Pyloric stenosis due to eosinophilic gastroenteritis: presentation of two cases in mono-ovular twins.

Abstract

Sir: Eosinophilic gastro-enteritis (EG) is a rare disorder (about 250 cases, overall in adults) de®ned by three criteria: (1) gastro-enteric symptoms; (2) eosinophilic in®ltration of one or more areas of gastro-intestinal tract; and (3) no evidence of parasitic or extraintestinal disease [1, 3, 4]. Peripheral blood eosinophilia is not considered a diagnostic criterion because it is not a universal ®nding (20%±90%) [5]. We present mono-ovular twins, 6-month-old males, with EG. One twin (6.750 kg) was admitted to hospital after a 1 week history of alimentary projectile vomiting. Abdominal ultrasonography showed an antropyloric muscular hypertrophy. The obstructive feature of the lesion was detected by upper gastro-intestinal radiography (Fig. 1). Oesophagogastroduodenoscopy, performed for the peculiar observation of an atypical hypertrophic pyloric stenosis, revealed a pylorospasm with normal oesophagogastric mucosa. Blood test showed WBC >30.000 (11.5% eosinophils). Identical leucocytosis was also noticed in his asymptomatic twin. The patient underwent a successful Fredet-Ramstedt extramucosal pyloromyotomy. A hyperemic, oedematous, hypertrophic pyloric olive, with a nearby di€use perivisceritis, was found. Therefore, a pyloric muscular biopsy was performed. Histology provided the diagnosis of EG showing muscular eosinophilic in®ltration. The postoperative course was uneventful and the normal feeding rapidly restored. Nevertheless, leucocytosis remained high (35,000), eosinophils increased (>15%) and echography revealed a persisting gastropyloric hypertrophy. Meanwhile, an analogous ultrasonographic picture was found in his twin, who began to present similar manifestations. Therefore, both brothers underwent corticosteroid therapy (methylprednisolone 1±2 mg/kg). One month later, eosinophilia disappeared and the gastropyloric muscle thickness became normal. The subocclusive phenomena recurred whenever the therapy was stopped. Both patients have been treated with cyclic methylprednisolone for almost 2 years, in the as yet fruitless search for the lowest e€ective dose. The aetiology of EG is unknown, but since the eosinophilic in®ltration is exclusive to the gastro-intestinal tract, it suggests the involvement of local factors in recruitment and activation of eosinophils: interleukin-3, granulocyte-macrophage colony-stimulating factor, and interleukin-5. The release by eosinophils of such cytokines might explain the persistence of eosinophilic in®ltration in EG. The genes for the mentioned cytokines share certain structural features and are clustered on the long arm of human chromosome 5 (5q23±22). This close linkage raises the chance of common mechanisms controlling the expression of these cytokines [2]. EG manifests with di€erent features according to the predominantly involved layer [5]: (a) mucosal disease induces nausea, vomiting, abdominal pain, diarrhoea and weight loss; (b) muscularis disease brings on intermittent obstructive symptoms because of the thickening and rigidity of the gastrointestinal tract; (c) subserosal disease, amongst others, produces eosinophilic ascites (hallmark). The de®nitive diagnosis of EG is based on biopsy. We suggest performing full-thickness biopsies, to avoid false-negative results, and taking more than one specimen, since EG occurs in a patchy distribution [1]. Steroids remain the mainstay of therapy. Most patients have a relatively benign course with recurrent symptoms controlled by short-term corticosteroid therapy. A few patients need chronic steroid therapy (as ours) and some have self-limited recurrences. Use of oral sodium chromoglycate is controversial, whereas elimination diets have generally been unsuccessful [1]. Signi®cant morbidity, including unnecessary surgery, is associated with delayed diagnosis.