Role of MR Csf flow study in NPH

By definition NPH is Ventriculomegaly on MRI with normal CSF pressure, altered CSF dynamics.

Of course NPH is a clinical diagnosis.
The classical clinical triad of NPH is
1) urinary incontinence
2 ) deterioration in cognition (dementia) and
3 ) Ataxia i.e. Gait disturbances.
As the name suggests mean CSF opening pressure in patients with NPH is within the normal range with a classic neurological sign is magnetic gait.

Best diagnostic clue on MRI is ventricles and Sylvian fissures symmetrically dilated out of proportion to sulcal enlargement, with normal hippocampus which distinguishes NPH from atrophy.
Ventriculomegaly is prominent in all 3 horns of lateral ventricles and 3rd ventricle, with relative sparing of 4th ventricle.

Role of MRI is now not confined only to support the clinical diagnosis of NPH but is to identify shunt-responsive NPH pts from non responsive by calculating aqueductal stroke volume non invasively.
This is because studies have shown that aqueductal stroke volume in patients of NPH decreases later in the disease process despite clinical progression and classifcal findings on MRI this has been theorized to be caused by cerebral atrophy, which indicates that the patient is unlikely to respond to shunt surgery.


Normal CSF flow and its dynamics

The stroke volume across aqueduct is the average CSF volume flowing through the aqueduct in one cardiac cycle, craniocaudad during systole OR caudocraniad during diastole.
Normal values of stroke volume is < .04 ml/cycle
It is assumed that the net flow over 1 cardiac cycle is [negligible enough to be considered as] zero
Aqueductal velocity (caudal): 3-8mm/s
The peak velocity was determined from the maximum value of the measured velocities of each cardiac phase.


CSF flow study findings in NPH
In NPH actually there is lack of flow from the cisternal and subarachnoid spaces with significant increase in amount of ventricular csf flow.
There increased flow void across aqueduct with increase stroke volume.
1) Increased aqueductal stroke volume is the average volume of CSF moving through the cerebral aqueduct calculated by summing the absolute values of stroke volume in systole and diastole and dividing by 2
i.e   Forward stroke volume +  Reverse stroke volume   /   2
2) Increased aqueductal peak velocity.

Various publications have set various normal and abnormal ranges.
Flow rate of > 24.5mL/min 95% specific for NPH.
Stroke volume of > 42 microL shown on one paper to predict good response from shunting was statistically significant (P < .05).Studies have shown that aqueductal stroke volume decreases later in the disease process despite clinical progression.
Sroke volume upper limit is now suggested to be variable between institutions due to intrinsic scanner differences, thus each centre should obtain their own "normal values", with the upper limit being suggested as two times the normal value.
There was no statistically significant relationship between aqueductal CSF flow void score and responsiveness to shunting.

Case 1

With clinical diagnosis of NPH

MR brain reveals diffuse cerebral and cerebellar cortical atrophy.
The ventricular dilatation is disproportionate to the amount of cortical atrophy, scalloping of inferior margin of corpus callosum and prominent flow void in posterior portion of third ventricle, across cerebral aqueduct and superior portion of fourth ventricle consistant with clinical diagnosis of NPH.

Csf flow study report :
Gated cine phase contrast study was performed to evaluate the CSF flow.
The diastolic phases are equal in number to the systolic.
The stroke volume is 41 microliters per cycle consistent with shunt responsive status (Reference : Patients with stroke volume more than 40 microliters respond well to VP shunt _ Bradley et al)

Case 2 

Atrophy Vs NPH clinically

MRI brain reveals moderate cerebral and cerebellar atrophy. The ventricular dilatation is mildly disproportionate to the amount of cortical atrophy.

Csf flow study report :
Gated cine phase contrast study was performed to evaluate the CSF flow.
The diastolic phases are equal in number to the systolic. The stroke volume is 30 microliters per cycle consistent with non-shunt responsive status. ( Reference : Patients with stroke volume more than 40 microliters respond well to VP shunt _ Bradley et al).

Contributors:
Dr Deepak Patkar (HOD, Nanawati Superspeciality Hospital, Vile Parle, Mumbai)
Dr Balaji Anvekar (Short visiting Fellowship in Neuroradiology and Recent Advances in Neuroimaging)