Described as an interruption of the continuity of the esophagus, esophageal atresia (EA) encompasses a group of congenital anomalies and occurs in association with tracheoesophageal fistula (TEF) or as an isolated entity 1. The knowledge of the size of the gap between the atretic esophageal ends before surgery is of clinical importance and it is also a prognostic indicator of mortality and morbidity 2,3. This clinical entity is divided into short-gap esophageal atresia (SGEA) and long-gap esophageal atresia (LGEA) depending on the distance between the atretic esophageal segments.
The distance between the esophageal ends, which constitutes SGEA and LGEA, lacks a strict numerical definition and there are no uniformly accepted criteria that define these entities. It has been reported that LGEA has gap intervals measuring greater than 2 cm, with a cutoff value of 2 cm or 2–3 vertebral bodies (VB) 4–6. There are other suggested classifications in the literature including the nomenclature of short, intermediate, and long gap intervals with 1, 2.5, and 3 cm as the cutoff points, respectively 3,7. A 3.5-cm ‘ultralong’ gap has also been described 8. Epidemiological studies concerning SGEA and LGEA are limited. Thus, we conducted the present study to analyze the characteristics of patients with SGEA and LGEA and to provide valuable data on this issue. It was also aimed in the current study to compare the demographic characteristics between patients with SGEA and LGEA in the light of relevant literature.
Patients and methods
Institutional ethical approval for the study was obtained. Medical records of all patients managed for EA spectrum in our department between 2003 and 2012 were evaluated retrospectively. In defining the type of atresia, SGEA and LGEA were described as a gap of less than three VB or three or more VB in length between the atretic esophageal ends, respectively. In patients with a gastrostomy tube, by introducing a bougie into the upper pouch and another into the lower pouch through gastrostomy, the gap was evaluated by radiological means (Fig. 1). In patients who underwent thoracotomy for ligation of TEF, assessment of the gap between the atretic esophageal segments was performed during the surgical intervention (Fig. 2). Demographic data included maternal age, the number of parities and deliveries, the presence of polyhydramnios and the prenatal diagnosis, sex, the gestational age and prematurity, the type of delivery and the birth weight, age at the time of diagnosis and treatment, the presence of associated anomalies including VACTERL-type and non-VACTERL-type anomalies, the type of EA according to Gross classification, and discrepancies between the diameters of atretic esophageal ends. VACTERL association was defined if the patient had two or more anomalies of the vertebral, anorectal, cardiac (excluding patent ductus arteriosus, patent foramen ovale), renal/genitourinary, and limb systems.
Statistical analyses were performed using IBM SPSS (ver. 18; IBM Co., Armonk, New York, USA). Data are presented as mean±SD, median (interquartile range) according to their distribution. Qualitative variables were expressed as percentages. Distribution of numerical data was assessed by graphical and statistical methods. Statistical analysis was performed with Student’s t-test to compare numerical values with normal distribution. To compare numerical values with an abnormal dispersal range, interquartile ranges were assessed using the Mann–Whitney U-test. Fisher’s exact test and χ2-tests were used for the comparison of categorical data. A P-value of less than 0.05 was considered statistically significant.
There were 99 patients treated for the diagnosis of EA spectrum during the study period, and the distribution of anatomical types of EA is shown in Table 1. Of the included patients, 81 were in the SGEA group and 18 were in the LGEA group. The characteristics of patients with SGEA and LGEA are depicted in Table 2. Most of the parameters studied did not differ between the two groups. Type-C EA (n=77/81) was more prevalent in patients with SGEA and type-A (n=8/18) was more frequent in children with LGEA (P<0.05). The frequency of prenatal diagnosis (2.5% for SGEA vs. 22.2% for LGEA) was more common in the LGEA group (P<0.05). Cesarean section was seen to occur more commonly than normal vaginal delivery in both groups (56.8% for SGEA vs. 66.7% for LGEA). Durations of ventilatory support (14.4±2.5 vs. 36.0±8.4 days; P<0.05) and hospital stay (30.0±3.8 vs. 69.7±15.2 days; P<0.05) were longer in the LGEA group. Patients with LGEA had a higher mortality rate (35.8 vs. 55.6%; P<0.05).
