Intestinal failure (IF) is associated with severe impairment of intestinal digestive-absorptive functions, of transepithelial electrolyte transport, or of both, and children with IF require parenteral nutrition (PN) for survival and growth (1). An increasing number of primary intestinal diseases are now recognized as being responsible for IF. These include short bowel syndrome, structural abnormalities of the enterocyte, inflammatory or allergic bowel diseases, autoimmune enteropathy, and disorders of intestinal motility (2). In the last few years, novel specific causes of IF have been described and their molecular mechanisms have been identified, although in a substantial number of cases, the cause of IF remains unknown. (3,4). Some of the diseases leading to IF may be reversible, whereas others are permanent or even progressive and are associated with a poor outcome. Because of better knowledge in the field, specifically progress in clinical nutrition techniques, life expectancy of children with IF has progressively increased (5–7). Yet, the severity of this condition, the associated complications, and the poor quality of life of children require better approaches to both short- and long-term problems, including intestinal transplantation (1,8–10). Although children are often seen in reference centers, optimal management of children with IF is hampered by the low number of cases seen in each individual center. Cooperation among centers in charge of these children generally is far from optimal and there is a lack of standard diagnostic and therapeutic protocols, including those for nutritional management.
In 1997, a collaborative network of centers involved in the care of children with IF was formally instituted by the Italian Society of Pediatric Gastroenterology and Hepatology with the aim to improve the management of children with IF (11). The Italian National Network for Intestinal Failure included all major Italian centers of pediatric gastroenterology. Specific diagnostic protocols were produced and specific advanced diagnostic procedures that were available in individual reference centers were provided to children enrolled in the whole country. Diagnostic procedures included identification of structural enterocyte abnormalities, such as proteoglycan or laminin deficiencies through immunohistochemistry. Individual reference centers for histologic analysis, immunohistochemistry studies, and electron microscopic examination were selected to examine all cases enrolled. Furthermore, reference centers for surgery and intestinal transplantation also were identified. The activity of the Network was reported annually during operative meetings, and data on all enrolled children were included in a database and made available to all centers participating the Network.
The aims of this study, based on the Network activity, were to provide epidemiologic data on IF, with insights into the natural history and cause-related outcomes. Based on these results, a classification system of IF was proposed and validated. Furthermore, we made available data suitable for designing appropriate intervention strategies to manage this complex condition.
PATIENTS AND METHODS
The Italian National Network for Pediatric Intestinal Failure was instituted as an open collaborative group of pediatric gastroenterologists, nutritionists, and surgeons seeing children with primary IF. The Network activity report was made available and updated on the Italian Society of Pediatric Gastroenterology and Hepatology web site (http://www.sigep-italia.org) and in the bulletin of the Society. Moreover, centers of pediatric surgery and nutrition, known to be in charge of PN in children, were contacted directly to share information and to set up collaborative work.
Children (0–14 years of age) with IF seen between January 1997 and December 2001 were enrolled. A consensus-based definition of IF was achieved among members of the Network and intestinal failure was defined as a clinical condition resulting from a primary intestinal disease for which PN, providing at least 75% of total calories for not less than 4 weeks or at least 50% of total calories for not less than 3 months, is needed. Exclusion criteria were: established extraintestinal cause, neoplastic diseases, and primary or secondary immune deficiencies. Children with inflammatory bowel diseases were excluded also.
Data were collected through web-provided enrollment and annual follow-up forms, using a numerical code for each patient. All forms were returned to the coordinating center for evaluation, validation, and statistical analysis. Items included cause of IF, child age at last follow-up, age at onset of symptoms, age at enrollment, duration of PN, and outcome.
Age at onset of symptoms was defined as the age at which initial intestinal signs or symptoms that were retrospectively linked to IF or to its cause were recorded. Age at enrollment was defined as the age at which PN was started. The outcomes were evaluated at the last follow-up and were identified as follows: progressing towards intestinal sufficiency, dependance on total or supplemental PN, death directly or indirectly associated with IF, and intestinal transplantation. Intestinal sufficiency was defined as full and permanent weaning from PN.
