Five Decades of Infusion Testing in Idiopathic Normal-Pressure Hydrocephalus: Resistance to CSF Outflow as a Biomarker in Search of a Gold Standard

Idiopathic normal-pressure hydrocephalus (iNPH) is a potentially reversible cause of gait impairment, cognitive decline, and urinary dysfunction in older adults, yet identifying patients most likely to benefit from cerebrospinal fluid (CSF) diversion remains a major diagnostic challenge. For more than five decades, the lumbar infusion test, introduced by Katzman and Hussey in 1970, has served as the principal method for assessing CSF absorptive capacity through measurement of resistance to CSF outflow (R_out). This narrative review examines the historical development, physiological basis, and contemporary clinical application of infusion testing in iNPH, with particular emphasis on the diagnostic and prognostic value of R_out. We critically review the evidence supporting the R_out thresholds recommended by current international, Japanese, and American Academy of Neurology guidelines and reassess their performance using a Bayesian framework. Recalculation of likelihood ratios from published meta-analyses demonstrates that none of the commonly proposed R_out thresholds (10–18 mmHg·min/mL) achieves the prognostic performance typically associated with a strong rule-in or rule-out test. We argue that this limited performance reflects not only methodological heterogeneity in infusion-testing protocols but also a more fundamental problem: the absence of a universally accepted reference standard for iNPH. We further examine the limitations of using postoperative clinical improvement as a surrogate reference standard, including variability in outcome definitions, delayed treatment effects, and the potential for circular reasoning. Finally, we review emerging CSF proteomic biomarkers of neuroinflammation, axonal injury, and concomitant Alzheimer-type pathology that may complement hydrodynamic assessment by capturing the balance between reversible hydrocephalus-related dysfunction and irreversible neurodegenerative burden. We conclude that R_out should be regarded as a valuable but incomplete biomarker and that future progress will likely depend on multimodal, probabilistically interpreted models integrating CSF hydrodynamics, neuroimaging, and molecular biomarkers to improve prognostic stratification and establish more robust diagnostic frameworks for iNPH.