’ experience, J. Pediatr. Surg. 38 (2003) 828–833. [6] R. Belsey, Reconstruction of the esophagus, Ann. R. Coll. Surg. Engl. 65 (1983) 360–364. [7] C.D. Sherman, D. Waterston, Esophageal reconstruction in children using intrathoracic colon, Arch. Dis. Child. 32 (1957) 11–16. [8] N.V. Freeman, D.T. Cass, Colon interposition: a modification of the Waterston technique using the normal esophageal route, J. Pediatr. Surg. 17 (1982) 17–21. [9] W.A. Dale, C.D. Sherman, Late reconstruction of congenital esophageal atresia by intrathoracic colon transplantation, J. Thorac. Surg. 29 (1955) 344–356. [10] T.G. Canty, B.E. LoSasso, One-stage esophagectomy and in situ colon interposition for esophageal replacement in children, J. Pediatr. Surg. 32 (1997) 334–337. [11] T.W. Rice, Right colon interposition for esophageal replacement, Oper. Tech. Thorac. Cardiovasc. Surg. 4 (1999) 210–221. [12] S. Ho, P. Patel, K. Mchugh. Imaging appearance following Paediatric oesophageal replacement surgery: an analysis of 72 cases at a tertiary referral centre. http://dx.doi.org/10.1594/ecr2012/C-2418. [13] J.L. Marks, W.L. Hofstetter, Esophageal reconstruction with alternative conduits, Surg. Clin. N. Am. 92 (2012) 1287–1297. [14] T.R. DeMeester, K.E. Johansson, I. Franze, E. Eypasch, C.T. Lu, J.E. McGill, et al., Indications, surgical technique, and long-term functional results of colon interposition or bypass, Ann. Surg. 208 (1988) 460–474. [15] J.C. Wain, C.D. Wright, E.Y. Kuo, A.C. Moncure, E.W. Wilkins, H.C. Grillo, et al., Long-segment colon interposition for acquired esophageal disease, Ann. Thorac. Surg. 67 (1999) 313–317. [16] G.S. Arul, D. Parikh, Oesophageal replacement in children, Ann. R. Coll. Surg. Engl. 90 (2008) 7–12. [17] Am Sieber, W.K. Seiber, Colon transplants as esophageal replacement: cieradiographic and manometric evaluation in children, Ann. Surg. 168 (1968) 116–122. [18] T. Yasuda, H. Shiozaki, Esophageal reconstruction with colon tissue, Surg. Today 41 (2011) 745–753. [19] H. Elshafei, E. Elshafei, M. ElDebeiky, N. Hegazy, A. Zaki, S.A. Hay, Colonic conduit for esophageal replacement: long-term endoscopic and histopathologic changes in colonic mucosa, J. Pediatr. Surg. 47 (2012) 1658–1661. [20] D.C. Strauss, M.J. Forshaw, R.C. Tandon, R.C. Mason, Surgical management of colonic redundancy following esophageal replacement, Dis. Esophagus 21 (2008) E1–E5. [21] A.F. Hamza, Colonic replacement in cases of esophageal atresia, Sem. Pediatr. Surg. 18 (2009) 40–43. [22] L. Burgos, S. Barrena, A.M. Andres, L. Martinez, F. Hernandez, P. Olivares, et al., Colonic interposition for esophageal replacement in children remains a good choice: 33-year median follow-up of 65 patients, J. Pediatr. Surg. 45 (2010) 341–345. [23] J.G. Raffensperger, S.R. Luck, M. Reynolds, D. Schwarz, Intestinal bypass of the esophagus, J. Pediatr. Surg. 31 (1996) 38–47.