Patients were grouped according to etiopathogenesis of IF. Five major groups were identified: short bowel syndrome (SBS); disorders of intestinal motility (MD), including chronic intestinal pseudo-obstruction and aganglionosis; primary structural defects of the enterocyte (SED), including microvillous inclusion disease and tufting enteropathy; multiple food intolerance (FI); and autoimmune enteropathy (AIE). Children who could not be classified in any of the identified groups or in whom the cause was not defined were included in a sixth group defined as others or unknown, and their data were included in the analysis of general features of children with IF.
The diagnostic approach was similar to that previously reported and was based on accepted standard criteria (11–13). A progressive systematic collaboration with diagnostic reference centers was achieved, thereby increasing the spectrum of etiologic investigations. Several diagnostic procedures were centralized in individual reference laboratories, including intestinal histologic and ultrastructure analysis, antienterocyte antibodies, intestinal hormones, and other specific investigations that were indicated in selected cases. In addition, a bank of biologic samples (stools, serum, intestinal biopsies) was started.
The vast majority of children on home PN were provided with parenteral feedings, peristaltic pumps, and software for parenteral nutrition by a private company (Baxter, S.p.A., Trieste, Italy). Nutrient intakes and supplements were prescribed by reference centers.
Data on age at onset of symptoms, age at enrollment, and duration of PN were expressed as mean ± standard error of each etiologic group and were compared with the averages of other children with IF.
The percentage of children reaching intestinal sufficiency was calculated as a function of time for each etiologic group, and curves of permanence on PN were constructed. Statistical differences between the curves were analyzed by log-rank test.
Epidemiologic Features and Etiologic Pattern of Intestinal Failure
Nine centers participating in the Network enrolled 109 children with primary IF in 5 years. Thirty children were already in follow-up at the beginning of the study, whereas 79 were prospectively enrolled. The number of new cases per year was quite constant: 16 in 1997, 17 in 1998, 18 in 1999, 14 in 2000, and 14 in 2001.
The age at presentation, the age at start of PN, and the duration of PN and of follow-up are reported in Table 1. Thirty-seven percent of children were started on PN at 1 month of age, 24% started between 1 and 3 months of age, 22% started between 3 and 12 months of age, and 17% started after the first year of age. The total cumulative follow-up was 457 years, and the mean duration was 50.3 ± 46.3 months per child (range, 2–252 months).
Causes of IF are listed in Table 2. Short bowel syndrome was the most frequent cause of intestinal failure, with 48 children enrolled, including 5 children with congenital short bowel and 43 who underwent surgery for SBS. Eight children with tufting enteropathy and six with microvillous inclusion disease were included in the SED group. Motility disorders included 11 children with chronic intestinal pseudo-obstruction and 5 with intestinal aganglionosis. Ten children were diagnosed as having multiple FI because they responded with severe diarrhea whatever they were fed. Diagnosis of FI was confirmed by challenge with cow's milk and with other dietary antigens in all cases. A number of more rare diseases, which could not be classified in any of the major causes, are listed in Table 2. A girl who was receiving laxatives from her mother was eventually diagnosed as Munchausen by proxy. In seven children, no cause was identified.
The features of patients in the five major etiologic groups were evaluated. The age of onset of symptoms showed a broad variability (Fig. 1). Overall, the mean age of onset was 8.4 ± 27.2 months. Several children had diarrhea immediately or shortly after birth. The earliest onset was observed in children with SED (range, 0–15 months), with all but two cases presenting in the first 2 months of life. Tufting enteropathy had a slightly higher mean age of onset than microvillous inclusion disease, and two of eight children had symptoms becoming evident at 5 and 15 months of age, respectively. Short bowel syndrome had the broadest age at onset among the causes considered (range, 0–192 months), reflecting the primary causes of intestinal surgery. Food intolerance and MD presented at older ages (ranges, 3–34 and 0–124 months, respectively). Finally, AIE had the highest mean age of onset among all groups (range, 1–121 months;Fig. 1).
The age of PN onset, was closely related to the age at onset of symptoms (Fig. 1). The interval between the onset of symptoms and the need for PN is a marker of a rapidly progressive disease and showed a broad variability in relation to the cause. The shortest interval between the onset of symptoms and the need for PN was observed in children with SED. Children with MD had a prolonged course of intestinal disease before being started on PN, and this was even longer in children with AIE (Fig. 1).
