Purpose: To improve the diagnostic precision of the lumbar tap test (LTT).
Subjects and methods: Thirty one patients (mean age, 65.4 years; male to female ratio, 12:19) suspected of having idiopathic normal pressure hydrocephalus (INPH) were used in the study. They underwent LTT (20-30 ml of cerebrospinal fluid was drained through a puncture with a 18 G needle; evaluation within 3 days after LTT). Based on changes in symptoms after LTT, including dementia (evaluated according to the duration of time needed to accomplish No. 7 items in the N method psychofunction test) and gait disturbance (evaluated according to the duration and the number of steps needed to walk 4 m in a straight line), mean cerebral blood flow (mCBF) monitored with 133Xe-CT, and electroencephalographic topography (EEG-T). The patients were divided into symptom-improved [LTT (+)] and no change [LTT(-)] groups. Changing patterns of EEG-T and mCBF after LTT and the efficacy of V-P shunt at 1 month after shunt administration (effective: E; non-effective: NE) were analyzed and compared. Changing patterns of EEG-T and mCBF after LTT were categorized as improvement of both parameters (type A), that of mCBF only (type B), that of EEG-T only (type C), and no change in either parameter (type D).
Results: (1) Twenty patients were classified as LTT(+) and eleven as LTT(-). (2) The false positive rate was 25% (5/20) for LTT(+); the false negative rate was 27.3% (3/11) for LTT (-). (3) In all the patients, in relation to changing patterns of EEG-T and mCBF, 100% of type A patients (9/9), 75% of type B (6/8), 42.9% of type C (3/7), and 0% of type D (0/7) responded to shunting. (4) Increased rates (IR) of mCBF in 17 patients with improvement of mCBF were 24.2 +/- 10.6% in E patients and 8.9 +/- 5.2% in NE patients, demonstrating a significantly higher percentage in E patients (p < 0.005). The borderline of IR between E and NE was around 15%.
Conclusion: (1) Although quantitative evaluation of symptoms (dementia and gait disturbance) before and after LTT, 27.3% of false negative and 25% of false positive were recognized. (2) According to changing patterns of EEG-T and mCBF after LTT, all type A patients responded to shunting (E), whereas type D patients were all categorized as NE. (3) When the IR of mCBF was 15% or more after LTT, such patients all responded to shunting. (4) The diagnostic precision of LTT in efficacy evaluation is improved when this test is combined with EEG-T and mCBF, in addition to quantitative evaluation of symptoms.