Keywords
calretinin stain - reoperation - enterocolitis - continence - aganglionosis
Introduction
Skip lesions have rarely been reported in patients with Hirschsprung disease (HD)
especially before the definitive pull-through procedure. We present a case of an infant
with HD with skip lesions known prior to their pull-through and managed with a pull-through
of the right colon.
Case Presentation
A full-term baby boy who had failed to pass meconium at 48 hours had persistent abdominal
distention and bilious emesis. An upper gastrointestinal contrast study was normal
but when contrast passed to the colon a concern for inspissated stool was entertained.
Due to persistent distention, he underwent exploratory laparotomy with enterotomy
(ileostomy, 24 cm from ileocecal valve) and terminal ileal washout. The baby was operated
on for presumed intestinal obstruction and no contrast enema was done due to significant
abdominal distention and to prevent delaying the surgery. At the time of surgery,
full-thickness sigmoid colon biopsies were obtained, with intraoperative frozen section
reporting presence of ganglion cells, but the tissue review on permanent was deemed
by the pathologists to be inadequate and no ganglion cells were seen. He initially
did well and was tolerating oral exclusively formula intake with about five spontaneous
bowel movements per day. Cystic fibrosis testing, both metabolic and genetic, was
negative. He was discharged home 28 days later. Four days after discharge, he presented
again with obstructive symptoms. He underwent a contrast enema that was suspicious
for long segment HD, microcolon, and the question mark-shaped colon ([Fig. 1]). Anorectal manometry showed an absent rectoanal inhibitory reflex (RAIR). Since
at patient's 22 weeks prenatal ultrasound, a left pelvic kidney with double ureteral
system was noted, a pelvic ultrasound was obtained. A pelvic ultrasound showed hydronephrosis
in a left pelvic kidney and a dimercaptosuccinic acid scan showed it had no function;
the right kidney was normal. A voiding cystourethrogram showed no vesicoureteral reflux.
Fig. 1 Repeat contrast enema concerning for long-segment Hirschsprung disease: microcolon,
the question mark-shaped colon, and the lack of features in an otherwise normal colon.
Rectal biopsy showed absence of ganglion cells in the submucosa, and loss of calretinin
expression in the lamina propria of the mucosa and no hypertrophic nerves. He underwent
a diverting ileostomy with colonic mapping with findings shown in [Fig. 2]. Notably, he had no ganglion cells in his cecum but had ganglion cells in the ascending
colon and hepatic flexure. The transverse colon and sigmoid were aganglionic. The
left colon had ganglion cells. This pathology was reviewed by two different pathology
departments and both agreed with the findings. This patient was then referred to an
HD referral center for another opinion, and the entire complete set of slides (966
sections) from the colonic mapping done in the first hospital were reviewed by a third
pathologist. After extensive discussion with the parents, the patient was taken to
the operating room for repeat colonic mapping as there was concern for the possibility
that the slides had been mislabeled. We repeated the mapping and confirmed that the
findings as originally determined were accurate ([Fig. 2]).
Fig. 2 Results of colonic mapping at the time of ileostomy creation, showing Hirschsprung
disease with skip lesions. (+) denotes presence of ganglion cells. (–) denotes absence
of ganglion cells.
With this ganglionic map in mind, we performed a transanal proctectomy, with transabdominal
(via laparotomy) excision of sigmoid, left colon, and transverse colon. The middle
colic vessels were divided, carefully preserving the right marginal vessels from the
ileocolic vessels to the point of good ganglion cells at the hepatic flexure with
the idea to preserve the ganglionic right colon (8 cm). Intraoperative biopsies from
hepatic flexure and cecum were done and ganglion cells were identified only in myenteric
plexus from hepatic flexure. We then performed a cecal resection, carefully preserving
the blood supply to the right marginal vessels from the ileocolic vessels. An ileocolic
anastomosis was performed (ileum to right colon with cecum now removed) and the right
colon was then derotated to perform the pull-through down the right pelvis ([Figs. 3] and [4]). Derotation has been described in adult literature extensively as Deloyers procedure.
It is safe and associated with good long-term results.[1] The diverting ileostomy was left in place. The patient underwent examination under
anesthesia 4 weeks postoperatively that showed a well-healed anastomosis and no stricture.
The parents were instructed to pass stool from the ileostomy into the distal segment
daily, and he was able to successfully control the stool out of his neorectum. He
then underwent ileostomy takedown 4 months later. On his 1 year follow-up (he is currently
2 years of age), the patient is stooling normally with Bristol stool chart type 4
and 5, 4 to 5 stools per day.
Fig. 3 Colonic vasculature and our plan for the resection and pull-through. The blue lines
mark the vasculature that will be taken down for the resection, derotation, and eventual
pull-through of the right colon.
Fig. 4 Cecal resection, extended left colectomy, ileocolic anastomosis, right colonic derotation
pull-through of the right colon. The last figure shows the final postoperative anatomy.
In the pathological study of the surgical specimen, the proximal margin (ileum) was
normal, with presence of ganglion cells in the submucosal and myenteric plexuses and
normal expression of calretinin in the mucosa. In the cecum, transverse colon, splenic
flexure, sigmoid colon, and rectum there were no ganglion cells in the plexuses and
there was loss of calretinin expression in the mucosa. In the ascending colon, hepatic
flexure, and descending colon, scattered ganglion cells were identified in both plexuses,
with focal expression of calretinin in the mucosa. The overall appearance of these
“skip areas” resembled “transition zones” and never showed a normal number or distribution
of ganglion cells ([Figs. 5A–C]).
