INTRODUCTION

Esophageal atresia (EA) is basically managed surgically by primary repair from several decades (1-4).  But still in developing countries, many patients are diagnosed lately and sent to our tertiary hospital for surgical correction. Because of presence of high death rates with primary repair at our hospital, we started to offer staged repair strategy involving tracheoesophageal fistula (TEF) division, placement of a feeding gastrostomy and cervical esophagostomy during the first weeks of life. After that, we did this research to investigate if there is any significant difference in the survival rate between primary repair and staged repair in the neonatal period.

PATIENTS AND METHODS

The clinical registrations of patients with late diagnosed esophageal atresia, referred to our hospital from February 2014 to February 2016, were studied. A retrospective analysis was done to study: Sex, age at diagnosis, type of atresia, associated congenital anomalies (VACTERL anomalies= vertebral, anal, cardiac, tracheo-esophageal, renal and limb anomalies), presence of pneumonia, Waterston classification, age at surgery, type of surgery, morbidity and mortality. The diagnosis was established on clinical signs like failure to pass naso-gastric tube, frothing, failure to feeds, respiratory distress, scaphoid or distended abdomen. Plain X-ray was used to find gaseous distention of stomach (with distal tracheoesophageal fistula) and coiling of 10 f Ryle tube in upper mediastinum. Water soluble contrast was used via Ryle tube to delineate blind proximal esophageal pouch.

Waterston groups classification was used to show the patients prognosis (Group A: Birth weight >2.5 kg. Group B: Birth weight 1.8-2.5 kg and otherwise healthy or birth weight >2.5kg with moderate pneumonia or moderate cardiac anomalies e.g. PDA, VSD and ASD. Group C: Birth weight <1.8 kg or higher with severe pneumonia or severe cardiac anomalies. Gross anatomical classification is used in the study (Type A; Isolated EA. Type B; EA with Proximal TEF. Type C; EA with Distal TEF. Type D; EA with proximal& Distal TEF. Type E; isolated TEF). Patients were classified into two groups regarding to the type of surgery.

The Surgical approach was primary repair in 15 cases (Thoracotomy then division of the tracheoesophageal fistula with proximal and distal esophageal anastomosis through extra-pleural approach).The first operation of staged repair strategy were done for the next 15 patients (thoracotomy with division of the tracheoesophageal fistula + feeding gastrostomy + approximation of proximal esophageal pouch to the distal esophageal end if the gap is long + cervical esophagostomy).

All analyses were executed by SPSS version 16.0. Comparison between groups was made with Chi-square (χ2) test and Fisher exact test when appropriate. Values of P<0.05 were considered statistically significant.

RESULTS

Thirty cases of delayed diagnosed esophageal atresia were managed over the study period. In total, 21 males (70%) and 6 females (20%) with three unknown gender (10%) were enrolled in the study. Two pre-term neonates (6.7%) were included with surgical mortality of 100%. The rest of cases (28 patients) were full-term with surgical mortality of 67.9% (19 patients). The mean weight at operation was 2.29±1.1 kg. The minimum birth weight was 1.2 kg and the maximum was 5 kg. The mean age of diagnosis was 8.9±6.23 days old.

Three cases (10%) were diagnosed at age of 4 days while the rest of cases (90%) were diagnosed at or after the age of 7 days. According to Gross classification, twenty six cases (86.7%) of the 30 patients had esophageal atresia with distal tracheoesophageal fistula (Type C). Two cases (6.7%) cases had esophageal atresia with both proximal and distal tracheoesophageal fistula (type D). Two cases (6.7%) cases had only isolated tracheoesophageal fistula (type E). Associated congenital anomalies were present in 10 patients (33.3%) It included cardiac diseases in eight patients (3 VSD, 3 PDA and 2 dextrocardia), renal disease in 1 patient (bilateral hydronephrosis) and limb anomaly in one patient (hypoplastic thumb). Number of patients with severe pneumonia and oxygen saturation less than 90% were 24 cases (80%); whereas 6 patients (20%) had mild to moderate respiratory distress and oxygen saturation more than 90%.

The mean age of operation was 12.8±8.7 days old. Long gap was an indication of staged repair and because of that it was not considered as a risk factor in this study. Post-operative complications were recorded which one patient had postoperative pneumothorax and subsequent serve hypoxia and was managed by urgent chest drain insertion. Development of serve form of pneumonitis followed by sepsis or development of sepsis alone were present in 80% of cases (24 patients) and was the direct cause of death in 70% of cases (21 patients were died in the study). Other postoperative complications such as major anastomotic leak in 3 patients (10%), feeding gastrostomy leak in 3 patients (10%) and surgical site infection in 1 patient (3.3%). Mortality rate was 80% with primary repair (12 from 15 patients) and 60% with staged repair (9 from 15 patients). The overall mortality rate was 70% (21 from 30 patients).

Patients were classified according to Waterston's groups to determine the correlation among the patients' classification, surgical technique and surgical outcome (table 1). Survival was 100% in group A of both techniques and 66.6% in group B. Little differences in the survival rate of group C appeared which staged repair survival reach 20% in comparison with no survival (zero%) with primary repair group at all. There was no statistical relationship between Waterston's groups, surgical techniques and outcomes (table 1).

