- Email: [email protected]
The pathology of infantile hypertrophic pyloric stenosis after healing
The Pathology of Infantile Hypertrophic Stenosis After Healing By Jean-Marie Vanderwinden, Hao Liu, Roberte Marc-Henri De Laet, and Jean-Jacques
Menu, Jean-Louis Vanderhaeghen
Pyloric Conreur,
Brussels, Belgium 0 Introduction: Infantile hypertrophic pyloric stenosis (IHPS) is a common surgical affection of unknown etiology. The muscular hypertrophy is known to resolve within a few months after pyloromyotomy (PM). The pathology of IHPS has been studied extensively at the time of PM, but the fate of the pylorus after healing remains unknown. Materials and Methods: We had the rare opportunity to study two pyloric biopsy specimens obtained 4 months and 2 years (respectively) after an uncomplicated PM for IHPS. They were compared with the initial specimen in one case, with 26 other specimens of IHPS, and with five normal controls. Immunohistochemistry using the avidin-biotin complex (ABC) system was performed for S-100 and nerve growth factor receptor, as markers for the enteric nervous system, and for the tyrosine kinase receptor c-kit. as a marker for the interstitial cells of Cajal (pacemaker cells). NADPH-diaphorase histochemim was performed as a marker for the neuronal enzyme nitric oxide synthase, which produces the inhibitory neurotransmitter nitric oxide. Resufts: In both cases of IHPS, after healing, the circular musculature was not hypertrophic. For all markers studied, the distribution appeared similar to that in the normal pylorus. In contrast, all specimens obtained at the time of PM displayed a severe reduction of the different markers in the hypertrophic musculature. Discussion: The pathological features observed in the circular layer in IHPS appear to resolve within a few months after PM. This suggests that the involvement of the enteric nervous system in IHPS might be milder than generally assumed. The etiology remains obscure, but our occasional observations may provide new insight into the pathophysiology of IHPS, and are in agreement with the excellent longterm clinical outcome for IHPS. Copyri9ht o 7996 by W.B. Saunders Company
From the Laboratoire de Neuropathologie et de Recherche sur les Neuropeptides, Fact& de Medecine et Departernent de Chirurgie Pediatnque, Hopital Universitaire des Enfants Reine Fabiola, Universite Libre de Bruxelles, Brussels, Be&m. Presented as a poster at the XLII Annual International Congress of the British Association of Paediatric Surgeons, Sheffield, England, July 25-28, 1995. J.M.V is Senior Research Assistant from the National Fund for Scientific Research (Belgium), 1995-1997. H.L. is a visiting scientist from the Department of Pediatric Surgery, Xian Children’s Hospital, Xian, People S Republic of China. Supported by Belgian Grants from the Fonds de la Recherche Scientifque Medicale (3.4582.94-97) Fondation Medtcale Reine Elisabeth (I 992-95) Ministere de la Politique Scientihque (Pole d ‘Attraction lnteruniversitaire 22, 1990-95), and National Lottery (1992, 1994). Address reprint requests to Jean-Mane Vanderwinden, Laboratoire de Neuropathologie et de Recherche sur les Neuropeptides, Faculte de Medecine, Campus Erasme CP 601, Universite Libre de Bruxelles, 808 route de Lennicc, B-1070 Brussels, Belgium. E-maik [email protected]. Copyright 0 I996 by U? B. Saunders Company 0022-3468/96/3111-0014$03.00i0 1530
INDEX WORDS: Pylorus, pathology, chemistry, immunohistochemistry, kit, interstitial cells of Cajal.
pyloric stenosis, NADPH-diaphorase,
histoc-
NFANTILE HYPERTROPHIC pyloric stenosis (IHPS) is a common surgical affection of infancy. Its diagnosis and treatment are now well standardized, and the outcome usually is benign.l The etiology of IHPS remains elusive,2 but ultrasonography has indicated that the muscular hypertrophy resolves within a few months after pyloromyotomy (PM).3 In patients operated on for IHPS more than 15 years earlier, the gastric emptying rate was normaL but in a recent study,5 persistent abnormalities of the serum gastrin level and mucosal somatostatin content were noted long term in children who had undergone PM, fueling the old debate on possible long-term sequelae of IHPS in terms of gastric function.6 The pathology of the enteric nervous system (ENS) in IHPS has been studied extensively at the time of PM, and the persistence of some abnormalities after clinical recovery might account for long-lasting problems. Unfortunately, the fate of the pylorus in IHPS after healing has not been documented, mainly because such specimens are not readily available. We had the rare opportunity to obtain pyloric biopsy specimens from two patients who had been operated on previously for IHPS at our institution, and we addressed the question of the pathology of the pylorus in IHPS after recovery.
I
MATERIALS
AND METHODS
Pyloric biopsy specimens were obtained from two patients, 4 months and 2 years (respectively) after PM for IHPS.
