Of the 160 patients (80 pirazolac/80 sulindac) who entered the study through 14 investigators, three-quarters completed a 12-weeks therapy and three-fifths completed the entire 24-weeks therapy. In the pirazolac group 15% of the patients and in the sulindac group 11% dropped out from the study due to adverse clinical experience. The drop-out rates due to unsatisfactory therapeutic response were respectively 15% and 16% in the pirazolac and the sulindac groups. Both treatment groups showed significant improvement from baseline for all parameters except for the erythrocyte sedimentation rate at weeks 4, 8, 12 and 16 for the sulindac group and weeks 4 and 8 for the pirazolac group. The two treatment groups were comparable as to effectiveness; however, the improvement rates in 36 out of 41 efficacy measurements based on the definition of clinically relevant changes in relation to baseline were estimated to be superior for the group under pirazolac therapy. The rate of improvement for the American Rheumatism Association functional class at the end of the study was 23% in the pirazolac group and 9% in the sulindac group (p less than 0.05). Of the patients in the pirazolac and sulindac groups, 45% and 44% respectively reported no adverse effects at all throughout the whole 24-weeks study. The rates of patients reporting at least one adverse reaction in a body system were not different between the two groups. An exception was the body as a whole where ten patients (12.5%) in the sulindac group and only two (2.5%) in the pirazolac group reported adverse reactions (p = 0.03). No differences occurred between the two treatment groups with regards to intensity, causality or the number of occurrences of adverse clinical experiences. One death in the sulindac treatment group was reported during the study. In both treatment groups, alterations in laboratory tests were minor or negligible or associated with abnormal pre-treatment values and, generally speaking, without any clinical relevance. There were some patients, who had increases from baseline in alkaline phosphate in both treatment groups. However, these were usually transient, occasionally complemented by a slight increase of serum glutamic oxaloacetic acid transaminase.