Contents lists available at ScienceDirect
European Journal of Radiology journal homepage: www.elsevier.com/locate/ejrad
The radiological assessment of colonic replacement of the esophagus in children: A review of 43 cases Amr Abdelhamid AbouZeid a,∗ , Shaimaa Abdelsattar Mohammad b , Leila Mohamed Rawash b , Ahmed Bassiouny Radwan a , Khaled M. El-Asmar a , Ehab El-Shafei a a b
Faculty of medicine, Department of Pediatric Surgery, Ain Shams University, Abbasia, Cairo 11657, Egypt Faculty of medicine, Department of Radiodiagnosis, Ain-Shams University, Abbasia, Cairo 11657, Egypt
a r t i c l e
i n f o
Article history: Received 5 May 2015 Received in revised form 10 September 2015 Accepted 17 September 2015 Keywords: Esophagus Colon bypass Pediatrics Corrosive injuries Thoracic surgery Barium swallow
a b s t r a c t Purpose: To define the characteristic radiological features following colonic replacement of the esophagus in children. Materials and methods: The upper gastro-intestinal contrast studies of 43 patients who underwent colonic replacement of the esophagus at our pediatric surgery unit were available for analysis. UGI contrast studies were performed routinely in the post-surgical period in 17 cases (first asymptomatic group), while the rest of contrast studies (26) belonged to a second group of out-patients complaining of dysphagia (18) or dyspepsia (8) following colonic replacement of the esophagus. Based on our observations, we proposed a grading system to describe the degree of colonic redundancy in the thorax. Results: Redundancy of the colonic conduit in the thoracic cavity was a common radiological finding (62.8%). The redundancy was mild (grade 1) in 18 patients, moderate (grade 2) in eight, and severe (grade 3) in only one patient. In 88.9%, the redundancy was in the right hemi-thorax. Patients presenting with postoperative dysphagia had a stricture at the site of the esophago-colic anastomosis in the neck, which should be differentiated from other sites of anatomical narrowing at the inlet and outlet of the thoracic cavity. Gastro-colic reflux was common among patients who underwent colonic replacement of the esophagus without an anti-reflux procedure. Conclusion: Colonic replacement of the esophagus in children results in considerable anatomical alterations. Knowledge about the normal post-surgical changes and imaging features of the commonly encountered complications can increase the diagnostic confidence among radiologists and clinicians when dealing with these cases. © 2015 Elsevier Ireland Ltd. All rights reserved.
1. Introduction The esophagus is the best conduit for food and fluids to the stomach, and every effort should be exhausted to keep children with their native esophagus. However, there are certain circumstances when esophageal replacement in children becomes unavoidable: long gap esophageal atresia, and resistant corrosive strictures [1]. In 1911, the colon was first introduced as a substitute for the esoph-
∗ Corresponding author. E-mail addresses: [email protected] (A.A. AbouZeid), [email protected] (S.A. Mohammad), [email protected] (L.M. Rawash), ahmed b [email protected] (A.B. Radwan), [email protected] (K.M. El-Asmar), ia [email protected] (E. El-Shafei). http://dx.doi.org/10.1016/j.ejrad.2015.09.014 0720-048X/© 2015 Elsevier Ireland Ltd. All rights reserved.
agus [2,3]; and in our pediatric surgery unit, the colon has been used to replace the esophagus since 1972 [4,5]. The procedure has been subjected to a continuous process of technical evolution starting by the historical subcutaneous placement of the colonic conduit [6], trans-pleural retro-hilar route [7], posterior mediastinal ‘transhiatal’ route [8], and lastly the anterior retro-sternal colon bypass (which is considered our preferred technique at the present time) [9]. There are two common complaints (complications) following colonic replacement of the esophagus: the first is usually related to mechanical obstruction ‘stricture’ at the site of esophago-colic anastomosis in the neck causing dysphagia; and the second is a functional problem related to the mechanism of gastric emptying and regurgitation (reflux) of gastric contents into the colonic conduit [10,11]. An upper gastrointestinal (UGI) contrast study is the
2626
A.A. AbouZeid et al. / European Journal of Radiology 84 (2015) 2625–2632
Fig. 1. Barium swallow of children with post-corrosive esophageal strictures for assessment before esophageal replacement. (a) Stricture in the mid-esophagus with a patent proximal esophagus suitable for the usual esophago-colic anastomosis in the neck. (b) A higher stricture in the proximal esophagus that would necessitate a higher anastomosis (pharyngeo-colic).