Case Reports This patient was a dysmature boy who had had an Patient A. uneventful operation for IHPS at 4 weeks of age. At 5 months of age he was readmitted, to the emergency ward, comatose and in severe shock. Immediate intubation and resuscitation were required. Parental anamnesis was confuse. The cardiorespiratoty condition was satisfactory, but a chest radiography showed multiple costal fractures, some with partial healing. The abdomen was extremely distended and painful. Emergency abdominal ultrasonography showed the fluid in the abdominal cavity, and several edematous, dilated intestinal loops, suggestive of intestinal vohulus. An emergency laparotomy was performed and showed internal contusions in the upper abdomen, associated with a retroperitoneat hematoma in the diaphragmatic region. The diagnosis of violent blunt trauma of the upper abdomen was established, and Journal
of Pedtafrtc
Surgery,
Vol31,
No 11 (November),
1996: pp 1530-l
534
PYLORIC
STENOSIS
AFTER
HEALING
parental attitude was seriously questioned. A left hemiparesia appeared on the second postoperative day. A cerebral computed tomography (CT) scan showed a right parietal fracture, bilateral subdural hematomas, and a zone of infarction in the right occipital region. The neurological signs abated subsequently. Despite the prophylactic administration of ranitidine since the admission of the patient, massive gastric bleeding occurred on the eighth postoperative day. The oozing of a large prepyloric stress ulcus could not be controlled by gastroscopy, and an emergency gastrotomy was required to stop the hemorrage. The lesion was located in the prepyloric region and a pyloric biopsy specimen was obtained at that time. The patient’s recovery was slow but eventually satisfactory. Severe parental abuse was estabhshed later. and appropriate measures were taken for follow-up. Patient B. This dysmature girl presented at birth with congenital stenosis of the lower esophagus. A segmental resection with end-to-end anastomosis was performed, without complication, on day 4 of life. At 6 weeks of age she was readmitted because of increasing vomiting and hematemesis. IHPS was diagnosed through echography, and a gastric ulcus was found during gastroscopy. Signs of colonic subobstruction appeared shortly after her admission, and congenital colonic stenosis was diagnosed. A PM and a segmental colonic resection with end-to-end anaostomosis were performed during the same laparotomy procedure. Recidives of hematemesis despite adequate antiacid therapy eventually led to the diagnosis of gastric duplication, which was resected at the age of 4 months. Severe gastroesophageal reflux persisted despite maximal medical therapy. and a Nissen fundoplication was performed when the patient was 2 years old. A pyloric biopsy specimen was obtained at that time because we routinely associate pyloromyectomy with the Nissen procedure. The patient’s development has been satisfactory since then. Although both these cases were complicated, both patients had classical uncomplicated IHPS. At the time of reintervention, they had no clinical sign of gastric emptying delay. The pylorus appeared normal during inspection and palpation. Full-thickness muscular biopsy specimens (with preservation of the mucosa and submucosa) were taken longitudmally, just outside the scar of the previous PM. to avoid fibrous tissue. Both patients had a normal, uncomphcated postoperative course.
Methods We compared the two specimens with a collection of 27 biopsies obtained at the time of PM for IHPS (including the initial specimen from patient A) and with five normal control specimens obtained at the time of autopsy from two newborns and three pediatric organ donors. The specimens were harvested and processed as described previously.’ Briefly, the tissues were hxed overnight in fresh 4% paraformaldehyde solution, cryopreserved in sucrose. embedded in Tissue-Tek OCT compound (Miles, Elkhart. IN), frozen in Z-methyl butane cooled on dry ice, and stored at -gO“C. Longttudinal sections (15 urn thick) were cut on a cryostat. mounted on slides coated with 0.1% poly-L-lysine (Sigma Chemical Co, St Louis. MO), and stored at -20°C until use. Several sections of each specimen were stained routmely with H&E. Immunohistochemical evaluation was carried out with commercially available kits using the avidin-biotin-complex (ABC) system (Vectastain ABC), according to the instructions of the supplier (Vector, Burlingame, CA). Differents kits were used, in accordance with the species of the primary antibodies employed. A mouse monoclonal antibody raised against the glial marker S-100 (reference no. SA2102, Affinity, Nottingham, UK; I ug/mL) and a mouse monoclonal antibody raised against the low-affinity receptor for nerve growth factor (NGFR) (reference no. 1198.637. Boeh-
1531
ringer Mannhein Biochemica, Germany; 5 ug/mL) were used to study the distribution of (respectively) the glial and neural components of the ENS.s9 An affinity purified polyclonal rabbit antiserum raised to a synthetic peptide corresponding to residues 961-976 within the C-terminal domain of the human c-ti receptor (reference no. SC39, Santa-Cruz Biotechnology, Santa-Cruz, CA; 0.4 ug IgG/mL) was used to study the distribution of the interstitial cells of Cajal (ICC), also known as the pacemaker cells of the gut.‘“~” Briefly, the sections were treated in 0.3% hydrogen peroxide in methanol, for 30 minutes, to block endogenous peroxydase activity. They were incubated in normal serum for 20 mmutes. incubated with the primary antibody diluted in normal serum for 3 hours, rinsed m phosphate-buffered saline (PBS) for 10 minutes, incubated with the biotinylated secondary antibody for 30 minutes, rinsed in PBS, and incubated with the ABC complex for 1 hour. ABC conjugated with horseradish peroxidase was used. The peroxidase activity was revealed for 10 to 15 minutes, at room temperature, with a solution containing 0.2 mg/mL 3,3’-diaminobenzidime (DAB) (Sigma), 2 mg/mL ammonium nickel sulfate (Fluka. Buchs, Switzerland), and 0.03% v/v Hz02 in 0.05 mol/LTris-buffered saline. Controls of the immunohlstochemical procedures. No staming was observed when the primary antibodies were omitted. The optimal working dilution was determined empirically for each antihody by serial dilutions. Liquid-phase preabsorption was carried out for the c-kit antiserum according to the instructions of the supplier (Santa Cruz). ie, overnight incubation at 4°C with a tenfold-by-weight excess (4 ug/mL) of the immunogenic peptide (Santa Cruz; reference no. SC39-P) abolishing the signal. NADPH-diaphorase histochemistry was performed (as described previously’) to localize the neuronal form of constitutive mtric oxide (NO) synthase. Briefly, the slides were incubated in the dark in a solution of 0.1 mol/L Tris HCI buffer containing 1 mmol/L NADPH (Sigma), 0.2 mmol/L Nitro Blue Tetrazolium (Sigma), and 10% dimethylsulfoxide, at 37°C for 75 to 90 minutes. Omission of NADPH was used as a negative control and resulted in the absence of staining.
