DIAGNOSIS AND GRADING OF ACUTE GRAFT-VERSUS-HOST DISEASE FOLLOWING ALLOGENEIC BONE-MARROW TRANSPLANTATION BY SIGMOIDOSCOPY

Objective: The diagnosis and grading of acute graft-versus-host diseas e (GVHD) are based on the relative unspecific symptoms of exanthema, j aundice and diarrhoea. Since acute GVHD grade greater-than-or-equal-to 2 requires intensive immumosuppresive therapy, a correct diagnosis is crucial. We investigated whether sigmoidoscopy is a useful method for a reliable and rapid diagnosis of acute GVHD. Patients and methods: W e performed sigmoidoscopies in patients following allogeneic bone marr ow transplantation (BMT). Macroscopic findings such as redness and gra nulation of the mucosa, aphthous lesions, erosions, ulcerations, confl uent defects and mucosal bleeding were regarded as criteria for intest inal involvement in acute GVHD. Depending on the extent of these sympt oms, a scale from grades 1 to 4 was proposed. To assess the diagnostic accuracy of these criteria, 42 sigmoidoscopies in 42 patients were re trospectively evaluated and a comparison between macroscopy and histol ogy, clinical grading and grading with regard to skin, liver, or gastr ointestinal symptoms was performed. Results: The sensitivity of macros copy compared with histological examination of the colonic mucosa for the diagnosis of intestinal acute GVHD grade greater-than-or-equal-to 2 was 100% [95% confidence interval (CI) 77-100%], the specificity was 96% (82-100%), the positive predictive value was 93% (68-100%), and t he negative predictive value was 100% (87-100%). Macroscopy had a sens itivity of 72% (47-90%), and a specificity of 92% (73-99%) for the det ection of acute GVHD greater-than-or-equal-to 2 according to the clini cal grading. As expected, there was a good correlation between macrosc opic findings and gastrointestinal symptoms. No complications related to the endoscopic procedure were observed. Conclusions: These data sug gest that sigmoidoscopy is a reliable method for objective diagnosis, grading, or exclusion of acute GVHD grade greater-than-or-equal-to 2 f ollowing allogeneic BMT. Furthermore, these results suggest that it ma y be possible to introduce a score for the objective assessment of acu te GVHD which includes endoscopic findings.