The first attempt of TILT insertion was from complicated operation with situational complexities. The baby was born at a gestational age of 27 weeks and with a birth weight of 980 g. He underwent an initial small bowel resection with end ileostomy at 10 cm above ileocecal valve at 6 days after birth due to necrotizing enterocolitis perforation. He was referred to our clinic for an early closure of the ileostomy due to excessive water loss and ileostomy retraction. Aside from an adhesiolysis, the ileostomy closure was uneventful. Though the proximal lumen was two times greater than distal lumen, I made end to end anastomosis as usual. Unfortunately, during the 10 days postoperation, gradual aggravation and distension of the abdomen became pronounced without stool passage (Fig. 1A). We completed a second operation, creating a side to side anastomosis for decompression because the anastomosis was healthy and intact. However, this second operation did not alleviate the abdominal distension, and the bowel movement was not recovered. At 5 days post the 2nd operation, wound dehiscence occurred, along with the exposure of the peritoneum. A third operation was planned (Fig. 1B). Due to concerns of creating a stoma due to complications from the previous enterostomy, I attempted a TILT insertion. I did an ileocecectomy and inserted a 12F nasogastric tube via the anus. The tube was passed through the whole colon and anastomosis and placed 10 cm proximal to the anastomosis. The total length of the inserted nasogastric tube was 60 cm. Stool started draining well via the TILT, and abdominal distension was markedly improved. He was able to start feeding at 10 days post TILT operation and defecated via the anus. The TILT was removed 68-day postoperation due to uncertainties of this new procedure. He is doing well now without complications.

The selection criteria for the TILT procedure in stoma closure were limited to those patients showing narrow caliber of ileum and microcolon, predicting a severe luminal size discrepancy between proximal lumen and distal lumen on contrast studies. In cases where the diameter of the lumen in the distal ileum and colon was found to be less than one third of the proximal intestine during operation, we decided to attempt the TILT procedure. In addition, the TILT procedure was attempted on patients exhibiting prolonged postoperative ileus after intestinal anastomosis. If the anastomosis would be so far from ileocecal valve, or if the luminal discrepancy is not dominant, we do not do TILT procedure.

To prepare the area for a surgical drape, the entire abdomen—including the inguinal area—was disinfected. After trimming the ileostomy, we resected the ileum and made an end to end anastomosis with an interrupt suture; starting with the posterior wall of the anastomosis (Fig. 2A). In preparation for insertion, we made an additional side hole on the nasogastric tube (we have used both 10F and 12F nasogastric tubes with success) to prevent occlusion of the hole with stool. The tube was inserted via the anus (Fig. 2B) and routed through the rectum and sigmoid colon with the help of intra-abdominal guidance. The tube easily passed through the entire colon to the proximal side of the anastomosis. After the intestinal tube passed over the anastomosis, the anterior wall of the intestine was closed with an interrupt suture. (Fig. 2C) The tube was fixed in place at the buttock with one tagging suture and taped down. The infantogram post this TILT procedure is shown below (Fig. 3).

The external portion of the tube was connected with a closed bag, allowing intestinal contents to drain naturally. We frequently checked the patency of the intestinal tube by milking and gently aspirating procedures, thus preventing occluding of the lumen. Irrigation with a warm saline solution was utilized 3 days post operation when the lumen became occluded.

We retrospectively reviewed the medical records of infant who were treated using the TILT procedure between March 2012 and February 2016 at the Inje University Haeundae Paik Hospital and the Dong-A Medical Center—both located in Busan, Korea. From these medical records, data surrounding gestational age, birth weight, underlying disease causing exploratory laparotomy, the timing of the 1st and 2nd operations, the maintenance period of the intestinal tube, and the time to full enteral feeding was retrospectively collected.

A series of 10 children was enrolled in this study, comprising of 7 male and 3 female patients. Median gestational age was 27⁺⁵ weeks (range, 23⁺²–39⁺¹ weeks - Nine out of the ten babies were premature) and birth weight was 940 g (range, 615–2,400 g). The first operations were done at a median of 4.5 days (range, 0–33 days) with a body weight of a median of 835 g (range, 720–2,360 g). The operations were initiated due to necrotizing enterocolitis perforation (4 cases), isolated intestinal perforation (3 cases), meconium related ileus (1 case), congenital ileal volvulus (1 case), and ileal atresia (1 case). Loop ileostomies were completed for the premature babies with solitary intestinal perforation, meconium related ileus or necrotizing enterocolitis and congenital ileal volvulus. For the case of ileal atresia, I did a resection and anastomosis (Table 1).

The second operation for ileostomy closure was planned at a median of 130.5 days (range, 86–226 days) with a median body weight of 3,060 g (range, 2,400–5,000 g). Before the 2nd operation, we performed distal contrast enemas and observed the passage of contrast in the distal ileum and colon. All of these operations were uneventful. For the ileal atresia case, postoperative obstructions were not improved over 3 weeks. We performed the 2nd operations as a TILT procedure without additional small bowel resection and anastomosis.

Nine out of ten TILT procedures were well functioned and defecation via natural anus was observed in a median of 4.5 days (range, 3–12 days). Milk feeding began at a median of 6 days (range, 2–15 days) and enteral feeding of 60 kcal/kg/day was reached at a median of 10 days (range, 5–27 days) after the TILT procedure. The long intestinal tube was removed in a median of 14.5 days (range, 3–68 days) and the time to full enteral feeding was a median of 15.5 days (range, 8–47 days) postoperatively. In 1 case, the malfunctioning tube was removed 3 days after the TILT procedure. In this patient, intestinal obstruction was prolonged for 3 days longer after the tube was removed. This was relieved after gastrograffin enema. Nine of the patients recovered well and have had no episodes of ileus. One patient with severe bronchopulmonary dysplasia died from respiratory distress with sepsis after 2 months (Table 2).