RESULTS
IHPS at the Time of PM (Refer to Fig 1, panel A.) All the specimens studied (n = 27) displayed the typical pathological features of IHPS. Upon H&E staining, the circular muscle layer appeared markedly enlarged whereas the myenteric plexus and longitudinal layers appeared normal. Immunohistochemical evaluation of the ENS markers S-100 and NGFR (Figs A, and AZ, respectively) indicated a severe reduction of neural elements in the hypertrophic circular musculature. Only a few nerve fibers were present, and they appeared abnormally large. The ordinate alignment of neural elements along the axis of the smooth musculature present in the normal pyloric musculature was lacking in all these specimens. In contrast, in the myenteric plexus layer, the patterns of labeling apparently were normal. Histochemical evaluation of NADPH-diaphorase (Fig A,) showed that nerve fibers expressing NO synthase were absent in the hypertrophic circular musculature. However, NO synthase was expressed in nerves bundles and in ganglionic cells in the myenteric plexus. c-kit immunoreac-
1532
VANDERWINDEN
ET AL
PYLORIC
STENOSIS
AFTER
HEALING
tivity labeled very few ICC in the hypertrophic pylorus (Fig &). Only a few ICC were observed in the inner third of the circular layer, near the submucosal edge and at the proximal (gastric) nonhypertrophied side of the biopsies (data not shown). After Healing
(Refer to Fig 1, panel B.) The two specimens’ muscular layers and myentric plexus appeared normal with H&E staining. No fibrous tissue was present, confirming the peroperative impression that the site of biopsy was chosen properly, outside the scar of the previous pyloromyotomy. The pathological features of the two cases were similar. Contrasting with the pattern observed at the time of PM (panel A), S-100 and NGFR immunoreactivity labeled numerous elements of the ENS in the circular muscular layer (Figs B1 and BZ, respectively). These fibers were thin and ordinately arranged in the general direction of the smooth muscular fibers. NADPH-diaphorase histochemistry (Fig B3) indicated the presence of neuronal NO synthase in a large number of nerve fibers in the circular layer. The density and the pattern of distribution of these markers of the ENS appeared similar to those of the normal pylorus. c-kit immunoreactivity (Fig B4) showed that ICC were densely and homogenously distributed in the circular layer, in a pattern similar to that observed in the normal pylorus. DISCUSSION
The pathology of the ENS in IHPS has been studied extensively at the time of PM. Morphological alterations of the ENS” as well as various abnormalities of distribution of neuropeptides and neurotransmitters have been reported,13 including reduction of the neuronal NO synthase in the hypertrophic circular layer. NO is a potent inhibitory neurotransmitter (ie, inducer of smooth muscle relaxation), and the absence of the enzyme producing NO suggested that
Fig 1. Pathological aspect of infantile hypertrophic pyloric stenosis at the time of pyloromyotomy (panel A) and after healing (panel B). Labeling of the enteric nervous system with immunoreactiidty for S- 100 and NGFR (Al & 131, A2 & B2 respectively) and histochemical reactivity for NADPH-diaphorase (A3 & B3): In IHPS, the labelings were severely reduced or absent in the circular muscular layer, indicating the reduction of supporting cells and nerve fibers, and the lack of expression of neuronal NO synthase in nerve fibers (panel A). After healing, supporting calls, nerve fibres, and NADPH-diaphorase reactivity were numerous in the circular layer and arranged in an apparently normal pattern of distribution (panel B). Labe/ing the pacemaker cells/intarstitial cells of Cajal (ICC) with c-kit immunoreactivity (A4& 84/: In IHPS (A4). ICC were observed only the inner part of the circular musculature near the submucosa; they were absent from the outer part of the hypertrophied musculature. After healing (B4). the distribution of ICC was homogenous in the entire circular layer, similar to the distribution in the normal pylorus. CM, circular muscular layer; LM. longitudinal muscular layer; MP, myenteric plexus layer; SM, submucosa. Scale bars = 100 pm.
