Intestinal atresia is one of the most common causes of intestinal obstruction in neonates, with reported frequency ranging from 0.4 to 3.1/10,000 live births (1–5). The finding of intestinal atresia is associated with both short- and long-term morbidity and mortality. In addition to short-term morbidities related to postoperative complications, these patients are at an increased risk for long-term complications and mortality from a variety of causes. Although the short-term mortality for children with intestinal atresias has been steadily improving over time, particularly in the developed world, the rates of long-term morbidities have been increasing in recent decades (1,6).
The predominant cause of late complications experienced by these infants is intestinal dysfunction and prolonged dependence on parenteral nutrition (PN) (6–8), resulting in cholestasis and liver dysfunction. In addition, the PN-dependent patients are at an increased risk for sepsis, thought to be caused by prolonged central-line use, and bacterial translocation caused by bacterial overgrowth in the abnormally functioning intestine (9). Moreover, the patients’ risk of complications is thought to be related not only to the surgical loss of portions of the intestine but also to the dilation and functional impairment of the remaining bowel (10).
In the past, studies have addressed some of the implications of the diagnosis of intestinal atresia, but they have typically focused on the perioperative and short-term sequelae experienced by these infants (6,11,12). Only a few studies of long-term outcomes have been performed, and almost all of those are from the 1990s or earlier, when treatment protocols differed from the ones used today (1,7,13).
At our institution, the intestinal rehabilitation program (IRP) was first implemented in August 2007, with the goal of decreasing long-term morbidity and mortality of patients with short bowel syndrome and other causes of intestinal failure. The philosophy of the program is to use multiple therapies to promote intestinal adaptation and the eventual weaning of patients off PN. The IRP optimizes the care of these patients through strict dietary management, parent education support, and prevention and treatment of acute and chronic complications as previously described (14–16).
Almost 5 years after the start of the IRP, our goal was to evaluate the outcomes of patients with atresias treated at our hospital before the program was in place, and assess how they may have changed with the arrival of the new program. In addition, these data would allow us to compare our progress with that of other institutions that have reported outcomes in similar groups of patients in recent years (1,6).
Thus, we chose to analyze the most recent decade's cohort of intestinal atresia patients, to determine what effect, if any, participation in an intensive IRP has on the health outcomes of these children. The focus of this study is the group of 20 children who were treated at our institution between July 2000 and June 2010 for prolonged PN dependence resulting from intestinal atresia. We first present background data for all children who were treated for intestinal atresia at our institution during that time period, noting the patients’ characteristics and outcomes at the time of neonatal intensive care unit (NICU) discharge. Subsequently, we evaluate the children who continued to be PN dependent at the time of NICU discharge, discuss their final outcomes and present health status, and compare the differences in their management before and after implementation of the IRP.
The study was approved by the institutional review board of Children's National Medical Center. The medical records of all patients treated at our institution for intestinal atresias between July 2000 and June 2010 were then reviewed. Patients with intestinal atresias were identified in a database maintained by the Department of Neonatology, as well as through records maintained by the IRP. These databases are text-based rather than International Classification of Disease-9 based, and patients were located by searching for the following diagnoses: duodenal atresia, ileal atresia, jejunal atresia, jejunoileal atresia, colonic atresia, anal atresia, small intestine atresia, large intestine atresia, intestinal atresia, or webs in any of the above-listed locations. Patients were excluded from the study population only if their preliminary diagnosis was rule-out intestinal atresia, but it was not confirmed by subsequent evaluation. No patients who truly had intestinal atresias were excluded because of associated diagnoses, but major comorbidities were noted. The information recorded for each patient included date of birth and gestational age, growth measurements, laboratory values correlated with liver function (bilirubin, albumin, coagulation parameters, and platelet count), bowel length after repair of initial atresia, duration of PN dependence and percentage of energy needs derived from PN, as well as coexistent diagnoses and late morbidities, such as repeat surgeries or hospitalizations, and central-line infections and replacements. For any patient who died, the age and cause of death were noted. Numerical data were shown using mean and standard deviation or median. Student t test was used to determine statistical significance for data that were normally distributed, and the Mann-Whitney test was used to compare data that were not normally distributed. Paired t test was used for data derived from the same patients before and after an intervention. Unpaired t test was used to compare the 6 non-IRP patients and their 14 counterparts enrolled in IRP.