first-line investigation; however, other imaging modalities (CT and MRI) may be of value [12]. The non-anatomical placement of the esophageal substitute can represent a source of confusion among radiologists and clinicians. The aim of this study is to define the characteristic radiological features of the commonly encountered complications as well as the normal appearance following colonic replacement of the esophagus in children. This may help to improve the diagnostic confidence when dealing with these cases. 2. Materials and methods Between January 2013 and April 2015, 23 children underwent colonic replacement of the esophagus at our pediatric surgery unit. The indication for operation was either corrosive strictures or long gap esophageal atresia. UGI contrast studies were performed in some of these patients routinely in the post-surgical period (first group). Also, UGI contrast studies were performed in another group referred from our outpatient clinic, who presented with dysphagia or dyspepsia following previous colonic replacement of the esophagus (second group). The UGI contrast and other radiological studies of both groups were collected, and retrospectively analyzed. Pediatric patients who undergo esophageal replacement require radiological assessment throughout the different stages of the procedure: pre-operative, post-operative, and in the follow up. 2.1. Pre-operative imaging For a case with long-gap esophageal atresia, an echo-cardiogram and renal ultrasound are performed to screen for possible associating anomalies. The patient undergoes esophagostomy and feeding
gastrostomy at birth, to be scheduled later for esophageal replacement (usually at 9–12 months of age). For patients with post-corrosive esophageal strictures, the preoperative esophagogram is important to examine for the patency and degree of scarring of the proximal esophagus (Fig. 1). Usually, the colonic conduit is anastomosed with the proximal esophagus in the neck (esophago-colic anastomosis) to bypass a mid or lower esophageal strictures. A scared proximal esophagus should be an indication to shift for a higher-level anastomosis (pharyngeo-colic); otherwise, dysphagia will persist after the esophageal replacement. Similarly, it is important to check for gastric emptying by a barium meal (with or without a follow-through study), especially with a history of acid ingestion that is known to cause antral scaring as well. 2.2. Early post-operative imaging Plain chest X-ray immediately after the colon bypass procedure is essential to check for complete lung inflation (Fig. 2), and for the position of the intravenous central line (if present). At the eighth to tenth post-operative day, a water-soluble esophagogram may be requested to check for any possible leakage at the esophago-colic anastomosis before starting oral intake. 2.3. Imaging in the follow-up An UGI contrast study is the investigation of choice for evaluating both the anatomical and functional results following an esophageal replacement procedure. This includes either a barium swallow, meal, or follow-through, according to the patient’s complaint. For uncooperative younger children, it may be necessary
A.A. AbouZeid et al. / European Journal of Radiology 84 (2015) 2625–2632
2627
Fig. 2. After esophageal replacement, early post-operative chest X-ray is mandatory to check for complete lung inflation. (a) 6-year-old boy showing post-operative complete lung inflation. (b) 9-month-old female showing post-operative bilateral lung collapse; arrows are pointing to the pleural line.
Fig. 3. Upper gastro-intestinal contrast studies of a group of children following colonic replacement of the esophagus, and showing the different degrees of colonic redundancy. (a) grade 0–straight colon; (b) grade 1–colonic conduit extending into the medial two thirds of one hemi-thorax; (c) grade 2–colonic conduit extending into the lateral third of one hemi-thorax; (d) grade 3–severe redundancy of the colonic conduit in both thoracic cavities; (e) grade 1 redundancy in the left hemi-thorax; (f) redundancy of the colonic conduit in both thoracic and abdominal cavities.
to inject the contrast material through a Ryles tube inserted into the upper cervical esophagus (Fig. 3a). The following are important points to be looked for in the contrast study: (1) the condition of the esophago-colic anastomosis in the neck (the usual cause of dysphagia); (2) the position, and degree of redundancy of the colonic conduit; and (3) abnormalities in gastric emptying (delay, or regurgitation into the colonic conduit). Based on our observations, we proposed a grading system to describe the degree of colonic redundancy in the thorax (Fig. 3): grade 0–straight colon (the colonic conduit is contained within the medial one third of each hemi-thorax); grade 1–colonic conduit extending into the medial two thirds of one hemi-thorax; grade
2–colonic conduit extending into the lateral third of one hemithorax; and grade 3–severe redundancy of the colonic conduit in both thoracic cavities. Contrast stagnation in the redundant colonic conduit (Fig. 4) hinders proper assessment of the gastro-colic regurgitation (reflux). Therefore, in patients with a feeding gastrostomy (which is usually kept for the first 2–3 months following the esophageal replacement), barium is injected through the gastrostomy tube (Fig. 5), to be followed by the swallow. This offers higher sensitivity for detection of gastro-colic reflux.