Fig. 5 (A) Section of normal ileum (proximal margin) with ganglion cells in the myenteric plexus
(hematoxylin and eosin [H&E] left, marked by arrows) and normal expression of calretinin
in lamina propria of the mucosa (right, marked by arrows). (B) Section of aganglionic segment of the cecum demonstrating absence of ganglion cells
in the submucosal and myenteric plexuses (H&E left) and loss of calretinin expression
in lamina propria of the mucosa (right). (C) Descending colon with isolated ganglion cell in the myenteric plexus (H&E left marked
by arrows) and few nerve fibers in lamina propria stained with calretinin (right,
marked by arrows)
Discussion
HD is a congenital intestinal obstruction due to distal aganglionosis, defined by
the lack of ganglion cells in the myenteric and submucosal plexuses of the distal
intestine to the level of proximal ganglionic bowel. It is seen in 1 in 5,000 live
births and its management involves resection of the aganglionic bowel and pull-through
of normal bowel to restore functional continuity.[2]
[3] All definitive Hirschsprung operations serve this vital purpose, no matter which
type of procedure is selected.
Segmental aganglionosis (skip lesions) is an extremely rare entity in HD, and in fact
many doubt its existence. The first reported case of a skip HD lesion was published
in 1954.[4] Skip segment HD refers to normally ganglionated intestine, surrounded proximally
and distally by aganglionosis.[5] On the other hand, in zonal aganglionosis, an aganglionic segment is sandwiched
or interposed between segments of normally innervated or ganglionated bowel.[6]
[7] Some of these cases used the appendix to show skip lesions but this is an incorrect
concept as many appendices are aganglionic.[8]
The pathology of typical HD is sustained contraction of the aganglionated bowel segment
due to absence of ganglion cells of the myenteric and submucosal plexuses. The lack
of ganglion cells, derived from neural crest cells, is usually due to loss of coordination
of the survival, migration, proliferation, and/or differentiation of the progenitor
cells in the gastrointestinal (GI) tract. Generally, this occurs either due to arrest
in migration of neural crest cells to reach their final GI destination[9] or due to failure of proliferation, survival or differentiation after the migration
has already occurred.[10] Neither of these hypotheses explains the skip lesion concept that has led to skepticism
of its existence. Even though there has been a relative paucity of data in reporting
skip lesions, a review identified 23 reported cases in the literature.[11] One theory to explain the skip lesions is that neuroblasts, when migrating in a
craniocaudal direction during neural crest migration, cross the mesenteric border
to a more distal part of the intestine and as a result end up well ahead of the wavefront,
colonizing an area within the aganglionic segment.[12]
[13]
Most cases of skip lesion HD are total colonic aganglionisis (TCA) reported to be
92%, and 8% were rectosigmoid HD. Of the TCA cases, 41% had a skip segment in the
transverse colon, 27% in the ascending colon, 9% in the cecum, and 23% had multiple
skip segments.[11] Skip segment thus should be suspected in patients with TCA. In cases with TCA, colonic
mapping should be performed. In these cases, biopsies are performed in the rectum,
sigmoid colon, descending colon, splenic flexure, transverse colon, hepatic flexure,
ascending colon, and the cecum.[14] Patients with skip lesion HD are described to present with failure to pass meconium,
chronic constipation, bilious emesis, and abdominal pain. Diagnosis of HD is made
in most patients using suction rectal biopsy findings of aganglionosis, and hypertrophic
nerve trunks in the submucosa.[15] Calretinin immunohistochemistry can be used as an adjunctive method to diagnose
HD, because of the loss of the staining pattern of small nerve fibers of the lamina
propria, muscularis mucosae, and submucosa of aganglionic zones.[16] We perform anorectal manometry in patients over 1 years old for an evaluation of
constipation in which Hirschsprung's is a possibility or internal sphincter achalasia.
Typically, if the manometry shows an absent RAIR, we perform a rectal biopsy at the
time. We prefer to perform manometry awake; however, this is not always possible.
The RAIR can be determined even in an anesthetized patient.
It is pertinent to understand the vascular anatomy of the colon especially in patients
with skip lesion HD to allow reconstruction with maximum colonic salvage. Resection
of the colon is based on the arterial supply to its various anatomic divisions ([Fig. 3]). The superior mesenteric artery, via ileocolic and right colic artery, supplies
the cecum and the ascending colon respectively. These arteries then further give off
vasa recta that form arcades adjacent to the colonic wall. We took advantage of the
colonic anatomy and resected the cecum only by dividing the vasa recta, and preserved
the marginal artery from ileocolic artery to supply the eventual pull-through segment
and allowed preservation of the ganglionated right colon (8 cm).
It is possible that many cases of skip lesion HD have been missed and dismissed by
surgeons and pathologists over the years because the idea contradicts the well accepted
theory of neural crest migration in the gut, and the disturbance of migration in HD.
This condition has divided pediatric surgeons, with many only believing this phenomenon
after experiencing this condition in their own patients. We can confirm with this
report that skip lesion HD does exist.
Skip lesion HD is an extremely rare entity in HD. If known prior to the surgical intervention,
careful preoperative planning may help preserve functional colon in patients with
this unique condition.