Our program consists of 2 gastroenterologists, who take care of hospitalized IRP children, as well as liver and small bowel transplant patients. One of these doctors also studies the IRP outpatients, whereas the other sees outpatients with liver disease. Two general gastroenterologists also help in taking care of IRP inpatients, and covering the service 30% of the time. The program also has 2 physician's assistants, 2 nurse coordinators, 1 social worker, 1 case manager, and 1 dietitian. We also have an exclusive inpatient service with staff trained and dedicated to these complex patients. Psychology, pharmacy, and the pain team support the IRP service as well. Finally, we have a partnership with the transplant group at Georgetown University, which can transplant any patients who fail rehabilitation, with a plan for us to continue studying these patients once they are stable posttransplant. All surgical procedures other than transplants are done by the general surgery service, but the lengthening procedures and bowel reconstruction are mainly performed by 1 surgeon.
A total of 118 patients with intestinal atresias were identified as having been treated at our hospital during the time period between July 2000 and June 2010. Forty-eight of these children (41%) were boys, and 70 (59%) were girls. Thirty-two patients (27%) were early preterm, defined as being born before 34 weeks’ gestation, 47 (40%) were late preterm (34–36 6/7 weeks), and 38 (32%) were full term (37 weeks or older) (17). The most common site of intestinal atresias for all patients was the duodenum (53 patients), followed closely by jejunum (47 patients). Sixteen patients (14%) had atresias at ≥2 different sites in the intestine. The most common diagnosis associated with intestinal atresia was trisomy 21, which was noted in 14 children. All 14 children had atresias of the duodenum, and 1 of the 14 had an additional site of atresia in the ileum. Other conditions frequently noted in association with intestinal atresias included vertebral anomalies, anal atresia, cardiac defects, tracheoesophageal fistula, esophageal atresia, renal and radial anomalies, limb defects (VACTERL) association (9 patients), tracheoesophageal fistula (6 patients), annular pancreas (6 patients), gastroschisis (4 patients), and cystic fibrosis (2 patients). From the initial group of 118 patients, 6 died before NICU discharge. All 6 deaths were caused by coexisting cardiopulmonary abnormalities and were not directly attributable to the intestinal atresia. An additional 2 patients died shortly after NICU discharge: 1 because of postdischarge onset of necrotizing enterocolitis totalis, and 1 from complications of cystic fibrosis. Ninety patients remained alive and well and off PN beyond NICU discharge.
Although the majority of patients successfully weaned off PN support during their NICU stay, 20 of 118 patients (17%) remained PN dependent beyond NICU discharge, for a mean of 88.5% of their energy needs. These 20 patients are the main focus of our study, and all of the data presented below pertain only to them. In this group of 20 children, 7 were born early preterm, 10 late preterm, and 3 after full-term gestation. Almost all of these patients had jejunoileal atresias, with the following distribution of atresia sites: 2 duodenal, 14 jejunal, 10 ileal, and 1 colonic (7 of the children had both jejunal and ileal atresias). After the initial surgical intervention, the median intestinal length for 15 of these 20 patients was 22.5 cm (range 11–70 cm; intestinal length was unknown for 5 of 20), and 10 of 20 no longer had an ileocecal valve.
Six of the 20 patients did not enroll in an intensive IRP, because it was not available at that time, and were managed by neonatologists and general gastroenterologists. Two of the 6 patients were born early preterm, 3 were late preterm, and 1 was full term. Their median birth weight z score was −1.47 (range −5.8 to 0.11, uncorrected for gestational age). One child had jejunoileal atresia, 1 had colonic atresia, and there were 2 patients each with duodenal and jejunal atresias. The median bowel length was 33.5 cm (range 12–54 cm) for 4 of these 6 patients, and unknown for the other 2. Three of 6 (50%) no longer had an ileocecal valve. Three of the 6 had their full colon, and the other 3 (50%) had their colon partially resected. The maximum direct bilirubin reached a mean of 23 mg/dL (range 12.3–37.2) for this group of patients. They were receiving an average of 92% of energy from PN at the time of NICU discharge. Only 1 of these 6 patients (who was later transplanted) underwent a bowel-lengthening surgery, which was done via a serial transverse enteroplasty (STEP) procedure at 7.5 years of age. Two of the 6 patients died at 10 months of age from worsening complications of liver failure. Another 2 patients received combined liver and small bowel transplants after a prolonged course of PN dependence, with recurrent infections and worsening liver function, and are now alive and well, but require immunosuppressive medication for life. Two additional patients were transferred to a transplantation center with worsening liver failure pending organ availability, and subsequently died without receiving transplants.