The combined analysis of the duration of PN and the time of weaning from PN provided a comparative evaluation of the natural history of IF in relation to the cause. A short duration of PN may be regarded as a measure of the reversibility of IF, or vice versa, of the risk of an early poor outcome. The overall mean duration of follow up was 35.1 months at the last follow-up (range, 1–252 months), but it showed a broad range among the etiologic groups (Fig. 2). The shortest duration of PN was observed in children with FI, indicating an early reversibility of IF. Structural enterocyte defects were the most severe conditions with a mean duration of PN of 60.5 ± 46.9 months (Fig. 2).
Overall, 42% of children could be weaned permanently from PN, including 38% who underwent full and permanent recovery, whereas 4% were still receiving enteral nutrition at the last follow-up. In contrast, a poor outcome was reported in 6% of children (Table 3). Four children (3.7%) underwent intestinal transplantation (Table 3).
The chance of gaining intestinal sufficiency largely depended on the primary disease leading to IF (Fig. 3) and was significantly different, related to individual etiology (P < 0.0001). The time needed to reach intestinal sufficiency was also related to individual cause (Fig. 3). All children with FI experienced prompt and full remission of the disease, whereas none of the children with SED could be weaned from PN (Fig. 3). The other three main causes showed a progressive trend toward achievement of intestinal sufficiency. However, the probability to reach such a benign outcome was more pronounced for AIE (70%) and less for MD (25%). Short bowel syndrome had an intermediate pattern, with 42% of recovery. In only 2 of the 14 children with SBS who remained on PN for more than 2 years was intestinal sufficiency ultimately reached.
The severity of SED was confirmed by the high rate of death (Table 3). Also, four children with SBS died (8.3%). One patient with MD had a poor outcome, whereas no fatal events were reported in children with multiple FI and AIE (Table 3). The terminal events of children with IF were lack of vascular accesses, liver failure, and central line infections.
Most children in the others or unknown group had more benign outcomes. No deaths were recorded in this group.
Four children underwent intestinal transplantation, including two with tufting enteropathy, one with microvillous inclusion disease, and one with SBS. They received the graft after 102, 42, 42, and 160 months of PN, respectively. Their follow-up was 90, 60, 47, and 178 months, respectively. Three eventually reached intestinal sufficiency. One child with tufting enteropathy received a second intestinal graft after rejection of the first but eventually died.
The aim of the Italian National Network for pediatric intestinal failure was to design new strategies for a condition in which a collaborative and multidisciplinary approach is essential. By selecting reference centers for diagnostic evaluation, a double advantage was obtained: gaining comparative data to prevent observer-related bias and improving individual specific expertise in fields where a high degree of subjective skills are needed, such as with histologic and electron microscopic analysis.
Over 5 years, we enrolled 109 children with IF, covering a total of 457 years of follow-up. We performed a cross control of the incidence and prevalence of SBS in Italy by accessing the National Institute of Statistics (http://www.istat.it) and reviewing the national disease codes of SBS, the most common cause of IF. Data obtained in this way indicated that children enrolled in the Network were not less than 75% of total children with IF in Italy.
Short bowel syndrome accounted for almost half of IF cases. Other causes included motility disorders, multiple food intolerance, tufting enteropathy, autoimmune enteropathy, and microvillous inclusion disease, with similar incidence. Other, more rare conditions, were also identified. The etiologic pattern was substantially changed in comparison with previously reported series, with an increased rate of primary enterocyte abnormalities and a decrease of infections (11–13). However, two children with severe and protracted Cytomegalovirus-induced diarrhea were seen, supporting a recently described infectious cause of IF (14).
The network approach allowed a reduction in the number of children whose disease was of unknown cause. In the present study, 6.5% of children remained without a cause identified, compared with 8% to 45% of children in other series (11–13,15).
The improved capacity to obtain an etiologic diagnosis and the availability of longitudinal data allowed us to investigate the outcome in relation to the primary intestinal cause. The natural history of children with IF, in relation to their cause, showed three distinct patterns. The first was the typical pattern of SED with an early onset of diarrhea and a high death rate, related to microvillous inclusion disease rather than tufting enteropathy. The second pattern of IF was a transient condition that ultimately resulted in either early or delayed weaning of PN. The former was observed in children with FI on dietary manipulations, the latter in those with AIE on immunosuppressive therapy. The third pattern was associated with a stable and nonreversible IF. This was the result of MD, mainly chronic intestinal pseudo-obstruction, and was frequently observed in children with cerebral palsy. In these children, the course of the disease was persistent with little chance to reach intestinal sufficiency, as reported previously (16).