1533
the gastric outlet obstruction of IHPS might be related to a defect of pyloric relaxation.’ Ultrastructural studies indicated that axonal changes were prominent features in IHPS; Few lesions were observed in the myenteric plexus. The markedly enlarged axons filled with organelles and the other features might be suggestive of neural lesions, because of compression associated with regeneration.14 The increase in muscular volume associated with IHPS is related to the hypertrophy, with little (if any) hyperplasia of the circular musculature,i5 but ultrastructural abnormalities of the smooth muscle cells have been reported in some cases.16It is still not clear whether these pathological alterations reflect a pathophysiological mechanism primarily affecting the ENS or are secondary to some phenomenon occurring primarily in the muscu1ature.l’ We had the opportunity to study pyloric biopsy specimens from two patients after healing of IHPS. Although the available material was limited, the observations were informative. Elements of the ENS (nerve fibers and glial cells) and the expression of neuronal NO synthase, which are reduced in the circular musculature in IHPS,’ apparently returned to normal within a few months after PM. One specimen was obtained 4 months after PM and already appeared normal. In addition, the distribution of ICC, which are observed only at the innermost part of the circular layer in IHPS also apparently returned to normal after healing. ICC are of mesenchymal origin, in contrast to the components of the ENS, which are neural-crest derivatives (10 for review). In the pylorus of the normal neonate, they are evenly distributed in a dense network.‘* Therefore, the hypothesis of the existence of a major structural defect in IHPS, involving the ENS and the ICC, that would result after healing in a pathological aspect similar to the normal pylorus, appears unlikely. Conversely, the hypothesis of a primary volume increase of the circular musculature appears to fit better with our observations and previous reports.“,15 The swelling of smooth muscle cells might explain the apparent reduction of density at the time of PM and the “restitutio ad integrum” of the ENS and of the network of ICC in the circular layer after healing, once the muscular hypertrophy abates. Additional observations, including quantitative studies, are needed to test that hypothesis and to confirm whether the pylorus of IHPS actually returns to normal in terms of its neuronal organization and function. The etiology of IHPS is still unknown, but our observations fill a gap in the knowledge of the natural history of IHPS, and the absence of persistent morphological lesions of the pylorus is in agreement with the general impression of an excellent clinical outcome for IHPS.1,4
VANDERWINDEN
1534
ET AL
REFERENCES 1. Zhang AL, Cass DT, Dawson KP, et al: A medium term follow-up study of patients with hypertrophic pyloric stenosis. J Paediatr Child Health 30:126-128,1994 2. Mitchell LA, Rich N: The genetics of infantile hypertrophic pyloric stenosis. A reanalysis. Am J Dis Child 147:1203-1211, 1993 3. Okorie NM, Dickson JA, Carver RA, et al: What happens to the pylorus after pyloromyotomy? Arch Dis Child 63:1339-1340, 1988 4. Liidtke FE, Berthus M, Voth E, et al: Gastric emptying 16 to 26 years after treatment of infantile hypertrophic pyloric stenosis. J Pediatr Surg 29:523-526,1994 5. Barrios V, Urrutia MJM, Hernandez M, et al: Serum gastrin level and gastric somatostatin content and binding in long-term pyloromyotomized children. Life Sci 55:317-325,1994 6. Nielsen OS: Congenital pyloric stenosis as a factor predisposing to the ulcer syndrome. Acta Paediatr 43:432-443,1954 7. Vanderwinden JM, Mailleux P, Schiffman SN, et al: Nitric oxide synthase activity in infantile hypertrophic pyloric stenosis. N Engl J Med 327:511-5151992 8. Kobayashi H, O’Brian DS, Puri P: Selective reduction in intramuscular nerve supporting cells in infantile hypertrophic pyloric stenosis. J Pediatr Surg 29:651-654, 1994 9. Kobayashi H, O’Brian DS, Puri P: Nerve growth factor receptor immunostaining suggests an extrinsic origin for hypertrophic nerves in Hirschsprung’s disease. Gut 35:1605-1607,1994 10. Christensen J: A commentary on the morphological identifi-
cation of interstitial cells of Cajal in the gut. J Auton Nerv Syst 37~7588,1992 11. Huizinga JD, Thuneberg L, Kiiippel M, et al: A.W:kit gene required for interstitial cells of Cajal and for pacemaker activity. Nature 373:347-349,1995 12. Rintoul JR, Kirkman NF: The myenteric plexus in infantile hypertrophic pyloric stenosis. Arch Dis Child 474-480, 1961 13. Wattchow DA, Cass DT, Furness JB, et al: Abnormalities of peptide containing nerve fibers in infantile hypertrophic pyloric stenosis. Gastroenterology 92:443-448,1987 14. Dieler R, Schroder JM: Myenteric plexus neuropathy in Infantile hypertrophic pyloric stenosis. Acta Neuropathol 78:649661,1989 15. Tam PK: Observation and perspectives of the pathology and possible aetiology of infantile hypertrophic pyloric stenosis-An histological, biochemical histochemical and immunohistochemical study. Ann Acad Med Singapore 14352%529,1985 16. Dieler R, Schroder JM, Skopnik H, et al: Infantile hypertrophic pyloric stenosis: Myopathic type. Acta Neuropathol 80:295306,199O 17. Rollins MD, Shields MD, Quinn RJ, et al: Pyloric stenosis: Congenital or acquired? Arch Dis Child 64:138-139,1989 18. Vanderwinden JM, Liu H, De Laet MH, et al: Study of the interstitial cells of Cajal in infantile hypertrophic pyloric stenosis. Gastroenterology 111:279-288,1996