The remaining 14 of the 20 PN-dependent patients were enrolled in the IRP, which was started in August 2007 at our institution. Five of these 14 patients were born before IRP initiation, and initially received conventional management, then were enrolled in the program once it started. The other 9 patients were enrolled in the IRP from early infancy. Five of the 14 IRP children were born early preterm, 7 late preterm, and 2 full term. This distribution of gestational ages was not significantly different from the non-IRP group. The median z score for the birth weights of these 14 children was −1.175 (range −3.75 to 2.8, uncorrected for gestational age), which was not significantly different from the birth weights of the non-IRP patients (comparing to both the 6 PN-dependent non-IRP patients [P = 0.31], and the larger group of all 118 patients with intestinal atresias [P = 0.25]). Eleven of these 14 children had jejunoileal atresias, 2 had only jejunal atresias, and the remaining one had an ileal atresia. None had atresias of the duodenum or colon. The median bowel length was 23 cm for 11 of the patients (range 11–70 cm, P > 0.5, no significant difference compared with the bowel length of the 6 non-IRP patients), and unknown for the other 3. Seven of 14 patients no longer had an ileocecal valve (50%, same as for the 6 non-IRP patients). Seven of 14 had their full colon, and the other 7 had their colon partially resected (50%, same as for non-IRP patients). Eleven of the 14 IRP patients had hyperbilirubinemia, with a mean highest direct bilirubin of 7.5 mg/dL (range 2.3–10.1), which was significantly lower than the peak bilirubin of the 6 non-IRP patients receiving PN (P < 0.001). Eight of the 14 patients underwent ≥1 intestinal lengthening operations. Six of the patients were treated with intestinal lengthening by the STEP procedure once only, 1 patient had the STEP procedure twice, and 1 underwent Bianchi lengthening followed by a STEP procedure. The average age at first surgery was 10 months. At the time of entry into the program, the mean energy requirement from PN was 87% for this group of patients, which was not significantly different from the 6 non-IRP children. At the most recent follow-up, the mean energy from PN had decreased to 10%, P < 0.001 (Fig. 1A, Table 1). Eleven of the 14 patients (78.5%) fully weaned off PN. The median time of weaning PN was 7 months (range 1–45 months). The remaining 3 PN dependent patients now require PN for only 47% of their energy needs, down from 100% at the time those 3 patients entered the IRP (Fig. 1B). These 3 patients have been PN dependent for an average of 12 months at the time of analysis.
Five of the 14 PN-dependent patients had already left the NICU by the time of IRP initiation and enrolled in the IRP at a later age. Thus, we were able to compare the status of these 5 children before and after participation in the IRP. Their characteristics are presented in Table 2. Of note, these children, who were transferred into the IRP at a median age of 19 months and with a mean PN dependence of 76%, were able to wean off PN since enrollment in the program.
In addition to the large improvement in their PN requirements, the 14 IRP patients with a history of intestinal atresias have also seen significant improvement in their laboratory values. The most striking of these is direct bilirubin, which was elevated in 11 of 14 patients at the time of entry into the IRP, with a mean value of 5.4 mg/dL among all 14 patients, and 7.5 mg/dL among the 11 children with hyperbilirubinemia (range 2.3–11 mg/dL). Notably, all 11 patients resolved their hyperbilirubinemia during an average of 12 weeks in the IRP, while still receiving PN. At the most recent measurement, the average level of direct bilirubin in all 14 patients was down to a normal value of 0.15 mg/dL (range 0–1 mg/dL, P < 0.001) (Fig. 2A). Similarly, 9 of 14 patients initially had hypoalbuminemia, with a mean albumin of 1.8 g/dL among the 14 children, which resolved in 7 of 9 patients during an average of 36 days. The remaining 2 patients have shown great improvement, with albumin that is only mildly low at 3.2 g/dL at this time. Overall, average albumin level has increased from 1.8 to 3.4 g/dL, P < 0.001 (Fig. 2B). The patients’ international normalized ratio (INR) and platelets also improved over time. Although 13 of the 14 patients initially had an elevated INR, with a mean of 1.6 for the entire group, 7 of these 13 patients normalized their INR during an average of 202 days, and the remaining 6 patients improved, with an average INR of 1.24 noted at most recent measurement, P < 0.05 (Fig. 2C). Likewise, even though 13 of the 14 patients initially had thrombocytopenia, with a mean platelet count of 117,000 cells per microliter for the entire group of 14, 12 of them normalized during an average of 15 days, and the last patient improved, with a recent platelet count of 147,000 cells per microliter, and a group average of 298,000 cells per microliter at most recent visit (P < 0.001) (Fig. 2D).