The rate of EA may be related to the maternal age, and in a study where maternal age was analyzed, both young (<20 years) and older (>35 years) mothers were shown to have an increased risk 9. In a recent study, compared with children of mothers younger than 20 years, children of women giving birth at 35–40 years and above 40 years showed a two-fold and three-fold increased risk of EA, respectively 10. In a most recent cohort from France, the median age of mothers giving birth to a newborn with EA at delivery was 30 years 11. In another study, the mean ages of mothers of children with LGEA and SGEA were 30.1 and 28.9 years, respectively 12. Average ages of mothers in both groups in the current study did not show a statistically significant difference, but as a whole, they were found to be lower compared with those reported previously 9,10,12. This finding may be explained by true population differences including biologic and environmental factors concerning the ages of the mothers.
Parity may be an independent risk factor for birth defects. Associations between maternal parity, the number of deliveries, and birth defects have been observed previously: compared with primiparous mothers, nulliparous mothers were more likely to have infants with a number of congenital birth defects including EA 13. Our series is dissimilar to this finding in that there were only one and three nulliparous mothers giving birth to patients with LGEA and SGEA, respectively. Although there is no definite explanation for this finding, variability in population compositions might have played a role. Although the mean number parities of mothers giving birth to children with LGEA was found to be higher than those of mothers with babies having SGEA, the median number of parities and deliveries in this study did not show statistical significance between the two groups. Research on the biologic or environmental factors that may be associated with the number of maternal parities and birth defects may be helpful in explaining some or all of these possible associations.
Polyhydramnios is an important clinical sign in the diagnosis of EA, but usually it is not apparent until the third trimester 14. The rate of polyhydramnios in mothers who give birth to a newborn with EA ranges from 20 to 30% 14–16. In a recent study, it was reported that EA associated with chromosomal or structural anomalies was associated with greater occurrence of polyhydramnios with a rate of 53% 17. In a recent cohort comprising 307 new esophageal cases, polyhydramnios was present in 53.5% of the cases 18. The 17.2% rate of polyhydramnios in the present study was lower than those reported previously 11,14–16. This may be explained by the relatively low number of prenatally diagnosed cases included in our series than have been included in previously published series. Although not statistically significant, another finding in our series was that children with LGEA were associated with a greater occurrence of polyhydramnios than patients with SGEA (27.8 vs. 14.8%).
Prenatal diagnosis of EA before the third trimester is difficult and the prenatal detection rate of EA is low 18. It was reported that prenatal detection was possible in 9.2 and 24% of cases, respectively 19,20. In a recent report comprising 23 registries, it was found that prenatal detection rates of EA varied by registry from more than 50% to less than 10% of cases, and it was also claimed that the prenatal detection rate increased from 26 to 36.5% over the last two decades 21. In the current study, the rate of prenatal diagnosis of LGEA was much higher than that of SGEA (P<0.05), a finding similar to a recent report 11. The ratio of 22.2% in the prenatal diagnosis regarding LGEA patients in our study is comparable to the literature, whereas that of patients with SGEA (2.5%) is lower than that reported 18,21. We do not have an explanation for this finding, which demands more aggressive attempts to conduct prenatal diagnosis. As regular antenatal check-up programs increase, it is anticipated that the prenatal detection rates of EA will increase in the future.
The ratio of male to female was 1.06 in our study. In 15 EUROCAT registries covering 1 546 889 births for EA, 62% of the cases were male 22. Other studies have reported either a small or no excess of male patients 9,23,24. In a most recent report, the sex ratio (male/female) was found to be 1.3 10. We found a sex ratio of 1.57 for LGEA in favor of male patients, a result similar to that found in the literature 25. In contrast, our finding of a sex ratio of 0.98 against male patients among children with SGEA is unlike that of Bianca and Ettore 25, who reported a ratio of 1.5 in favor of male patients. Although there is no definite explanation for this finding in the present study, variability of causes linked to genetics and the gene–environment interaction may be involved.
