BACKGROUND: Immunodeficiency predisposes to invasive esophageal infections. The treatment of perforation, respiratory fistula, and necrosis due to transmural esophageal infection is guided by anecdote. We wish to determine treatment and outcome of local complications of necrotizing esophagitis. METHODS: We report our experience over a 7-year period and review published reports since 1976. We treated 4 patients and found 21 reported patients with perforation (11/25), fistula (8/25), and necrosis (6/25) at a mean age of 35 years. Twenty-one patients were immunodeficient (84%) due to acquired immunodeficiency syndrome in 8, acute leukemia in 6, renal transplant in 3, diabetes mellitus, renal failure, and corticosteroids in 1 each. Pathogenic organisms were fungal in 15 cases, viral in 7, and bacterial in 7. RESULTS: Treatment consisted of antibiotic therapy in 13 patients and surgical intervention combined with antibiotic therapy in 12: esophagectomy in 6, esophageal stenting and drainage in 2, drainage alone in 2, and salivary diversion in 2. Overall mortality was 48% (12/25). Mortality without surgical intervention was 90% (9/10) and with surgical intervention 27% (3/11). One of 6 patients undergoing esophagectomy (17%) died. The difference in mortality was due to sepsis, which was the cause of death in 8 patients treated with medical intervention and only 1 treated with surgical intervention. CONCLUSIONS: Local complications of necrotizing esophagitis have a high mortality due to sepsis. Surgical intervention, in particular esophagectomy, controls sepsis in published case reports and should be considered in selected patients. Further study is required to determine the true prevalence of these complications and the outcome of intervention.
BACKGROUND:Immunodeficiency predisposes to invasive esophageal infections. The treatment of perforation, respiratory fistula, and necrosis due to transmural esophageal infection is guided by anecdote. We wish to determine treatment and outcome of local complications of necrotizing esophagitis. METHODS: We report our experience over a 7-year period and review published reports since 1976. We treated 4 patients and found 21 reported patients with perforation (11/25), fistula (8/25), and necrosis (6/25) at a mean age of 35 years. Twenty-one patients were immunodeficient (84%) due to acquired immunodeficiency syndrome in 8, acute leukemia in 6, renal transplant in 3, diabetes mellitus, renal failure, and corticosteroids in 1 each. Pathogenic organisms were fungal in 15 cases, viral in 7, and bacterial in 7. RESULTS: Treatment consisted of antibiotic therapy in 13 patients and surgical intervention combined with antibiotic therapy in 12: esophagectomy in 6, esophageal stenting and drainage in 2, drainage alone in 2, and salivary diversion in 2. Overall mortality was 48% (12/25). Mortality without surgical intervention was 90% (9/10) and with surgical intervention 27% (3/11). One of 6 patients undergoing esophagectomy (17%) died. The difference in mortality was due to sepsis, which was the cause of death in 8 patients treated with medical intervention and only 1 treated with surgical intervention. CONCLUSIONS: Local complications of necrotizing esophagitis have a high mortality due to sepsis. Surgical intervention, in particular esophagectomy, controls sepsis in published case reports and should be considered in selected patients. Further study is required to determine the true prevalence of these complications and the outcome of intervention.