Search

The clinical course of patients with leukemia, specifically after treatment, is complicated by opportunistic infections, often derived from the gastrointestinal tract. Four hundred and thirty-eight patients with leukemia were treated at the Arizona Health Sciences Center from 1976 to 1985. Fifty-five (13 per cent) had 60 major gastrointestinal complications develop. Thirty-seven were treated medically with a mortality rate of 51 per cent, while 18 who were treated surgically had a 17 per cent mortality rate. These infections usually occur during induction or consolidation chemotherapy. Positive blood cultures and white blood cell counts of less than 2,500 per millimeter are a frequent occurrence. The most common complications of the gastrointestinal tract are enterocolitis, perirectal sepsis and bleeding. Because of the combined defects in phagocytosis (neutropenia), antibody production and cell mediated immunity, opportunistic infections (such as viral, fungal, parasitic or bacterial) are frequent and often lethal, despite routine antibiotic coverage. A protocol to identify a subset of patients with septic leukemia who may require emergency surgical treatment is vital since death in this group is most commonly from undiagnosed sepsis or progression of hematologic defects. This protocol should include repeated physical examinations, daily roentgenograms of the abdomen, liberal use of endoscopy, contrast roentgenography and computed tomography (CT) or ultrasound. Extensive surgical procedures can be safely carried out in patients with leukemia. The diagnosis of leukemia should not be a deterrent to emergency surgical exploration.

Twelve patients underwent a takedown of the jejunoileal bypass with restoration of normal continuity of the intestine. Specimens were obtained for biopsy from the excluded proximal jejunum and distal ileum as well as the distal in continuity jejunum. Hypertrophy of the mucosa was observed in the in continuity segment, as previously described. Villi of the bypassed jejunum displayed no evidence of hypertrophy or of atrophy and were identical in size to villous measurements prior to creation of the jejunoileal bypass. At the ultrastructural level, absorptive cells of bypassed jejunal segments exhibited normal morphologic characteristics. There was no indication of cellular immaturity above the level of the intestinal crypts. In contrast, villi of bypassed ileum, near the colonic anastomosis, demonstrated hypertrophy, presumably reflecting reflux of luminal material into this segment of bypassed intestine. The results of this investigation support the concept that significant morphologic alterations of the mucosa do not occur in totally bypassed segments of intestine after the jejunoileal bypass procedures with attainment of weight stabilization.

With better methods of diagnosis, patients will be identified earlier in the course of their disease and will often have atypical and borderline manifestations of the syndrome. Serum gastrin measurements with calcium and especially with secretin challenge will be the most important method of diagnosis. Any patient with acid hypersecretion who has a high serum gastrin level that does higher on secretin infusion should be considered to have the Zollinger-Ellison syndrome. A firm diagnosis of the Zollinger-Ellison syndrome should be made, if at all possible, prior to operation. At operation, a thorough search of the pancreas, duodenum, stomach, greater and lesser omentum and liver should be made for primary and secondary gastrinomas. If the preoperative data firmly establish the diagnosis of the Zollinger-Ellison syndrome, a total gastrectomy should be carried out even if no primary tumor is found. Similarly, a total gastrectomy should be done even if there are massive hepatic metastases. If total gastrectomy is not performed, the patient is apt to die of complications of acid hypersecretion. The only possible exceptions to the rule of always performing a total gastrectomy are in asymptomatic patients with easily excisable tumors or patients with tumors of the duodenum that are easily excisable, providing that in both instances after the excision of the tumor the output of gastric acid as measured at operation is immediately halted. All possible metastatic tumor tissue should be removed. The more tumor tissue removed, the longer the patient will survive. Metastases should be treated aggressively. They do not disappear after total gastrectomy in our experience, and they may kill patients. Patients should be followed after operation with serial measurements of serum gastrin concentrations and by hepatic scintillation scans and hepatic angiography. If hepatic metastases develop, intrahepatic artery infusions of 5-fluorouracil may slow tumor growth.