Esophageal atresia associated with anorectal malformation: Is the outcome better after surgery in two stages in a limited resources scenario?

AIMS: To analyze whether outcome of neonates having esophageal atresia with or without tracheoesophageal fistula (EA±TEF) associated with anorectal malformation (ARM) can be improved by doing surgery in 2 stages.
MATERIALS AND METHODS: A prospective study of neonates having both EA±TEF and ARM from 2004 to 2011. The patients with favorable parameters were operated in a single stage, whereas others underwent first-stage decompression surgery for ARM. Thereafter, once septicemia was under control and ventilator care available, second-stage surgery for EA±TEF was performed.
RESULTS: Total 70 neonates (single stage = 20, 2 stages = 30, expired after colostomy = 9, only EA±TEF repair needed = 11) were enrolled. The admission rate for this association was 1 per 290. Forty-one percent (24/70) neonates had VACTERL association and 8.6% (6/70) neonates had multiple gastrointestinal atresias. Sepsis screen was positive in 71.4% (50/70). The survival was 45% (9/20) in neonates operated in a single stage and 53.3% (16/30) when operated in 2 stages (P = 0.04). Data analysis of 50 patients revealed that the survived neonates had significantly better birth weight, better gestational age, negative sepsis screen, no cardiac diseases, no pneumonia, and 2-stage surgery (P value 0.002, 0.003, 0.02, 0.02, 0.04, and 0.04, respectively). The day of presentation and abdominal distension had no significant effect (P value 0.06 and 0.06, respectively). This was further supported by stepwise logistic regression analysis.
CONCLUSIONS: In a limited resources scenario, the survival rate of babies with this association can be improved by treating ARM first and then for EA±TEF in second stage, once mechanical ventilator care became available and sepsis was under control.

INTRODUCTION

In the developing countries, still, the availability of neonatal intensive care units (NICUs) is meager. The aim of our study was to find out whether survival of the neonates suffering from anorectal malformations (ARM) and esophageal atresia with or without tracheoesophageal fistula (EA±TEF) can be improved by performing emergency surgery for ARM by bowel decompression first and deferring EA±TEF repair, until septicemia resolves and postoperative ventilator care is available. We also studied the effects of other variables, such as gestation age, birth weight, septicemia, day of presentation to the hospital, cardiac disease, abdominal distension, and pneumonia on the survival.

MATERIALS AND METHODS

Study design and statistics

This is a prospective study on 70 neonates (having EA±TEF with ARM) admitted from December 2004 to December 2011. The ethical approval was taken from ethical committee of the university. Informed and written consent was taken from parents. The data were analyzed using SPSS 17.0 version for Windows. Continuous variables were expressed as median values and ranges (minimum to maximum), as well as in percentages. The analysis was performed using the Pearson χ2 for categorical variables and Student's t test for continuous variables. Simultaneous multiple logistic regression analysis was also applied for different variables. P value < 0.05 was considered as significant.

Data collection and categorization procedure

After baseline investigations, the babies were nursed in semi-recumbent position with intermittent low-pressure oropharyngeal suction, intravenous fluids, antibiotics, and vigorous vital monitoring. In patients where both surgeries were performed simultaneously, enteral nutrition through nasogastric tube was started from next day. Total parenteral nutrition was given to all extremely premature neonates and to those where enteral nutrition was deferred for more than 5 days due to any reason (awaiting surgery for EA±TEF, paralytic ileus, necrotizing enterocolitis, and others). The demographic data (sex, gestation age, birth weight, and day of presentation to the hospital) were recorded.

The neonates were selected for surgery either in single or 2 stages. The selection of a neonate for staging of surgery was relative, that is, each neonate didn′t fulfill all the criteria of one group. Relatively full term, appropriate for gestation age, sepsis screen negative, earlier presenting (1st or 2nd day of birth) babies who had adequate length of upper esophageal pouch, and not much abdominal distension or respiratory distress were operated in single stage for both (ARM and EA±TOF). Conversely, low birth weight, small for gestation age, sepsis screen positive, late presenting (beyond 2 days) babies who had short upper esophageal pouch with abdominal distension, or respiratory distress underwent surgery in 2 sittings. First-stage surgery for ARM was done under local anesthesia. Second-stage surgery for EA±TEF was undertaken after stabilization of vitals, control of sepsis, and availability of postoperative ventilator care. Conversely, if the neonatal ventilator care/NICU care was immediately available, a few small for gestation age and delayed presenting patients were also operated in a single stage.

Procedure

For nondecompressing low ARM, anoplasty was performed. For male (intermediate and high) ARM and nondecompressing female ARM diversion colostomy was performed. The decompressing (anovestibular, rectovestibular, anocutaneous fistulas) ARMs were not treated in neonatal period. The patients having Gross's type A EA underwent cervical esophagostomy and feeding gastrostomy, whereas Gross's type C EA underwent end-to-end esophagoesophageal anastomosis. Two babies having difficult intubations during induction of anesthesia also needed tracheostomy. Postoperatively, as per indications, the babies were kept on mechanical ventilation. The final outcome was determined by the survival of the neonate. In followup, once the baby gained adequate weight, definitive management for ARM was undertaken.

RESULTS

During the study period of 7 years, 1050 neonates of EA±TEF and 3000 neonates of ARM were admitted. Seventy-six neonates had both EA±TEF and ARM. Three neonates expired without any intervention (cyanotic heart disease and severe sepsis), whereas parents of 3 patients refused further treatment after colostomy. The sepsis screen was positive in 71.4% (50/70) patients. Twenty neonates underwent single-stage surgery, 30 neonates were operated in 2 stages. Nine neonates expired only after colostomy, whereas 11 neonates needed only EA±TEF repair because of associated decompressing anovestibular, rectovestibular, and anocutaneous fistula. Thus, the final outcome analysis involved 50 neonates.