2628
A.A. AbouZeid et al. / European Journal of Radiology 84 (2015) 2625–2632
Fig. 4. Six-year-old girl with post-corrosive esophageal stricture underwent colonic replacement of the esophagus. Upper gastro-intestinal contrast study showing stagnation of the contrast in both the stomach and the colonic conduit in delayed films.
3. Results The UGI contrast studies of 43 patients who underwent colonic replacement of the esophagus were included in the study. Their age ranged from 7 months to 17 years. Most cases (41) underwent retro-sternal colon bypass, while only two had a posterior trans-hiatal colonic replacement of the esophagus (Fig. 6). UGI contrast studies were performed routinely following the procedure in 17 cases (first asymptomatic group), while the rest of contrast studies (26) belonged to the second group of outpatients complaining of dysphagia (18) or dyspepsia (8) following previous colonic replacement of the esophagus. One patient underwent CT chest that was ordered by her pediatrician who misinterpreted the colonic conduit in the chest X-ray for a diaphragmatic hernia. A retention mucous cyst was an incidental finding in the study at the site of the retained scared esophagus (Fig. 7). Redundancy of the colonic conduit in the thoracic cavity was a common radiological finding (27 patients; 62.8%). The redundancy was mild (grade 1) in 18 patients, moderate (grade 2) in eight, and severe (grade 3) in only one patient. Higher grades of redundancy were more commonly seen in patients presenting with dyspeptic symptoms (vomiting; aspiration especially on bending and at night) (Table 1). In 88.9% (24 patients), the redundancy was in the right hemi-thorax. On the contrarily, redundancy of the colonic conduit in the abdomen was not a common finding (only 4 cases) (Fig. 3f). We examined the UGI contrast studies of 18 patients who presented with dysphagia following colonic replacement of the esophagus; all of them gave a history of post-corrosive esophageal
stricture as an indication for esophageal replacement. A persistent narrowing could always be depicted in their esophagrams at the site of the esophago-colic anastomosis in the neck (Fig. 8). Finding a stricture at the site of anastomosis provided a clue to the cause of dysphagia, which was managed either by endoscopic dilatation, or by surgical revision of the anastomosis. This should be differentiated from other sites of anatomical narrowing, which are usually seen along the course of the retro-sternal colonic conduit in the contrast studies of asymptomatic cases as well: the first at the thoracic inlet, behind the upper border of the manubrium, and the second at the site of exit of the colonic conduit through the diaphragm (Figs. 6 a,b and 8 c,f). The gastro-colic reflux was difficult to be assessed retrospectively in the contrast studies that were performed the conventional way (baruim swallow, then meal), due to contrast stagnation in the redundant colonic conduit (Fig. 4). To overcome this problem, we modified our technique by performing the barium meal first through the gastrostomy (14 patients), to be followed by the swallow. Reflux of the contrast from the stomach into the colonic conduit was evident in five cases, while it was absent in nine cases that had an anti-reflux mechanism added to the procedure (Fig. 5). In two patients who presented with severe vomiting and hematemesis following a previous retro-sternal colonic bypass procedure, their gastrograms showed gastro-colic reflux with marked distention and loss of haustrations of the colonic conduit (similar to the radiological picture of a toxic megacolon, Fig. 5b). These patients were successfully managed by a re-operation to add an anti-reflux mechanism to the procedure.
