Our present results highlight the value of pcMRI for the preoperative evaluation of CMI. Indeed, high preoperative intraventricular CSF pulsatility has prognostic value with regard to postoperative headache relief. This finding might help physicians to select patients for surgery or, in contrast, to recommend drug treatment alone, with a view to avoiding potential post-operative morbidity in the absence of likely clinical benefits.
Headaches
According to some research, the typically pulsatile CMI-associated headaches appear to respond better to surgery than atypical headaches do4. Our study included both patients with typical headaches (pulsatile, accentuated by effort and by coughing, with radiation to the vertex) and patients with atypical headaches (all other types). However, some patients with atypical headaches experience clinical relief after surgery, and so it is difficult to base selection for surgery purely on the patient’s subjective description of headache; objective variables (such as those based on pcMRI data) are needed. The preoperative profile of symptoms other than headache did not appear to have prognostic value because it was similar in patients with vs. without postoperative headache relief.
Hydrodynamics and prognostic assessment
A few research have studied the relationship between craniospinal hydrdynamics and symptoms. Several studies have shown the link between the presence of symptoms and MICs, without reaching the point of being a definitive marker19,24.
However, these studies focused solely on the maximum flow velocity of the CSF or blood. The volume displaced depends directly in the pressure gradient25. Hence, an analysis of the displaced volume appears to be entirely appropriate. Some research have already described the value of pcMRI for predicting postoperative clinical improvement in patients with CMI19–21,24. McGirt et al.24 demonstrated that normal preoperative CSF flow in the hindbrain, assessed by pcMRI, was an independent risk factor for failure to achieve postoperative headache relief in patients with CMI. PcMRI may prove useful in selecting Chiari I malformation patients who are at higher risk of poor response to surgical decompression.24 A high preoperative SVaqu appears to predict postoperative headache relief. This might be due to the recruitment of ventricular compliance during CMI subarachnoid compliance might change as a result of ptosis of the cerebellar tonsils, with greater flow resistance in this area.
Our present results also indicate that changes in the dynamics of the cervical subarachnoid CSF might be related to the postoperative prognosis for headaches. According to other studies, a decrease in postoperative CSF dynamics (a decrease of at least 20% in SVc2c3, specifically) is associated with headache relief21. Nevertheless, this significant difference appears to have less clinical relevance than an analysis of intraventricular CSF dynamics. Indeed, an analysis of the distribution of CSF hydrodynamic profiles (Fig. 4) shows that most of patients presented similar profiles. This similarity might be linked to different hemodynamic profiles. Indeed, CSF pulsatility is directly related to variations in intracranial vascular volume16.
Aqueduct area and Evans index
According to Poiseuille's law, fluid flow depends on the pressure gradient, flow area, and fluid viscosity. Intraventricular CSF dynamics result from variations in the pressure gradient between the 3rd and 4th ventricles. The viscosity of CSF does not change. There was no significant difference in the cross-sectional area of the midbrain aqueduct between patients with headache relief and those without. The increase in intraventricular CSF dynamics therefore appears to be related to variations in the pressure gradient between the 3rd and 4th ventricles. It would therefore be interesting to measure the pressure gradient on either side of the aqueduct of Sylvius, in order to determine whether or not this variable is related to clinical relief of headache after surgery. This gradient might be responsible for high intraventricular pulsatility.
The results of several studies of patients with normal pressure hydrocephalus have demonstrated that the SV at the aqueduct of Sylvius depends essentially on the ventricular volume26. In our study, we observed a higher preoperative CSF SV in the group of patients with headache relief. However, this SV was not linked to an increase in ventricular volume because the two groups of patients did not differ significantly with regard to the Evans index - a good marker of ventriculomegaly. Furthermore, recent studies suggest that combining preoperative assessments of cerebrospinal fluid hydrodynamics with morphological analysis of the fourth ventricle outlet improves the predictive accuracy of postoperative clinical outcomes in patients undergoing decompression surgery CMI27. An increase in ventricular pulsatility (highlighted in our study by the increase in the SV of the aqueduct of Sylvius) in patients with postoperative improvement does not therefore appears to be linked to an increase in ventricular volume: whatever the preoperative ventricular volume, the ventricular pulsatility increases. Pulsatility does not therefore appear to be linked to ventricular volume.
Study limitations
Firstly, our study included a small number of patients; further studies of larger populations are needed to confirm the present results. Secondly, our study was retrospective study, and so further prospective studies are required including headache-specific score assessment. Thirdly, our analyses were limited to CSF dynamics, which are secondary to variations in intracranial arteriovenous volume. Hence, studies focusing on arteriovenous variables might provide further prognostic evidence of postoperative clinical improvements in patients. Future studies could assess both variations in CSF volumes and changes in arterial and venous volume dynamics during the CC. Breathing also influences hemodynamics during the CC. It would be interesting to study of respiratory variations during the CC in this context.