The admission rate of this association was 1 per 290. Seven percent (70/1050) of EA±TEF neonates had ARM, whereas 2% (70/3000) of ARM neonates had EA±TEF. Sixty-six neonates had Gross's type C and 4 neonates had Gross's type A EA. Of the male patients, 43, 5, and 15 neonates had high, intermediate, and low-type ARM, respectively. Of female patients, 4, 2, and 1 had anovestibular fistula, anterior ectopic anus, and rectovaginal fistula, respectively. Forty-one percent (24/70) neonates had VACTERL association and 8.6% (6/70) neonates had multiple gastrointestinal atresia syndrome (MGIA). The patients having MGIA included 3 cases of anovestibular fistulas, 1 case of anterior ectopic opening, and 2 intermediate male ARMs. The diagnosis of intermediate male ARM anomalies was made only by colostogram at second admission. Association of anomalies is shown in Table 1. The mean interval between 2 surgeries in second group was 4±0.8 days (range 3–7 days).

Table 1

Associated anomalies besides EA±TEF and anorectal malformation

Distribution of neonates according to the treatment they received, and difference of variables between both the groups is shown in Table 2. The overall survival rate was 60% (42/70) . The mortality rate for VACTERL was 88% (21/24). The survival was 45% (9/20) in single stage and 53.3% (16/30) in patients operated in 2 stages. On applying Pearson χ2 test, the difference was significant (P = 0.04). The data analysis of 50 neonates showed that neonates who survived had significantly better birth weight, gestational age, negative sepsis screen, absent cardiac diseases, absent pneumonia, and underwent 2-stage surgery (P value 0.002, 0.003, 0.02, 0.02, 0.04 and 0.04, respectively), whereas day of presentation and abdominal distension had no significant effect (P value 0.06 and 0.06, respectively) on the survival. The comparison of survived neonates in both groups showed significant difference in the birth weight, gestation age, septicemia, abdominal distension, and cardiac disease (P < 0.05, Table 3), which indicated that even in the presence of adverse parameters the survival may be improved by staging the surgery.

Table 2

The characteristics of neonates, who underwent surgery for both EA±TEF and anorectal malformation in single and two stages

Table 3

The data of survived patients after surgery for both anorectal malformation and EA±TEF

Furthermore, stepwise logistic regression analysis showed that gestation period, weight, and positive sepsis screening had a strong positive effect on the survival of neonates. Besides them, presence of pneumonia, cardiac disease, and stages of surgery also had a significant effect on the survival but the day of presentation to the hospital and degree of abdominal distension did not have much effect on the survival.

All patients having pure EA (n = 4) were negative for sepsis screen, and they were operated in single stage. Therefore, direct comparison of stages of surgery on survival could not be ascertained in these patients. On comparing the survival among patients having decompressing (11/70) and nondecompressing ARM (59/70) with EA±TEF, we found that the survival of former group (90.9%, 10/11) was better compared with latter (42.8%, 25/59, P = 0.009).

DISCUSSION

The most frequently associated anomalies with EA±TEF are cardiac (49%) and anorectal malformations (15%).[1-4] We had relatively less cases (7%) of associated ARM. Earlier, most of the mortality in EA±TEF was due to respiratory failure (soiling of lung, hyaline membrane disease), prematurity, and complications of surgery.[1] Based on these, Waterston gave a criteria regarding risk stratification and survival of EA±TEF neonates.[13] However, with the advancement of NICU and treatment of cardiac diseases, associated anomalies (chromosomal anomaly) have taken the lead role.[15] This may be true in western world but the situation is different in developing countries, such as ours where deliveries are still conducted by untrained personnel, and neonates are septicemic at admission.[6-8] The congenital malformations as a primary cause of mortality contribute 9.2% of all NICU mortalities.[67] Furthermore, in the developing countries there is no adequate availability of NICU everywhere. Even in most referral centers sometimes the neonates′ input is so much that it cannot fulfill the requirements as needed.[9-12] In resource-limited situations similar to ours, we have attempted to develop a strategy, which decreases the septicemic load of these neonates and with best utilization of available resources, survival of neonates can be improved. Till date, in the literature, there is no such study, which has tried to improve the survival of EA±TEF with ARM neonates by performing surgery in 2 stages.

In our 2 study groups, despite the differences in birth weight, gestation age, septicemia, abdominal distension, and congenital heart disease, the survival was comparable. Probably, it was the stages of surgery, which may have improved the survival in adverse conditions. The hypothesis was further tested by stepwise logistic regression analysis, which also proved that staging of surgery had its independent effect on the survival.

The overall survival rate (50%, 25/50) of association was higher than that in Kumar et al.'s study (43%) but much lower than neonates having only EA±TEF (86%).[128] We had a very high mortality rate (88%) in VACTERL association compared with 58% in Kumar et al.'s study.[8] The reason was high proportion of septicemic neonates in the study group due to aspiration pneumonia/chemical pneumonitis and increased fecal load to the bowel.

To conclude, in adverse conditions (especially positive sepsis screen, and unavailability of neonatal mechanical ventilator), primary surgery for decreasing fecal load (for ARM) and deferring surgery for EA±TEF may improve the survival of neonates having this association. Although the study couldn’t make direct comparison between pure EA and EA with distal tracheoesophageal fistula, the former group has better prognosis as all our patients were negative for sepsis screen at the time of presentation. The study also indicates that decompressing ARM had better survival than nondecompressing ARM, if both are associated with EA±TEF.

Based on various parameters, the categorization of patients of EA±TEF can be done, so as to decide the type of surgery. This may have a favorable affect on the outcome of babies suffering from this entity.