Menu, Jean-Louis Vanderhaeghen
Pyloric Conreur,
Brussels, Belgium 0 Introduction: Infantile hypertrophic pyloric stenosis (IHPS) is a common surgical affection of unknown etiology. The muscular hypertrophy is known to resolve within a few months after pyloromyotomy (PM). The pathology of IHPS has been studied extensively at the time of PM, but the fate of the pylorus after healing remains unknown. Materials and Methods: We had the rare opportunity to study two pyloric biopsy specimens obtained 4 months and 2 years (respectively) after an uncomplicated PM for IHPS. They were compared with the initial specimen in one case, with 26 other specimens of IHPS, and with five normal controls. Immunohistochemistry using the avidin-biotin complex (ABC) system was performed for S-100 and nerve growth factor receptor, as markers for the enteric nervous system, and for the tyrosine kinase receptor c-kit. as a marker for the interstitial cells of Cajal (pacemaker cells). NADPH-diaphorase histochemim was performed as a marker for the neuronal enzyme nitric oxide synthase, which produces the inhibitory neurotransmitter nitric oxide. Resufts: In both cases of IHPS, after healing, the circular musculature was not hypertrophic. For all markers studied, the distribution appeared similar to that in the normal pylorus. In contrast, all specimens obtained at the time of PM displayed a severe reduction of the different markers in the hypertrophic musculature. Discussion: The pathological features observed in the circular layer in IHPS appear to resolve within a few months after PM. This suggests that the involvement of the enteric nervous system in IHPS might be milder than generally assumed. The etiology remains obscure, but our occasional observations may provide new insight into the pathophysiology of IHPS, and are in agreement with the excellent longterm clinical outcome for IHPS. Copyri9ht o 7996 by W.B. Saunders Company
From the Laboratoire de Neuropathologie et de Recherche sur les Neuropeptides, Fact& de Medecine et Departernent de Chirurgie Pediatnque, Hopital Universitaire des Enfants Reine Fabiola, Universite Libre de Bruxelles, Brussels, Be&m. Presented as a poster at the XLII Annual International Congress of the British Association of Paediatric Surgeons, Sheffield, England, July 25-28, 1995. J.M.V is Senior Research Assistant from the National Fund for Scientific Research (Belgium), 1995-1997. H.L. is a visiting scientist from the Department of Pediatric Surgery, Xian Children’s Hospital, Xian, People S Republic of China. Supported by Belgian Grants from the Fonds de la Recherche Scientifque Medicale (3.4582.94-97) Fondation Medtcale Reine Elisabeth (I 992-95) Ministere de la Politique Scientihque (Pole d ‘Attraction lnteruniversitaire 22, 1990-95), and National Lottery (1992, 1994). Address reprint requests to Jean-Mane Vanderwinden, Laboratoire de Neuropathologie et de Recherche sur les Neuropeptides, Faculte de Medecine, Campus Erasme CP 601, Universite Libre de Bruxelles, 808 route de Lennicc, B-1070 Brussels, Belgium. E-maik [email protected]. Copyright 0 I996 by U? B. Saunders Company 0022-3468/96/3111-0014$03.00i0 1530
INDEX WORDS: Pylorus, pathology, chemistry, immunohistochemistry, kit, interstitial cells of Cajal.
pyloric stenosis, NADPH-diaphorase,
histoc-
NFANTILE HYPERTROPHIC pyloric stenosis (IHPS) is a common surgical affection of infancy. Its diagnosis and treatment are now well standardized, and the outcome usually is benign.l The etiology of IHPS remains elusive,2 but ultrasonography has indicated that the muscular hypertrophy resolves within a few months after pyloromyotomy (PM).3 In patients operated on for IHPS more than 15 years earlier, the gastric emptying rate was normaL but in a recent study,5 persistent abnormalities of the serum gastrin level and mucosal somatostatin content were noted long term in children who had undergone PM, fueling the old debate on possible long-term sequelae of IHPS in terms of gastric function.6 The pathology of the enteric nervous system (ENS) in IHPS has been studied extensively at the time of PM, and the persistence of some abnormalities after clinical recovery might account for long-lasting problems. Unfortunately, the fate of the pylorus in IHPS after healing has not been documented, mainly because such specimens are not readily available. We had the rare opportunity to obtain pyloric biopsy specimens from two patients who had been operated on previously for IHPS at our institution, and we addressed the question of the pathology of the pylorus in IHPS after recovery.
I
MATERIALS
AND METHODS
Pyloric biopsy specimens were obtained from two patients, 4 months and 2 years (respectively) after PM for IHPS.
