Hirschsprung Disease Part I with Marc Levitt

The best treatment for Hirschsprung disease is irrigations using a large bore tube (20 French Foley) with warm saline, instilling 10-20 cc aliquots at a time and allowing fluid mixed with stool to drain back.

It is very rare that Hirschsprung disease is a surgical emergency, but if you don't irrigate and overcome the distal obstruction, it will become an emergency.

For rectal biopsy in suspected Hirschsprung disease, you must be at least 1 centimeter in from the dentate line. If you biopsy too close to the dentate line, everyone has an aganglionic segment there and you could get the wrong answer.

The pathologist must report both the absence of ganglion cells AND the presence of hypertrophic nerves. The absence of ganglion cells alone is not Hirschsprung disease - that could be a biopsy taken too low.

Hirschsprung disease has an immune component and the lining of the bowel (mucosa) is much more susceptible to bacterial translocation. Stasis in a Hirschsprung patient leads to bacterial translocation and a very sick baby from bacteremia.

If irrigations are not working and the baby is ill, you need to divert. The recommendation is to divert in the ileum rather than doing a leveling colostomy, because frozen section can be inaccurate, particularly as you move higher in the colon.

The Swenson operation involved full-thickness transabdominal dissection down below the peritoneal reflection. The operation was often done incorrectly with perirectal dissection done too wide, leading to fecal incontinence, urinary incontinence, sexual problems, and impotence.

The Suave operation used a mucosal dissection to keep the outer rectal wall intact and avoid injury. The original Suave would leave the colon coming through the anus for a week before doing the coloanal anastomosis; Dr. Scott Boley modified it to do everything in one stage.

The Duhamel operation leaves the original rectum in place, removes the aganglionic portion from the peritoneal reflection, and pulls ganglionic bowel in a retrorectal position, connecting the two lumens with a stapler.

Of the four classic procedures (Swenson, Suave, Duhamel, Rabine), only the Swenson actually leaves behind virtually no Hirschsprung tissue. The others leave behind outer rectal wall or original rectum. Many patients did well despite this because ganglionic bowel, if it's good, can overcome a lot.

Dr. Henrys in the Philippines pioneered primary pull-through operations (without initial colostomy) out of necessity because patients would never return to clinic due to social stigma against colostomies.

Keith Jorgeson applied laparoscopic skills to Hirschsprung disease, doing the colonic work laparoscopically with a Suave dissection from above. Luis de la Torre and Jack Langer developed the transanal approach, starting the dissection from below.

The current preferred approach is a transanal Swenson (full-thickness dissection). It is the purest operation, leaving behind no Hirschsprung except the very bottom just above the dentate line. If you find the right plane, it's elegant and bloodless.

Dr. Orvar Swenson, who recently died at age 105, maintained that the Swenson operation was good all along and people just weren't doing it right, which gave it a bad reputation.

Even Suave enthusiasts are making shorter and shorter cuffs over time. Jorgeson's original laparoscopic Suave recommended a 5 cm cuff; nowadays Langer and de la Torre do 1-1.5 cm cuffs, essentially approaching a Swenson.

Laparoscopy is appropriate for all cases to find the transition zone level. A transanal-only approach is only appropriate when you have a very reachable, obvious transition zone at mid-sigmoid that you can comfortably reach transanally.

There is significant morbidity from an overly aggressive transanal-only approach trying to reach the transition zone without going into the abdomen. If you're too aggressive transanally trying to visualize the dissection, you would have been better off with laparoscopic dissection.

Total colonic Hirschsprung patients present differently: the diagnosis isn't made right away, the contrast study is not typical, and irrigations don't go well. These patients need laparoscopy first to find the biopsy location.

The biggest technical problems are: (1) surgeons don't give themselves good exposure and start dissection too low, injuring or resecting the dentate line, or (2) they give very aggressive exposure and overstretch the sphincters.

Proper technique requires marking 1 cm proximal to the dentate line with a purple mark and placing 5-0 silk stitches circumferentially at that level. This preserves the anal canal, dentate line, plus an additional 1 cm of columnar epithelium before starting dissection.

By definition, you're leaving behind 1 cm of columnar epithelium that is Hirschsprung disease, plus the internal sphincter which has a problem with relaxation. However, good ganglionic bowel is able to overcome that, and the baby eventually learns to relax the sphincter.

The rectum doesn't really have a mesentery; the rectum's blood supply is intramural. As you get higher and reach the peritoneal reflection, that's when you start to see sigmoidal vessels where rectum transitions to sigmoid.

The anterior rectum frees up much quicker than the posterior rectum. Break into the peritoneal reflection anteriorly first. In transanal-only cases, you can pull the sigmoid out anteriorly, do a full-thickness biopsy, and send it while continuing the posterior dissection.

For laparoscopic biopsies through the umbilicus, a seromuscular biopsy (without violating mucosa) may show ganglion cells in the seromuscular layer but miss hypertrophic nerves in the submucosa. If using this technique, send a full-thickness biopsy later to confirm the level is appropriate.

Pathologists must report not only ganglion cells but also the quality of nerves with actual micron measurements. Nerves should be 40 microns or less. Anything bigger than 40 microns is transition zone bowel.

The concept of 'go 5 cm above the transition zone' is inaccurate. Transition zone is a spectrum - some are 10 cm, some are 3 cm. You need confirmatory biopsy with ganglion cell information and nerve quality measured in microns.

It's preferable to take the IMA, preserve the arcade, and have the left colon and sigmoid nice and straight down into the perineum. This makes for an easy-to-irrigate baby. Many patients have not had enough of a pull-through with the entire sigmoid loop still there, requiring redo to remove more.

For anything proximal to the splenic flexure, do colonic biopsies and an ileostomy and wait, because frozen section has been notoriously fraught with errors in those cases. There is no urgency. Another option is to take biopsies, quit without diverting, and return 3-4 days later for pull-through with permanent section results.

Postoperatively, wait until the belly is absolutely soft and flat with bowel function before feeding. Get an X-ray because abdominal distention is sometimes subclinical. This usually takes 3-4 days. Feeding too early risks the baby going home distended and returning with enterocolitis.

Routine postoperative management: at one month, check the anus with Hagar dilators (not finger). Most babies need calibration rather than true dilation. The stimulation of passing the Hagar has value to help the baby more successfully empty.

Flagyl is only used to treat enterocolitis, not routinely postoperatively. Give a pre-op dose of second-generation cephalosporin and maybe 2 post-op doses.

Routine irrigations are only done postoperatively if the baby develops significant distention, which is rare. However, for total colonic patients after ileoanal pull-through, send all patients home on irrigations for 3 months.

For hepatic flexure transition zone (rare), open the patient rather than laparoscopy. Take down the entire right colon, recognize the ileocolic vessel and how it feeds the vessel paralleling the right colon. Often need to take the right colic artery. De-rotate the colon so cecum is at the hepatic liver bed, then do pull-through down the right side of abdomen.

If bringing de-rotated colon down the left side of abdomen, must mobilize the ligament of Treitz so the mesenteric vessel is not draped across the third portion of duodenum, which can cause duodenal obstruction.