Short summary

Long-gap esophageal atresia (LGEA) is a very rare birth defect with unknown cause in which a baby is born without part of the esophagus. Several surgical procedures to obtain esophageal continuity exist but both short- and long-term complications are common. To date it is unclear which treatment is preferable and the literature contains only small patient series. This multi-centre study will be the largest series published on this topic and aims to investigate which surgical treatment in Denmark, Norway, Sweden and Finland is preferred and to examine the corresponding complications. It attempts to find a gold standard in LGEA treatment to guide the future management of this malformation.

Rationale

Esophageal atresia (EA) occurs with a prevalence of 1 in every 2500-4500 live births. The malformation originates from a developmental disruption with incorrectly separation of the embryogenic foregut into the trachea and the esophagus, resulting in an upper and a lower esophageal pouch and sometimes in a tracheo-esophageal fistula (TEF). However, the embryology is not completely understood and the etiology is unknown.

There are five different types of congenital esophageal atresia classified by two different systems, the Gross and the Vogt classification systems, which are used interchangeably. The most common type is Gross type C (Vogt type 3B) consisting of a distal TEF with a proximal EA - it comprises approximately 88.5% of all cases. Isolated EA (Gross type A/Vogt type 2) occurs in approximately 8% of all cases.

Recent advances in surgery, pediatric anaesthesia, neonatal intensive care and parenteral nutrition have improved survival rate for children born with EA to approximately 95%. Even new-borns with very low birth weight and associated congenital anomalies have a better chance at survival.

Results from a study based on routinely collected data from twenty-three European registries of congenital anomalies (EUROCAT) show, that EA is often associated with other major anomalies. A common association of congenital anomalies is the VACTERL association consisting of vertebral, anal, cardiac, tracheal, esophageal, renal and limb anomalies. The study found a high proportion of associated anomalies including congenital heart defects, urinary tract anomalies, other gastrointestinal anomalies and limb anomalies being the most frequent.

The preferred surgical treatment of EA is a primary anastomosis between the two pouches and closure of TEF if present. However, primary anastomosis may not be possible if the distance between the 2 esophageal pouches is too large.

Long-gap esophageal atresia (LGEA) occurs in approximately 15% of those born with EA. Throughout the literature the definition of LGEA is inconsistence. Some define it as either 2 or more centimetres or 2 vertebral body spaces between the 2 esophageal pouches, whereas others define a gap of 3.5 centimetres or more as an "ultra long gap". When speaking of LGEA in this study, it means that the gap between the 2 esophageal pouches was too long to perform a primary end-to-end anastomosis. All these patients will be registered, but the primary focus in the present study will be on Gross type A and B (gasless abdomen at plain X-ray preoperatively).

In the case of LGEA, several procedures have been developed with the aim to obtain esophagogastric continuity. They include delayed oesophageal anastomosis and organ interposition (colon, jejunum and ventricle).

Delayed primary oesophageal anastomosis (DPA) is based on spontaneous natural growth and hypertrophy of the esophageal segments. Puri et al. observed, that the spontaneous growth and hypertrophy of esophageal segments occur at a faster rate than overall somatic growth in the absence of any stretching or bougienage of the esophageal ends. The theory is that the swallowing reflex and the reflux of gastric contents into the lower esophageal pouch stimulates natural growth. The maximal growth occurs during the first 8 to 12 weeks of life.

A cumulative meta-analysis presents complications and long-term outcome in new-borns with LGEA managed by DPA. In most studies, early complications were leaks of which some needed drainage or reoperation. Stricture formation was also a common complication, previous anastomotic leak being one of the most important risk factors. Other common complications were symptomatic gastroesophageal reflux (GER), esophagitis, dysphagia, aspiration pneumonia, growth retardation, Barrett´s metaplasia and recurrent fistula.

The survival rate after DPA is reported to be approximately 90%. Long-term follow-up showed that the majority of patients have normal growth and developmental curves after the procedure. A disadvantage of DPA consists of waiting for the esophageal segments to grow and hypertrophy, resulting in prolonged hospital stay and constant threat of aspiration pneumonia.