Case Reports This patient was a dysmature boy who had had an Patient A. uneventful operation for IHPS at 4 weeks of age. At 5 months of age he was readmitted, to the emergency ward, comatose and in severe shock. Immediate intubation and resuscitation were required. Parental anamnesis was confuse. The cardiorespiratoty condition was satisfactory, but a chest radiography showed multiple costal fractures, some with partial healing. The abdomen was extremely distended and painful. Emergency abdominal ultrasonography showed the fluid in the abdominal cavity, and several edematous, dilated intestinal loops, suggestive of intestinal vohulus. An emergency laparotomy was performed and showed internal contusions in the upper abdomen, associated with a retroperitoneat hematoma in the diaphragmatic region. The diagnosis of violent blunt trauma of the upper abdomen was established, and Journal
of Pedtafrtc
Surgery,
Vol31,
No 11 (November),
1996: pp 1530-l
534
PYLORIC
STENOSIS
AFTER
HEALING
parental attitude was seriously questioned. A left hemiparesia appeared on the second postoperative day. A cerebral computed tomography (CT) scan showed a right parietal fracture, bilateral subdural hematomas, and a zone of infarction in the right occipital region. The neurological signs abated subsequently. Despite the prophylactic administration of ranitidine since the admission of the patient, massive gastric bleeding occurred on the eighth postoperative day. The oozing of a large prepyloric stress ulcus could not be controlled by gastroscopy, and an emergency gastrotomy was required to stop the hemorrage. The lesion was located in the prepyloric region and a pyloric biopsy specimen was obtained at that time. The patient’s recovery was slow but eventually satisfactory. Severe parental abuse was estabhshed later. and appropriate measures were taken for follow-up. Patient B. This dysmature girl presented at birth with congenital stenosis of the lower esophagus. A segmental resection with end-to-end anastomosis was performed, without complication, on day 4 of life. At 6 weeks of age she was readmitted because of increasing vomiting and hematemesis. IHPS was diagnosed through echography, and a gastric ulcus was found during gastroscopy. Signs of colonic subobstruction appeared shortly after her admission, and congenital colonic stenosis was diagnosed. A PM and a segmental colonic resection with end-to-end anaostomosis were performed during the same laparotomy procedure. Recidives of hematemesis despite adequate antiacid therapy eventually led to the diagnosis of gastric duplication, which was resected at the age of 4 months. Severe gastroesophageal reflux persisted despite maximal medical therapy. and a Nissen fundoplication was performed when the patient was 2 years old. A pyloric biopsy specimen was obtained at that time because we routinely associate pyloromyectomy with the Nissen procedure. The patient’s development has been satisfactory since then. Although both these cases were complicated, both patients had classical uncomplicated IHPS. At the time of reintervention, they had no clinical sign of gastric emptying delay. The pylorus appeared normal during inspection and palpation. Full-thickness muscular biopsy specimens (with preservation of the mucosa and submucosa) were taken longitudmally, just outside the scar of the previous PM. to avoid fibrous tissue. Both patients had a normal, uncomphcated postoperative course.
Methods We compared the two specimens with a collection of 27 biopsies obtained at the time of PM for IHPS (including the initial specimen from patient A) and with five normal control specimens obtained at the time of autopsy from two newborns and three pediatric organ donors. The specimens were harvested and processed as described previously.’ Briefly, the tissues were hxed overnight in fresh 4% paraformaldehyde solution, cryopreserved in sucrose. embedded in Tissue-Tek OCT compound (Miles, Elkhart. IN), frozen in Z-methyl butane cooled on dry ice, and stored at -gO“C. Longttudinal sections (15 urn thick) were cut on a cryostat. mounted on slides coated with 0.1% poly-L-lysine (Sigma Chemical Co, St Louis. MO), and stored at -20°C until use. Several sections of each specimen were stained routmely with H&E. Immunohistochemical evaluation was carried out with commercially available kits using the avidin-biotin-complex (ABC) system (Vectastain ABC), according to the instructions of the supplier (Vector, Burlingame, CA). Differents kits were used, in accordance with the species of the primary antibodies employed. A mouse monoclonal antibody raised against the glial marker S-100 (reference no. SA2102, Affinity, Nottingham, UK; I ug/mL) and a mouse monoclonal antibody raised against the low-affinity receptor for nerve growth factor (NGFR) (reference no. 1198.637. Boeh-
1531
ringer Mannhein Biochemica, Germany; 5 ug/mL) were used to study the distribution of (respectively) the glial and neural components of the ENS.s9 An affinity purified polyclonal rabbit antiserum raised to a synthetic peptide corresponding to residues 961-976 within the C-terminal domain of the human c-ti receptor (reference no. SC39, Santa-Cruz Biotechnology, Santa-Cruz, CA; 0.4 ug IgG/mL) was used to study the distribution of the interstitial cells of Cajal (ICC), also known as the pacemaker cells of the gut.‘“~” Briefly, the sections were treated in 0.3% hydrogen peroxide in methanol, for 30 minutes, to block endogenous peroxydase activity. They were incubated in normal serum for 20 mmutes. incubated with the primary antibody diluted in normal serum for 3 hours, rinsed m phosphate-buffered saline (PBS) for 10 minutes, incubated with the biotinylated secondary antibody for 30 minutes, rinsed in PBS, and incubated with the ABC complex for 1 hour. ABC conjugated with horseradish peroxidase was used. The peroxidase activity was revealed for 10 to 15 minutes, at room temperature, with a solution containing 0.2 mg/mL 3,3’-diaminobenzidime (DAB) (Sigma), 2 mg/mL ammonium nickel sulfate (Fluka. Buchs, Switzerland), and 0.03% v/v Hz02 in 0.05 mol/LTris-buffered saline. Controls of the immunohlstochemical procedures. No staming was observed when the primary antibodies were omitted. The optimal working dilution was determined empirically for each antihody by serial dilutions. Liquid-phase preabsorption was carried out for the c-kit antiserum according to the instructions of the supplier (Santa Cruz). ie, overnight incubation at 4°C with a tenfold-by-weight excess (4 ug/mL) of the immunogenic peptide (Santa Cruz; reference no. SC39-P) abolishing the signal. NADPH-diaphorase histochemistry was performed (as described previously’) to localize the neuronal form of constitutive mtric oxide (NO) synthase. Briefly, the slides were incubated in the dark in a solution of 0.1 mol/L Tris HCI buffer containing 1 mmol/L NADPH (Sigma), 0.2 mmol/L Nitro Blue Tetrazolium (Sigma), and 10% dimethylsulfoxide, at 37°C for 75 to 90 minutes. Omission of NADPH was used as a negative control and resulted in the absence of staining.