A.A. AbouZeid et al. / European Journal of Radiology 84 (2015) 2625–2632
2629
Fig. 5. Barium-meal-first through the gastrostomy tube (gastrograms) to test for the presence of gastro-colic reflux in a group of children following colonic replacement of the esophagus. (a) and (b) showing reflux of contrast into the colonic conduit in the thorax. Note the marked distention of the colonic conduit in (b) with loss of colonic haustrations. (d) and (c) Gastrograms showing absence of gastro-colic reflux in those who had an anti-reflux mechanism added to the esophageal replacement procedure.
Table 1 Correlation between the post-operative symptoms and the degree of redundancy of the colonic conduit used to replace the esophagus in children. Degree of colonic redundancy
Asymptomatic group(17 cases)
Dysphagia group(18 cases)
Dyspepsia group(8 cases)
Grade 0 (no redundancy) Grade 1 (mild redundancy) Grade 2 (moderate redundancy) Grade 3 (severe redundancy)
10 cases (58.8%) 5 cases (29.4%) 2 cases (11.8%) –
5 cases (27.8%) 10 cases (55.6%) 3 cases (16.6%) –
1 case (12.5%) 3 cases (37.5%) 3 cases (37.5%) 1 case (12.5%)
4. Discussion Esophageal replacement is one of the technically challenging procedures in pediatric surgery. Different organs can be used to replace the esophagus: the stomach, the jejunum, or the colon [13]. For many years, the colon has been the organ of choice for replacing the esophagus in children, for its durability and good functional
outcome [14,15]. In the last two decades, the colon has lost some of its popularity against the gastric pull-up procedure that became more predominant [11,12,13,15][11–13,15]. However, the colon has remained the preferred organ for replacing the esophagus at our institution, with satisfactory outcome and acceptable morbidity. It has been shown that choosing between the different techniques tends to be based on personal preference and local experience
2630
A.A. AbouZeid et al. / European Journal of Radiology 84 (2015) 2625–2632
Fig. 6. Upper gastro-intestinal contrast study (lateral view) of three cases following colonic replacement of the esophagus. (a) and (b) The retro-sternal position of the colonic conduit showing ‘anatomical’ narrowing at two points (white arrows): the first behind the upper border of the manubrium at the thoracic inlet, and the second at the site of exit of the colonic conduit through the diaphragm. (c) Posterior mediastinal position of the colonic conduit.
Fig. 7. Three-year-old girl underwent colonic replacement of the esophagus for corrosive stricture. A CT scan was ordered by her pediatrician: (a) mediastinal window, (b) lung window showing a mucous cyst at the site of the retained scared esophagus (*). Five years later, a follow-up MRI (c) and (d) was performed showing reduction in the size of the cyst (*). (c) Axial T1-weighted spin echo showing the cyst exhibiting hypo-intense signal. (d) Axial T2-weighted fast spin-echo showing hyper-intense signal of the cyst. (Cd.: colonic conduit)
rather than on controlled studies and objective data [16]. There are many reports discussing the variable operative techniques and the different routes used for the colonic replacement of the esophagus [1,6,11,15,16]. However, to our knowledge, there is paucity of data describing the post-surgical imaging findings [12,17]. In this report, we presented the expected normal radiological findings following the retro-sternal colon bypass procedure (which is the commonly practiced technique at our institute), as well as the radiological features of the commonly encountered post-surgical complications.
Redundancy of the colonic graft has been a major concern for thoracic surgeons who perform esophageal replacement procedures [1,11,18,19]. In our series, redundancy of the colonic conduit in the thorax was present in 62% of the studied cases. However, redundancy of the colonic conduit is not always symptomatic. That is why we proposed a new grading scale to differentiate between different degrees of redundancy. Mild degrees (grade 1) were the most common (66%), and were usually asymptomatic. More severe degrees are more liable to produce symptoms (delay in transit of food; regurgitation; dysnea with exertion). Several theories about
A.A. AbouZeid et al. / European Journal of Radiology 84 (2015) 2625–2632
2631
Fig. 8. Barium swallow of a group of children who underwent colonic replacement of the esophagus. (a)–(c) belong to three patients complaining of dysphagia, demonstrating strictures at the site of the esophago-colic anastomosis (white arrows). Note in (b) for the diffuse scaring of the proximal esophagus and pharynx due to corrosive injury. (d)–(f) belong to another three patients without dysphagia. Note the site of anatomical narrowing (*) in the retro-sternal colonic conduit at the thoracic inlet behind the manubrium in (c) and (f).
