Role of chronic infection and inflammation in the gastrointestinal tract in the etiology and pathogenesis of idiopathic parkinsonism. Part 2: response of facets of clinical idiopathic parkinsonism to Helicobacter pylori eradication. A randomized, double-blind, placebo-controlled efficacy study.

BACKGROUND: Links between etiology/pathogenesis of neuropsychiatric disease and infection are increasingly recognized. AIM: Proof-of-principle that infection contributes to idiopathic parkinsonism. METHODS: Randomized, double-blind, placebo-controlled efficacy study of proven Helicobacter pylori eradication on the time course of facets of parkinsonism. Intervention was 1 week's triple eradication therapy/placebos. Routine deblinding at 1 year (those still infected received open-active), with follow-up to 5 years post-eradication. Primary outcome was mean stride length at free-walking speed, sample size 56 for a difference, active vs. placebo, of 3/4 (between-subject standard deviation). Recruitment of subjects with idiopathic parkinsonism and H. pylori infection was stopped at 31, because of marked deterioration with eradication failure. Interim analysis was made in the 20 who had reached deblinding, seven of whom were receiving antiparkinsonian medication (long-t(1/2), evenly spaced) which remained unchanged. RESULTS: Improvement in stride-length, on active (n = 9) vs. placebo (11), exceeded size of effect on which the sample size was calculated when analyzed on intention-to-treat basis (p = .02), and on protocol analysis of six weekly assessments, including (p = .02) and excluding (p = .05) those on antiparkinsonian medication. Active eradication (blind or open) failed in 4/20, in whom B-lymphocyte count was lower. Their mean time course was: for stride-length, -243 (95% CI -427, -60) vs. 45 (-10, 100) mm/year in the remainder (p = .001); for the ratio, torque to extend to flex relaxed arm, 349 (146, 718) vs. 58 (27, 96)%/ year (p < .001); and for independently rated, visual-analog scale of stance-walk videos (worst-best per individual identical with 0-100 mm), -64 vs. -3 mm from anterior and -50 vs. 11 lateral (p = .004 and .02). CONCLUSIONS: Interim analysis points to a direct or surrogate (not necessarily unique) role of a particular infection in the pathogenesis of parkinsonism. With eradication failure, bolus release of antigen from killed bacteria could aggravate an effect of ongoing infection.