Dementia due to normal-pressure hydrocephalus by single photon emission computer tomography

Aims: Once a diagnosis of dementia is established, the irreversible and degenerative types of dementia should be differentiated from those with potentially treatable causes. Normal-pressure hydrocephalus (NPH) is a reversible and poorly understood cause of mental impairment. Ventricular enlargement as registered by Computer Tomography (CT) or Magnetic Resonance Imaging (MRI) might be neurologically stable or progressive in nature. CT/MRI can display the exact size and shape of the ventricles. However, either CT or MRI cerebrospinal fluid (CSF) flow studies are unable to predict the response to shunt operation. The purpose of this study was to determine the role of 99mTc-HMPAO regional cerebral blood flow (rCBF) SPECT combined with 99mTc-DTPA CSF-SPECT in the preoperative evaluation of patients with chronic hydrocephalus. Methods: 103 consecutive studies were carried out where patient management was of concern with a standard technique for each patient (555 MBq 99mTc-HMPAO/ECD; 250 MBq 99mTc-DTPA, 1/3/6/24 h planar, 6/24h SPECT; double head ADAC SPECT camera). RCBF SPECT studies were analysed visually and by a special region of interests (ROIs) program. Circular ROIs were placed over the striatum, thalamus, white matter, frontal, temporal, parietal, occipital cortex, and cerebellum. ROIs were normalized to the whole brain average. The statistical analysis was considered significant at P≤0.05. The results of combined SPECT studies were correlated with CT/MRI, and clinical findings. Results: In 75 cases the hydrocephalus appeared to be communicating. In 19 cases there was a contradiction between the results of CT/MRI and CSF SPECT. 67 ventriculoperitoneal shunt implantations were carried out on the basis of combined SPECT studies. In 67 patients the rCBF SPECT except for hydrocephalus, demonstrated decreased rCBF in the periventricular white matter, thalamus and basal ganglia. There were additional bilateral rCBF abnormalities in the frontal or fronto-temporal regions in 26 patients. 36 patients were not operated upon due to the rCBF and/or CSF SPECT results. One year patients follow-up demonstrated a positive correlation between rCBF SPECT patterns and clinical findings. Conclusion: The sensitivity and specificity of CSF SPECT is higher than that of CT/MRI in the evaluation of CSF abnormalities. The combination of CSF and rCBF SPECT seems to be highly useful for selection of candidates of shunt implantation and timing of surgery