Background: Idiopathic Normal Pressure Hydrocephalus (iNPH) is a disease mainly affecting the elderly with the core symptoms of incontinence, gait and cognitive impairment. The standard treatment is cerebrospinal fluid diversion, with ventriculoperitoneal shunts being the most common choice. The main radiological finding in iNPH is the expansion of the cerebral ventricles. Recent radiological and computational advances have made the measurement of ventricular volume increasingly feasible. This thesis, comprising of four papers, focuses on the potential use of ventricular volume as a clinical tool in the postsurgical setting in iNPH. Primarily for assessing suspected shunt malfunction (paper II), but also the individualization of shunt opening pressure settings (paper IV). Additional aims were to investigate associations between ventricular volume and cerebrospinal fluid biomarkers to assess whether there is a dilution effect (paper I), and investigating improvement in excessive daytime sleepiness after surgery (paper III).

Methods: For Paper I, participants with iNPH were examined (n=136). Lumbar CSF was sampled as a part of routine care and concentrations of Amyloid β 1-42, total tau and phosphorylated tau results extracted from medical charts. Ventricular volume for the lateral and third ventricles was manually segmented on presurgical MRI using ITK-SNAP. Papers II-IV were based on one study population of participants with iNPH (n=36). All participants received Strata II shunts with adjustable performance levels (PL) corresponding to different opening pressures. Clinical evaluation and MRI were performed repeatedly: before surgery and at one, two, three months as well as three months plus one day after surgery. At surgery, all shunts were set at PL 1.5. After the first follow-up, participants were randomly assigned to either PL 1.0 or 2.5. After the two months follow-up, the assignments were crossed over. At the third follow-up, all shunts were set to PL 0.5 for the final 24 hours. Clinical assessment included the Timed Up and Go test, 10-meter walking test, Mini-Mental State Examination, continence scale, gait scale, balance scale, modified Rankin scale, and Epworth Sleepiness Scale. Ventricular volume was manually segmented using SyMRI.

Results: Ventricular volume and Amyloid β 1-42 were weakly associated (β=-0.20 p=0.027). Ventricular volume was not significantly associated with the tau forms, though the estimates were positive. Ventricular volume decreased gradually according to lower opening pressures of the shunts: the median change from presurgical to PL 2.5 was 6.9 mL (p<0.001), from PL 2.5 was PL 1.5 of 3.5 mL (p=0.01), from PL 1.5 was PL 1.0 of 9.6 mL (p<0.001). Daytime sleepiness significantly improved following surgery with a median change of -1.5 points, p=0.026, though it did not change further at subsequent follow-ups. Clinical outcomes did not change following randomization to PL 1.0 versus PL 2.5, despite an overall significant clinical improvement following surgery. Clinical outcomes were not consistently associated with change in ventricular volume. There was an increased incidence of subdural effusion after randomization to PL 1.0, occurring in eight participants (p=0.005).

Conclusions: In this thesis, ventricular volume shows substantial and predictable change with shunt adjustments, supporting its usefulness as a marker of shunt patency. In contrast, the pattern of associations between ventricular volume and CSF biomarkers does not support dilution as a primary cause of lower biomarker concentrations in iNPH, and ventricular volume does not meaningfully relate to clinical outcomes within a one-month time frame. Lower shunt opening pressures increase risk without providing clinical benefit. Another key finding was that excessive daytime sleepiness appears to be a symptom of iNPH that is amenable to treatment.