Magnetic Compression Anastomosis (Magnamosis) for Functional Undiversion of Ileostomy in Pediatric Patients

Introduction

Magnamosis forms a compression anastomosis using a pair of self-aligning magnetic Harrison rings. The magnets' specially engineered geometry ensures reliable anastomosis formation and healing before passage of the coupled rings. ¹ Although magnetic compression anastomoses have been previously reported,^2,3 only recently has a device been engineered to optimize the geometry of the mating surfaces. After extensive testing in animal models, the device has been approved by the Food and Drug Administration (FDA) for first-in-human testing (IDE: G130046). A Phase I clinical trial is currently underway in adults for intestinal anastomosis, and the device has been successfully used to reconstruct small intestine during complex urologic reconstructions. ⁴ We now report a novel procedure to functionally undivert the fecal stream with magnamosis from a previously created loop ileostomy in pediatric patients.

Materials and Methods

Technique

Parental and institutional approval was obtained after detailed description of the procedure.

The magnamosis FUN technique involves introducing a Harrison ring through each stomal limb 5 cm deep from the skin under general anesthesia with X-ray guidance to check for a correct magnet alignment. Magnets are each tied with nonabsorbable sutures that exit the stoma and are then tied to each other externally (Fig. 1). The device is removed when patency is detected.

OPEN IN VIEWER

FIG. 1.

Magnamosis undiversion technique: A side-to-side small bowel ostomy is demonstrated. Sutures fix both sides of the proximal and distal bowel loops, so that mesentery is confined to one side, rather than between the two loops of the bowel. The magnetic anastomosis is created by placing two Harrison rings, tied to two nonabsorbable sutures, ∼5 cm deep from the skin. Both sutures are tied together outside of the stoma. After anastomosis formation, the Harrison magnets pass through the distal loop and can be retrieved with the string.

Results

In both cases, placing and mating the magnets took less than 20 minutes, and there were no complications. They were performed under general anesthesia with X-ray guidance. Enteral feeding was initiated 24 and 6 hours postoperatively, and passage of stool through the anus occurred by the fourth and fifth days, respectively. Magnets were removed 14 and 15 days postoperatively, with no evidence of leak or any other complication (Fig. 2).

OPEN IN VIEWER

FIG. 2.

Removal of the mated magnets through one of the stomal limbs, done at the office.

In case 1, enteral nutrition was advanced and parenteral nutrition discontinued. Chemotherapy proved successful and the extrahepatic tumor disappeared, so the patient underwent cadaveric liver transplantation with a choledochal anastomosis. The magnetic anastomosis was left in situ and was closed by stitching shut the “chimney” opening. During follow-up, the patient presented with symptoms of bowel obstruction. At exploratory laparotomy, the obstruction, secondary to adhesions, was found to be well proximal to the ileal magnamosis, which remained patent. At 5 months posttransplantation and stoma closure, he remains tumor free, continues with chemotherapy, and is tolerating exclusive enteral nutrition.

In case 2, after the FUN procedure, the ostomy output decreased from 4 to 1.25 L/day. She started having 15 highly acid stools a day, so intense preventive local perineal measures were undertaken. Three months later, as the stools improved in number and quality, thanks to dietary adjustments and high-dose loperamide, the stoma was taken down (Fig. 3). She had a favorable outcome without burns in the perineal area or infections at the stoma site.

OPEN IN VIEWER

FIG. 3.

Ostomy takedown in case 2. Magnamosis is shown 5 cm deep from the stomal chimney.

Discussion

The magnamosis device has been tested in over 60 pigs in a variety of gastrointestinal sites and has proven that it generates stronger anastomoses than handsewn or stapled ones.^1,5–7 After FDA approval, five adult patients have had successful magnetic compression anastomoses using the magnamosis device during bladder augmentation procedures. All had side-to-side, and functional end-to-end small bowel anastomoses without complications. ⁴ In this study, we describe a novel procedure: FUN of a loop ileostomy created by magnetic compression anastomosis (Fig. 4).

OPEN IN VIEWER

FIG. 4.

Steps of FUN procedure. (a) Create ostomy of small or large bowel. (b) The ostomy may be a loop or a double barrel. (c) If possible, we recommend tacking the antimesenteric or the lateral aspects of the stoma limbs together for a length of 5 cm deep so that mesentery will not be trapped between the magnets. (d) Magnets are deployed into each stomal limb 5 cm deep into the skin. (e) Magnamosis is created. One or both of the magnets may be tied and fixed externally so that the mated magnets come out through the stoma. (f) Magnets are retrieved through the stoma once magnamosis is formed. (g) The bowel contents tend to pass through the magnetic anastomosis and the stoma opening acts as a pop-off valve if distal passage meets resistance. (h) Stoma opening is closed (both handsewn or stapled as preferred). FUN, functional undiversion.

These are the first reported cases of magnetic undiversion of an ostomy in either humans or animals. This procedure can be compared with others designed to allow distal bowel functionalization like Bishop-Koop or Santulli procedures. All these techniques have the advantage of permitting distal passage of content while maintaining a “pop-off valve” if the bowel does not function properly. In our practice, we sometimes opt for functional ostomies in patients in whom we think closure might have complications, so magnamosis is an interesting option to consider in these cases. In the first patient, we did the FUN procedure only a month after the stoma creation, and we did not trust the large bowel to be completely patent. On the other hand, it became necessary to consider refunctionalization of the excluded bowel (i.e., defunctionalization of the ostomy) to optimize the patient's poor nutritional status. Magnetic compression anastomosis became a very attractive option since we could increase bowel absorptive surface while keeping a pop-off valve, and without requiring tissue dissection and surgical handling of the ostomy. In the months after the magnetic undiversion, fecal matter tended to pass through the anastomosis rather than out of the ostomy and allowed for resumption of enteral nutrition. In the second patient, FUN allowed for a gradual refunctionalization for a transient short bowel syndrome caused by extensive resection of rectum, colon, and distal ileum. It was a less invasive way to facilitate adaptation to her new intestinal absorptive condition.

These are also the first pediatric cases of magnamosis. We have proved the procedure is feasible in children, and we had no complications. Although the 23 mm magnets were appropriate for use even in a 4-year-old child, it is unlikely that younger children are amenable for those magnets in terms of diameter. We have developed a variety of magnamosis sizes from 8 to 30 mm diameter to accommodate essentially any surgical indication.

Although studies in swine of the magnamosis device, the anastomoses were formed and the rings excreted after 5–7 days on average, it appears that it takes longer to pass the mated magnets in these first human cases, about 12–15 days. These time lapses have proven the same both in children and in adults. We are exploring the use of stronger magnets to achieve more rapid anastomosis formation.

Regarding technique, if magnamosis undiversion is being considered, the surgeon should tack the antimesenteric serosal surfaces of the stomal limbs together when constructing the diverting loop ostomy. This preventive measure would eliminate the possibility of having mesentery trapped between the two limbs when the magnets are placed. However, failure to do this should not be limiting, since the rings are applied manually, and their final location can be determined by the surgeon by palpation and X-ray. Although in our cases we deployed the magnets in the operating room with general anesthesia to achieve adequate abdominal muscle relaxation, we think bedside deployment of the magnets may be possible if the patient is relaxed, since the procedure is painless.

We conclude that Magnetic FUN is a safe, minimally invasive procedure that gradually refunctionalizes the excluded distal bowel after previous diverting ostomy. These are the first pediatric cases of undiversion using the magnamosis device.