Radiographic appearance of mucosal dysplasia associated with ulcerative colitis.

The radiographic appearance of dysplasia, a precancerous histologic change that is frequently present in colitic colons at high risk for cancer, was analyzed. Over a 5-year period, 170 patients with long-standing ulcerative colitis participated in a cancer surveillance program at our institution. On entry, all patients underwent radiography with a double-contrast barium enema and colonoscopy with multiple random biopsies. All radiographs were reviewed retrospectively and were correlated with the colonoscopic and histologic findings. Dysplasia, ranging from mild to severe, was found in the biopsy specimens from 26 (15%) of these patients (in 59 different regions). Thirteen of the 26 patients with dysplasia had colectomies, and the specimens were available for review. Most dysplasia is not radiographically visible, as was the case in 40 (68%) of the 59 regions found histologically to have dysplasia. In the 19 regions in which dysplasia was visible radiographically, it appeared as a solitary nodule or as several separate nodules in 11 (19%) of the 59 regions or as a close grouping of multiple adjacent nodules with apposed, flattened edges in 8 (14%) of the 59 regions. The radiographic finding of a nodule or of several separate nodules was not specific because these nodules were indistinguishable from the inflammatory nodules often present in colitic colons. On the other hand, the radiographic finding of a close grouping of adjacent nodules with apposed, flattened edges was associated with dysplasia 50% of the time. This radiographic appearance was seen in five of our 26 patients who had dysplasia. Our study shows that barium examinations are useful in some patients with chronic ulcerative colitis by suggesting the presence of dysplasia and directing the endoscopist to specific locations for biopsy.

Fractures after rheumatoid arthritis. A population-based study.

In a population-based study, the incidence of osteoporotic fractures in patients who have been diagnosed as having rheumatoid arthritis was investigated. This incidence was found to be increased, though not dramatically so: the relative risk for hip fracture, for example, was 1.5. Univariate analyses generally indicated increased risk associated with increasing age, earlier age at diagnosis of rheumatoid arthritis, disability, impaired ambulation, steroid use, and thinness, and decreased risk associated with obesity and estrogen use. In multivariate analyses, only aging, impaired ambulation, and thinness were identified as independent risk factors.