Table 1

Waterston's classification groups and surgical techniques outcomes.

Table 2

Survived & died cases with risk factors relationship.

Table (2) compares between survival and death rate from side and its association with risk factors and its significance by P-value calculations.  It revealed significant association of low birth weight, delayed diagnosis, lower preoperative O2% saturation and presence of anomalies, all with higher incidence of mortality rate.

DISCUSSION

This study shows predominance of male sex (70%) as it is mentioned in many reports before (5-9).   Pre-term babies are represented by 6.7% with 100% death rate. It means strong correlation between the prematurity and the mortality in this review but it is not exclusively due to low number of the preterm babies. However, many studies reported significant statistical relationship between prematurity and surgical results (6, 10).  Mean age at diagnosis was 8.9±6.23 days old in our review which it is older in comparison with reports like Osei-Nketiah at al. (5.12±4.04 days old) from Nigeria (11) and Verma et al. (3.12 ± 1.87 days old) from India (12).

There is extremely strong relationship between the lower O2 saturation and presence of pneumonia with surgical mortality in our review (P=0.0002).  Those findings is like what was mentioned by Narasimman et al. (10) that they found a significant correlation (P= 0.025).

In our review, we noticed significant association between age at surgery and surgical outcome which the survival rate is significant better when the age at operation was 7 days old or less. This is mostly due to exposure of the cases to longer time of malnutrition, progressive aspiration and pneumonia with poorer general condition. In contrary to our findings, many reports said that is no significant link between age at surgery and mortality (10, 11). But, we consider that the higher incidence of delayed presentation, and subsequent delayed surgical intervention, has a direct relationship with higher mortality rate.

Staged repair outcome (6 survived patients from 15) is better, numerically, than with primary repair (3 survived patients from 15). But, statistically, it didn’t show significant differences in the surgical results. This outcome is supported by the results which were reported by many studies (10, 11). Our study did not show any correlation between Waterston's classification, type of operation and surgical outcome. It means that whatever the surgical strategy, the outcome is directly proportionate with the preoperative general condition. But even with delayed presented cases, Waterston group A had still the best survival rate in the study (100%). It means that is no need to add the delayed diagnosis as a separate item in the prognostic classification as some authors suggested before (13).

Eighty percent of cases suffered from postoperative complications that were the cause of death (70% of the patients). The commonest complication was sepsis with or without pneumonitis leading to low cardiac output and cardiac arrest. The postoperative complications usually are not related to surgical interventions. Postoperative pneumothorax, pneumonitis, sepsis, major anastomotic, feeding gastrostomy leak and surgical site infection, all those postoperative complications are also reported in many other studies before (14, 15).

The postoperative survival rate of delayed presentation of esophageal atresia is 30% (40% with staged repair and 20% with primary repair).  We considered that the delayed diagnosis (older than 4 days old) and delayed surgical interference (older than 7 days old)   are strongly associated with higher mortality rate. Some studies said that the age at admission is not a worse prognostic factor but the cutpoint of age at admission was 48 hours old (16). Those kinds of problems are present only at hospitals in developing countries. For example, Ameh et al. reported 80% mortality and Osei-Nketiah et al mentioned that the overall postoperative mortality rate was 58.7% at tertiary Centers in Nigeria due to similar reasons (11, 17). The leading causes of sepsis and higher death rate in cases of delayed diagnosis are malnourishment with chemical pneumonitis and aspiration pneumonia (13).

Many metabolic complications faced the neonatal ICU staff during the management of those cases like dehydration, ketosis and electrolytes disturbance (18). Worse pulmonary conditions of those cases pushed us for staged repair strategy, as pulmonary complications can be optimized with better outcome. Preoperative optimization of delayed diagnosed esophageal atresia has a better effect than providing surgery in unstable patient (19). Cameron Haight recorded the first successful primary repair in 1941 (1, 20). The survival rate of Waterston series was 56.7% from nearly seven decades (20, 21). At 1980's, survival had improved to be above 85% and reached over 90% at some hospitals (20). The development in neonatal intensive care and neonatal anesthesia were the keys to enhance the survival rate beside the early diagnosis of the anomaly (1, 22, 23).  Despite of presence of patients with severe associated anomalies, mortality rate is lower than 5% at some centers in the developed countries which are providing the best neonatal intensive care (1, 24).

One of limitations of this study is the fewer number of the study cases and the research should be continued to contain larger number to pick up more significant conclusions.

CONCLUSIONS

The delayed presentation of esophageal atresia, by itself, is a leading cause of higher mortality rate. It is described hugely in the developing countries.  Most of the mortality is directly related to presence of pneumonia and sepsis.  There is no statistically significant superiority of staged surgical repair over the primary repair in the surgical outcome. The sample size is small and a prospective larger sample size study is needed in the future to ensure a representative distribution of the cases. Waterston prognostic classification is still providing an efficient method to predict the esophageal atresia fate.