The 14 patients enrolled in the IRP also showed significant improvement in their growth parameters over time. Their average z scores were approximately −2 for both height and weight at the time of NICU discharge. With close nutritional follow-up and careful management, their z scores for both measurements improved over time, approaching 0 for height (P < 0.05) and 1 for weight (P < 0.01) (Fig. 3).
In the group of patients enrolled in the IRP, who have been studied for a minimum of 1 year and a maximum of 7 years at this point (4 of those years within the IRP itself), none of the patients have died or required a transplant. This is in contrast (P < 0.001) to the 6 patients who were treated outside the IRP, 4 of whom died at ages ranging from 10 to 14 months, and 2 received intestinal and liver transplants, 1 at 19 months and 1 at 8 years of age (Fig. 4).
With modern advances in medical care, such as improved surgical techniques and PN, the survival of neonates with intestinal atresias improved dramatically in the 20th century. As more of these infants were able to survive, long-term morbidity began to play a more important role because children with more severe disease were often unable to resume enteric feeding. With the survival of short bowel patients, who required prolonged use of PN, the resulting complications such as sepsis and liver failure were seen more frequently. Standard care and prolonged PN use in patients with short bowel or intestinal failure typically led to either death or liver and intestinal transplantation.
In view of this dismal prognosis, our institution implemented the IRP in 2007. The goal of the IRP is to facilitate bowel adaptation and enteral autonomy of the PN-dependent children, and thus avoid chronic complications and need for transplantation. The approaches to managing complications include using antibiotic prophylaxis for bacterial overgrowth, blood cultures and prompt antibiotic treatment when needed, ethanol or antibiotic locks for high-risk patients, and evaluation and monitoring for immunodeficiency. Surgical interventions, such as stoma closure, stricturoplasty, enteroplasty, and tapering or lengthening procedures, offer great therapeutic benefits as well. The main indications for surgical intervention in the IRP for children with short bowel syndrome are failure to progress in enteral feedings, a decrease in the previous level of enteral tolerance, life-threatening complications such as intestinal failure–associated liver disease, and recurrent septic episodes with enteric bacteria in a child with a dilated bowel. Some patients with intestinal atresia have severe dysmotility owing to a mismatch in intestinal diameter before and after the site of obstruction. In these situations, patients are managed in the IRP using ostomy in continuity to decompress the small bowel but continue using the colon, improving tolerance of enteric feeding (16).
The patients enrolled in the IRP and their PN-dependent counterparts who were unable to enroll in the program were similar in gestational age and intestinal length after initial repair of atresias. In addition, there were no associated diagnoses that disproportionately affected the non-IRP or the IRP group. One difference was the location of the patients’ atresias. PN-dependent patients in the IRP era had only jejunal and/or ileal atresias, which are normally more difficult to manage and lead to more complications, whereas atresia sites in the non-IRP patients were evenly distributed throughout the intestine (although the patients with duodenal and colonic atresias in the non-IRP group did have associated congenital anomalies, namely esophageal atresia, venous abnormalities, and omphalocele, respectively). Because jejunal and ileal atresias have been reported (6,7) to have a higher likelihood of leading to prolonged PN dependence than duodenal and colonic atresias, the fact that they are overrepresented in the group with better long-term outcomes suggests that the IRP is effective in reducing PN dependence and associated morbidity and mortality, even in the patients with more severe atresias.
Another significant difference between IRP and non-IRP patients who were dependent on PN is the degree of cholestasis, with significantly higher peak levels of direct bilirubin noted in the non-IRP patients. Given that the patients enrolled in IRP had a similar bowel length, gestational age, and comorbidities when compared with non-IRP patients, the most likely reason for their lower levels of cholestasis is not a difference in the underlying disease process, but intensive nutritional management and prevention of complications. Factors that contribute to the reduction in hyperbilirubinemia in IRP patients are close control of infections, lipid minimization, and redistribution (providing a week's worth of lipids for 3 days out of each week, rather than continuous daily lipid infusion), as well as aggressive dietary management with precise control of the metabolic balance, and prompt and effective treatment of complications. The patients who were born after implementation of the IRP and admitted to our NICU were followed by consultants from the IRP team starting soon after their first surgery, to maximize enteral nutrition and minimize the use of PN to the extent possible. After the resolution of postoperative ileus, these children were started on continuous enteral feeding and gradually advanced as much as tolerated, with careful supervision of their nutrient and fluid balance, as previously described (16).