Short bowel syndrome encompassed all the three patterns but with a time-related prevalence, in that instability of water and electrolyte homeostasis and severe nutritional failure were observed early in the course of the disease, but when this phase was overcome, stability was usually achieved. Ultimately, intestinal sufficiency was obtained in approximately 40% of children. As reported in other studies, the outcome was associated critically with the length of residual gut and to early intestinal feeding (1,17–20).
Limited data are available to predict reliably the outcome of IF. Our data show that the probability of intestinal sufficiency had a cause-related and also a time-dependent pattern. For SBS, specifically, the chance of recovery decreased dramatically after 2 years of PN. The time of permanence in PN thus could be regarded as a major indication to therapeutic options such as intestinal transplantation.
In recent years, the outcome of children receiving an intestinal graft substantially improved because of better postsurgical management and more effective immunosuppression (8–10,21–25). However, late-onset complications of PN, such as cirrhosis, are now becoming evident (26).
Of the four children who received intestinal transplantation in our population, three had SED, and two of these three children ultimately discontinued PN. Two of the six children with microvillous inclusion disease died, similar to what was reported in previous series (27). We thus believe that children with congenital enterocyte defects should be referred for intestinal transplantation as early as possible.
The data in this series allowed an estimate of the number of IF patients with particular diagnoses who would potentially become candidates for intestinal transplantation. In our study, all children with SED and those with SBS on PN for more than 2 years appear to be potential candidates for intestinal transplantation. Thus, our gross estimate for intestinal transplantation is 20 children in 5 years, corresponding to a range of 3 to 5 children per year for IF secondary to primary intestinal diseases in Italy. The proportional estimate for European children would be 35 to 60 intestinal transplantations per year. In light of this estimate, it would be sufficient to have a maximum of two to four intestinal transplantation reference centers in Europe, to have an efficient management of this complex condition (9–10).
In conclusion, IF is the endpoint of several rare diseases of heterogeneous origin. The primary cause of IF is a major determinant of its outcome, and its early identification and proper management are crucial for planning therapeutic approaches, including intestinal transplantation (10).
We believe that the network model proved to be highly effective, and it may be extended at the European level with several advantages: 1) to obtain larger and comparable data on IF, on which the needs for health care should be based; 2) to upgrade the management ability in individual centers involved in the care of children; 3) to promote basic and applied research in the field; and 4) to improve the quality of life of children.
1. Goulet O. Intestinal failure in children. Transplant Proc 1998; 30:2523–5.
2. Walker-Smith JA. Intractable diarrhoea of infancy: a continuing challenge for the paediatric gastroenterologist. Acta Paediatr 1994; 83:6–9.
3. Murch SH, Winyard PJD, Koletzko S, et al. Congenital enterocyte heparin sulphate deficiency with massive albumin loss, secretory diarrhoea, and malnutrition. Lancet 1996; 347:1299–301.
4. Murch SH. The molecular basis of intractable diarrhoea of infancy. Baillières Clin Gastroenterol 1997; 1:413–40.
5. Dorney SF, Ament ME, Berquist WE, et al. Improved survival in very short small bowel of infancy with the use of long-term parenteral nutrition. J Pediatr 1985; 107:521–5.
6. Zlotkin SH, Stallings VA, Pencharz PB. Total parenteral nutrition in children. Pediatr Clin North Am 1985; 32:381–400.
7. Burnes JU, O'Keefe SJ, Fleming CR, et al. Home parenteral nutrition—a 3-year analysis of clinical and laboratory monitoring. J Parenter Enteral Nutr 1992; 16:327–32.
8. Jan D, Michel JL, Goulet O, et al. Up-to-date evolution of small bowel transplantation in children with intestinal failure. J Pediatr Surg 1999; 34:841–4.
9. Nishida S, Levi D, Kato T, et al. Ninety-five cases of intestinal transplantation at the University of Miami. J Gastrointest Surg 2002; 6:233–9.
10. Kaufman SS, Atkinson JB, Bianchi A, et al. Indications for pediatric intestinal transplantation: a position paper of the American Society of Transplantation. Pediatr Transplant 2001; 5:80–7.
