gms | German Medical Science

Objective: Acute hydrocephalus is readily addressed by implantation of an external ventricle drain (EVD). To determine the need for permanent CSF diversion, probatory closure of the EVD is performed after the acute stage. Typically, the EVD is re-opened if clinical deterioration or progressive dilatation of the ventricles is observed, or removed if secondary deterioration is not observed. Automated infrared pupillometry (aiPM) has been shown to facilitate early detection of elevated ICP non-invasively. This technique may hold the potential to bypass the risk of exposing a patient to a potentially harmful secondary deterioration by timely prediction of hydrocephalic enlargement.

Methods: We prospectively enrolled 65 consecutive adult patients (56,89±12,10 yrs) with EVD due to acute hydrocephalus from May 2016 - October 2019. Patients were stratified into two groups based on whether weaning of EVD was successful or failed (groupS, groupF, resp.). Pupillometry data were collected before EVD closure until after eventual reopening or removal. The following parameters were recorded: Neurological Pupil index (NPi), pupil size (SIZE), minimum pupil diameter (MIN), decrease in pupil diameter after light stimulation (DIA), constriction velocity (CV), maximum constriction velocity (MCV), dilation velocity (DV) and latency (LAT).

Results: Weaning of EVD was successful in 44 cases (67.7%; groupS) and failed in the remaining third of all patients (n=21, 32.3%; groupF). Baseline parameters of aiPM were comparable in both groups before probatory closure of the EVD. Two days after closure, patients with eventual weaning failure showed significantly reduction of both constriction velocity (CV, MCV) and dilation velocity (DV) compared to groupS (p<0.05, p<0.05, p<0.05). A trend towards an early, incipient reduction of constriction velocity was already observed 24hrs after closure (p=0.164, p=0.114). NPi, SIZE, MIN, DIA, LAT showed no differences. After reopening the EVD, all pupillometry parameters of the failure group recovered to comparable values.

Conclusion: During EVD weaning, aiPM detects specific fluctuations in pupillary function, suggesting progressive impairment of neuronal circuitry in patients requiring permanent CSF diversion. If these results are confirmed in an even larger cohort, this non-invasive monitoring technique may enable us to shorten the weaning process, thus avoiding prolonged probatory closure and provocation of significant clinical deterioration.