The current study did not show differences between the groups with regard to the gestational age and the age at diagnosis and treatment. This finding is consistent with that of the studies by Bagolan et al. 6. In contrast, studies by Al-Shanafey and Harvey 12 and Lopes and Botelho 2 demonstrated that patients with LGEA were born at an earlier gestational age. In a most recent cohort, children with LGEA had a lower birth weight than children with SGEA 11. Although children with LGEA had a lower birth weight and a high rate of prematurity compared with patients with SGEA in this study, we did not find any statistically significant difference between the groups, a finding unlike the study by Lopes and Botelho 2. Dissimilarities in the present series concerning the birth weight and prematurity may be attributed to discrepancies pertaining to the demography of the study population.
Concerning the type of delivery in our study, most of the patients in both groups were delivered by cesarean section. This finding is consistent with that of the study by Chang et al. 15, who reported that 44 out of 72 babies with EA were born through cesarean section. In contrast, in a study by Aslanabadi et al. 26, it was found that patients with LGEA but not SGEA were delivered more often by a cesarean section. We speculate that maternofetal factors might have been involved in the type of delivery in these patients. Further comprehensive investigations on the choice of the delivery type in these patients may give clear knowledge on this issue.
Full or partial VACTERL spectrum defects have been reported in 10–67% of the neonates with EA 27–32. Similar to this range, we found VACTERL association in 67% of the children with EA in our study. The prevalence of non-VACTERL-type anomalies in patients with EA ranges from 20 to 70%, and our rate of 9% for non-VACTERL-type anomalies is lower compared with previous studies 27,33–35. Our study revealed that the prevalence of associated anomalies including VACTERL type were comparable between the groups. This finding is consistent with that of the study by Lopes and Botelho 2. Concerning the number of ribs, two of the patients in the LGEA group had 13 pairs of ribs, whereas none of those with SGEA had such a skeletal anomaly. Also, both groups in the current study showed similarities concerning the prevalence of associated non-VACTERL-type anomalies. This finding is different from that of study by Aslanabadi et al. 26, which reports more frequent non-VACTERL-type anomalies in patients with LGEA. Variations in our series may be due to genetic and geographic differences pertaining to our study population.
The other noticeable finding of this study regarding the type of EA according to the Gross classification was that type-C was more prevalent in patients with SGEA and type-A was more frequent in children with LGEA. Although there are conflicting reports, this finding has been considered as a description of LGEA by some authors in that there may be an association between the absence of TEF and LGEA 12. Apart from the type of EA, both groups of patients did not show any difference with regard to the discrepancy in the diameter of atretic esophageal ends.
As may be expected, both the duration of ventilatory support and the length of hospital stay were found to be higher in children with LGEA compared with those of patients with SGEA in the current study (P<0.05). In a previous study, the median length of hospital stay for patients with LGEA was found to be 83 days, a finding higher than that of ours 36. In a recent cohort, the median durations of ventilation and of first hospital stay were 3 and 22 days, respectively 11. Nevertheless, infants with LGEA spend a long time in hospital and need more time under ventilatory support compared with those with SGEA. Complications and long-term problems are frequently seen in patients with EA during treatment. The incidence of early and late complications during treatment and the follow-up duration did not differ between the two groups in the current study. Although of a limited time period, the follow-up duration in our study was comparable to that reported previously 14. Long-term multidisciplinary follow-up seems to be beneficial in these patients 37,38.
Nevertheless, there are limitations in this report. The retrospective design and the limited number of patients could impact the validity of the data. Moreover, eight different surgeons managed our patients. This could bring up individual variations between the surgeons with different degrees of experience and technical skills. However, our discrimination of LGEA and SGEA was objective considering a gap of three VB as a cutoff value. We believe that using uniform criteria to measure the gap length in prospective studies involving more patients may allow us to compare results from different institutions. With this approach, it may be possible to shed light on this issue, providing fundamental information.
Most of the demographic parameters studied were similar between the two groups of patients in this series. However, the frequency of prenatal diagnosis was more common in patients with LGEA. Most of the patients in both groups were delivered by a cesarean section. Concerning the type of EA according to the Gross classification, type-C was more prevalent in patients with SGEA and type-A was more frequent in children with LGEA. The duration on ventilatory support and the hospital stay were higher in children with LGEA. Further prospectively designed population-based registries are necessary and may be beneficial in revealing the true demographic differences between patients with SGEA and LGEA.