the cause of redundancy have been proposed [20]. Surgeons are advised to avoid using too long grafts to bridge the gap, and to excise any excess length before completing the neck anastomosis [21]. In our series, we have noticed that the colonic redundancy was usually in the right hemi-thorax (89%), which may be explained by the position of the heart and pericardium normally toward the left. Also, we believe that major regurgitation of gastric contents into the colonic conduit would further aggravate existing redundancy through distension (Fig. 5b). Dysphagia was the most common complaint after colonic replacement of the esophagus. This was especially evident among patients with a history of post-corrosive esophageal stricture, probably due to the fibrosis and compromised blood supply of their proximal esophagus. The cause of dysphagia was usually a stricture
at the site of the esophago-colic anastomosis in the neck. Occasionally, the whole proximal esophagus was scared, which would emphasize on the importance of proper pre-operative assessment of the proximal esophagus (either radiologically or endoscopically). A scared proximal esophagus may indicate the need to perform the bypass at a higher level (with the pharynx). In colonic replacement of the esophagus for corrosive injuries, there is a debate about the necessity of excising the scared esophagus [1,4,5,22]. A colonic bypass procedure via a retro-sternal route saves a technically challenging step (esophageal resection) and considerably simplifies the procedure with comparable functional results. However, there is a concern about possible late complications in the retained scared esophagus (malignant transformation and mucous cyst formation) [4,5,22]. Usually the residual scared
2632
A.A. AbouZeid et al. / European Journal of Radiology 84 (2015) 2625–2632
esophagus is a harmless fibrotic tube [23], with only very few reports on patients developing carcinoma [22]. In our series, we had an incidental finding of a mucous cyst in the retained esophagus in one patient. This cyst was small in size (2.5 × 2.4 cm), and got even smaller on follow-up. The last complication is concerning the gastric emptying and regurgitation into the colonic conduit. Early reports highlighted the advantage of utilizing the colon in replacing the esophagus for having a ‘relatively’ resistant mucosal lining to peptic ulceration [4,13]; and there was some reluctance for adding a routine anti-reflux mechanism to the procedure [9]. However, reflux problems are not only restricted to peptic ulceration. We have seen fatal respiratory complications especially in patients with long gap esophageal atresia who underwent the colon bypass procedure early in infancy. Also, in some patients with major gastro-colic reflux, we have noticed severe interference with the normal gastric emptying. The stomach is emptying a majority of contents into the colonic conduit in the chest, which then goes back again down to the stomach (yo–yo movement). This abnormal ‘yo–yo movement’ of the gastric contents will be associated with little amounts passing each time through the normal path into the duodenum and small bowel. With our increased awareness of the possible adverse effects of the gastro-colic regurgitation, it is now our routine practice to add an anti-reflux mechanism to the retro-sternal colon bypass procedure. This study is limited by its retrospective nature. Also, it mainly included a single technique for esophageal replacement ‘the colon bypass procedure’. However, through collecting a relatively large number of contrast studies that belong to both symptomatic and asymptomatic cases, we could provide clear illustration of the typical radiological features of the commonly encountered complications as well as the normal appearance following colonic replacement of the esophagus in children. 5. Conclusion Colonic replacement of the esophagus in children results in considerable anatomical alterations. Knowledge about the normal post-surgical changes and imaging features of the commonly encountered complications can increase the diagnostic confidence among radiologists and clinicians when dealing with these cases. Conflict of interest None. References [1] S.A. Ahmad, K.G. Sylvester, A. Hebra, A.M. Davidoff, S. McClane, P.W. Stafford, et al., Esophageal replacement using the colon: is it a good choice? J. Pediat. Surg. 31 (1996) 1026–1031.
