Revisiting secondary normal pressure hydrocephalus: does it exist? A review

OBJECTIVE There are several etiologies that can lead to the development of secondary normal pressure hydrocephalus (sNPH). The aim of this study was to evaluate the etiology, diagnosis, treatment, and outcome in patients with sNPH and to highlight important differences between the separate etiologies. METHODS A comprehensive review of the literature was performed to identify studies conducted between 1965 and 2015 that included data regarding the etiology, treatment, diagnosis, and outcome in patients with sNPH. Sixty-four studies with a total of 1309 patients were included. The inclusion criteria of this study were articles that were written in English, included more than 2 patients with the diagnosis of sNPH, and contained data regarding the etiology, diagnosis, treatment, or outcome of NPH. The most common assessment of clinical improvement was based on the Stein and Langfitt grading scale or equivalent improvement on other alternative ordinal grading scales. RESULTS The main etiologies of sNPH were subarachnoid hemorrhage (SAH) in 46.5%, head trauma in 29%, intracranial malignancies in 6.2%, meningoencephalitis in 5%, and cerebrovascular disease in 4.5% of patients. In 71.9% of patients the sNPH was treated with ventriculoperitoneal shunt placement, and 24.4% had placement of a ventriculoatrial shunt. Clinical improvement after shunt placement was reported in 74.4% and excellent clinical improvement in 58% of patients with sNPH. The mean follow-up period after shunt placement was 13 months. Improvement was seen in 84.2% of patients with SAH, 83% of patients with head trauma, 86.4% of patients with brain tumors, 75% of patients with meningoencephalitis, and 64.7% of patients with NPH secondary to stroke. CONCLUSIONS Secondary NPH encompasses a diverse group of clinical manifestations associated with a subset of patients with acquired hydrocephalus. The most common etiologies of sNPH include SAH and traumatic brain injury. Secondary NPH does indeed exist, and should be differentiated from idiopathic NPH based on outcome and on clinical, pathophysiological, and epidemiological characteristics, but should not be considered as a separate entity.