Conflicts of interest
There are no conflicts of interest.
1. Aslanabadi S, Jamshidi M, Tubbs RS, Shoja MM. The role of prophylactic chest drainage in the operative management of esophageal atresia with tracheoesophageal fistula. Pediatr Surg Int 2009; 25:365–368.
2. Lopes MF, Botelho MF. Midterm follow-up of esophageal anastomosis for esophageal atresia repair: long-gap versus non-long-gap. Dis Esophagus 2007; 20:428–435.
3. Brown AK, Tam PKH. Measurement of gap length in esophageal atresia: a simple predictor of outcome. J Am Coll Surg 1996; 182:41–45.
4. Hands LJ, Dudley N. A comparison between gap length and Waterston classifications as guides to mortality and morbidity after surgery for esophageal atresia. J Pediatr Surg 1986; 21:404–406.
5. Hollands CM, Lankau CA Jr. Preoperative home care for esophageal atresia: a survey. J Pediatr Surg 2000; 35:279–281.
6. Bagolan P, Iacobelli BdB, De Angelis P, di Abriola GF, Laviani R, Trucchi A, et al.. Long gap esophageal atresia and esophageal replacement: moving toward a seperation? J Pediatr Surg 2004; 39:1084–1090.
7. Foker JE, Linden BC, Boyle EM, Marquardt C. Development of a true primary repair for the full spectrum of esophageal atresia. Ann Surg 1997; 226:533–543.
8. Boyle EM Jr, Irwin ED. Primary repair of ultra-long gap esophageal atresia: results without a lengthening procedure. Ann Thorac Surg 1994; 57:576–579.
9. David TJ, O’Calaghan SE. Oesophageal atresia in the South west of England. J Med Genet 1975; 12:1–11.
10. Oddsberg J, Jia C, Nilsson E, Lagergren J. Influence of maternal parity, age, and ethnicity on risk of esophageal atresia in the infant in a population-based study. J Pediatr Surg 2008; 43:1660–1665.
11. Sfeir R, Bonnard A, Kehn-Dunlop N, Auber F, Gelas T, Michaud L, et al.. Esophageal atresia: data from a national cohort. J Pediatr Surg 2013; 48:1664–1669.
12. Al-Shanafey S, Harvey J. Long gap esophageal atresia: an Australian experience. J Pediatr Surg 2008; 43:597–601.
13. Duong HT, Hoyt AT, Carmichael SL, Gilboa SM, Canfield MA, Case A, et al.. Is maternal parity an independent risk factor for birth defects? Birth Defects Res A Clin Mol Teratol 2012; 94:230–236.
14. Brantenberg A, Blaas HGK, Haugen SE, Eik-Nes SH. Esophageal obstruction–prenatal detection rate and outcome. Ultrasound Obstet Gynecol 2007; 30:180–187.
15. Chang EY, Chang HK, Han SJ, Choi SH, Hwang EH, Oh JT. Clinical characteristics and treatment of esophageal atresia: a single institutional experience. J Korean Surg Soc 2012; 83:43–49.
16. Seo J, Kim do Y, Kim AR, Kim DY, Kim SC, Kim IK, et al.. An 18-year experience of tracheoesophageal fistula and esophageal atresia. Korean J Pediatr 2010; 53:705–710.
17. de Jong EM, de Haan MA, Gischler SJ, Hop W, Cohen-Owerbeek TE, Bax NM, et al.. Pre- and postnatal diagnosis and outcome of fetuses and neonates with esophageal atresia and tracheoesophageal fistula. Prenat Diagn 2010; 30:274–279.
18. Stoll C, Dott B, Alembik Y, Roth MP. Evaluation of routine prenatal diagnosis by a registry of congenital anomalies. Prenat Diagn 1995; 15:791–800.
19. Sparey C, Jawaheer G, Barrett AM, Robson SC. Esophageal atresia in the Northern Region Congenital Anomaly Survey, 1985–1997: prenatal diagnosis and outcome. Am J Obstet Gynecol 2000; 182:427–431.
