Search

BACKGROUND: Feeding difficulties and gastroesophageal reflux (GER) are major problems in severely neurologically impaired children. Many patients are managed with a simple gastrostomy, with or without fundoplication. Unfortunately, fundoplication and gastrostomy are not devoid of complications, indicating the need for other options in the management of these patients. METHODS: Since January 2002, seven patients (age range, 5-14 years) have been treated by creating a jejunostomy with the laparoscopic-assisted procedure. The procedure was performed using two 10-mm trocars. The technique consists of identifying the first jejunal loop, grasping it 20-30 cm away from the Treitz ligament with fenestrated atraumatic forceps, and exteriorizing it to the trocar orifice under visual guide. The jejunostomy was created outside the abdominal cavity during open surgery. At the end of the jejunostomy, the correct positions of the intestinal loop and feeding tube were evaluated via laparoscopy. RESULTS: Surgery lasted 40 min on average, the laparoscopic portion only 5 min. There were no perioperative complications; hospital stay was 3 or 4 days for all patients. At the longest follow-up (18 months), all patients had experienced a significant weight gain, with a high level of parental satisfaction. One patient died 1 year after the procedure of unknown causes. All the others are well, without complications or problems, and their parents are extremely satisfied with the improved quality of life of their children. CONCLUSIONS: Laparoscopic-assisted jejunostomy is a safe and effective procedure to adopt in neurologically impaired children with feeding problems and GER. This procedure solves these patients' feeding problems even if the reflux is not completely eliminated. We advocate the use of this procedure in neurologically impaired patients with feeding problems and reflux due to its overall practicability and because there is minimal surgical trauma. This technique is extremely safe because the surgeon is able to verify, at the end of procedure, the status of the jejunostomy from outside and inside the abdominal cavity. The improvement in the quality of life of these children after the jejunostomy seems to be the major advantage of this procedure.

Citrobacter species are often present in the stools of children and are generally considered a normal component of the intestinal microflora. Previous reports suggested that they might act as enteric pathogens. Aiming at defining the role of Citrobacter species in inducing diarrhea, we looked for their presence in the stools of 328 children with diarrhea and in 108 controls. Citrobacter strains were isolated from 46 patients (14%) and 7 controls (6.5%) (P less than 0.05). All isolates were tested for heat-stable (ST) and heat-labile (LT) enterotoxin. No LT-producing organisms were found. Three C. freundii strains, all isolated from children with diarrhea, elaborated an enterotoxin detected by the suckling mouse assay. A crude extract was prepared by acetone precipitation and a sequential ultrafiltration technique. The enterotoxin was heat stable, and its estimated molecular weight was between 2,000 and 10,000. Citrobacter enterotoxin was soluble in methanol and stable at acid and neutral pHs but not above pH 8, and its activity was destroyed by treatment with 2-mercaptoethanol. Citrobacter enterotoxin was inactive in the 18-h rabbit ileal loop test. All these characteristics closely resemble STa produced by Escherichia coli. The time course of Citrobacter enterotoxin-induced intestinal secretion in suckling mice was similar to that of E. coli STa. The enterotoxin produced by C. freundii cross-reacted with monoclonal antibodies raised against E. coli STa. These results suggest that C. freundii is capable of inducing diarrhea through the production of an E. coli-like STa, and its presence in the stools of patients with diarrhea should be considered as that of a possible etiologic agent.

Escherichia coli heat-stable enterotoxins (ST) are classified into STa and STb according to their physicochemical and biologic characteristics. STa induces diarrhea, activating the guanylate cyclase-cGMP system. ST-like enterotoxins can be produced by bacteria other than E. coli, including Klebsiella pneumoniae. A Klebsiella ST has previously been shown to share some chemical and immunologic characteristics with E. coli ST. Aiming to define better the nature of Klebsiella ST, we have screened 237 children with diarrhea and 179 controls for ST-producing Klebsiella, using the SMA. We detected 26 Klebsiella strains from patients, two of which were positive in the SMA, and 36 from controls, all negative for ST. A partial purification was performed using an acetone precipitation followed by ultrafiltration and gel filtration techniques. Klebsiella toxin was heat-stable, methanol-soluble, sensitive to mercaptoethanol, active at acid pH values, but not at pH greater than 8. The time course of Klebsiella toxin in the SMA resembled that of E. coli STa. Klebsiella ST caused reduced Na absorption and net Cl secretion in rabbit ileal mucosa mounted in Ussing chambers. It was found to increase the cGMP but not the cAMP concentration. Finally, Klebsiella ST did not react with anti-E. coli STa MAb in a competitive ELISA. We conclude that K. pneumoniae may induce diarrhea through the production of an STa similar but not identical to E. coli STa.

Summary. A cell-culture assay was used to detect toxins directly in stools from sporadic cases of infantile diarrhoea. Cytotoxins were revealed in 11 out of 58 samples from children with diarrhoea, nine of whom had no common enteric pathogens in their stools. A preliminary characterisation of the cytotoxins was obtained by neutralisation tests with clostridial antitoxins.