The Foker technique is based upon the basic biological principal that axial tension provides the growth signal. Traction may be external or internal only or a combination of both. The original study by Foker showed that even the most rudimentary segment has the potential to achieve a length, which makes a delayed oesophageal anastomosis possible. 

Interposition of the colon is another option for bridging the long gap between the two esophageal pouches. The early complications of colon replacement are anastomotic leakage, upper anastomotic stricture and pneumothorax. The majority achieved normal swallowing one month postoperative. Late complications included redo of the upper anastomosis due to persistent stricture or redundancy of the colon. Another problem was reflux.

Interposition of the jejunum and gastric transposition provide other options to overcome the gap in LGEA. Gallo et al. reported the experience from gastric pull-up (GPU) and jejunal interposition (JI) in a comparative study. There was no perioperative mortality in both surgical procedures and it showed that the blood supply is less generous in the jejunal grafts than in the pulled up stomach, which makes it more challenging using jejunum. While long pedicle grafts of the jejunum can be constructed they tend to curl and are constructed at the expense of the remaining jejunum. JI had a higher perioperative total anastomotic complication rate (leak, stenosis) than GPU. Anastomotic leak and stenosis as well as pneumothorax represent the most common complications in both GPU and JI. Long-term outcome showed that no patients reported functional obstruction after GPU, whereas it was present in 46% of the patients who underwent JI. The rate of dysphagia was higher in the GPU group. Gastric transposition has the least short-term morbidity and mortality compared with the alternative procedures.

Another option to repair the esophagus is primary reversed gastric tube (RGT) reconstruction, where the gastric tube is composed of the greater curvature of the stomach. Borgnon et al. reviewed the experience using such an isoperistaltic gastric tube to replace the esophagus and concluded, that this approach seems to have fewer ischemic complications in the early postoperative period compared to other approaches, whereas long-term results are similar. A retrospective study compared the techniques delayed primary anastomosis and esophageal replacement with greater curvature gastric tube in matters of perioperative and long-term complications and outcomes. It showed similar risks of perioperative complications between the two treatment options but the gastric tube caused more long-term complications than delayed primary anastomosis, especially stricture. No differences in long-term outcomes were found.

In conclusion there is no gold standard in the surgical treatment of long-gap esophageal atresia, and the literature contain only small patients series. Further multi-centre studies including higher number of patients are necessary to evaluate the different methods used. The Nordic countries are ideal for such studies at the health-care systems are comparable.

The primary aim of this study is to investigate which surgical approach for long-gap esophageal atresia Gross type A and B pediatric surgeons in Denmark, Norway, Sweden and Finland prefer. The secondary aim is to investigate postoperative complications within one year including total hospital stay, stay at intensive care unit and duration on parenteral nutrition. The tertiary aim is also to register all patients with different types of esophageal atresia (Gross type C and D) where primary anastomosis was not performed. The intention is to analyse these patients in a separate study-protocol.  

Description of the cohort

The population will consist of patients born with long-gap esophageal atresia Gross type A or B in a 15-year period from the 1st of January 2000 to the 31st of December 2014 where primary anastomosis was attempted but not succeeded or where a delayed anastomosis was planed. All departments of pediatric surgery in the Nordic countries will be invited to participate in the study under the auspices of the Nordic Council. We expect to include at least 200 children.

Data and biological material

Data from the participating countries will be collected by retrospective medical records review. We wish to collect data on the following:

• Date of birth
• Gestational age
• Birth weight
• Birth length
• C-section
• Vaginal delivery
• Type of congenital esophageal atresia
• Other congenital anomalies
• Attempted primary anastomosis
• Delayed primary anastomosis
• Active elongation technique
• Interposition procedure (gastric pull-up, colonic interposition, jejunal interposition, gastric tube)
• Complications (anastomotic stricture, anastomotic leakage, gastroesophageal reflux)
• Mortality (pre-, intra- and postoperative)
• Total hospital stay
• Stay at intensive care unit
• Duration of total parenteral nutrition
• 1-year follow-up

Collaborating researchers and departments

Department of Surgery of Gastroenterology, Paediatrics and Endocrinology, Odense University Hospital

• Professor and Head of Research Niels Qvist, MD, DMSc
• Rikke Neess Pedersen, MD, PhD
• Lars Rasmussen, MD