Probable normal pressure hydrocephalus with cardiac and renal comorbidity

The history of gait disturbance, urinary urgency, and mild cognitive decline with ventriculomegaly and DESH is classic for probable normal pressure hydrocephalus, but the differential includes Alzheimer's disease, vascular dementia, and Parkinsonism. A 30 mL lumbar puncture with a 4-second improvement in the 10-metre walk at 48 hours is a positive but modest tap test; it supports CSF diversion but does not guarantee shunt response. I would document gait speed and video before and after, and consider a large-volume tap (40–50 mL) or external lumbar drainage for 24–48 hours if the response is equivocal. I would assess gait with the 10-metre walk and Timed Up and Go, and check for comorbid cervical myelopathy or lumbar stenosis, which can mimic NPH. The DESH pattern and tight high-convexity sulci increase the probability. I would not diagnose NPH on imaging alone. The next step is a multidisciplinary discussion with the patient and family about a shunt, while optimising heart failure and renal function.

I would choose a ventriculoperitoneal shunt with a programmable valve and an antisiphon or gravitational device, because this patient has a 2-year history, DESH, and a positive tap test, and over-drainage is a major risk in an elderly patient with altered brain compliance. A fixed medium-pressure valve is available only; if that is the only option, I would use it with a small ventricular catheter and a distal antisiphon device if stock allows, and set the valve to medium pressure rather than low. The goal is not maximum drainage but symptom improvement without subdural collections. I would avoid a lumboperitoneal shunt if there is any question of communicating hydrocephalus, and I would not choose ETV because the aqueduct is patent and ETV success is lower in NPH. Postoperatively, I would monitor for subdural haematoma, headache, and gait deterioration, and image if symptoms change. If over-drainage occurs, a programmable valve can be adjusted; with a fixed valve, revision may be needed.

Apixaban should be stopped 48 hours before elective surgery if creatinine clearance is normal, but this patient has stage 3 CKD, so I would stop it at least 72 hours before and check renal function; bridging with low-molecular-weight heparin is not routinely needed for a shunt and may increase bleeding. I would involve cardiology because of atrial fibrillation and heart failure with ejection fraction 35%. I would not use andexanet or PCC unless there is urgent surgery or major bleeding, because of thrombotic risk. Postoperatively, I would restart apixaban once haemostasis is secure, usually after 24–72 hours, with mechanical prophylaxis in the interim. I would reconsider surgery if the gait improvement were not reproducible, if the patient had a high fall risk that could not be managed, or if the cardiac risk were prohibitive. The decision is a balance between a modest chance of gait improvement and the risk of subdural haematoma, infection, and shunt failure; a frank discussion with the family is essential.