RESULTS
IHPS at the Time of PM (Refer to Fig 1, panel A.) All the specimens studied (n = 27) displayed the typical pathological features of IHPS. Upon H&E staining, the circular muscle layer appeared markedly enlarged whereas the myenteric plexus and longitudinal layers appeared normal. Immunohistochemical evaluation of the ENS markers S-100 and NGFR (Figs A, and AZ, respectively) indicated a severe reduction of neural elements in the hypertrophic circular musculature. Only a few nerve fibers were present, and they appeared abnormally large. The ordinate alignment of neural elements along the axis of the smooth musculature present in the normal pyloric musculature was lacking in all these specimens. In contrast, in the myenteric plexus layer, the patterns of labeling apparently were normal. Histochemical evaluation of NADPH-diaphorase (Fig A,) showed that nerve fibers expressing NO synthase were absent in the hypertrophic circular musculature. However, NO synthase was expressed in nerves bundles and in ganglionic cells in the myenteric plexus. c-kit immunoreac-
1532
VANDERWINDEN
ET AL
PYLORIC
STENOSIS
AFTER
HEALING
tivity labeled very few ICC in the hypertrophic pylorus (Fig &). Only a few ICC were observed in the inner third of the circular layer, near the submucosal edge and at the proximal (gastric) nonhypertrophied side of the biopsies (data not shown). After Healing
(Refer to Fig 1, panel B.) The two specimens’ muscular layers and myentric plexus appeared normal with H&E staining. No fibrous tissue was present, confirming the peroperative impression that the site of biopsy was chosen properly, outside the scar of the previous pyloromyotomy. The pathological features of the two cases were similar. Contrasting with the pattern observed at the time of PM (panel A), S-100 and NGFR immunoreactivity labeled numerous elements of the ENS in the circular muscular layer (Figs B1 and BZ, respectively). These fibers were thin and ordinately arranged in the general direction of the smooth muscular fibers. NADPH-diaphorase histochemistry (Fig B3) indicated the presence of neuronal NO synthase in a large number of nerve fibers in the circular layer. The density and the pattern of distribution of these markers of the ENS appeared similar to those of the normal pylorus. c-kit immunoreactivity (Fig B4) showed that ICC were densely and homogenously distributed in the circular layer, in a pattern similar to that observed in the normal pylorus. DISCUSSION
The pathology of the ENS in IHPS has been studied extensively at the time of PM. Morphological alterations of the ENS” as well as various abnormalities of distribution of neuropeptides and neurotransmitters have been reported,13 including reduction of the neuronal NO synthase in the hypertrophic circular layer. NO is a potent inhibitory neurotransmitter (ie, inducer of smooth muscle relaxation), and the absence of the enzyme producing NO suggested that
Fig 1. Pathological aspect of infantile hypertrophic pyloric stenosis at the time of pyloromyotomy (panel A) and after healing (panel B). Labeling of the enteric nervous system with immunoreactiidty for S- 100 and NGFR (Al & 131, A2 & B2 respectively) and histochemical reactivity for NADPH-diaphorase (A3 & B3): In IHPS, the labelings were severely reduced or absent in the circular muscular layer, indicating the reduction of supporting cells and nerve fibers, and the lack of expression of neuronal NO synthase in nerve fibers (panel A). After healing, supporting calls, nerve fibres, and NADPH-diaphorase reactivity were numerous in the circular layer and arranged in an apparently normal pattern of distribution (panel B). Labe/ing the pacemaker cells/intarstitial cells of Cajal (ICC) with c-kit immunoreactivity (A4& 84/: In IHPS (A4). ICC were observed only the inner part of the circular musculature near the submucosa; they were absent from the outer part of the hypertrophied musculature. After healing (B4). the distribution of ICC was homogenous in the entire circular layer, similar to the distribution in the normal pylorus. CM, circular muscular layer; LM. longitudinal muscular layer; MP, myenteric plexus layer; SM, submucosa. Scale bars = 100 pm.
