ObjectiveTo describe the development of progressive multifocal leukoencephalopathy (PML) in patients with rheumatoid arthritis (RA) treated with rituximab.DesignCase study.SettingClinical care for patients with rheumatologic diseases. Most were referred to academic centers for care after diagnosis (Washington University, St Louis, Missouri; Karolinska Insitute, Stockholm, Sweden; and Royal Melbourne Hospital, Melbourne, Australia) while one was cared for in a neurology practice in Dallas, Texas, with consultation by an academic neurovirologist from the University of Colorado in Denver.PatientsFour patients developing PML in the setting of rituximab therapy for RA.InterventionRituximab therapy.Main Outcome MeasuresClinical and pathological observations.ResultsFour patients from an estimated population of 129 000 exposed to rituximab therapy for RA are reported in whom PML developed after administration of this drug. All were women older than 50 years, commonly with Sjögren syndrome and a history of treatment for joint disease ranging from 3 to 14 years. One case had no prior biologic and minimal immunosuppressive therapy. Progressive multifocal leukoencephalopathy presented as a progressive neurological disorder, with diagnosis confirmed by detection of JC virus DNA in the cerebrospinal fluid or brain biopsy specimen. Two patients died in less than 1 year from PML diagnosis, while 2 remain alive after treatment withdrawal. Magnetic resonance scans and tissue evaluation confirmed the frequent development of inflammatory PML during the course of the disease.ConclusionThese cases suggest an increased risk, about 1 case per 25 000 individuals, of PML in patients with RA being treated with rituximab. Inflammatory PML may occur in this setting even while CD20 counts remain low.

The authors investigated the potential utility of artificial neural networks as a decision-making aid to radiologists in the analysis of mammographic data. Three-layer, feed-forward neural networks with a back-propagation algorithm were trained for the interpretation of mammograms on the basis of features extracted from mammograms by experienced radiologists. A network that used 43 image features performed well in distinguishing between benign and malignant lesions, yielding a value of 0.95 for the area under the receiver operating characteristic curve for textbook cases in a test with the round-robin method. With clinical cases, the performance of a neural network in merging 14 radiologist-extracted features of lesions to distinguish between benign and malignant lesions was found to be higher than the average performance of attending and resident radiologists alone (without the aid of a neural network). The authors conclude that such networks may provide a potentially useful tool in the mammographic decision-making task of distinguishing between benign and malignant lesions.

No abstract

http://radiology.rsnajnls.org/cgi/content/full/245/3/684/DC1.

No abstract