A difference in operative management that likely also contributed to improved long-term outcomes in the IRP group is more aggressive use of intestinal lengthening surgeries. Unlike the pre-IRP group, in which only 1 patient underwent intestinal lengthening, and not until age 7.5 years, the majority of IRP patients were treated with this type of surgery, and at much younger ages (10 months on average), leading to improved tolerance of enteral nutrition before the liver was subjected to irreversible damage from prolonged use of PN.
Based on the data presented here, we conclude that our results compare favorably with those of other institutions (1,7,11). Mortality was noted to be rare in our patient population, and almost always, because of comorbid cardiorespiratory abnormalities, rather than intestinal atresias. Several children did die or receive transplants because of complications of intestinal atresias, but these were patients who had not been able to participate in the IRP. For those who did enroll in the IRP, the results were extremely encouraging, with most patients weaning off PN completely and the remaining few progressively decreasing their PN dependence over time. Similarly, the group of patients enrolled in the IRP showed improvement in laboratory results, with normalization of bilirubin, and normalization or near-normalization of albumin levels, coagulation parameters, and platelet counts. These patients also demonstrated adequate growth in both weight and height, even as they required progressively smaller amounts of PN.
Our study has some limitations. First of all, it is a retrospective study using historical controls, and only information that had been previously recorded and is available through our hospital's medical records. In addition, the size of our population of PN-dependent patients is small, owing to the relative rarity of this outcome in children with intestinal atresias. Despite this, the findings of this study are unlikely to be caused by random differences stemming from small sample size, given the magnitude of the observed outcome differences between IRP and non-IRP patients.
We conclude that with programs such as the IRP, patients with short bowel syndrome and bowel dysfunction secondary to intestinal atresia can show improvement in liver functions and nutritional parameters and have the ability to discontinue PN, avoid need for transplantation, and improve their chances of survival.
1. Stollman TH, de Blaauw I, Wijnen MH, et al. Decreased mortality but increased morbidity in neonates with jejunoileal atresia; a study of 114 cases over a 34-year period. J Pediatr Surg
3. Cragan JD, Martin ML, Moore CA, et al. Descriptive epidemiology of small intestinal atresia
, Atlanta, Georgia. Teratology
4. Forrester MB, Merz RD. Population-based study of small intestinal atresia
and stenosis, Hawaii, 1986–2000. Public Health
5. Martínez-Frías ML, Castilla EE, Bermejo E, et al. Isolated small intestinal atresias in Latin America and Spain: epidemiological analysis. Am J Med Genet
6. Piper HG, Alesbury J, Waterford SD, et al. Intestinal atresias: factors affecting clinical outcomes. J Pediatr Surg
7. Dalla Vecchia LK, Grosfeld JL, West KW, et al. Intestinal atresia
and stenosis: a 25-year experience with 277 cases. Arch Surg
8. Burjonrappa SC, Crete E, Bouchard S. Prognostic factors in jejuno-ileal atresia. Pediatr Surg Int
9. Duro D, Kamin D, Duggan C. Overview of pediatric short bowel syndrome
. J Pediatr Gastroenterol Nutr
10. Tongsin A, Anuntkosol M, Niramis R. Atresia of the jejunum and ileum: what is the difference? J Med Assoc Thai
11. Prasad TR, Bajpai M. Intestinal atresia
. Indian J Pediatr
12. Sato S, Nishijima E, Muraji T, et al. Jejunoileal atresia: a 27-year experience. J Pediatr Surg
13. Escobar MA, Ladd AP, Grosfeld JL, et al. Duodenal atresia and stenosis: long-term follow-up over 30 years. J Pediatr Surg
14. Torres C, Sudan D, Vanderhoof J, et al. Role of an intestinal rehabilitation
program in the treatment of advanced intestinal failure. J Pediatr Gastroenterol Nutr
15. Torres C, Vanderhoof J. Chronic complications of short bowel syndrome
. Curr Paediatr
16. Torres C. Duggan CP, Gura KM, Jaksic T. Medical and nutritional management. Clinical Management of Intestinal Failure
. Boca Raton, FL:CRC Press; 2012. 107–116.
17. Engle WA. A recommendation for the definition of “late preterm” (near-term) and the birth weight-gestational age classification system. Semin Perinatol
Keywords:© 2013 by European Society for Pediatric Gastroenterology, Hepatology, and Nutrition and North American Society for Pediatric Gastroenterology,
intestinal atresia; intestinal rehabilitation; short bowel syndrome