Clinical tolerance of corticospinal tracts in convection-enhanced delivery to the brainstem.

OBJECTIVE: Convection-enhanced delivery (CED) has been explored as a therapeutic strategy for diffuse intrinsic pontine glioma (DIPG). Variables that may affect tolerance include infusate volume, infusion rate, catheter trajectory, and target position. Supratentorial approaches for catheter placement and infusate distribution patterns may conflict with corticospinal tracts (CSTs). The clinical relevance of these anatomical constraints has not been described. The authors report their experience using CED in the brainstem as it relates to anatomical CST conflict and association with clinical tolerance. METHODS: In a phase I clinical trial of CED for DIPG (clinical trial registration no. NCT01502917, clinicaltrials.gov), a flexible infusion catheter was placed with MRI guidance for infusion of 124I-8H9, a radioimmunotherapeutic agent. Intra- and postprocedural MR images were analyzed to identify catheter trajectories and changes in T2-weighted signal intensity to approximate volume of distribution (Vd). Intersection of CST by the catheter and overlap between Vd and CST were recorded and their correlation with motor deficits was evaluated. RESULTS: Thirty-one patients with a mean age of 7.6 years (range 3.2-18 years) underwent 39 catheter insertions for CED between 2012 and 2017. Thirty catheter insertions had tractography data available for analysis. The mean trajectory length was 105.5 mm (range 92.7-121.6 mm). The mean number of intersections of CST by catheter was 2.2 (range 0-3) and the mean intersecting length was 18.9 mm (range 0-44.2 mm). The first 9 infusions in the highest dose level (range 3.84-4.54 ml infusate) were analyzed for Vd overlap with CST. In this group, the mean age was 7.6 years (range 5.8-10.3 years), the mean trajectory length was 109.5 mm (range 102.6-122.3 mm), and the mean overlap between Vd and CST was 5.5 cm3. For catheter placement-related adverse events, 1 patient (3%) had worsening of a contralateral facial nerve palsy following the procedure with two CST intersections, an intersecting distance of 31.7 mm, and an overlap between Vd and CST of 3.64 cm3. For infusion-related adverse events, transient postinfusion deficits were noted in 3 patients in the highest dose level, with a mean number of 2 intersections of CST by catheter, mean intersecting length of 12.9 mm, and mean overlap between Vd and CST of 6.3 cm3. CONCLUSIONS: A supratentorial approach to the brainstem crossing the CST resulted in one worsened neurological deficit. There does not appear to be a significant risk requiring avoidance of dominant motor fiber tracts with catheter trajectory planning. There was no correlation between Vd-CST overlap and neurological adverse events in this cohort.Clinical trial registration no.: NCT01502917 (clinicaltrials.gov).