20. De Vigan C, Goujard J, Vadovar V, Uzan S. Management of the fetus with a correctable malformation in Paris maternity units: evolution 1985–1994. Fetal Diagn Ther 1997; 12:216–220.
21. Pedersen RN, Calzolari E, Husby S, Garne EEUROCAT Working Group. Oesophageal atresia: prevalance, prenatal diagnosis and associated anomalies in 23 European regions. Arch Dis Child 2012; 97:227–232.
22. Depaepe A, Holk D, Lechat MF. The epidemiology of tracheo-esophageal fistula and aoesophageal atresia in Europe. Arch Dis Child 1993; 68:743–748.
23. Keckler SJ St, Peter SD, Valusek PA, Tsao K, Snyder CL, Holcomb GW III, et al.. VACTERL anomalies in patients with esophageal atresia: an updated delineation of the spectrum and review of the literature. Pediatr Surg Int 2007; 23:309–313.
24. Fraser C, Baird PA, Sadovnick AD. A comparison of incidence trends for esophageal atresia and tracheoesophageal fistula, and infectious disease. Teratology 1987; 36:363–369.
25. Bianca S, Ettore G. Isolated esophageal atresia and perinatal risk factors. Dis Esophagus 2003; 16:39–40.
26. Aslanabadi S, Ghabili K, Rouzrokh M, Hosseini MB, Jamshidi M, Adl FH, et al.. Associated congenital anomalies between neonates with short-gap and long gap esophageal atresia: a comparative study. Int J Gen Med 2011; 4:487–491.
27. de Jong EM, Felix JF, Deurloo JA, van Dooren MF, Aronson DJ, Torfs CP, et al.. Non-VACTERL-type anomalies are frequent in patients with esophageal atresia tracheoesophageal fistula and full or partial VACTERL association. Birth Defects Res A Clin Mol Teratol 2008; 82:92–97.
28. Botto LD, Khoury MJ, Mastroiacovo P, Castilla EE, Moore CA, Skjaerven R, et al.. The spectrum of congenital anomalies of the VATER association: an international study. Am J Med Genet 1997; 71:8–15.
29. Weaver DD, Mapstone CL, Yu PL. The VATER association. Analysis of 46 patients. Am J Dis Child 1986; 140:225–229.
30. Chittmittrapap S, Spitz L, Kiely EM, Brereton RJ. Oesophageal atresia and associated anomalies. Arch Dis Child 1989; 64:364–368.
31. Spitz L. Esophageal atresia. Lessons I have learned in a 40-year experience. J Pediatr Surg 2006; 41:123–128.
32. Teich S, Barton DP, Ginn-Pease ME, King DR. prognostic classification for esophageal atresia and tracheoesophageal fistula: Waterston versus Montreal. J Pediatr Surg 1997; 32:1075–1079.
33. Khoury MJ, Cordero JF, Greenberg F, James LM, Erickson JD. A population study of the VACTERL association evidence for its etiologic heterogeneity. Pediatrics 1983; 71:815–820.
34. Yang CF, Soong WJ, Jeng MJ, Chen SJ, Lee YS, Tsao PC, et al.. Esophageal atresia with tracheoesophageal fistula: ten years of experience in an institute. J Clin Med Assoc 2006; 69:317–321.
35. Czeizel A, Telegdi L, Tusn dy GCzeizel A, Telegdi L, Tusn dy G. VACTERL association. Multiple congenital anomalies 1988. Budapest: Akademiai Kiad; 247–280.
36. Maheshwari R, Trivedi A, Walker K, Holland AJ. Retrospective cohort study of long-gap oesophageal atresia. J Paediatr Child Health 2013; 49:845–849.
37. Gottrand F, Sfeir R, Coopman S, Deschildre A, Michaud L. Outcome of children with repaired oesophageal atresia. Arch Pediatr 2008; 15:1837–1842.
38. Rintala RJ, Sistonen S, Pakarinen MP. Outcome of oesophageal atresia beyond childhood. J Pediatr Gastroenterol Nutr 2011; 52Suppl 1S35–S36.