11. Catassi C, Fabiani E, Spagnuolo MI, et al. Severe and protracted diarrhea: results of the 3-year SIGEP multicenter survey. J Pediatr Gastroenterol Nutr 1999; 29:63–8.
12. Guarino A, Spagnuolo MI, Russo S, et al. Etiology and risk factors of severe and protracted diarrhea. J Pediatr Gastroenterol Nutr 1995; 20:173–8.
13. Ventura A, Dragovich D. Intractable diarrhoea in infancy in the 1990s: a survey in Italy. Eur J Pediatr 1995; 154:522–5.
14. Fox LM, Gerber MA, Penix L, et al. Intractable diarrhea from cytomegalovirus enterocolitis in an immunocompetent infant. Pediatrics 1999; 103:E10.
15. Scott NA, Leinhardt DJ, O'Hanrahan T, et al. Spectrum of intestinal failure in a specialised unit. Lancet 1991; 337:471–3.
16. Heneyke S, Smith VV, Spitz L, et al. Chronic intestinal pseudo-obstruction: treatment and long term follow up of 44 patients. Arch Dis Child 1999; 81:21–7.
17. Sondheimer JM, Cadnapaphornchai M, Sontag M, et al. Predicting the duration of dependence on parenteral nutrition after neonatal intestinal resection. J Pediatr 1998; 132:80–4.
18. Wasa M, Takagi Y, Sando K, et al. Intestinal adaptation in pediatric patients with short bowel syndrome. Eur J Pediatr Surg 1999; 8:207–9.
19. Gambarara M, Ferretti F, Papadatou B, et al. Intestinal adaptation in short bowel syndrome. Transplant Proc 1997; 29:1862–3.
20. Andorsky DJ, Lund DP, Lillehei CW, et al. Nutritional and other postoperative management of neonates with short bowel syndrome correlates with clinical outcomes. J Pediatr 2001; 139:27–33.
21. Goulet O, Michel JL, Jan D, et al. Intestinal transplantation in pediatric patients: the European experience. Transplant Proc 1997; 29:1785–6.
22. Abu-Elmagd KM, Reyes J, Fung JJ, et al. Evolution of clinical intestinal transplantation: improved outcome and cost effectiveness. Transplant Proc 1999; 31:582–4.
23. Pinna AD, Weppler D, Nery J, et al. Intestinal transplantation at the University of Miami. Transplant Proc 2000; 32:1226–7.
24. Farmer DG, McDiarmid SV, Yersiz H, et al. Improved outcome after intestinal transplantation: an 8-year, single-center experience. Transplant Proc 2000; 32:1233–4.
25. Kaufmann SS. Small bowel transplantation: selection criteria, operative techniques, advances in specific immunosuppression, prognosis. Curr Opin Pediatr 2001; 13:425–8.
26. Kelly DA. Liver complications of pediatric parenteral nutrition-epidemiology. Nutrition 1998; 14:153–7.
27. Phillips AD, Schmitz J. Familial microvillous atrophy: a clinicopathological survey of 23 cases. J Pediatr Gastroenterol Nutr 1992; 14:380–96.
The Italian National Network for Pediatric Intestinal Failure
The following colleagues participated in the study design, discussion of the results, and provided cases (numbers indicated in parenthesis): M. Castro, M. Gambarara, A. Diamanti, Rome (n = 44); A. Barabino, P. Gandullia, Genoa (n = 20); P. Roggero, Milan (n = 16); E. Bruzzese, R. Berni Canani, A. Tramontano, Naples (n = 9); S. Martelossi, D. Faraguna, Trieste (n = 8); F. Indrio, Bari (n = 7); A. Campanozzi, S. Giovanni, Rotondo (n = 3); S. Amarri, Modena (n = 1); P. Lionetti, Florence (n = 1).
The following colleagues participated in the network with the following specific functions: E. Bazzigaluppi, E. Bosi (Milan), anti-enterocyte antibodies; R. S. Collins, G. Piana (Baxter, Rome), home parenteral nutrition; O. Goulet (Paris), intestinal transplantation; C. Marino (Genova), histologic analysis; M. Morroni (Ancona), electron microscopic analysis; D. Sordino, A. Settimi (Naples), short bowel syndrome revision and analysis.