1533
the gastric outlet obstruction of IHPS might be related to a defect of pyloric relaxation.’ Ultrastructural studies indicated that axonal changes were prominent features in IHPS; Few lesions were observed in the myenteric plexus. The markedly enlarged axons filled with organelles and the other features might be suggestive of neural lesions, because of compression associated with regeneration.14 The increase in muscular volume associated with IHPS is related to the hypertrophy, with little (if any) hyperplasia of the circular musculature,i5 but ultrastructural abnormalities of the smooth muscle cells have been reported in some cases.16It is still not clear whether these pathological alterations reflect a pathophysiological mechanism primarily affecting the ENS or are secondary to some phenomenon occurring primarily in the muscu1ature.l’ We had the opportunity to study pyloric biopsy specimens from two patients after healing of IHPS. Although the available material was limited, the observations were informative. Elements of the ENS (nerve fibers and glial cells) and the expression of neuronal NO synthase, which are reduced in the circular musculature in IHPS,’ apparently returned to normal within a few months after PM. One specimen was obtained 4 months after PM and already appeared normal. In addition, the distribution of ICC, which are observed only at the innermost part of the circular layer in IHPS also apparently returned to normal after healing. ICC are of mesenchymal origin, in contrast to the components of the ENS, which are neural-crest derivatives (10 for review). In the pylorus of the normal neonate, they are evenly distributed in a dense network.‘* Therefore, the hypothesis of the existence of a major structural defect in IHPS, involving the ENS and the ICC, that would result after healing in a pathological aspect similar to the normal pylorus, appears unlikely. Conversely, the hypothesis of a primary volume increase of the circular musculature appears to fit better with our observations and previous reports.“,15 The swelling of smooth muscle cells might explain the apparent reduction of density at the time of PM and the “restitutio ad integrum” of the ENS and of the network of ICC in the circular layer after healing, once the muscular hypertrophy abates. Additional observations, including quantitative studies, are needed to test that hypothesis and to confirm whether the pylorus of IHPS actually returns to normal in terms of its neuronal organization and function. The etiology of IHPS is still unknown, but our observations fill a gap in the knowledge of the natural history of IHPS, and the absence of persistent morphological lesions of the pylorus is in agreement with the general impression of an excellent clinical outcome for IHPS.1,4
VANDERWINDEN
1534
ET AL
REFERENCES 1. Zhang AL, Cass DT, Dawson KP, et al: A medium term follow-up study of patients with hypertrophic pyloric stenosis. J Paediatr Child Health 30:126-128,1994 2. Mitchell LA, Rich N: The genetics of infantile hypertrophic pyloric stenosis. A reanalysis. Am J Dis Child 147:1203-1211, 1993 3. Okorie NM, Dickson JA, Carver RA, et al: What happens to the pylorus after pyloromyotomy? Arch Dis Child 63:1339-1340, 1988 4. Liidtke FE, Berthus M, Voth E, et al: Gastric emptying 16 to 26 years after treatment of infantile hypertrophic pyloric stenosis. J Pediatr Surg 29:523-526,1994 5. Barrios V, Urrutia MJM, Hernandez M, et al: Serum gastrin level and gastric somatostatin content and binding in long-term pyloromyotomized children. Life Sci 55:317-325,1994 6. Nielsen OS: Congenital pyloric stenosis as a factor predisposing to the ulcer syndrome. Acta Paediatr 43:432-443,1954 7. Vanderwinden JM, Mailleux P, Schiffman SN, et al: Nitric oxide synthase activity in infantile hypertrophic pyloric stenosis. N Engl J Med 327:511-5151992 8. Kobayashi H, O’Brian DS, Puri P: Selective reduction in intramuscular nerve supporting cells in infantile hypertrophic pyloric stenosis. J Pediatr Surg 29:651-654, 1994 9. Kobayashi H, O’Brian DS, Puri P: Nerve growth factor receptor immunostaining suggests an extrinsic origin for hypertrophic nerves in Hirschsprung’s disease. Gut 35:1605-1607,1994 10. Christensen J: A commentary on the morphological identifi-
cation of interstitial cells of Cajal in the gut. J Auton Nerv Syst 37~7588,1992 11. Huizinga JD, Thuneberg L, Kiiippel M, et al: A.W:kit gene required for interstitial cells of Cajal and for pacemaker activity. Nature 373:347-349,1995 12. Rintoul JR, Kirkman NF: The myenteric plexus in infantile hypertrophic pyloric stenosis. Arch Dis Child 474-480, 1961 13. Wattchow DA, Cass DT, Furness JB, et al: Abnormalities of peptide containing nerve fibers in infantile hypertrophic pyloric stenosis. Gastroenterology 92:443-448,1987 14. Dieler R, Schroder JM: Myenteric plexus neuropathy in Infantile hypertrophic pyloric stenosis. Acta Neuropathol 78:649661,1989 15. Tam PK: Observation and perspectives of the pathology and possible aetiology of infantile hypertrophic pyloric stenosis-An histological, biochemical histochemical and immunohistochemical study. Ann Acad Med Singapore 14352%529,1985 16. Dieler R, Schroder JM, Skopnik H, et al: Infantile hypertrophic pyloric stenosis: Myopathic type. Acta Neuropathol 80:295306,199O 17. Rollins MD, Shields MD, Quinn RJ, et al: Pyloric stenosis: Congenital or acquired? Arch Dis Child 64:138-139,1989 18. Vanderwinden JM, Liu H, De Laet MH, et al: Study of the interstitial cells of Cajal in infantile hypertrophic pyloric stenosis. Gastroenterology 111:279-288,1996