[2] G. Kelling, Oesophagoplastik mit hilfe des querkolon, Zbl. Chir. 38 (1911) 1209–1212. [3] H. Vulliet, De l’oesophagoplastie et des diverses modifications, Sem. Med. 31 (1911) 529–534. [4] A.F. Bahnassy, I.E. Bassiouny, Esophagocoloplasty for caustic stricture of the esophagus: changing concepts, Pediatr. Surg. Int. 3 (1993) 103–108. [5] A.F. Hamza, S. Abdelhay, H. Sherif, T. Hasan, H. Soliman, A. Kabesh, et al., Caustic esophageal strictures in children: 30 years’ experience, J. Pediatr. Surg. 38 (2003) 828–833. [6] R. Belsey, Reconstruction of the esophagus, Ann. R. Coll. Surg. Engl. 65 (1983) 360–364. [7] C.D. Sherman, D. Waterston, Esophageal reconstruction in children using intrathoracic colon, Arch. Dis. Child. 32 (1957) 11–16. [8] N.V. Freeman, D.T. Cass, Colon interposition: a modification of the Waterston technique using the normal esophageal route, J. Pediatr. Surg. 17 (1982) 17–21. [9] W.A. Dale, C.D. Sherman, Late reconstruction of congenital esophageal atresia by intrathoracic colon transplantation, J. Thorac. Surg. 29 (1955) 344–356. [10] T.G. Canty, B.E. LoSasso, One-stage esophagectomy and in situ colon interposition for esophageal replacement in children, J. Pediatr. Surg. 32 (1997) 334–337. [11] T.W. Rice, Right colon interposition for esophageal replacement, Oper. Tech. Thorac. Cardiovasc. Surg. 4 (1999) 210–221. [12] S. Ho, P. Patel, K. Mchugh. Imaging appearance following Paediatric oesophageal replacement surgery: an analysis of 72 cases at a tertiary referral centre. http://dx.doi.org/10.1594/ecr2012/C-2418. [13] J.L. Marks, W.L. Hofstetter, Esophageal reconstruction with alternative conduits, Surg. Clin. N. Am. 92 (2012) 1287–1297. [14] T.R. DeMeester, K.E. Johansson, I. Franze, E. Eypasch, C.T. Lu, J.E. McGill, et al., Indications, surgical technique, and long-term functional results of colon interposition or bypass, Ann. Surg. 208 (1988) 460–474. [15] J.C. Wain, C.D. Wright, E.Y. Kuo, A.C. Moncure, E.W. Wilkins, H.C. Grillo, et al., Long-segment colon interposition for acquired esophageal disease, Ann. Thorac. Surg. 67 (1999) 313–317. [16] G.S. Arul, D. Parikh, Oesophageal replacement in children, Ann. R. Coll. Surg. Engl. 90 (2008) 7–12. [17] Am Sieber, W.K. Seiber, Colon transplants as esophageal replacement: cieradiographic and manometric evaluation in children, Ann. Surg. 168 (1968) 116–122. [18] T. Yasuda, H. Shiozaki, Esophageal reconstruction with colon tissue, Surg. Today 41 (2011) 745–753. [19] H. Elshafei, E. Elshafei, M. ElDebeiky, N. Hegazy, A. Zaki, S.A. Hay, Colonic conduit for esophageal replacement: long-term endoscopic and histopathologic changes in colonic mucosa, J. Pediatr. Surg. 47 (2012) 1658–1661. [20] D.C. Strauss, M.J. Forshaw, R.C. Tandon, R.C. Mason, Surgical management of colonic redundancy following esophageal replacement, Dis. Esophagus 21 (2008) E1–E5. [21] A.F. Hamza, Colonic replacement in cases of esophageal atresia, Sem. Pediatr. Surg. 18 (2009) 40–43. [22] L. Burgos, S. Barrena, A.M. Andres, L. Martinez, F. Hernandez, P. Olivares, et al., Colonic interposition for esophageal replacement in children remains a good choice: 33-year median follow-up of 65 patients, J. Pediatr. Surg. 45 (2010) 341–345. [23] J.G. Raffensperger, S.R. Luck, M. Reynolds, D. Schwarz, Intestinal bypass of the esophagus, J. Pediatr. Surg. 31 (1996) 38–47.